Team 32 - Overall Team Report

CN 4120: DESIGN II Project PRODUCTION OF HYDROGEN VIA SYNGAS ROUTE

Overall Team Design Report TEAM 32:

Lim Yueh Yang Ng Su Peng Ong Song Kun Tham Zhi Yong, Andrew Zhang Zihong (Leader) Sin Yew Leong Heng Chee Hua

U046787U U046929L U046829M U046754W U046816H U046835M U046793U

(Steam Methane Reformer) (Furnace) (High Temp. Shift Reactor) (Low Temp. Shift Reactor) (Pressure Swing Adsorption) (Heat Exchanger Network) (Cooling Tower)

This report is submitted in partial fulfillment of the requirements for the Degree of Bachelor of Engineering (Chemical) Department of Chemical & Biomolecular Engineering National University of Singapore 2007/2008

CN 4120: Design II Team 32

Executive Summary

EXECUTIVE SUMMARY

Hydrogen is vital for daily operations in refineries worldwide due to its primary usage in hydrotreaters, as environmental regulations on sulphur emissions are strictly enforced. Furthermore, secondary units, such as hydrocrackers are constructed to boost the margins of refineries through upgrading of middle distillates (kerosene and diesel), which requires the hydrogen. Thus this report aims to develop a preliminary design for a hydrogen plant (1.25e9 m3 (STP/year)), whose operation is based on the syngas route that involves the coupling of steam-methane-reforming with low temperature and high temperature shift reactions. Further assumptions such as siting the plant in Singapore and an 8000 h/year operation time were also considered in the production of hydrogen with at least 99.9% in product purity. The preliminary design consists of seven main units – furnace, steam-methanereformer (SMR), high-temperature-shift-reactor (HTS), low-temperature-shift-reactor (LTS), pressure swing adsorber (PSA), heat exchanger network (HEN) and cooling tower. The main purpose of the furnace was to supply heat to the feed mixture of steam and natural gas so that the endothermic reaction can proceed in the SMR. Optimal design indicated the requirement of a four-chamber single-tube-pass side-fired heater which comprised of a radiant section (33m x 27m x 13m), a convection section (7m x 7m x 1m) and a stack (diameter = 4m, height = 8m). A thermal efficiency of 94% was achieved. The refractory walls comprised of firebricks with silicon carbide linings.

450 SMR tubes (material = HK-40) were housed in the radiant section of the furnace where 6.0e8 kJ/h was supplied via the combustion of the tail-gas directed from PSA and excess air. A steam-to-carbon ratio of 3:1 was stipulated, which would also minimize coking. Prior to entering SMR, the feed was preheated to an inlet temperature of 539oC. In the presence of Ni/Mg-Al2O4, methane would react with steam to produce an effluent that contained primarily carbon monoxide and hydrogen. A methane conversion of 80.1% was attained with an exit temperature of 852oC. The total cost of the SMR tubes and furnace were estimated to be US$730,000 and US$37 million respectively. The SMR effluent was subsequently cooled before entering HTS at a temperature of 354oC,

Production of Hydrogen via Syngas Route

CN 4120: Design II Team 32

Executive Summary

with make-up steam to achieve a steam-to-carbon ratio of 5:1. The HTS vessel (diameter = 3.46 m, height = 12.11 m, material = ASTM A387) served to increase the hydrogen yield through the oxidation of carbon monoxide to carbon dioxide, in the presence of chromium promoted iron oxide. The stream composition of carbon monoxide was subsequently reduced from 13.3% to 3%. The calculated bare module cost was approximately US$3.6 million.

As the oxidation process was slightly exothermic, a lower temperature operation would favour a higher conversion, thus justifying further process cooling to 220oC prior to entering LTS. Through the optimal design of the LTS vessel (diameter = 3.31 m, height = 5.07 m, material = ASTM A387) with copper-zinc oxide catalyst supported on alumina, the carbon monoxide level was further lowered to 0.5% at the LTS exit. To prevent poisoning of PSA catalyst downstream by condensate, a knock-out drum and a bed of silica gel was installed after LTS, prior to the entry into PSA. The bare-module cost of LTS was estimated at US$1.4 million. Further cooling of the LTS effluent to 50oC was effected before entering into the knock-out drum and subsequently into PSA. The composition of the PSA feed was roughly 75% hydrogen and 18% carbon dioxide. A Polybed system of 8 columns (diameter = 3 m, height = 8.5 m, material = SS clad), operating between 1 and 25 bar at 50oC, was adopted for the purification of hydrogen with the use of activated carbon and zeolite 5A at a ratio of 5:1. A hydrogen recovery of 85% with a product purity of 99.9% was subsequently achieved. The estimated bare module cost of PSA was $25 million.

Extensive heat integration was performed for maximum energy recovery in this design. Only cold utilities, such as high pressure steam and cooling water were needed, as the furnace had fulfilled all the heating requirements of the plant. This resulted in the presence of a utility pinch, which requires the adoption of pinch analysis. 3 networks each satisfying the maximum energy recovery criterion was designed using HX-Net. The selected network was chosen based on the lowest total annual cost and operational considerations, attaining 100.6 % of the total cost target. The chosen TEMA configuration of the heat exchangers was that of AES shell and tube exchangers (split-ring floating head). For thermal design, the heat

Production of Hydrogen via Syngas Route

CN 4120: Design II Team 32

Executive Summary

exchanger chosen possessed a heat transfer area of 519.2 m2 (calculated by HX-Net) with 456 SS tubes of length 4.88 m while the shell was fabricated from carbon steel. The cost of this heat exchanger was about US$526,000.

A cooling tower was designed based on an induced draft counter-flow configuration. A filled height of 6.0 m was essential for the rejection of heat into the atmosphere via both evaporation and sensible means. Replenishment of water (188 m3/h) was required for continuous operation due to evaporative losses. Construction costs were estimated at US$941,000. The lifespan of all catalysts was assumed to be 3 years and their cost amounted to roughly US$2.2 million/year. Based on a discounted cash flow rate of return of 10% and a payback period of 15 years (inclusive of 2 years of construction), the selling price of hydrogen calculated was US$2.43 / kg (STP), which was less than US$2.70 / kg (STP)1. Therefore, we recommend the construction of the plant due to the profitability of the product.

Safety is paramount, thus a HAZOP worksheet was generated to identify potential hazards due to possible deviations in both SMR and furnace operations. Recommended safeguards and actions were also highlighted. A summary of occupational safety and health, environmental impact assessment and plant layout was further discussed in this report. Lastly, an implementation of process controls and instrumentations was performed on both SMR and furnace. A piping and instrumentation diagram (P&ID) was subsequently developed with further discussions centering on the various reflected control strategies.

Production of Hydrogen via Syngas Route

CN 4120: Design II Team 32

Production of Hydrogen via Syngas Route

Executive Summary

CN 4120: Design II Team 32

Production of Hydrogen via Syngas Route

Executive Summary

CN 4120: Design II Team 32

Production of Hydrogen via Syngas Route

Executive Summary

CN 4120: Design II Team 32

Executive Summary

REFERENCE 1. Hydrogen and Clean Fuels: Systems Studies. Retrieved on April 17, 2008 from National Energy Technology Laboratory Web site: http://www.netl.doe.gov/technologies/hydrogen_clean_fuels/systems_studies.html

ACKNOWLEDGEMENTS

This section dedicates acknowledgements to all who have helped our team by offering their valuable advice. In particular, we would like to express our heart-felt gratitude to our professors, Prof Karimi, Prof Rangaiah, Prof Farooq, A/P Kawi, A/P M.P. Srinivasan, A/P R. Srinivasan, A/P Krishnaswamy, A/P Hidajat and A/P Borgna, for their valuable insights.

Last but not least, this work would not have been possible without the coordination and teamwork in Team 32 (CN4120 – Sem 2 of AY2007/08), hence we would like to thank all of them for their assistance and understanding.

Production of Hydrogen via Syngas Route

CN 4120: Design II Team 32

Process Flow Diagram (P.F.D.)

SMR HTS

Furnace LTS & K/O Drum

PSA

Production of Hydrogen via Syngas Route

CN 4120: Design II Team 32

TABLE OF CONTENTS Chapter 1 : PROBLEM DESCRIPTION........................................................................ 1-1 1.1 PROBLEM STATEMENT FOR PLANT DESIGN SPECIFICATIONS ............ 1-1 1.2 BACKGROUND FOR DEVELOPMENT IN HYDROGEN PRODUCTION .... 1-1 1.2.1 Energy Woes – Away from Fossil Fuels Era ................................................... 1-1 1.2.2 Identifying & Justifying the Production Route – SMR .................................. 1-2 1.2.3 Choice and Significance of Reforming Feedstock – Natural Gas................... 1-4 1.2.4 Steam Methane Reforming (SMR) Reactor .................................................... 1-5 1.2.5 Furnace ............................................................................................................ 1-6 1.2.6 Shift Reactions ................................................................................................. 1-7 1.2.7 Product Purifications ....................................................................................... 1-7 1.2.8 Heat Integration ............................................................................................... 1-8 1.2.9 Cooling Requirements ..................................................................................... 1-9 1.2.10 Use of HYSYS Simulation ........................................................................... 1-10 1.3 REFERENCES ..................................................................................................... 1-10 Chapter 2 STEAM METHANE REFORMER ............................................................... 2-1 2.1 PROBLEM STATEMENT .................................................................................... 2-1 2.1.1 Problem and Specifications ............................................................................. 2-1 2.1.2 Justifications for using SMR ........................................................................... 2-1 2.2 DESIGN METHODOLOGY & PROCESS DESIGN........................................... 2-2 2.2.1 Outline of Design Methodology ....................................................................... 2-2 2.2.2 Reaction Chemistries ....................................................................................... 2-2 2.2.2.1 Effects of Temperature and Pressure – Revisiting Le Chatelier’s Principle ............................................................................................................ 2-3 2.2.2.2 Coke Formation, Steam:Methane ratio & Inclusion of CO2 in feed ... 2-3 2.2.3 Choice of Reactor – Tubular Reformer .......................................................... 2-4 2.2.4 Justifications for Choice of Firing Configuration – Side-fired reformer furnace ...................................................................................................................... 2-5 2.2.5 Justifications for Choice of Fluid Package – Peng-Robinson ......................... 2-6 2.2.6 Choice of Catalyst ............................................................................................ 2-6 2.2.7 Kinetics, Ni-based Catalyst & Role of Support .............................................. 2-7 2.2.8 Justification for Choice of Reactor Inlet Conditions ...................................... 2-7 2.3 PRELIMINARY DESIGN ..................................................................................... 2-8 2.3.1 Establishment of Base Case ............................................................................. 2-8

Production of Hydrogen via Syngas Route

i

CN 4120: Design II Team 32

2.3.2 Preliminary Simulation using HYSYS ............................................................ 2-8 2.4 DETAILED DESIGN ........................................................................................... 2-10 2.4.1 Development of Critical Profiles via MATLAB & Optimisation ................. 2-10 2.4.2 Design Equations & Key Assumptions.......................................................... 2-10 2.4.3 Fitting into HYSYS Simulation Environment using Plug-Flow Reactor (PFR) ................................................................................................................................. 2-12 2.4.4 Results and Discussions ................................................................................. 2-12 2.4.4.1 Conversion profiles for CH4 and CO2 ............................................... 2-12 2.4.4.2 Temperature and Pressure Variations ............................................... 2-13 2.4.4.3 Component Mole Fractions ................................................................. 2-14 2.4.5 Optimization .................................................................................................. 2-15 2.4.6 Operating Conditions & Streams Conditions ............................................... 2-17 2.5 MATERIALS OF CONSTRUCTION & SIZING .............................................. 2-18 2.5.1 Selection Methodology ................................................................................... 2-18 2.5.2 Justifications for selecting from different grades of stainless steels ............. 2-18 2.5.3 Tube life estimation, Minimum Stress Rupture & Identification of Choice Material................................................................................................................... 2-19 2.5.4 Sizing – Computation for Tube Thickness .................................................... 2-21 2.5.5 Sizing – Summary .......................................................................................... 2-21 2.6 ECONOMICS & SAFETY CONSIDERATIONS .............................................. 2-22 2.6.1 Economic Analysis (Brief) ............................................................................. 2-22 2.6.2 Safety Consideration for Reactor Design ...................................................... 2-22 2.7 LEARNING & CONCLUSIONS ......................................................................... 2-23 2.8 NOTATIONS ........................................................................................................ 2-24 2.9 FIGURES AND TABLES .................................................................................... 2-25 2.10 ACKNOWLEDGEMENTS................................................................................ 2-25 2.11 REFERENCES ................................................................................................... 2-26 2.12 APPENDIX ......................................................................................................... 2-28 2.12.1 MATLAB Code ............................................................................................ 2-28 2.12.1.1 Main m-file to resolve O.D.E.s .......................................................... 2-28 2.12.1.2 Function m-file to define reactions conditions and O.D.E.s ............. 2-29 2.12.2 List of Equations .......................................................................................... 2-33 2.12.2.1 Rate Equations for reactions and species for 4 O.D.E.s ................... 2-33 2.12.2.2 Mole Fractions for species ................................................................. 2-33 Production of Hydrogen via Syngas Route

ii

CN 4120: Design II Team 32

2.12.2.3 Effectiveness Factors for reactions and species ................................ 2-33 2.12.2.4 Rate & Adsorption constants for reactions 1, 2 and 3 ................................. 2-34 2.12.2.5 Adsorption constants for species .................................................................. 2-34 2.12.2.6 Heat Capacities ............................................................................................. 2-35 2.12.3 Sample Calculations..................................................................................... 2-35 2.12.4 Typical Natural Gas Compositions ............................................................. 2-35 Chapter 3 : FURNACE .................................................................................................... 3-1 3.1 INTRODUCTION .................................................................................................. 3-1 3.1.1 Furnace design methodology ........................................................................... 3-1 3.1.2 Heat transfer process in fired heater .............................................................. 3-2 3.2 RADIATION ZONE DESIGN ............................................................................... 3-2 3.2.1 Thermal Efficiency of Fired Heater ................................................................ 3-2 3.2.2 Calculation for the number of reformer tubes ............................................... 3-6 3.2.3 Calculation for mass velocity in reformer tubes ............................................. 3-7 3.2.4 Calculation of reformer tube thickness ........................................................... 3-8 3.2.5 Selection of material for reactor tube in radiation section ............................. 3-9 3.2.6 Reformer inner tube diameter....................................................................... 3-11 3.2.7 Furnace layout and design ............................................................................. 3-11 3.2.7.1 Side Fired Heater................................................................................. 3-11 3.2.7.2 Distance between burners ................................................................... 3-12 3.2.7.3 Burners used at Side Walls ................................................................. 3-13 3.2.7.4 Determination of number of burners.................................................. 3-14 3.2.8 Computations for flue gas temperature ........................................................ 3-15 3.2.8.1 Cold plane area .................................................................................... 3-15 3.2.8.2 Refractory area .................................................................................... 3-15 3.2.8.3 Absorptivity, α ..................................................................................... 3-15 3.2.8.4 Sum of product of area and the absorptivities in the radiant zone.... 3-15 3.2.8.5 Mean beam length ............................................................................... 3-16 3.2.8.6 Partial pressure of CO2 and H2O ....................................................... 3-16 3.2.8.7 Product of partial pressure and mean beam length ........................... 3-16 3.2.8.8 Mean refractory tube wall temperature ............................................. 3-16 3.2.8.9 Two main equations that will be used for iteration to find Tg (flue gas temp) ................................................................................................................ 3-16 3.2.8.9.1 Radiant zone heat transfer ........................................................... 3-16

Production of Hydrogen via Syngas Route

iii

CN 4120: Design II Team 32

3.2.8.9.2 Radiant zone heat balance ............................................................ 3-16 3.2.8.9.3 Enthalpy of the flue gas as a function of Tg (flue gas temp)........ 3-17 3.2.8.9.4 Emissitivity of the gas Ф ............................................................... 3-17 3.2.8.9.5 Exchange factor F ......................................................................... 3-17 3.2.9 Residence Time .............................................................................................. 3-18 3.3 CONVECTION SECTION .................................................................................. 3-19 3.3.1 Convection design – Finned tubes ................................................................. 3-19 3.3.2 Design parameters for convection tubes ....................................................... 3-21 3.3.3 Pressure drop in the tubes present in furnace .............................................. 3-22 3.4 STACK DESIGN .................................................................................................. 3-24 3.4.1 Stack diameter ............................................................................................... 3-24 3.4.2 Pressure Drop across stack............................................................................ 3-24 3.4.2.1 Stack exit loss ....................................................................................... 3-24 3.4.2.2 Frictional Loss in stacks and ducts ..................................................... 3-24 3.4.2.3 Stack entrance loss .............................................................................. 3-25 3.4.2.4 Flue gas pressure drop through the convection section ..................... 3-25 3.4.2.5 Pressure drop at the top of the radiant section .................................. 3-25 3.4.2.6 Pressure gain at the convection section............................................... 3-25 3.4.3 Stack Height ................................................................................................... 3-26 3.5 MATERIALS FOR CONSTRUCTION OF FURNACE BODY & ADDITIONAL AUXILIARIES ........................................................................................................... 3-27 3.5.1 Refractory walls ............................................................................................. 3-27 3.5.2 Stack Walls..................................................................................................... 3-28 3.5.3 Additional auxiliaries..................................................................................... 3-28 3.5.3.1 Air Preheaters ...................................................................................... 3-28 3.5.3.2 Forced Draft Fan ................................................................................. 3-29 3.5.3.3 Induced Draft Fan ............................................................................... 3-29 3.6 COST ANALYSIS ................................................................................................ 3-30 3.6.1 Purchased Equipment Costs.......................................................................... 3-30 3.6.1.1 Costing for Furnace ............................................................................. 3-30 3.6.1.2 Costing for Air Preheater .................................................................... 3-31 3.6.1.3 Costing for Induced Draft Fan and Forced Draft Fan for Air Preheating System ........................................................................................... 3-31 3.6.1.4 Burners ................................................................................................ 3-32 Production of Hydrogen via Syngas Route

iv

CN 4120: Design II Team 32

3.6.2 Utility Cost ..................................................................................................... 3-32 3.6.2.1 Electricity cost ..................................................................................... 3-32 3.6.3 Total Annual cost ........................................................................................... 3-32 3.7 SUMMARY & CONCLUSION ........................................................................... 3-33 3.8 SPECIFICATION OF FIRED-HEATER ........................................................... 3-34 3.9 REFERENCES ..................................................................................................... 3-35 Chapter 4 : HIGH TEMPERATURE SHIFT REACTOR............................................. 4-1 4.1 INTRODUCTION .................................................................................................. 4-1 4.1.1 Water gas shift ................................................................................................. 4-1 4.1.2 High temperature shift .................................................................................... 4-2 4.2 PROBLEM DESCRIPTION.................................................................................. 4-3 4.3 REACTION THERMODYNAMICS .................................................................... 4-6 4.3.1 Criteria for Chemical Reaction Equilibrium .................................................. 4-6 4.3.2 Effects of Pressure on Reaction Equilibrium.................................................. 4-7 4.3.3 Effects of Temperature on Reaction Equilibrium .......................................... 4-8 4.4 REACTION KINETICS ...................................................................................... 4-11 4.5 CATALYST .......................................................................................................... 4-12 4.6 REACTOR............................................................................................................ 4-13 4.6.1 Type of reactor ............................................................................................... 4-13 4.6.2 Reactor design ................................................................................................ 4-13 4.7 METHODOLOGY AND CALCULATIONS ...................................................... 4-16 4.7.1 Weight of catalyst .......................................................................................... 4-16 4.7.2 Pressure drop ................................................................................................. 4-19 4.7.3 Thickness of vessel ......................................................................................... 4-24 4.7.4 Reactor size and cost ...................................................................................... 4-24 4.8 HEAT EXCHANGER .......................................................................................... 4-26 4.8.1 Heat Exchanger Design Considerations ........................................................ 4-27 4.8.1.1 Physical properties extraction ........................................................................ 4-27 4.8.1.2 Determination of overall heat transfer coefficient......................................... 4-28 4.8.1.3 Exchanger type and dimensions ..................................................................... 4-28 4.8.1.4 Heat transfer area........................................................................................... 4-29 4.8.1.5 Layout and tube size ....................................................................................... 4-29 4.8.1.6 Number of tubes ............................................................................................. 4-29 4.8.1.7 Bundle and shell diameter .............................................................................. 4-29

Production of Hydrogen via Syngas Route

v

CN 4120: Design II Team 32

4.8.1.8 Tube-side heat transfer coefficient ................................................................. 4-30 4.8.1.9 Shell-side heat transfer coefficient ................................................................. 4-30 4.8.1.10 Overall coefficient ......................................................................................... 4-31 4.8.1.11 Pressure drop ................................................................................................ 4-31 4.9 CONCLUSION ..................................................................................................... 4-32 4.10 NOTATIONS ...................................................................................................... 4-33 4.11 REFERENCES ................................................................................................... 4-35 4.12 APPENDICES .................................................................................................... 4-36 Appendix 4.12.1 .......................................................................................................... 4-36 Appendix 4.12.2 .......................................................................................................... 4-38 Appendix 4.12.3 .......................................................................................................... 4-39 Chapter 5 : LOW TEMPERATURE SHIFT REACTOR .............................................. 5-1 5.1 INTRODUCTION .................................................................................................. 5-1 5.2 LTS DESIGN CONSIDERATIONS ...................................................................... 5-2 5.2.1 Current Status ................................................................................................. 5-2 5.2.2

Kinetics of Low-Temperature Water-Gas-Shift (LTWGS) ........................ 5-3

5.2.2.1 Assumption made for equation (5-3) : .................................................. 5-4 5.2.3 LTS Catalyst .................................................................................................... 5-5 5.2.3.1 Characteristics of the industrial LTS catalyst ...................................... 5-6 5.2.3.2 Preparation ............................................................................................ 5-6 5.2.3.3 Supply .................................................................................................... 5-6 5.2.3.4 Deactivation of LTS Catalyst ................................................................ 5-7 5.2.3.5 LTS catalyst in operation ...................................................................... 5-8 5.2.3.6 Assumptions made for LTS Catalyst .................................................... 5-9 5.2.3.7 Mass balance on the Copper-Zinc catalyst pellet ................................. 5-9 5.2.3.8 Heat balance on the Copper-Zinc catalyst pellet................................ 5-10 5.2.4

Modeling the converter ............................................................................. 5-12

5.2.4.1 Assumptions made for the converter ................................................ 5-12 5.2.4.2 Reactor mass balance .......................................................................... 5-12 5.2.4.3 Reactor mass balance .......................................................................... 5-13 5.3 DESIGN CONDITIONS ...................................................................................... 5-17 5.3.1 Temperature .................................................................................................. 5-17 5.3.2 Pressure .......................................................................................................... 5-17 5.3.3 Steam to CO ratio .......................................................................................... 5-17

Production of Hydrogen via Syngas Route

vi

CN 4120: Design II Team 32

5.3.4 Design Procedure for LTS outlet compositions and Mass of Catalyst used 5-19 5.3.5 LTS outlet compositions and Mass of Catalysts used ................................... 5-20 5.3.6 Design Procedure for Aspect Ratio ............................................................... 5-21 5.3.7 Results for Aspect Ratio ................................................................................ 5-22 5.3.8 Design Procedure for the dimensions of bed and thickness of vessel wall ... 5-23 5.3.9 Results for bed dimensions and wall thickness ............................................. 5-24 5.3.10 Allowances set for design ............................................................................. 5-25 5.3.11 Study of controlling parameters ................................................................ 5-26 5.4 CHOICE OF A REACTOR BED ........................................................................ 5-27 5.4.1 Cost estimation for the LTS converter .......................................................... 5-28 5.5 DESIGN OF THE KNOCK-OUT DRUM........................................................... 5-30 5.5.1 Working principle of the knock-out drum .................................................. 5-30 5.5.2 Sizing of the knock-out drum ........................................................................ 5-30 5.5.3 Results and cost estimation............................................................................ 5-31 LITERATURE REVIEW .......................................................................................... 5-33 CONCLUSION .......................................................................................................... 5-35 BIBLIOGRAPHY ...................................................................................................... 5-36 APPENDIX A1 ........................................................................................................... 5-37 APPENDIX A2 ........................................................................................................... 5-39 APPENDIX A3 ........................................................................................................... 5-39 Chapter 6 : PRESSURE SWING ABSORPTION .......................................................... 6-1 6.1 INTRODUCTION .................................................................................................. 6-1 6.2 PROBLEM STATEMENT .................................................................................... 6-2 6.3 THEORETICAL BACKGROUND ....................................................................... 6-2 6.3.1 Separation via adsorption................................................................................ 6-2 6.3.2 Pressure-Swing Adsorption (PSA) .................................................................. 6-3 6.3.3 Skarstrom Cycle............................................................................................... 6-3 6.3.4 Adsorbents ....................................................................................................... 6-4 6.4 DESIGN CONSIDERATIONS .............................................................................. 6-5 6.5 ACTUAL MODELING OF PSA ........................................................................... 6-7 6.5.1 Component Mass Balance ............................................................................... 6-8 6.5.2 Overall Mass Balance ...................................................................................... 6-8 6.5.3 Pressure terms.................................................................................................. 6-8 6.5.4 Adsorption rates .............................................................................................. 6-8

Production of Hydrogen via Syngas Route

vii

CN 4120: Design II Team 32

6.5.5 Overall Mass Balance in Dimensionless Form ................................................ 6-9 6.5.6 Component Mass Balance in Dimensionless Form ......................................... 6-9 6.5.7 Dimensionless Pressure terms ......................................................................... 6-9 6.5.8 Dimensionless Langmuir Adsorption Isotherms ............................................ 6-9 6.5.9 Boundary Conditions ..................................................................................... 6-10 6.6 MODEL OPTIMIZATION.................................................................................. 6-11 6.6.1. Process Methodology .................................................................................... 6-12 6.6.2 Initial approximation of the adsorption time from the breakthrough curve... 613 6.6.3 Determination of Cyclic steady state ............................................................. 6-14 6.6.4 Refinement of the pressurization time .......................................................... 6-15 6.6.5 Possible optimization of feed superficial velocity and diameter of the bed . 6-16 6.7 FINAL RESULTS AND DISCUSSIONS ............................................................ 6-18 6.8 COST ESTIMATIONS ........................................................................................ 6-19 6.9 CONCLUSION ..................................................................................................... 6-22 6.10 NOTATIONS ...................................................................................................... 6-23 6.11 APPENDIX ......................................................................................................... 6-25 6.12 CONSTANTS APPLIED IN COMSOL SIMULATION .................................. 6-29 6.13 REFERENCES ................................................................................................... 6-30 Chapter 7 : HEAT EXCHANGER NETWORK ............................................................ 7-1 EXECUTIVE SUMMARY .......................................................................................... 7-1 ACKNOWLEDGEMENTS ......................................................................................... 7-1 7.1 DESIGN METHODOLOGY OF A HEAT EXCHANGER NETWORK ............ 7-2 7.1.1 Determination & Verification of Stream Data Properties Extracted from Hysys ......................................................................................................................... 7-2 7.1.1.1 Calculations of Maximum Design Velocities ........................................ 7-2 7.1.1.2 Determination of Flow Area Diameter ................................................. 7-3 7.1.1.3 Calculations of Convective Heat Transfer Coefficients (HTC) ........... 7-4 7.1.1.4 Fouling Factors ...................................................................................... 7-5 7.2 TARGETING ......................................................................................................... 7-6 7.2.1 Cost Considerations ......................................................................................... 7-6 7.2.2 Utility Cost Calculations .................................................................................. 7-8 7.2.3 Heat Exchanger Capital Cost Estimations...................................................... 7-9 7.2.4 Supertargeting ................................................................................................. 7-9

Production of Hydrogen via Syngas Route

viii

CN 4120: Design II Team 32

7.2.5 Comparison between the usage of HP and LP Steam Generation ............... 7-10 7.2.6 Calculation of Utility Targets ........................................................................ 7-10 7.3 MER NETWORK DESIGN................................................................................. 7-12 7.3.1 Stream matching above pinch ....................................................................... 7-14 7.3.2 Stream matching below pinch ....................................................................... 7-14 7.3.3 Number of units in MER network................................................................. 7-15 7.3.4 Alternative MER Network Designs for Consideration ................................. 7-15 7.3.4.1 Network 1a ........................................................................................... 7-15 7.3.4.2 Network 1b........................................................................................... 7-16 7.4 NETWORK EVOLUTION .................................................................................. 7-20 7.4.1 Steps involved in network evolution.............................................................. 7-20 7.4.2 Evolution of 1st loop ....................................................................................... 7-20 7.4.3 Evolution of 2nd loop ...................................................................................... 7-21 7.5 HEAT EXCHANGER DESIGN .......................................................................... 7-26 7.5.1 Stream Data ................................................................................................... 7-26 7.5.2 Material of Construction ............................................................................... 7-27 7.5.3 Shell and Tube-Side Fluid Allocation ........................................................... 7-27 7.5.4 Exchanger Type ............................................................................................. 7-28 7.5.5 Baffles ............................................................................................................. 7-28 7.5.6 Tube Dimensions ............................................................................................ 7-28 7.5.7 Tube Arrangements ....................................................................................... 7-29 7.5.8 Calculations .................................................................................................... 7-29 7.5.8.1 Tube-Side Heat Transfer Coefficient Calculations ............................ 7-31 7.5.8.2 Shell-Side Heat Transfer Coefficient Calculations ............................. 7-31 7.5.8.3 Overall Heat Transfer Coefficient Calculations ................................. 7-33 7.5.8.4 Tube-Side Pressure Drop Calculations ............................................... 7-33 7.5.8.5 Shell-Side Pressure Drop Calculations ............................................... 7-34 7.5.9 Modification of Design ................................................................................... 7-34 7.5.10 Exchanger Cost ............................................................................................ 7-35 7.6 RECENT DEVELOPMENTS ................................................................................. 7-36 7.7 HEAT EXCHANGER SPECIFICATION SHEET............................................. 7-37 7.8 INTEGRATED HEN WITH PFD OF PROPOSED HYDROGEN PLANT ..... 7-38 APPENDIX A – STREAM DATA ............................................................................. 7-39 REFERENCE ............................................................................................................. 7-40

Production of Hydrogen via Syngas Route

ix

CN 4120: Design II Team 32

Chapter 8 : COOLING TOWER .................................................................................... 8-1 8.1 PROBLEM STATEMENT .................................................................................... 8-1 8.2 WORKING PRINCIPLES OF COOLING TOWER ........................................... 8-2 8.3 Preliminary Design ................................................................................................. 8-3 8.3.1 Selection of cooling tower ................................................................................ 8-3 8.3.1.1 Justification to reject the use of natural draft tower ............................ 8-3 8.3.1.2 Justification to use induced draft tower ............................................... 8-4 8.3.2 Comparison between counter-flow and cross-flow Pattern ........................... 8-4 8.4 DETAILED DESIGN OF COOLING TOWER ................................................... 8-5 8.4.1 Specification of cooling tower design parameters........................................... 8-5 8.4.1.1 Wet bulb temperature ........................................................................... 8-5 8.4.1.2 Range ..................................................................................................... 8-5 8.4.1.3 Cooling water requirement ................................................................... 8-6 8.4.1.4 Approach ............................................................................................... 8-6 8.4.2 Exit air temperature and water to air flow ratio (L/G) .................................. 8-6 8.4.2.1 Exit air temperature .............................................................................. 8-6 8.4.2.2 Water to air flow (L/G) ratio................................................................. 8-7 8.4.3 Cooling tower characteristic............................................................................ 8-7 8.4.4 Loading factor .................................................................................................. 8-8 8.4.5 Dimensions of Tower ......................................................................................... 8-10 8.4.5.1 Fill Height ............................................................................................ 8-10 8.4.5.2 Base area .............................................................................................. 8-10 8.4.5.3 Fill volume ........................................................................................... 8-10 8.4.6 Make-up Water Requirement ....................................................................... 8-11 8.4.6.1 Evaporation loss (E) ............................................................................ 8-11 8.4.6.2 Drift loss (D)......................................................................................... 8-11 8.4.6.3 Blow-down (B) ..................................................................................... 8-11 8.4.6.4 Makeup water requirement (M) ......................................................... 8-12 8.4.7 Power Requirement ....................................................................................... 8-12 8.4.7.1 Pump power (Pp).................................................................................. 8-12 8.4.7.2 Fan Power (PF)..................................................................................... 8-13 8.5 COOLING TOWER INTERNALS ..................................................................... 8-14 8.5.1 Liquid Distributor.......................................................................................... 8-14 8.5.2 Fill................................................................................................................... 8-15

Production of Hydrogen via Syngas Route

x

CN 4120: Design II Team 32

8.5.3 Drift Eliminators ............................................................................................ 8-15 8.5.4 Supports ......................................................................................................... 8-16 8.5.5 Cooling tower basin ....................................................................................... 8-16 8.6 MATERIAL OF CONSTRUCTION ................................................................... 8-17 8.6.1 Liquid Distributor.......................................................................................... 8-17 8.6.2 Fills ................................................................................................................. 8-18 8.6.3 Drift eliminator .............................................................................................. 8-18 8.6.4 Mechanical support ....................................................................................... 8-18 8.7 COST ANALYSIS ................................................................................................ 8-19 8.7.1 Construction cost of for cooling tower .......................................................... 8-19 8.7.2 Operating Cost ............................................................................................... 8-20 8.7.2.1 Cost of makeup water .......................................................................... 8-20 8.7.2.2 Cost of Electricity ................................................................................ 8-20 8.7.2 Optimization between the operating and construction cost ......................... 8-21 8.8 ADDITIONAL CONSIDERATIONS TO COOLING TOWER DESIGN ........ 8-22 8.8.1 Water Treatment ........................................................................................... 8-22 8.8.1.1 Corrosion control................................................................................. 8-22 8.8.1.2 Scale control......................................................................................... 8-23 8.8.1.3 Biological control ................................................................................. 8-23 8.8.2 Environmental Concerns ............................................................................... 8-24 8.9 CONCLUSION ..................................................................................................... 8-25 REFERENCES........................................................................................................... 8-27 APPENDIX A IMPURITIES FOUND IN COOLING WATER .............................. 8-28 Chapter 9 : ECONOMICS & PROFITABILITY........................................................... 9-1 9.1 INTRODUCTION .................................................................................................. 9-1 9.2 ASSUMPTIONS ..................................................................................................... 9-1 9.3 CAPITAL COSTS .................................................................................................. 9-2 9.3.1 Computations for Fixed Capital ......................................................................... 9-2 9.3.2 Computations for Total Module Costs ............................................................ 9-7 9.3.3 Computations for Grassroots Costs (FCI) ...................................................... 9-7 9.3.4 Computations for Working Capital ................................................................ 9-8 9.4 MANUFACTURING COSTS ................................................................................ 9-8 9.4.1 Operating labour costs, COL .......................................................................... 9-10 9.4.2 Utility costs, CUT............................................................................................. 9-11

Production of Hydrogen via Syngas Route

xi

CN 4120: Design II Team 32

9.4.2.1 Electricity ............................................................................................. 9-11 9.4.2.2 Cooling water cost ............................................................................... 9-12 9.4.2.3 Waste treatment costs, CWT................................................................. 9-12 9.4.3 Raw materials costs, CRM .............................................................................. 9-13 9.4.4 Land lease, CL ................................................................................................ 9-13 9.4.5 Computation of manufacturing costs ............................................................ 9-14 9.4.6 Salvage value .................................................................................................. 9-15 9.4.7 Depreciation ................................................................................................... 9-15 9.4.8 Revenues......................................................................................................... 9-15 9.5 PROFITABILITY ANALYSIS............................................................................ 9-16 9.5.1 Land Cost ....................................................................................................... 9-17 9.5.2 After Tax Cash Flow ...................................................................................... 9-17 9.5.2 Rate of Return on Investment (ROROI) ....................................................... 9-18 9.5.3 Net Present Value (NPV) ............................................................................... 9-18 9.5.4 Discounted Cash Flows in Project ................................................................. 9-18 9.6 FEASIBILITY OF STORAGE FACILITIES FOR NATURAL GAS FEED ... 9-21 9.6.1 Capital Costs .................................................................................................. 9-24 9.6.2. Operating Costs ............................................................................................ 9-25 9.6.3 Overall Costs .................................................................................................. 9-25 9.6.4 Economic Compensation ............................................................................... 9-26 9.7 RECOMMENDATIONS ..................................................................................... 9-27 9.8 CONCLUSION ..................................................................................................... 9-29 REFERENCES........................................................................................................... 9-30 Chapter 10 : SAFETY, HEALTH & ENVIRONMENT (S.H.E.) ................................ 10-1 10.1 INTRODUCTION .............................................................................................. 10-1 10.2 HAZARDS AND OPERABILITY STUDIES (HAZOP) REVIEW ................. 10-2 10.3 PLANT LAYOUT ............................................................................................ 10-23 10.3.1 Segregation ................................................................................................. 10-23 10.3.2 Transportation Considerations ................................................................. 10-24 10.3.3 Administration ........................................................................................... 10-24 10.3.4 Laboratory ................................................................................................. 10-25 10.3.5 Workshop ................................................................................................... 10-25 10.3.6 Control Room............................................................................................. 10-25 10.3.7 Transformer Substation ............................................................................ 10-26

Production of Hydrogen via Syngas Route

xii

CN 4120: Design II Team 32

10.3.8 Emergency Services ................................................................................... 10-26 10.3.9 Amenities (Medical Centre and Canteen) ................................................. 10-26 10.3.10 Process and Auxiliary Units .................................................................... 10-27 10.3.10.1 Furnace (Housing SMR)................................................................ 10-27 10.3.10.2 Reactors (HTS, LTS), PSA and Knockout Drum ......................... 10-27 10.3.10.3 Cooling Tower................................................................................ 10-28 10.3.10.4 Heat Exchangers ............................................................................ 10-29 10.3.10.5 Flares .............................................................................................. 10-29 10.3.10.6 Wastewater Treatment Plant ........................................................ 10-30 10.4 OCCUPATIONAL SAFETY ........................................................................... 10-31 10.4.1 Personal Protection Equipment (PPE) ...................................................... 10-31 10.4.2 Noise ............................................................................................................... 10-32 10.4.3 Ventilation .................................................................................................. 10-33 10.5 OCCUPATIONAL HEALTH HAZARD IDENTIFICATION ...................... 10-34 10.6 ENVIRONMENTAL IMPACT ASSESSMENT ............................................. 10-38 10.6.1 Objectives ................................................................................................... 10-38 10.6.2 Risk Assessment Matrix............................................................................. 10-38 10.6.3 Elements of Environmental Impact Assessment ....................................... 10-46 10.6.3.1 Gaseous emissions ............................................................................ 10-46 10.6.3.2 Effluent discharge ............................................................................ 10-46 10.6.3.3 Waste management & minimization ............................................... 10-47 10.6.3.4 Energy efficiency ............................................................................. 10-47 10.6.4 Hydrogen Product Life Cycle Assessment ................................................ 10-48 10.6.4.1 Ramifications of Hydrogen LCA .................................................... 10-49 10.7 CONCLUSION ................................................................................................. 10-50 REFERENCES......................................................................................................... 10-51 Chapter 11 : INSTRUMETNATION & CONTROL.................................................... 11-1 11.1 INTRODUCTION .............................................................................................. 11-1 11.2 PROCESS CONSIDERATION AND DESCRIPTION .................................... 11-2 11.3 PROCESS CONTROL METHODOLOGY ...................................................... 11-3 11.4 SELECTION OF CONTROLLED, MANIPULATED AND MEASURED VARIABLE ................................................................................................................ 11-4 11.5 DETAILED CONTROL DESIGN FOR REFORMER FEED ......................... 11-5 11.5.1 Steam-to-Methane Ratio Control ................................................................ 11-5

Production of Hydrogen via Syngas Route

xiii

CN 4120: Design II Team 32

11.5.2 Pressure Control Loop for Expander.......................................................... 11-7 11.5.3 Temperature Control Loop to Preheat SMR Feed ..................................... 11-7 11.5.4 Composition Analyzer for SMR Effluent.................................................... 11-8 11.6 DETAILED CONTROL DESIGN FOR SMR FURNACE .............................. 11-9 11.6.1 Air-to-Fuel Ratio Control ............................................................................ 11-9 11.6.2 Temperature Control Loop to Regulate Effluent Exit Temperature....... 11-10 11.6.3 Pressure Control Loop to Regulate Furnace Draft .................................. 11-11 11.6.4 Flue Gas Exit Temperature Control ......................................................... 11-12 11.6.6 Analyzers for Furnace Control.................................................................. 11-14 11.7 Safety Devices ................................................................................................... 11-14 11.7.1 Pressure Relief Valves................................................................................ 11-14 11.7.2 Process Alarms ........................................................................................... 11-15 11.7.3 Safety Interlocks or Emergency Shutdown System (SIS or ESD)............ 11-16 11.7.3.1 Implementation of SIS or ESD for the protection of nickel catalyst . 1117 11.8 Additional Considerations in Process Control ................................................ 11-18 11.8.1 Redundancy of Air Blowers and Expanders ............................................. 11-18 11.8.2 Isolation Valves and Bypass ...................................................................... 11-19 11.9 REFERENCES ................................................................................................. 11-19

Production of Hydrogen via Syngas Route

xiv

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

Chapter 1 : PROBLEM DESCRIPTION 1.1 PROBLEM STATEMENT FOR PLANT DESIGN SPECIFICATIONS This project requires the production of hydrogen in Singapore. This has to be accomplished via a syngas route, which involves the governing Steam Methane Reforming (SMR) reactions, as well as the low and high temperature shift reactions. The following design specifications have been given and the following plant design endeavours to meet these criteria:
Location of Plant: Singapore
Operation Time: 8000 hours / year
Plant Capacity (PC): 1.25 × 109 m3(STP) / year
Feed Composition (FC) to SMR reactor: 3 : 1 (H2O : CH4)
% CO in H2 Specification at the exit of the shift converter: 0.7%
Purity of hydrogen product: > 99.9% (mole)
Natural Gas Feed: C1 = 97.7%; C2+ = 1.2%; CO2 = 0.7%; N2 = 0.4%

1.2 BACKGROUND FOR DEVELOPMENT IN HYDROGEN PRODUCTION

1.2.1 Energy Woes – Away from Fossil Fuels Era Recent years saw the rapid developments on alternative energies, in place of their conventional fossil fuels counterpart. The latter has several disadvantages [R4] associated with it, including: (i)

Air pollution (formation of NOx, CO & Unburned hydrocarbons contributing to urban ozone);

(ii)

Environmental pollutions (e.g. oil spill during transport)

(iii)

Global warming (emission of greenhouse gases) during combustion

(iv)

Dependence of fuel supply on oil-producing nations, which could result in dominance in oil prices

Production of Hydrogen via Syngas Route

1-1

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

Given these drawbacks, the search for an alternative fuel becomes more pertinent [R4]. One of the possible solutions is the production of hydrogen. In contrast, the latter promises: (i)

Exclusion pollution due to fossil fuels (by-product is H2O & hazards associated with spills are minimal)

(ii)

Exclusion of greenhouse gases

(iii)

Removal of price dominance, from the oil-producing nations

(iv)

Well-distributed production due to the ease of manufacture.

1.2.2 Identifying & Justifying the Production Route – SMR The uses of hydrogen extend way beyond the supply for fuels. For instance, hydrogen could be used in the petrol-chemical industries to make plastics products or it could be used to produce ammonia in the Haber process. In addition, it has been employed in the refineries to remove unwanted sulfur contents in crudes via the hydro-de-sulfurization (HDS) units. Nonetheless, hydrogen does not exist on Earth naturally. To harness of the above-mentioned uses, a plant has to be designed to produce hydrogen efficiently and safely. Typically, several methods (Gross, 2005) [R1] are available for hydrogen production. In the refineries, H2 can be produced in its in-house hydrogen plant or from the CRU (Catalytic Reformer Unit). H2 produced via the coal gasification route is not aimed at H2 production, rather, it is a by-product of coke production, such as the steel industry in Asia & Europe [R1]. With more advanced gasification processes, it could also increase the amount of H2 from coal by a considerable extent. Meanwhile, electrolysis of water promises H2 product of high purity, but this is dependent on the local costs for electricity. To make it economically more viable, electricity has to be available at a lower cost. Another instance of using electricity is the production of Cl2 and NaOH, namely the Chloroalkali process, whereby H2 is produced as a by-product. In fact, more recently, experimental works have gone underway to produce H2 via photo-electrolysis and biomass gasification. In this work, one of the most commonly used industrial processes has been adopted, which is the Steam Methane Reforming (SMR), which accounts for about 45% of world H2 production. This has been illustrated by the following diagram.

Production of Hydrogen via Syngas Route

1-2

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

Fig 1-1: Distribution of hydrogen production methods in industries Looking from the perspective of production on an industrial scale, the use of SMR would provide the economy-of-scale by providing more opportunities for heat integration (i.e. within SMR unit itself). This reflects a higher degree of optimization for the usage of utilities. Typically, this can be achieved via steam generation. The latter can be used for (i) for sale; (ii) for recycle as feed into SMR. Meanwhile, comparing to other methods (e.g. partial oxidation, auto-thermal reforming) of syngas production, the SMR route offered the following [R2] competitive advantages:
Lowest Tprocess required (better cost-savings)
Extensive industrial experience
Best ratio of H2 : CO for production applications of hydrogen
Does not require O2 (cost-savings & safety enhanced) With such encouraging advantages, the steam reforming process remains as the most mature and established form of technology to produce hydrogen [R3]. And indeed, several companies world-wide like Haldor-Topsøe, Howe-Baker, Foster Wheeler, Tokyo Gas Company, McDermott Technology Inc. and IDATech are employing SMR technology to manufacture hydrogen [R3].

Production of Hydrogen via Syngas Route

1-3

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

1.2.3 Choice and Significance of Reforming Feedstock – Natural Gas One of the major factors contributing to the operating characteristics of the reforming applications is the choice of processing feed for the reformer. In this work, natural gas (containing predominantly Methane, CH4) has been designated. Upon further research [R3], few possible reasons for using natural gas include:
Most economic & mature reforming technology
Lower environmental impact (few emissions, except CO2)
Supply of natural gas more readily available
Lower risk of coking (carbon formation) The following table adapted from literature shows the some of the noteworthy features for the various choices of reforming feedstock:

Table 1-1: Comparison on different steam reforming feedstock

Production of Hydrogen via Syngas Route

1-4

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

1.2.4 Steam Methane Reforming (SMR) Reactor In this design, the SMR unit has been employed to manufacture the product of interest – Hydrogen (H2). The earlier text has described and justified the need to use SMR and to have Methane (CH4) as the feedstock. In order to design an efficient reactor to meet the high CH4 conversion, it is imperative to consider the key factors that played an instrumental role in influencing the performance of the reactor. (I) Tube Geometry (related to tube length & diameter, average heat flux & space velocity) •

↑ Tube Length more economical than ↑ No. of tubes
↑ No. of tubes complicate design at reactor’s inlet and outlet. • Limit for Tube Length
Threat of tube bending.
Risk of too drastic pressure drop over the catalyst bed. • ↑ Tube diameter to be accompanied with ↑ Tube Wall thickness
For thinner tubes, ↓ temperature required & better heat transfer (↑cost savings).
Also, less tubes need to be used to meet required conversion. (II) Firing Configuration (Bottom vs Top vs Terrace vs Side) •

Side-Fired Configuration (with short flames distributed along reactor wall)
Higher level of regulatory control over Tube Wall Temperature.
↑ Design and operational flexibilities.
↑ Average Heat Flux for higher conversion.
Endure more severe reaction conditions.
Lower NOx levels produced in flue gas stream.

•

Construction of tubes
Higher level of regulatory control over Tube Wall Temperature.
Creeping strength is a strong function of the choice for tube material

(III) Catalyst (intrinsic activity, surface area, microstructure, porosity, mechanical resistance, thermal & chemical stability, resistance to carbon deposition) •

Catalyst Structure
Provision of support to give stable micropore system, overcoming sintering issue when process temperature is above Tamman temperature (Ni: 863K)
Low surface area carriers employed due to high temperatures involved.
Crux: Maximize catalyst activity & heat transfer; Minimize Pressure Drop Table 1-2: Key design considerations for SMR reactor unit

Production of Hydrogen via Syngas Route

1-5

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

The SMR reactor typically consists of multiple catalyst-loaded tubes housed within a furnace. The latter provides the much-needed heat duty due to the characteristic endothermic reforming reactions, in which CH4 reacts with steam to give the desired H2 product. Given that the SMR unit represents the heart of the operations for the plant, it is essential for us to fulfill the key design considerations, as given in Table 1-2.

1.2.5 Furnace The furnace provides heat to support the endothermic SMR reactions. In this design, the fuel feed used for the combustion is harnessed from the purge stream of the Pressure Swing Adsorption (PSA) unit, considering its high H2 (as compared to CH4) content. However, this purge stream has high carbon dioxide (CO2) content, which does not support combustion. Consequently, an amine scrubber is also proposed to remove this undesired CO2. Nonetheless, in the event of insufficient fuel supply by the PSA purge stream, it is recommended to make up with a natural gas fuel feed. This could originate from the feedstock of SMR reactor. It is noteworthy that combustion typically occurs at atmospheric pressure, hence, an expander is to be installed to decrease the pressure of the SMR natural gas feedstock, before allowing the fuel to proceed to the burner.

To demonstrate the advantages conferred by the side-fired configuration, small premix burners would line up along the walls of the straight wall furnace, as such burners provide short flame length and ease for temperature control. The flue gas generated from the combustion process carries a net amount of heat for which is transported upward to the convection section through the use of induced draft fan. The heat carried by the flue gas is then used to heat up the process streams passing through the convection section.

In the convection section, steam is generated within the tubes closest to the radiation section. This is followed by two other process streams, namely, SMR feed and combustible air. Steam generation is situated closest to the radiation section because heat transfer is most efficient for heat exchange between two different phases. Through effective process control and instrumentation, the process variables within the furnace are kept constant. This helps to maintain the product yield, while keeping the operating environment safe.

Production of Hydrogen via Syngas Route

1-6

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

1.2.6 Shift Reactions The main purpose of the high temperature shift reactor is to derive more H2 from the one of the SMR products, carbon monoxide (CO). The feed into the High Temperature Shift (HTS) reactor is at a relatively high level of CO as it exits from the steam methane reformer. This CO was reacted with steam to form more H2 with the following water gas shift equation: CO + H 2 O ⇔ CO2 + H 2

∆H rxn = −44.447 kJ / mol

There is a need to couple the high temperature shift reaction with a low temperature shift (LTS) reaction because of the exothermic nature of the water gas shift reaction. Therefore, high conversion occurs at low temperatures. However, the rate of reaction is too slow (i.e. compromised) at low temperatures. Thus, the HTS reactor is employed to ensure a high reaction rate, while its LTS counterpart maintains the required conversion. Based on iron oxide as the catalyst, the design of the HTS was able to convert a 13.34% CO feed, to 3.0% CO, after which it is fed into the LTS. The designed conversion of the high temperature shift reactor was 75.27%. Due to the adiabatic reaction in high temperature shift, the temperature of the feed was raised from 627K to 692K. The feed was cooled to 493K prior to entry into the low temperature shift reactor, which employed the Copper-Zinc Oxide catalyst supported on alumina. A CO conversion efficiency of 82.9% was obtained, which corresponds to a 0.5mol% CO (dry basis) in the outlet stream of the LTS. This product stream was then transferred to the knock-out drum, where liquid water was separated from the other gaseous products. The latter then flowed to the PSA columns for further purification.

1.2.7 Product Purifications For this design, pressure swing adsorption (PSA) was adopted as the preferred mode of purification due to the stated requirement of attaining 99.9% in product purity, which otherwise was not achievable through the conventional use of a CO2 scrubber and a methanator (95–97%).

Production of Hydrogen via Syngas Route

1-7

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

The high concentration of hydrogen (75% at the entrance) could lead to possible hydrogen embrittlement, thus the material of construction chosen was carbon steel clad with stainless steel as this material possessed a lower material factor as compared to stainless steel, which translated to a lower bare module cost (1.8 vs 3). Through the prior installation silica gel, the amount of water entering PSA after exiting the knockout drum was assumed to be negligible. Thus the chosen adsorbents for PSA were activated carbon and zeolite 5A. Activated carbon was utilized to remove hydrocarbons, such as CH4, C2+ and CO2 due to the preferential adsorption isotherms that these components exhibit with activated carbon. Similarly, zeolite 5A was employed to remove CO and N2. According to the Polybed design, this comprises of 7-10 beds with the incorporation of various operation steps, such as pressurization, high pressure adsorption, blowdown and purge, a final product of 99.9% purity and 85% hydrogen recovery was attainable. Subsequently, the PSA tail gas was routed to the furnace as a source of fuel for combustion.

1.2.8 Heat Integration Energy integration involves the usage of process streams within the plant itself to fulfil the heating and cooling requirements at various points of the process. An optimal solution would be of utmost importance in a chemical plant, as this would help to mitigate the rising cost of utilities associated with increased fuel cost. Therefore, to achieve optimal heat integration, the systematic development of a heat exchange network (HEN) would have to be carried out. The usage of a HEN would be an integral step in the maximization of energy recovery. The use of pinch analysis would be critical in lessening the requirements for hot and cold utilities, which are major components of the operating cost of the hydrogen plant. However, this must be balanced with the increased capital investment associated with the installation of heat exchangers. It was found that a furnace was required to provide the necessary heat of reaction for SMR in normal operation. Preliminary calculations showed that the large amount of heat produced by the furnace provided for the entire heating requirement in the hydrogen plant.

Production of Hydrogen via Syngas Route

1-8

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

This results in the lack of need for hot utilities, i.e. a threshold problem ensued with only cold utilities required. Hence, possible cold utilities to be considered would be the generation of high or low pressure steam, and the usage of cooling water to cool low-grade heat. This could result in the presence of a utility pinch, which would be tackled by a similar application of pinch analysis, treating the utility stream as a dummy process stream. A multitude of network variations and possible evolution of the network would also be considered to obtain the most economical and practical solution for the energy integration of the designed hydrogen plant.

1.2.9 Cooling Requirements The main purpose of the cooling tower is to reject the low grade heat absorbed from process stream into the atmosphere by means of latent heat of evaporation and sensible heat transfer. The cooling tower in this hydrogen plant is designed to provide a continuous flow of cooling water required for the condensation and elimination of water vapor in the outlet stream of the LTS reactor, before it is fed into the PSA columns for purification of H2 and removal of CO2. The design of the cooling tower is based on an induced draft counter-flow configuration. This is because this type of configuration does not experience any recirculation which can cause a drop in cooling tower efficiency due to higher wet bulb temperature and in the long run, it is more economical due to lower power requirement for auxiliary units such as fans and pumps. In the design of this cooling tower, it is assumed by heuristic that the maximum inlet temperature of cooling water to be 120˚F and cooling water exit temperature to be 90˚F and the ambient wet bulb temperature is derived from the average daily maximum dry bulb temperature and mean humidity. Hence, the performance of the cooling tower can be optimized by manipulating the exit air temperature and it is found to be 105˚F, which is the average of the inlet and outlet water temperature.

Production of Hydrogen via Syngas Route

1-9

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

1.2.10 Use of HYSYS Simulation The simulation of the hydrogen plant was performed in Hysys. Care must be exercised in the selection of the fluid package of choice as any unsuitability would be reflected in the obtainment of inaccurate simulation results. Thus Peng-Robinson (PR) Equation Of State (EOS) was adopted as the preferred fluid package. AspenTech recommended it for oil, gas and petrochemical applications due to its special enhancement in HYSYS for the generation of accurate phase calculations over a wide range of operating conditions (T > -271°C, P < 1000kPa). Our reaction conditions were well within the range. Furthermore, literature values obtained for the reactor units had been based primarily on the PR EOS. The PSA was reflected as a component splitter in the PFD. PSA was a process unit that could not be adequately simulated in Hysys, thus its simulation was performed in COMSOL. The target specifications for the various major units had been met with the convergence of the Hysys simulations, which also implied an overall satisfactory plant design.

1.3 REFERENCES [R1]: Tom Gross. (2005). Hydrogen – An Overview. Foundation for Nuclear Studies Briefing. [R2]: Wilhelm, D., Simbeck, D., Karp, A., Dickenson, R. (2001). Syngas production for gasto-liquids applications: technologies, issues and outlook. Fuel Proc. Tech., Vol 71 – P139 [R3]: Ferreira-Aparicio, P., Benito, M. J. & Sanz, J. L. (2005). New Trends in Reforming Technologies: from Hydrogen Industrial Plants to Multifuel Microreformers. Catalysis Reviews, 47:4, P491-588. [R4]: Marshall Brain. How the Hydrogen Economy Works. Adapted on 15th Apr 2008 from: http://auto.howstuffworks.com/hydrogen-economy.htm

Production of Hydrogen via Syngas Route

1-10

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design

Chapter 2 STEAM METHANE REFORMER 2.1 PROBLEM STATEMENT 2.1.1 Problem and Specifications In this report, a Steam Methane Reformer (SMR) reactor unit is to be designed. The plant is to produce hydrogen via the syngas route. The SMR reactor is one of the first units in the process stream, and hence its design would be critical for the downstream process units, in a bid to achieve an overall economical, safe and efficient plant for the hydrogen production.

Amongst all, the design specifications for Team 32 are shown as follow: •

Location of Plant: Singapore

•

Operation Time: 8000 hours / year

•

Plant Capacity (PC): 1.25 × 109 m3(STP) / year

•

Feed Composition (FC) to SMR reactor: 3 : 1 (H2O : CH4)

•

% CO in H2 Specification at the exit of the shift converter: 0.7%

•

Purity of hydrogen product: > 99.9% (mole)

•

Natural Gas Feed: C1 = 97.7%; C2+ = 1.2%; CO2 = 0.7%; N2 = 0.4%

2.1.2 Justifications for using SMR Justifications to leverage upon the SMR reactor unit for hydrogen production have been found in literature. For instance, Wilhelm et. al. (2001) [R8] described the following advantages, which are aligned with the current intention of the usage of the SMR unit. These advantages have made SMR the chosen reforming concept. Hence, this project endeavours to produce hydrogen via this syngas route.
Lowest Tprocess required (better cost-savings)
Extensive industrial experience
Best ratio of H2 : CO for production applications of hydrogen
Does not require O2 (cost-savings & safety enhanced)

Production of Hydrogen via Syngas Route

2-1

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.2 DESIGN METHODOLOGY & PROCESS DESIGN 2.2.1 Outline of Design Methodology

Fig 2.2.1a Flowchart to illustrate design methodology 2.2.2 Reaction Chemistries CH4 CO CH4

+ + +

H2 O H2 O 2 H2 O

↔ ↔ ↔

CO CO2 CO2

+ + +

3 H2 H2 4 H2

Eqn (2-1) Eqn (2-2) Eqn (2-3)

3 governing equations responsible for the reactions in the reactor are given as above. At this point, it is crucial to note that Beurden (2004) [R24] described that Eqn (2-3) is not a combination of the Eqn (2-1) and Eqn (2-2) as CO2 is produced in both Eqn (2-2) and (2-3), implying that the latter itself does not represent an overall reaction.

Production of Hydrogen via Syngas Route

2-2

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.2.2.1 Effects of Temperature and Pressure – Revisiting Le Chatelier’s Principle La Chatlier’s Principles can be used to explain the effects of the operating conditions for a typical SMR. A higher inlet temperature (typically 723 – 923K [R25]) would drive the endothermic reactions (Eqn (2-1) & (2-3)) forward to produce more H2 product. This is in contrast with that of the water-gas- shift (WGS) reaction (Eqn (2-2)), which is favoured at lower temperature and not affected by pressure (same molar ratio on both sides of reaction  no volume expansion). Meanwhile, the stoichiometries of these 2 reforming reactions also suggested that forward reactions are favoured when a lower pressure is used. This is to allow for volume expansion to occur since the number of moles of product is greater than that of reactants.

2.2.2.2 Coke Formation, Steam:Methane ratio & Inclusion of CO2 in feed Also, the Steam:Methane ratio (sc) used is 3. This coincides with what is typically found in industrial practices, which suffice in suppressing coke formation [R24] during the reaction. The presence of the carbon deposits during coke formation is detrimental to the process as it would result in tube blockage, forming hot spots that can very well destroy the tubes, threatening both the economics and safety of the process. Since this SMR reactor unit design does not consider formation of coke, the choice of sc = 3 is made during the start of the project to favour the design considerations of not involving coking as one of the reactions. The suppression of coke formation is further promoted by the inclusion of CO2 [R25] in the feed gas (Boudouard reaction during coking: 2CO = C + CO2), as mentioned in the design brief. This shifts the Boudouard reaction backwards and thereby suppressing coke formation. In addition, adding CO2 at the inlet of the reformer helps to save on hydrocarbon feedstock and decrease the H2:CO formed in the SMR product stream. With these advantages in mind, in industrial practice, some of these CO2 are typically being recycled from the SMR product stream or being imported from another source.

Production of Hydrogen via Syngas Route

2-3

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.2.3 Choice of Reactor – Tubular Reformer Nielsen (1993) indicated that the steam-reforming reaction typically involves catalysts being loaded into tubes, which are in turn housed in a furnace to satisfy the highly endothermic reaction. The tube material has to be capable to withstand the high temperature and the temperature gradient (e.g. 1223 K at outlet [R25]).

As such, these tubular reactors typically experienced very huge stresses. Given that upper limit of the tolerable stress value for the tubes is very much affected by the maximum tube wall temperature and heat flux, a small rise in the maximum tube wall temperature could very well resulting a reduction of life expectancy for these tubes. Typical average lifespan of these reformer tubes can be around 100,000 hours. Given that the current plant is designed to run at 8000 hours/yr, this would allow use for up to a good 12.5 years. Such tubular reformer would be choice reactor for the current design because it allows catalysts

Fig 2.2.3a: Typical Natural Gas & Reformer Catalysts. Retrieved from Midrex on World Wide Web: http://www.midrex.com/uploads/documents/Catalyst(1)1.pdf

Production of Hydrogen via Syngas Route

2-4

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Fig 2.2.3b: Tubular reformer configurations [R23, R25]

2.2.4 Justifications for Choice of Firing Configuration – Side-fired reformer furnace The side-fired heating configuration is chosen because this has:
Provided greater degree of control for Ttube wall to allow a more robust operation, to meet the demands of the production by enduring more severe operating conditions. Also, a higher average heat flux of 313800 kJ/h/m2 [R26] can be allowed.
Shorter residence time discouraged [R25] formation of nitrogen oxides (NOx), up to <50ppm, which is critical for the current design since it is assumed that no NOx is formed in the SMR unit, even though there is presence of some N2 in the feed gas and that the reactor is to be operating at high temperatures.

Production of Hydrogen via Syngas Route

2-5

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.2.5 Justifications for Choice of Fluid Package – Peng-Robinson The Peng-Robinson fluid package is chosen because it gives a better prediction of liquid densities than the Soave-Redlich-Kwong equation of state [R19]. In fact, the PengRobinson equation of state is the most commonly used for systems containing non-polar components [R20]. From the Table 2.2.4a below, it can be determined that Peng Robinson fluid package remains the best choice as the system [R21].

Table 2.2.4a: Recommended Property Package based on type of system

2.2.6 Choice of Catalyst Catalyst for steam reforming is typically nickel-based, which is cheaper than its noble metal counterparts. Key criteria (besides costing) are to maximise activity and heat transfer, while not causing a huge pressure drop. Non-metallic options are yet to be made commercial because of the inherent low catalyst activity and thereby the impact of pyrolysis [R27]

. The current work is based on that of Rajesh et. al. (2000) but there is no clear mention

of the catalyst used. Hence, the catalyst would follow that used by Xu and Froment (1989) [R3,4]

, which is mentioned by Rajesh et. al, namely Ni/MgAl2O4. In this case, note that the

stability of the MgAl2O4 catalyst support is also very crucial in order to withstand conditions during operation, start-up and shut-down [R27].

Production of Hydrogen via Syngas Route

2-6

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.2.7 Kinetics, Ni-based Catalyst & Role of Support To facilitate the reforming reactions, the Ni-based catalyst is added to lower the activation energies requirements. Intrinsic activity of the catalyst depends [R25] on the catalyst surface available for chemisorption to occur. To do so, the cleavage of a C–H is needed, and this entails overcoming the barrier of 52 kJ/mol and contacting a free Ni adsorbent site with free neighbouring site. This chemisorption is of CH4 plays an instrumental role in determining the reaction rates of the SMR process (typically given by CH4=CH3*  CH2*  CH*=C*). Indeed, this is also aligned [R25] with the typical 1st order kinetics mentioned by many [R2,3,4,5].

Nonetheless, as observed from Eqn (2-17) in Chapter 2.12.2.1, the presence of the equilibrium adsorption constant for water (KH2O) may explain for a negative reaction order with respect to the feed steam, giving rise to a ‘retarding effect’ whose extent may vary with the choice of catalyst. In view of that, it is thus important to include alkali or use of magnesia as support for the Ni catalyst, which provide for an enhanced level of adsorption of steam molecules to avoid carbon formation.

2.2.8 Justification for Choice of Reactor Inlet Conditions Rajesh et. al. (2000) [R1] suggested that inlet temperature for the SMR reactor should not be lower than 725 K due to thermodynamic limitations, which thus prevent possible formation of gum on the reformer catalyst. This would block the catalyst surface [R24] via polymerisation of the adsorbed CnHm radicals. Such progressive deactivation is possible since species like ethane (C2H6) is one of the components in the natural gas feed for the current project. Meanwhile, operating at temperatures higher than 900 K is also not probable because of the limitations for the maximum amount of heat energy that can be harnessed form the flue gas. Meanwhile, it is also suggested [R1] that the inlet pressure is typically between 2400 – 3000 kPa, after accounting for the normal pressures at which H2 is to be produced, and the presence of natural gas as the feed. With these in mind, the arithmetic averages for the inlet temperature and pressure (812.5 K and 2700 kPa) are used during the detailed design of the SMR reactor unit.

Production of Hydrogen via Syngas Route

2-7

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.3 PRELIMINARY DESIGN 2.3.1 Establishment of Base Case The governing reactions, key reactions and products for the reactor are first identified. From the specified plant capacity provided in the design brief, and given the conversion specified in literature (see trailing paragraph), a preliminary mass balance is then performed across the reactor unit, after obtaining the SMR outlet flow rates from mass balance for the inlet of trailing High Temperature Shift (HTS) unit. This gave an initial value for the flow rates for the feed to be used.

Research efforts have been invested to find the conversion of the major component, namely Methane (CH4). Coincidentally, for a Steam:Methane ratio of 3, Moulijn et. al. (2001) [R9]

has reported on the expected methane slip is to be 21% (assuming P = 2700 kPa (27

bar); T = 1123.15 K), corresponding to a 79% (i.e. 100% - 21%) CH4 conversion. Fig 2.3.1a: Literature data Moulijn et. al. (2001) [R9] to support conversion obtained during preliminary design is valid at the assumed conditions (2700kPa & 1123.15K)

2.3.2 Preliminary Simulation using HYSYS Initial efforts for SMR reactor design has been made via data from Hou & Hughes (2001) [R6] literature data and using the Equilibrium Reactor module in HYSYS as a simulation tool.

Production of Hydrogen via Syngas Route

2-8

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Accordingly, the reactors were stainless steel tubes with 0.01m internal diameter and 0.338m long. Catalyst used was ICI steam reforming catalyst 57-4, which is of a cylindrical fashion with 4 axial holes and provided by ICI Katalco. These catalysts have been crushed prior to their usage. Amongst the various reaction equations given, the 3 predominant ones haven seen selected and their respective equilibrium constants have been given.

Reaction

Reaction Equilibrium Rate constant, Ki

1) CH4 + H2O = CO + 3H2

1.198 × 1017 exp(-26830/T)

Eqn (2-4)

2) CO + H2O = CO2 + H2

1.767 × 10-2 exp(4400/T)

Eqn (2-5)

3) CH4 + 2H2O = CO2 + 4H2

2.117 × 1015 exp(-22430/T)

Eqn (2-6)

Table 2.3.2a: Reaction Equilibrium Rate constants, from Hou & Hughes (2001) [R6]

For a preliminary estimate, assume if TSMR, Exit = TSMR, Equilibrium = 1123.15K, then: K1 = 5.058 × 106 (kPa)2; K2 = 8.884 × 10-1; K3 = 4.484 × 106 (kPa)2 With these Ki expressions, an Equilibrium Reactor module is specified in the HYSYS Simulation Environment. Inlet pressure is assumed to be 270kPa (arithmetic average of the TSMR, Inlet from the Rajesh et. al. [R1]), while inlet temperature is set to be 923.15K [R??]. Further research depicts an allowable pressure drop of 200 kPa [R27] may be typically used across the SMR reactor unit. Overall, the simulation has achieved a CH4 conversion of ≈ 79 mol%, which corresponds to that stipulated by the literature, as mentioned in Chapter 2.3.1. The resultant process flow conditions are then presented in the interim report. The latter would be used as an initial estimate for the respective downstream units. Note: Streams and reactors data obtained at this stage from the HYSYS Process Flow Diagram (PFD) via use of the Equilibrium Reactor) module merely provide initial iteration values for solving of MATLAB Ordinary Differential Equations (O.D.E.s) for the detailed design.

Production of Hydrogen via Syngas Route

2-9

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.4 DETAILED DESIGN 2.4.1 Development of Critical Profiles via MATLAB & Optimisation To materialise the design of the reactor in detail, MATLAB is employed to solve for the O.D.E.s given in literature [R1], after obtaining initial iteration values from the preliminary design. The intention is to develop critical profiles and useful data via literature data and assumptions, and fit the findings into the HYSYS Simulation Environment (using a PFR module) for the overall plant design.

The MATLAB code developed could be found in the Appendix (Chapter 2.12.1). 2 m-files are written. The first one (Chapter 2.12.1.1) specified the reaction inlet conditions and the tube dimensions, while using ode15s to resolve the O.D.E.s along the axial direction. The second m-file (Chapter 2.12.1.2) specified the 4 O.D.E.s to be solved, as well as how each parameter is obtained from literature.

Optimization is then performed by manipulating the various parameter like heat flux, number of tubes and tube inner diameter being used. Since bulk of the written code followed closely to the O.D.E.s developed by Rajesh et. al. (2000) [R1], the arithmetic averages for the inlet pressure (2700 kPa) and temperature (812.5 K) of the range recommended by the same journal are used to be the base case in this case.

2.4.2 Design Equations & Key Assumptions Rajesh et. al. [R1] summarised the kinetic model and energy balance of the steam reforming process, using a side-fired configuration with Mg-supported Ni catalyst. The 4 major O.D.E.s to be resolved via MATLAB are shown as follow. dχ CH 4 dt dχ CO2 dt

=( =(

πd i 2 Rη CH rCH 4

)

4

F

πdi 2 RηCO rCO 4

)

2

F

2

4

; χ CH 4

; χCO2

t =0

t =0

=0

=0

1.75G s (1 − catbedvoid ) dP =− ; P t =0 = Pi dt φ s D p (catbedvoid ) 3 ρ g

CH4 Differential Mass Balance

Eqn (2-7)

CO2 Differential Mass Balance

Eqn (2-8)

Momentum Balance

Eqn (2-9)

Production of Hydrogen via Syngas Route

2-10

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U) 3  4H F  dT 1 = + ρ b ∑ ( − ∆H i )η i ri ) ; T  dt Gmass Cp overall  d i i =1 

SMR Unit Design Report

t =0

= Ti

Energy Balance

Eqn (2-10)

There 4 O.D.E.s have to be resolved to observe the profiles of how the CH4 conversion, CO2 conversion, pressure and temperature vary along the length of the reactor (t). Expanded form of the expressions for the rates, effectiveness factors, adsorptions and rates constants are available in Chapter 2.12.2.

Several key assumptions are made, including: 1) Constant heat flux is assumed, which implied [R28] that a well-controlled firing rate is assumed. This is more probable [R27] in side-fired furnace used in this work due to greater degree of adjustments and control over the tube wall temperature. Consensus is also reached by Furnace Team to use a constant heat flux for the current work. 2) Effectiveness factor <ηi > for the ith reactions, which vary along the tube length, are fitted with Excel, whose coefficients of the polynomial equations are then input into MATLAB. 3) There is negligible (0.001H2:1CH4) H2 recycle from the Pressure Swing Adsorption (PSA) unit to the reformer feed. 4) Feed ratios <sc, dc & nc> and the total mass flow rate are obtained from HYSYS PFD. The Adjust Function is used for more accurate values of the component molar flow rates. 5) No formation of NOx due to use of short flames of the side-fired furnace (Chapter 2.2.4). 6) No coking due to presence of CO2 in the reformer feed (Chapter 2.2.2). 7) N2 and higher carbons (C2+ = C2H6 in this work) remained inert in the SMR unit. Typical compositions of Natural Gas feed may be found in Chapter 2.12.4 of the Appendix. 8) Catalyst dimensions and characteristics follow that of Rajesh et. al. (2000) [R1]. 9) Density of process gas < ρ g > varies along the tube length , which is taken to be the division of the mass flow rate over volumetric flow rate. 10) Volumetric flow rate is taken to Molar flow rate × Universal gas cons tan t × Temperature Pr essure

for any axial position , hence Ideal Gas Law is assumed here. 11) For the MATLAB code being written, expressions for C2 is not available in the literature used [R1], hence this is lumped the other inert species N2. However, in the HYSYS Simulation, these two species are distinguished.

Production of Hydrogen via Syngas Route

2-11

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.4.3 Fitting into HYSYS Simulation Environment using Plug-Flow Reactor (PFR) The HYSYS PFR module is employed to simulate the findings from MATLAB by: 1) Obtaining conditions of inlet and outlet streams, and also tubing and catalyst specifications [R1,2,3,4] from optimisation in MATLAB, which are then fitted into the HYSYS Simulation Environment. 2) Upon convergence in HYSYS, initial flow rates are then fitted back into MATLAB to regenerate the critical profiles. These profiles are to be aligned with that in HYSYS. As the flow fashion is now being modelled as plug flow [R7] in HYSYS, it is assumed that no axial mixing occurs. This coincides with the intention of generating 1-D critical profiles (with respect to Length of Reactor, t). Also, a Heterogeneous Catalytic reaction set is chosen since the SMR reactor unit involves In the current HYSYS simulation, the PFR is being segmented to 50, instead of the default value of 20. Hesketh (2003) [R7] described that the increased number of steps conferred higher accuracy when resolving the O.D.E.s, since now more steps used to resolve the O.D.E.s.

2.4.4 Results and Discussions 2.4.4.1 Conversion profiles for CH4 and CO2 Fig 2.4.4.1a: Conversion Profiles from MATLAB

Production of Hydrogen via Syngas Route

2-12

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

A desirable conversion of up to 80% is achieved for methane (CH4). This is in close agreement with results obtained in the preliminary findings (Chapter 2.3.2), as well as that in literature (Chapter 2.3.1). This also affirms that the SMR unit design is able to meet its conversion targets for H2 production, as proposed in the interim report.

2.4.4.2 Temperature and Pressure Variations

Fig 2.4.4.2a: Pressure and Temperature profiles along length of reformer tube Nielsen (1993) [R25] reported that typical outlet can be as high as 1223 K (9500C). Hence, the current SMR exit temperature of 1100 K is still lower than the literature value. As mentioned earlier in Chapter 2.2.3, it is noteworthy that neither very high temperature nor great temperature gradient is encouraged since this may increase stress on the reformer tubes, which greatly reduces the lifespan of the tubes.

Meanwhile, the pressure drop is about 65 kPa, which is lower than the proposed drop of 200 kPa (2 bar) in the interim report. A lower pressure drop would mean that downstream compressions could be avoided/minimised [R26], resulting in cost savings.

Production of Hydrogen via Syngas Route

2-13

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.4.4.3 Component Mole Fractions

Fig 2.4.4.3a: Component Mole Fractions from MATLAB

Fig 2.4.4.3b: Component Mole Fractions from HYSYS

Production of Hydrogen via Syngas Route

2-14

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Comparing Fig 2.4.4.3a & b with that of Fig 2.4.4.3c, it can be observed that the simulated profiles from both MATLAB and HYSYS PFR are aligned with that in literature by Rajesh et. al. (2001) [R12]

on multi-objective optimization.

This may signified that the current chosen design configuration has also been optimized.

Fig 2.4.4.3c: Graph from literature on mole fraction profiles [R12] 2.4.5 Optimization Optimization has been performed to obtain the desirable reformer tube configuration and operating conditions. At this juncture, since costs of the various materials (e.g. catalyst, tube materials) are typically proprietary information, hence, the chosen tube dimensions and the conditions of operation is based on other parameters. Through MATLAB, graphs of increasing and decreasing a particular parameter (e.g. inner diameter, di) are painstakingly plotted. One example is as shown in Fig 2.4.5a.

Fig 2.4.5a: Effect of varying tubular inner diameter on CH4 conversion profile

Production of Hydrogen via Syngas Route

2-15

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Fig 2.4.5b: Effect of varying pressure on CH4 conversion profile

Table 2.4.5a: Effect of Manipulating Parameters on Critical Profiles (from MATLAB)

As illustrated above, trends observed by increasing and decreasing selected parameters are summarised in Table 2.4.5a. After several rounds of optimizing and fine adjustments, the operating conditions and stream properties are presented in Chapter 2.4.6.

Production of Hydrogen via Syngas Route

2-16

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.4.6 Operating Conditions & Streams Conditions

Table 2.4.6a: Stream Properties & Operating Conditions (HYSYS PFR & MATLAB) For overall integration purposes and in view that both HYSYS & MATLAB values are in close agreement, these values from HYSYS are passed down to the downstream units. Note that from HYSYS, a heating duty of 5.949 × 108 kJ/hr is required.

Production of Hydrogen via Syngas Route

2-17

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.5 MATERIALS OF CONSTRUCTION & SIZING 2.5.1 Selection Methodology Selection of reactor materials has been made via the following considerations [R15]. •

Conditions of exposure for reactor equipment being determined.

•

Availability of materials being explored.

•

Suitable material being identified.

•

Ensured choice of material being substantiated with certifications.

The reactions involved exposed the SMR reactor unit to high temperature and pressure. Comparisons between operating conditions for using MATLAB and HYSYS PFR module revealed that these are typically as high as 1130 K (15750F) and 2700 kPa (391 psi) [R16]

. In addition, there are chances of hot gas corrosion due to the high mass velocity.

2.5.2 Justifications for selecting from different grades of stainless steels In view of these conditions, stainless steel is a suitable material for the construction of these SMR reactor tubes, which can be summarised as such:
Higher C Content  Offers greater creep resistance than other metals
Addition of Ni and Cr  Resistance to carburization and creep being enhanced. Different grades of heat resistance steel, namely, HH, HK, HD and HF. HK have been specifically found to be of great use for SMR due to their high creep and rupture strength even up to 1150°C. Most importantly, they offer resistance to hot gas corrosion. Literature [R16]

revealed typical material of choice for SMR is HK40. Fig 2.5.2a illustrates the relative

tensile strengths of the different stainless steel grades.

Production of Hydrogen via Syngas Route

2-18

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Fig 2.5.2a: Relative tensile strengths of different stainless steel grades

2.5.3 Tube life estimation, Minimum Stress Rupture & Identification of Choice Material From Fig 2.5.3a [R16], at the high inlet pressure of 2700kPa (391 psi) and at the high outlet temperature of 1130 K (1575 °F) for the reactor, the furnace tube life is found to more than 20.3 years. This provides a confirmation that the material can be employed as suitable for use in steam methane reforming processes. Meanwhile, Fig 2.5.3b [R16] illustrates the minimum stress to rupture for the HK40 material as compared to other grades. The figure implies a lower performance of HK40 grade compared to HP grades. However, since the maximum temperature is 1130 K, which is low compared to the maximum temperature at which these stainless steel grades can withstand, the pressure factor is taken for higher consideration in selection of the suitable metal type. Since HK40 is capable of withstanding high pressure, HK40 would thus be chosen as the choice of material for the reformer tubes (assuming economic considerations are ignored).

Production of Hydrogen via Syngas Route

2-19

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

Fig 2.5.3a: Estimation of furnace tube life for a given set of operating temperature & pressure

Fig 2.5.3b: Minimum Stress to rupture for chosen material

Production of Hydrogen via Syngas Route

2-20

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.5.4 Sizing – Computation for Tube Thickness HK40 (25% Cr, 20% Ni) are usually used for steam reformer tubes [R16]. It has been chosen for its high creep and rupture strengths. It is also resistant to hot gas corrosion and hence is usually employed in steam methane reforming processes [R15]. Meanwhile, it provides creep resistance up to 980°C (1253 K) [R17], making it a good candidate for the current design whereby the highest temperature 856.85°C (1130 K). A standard code formula is employed here to compute the minimum wall thickness required [R18], for sizing purpose and specification of the PFR module in the HYSYS Simulation Environment.

min t wall

d  P i + FCA  2  =  S a E − 0.6 P 0.1448 psi   0.127 m 1in   × + 0.039in   2700kPa × × 1kPa   2 0.0254m  = 1000 psi 0.1448 psi 3.2ksi × × 0.85 + 0.6 × 2700kPa × 1ksi 1kPa 0.0254m = 0.399in × 1in −2 = 1.01 × 10 m

P = Max Pressure (to be in psig) = 2700 kPa (highest pressure at inlet); di = Inner Diameter (to be in inches) = 0.127 m; FCA = 10-year corrosion allowance (to be in inches) = 0.039 in [R18]; Sa = Minimum Creep Stress for HK40 (to be in psi) = 3.2 ksi; E = Weld Efficiency Factor = 0.85 [R17]. Hence, a tube wall thickness of 1.01×10-2 m would be used.

2.5.5 Sizing – Summary

Inner Diamter (di)

: 0.127 m (5”)

Outer Diameter (do) : 0.147 m

Wall thickness

: 0.0101 m

No. of tubes needed : 450 (HK-40 Steel)

Further details of sizing of SMR reactor are to be done with Furnace unit counterpart since these reformer tubes are housed in the furnace itself. These would then be presented in the Final Team Report.

Production of Hydrogen via Syngas Route

2-21

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.6 ECONOMICS & SAFETY CONSIDERATIONS 2.6.1 Economic Analysis (Brief)

Table 2.6.1a: Costing Analysis for SMR unit [R29,30,31,32] Due to lack of credible literature data to support the costing analysis, an extremely rough estimate is given above in Table 2.6.1a. Computation efforts are done to illustrate how costing analysis can be done if there is access to proprietary pricing information while working as a real engineer. Nonetheless, further research effort would be done and cost estimations (with other considerations) would be put forth in the Final Team Report. 2.6.2 Safety Consideration for Reactor Design

Reactor is the heart of the plant design. Given the high speed steam and natural gas to be fed into the reactor, and the huge amount of heat is needed to supply to this endothermic reforming process, a great deal of safety consideration has to be in place to ensure that the plant and its operators can operate safety and efficiently.

In the Final Team Report where Process & Instrumentation analysis is done, more findings on safety considerations would be reported, with collaborative efforts with the furnace counterpart.

Production of Hydrogen via Syngas Route

2-22

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.7 LEARNING & CONCLUSIONS

In this report, the tubular reformer used for to produce hydrogen via a syngas route is designed. Both preliminary and detailed design achieved similar conversions of the major reactant component, namely Methane (CH4). Also, the detailed design is done via resolving Ordinary Differential Equations in MATLAB by obtaining information from literature research and making key assumptions with appropriate justifications. Besides, integration into the team’s overall process flow diagram has also been done by inserting values from the MATLAB model (which takes into account of both intrinsic kinetics and diffusional limitations) into a Plug-Flow-Reactor module in the HYSYS Simulation Environment. Results are considerably satisfactory since both values from MATLAB and HYSYS are in close agreement with each other, and aligned with that found in literature. Meanwhile, tube dimensions and material of constructions, brief economic analysis and safety considerations have also been covered in this report

Several learning can be derived from the current work of design a Steam Methane Reforming (SMR) reactor unit. Besides the need to plough through several literature data, the author has learnt to exercise discretion when researching through the available information, via perform the Principles of Chemical Engineering taught earlier.

Production of Hydrogen via Syngas Route

2-23

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.8 NOTATIONS Sym.

Description

Units

Ki Kj

Equilibrium rate constant for i rxn Equilibrium adsorption constant for specie ‘j’

T P ki

Temperature Pressure Rate constant for ith rxn

yj

Mole Fraction for specie ‘j’ Conversion for specie ‘j’

K1 & K3: (kPa)2; K2: unitless KCH4,H2,CO: (kPa)-1; KH2O: unitless Kelvin (K) kPa k1,3: (kmol•kPa0.5)/(kg•h) k2: (kmol•kPa-1)/(kg•h) molj / moltotal j = CH4, CO2

Molar feed ratio of H2O:CH4 Molar feed ratio of H2(recycle stream):CH4 Molar Feed ratio of CO2:CH4 Molar Feed ratio of N2:CH4 Rate of ith rxn at catalyst surface Rxn rate for specie ‘j’ at catalyst surface Effectiveness factor: ith rxn Effectiveness factor: Conversion for jth specie

unitless unitless unitless unitless kmol/(h•kgcat) j = CH4, CO2; kmol/(h•kgcat) unitless j = CH4, CO2; unitless

Dp

Mean Specific Heat Capacity of ith rxn Overall Specific Heat Capacity of ith rxn Axial position in reformer tube Inner diameter of reformer tube Outer diameter of reformer tube Catalyst bed void fraction Mass velocity of process gas (from HYSYS) Mass velocity of process gas (from HYSYS) Sphericity of catalyst pellet Equivalent length for catalyst pellet

kJ/(kmol•K) kJ/(kmol•K) m m m 0.605 [R1] kg/(h•m2) kg/(s•m2) 0.6563 [R1] 0.0174131m [R1]

ρb ρg

Bulk density of catalyst Density of process gas at any axial position

1362.0 kg/m3 [R1] kg/m3 [R1]

F R HF

Reformer feed rate Sum of all molar ratios in feed Heat Flux (= Heat Transfer Coefficient × Temp. Diff b/w Tinner tube wall & Touter tube wall)

kmol/h unitless kcal/(h•m2) Assume to be: 25000 Btu/h/ft2

χj sc hc dc nc ri rj

ηi ηj Cpmean,i Cpoverall t di do catbedvoid Gs Gmass

φs

th

th

or 283913.167 kJ/h/m2

- ∆H i ∆ H R ,i

Heat of the i reaction Heat of ith reaction

kcal/kmol kcal/kmol

υi

Stoichiometric coefficient for ith reaction

‘-’  reactants; ‘+’  products

Production of Hydrogen via Syngas Route

2-24

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.9 FIGURES AND TABLES

•

Fig 2.2.1a Flowchart to illustrate design methodology

•

Fig 2.2.3a: Typical Natural Gas & Reformer Catalysts

•

Fig 2.2.3b: Tubular reformer configurations

•

Table 2.2.4a: Recommended Property Package based on type of system

•

Fig 2.3.1a: Literature data to support conversion obtained during preliminary design is valid at the assumed conditions (2700kPa & 1123.15K)

•

Table 2.3.1a: Reaction Equilibrium Rate constants, from Hou & Hughes (2001)

•

Fig 2.4.4.1a: Conversion Profiles from MATLAB

•

Fig 2.4.4.2a: Pressure and Temperature profiles along length of reformer tube

•

Fig 2.4.4.3a: Component Mole Fractions from MATLAB Fig 4.4.3b: Component Mole Fractions from HYSYS

•

Fig 2.4.4.3c: Graph from literature on mole fraction profiles

•

Fig 2.4.5a: Effect of varying tubular inner diameter on CH4 conversion profile

•

Fig 2.4.5b: Effect of varying pressure on CH4 conversion profile

•

Table 2.4.5a: Effect of Manipulating Parameters on Critical Profiles (from MATLAB)

•

Table 2.4.6a: Stream Properties & Operating Conditions (HYSYS PFR & MATLAB)

•

Fig 2.5.2a: Relative tensile strengths of different stainless steel grades

•

Fig 2.5.3a: Estimation of furnace tube life for a given set of operating temp. & pressure

•

Fig 2.5.3b: Minimum Stress to rupture for chosen material

• Table 2.6.1a: Costing Analysis for SMR unit 2.10 ACKNOWLEDGEMENTS

This section dedicates acknowledgements to all who have helped the author by offering their valuable insights and advices. In particular, the author would like to express gratitude to Prof. Kawi for his advice, as well as to Mr Thanneer for his consultation on the MATLAB codes and functions. Last but not least, this work would not have been possibly done without the coordination and teamwork in Team 32 (CN4120 – Sem 2 of AY2007/08), hence the author would like to thank all of them for their assistance and understanding.

Production of Hydrogen via Syngas Route

2-25

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.11 REFERENCES

[R1]: J.K. Rajesh, Santosh K.Gupta, G.P.Rangaiah & Ajay K. Ray. (2000). Multiobjective Optimization of Steam Reformer Performance Using Genetic Algorithm. Ind. Eng. Chen.

Res.: Vol 39 – P706-717. [R2]: S.S.E.H. Elnashaie & S.S. Elshishini. (1993). Modelling, Simulation And Optimization Of Industrial Fixed Bed Catalytic Reactors. Gordon And Breach Science Publishers. [R3]: Jianguo Xu & Gilbert F. Froment (1989). Methane Steam Reforming, Methanation and Water-Gas Shift: I. Instrinsic Kinetics. AIChE Journal: Vol 35 – No.1 [R4]: Jianguo Xu & Gilbert F. Froment (1989). Methane Steam Reforming: II. Diffusional Limitations and Reactor Simulation. AIChE Journal: Vol 35 – No.1 [R5]: Kaihu Hou & Ronald Hughes. (2001). The kinetics of methane steam reforming over a Ni/α-Al2O catalyst. Chemical Engineering Journal: Vol 82 – P311-328. [R6]: Kaihu Hou & Ronald Hughes. (2001). The kinetics of methane steam reforming over a Ni/α-Al2O catalyst. Chemical Engineering Journal.: Vol 82 – P311-328. [R7]: Robert P. Hesketh. (2003). Catalytic Rates & Pressure Drops in PFR Reactors: HYSYS 3.0. [R8]: Wilhelm, D., Simbeck, D., Karp, A., Dickenson, R. (2001). Syngas production for gasto-liquids applications: technologies, issues and outlook. Fuel Proc. Tech., Vol 71 – P139 [R9]: Moulijn, J., Makkee, M., van Diepen, A. (2001). Chemical Process Technology. John

Wiley & Sons Ltd (England). [R10]: J.A. Moulijn, A.E. van Diepen & F. Kapteijn. (2001). Catalyst deactivation: is it predictable? What to do? Applied Catalysis A: General 212 – P3-16. [R11]: Chang Samuel Hsu & Paul R. Robinson. (2006). Practical Advances in Petroleum Processing. Springer Science+Business Media, Inc. [R12]: J. K. Rajesh, S. K. Gupta, G. P. Rangaiah & A. K. Ray. (2001). Multi-objective optimization of industrial hydrogen plants. Chemical Engineering Science: Vol56–P999-1010. [R13]: J.M. Smith, H.C. Van Ness & M. M. Abbott. (2005). Introduction to Chemical Engineering Thermodynamics – 7th Edition. McGraw-Hill International Edition. [R14]: Dilton, C.P. (1992). Materials selection for the chemical process industries. McGraw-

Hill. [R15]: Retrieved on 16th March 2008 from World Wide Web: http://www.valve-world.net/pdf/11022.pdf.

Production of Hydrogen via Syngas Route

2-26

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

[R16]: V. Ganapathy. (1982). Applied heat transfer. PennWell. [R17]: Retrieved on 10th March 2008 from World Wide Web: http://www.fitness4service.com/publications/pdf_downloads/Jaske-Shannon%20Paper.PDF [R18]: Retrieved on 18th March 2008 from World Wide Web: http://www.kubotametal.com/alloys/heat_resistant/HK-40.pdf [R19]: Retrieved on 18th March 2008 from World Wide Web: http://www.tu-harburg.de/vt2/pe2000/Dokumentation/PE2000_Kap7A1.htm [R20]: Chorng H. Twu, John E. Coon & David Bluck. (1997). A Comparison of the PengRobinson and Soave-Redlich-Kwong. Equations of State Using a New Zero-Pressure-Based Mixing Rule for the Prediction of High Pressure and High Temperature Phase Equilibria. Simulation Sciences Inc. [R21]: Retrieved on 17th March 2008 from World Wide Web: http://che.sut.ac.ir/People%5CCourses%5C65%5CCHEM_2_3.PDF [R22]: Retrieved on 10th March 2008 from World Wide Web: http://encyclopedia.airliquide.com/Encyclopedia.asp?GasID=41 [R23]: Kelly Ibsen. (2006). Equipment Design and Cost Estimation for Small Modular Biomass Systems, Synthesis Gas Cleanup, and Oxygen Separation Equipment. Nexant Inc. [R24]: P. van Beurden. (2004). On the Catalytic Aspects Of Steam-Methane Reforming. [R25]: J.R. Rostrup-Nielsen. (1993). Production of synthesis gas. Catalysis Today: Vol 18. [R26]: Ib Dybkjaer. (1995). Tubular reforming and authothermal reforming of natural gas – an overview of available processes. Fuel Processing Technology: Vol 42 – P85-101. [R27]: H.I.deLasa, G.Dogu & A.Ravella. (1991). Chemical Reactor Technology for Environmentally Safe Reactors and Products. Applied Sciences: NATO ASI Series Vol. 225 [R28]: J.R.Rostrup-Nielsenn, L.J.Christiansen & J.H.Bak Hansen. (1988). Activity of Steam Reforming Catalysts: Role and Assessment. Applied Catalysis: Vol43–P287-303. [R29]: Price of Nickel and Magnesium. Retrieved from World Wide Web on 19th March 2008: http://www.sciencelab.com/page/S/PVAR/10-807 [R30]: Price of HK-40 alloy (approximate): Retrieved from World Wide Web on 19th March 2008: http://www.meps.co.uk/Stainless%20Prices.htm [R31]: Price of Al2O4: Retrieved from World Wide Web on 19th March 2008: http://www.encyclopedia.com/doc/1G1-104622322.html [R32]: Density of HK40 alloy: Retrieved from World Wide Web on 19th March 2008: http://sg.search.yahoo.com/search?p=density+of+HK40+alloy&fr=yfp-tweb&toggle=1&cop=&ei=UTF-8 Production of Hydrogen via Syngas Route

2-27

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.12 APPENDIX 2.12.1 MATLAB Code 2.12.1.1 Main m-file to resolve O.D.E.s

clear all clc close all format long global catbedvoid phis Dp tube_length rhob rhos Pi Ti num_tubes_total di di_inch do global Ac sc hc dc nc R %SMR Inlet Conditions================================================= %kPa; Inlet Pressure Pi = 2700; Ti = 812.5; %K; Inlet Temperature; r/f [R1], this is b/w 725K and 900K %Tubing Dimensions & Number=========================================== num_tubes_total = 450; %TOTAL number of tubes di_inch = 5; %inch; Specify di in inches di = di_inch*0.0254; %m; Inner tube diameter Ac = pi*(di^2)/4; %m^2; Tube Cross-Sectional Area tube_length = 11.95; %m; Length of tube %Catalyst and Bed properties============================================= Dp = 0.0174131; %m; Pellet equivalent diameter catbedvoid = 0.605; %unitless; Catalyst bed void fraction rhob = 1362.0; %kg/m^3; Catalyst bed density rhos = 2355.2; %kg/m^3; Solid catalyst density phis = 0.6563; %unitless; Pellet sphericity %Molar Feed Compositions & Ratios======================================= sc = 3; % steam/CH4 molar feed ratio ==> FIXED hc = 0.0001; % H2/CH4 molar feed ratio; H2 from PSA RECYCLE dc = 0.00716496; % mol.CO2 / mol.CH4; from HYSYS PFR nc = 0.004094094; % mol.N2 / mol.CH4; from HYSYS PFR %Specify conditions and solve for the 4 ODES================================= tspan = [0 tube_length]; [t,y] = ode15s('smrodes',tspan,[0,0,Pi,Ti]); %Simulation results==================================================== figure subplot(2,2,1) hold on plot(t,y(:,1)) xlabel('Length of Reactor, m'); ylabel('CH4 conversion, xCH4'); subplot(2,2,2) hold on plot(t,y(:,2)) xlabel('Length of Reactor, m'); ylabel('CO2 conversion, xCO2');

Production of Hydrogen via Syngas Route

2-28

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

subplot(2,2,3) hold on plot(t,y(:,3)) xlabel('Length of Reactor, m'); ylabel('Pressure, kPA'); subplot(2,2,4) hold on plot(t,y(:,4)) xlabel('Length of Reactor, m'); ylabel('Temperature, K'); 2.12.1.2 Function m-file to define reactions conditions and O.D.E.s

function dy = smrodes(t,y) % Only declare those used as CONSTANTS to be "global" global R sc hc dc nc num_tubes_total Ac HF HF_btuperft2perhr do di di_inch rhob catbedvoid phis Dp global Fo FN2o FN2 G global MWCH4 MWH2O MWH2 MWCO MWCO2 MWN2 MWC2H6 FN2 FN2o global deltaA1 deltaB1 deltaC1 deltaD1 global deltaA2 deltaB2 deltaC2 deltaD2 global deltaA3 deltaB3 deltaC3 deltaD3 global H10 H20 H30 %These computations are done according to [R1], if there exists deviations, %these would be typically be mentioned as comments. %This m-file computes the 4 ODES to be resolved, whose solutions are then %input in the matrix y, so as to resolve them al xCH4 = y(1); %CH4 molar conversion at any axial position xCO2 = y(2); %CO2 molar conversion at any axial position P = y(3); %Pressure at any axial position T = y(4); %Temperature at any axial position %Computation for R=================================================== R = 1 + sc + hc + dc + nc; %sum of molar feed ratios %Adsorption constants for Individual Species============================== KCH4 = (6.65*10^(-6)).*exp(4604.28./T); %kPa^-1 KH2O = (1.77*10^(3)).*exp(-10666.35./T); %unitless; r/f [R1] & [R5] KH2 = (6.12*10^(-11)).*exp(9971.13./T); %kPa^-1 KCO = (8.23*10^(-7)).*exp(8497.71./T); %kPa^-1 %Equilibrium constants for Rxn I, II & III================================ K1 = 10266.76.*exp(-26830./T + 30.11); %kPa^2 K2 = exp(4400.0./T - 4.063); %unitless K3 = K1.*K2; %kPa^2 %Rate Coefficients for Rxn I, II & III==================================== k1 = 9.490*10^16.*exp(-28879./T); %kmol.kPa^0.5/kg.h k2 = 4.390*10^4.*exp(-8074.3./T); %kmol.kPa^-1/kg.h k3 = 2.290*10^16.*exp(-29336.0./T); %kmol.kPa^0.5/kg.h

Production of Hydrogen via Syngas Route

2-29

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

%Effectiveness Factors for Rxn I, II & III================================ n1 = (-7*10^-7).*t.^6 + (3*10^-5).*t.^5 - 0.0004.*t.^4 + 0.0029.*t.^3 - 0.0102.*t.^2 + 0.015.*t + 0.0165; %excel ==> deg 6 if (t <= 3.4) n2 = -0.0059.*t.^5 + 0.0559.*t.^4 - 0.1971.*t.^3 + 0.3263.*t.^2 - 0.2316.*t + 0.0889; %excel ==> deg 5 elseif (t == 3.4) n2 = 0; else n2 = (-7*10^-6).*t.^4 + 0.0004.*t.^3 - 0.0074.*t.^2 + 0.0651.*t - 0.2158; %excel ==> deg 6 end n3 = (-6*10^-7).*t.^6 + (3*10^-5).*t.^5 - 0.0004.*t.^4 + 0.0033.*t.^3 - 0.0132.*t.^2 + 0.0229.*t + 0.0102; %excel ==> deg 6 %GASEOUS Mole Fraction Basis for ALL SPECIES============================ yCH4 = (1-xCH4)./(R+2.*xCH4); yH2O = (sc-xCH4-xCO2)./(R + 2.*xCH4); yCO = (xCH4 - xCO2)./(R+2.*xCH4); yCO2 = (dc + xCO2)./(R + 2.*xCH4); yH2 = (hc + 3.*xCH4 + xCO2)./(R+2.*xCH4); yN2 = nc./(R+2.*xCH4); %Molecular Weights for ALL SPECIES===================================== MWCH4 = 16.043; %kg/kmol; from [R2] MWH2O = 18.01524; %kg/kmol; from [R2] MWH2 = 2.016; %kg/kmol; from [R2] MWCO = 28.01; %kg/kmol; from [R2] MWCO2 = 44.01; %kg/kmol; from [R2] MWN2 = 28.0134; %kg/kmol; from [R2] %Flowrates (Mass & Molar); Density of Process Gas========================= Fo = 1.040E4/num_tubes_total; % kmol/h %Reformer Molar Feed Flow Rate at inlet for ONE TUBE; from HYSYS FN2o = 10.4; % kmol(N2)/h %yN2o * Fo = FN2o (for ONE TUBE) where yN2o is initial N2 mole fraction; from HYSYS FN2 = FN2o; % kmol(N2)/h %FN2o is the N2 molar flow rate at inlet = FN2 is the N2 molar flow rate at any axial length % kmol/h F = FN2./(yN2*num_tubes_total); %Total molar flow rate at any axial length for ONE TUBE sv = ((F.*8.314.*T)./P)/(Ac); % m/h %Superficial Velocity = Volumetric Flow Rate (ASSUME Ideal Gas) / Ac G = 183287.978554998/num_tubes_total; % kg/h %Total mass flow rate for ONE TUBE; from HYSYS rhog = G./((F.*8.314.*T)./P); % kg/m^3; %Density of gas mixture = Mass flow rate per tube / Volumetric Flow Rate per tube Gmass = rhog.*sv; % kg/h/m^2; Mass velocity in per HOUR basis, to be used for dy(4) Gs = Gmass/3600; % kg/s/m^2; mass velocity in per SECOND basis, to be used for dy(3)

Production of Hydrogen via Syngas Route

2-30

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

%Rate of Rxns at catalyst surface at any axial position kmol/h/kgcat================== E = 1 + P.*(KCO.*yCO + KCH4.*yCH4 + KH2.*yH2) + KH2O.*yH2O./yH2; r1 = (k1./(E.^2.*yH2.^2.5.*P.^0.5)).*(yCH4.*yH2O-(P.^2.*yH2.^3.*yCO./K1)); r2 = ((k2.*P)./(E.^2.*yH2)).*((yCO.*yH2O)-((yH2.*yCO2)./K2)); r3 = (k3./(E.^2.*yH2.^3.5.*P.^0.5)).*(yCH4.*yH2O.^2(((yH2.^4.*yCO2).*(P.^2))./(K1.*K2))); rCH4 = r1+r3; rCO2 = r2+r3; nch4 = (n1.*r1 + n3.*r3)./(r1+r3); %effectiveness factor for adsorption of CH4 nco2 = (n2.*r2 + n3.*r3)./(r2+r3); %effectiveness factor for adsorption of CO2 %Specific Heat Capacities=============================================== %Compute constants to find Cp(mean) for Rxn I, II & III [R13] %Recall Rxn1: CH4 + H2O = CO + 3H2 deltaA1 = (-1)*(1.702)+(-1)*(3.470)+(1)*(3.376)+(3)*(3.249); deltaB1 = (-1)*(9.081/10^3)+(-1)*(1.450/10^3)+(1)*(0.557/10^3)+(3)*(0.422/10^3); deltaC1 = (-1)*(-2.164/10^6) +(-1)*(0)+(1)*(0)+(3)*(0); deltaD1 = (-1)*(0)+(-1)*(0.121/10^-5)+(1)*(-0.031/10^-5)+(3)*(0.083/10^-5); %Recall Rxn2: CO + H2O = CO2 + H2 deltaA2 = (-1)*(3.376)+(-1)*(3.470)+(1)*(5.457)+(1)*(3.249); deltaB2 = (-1)*(0.557/10^3)+(-1)*(1.450/10^3)+(1)*(1.045/10^3)+(1)*(0.422/10^3); deltaC2 = (-1)*(0)+(-1)*(0)+(1)*(0)+(1)*(0); deltaD2 = (-1)*(-0.031/10^-5)+(-1)*(0.121/10^-5) +(1)*(-1.157/10^-5) +(1)*(0.083/10^-5); %Recall Rxn3: CH4 + 2H2O = CO2 + 4H2 deltaA3 = (-1)*(1.702)+(-2)*(3.470)+(1)*(5.457)+(4)*(3.249); deltaB3 = (-1)*(9.081/10^3)+(-2)*(1.450/10^3)+(1)*(1.045/10^3)+(4)*(0.422/10^3); deltaC3 = (-1)*(-2.164/10^6) +(-2)*(0)+(1)*(0)+(4)*(0); deltaD3 = (-1)*(0)+(-2)*(0.121/10^-5) +(1)*(-1.157/10^-5) +(4)*(0.083/10^-5); %Note that 1.987 is multiplied to convert kJ/mol.K to kcal/kmol.K %Also, note that 298.15 K is reference temperature %Also, T/298.15 is tile in [R13] Pg. 141 Eqn (4-20) Cpmean1 = 1.987*(deltaA1 + (deltaB1/2)*(298.15).*(T/298.15+1) + (deltaC1/3).*((298.15)^2).*((T/298.15).^2 + (T/298.15) + 1) + deltaD1./((T/298.15).*298.15^2)); Cpmean2 = 1.987*(deltaA2 + (deltaB2/2)*(298.15).*(T/298.15+1) + (deltaC2/3).*((298.15)^2).*((T/298.15).^2 + (T/298.15) + 1) + deltaD2./((T/298.15).*298.15^2)); Cpmean3 = 1.987*(deltaA3 + (deltaB3/2)*(298.15).*(T/298.15+1) + (deltaC3/3).*((298.15)^2).*((T/298.15).^2 + (T/298.15) + 1) + deltaD3./((T/298.15).*298.15^2)); %Note that divide by 4.196 such that kJ/kmol ==> kcal/kmol %ENDOTHERMIC; [R1] H10 = (2.061*10^5) /(4.186); H20 = (-4.11*10^4) /(4.186); %EXOTHERMIC; [R1] H30 = (1.650*10^5) /(4.186); %ENDOTHERMIC; [R1] %Heats of Reactions=================================================== H1 = H10 + Cpmean1.*(T-298.15); %kcal/kmol H2 = H20 + Cpmean2.*(T-298.15); %kcal/kmol if (n2<0) H2 = -H2; %kcal/kmol else H2 = H2; %kcal/kmol end

Production of Hydrogen via Syngas Route

2-31

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

H3 = H30 + Cpmean3.*(T-298.15); %kcal/kmol sum1 = -H1.*n1.*r1; %for dy(4) sum2 = -H2.*n2.*r2; %for dy(4) %for dy(4) sum3 = -H3.*n3.*r3; %Heat Flux (CONSTAN VALUE IS ASSUMED HERE; r/f to Chapter 2.4.2 for explanation) %btu/h/ft^2; from [R11] HF_btuperft2perhr = 25000; HF = HF_btuperft2perhr*2.71427502; %kcal/h/m^2; %Weight Fraction for ALL SPECIES======================================= %From ABOVE ==> G is Total mass flow rate for ONE TUBE; from HYSYS wCH4 = (yCH4.*F)*MWCH4./G; wH2O = (yH2O.*F)*MWH2O./G; wH2 = (yH2.*F)*MWH2./G; wCO = (yCO.*F)*MWCO./G; wCO2 = (yCO2.*F)*MWCO2./G; wN2 = (yN2.*F)*MWN2./G; %Specific Heat Capacities for ALL SPECIES================================= CpCH4 = 1.987*(1.702 + (9.081/10^3).*T + (-2.164/10^6).*T.^2 + (0).*T.^%kcal/kg.K 2)/MWCH4; CpH2O = 1.987*(3.470 + (1.45/10^3).*T + (0/10^6).*(T.^2) + (0.121/10^5).*(T.^-2))/MWH2O; %kcal/kg.K CpH2 = 1.987*(3.249 + (0.422/10^3).*T + (0/10^6).*(T.^2) + (0.083/10^-5).*(T.^2))/MWH2; %kcal/kg.K CpCO = 1.987*(3.376 + (0.557/10^3).*T + (0/10^6).*(T.^2) + (-0.031/10^5).*(T.^-2))/MWCO; %kcal/kg.K CpCO2 = 1.987*(5.457 + (1.045/10^3).*T + (0/10^6).*(T.^2) + (-1.157/10^5).*(T.^-2))/MWCO2; %kcal/kg.K CpN2 = 1.987*(3.280 + (0.593/10^3).*T + (0/10^6).*(T.^2) + (0.040/10^-5).*(T.^2))/MWN2; %kcal/kg.K Cpoverall = wCH4.*CpCH4 + wH2O.*CpH2O + wH2.*CpH2 + wCO.*CpCO + wCO2.*CpCO2 + wN2.*CpN2; %kcal/kg.K %Ordinary Differential Equations========================================== dy(1) = Ac.*R.*rhob.*nch4.*rCH4./Fo; dy(2) = Ac.*R.*rhob.*nco2.*rCO2./Fo; dy(3) = -(1.75*(Gs^2)*(1-catbedvoid))./(phis*Dp*((catbedvoid)^3).*rhog)/1000; %/1000 is to account to change Pa to kPa so as to use P for other functions dy(4) = (1./(Gmass.*Cpoverall)).*((4*HF)/di + rhob.*(sum1+sum2+sum3)); % note that units for Cpoverall is kcal/kg.K dy = dy'; %Note: This is skeletal MATLAB developed to solve for the 4 O.D.E.s. Additional strings of code used to generate the various plots shown in Chapter 2.4.4, 2.4.5 and 2.4.6 are not shown here due to space constraint. In general, just need to comment off the ‘clear all’ command, and then vary the parameter(s) (e.g. num_tubes_total), and record the figures in a new matrix after each run. Thus, graphs with different num_tubes_total can then be plotted on one plot.

Production of Hydrogen via Syngas Route

2-32

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.12.2 List of Equations 2.12.2.1 Rate Equations for reactions and species for 4 O.D.E.s

r1 =

r2 =

r3 =

3

k1 2

2.5

E yH2 P

( y CH 4 y H 2O −

0.5

P 2 y H 2 y CO K1

kmol h • kg cat

);

Eqn (12-11)

y H y CO2 k2 kmol ( y CO y H 2O − 2 ); K2 h • kg cat E yH2

Eqn (12-12)

2

4

k3 3.5

2

E yH2 P

2

( y CH 4 y H 2O −

0.5

rCH 4 = r1 + r 3

y H 2 y CO2 P 2 K3

Eqn (2-14)

kmol h • kg cat

);

Eqn (12-13)

rCO2 = r2 + r 3

R = 1 + sc + hc + dc + nc (used in Eqn (4-1) & (4-2)) E = 1 + P( K CO yCO + K CH 4 yCH 4 + K H 2 yH 2 ) + K H 2 O

Eqn (2-15) Eqn (2-16)

y H 2O

Eqn (2-17)

yH2

2.12.2.2 Mole Fractions for species

yCH 4 =

yCO =

1 − χ CH 4 R + 2 χ CH 4

χ CH − χ CO R + 2 χ CH 4

2

Eqn (2-18)

y H 2O =

Eqn (2-20)

yCO2 =

Eqn (2-22)

y N2 =

4

yH2 =

hc + 3χ CH 4 + χ CO2 R + 2 χ CH 4

sc − χ CH 4 − χ CO2 R + 2 χ CH 4 dc + χ CO2 R + 2 χ CH 4

nc R + 2 χ CH 4

Eqn (2-19)

Eqn (2-21)

Eqn (2-23)

2.12.2.3 Effectiveness Factors for reactions and species

η1 , η 2 & η 3 are obtained via fitting polynomials using Microsoft Excel. Points are specified via identifications of coordinates for these 3 curves via vigorous read-off.

η1 = ( −7 × 10 −7 )t 6 + (3 × 10 −5 )t 5 − (0.0004)t 4 + (0.0029)t 3 − (0.0102)t 2 + 0.015t + 0.0165 Polynomial of degree 6

Eqn (2-24)

η 2 = (−0.0059)t 5 + (0.0559)t 4 − (0.1971)t 3 − (0.3263)t 2 + 0.2316t + 0.0889 Polynomial of degree 5 (for t < 3.4)

Production of Hydrogen via Syngas Route

Eqn (2-25)

2-33

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

η 2 = (−7 × 10 −6 )t 4 + (0.0004)t 3 − (0.0074)t 2 + 0.0651t − 0.0.2158 Polynomial of degree 4 (for t =3.4 and t > 3.4)

Eqn (2-26)

η 3 = ( −6 × 10 −7 )t 6 + (3 × 10 −5 )t 5 − (0.0004)t 4 + (0.0033)t 3 − (0.0132)t 2 + 0.0229t + 0.0102 Polynomial of degree 6

η CH = 4

η CO = 2

η1 r1 + η 3 r3 r1 + r3

η 2 r2 + η 3 r3 r2 + r3

Eqn (2-27) Eqn (2-28)

Eqn (2-29)

2.12.2.4 Rate & Adsorption constants for reactions 1, 2 and 3 0 .5  − 28879 .0  kmol • kPa k1 = 9.490 × 10 6 exp ; T kg • h  

Eqn (2-30)

−1  − 8074 .3  kmol • kPa k 2 = 4.390 × 10 4 exp ; T kg • h  

Eqn (2-31)

0 .5  − 29336 .0  kmol • kPa k 3 = 2.290 × 10 16 exp ; T kg • h  

Eqn (2-32)

 − (26830.0)  K 1 = 10266.76 exp + 30.11; kPa 2 T  

Eqn (2-33)

 − (−4400.0)  K 2 = exp − 4.063 ; unitless T  

Eqn (2-34)

K 3 = K 1 × K 2 ; kPa 2

Eqn (2-35)

2.12.2.5 Adsorption constants for species

 − ( −4604.28)  −1 K CH 4 = 6.65 × 10 −6 exp ; kPa T  

Eqn (2-36)

 − (10666.35)  K H 2O = 1.77 × 10 3 exp ; unitless T  

Eqn (2-37)

 − ( −9971.13)  −1 K H 2 = 6.12 × 10 −11 exp ; kPa T  

Eqn (2-38)

 − ( −8497 .71)  −1 K CO = 8.23 × 10 −7 exp ; kPa T  

Eqn (2-39)

Production of Hydrogen via Syngas Route

2-34

CN 4120: Design II Team 32: Lim Yueh Yang (U046787U)

SMR Unit Design Report

2.12.2.6 Heat Capacities

∆H i = ∆H R ,i + Cp mean ,i (T − TR )

Cpmean , i

Eqn (2-40)

  2 2    (υ B )T  T  (υ C )(T )  T  T (υ D )  = R (υi Ai ) + i i 0  − 1 + i i 0    + + 1 + i i    T0  T0  T 2  T0 3   (T0 )2      T0

Values of A, B, C and C for the respective species for the ith reaction are found in [R13]. Eqn (2-40)

2.12.3 Sample Calculations

Most of the calculations are performed via MATLAB and the HYSYS Simulation Environment, so long as the relevant parameters are specified. Hence, sample calculations would not be shown in this work. All the MATLAB written have comment statements intended to make the code self-explanatory.

2.12.4 Typical Natural Gas Compositions

Figure is retrieved from Midrex from World Wide Web on 19th March 2008 at: http://www.midrex.com/uploads/documents/Catalyst(1)1.pdf.

Production of Hydrogen via Syngas Route

2-35

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Chapter 3 : FURNACE 3.1 INTRODUCTION 3.1.1 Furnace design methodology

Furnaces serve various purposes in process industries such as column reboilers, reactor-feed preheaters etc. Unlike typical furnaces; the furnace used for the steam methane reforming has additional design considerations compared to conventional furnaces. The furnace in this project is used to provide energy for the steam methane reaction. It provides single-phase/multiple-component heating. In addition, the convection section of the furnace serves to extract excess heat from the flue gas to heat up process streams from other part of the plant. Typically, 70% of the heat generated by the burner goes to the radiation section while the remaining 30% goes to the convection. Single phase multiple components heating will be carried out by the furnace. Catalysts would be placed in the reactor tubes lining the refractory. The furnace design would incorporate the following considerations: (1) Capacity and size of furnace

(2) Dimensions of reactor tubes

(3) Material selection

(4) Safety considerations

Fig 3.1.1a: Furnace design methodology

Production of Hydrogen via Syngas Route

3-1

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.1.2 Heat transfer process in fired heater

There are two main heat transfer process – radiation and convection. Radiation occurs within the radiation zone where fire from the burners heats up the tubes containing the process fluid, in this case the reactants and products of the steam methane reform reaction. In the convection zone, heat transfer is a combination of non-luminous and convective heat transfer. The flue gas is the main medium for convective heat transfer to take place.

3.2 RADIATION ZONE DESIGN

Heat transfer to the radiant zone is the most important aspect of design for a fired heater. An acceptable heat flux and metal tube temperature has to be achieved during design4.

3.2.1 Thermal Efficiency of Fired Heater

Heater efficiency is essential for determining the energy to be supplied through the combustion process in the fired heater. It is the ratio of the amount of heat transferred to the tubes to the amount of heat generated through combustion in the fired heater. The heater’s efficiency is dependent on the following factors: •

Flue-gas stack temperature

•

Excess air or oxygen

•

Heat lost to the surrounding

•

Design of the convection section in the fired heater

The flue gas stack temperature can be computed using the approach temperature, which is the difference in the stack temperature to the inlet fuel temperature. Typical approach temperature varies between 100-150°F1. Through HYSYS simulation, the stack gas temperature is 565.6°C. The percentage heat available (thermal efficiency) can be derived from the graph as shown below. Typically, heat efficiency can also be computed from the following equation: Heater Efficiency =

Heat available at flue gas temperature Lower Heating Value of fuel gas

Production of Hydrogen via Syngas Route

3-2

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

To obtain the furnace efficiency, a theoretical flame temperature has to be found. However, a few assumptions have to be made to simplify calculations. •

Combustion of nitrogen is negligible

•

No carbon monoxide is formed

To calculate the heat released from combustion and the temperature of the products formed, the enthalpy change of the combustion process can be considered2.

Fig 3.2.1a: Thermodynamic flow of combustion reaction The total heat of combustion can be given as  Heat of combustion = ∆HR + ∆HP + ∆H0C Assuming adiabatic combustion, heat of combustion = 0  ∆HP = -∆HR - ∆H0C Composition of fuel gas from PSA outlet consists mainly of CH4 and H2 where number of moles of H2 is 3 times the number of moles of CH4. The other components will be ignored for furnace efficiency computations as they are present in small quantities. Hence, the main combustion reactions considered for calculations are (1) CH4 + 2O2 → CO2 + 2H2O

Production of Hydrogen via Syngas Route

(2) 2H2 + O2 → 2H2O

3-3

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Hence, Number of Moles of CH4= 1; H2 = 3; O2 = 3.5; CO2 = 1; H2O = 5; N2 = (0.79/0.21) x 3.5 = 13.17 To compute ∆HR, assume •

Temperature of air is preheated to 150°C and

•

Temperature of fuel feed from PSA unit is 40°C

•

Flame temperature is 1900°C

•

1 mol of CH4 present

•

Air comprise of 79% N2 and 21% O2

The Cp of the gases present at a flame temperature is obtained from literature. Using the table above and with excel spreadsheet, iteration is performed to obtain the flame temperature. A flame temperature is first assumed. No. of moles Reactant

Fuel

Air

Product

% excess air

Cp (KJ/mol-K)

CH4

1

68.05

H2

3

30.5

O2

3.5

1.15

34.25

N2

13.16666667

1.15

32.39

CO2

1

52.31

H2O

5

40.93

N2

13.16666667

1.15

32.39

O2

3.5

0.15

34.25

Table 2.1a: Excel spreadsheet used in calculation for flame temperature ∆HR = (68.05+3 x30.5) (50-25) + (34.25 x 3.5 x 1.15 + 13.17 x 1.15 x 32.39) (150-25) =

82525 KJ/mol ∆HP = (1 x 52.31 + 5 x 40.93 + 13.17 x 1.15 x 32.39 +3.5 x 0.15 x 34.25) x (flame T – 25) =

765.4 (flame T -25) ∆H0C can be computed as: ∆H0C = 802800 + 241800 = 1044600 KJ/mol

Hence,

765.4 (flame T -25) = 1044600 – 82525

Production of Hydrogen via Syngas Route



Flame T = 1389°C

3-4

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

The flame T is quite close to the original flame temperature calculated and hence flame T will be taken to be 1390°C. Fig 2.1: Flue gas profile of fired heater

Given the furnace profile: Furnace Effficiency = (Heat to process) / (Heat released by fuel) Assuming stack temperature is 150°C, Furnace efficiency = 765.4 (1390 -150) / (1044600 – 80930) = 0.9449 As it is usually not an adiabatic combustion process, heat is also lost to the surrounding through the refractory walls. The value of heat loss is usually 2%3. Hence, the overall thermal efficiency of the furnace is: 94.49 – 2 = 92.49% Calculating the amount of heat to be supplied by the furnace, based on the energy requirement specified by the SMR personnel,

Energy required =

100 × 5.97 × 10 8 = 6.45 x 108 KJ/h. 92.49

Assuming purged product from PSA contain a majority of methane gas for combustion, the LHV of the fuel feed to the furnace will be approximately 50MJ/kg (5 x 104 KJ/kg). Hence amount of fuel feed needed is =

Production of Hydrogen via Syngas Route

5.82 × 10 8 = 1.29 x 104 kg/h. 4 5 × 10

3-5

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Given simulation from HYSYS, amount of PSA outlet is approximately 1 x 105 kg/h. Since the CH4 content is not exactly 100%, despite the excess in quantity of fuel from PSA outlet to that required, the totally energy that can be supplied will be the same as that of a pure methane feed of lower quantity. Hence, the amount of fuel supplied from PSA outlet is sufficient for supporting furnace combustion. However, in case insufficient fuel is supplied from the PSA outlet due to equipment fault, a makeup fuel feed will be fed to the furnace. This will be done through control instrumentation design. The amount of makeup feed will then be 1.163 x 104 kg/h. Fig 3.2.1d: Suggested instrumentation control for fuel gas inlet control

3.2.2 Calculation for the number of reformer tubes

The number of reactor tubes within the furnace can be computed from an average heat flux. Typical heat flux value for reformer unit is3 25000 BTU/h-ft2. With the number of tubes computed, the mass velocity within the reformer tubes can then be computed. An excel spreadsheet was used to compute the number of tubes from the heat flux value. The value of heat flux is found in literature6 to be 25 000 BTU/hr-ft2.

Production of Hydrogen via Syngas Route

3-6

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Length of tube (ft)

39.2

Internal diameter of pipe (ft)

0.417

Surface area of tube (ft2)

51.3277

Total energy required in KJ/h as derived from HYSYS

5.97 x 108

Average flux in KJ/hr-ft2 as specified in literature

25000

No. of tubes

465

Table 3.2.2a: Computation for no. of tubes with heat flux The number of tubes was calculated using the following equation: Total energy required as derived ( HYSYS ) = Average heat flux (literature) π × Internal diameter of pipe × length of tube

Hence it can be concluded that the number of tubes to be used for the reactor is approximately 465 tubes. However, since the SMR personnel have obtained good conversion with 450 tubes, 450 tubes will be used for further design considerations.

3.2.3 Calculation for mass velocity in reformer tubes

Mass velocity ( per reformer tube) =

total mass flow of reac tan ts no. of tubes × cross sec tion per tube

Total SMR feed load (kg/h) Total SMR feed load (lb/s)

182400 111.4667

Internal diameter of pipe (inches)

5

Internal diameter of pipe (ft)

0.417

cross section of pipe in (ft2)

0.0137

mass flow velocity in lb/s (ft2) No. of tubes

18 450

Table 3.2.3a: Computation for no. of tubes from mass velocity

The mass velocity of the fluid in the tube can be found to be around 18 lb/s ft2. The minimum mass flow velocity required of 15 lb/s ft2 is satisfied.

Production of Hydrogen via Syngas Route

3-7

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Turndown consideration

It is important to consider turn-down and possible stream recycling. Usually, turndown of 60%16 is taken into account. 60% x 15 = 10.8 lb/s ft2. However, as the furnace supports a reaction process, recycling will not be considered.

3.2.4 Calculation of reformer tube thickness

HK40 (25% Cr, 20% Ni) are usually used for steam reformer tubes4. HK40 is chosen for its high creep and rupture strengths. It is also resistant to hot gas corrosion and hence is usually employed in steam methane reforming processes5. It provides creep resistance up to 980°C. This makes it suitable for the current design where the highest temperature 826.85°C6. A standard code formula is employed to calculate the minimum wall thickness required7.

min t wall

d  P i + FCA  2 = =  S a E − 0.6 P

0.1448 psi   0.127m 1in   × + 0.039in   2700kPa × × 1kPa   2 0.0254m   = 0.399in 1000 psi 0.1448 psi 3.2ksi × × 0.85 + 0.6 × 2700kPa × 1ksi 1kPa

P = Max Pressure (to be in psig) = 2700 kPa (highest pressure at inlet); di = Inner Diameter (to be in inches) = 0.127 m; FCA7 = 10-year corrosion allowance (to be in inches) = 0.039 in; Sa7 = Minimum Creep Stress for HK40 (to be in psi) = 3.2 ksi; E 6 = Weld Efficiency Factor = 0.85.

Production of Hydrogen via Syngas Route

3-8

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.2.5 Selection of material for reactor tube in radiation section

A series of procedure has been developed to aid the selection of material for reactors8. 1. Define the conditions of exposure (eg. Temperature and pressure) 2. Explore available materials 3. Identify the suitable material 4. Evaluate the material In the steam methane reforming process, the reactor will be exposed to high temperature of approximately 1130K and pressure of 2700kPa9. In addition, there are chances of hot gas corrosion due to the high mass velocity. In view of these conditions, stainless steel is a suitable material for the construction of the steam methane reformer tubes. With higher carbon content, stainless steel offers greater creep resistance than other metals. With the addition of nickel and chromium, resistance to carburization and creep is enhanced. There are different grades of heat resistance steel, namely, HH, HK, HD and HF. HK has been specifically found to be of great use in steam methane reforming due to their high creep and rupture strength even up to 1150°C. Most importantly, it offers resistance to hot gas corrosion. The following figure shows the superiority of HK40 metal compared to other grades. It shows the relative tensile strength of the different stainless steel grades. Fig 3.2.5a: Tensile strength

Production of Hydrogen via Syngas Route

3-9

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

In addition, the tube life can be determined as follows.

Fig 3.2.5b: Tube life of HK40 tube From the figure, at pressure of 2700kPa (391psi) and temperature of 1130K, the furnace tube life can be found to be more than 20.3 years. This provides a confirmation that the material is suitable for use in steam methane reforming processes. The follow figure shows the minimum stress to rupture for HK40 piping as compared to other grades. The figure implies a lower performance of HK40 grade compared to HP grades. However, since the maximum temperature is 1130K, which is low compared to the maximum temperature at which these stainless steel grades can withstand, the pressure factor is taken for higher consideration in selection of the suitable metal type. Since HK40 is capable of withstanding higher pressure, will be the final choice of material for the reformer tubes.

Fig 3.2.5c: Minimum stress Vs Temperature

Production of Hydrogen via Syngas Route

3-10

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.2.6 Reformer inner tube diameter

Tube sizes ranging from 4 to 8 inches are usually used based on the standard nominal pipe size. A 5 inch inner tube diameter has been chosen based on the tube thickness calculated (0.399 inches) and the standard nominal tube sizes1. Hence a Schedule 8016 tube constructed from HK40 stainless steel will be used for the reformer tubes.

3.2.7 Furnace layout and design

3.2.7.1 Side Fired Heater

A side fired heater with vertical tubes has been used for simulation of the SMR reaction. Hence, a side fired heater design will be proposed for the furnace type. Side fired furnace has a few advantages. It allows the adjustment and control of the tube wall temperature. The maximum temperature will be at the outlet of the reformer tube while the highest heat flux is at a relatively low temperature. The side fired furnace offers more flexibility in design and operation10. Side fired configuration also allows a countercurrent flow of flue gas and process fluid which yields a higher heater efficiency. A typical side fired heater has the following configuration as found in literature10.

However, given the large number of tubes, it is not economically feasible to line the tubes in two rows as a large amount of space will be needed. Hence, 4 rows of tubes will be proposed, each row comprising 450 /4 = 112 or 113 tubes.

Production of Hydrogen via Syngas Route

3-11

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.2.7.2 Distance between burners

The distance between the burners is kept at 4m. This is to ensure a safe distance between the tubes and the two burners. The 2-D sketch is shown below.

Fig 3.2.7.2a: Proposed side-fired heater design (radiation + convection zone) Given the tube dimensions as computed and that tube pitch is taken as twice the tube outer diameter, and taking the allowance from the refractory wall to be 1 m in total, Tube dimensions

Length (m)

11.95

I.D (m)

0.127

O.D (m)

0.147

Number of tubes per row

113

Tube pitch (where D = outer diameter)

Length of

2 x O.D

furnace

= No. of tubes per rows × Tube pitch × Outer diameter per tube + Allowance Hence length of furnace is approximately 33m.

Production of Hydrogen via Syngas Route

3-12

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

The pigtails will be approximated to be 0.5m each. Pigtails are used to allow expansion and contraction of tubes during start up. It minimises the need for joints and welding that may fail when there is too much stress. Inlet pigtails are silicon killed while outlet pigtail is made up of high alloy material. Height of the radiant section will be taken to be: 11.95m + length of pigtails = 13m. A 3-dimensional proposed design is as shown below.

Fig 3.2.7.2a: 3-D view radiation zone of proposed side-fired heater

3.2.7.3 Burners used at Side Walls

Premix burners will be used for the side wall. This is because they offer better linearity, where excess air remains more nearly constant at turndown. Air will be drawn in through the primary box register and mixed with the fuel before it flows to the furnace firebox. Good mixing has to be ensured so that a short non-yellow flame can be obtained. This is to prevent the flame from being in contact with the reformer tubes and cause locus increase of temperature on the reformer tube. Long flames cause tube failure in the long run and soot blower may be necessary to clean the heating surface. The figure below shows a typical premix burner11.

Production of Hydrogen via Syngas Route

3-13

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

Fig 3.2.7.3a: Pre-mix burner However, usually gaseous fuels provide non-luminous flames4. 3.2.7.4 Determination of number of burners

The length of the furnace box is given to be 33m. For maximum heat distribution, the centre to centre distance between burners should be 1m. Hence there would be approximately 32 burners along the length of the furnace. Since the height of the furnace is 13m, the number of burners along the height of the furnace is 12. The layout on the refractory wall is shown below. The total number of burners used will be 3072 burners.

Fig 2.7.4a: Side-fired heater burner arrangement

Production of Hydrogen via Syngas Route

3-14

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.2.8 Computations for flue gas temperature

The flue gas temperature as obtained from HYSYS is 565.6°C. However, the conversion reactor is used in the HYSYS design, which is a steady state module. Hence, more calculations should be done to clarify the flue gas temperature. These set of calculations take into account the dynamic state of the furnace.

3.2.8.1 Cold plane area

The cold plane area, which is the projected area of reformer tubes, is calculated as follows13:

Acp= exposed tube length x centre to centre spacing x number of tubes excluding shield tubes. = 11.95 x 2 (pitch) x 0.147 (outer diameter of pipe) x 450 = 1584 m2.

3.2.8.2 Refractory area

The refractory area is defined as the inside surface of the shell minus the cold plane area. The equation for computation of the refractory area is as follows: Aw = 2[W(H+L) + H x L)] = 2[16(13+33)+33 x 13] - Acp = 746 m2.

3.2.8.3 Absorptivity, α α = 1- [0.0277 + 0.927 (x -1)] (x-1) ; where x refers to the pitch. Since pitch is 2, α = 0.879.

3.2.8.4 Sum of product of area and the absorptivities in the radiant zone

The equation for calculation is shown below:

αAR = αAshield+ αAcp

Assuming the Ashield is negligible, then

αAR = αAshield + αAcp

αAR = αAcp

AR = Acp

Hence AR = 1584m2.

Production of Hydrogen via Syngas Route

3-15

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.2.8.5 Mean beam length

L = (2/3)(furnace volume)1/3 = 12.7m

3.2.8.6 Partial pressure of CO2 and H2O

The main combustion products are CO2 and H2O. P = 0.288 – 0.229x + 0.0090x2; where x is the fraction of excess air taken to be 0.15. Hence P = 0.256.

3.2.8.7 Product of partial pressure and mean beam length

PL = 0.256 x 12.7 = 3.24

3.2.8.8 Mean refractory tube wall temperature

Tt = 100 + 0.5 (T1 + T2) From the SMR personnel: T1 = 539.4°C and T2 = 862.85°C. Hence Tt = 783°C = 1384°F

3.2.8.9 Two main equations that will be used for iteration to find Tg (flue gas temp) 3.2.8.9.1 Radiant zone heat transfer

 Tg + 460  4  T + 460  4  QR  −  t = 1730    + 7(Tg − Tt ) αAR F 1000    1000  

3.2.8.9.2 Radiant zone heat balance

Qn  Qa Q f Q L Q g  QR 1 +  = + − − αAR F αAR F  Qn Qn Qn Qn  The unknowns in the equations also require approximation of Tg to be made.

Production of Hydrogen via Syngas Route

3-16

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Qf Qn

Furnace Unit Design Report

refers to the enthalpy of the fuel feed and is taken to be negligible as it is not preheated.

QR is related to Qn by the efficiency. Qn = QR/efficiency where efficiency = 94% as computed earlier Qa refers to the enthalpy of the preheated air and will be taken from HYSYS simulation. Qn

3.2.8.9.3 Enthalpy of the flue gas as a function of Tg (flue gas temp)

  T  T  = a + b − 0.1 − 0.1 Qn  1000 1000   

Qg

Z = fraction excess air a= 0.22048-0.35027*z+0.92344*(z)^2;

b=0.016086+0.29393*z-0.48139*(z^2)

3.2.8.9.4 Emissitivity of the gas Ф Ф = a + b(PL) + c(PL)2

where PL was calculated earlier on

Z = (Tg+460)/1000 a= 0.47916-0.1984*z+0.022569*(z^2);

b= 0.047029+0.0699*z-0.01528*(z^2)

c= -0.000803-0.00726*z+0.001597*(z^2)

3.2.8.9.5 Exchange factor F

F = a + b Ф + c Ф2

Z = Aw/αAR

a=0.00064+0.0591*z+0.00101*(z^2);

b=1.0256+0.4908*z-0.058*(z^2)

c=-0.144-0.552*z+0.04*(z^2)

Production of Hydrogen via Syngas Route

3-17

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

After performing iteration using goal-seek tool in excel spreadsheet, the flue gas temperature derived is 417°C. However the temperature as obtained from HYSYS is 565.6°C. However, since the heat exchanger network person-in-charge has decided to use 565.6°C for calculations in stream-matching, further computations and design for convection section will make use of this value.

Since the temperature is low, a check is carried out to ensure that the dew point of the flue gas is not reached. The graph below shows the dew point temperature of flue gas at different temperatures.13.

Fig 3.2.8.9.5a14: Dew point of flue gases versus fuel sulphur

Given that the excess air is 15%, and that there is 0wt% sulphur in fuel, the dew point if about 130°F, which is lower than the flue, gas temperature computed (410°C). Hence the flue gas temperature computed is reasonable.

3.2.9 Residence Time

Residence time = volume of each reformer tube / volumetric flow rate of reactant gas = π x (D/2)2 / (mass flow per tube/ density) = 1.286s

Production of Hydrogen via Syngas Route

3-18

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.3 CONVECTION SECTION

The convection section is used to preheat streams from other parts of the plant. This helps to maximise the amount of energy that can be extracted from the furnace. Shield tubes are omitted as convection tubes are not receiving direct heat from the flame from the proposed design of the fired heaters. As discussed with the Heat Exchanger Network personin-charge, three process streams will have to be heated and the heating scheme will be as shown below.

Steam will first be generated followed by heating up the SMR feed and finally preheating the air fuel feed.

3.3.1 Convection design – Finned tubes

In the design of finned tubes, the following equation will be used. Ac =

Qc ; U c (LMTCD )

where LMTD =

(Tg1 − TL1 ) − (Ts − TL 0 )

[

ln (Tg1 − TL1 ) /(Ts − TL 0 )

];

T L0 and TL1 = inlet and outlet temperature of process fluid (respectively) Tg1 and Ts = temperature of incoming and outgoing flue gas (respectively)

Production of Hydrogen via Syngas Route

3-19

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

From the Heat Exchanger Person-in-charge, the following information is derived.

Process streams

Heat duty KW

LMTD

Uc (W/m2K)

Steam

32090.7

87.5,360.7

10

Natural Gas

7307.5

61.4, 334.6

6.68

Air

4720.5

119.5,392.7

8.78

The total area is then calculated to be 8858.7 + 3268.6 + 1368.1 = 13495.4m2. From the HEN person-in-charge, the desired outer diameter is 0.01905m = 0.75 in. From the vendor of finned tubes (Vulcan Tubes), an appropriate fin tube is chosen. The fin dimensions17 are shown below:

Number of fins per inches

7

Fin thickness (in.)

0.06

Fin height (in.)

0.625 = 0.015m

Surface area (sq ft per linear foot)

3.39

Total length of tubes needed = 145251 / 3.39 = 44015ft = 13338m

Section

Total Length of each section (m)

Steam

8755

SMR feed

3126

Air

1352

Since the length of each tubes is very long, pressure drop will be high and hence the stream has to be split into different tubes to prevent high pressure drop. Section

Number of tubes

steam

1250

SMR feed

446.47

Air

193

Number of tubes required = 13338 / 7 = 1905 tubes

Production of Hydrogen via Syngas Route

3-20

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

To ensure a mass flow of flue gas to be 1.7kg/s-m2, the number of tubes is computed as: Flow rate of flue gas is 86.88kg/s; Cross section area of tubes = 86.88/1.7 = 51m2. Width of cross section = 51/7 = 7.29m Number of tubes along the cross section = 7.29 / (2 x (0.01905+0.015=0.03405)) = 107 tubes

Hence the numbers of rows of tubes are 404 / 22 = 17.8 rows Assuming the same pitch, height of convection section = 18 x 2 x 0.03405 =1.21m Final dimension of the convection box is: 1.21m (Height) × 7m (Length) × 7.29m (Width)

3.3.2 Design parameters for convection tubes

Dimensions and tube material as provided by Heat Exchanger Network (HEN) counterpart:

Thickness (m)

0.002

Outer diameter (m)

0.01483

Tube nominal size15

Schedule 10

Tube material

Carbon Steel

The minimal thickness to withstand the creep of carbon steel is found using the equation as used for HK40 calculated above (for thickness of reformer tube minimum wall

thickness),

min t wall

d  P i + FCA  2  =  S a E − 0.6 P

Given that the creep rupture strength of carbon steel is 54000 psi, the thickness is 0.00285in. Hence the minimum requirement is satisfied.

Production of Hydrogen via Syngas Route

3-21

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.3.3 Pressure drop in the tubes present in furnace

The typical pressure drop for crude unit heaters is between 150-250 psi. An additional 20-25 psi is added for fouled tubes. In order to satisfy this condition, the number of passes and tube size has to be optimized. Flow of the fluid within the furnace pipes is turbulent, via Reynolds number: Re = Di ρV µ . Thus, a correlation developed by Haaland19 was used to determine the Fanning friction factor: 1/

[

f f = −3.6 log 10 (6.9 / Re ) + (e / 3.7 Di )

10 / 9

]

The following conditions must be satisfied to accurately determine the friction factor with this correlation: (a) 10 8 ≥ Re ≥ 4 × 10 4 ; (b) 0.05 ≥ e / Di ≥ 0 Otherwise ff = 16/Re (for laminar flow) Computation of the frictional head loss for a straight pipe is evaluated using the relation:

hL = 2 f f LV 2 / (2 gDi )

The pressure drop across the straight pipe is then given by: ∆Pp = ρhL

Presence of 180° bends within the 2 sections also contribute to the pressure drop because the direction of flow changes. For each bend, a friction loss factor of K=1.6 is used to compute the head loss. Subsequently, the pressure drop is obtained: ∆PB = ρhL = KV 2 / (2 g )

The total pressure drop in the tubes is evaluated by addition of the pressure drop across straight pipes and the bends: ∆PT = ∆P1 + ∆P2

Production of Hydrogen via Syngas Route

3-22

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

The pressure drop for the following streams is computed. For the SMR stream, pressure drop through bends is neglected as the tubes are straight.

Stream

Re

SMR tubing

Steam in

SMR feed in

Preheated air

in radiation

convection

convection

in convection

section

section

tubes

tubes

852

7934

7334

230.67

Fanning friction factor

0.0188

0.00202

0.00218

0.0694

(assuming smooth tube) Friction head loss

0.103

0.142

346.6

154062

Pdrop through straight

-

112

3982

180247

tube (psi) Pdrop through bends

-

-

-

-

0.109

112

3982

180247

(psi) Total Pdrop (psi)

It is noted that the pressure drop across the tube for preheating SMR feed and for preheating air feed is much higher than the typical value. However, as this design is based on the inner diameter as supplied by the heat exchanger network person-in-charge, this problem will only be brought up for further mitigation on the best diameter for the convection tubes. The pressure drop for SMR tubing is negligible, which is ideal for the steam methane reforming reaction.

Production of Hydrogen via Syngas Route

3-23

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.4 STACK DESIGN

The stack is designed to direct the flue gas out of the furnace into the atmosphere as well as to achieve a draft of required combustion air through the furnace. The stack height must be sufficient to achieve this flow without imposing a positive pressure on any part of the furnace chamber18. The usual practice is to maintain a small negative pressure in the furnace to enable the introduction of air from the atmosphere. It also allows for the removal of undesirable products from the furnace. The required stack height is dependent on the temperature of the flue gas leaving the convection section and the difference in density of the flue gas and the atmospheric air.

3.4.1 Stack diameter

An acceptable velocity for the flue gas velocity is found12 to be 7.6m/s. Assuming that the stack is a uniform cylinder, Diameter = [(volumetric flow rate of flue gas) / (π x flow rate)] ^ 0.5 = 4.96m

3.4.2 Pressure Drop across stack

3.4.2.1 Stack exit loss

The stack exit loss is computed as follows: ∆P1 = 0.176 KV g2 / (Ta + 273) Velocity of flue gas = 11.5m/s

∆P1= 0.0783kPa

3.4.2.2 Frictional Loss in stacks and ducts

The flow in the stack is turbulent and hence the von Karman’s equation is used. 1/

f f = 4 log10 [Ds / e] + 2.28

∆P2 = 2 f f HV g2 / ( Ds g )

Assume the roughness factor is 0.5, ff= 0.86. H is taken to be 4m as an initial guide. ∆P2= 0.0186 kPa

Production of Hydrogen via Syngas Route

3-24

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.4.2.3 Stack entrance loss

Stack entrance loss takes into account the full velocity head loss due to a change in direction as the stack gas exits the furnace. ∆P3= 5.5×10-5 kPa

∆P3 = 0.176V g2 / (Tstack + 273)

3.4.2.4 Flue gas pressure drop through the convection section

Gunter Shaw’s correlation is used for pressure drop of a bank of helical bank tubes of staggered arrangements. fG 2 L p

 d ev  ∆P4 = 10 ( ρ g / ρ w )5.22 × 10 d evφ s  S T

  

0 .4

 SL   ST

  

0.6

∆P4 = 1.11×10-5 kPa

3.4.2.5 Pressure drop at the top of the radiant section

A vacuum of 2 mm H2O gauge just below the convection section is to be maintained to prevent leakage of flue gas through the casing of the furnace. Hence ∆P5 = 0.0020kPa.

3.4.2.6 Pressure gain at the convection section

The stack effect at the convection section brings about a pressure gain in the furnace. This gain is caused by the density difference between the hot flue gas and the ambient air outside. ∆P6= 2.7 x 10-2 kPa ;

Total Pdrop across stack = ∆P1 + ∆P2 +∆P3 + ∆P4 +∆P5 +∆P6 = 0.0720kPa

Production of Hydrogen via Syngas Route

3-25

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.4.3 Stack Height

A stack height of 4m has been approximated for calculation of pressure drop through the stack. The stack height is calculated again to ensure that the approximation is correct. According to the Code of Practice on Pollution Control by National Environment Agency (NEA), the stack height should be at least 15m from the ground. This is so that the hot stack gases are discharged at a safe height with respect to the surrounding equipment in the plant. In addition, the flue gas may contain pollutants such as SOx, NOx and particulates. Hence, the stack must be designed to discharge these gases in a manner that avoids causing a local pollution problem.

The equation used to calculate the stack height:

  Pd = 0.35 H g Patm  1 − 1  T T ga   a

Pd = 0.0720kPa. H is found to be 10m. Hence the average height is taken to be 8m.

Production of Hydrogen via Syngas Route

3-26

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.5 MATERIALS FOR CONSTRUCTION OF FURNACE BODY & ADDITIONAL AUXILIARIES 3.5.1 Refractory walls

The refractory walls are to be made of strong material that can withstand high temperature. It also must resist abrasion and flue gas and most importantly, it should have a high insulation to prevent heat loss to the surrounding.

In this design, silica brick (97-98% silica) with a thickness of 5-8 inches is selected to line the furnace walls. It has the ability to retain its strength at high temperatures. A highly porous fire clay insulating firebrick (1”) is placed between this lining & the metal casing. The silicon carbide coating is light, low in thermal conductivity and sufficiently resistant to temperature for the use on the hot side of the furnace wall. Thus, it permits thin walls of low thermal conductivity and low heat content. The low heat content is particularly valuable in saving fuel and time on heating time.

The properties of the silicon carbide and insulating wall are shown below. Properties

Silicon Carbide

Insulating Brick

Thermal shock resistant

Excellent

Excellent

Hot strength/

Excellent/

Poor/

Deformation under hot loading/

Excellent/

Poor/

Permeability

Very Low

High

4175

Varies

Bulk density lb/ft

160

30-75

Composition

SiC 80-90%

Varies

o

Fusion pt ( F) 3

To further confirm that the refractory material chosen will be able to withstand high temperature from the flames, Stefan-Boltzman equation will be used:

q r = σT 4

Given that the radiant heat flux is about 25,000 BTU/h-ft2, which is 7.9 × 104 W/m2, T (wall temperature) = 1086K = 1495F which is less that the fusion pt of silicon carbide and hence this material is suitable for use.

Production of Hydrogen via Syngas Route

3-27

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.5.2 Stack Walls

Stainless steel will be used15 for construction of stack walls instead of insulating brick. This reduces the cost and will exert less stress to lesser weight of the material. Stainless steel melts at around 1370°C. Since the maximum temperature within the stack is approximately 565°C, stainless steel is suitable to form the stack wall. However, it is important to note that since metal is involved, the temperature within the stack should be kept above 150°C, which is above the dew point of water to prevent condensation and thereby the formation of acid which will corrode the metal.

3.5.3 Additional auxiliaries

3.5.3.1 Air Preheaters

There are commercially available air preheaters to heat up the furnace air feed. One of the commercially available air preheater is the Rekuluvo® Recuperative Air Preheater. Air is preheated prior to burning in the furnace to ensure higher heat recovery. This The good accessibility to heating surfaces allows easy maintenance. In addition, it is corrosion resistant and does not have any mechanical moving parts that need additional power supply.

Production of Hydrogen via Syngas Route

3-28

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.5.3.2 Forced Draft Fan

Forced draft has to be installed at the inlet of the furnace to draw air supply to the furnace. This is because after being preheated by heat exchangers, the pressure of the air supply drop by 3 psi for each heat exchanger. Having passed through 3 heat exchangers, the pressure drop would be 9 psi in total. The final pressure before entering the fired heater might be 14.5 – 9 = 5.5 psi, which is very low. The forced draft will be used to increase the pressure of the air supply to 1atm prior to feeding into the fired heater.

3.5.3.3 Induced Draft Fan

Induced draft fan is placed at the outlet of the furnace to draw the flue gas out of the stack. A pressure of 2mmH2O less than atmospheric pressure is maintained. The proposed force and induced draft fan are shown below. Since both types of draft are used, the set-up is known as balanced draft. The fans will be chosen in a way that the pressure is slightly below atmospheric pressure. This ensures safe operation and reduces leakage of air into the furnace.

Production of Hydrogen via Syngas Route

3-29

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.6 COST ANALYSIS 3.6.1 Purchased Equipment Costs

Preliminary cost estimation was done to estimate the furnace and its auxiliary equipment cost using the CAPCOST program developed by Turton.

3.6.1.1 Costing for Furnace

The bare module cost of reformer furnace before accounting for inflation is calculated from Equation (3-1). Equation (3-2) gives the pressure factor (Fp) for the furnace. As carbon steel is the base material used, the material factor, FM, is 2.1 . (3-1) for P < 10 barg

(3-2) (3-3)

where Ft (superheat correction factor for steam boilers) = 1 for heaters and furnaces Identification number for HK40 alloy steel is 54, hence bare module factor FBM = 2.5 The various parameters that will be used for cost estimation: Parameter

Unit

Value

A, Heat Duty

KW

1.65 x 105

Pbarg

barg

1

FM

2.5

FP

1

CBM= USD 1.545 x 1012 As the data for the equations were obtained during May to September 2002 when Chemical Engineering Plant Cost Index (CEPCI) was 395.6, inflation should be accounted for using the CEPCI of 595.1 in the last quarter of 2007 to USD 2.32 x 1012.

Production of Hydrogen via Syngas Route

3-30

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.6.1.2 Costing for Air Preheater

The bare module cost of the air preheater was approximated to that of a flat plate heat exchanger because of the difficulty of getting the actual cost from vendors. Carbon steel was chosen as the base material. Similarly, the bare module cost is calculated with equation 3-4. C BM = C op (0.96 + 1.21FM Fp ) Parameter

Unit 2

A, Area

M

Pbarg

barg

(3-4) Value 160 1

FM

1

FP

1

After taking inflation into account by using the CEPCI of 512 in the last quarter of 2006, the estimated cost of the air preheater is US$256,000.

3.6.1.3 Costing for Induced Draft Fan and Forced Draft Fan for Air Preheating System

The induced and forced draft fans selected are centrifugal fans, thus the bare module cost can be approximated with that of the centrifugal radial fan. For both, carbon steel was selected as the bare material since only flue gas and combustion air will be in contact with them at relatively low temperatures. C BM = C op FM Fp

(3-5)

The table below shows the Fans Bare Module cost parameters

Using the same CAPCOST Program, and after taking inflation into account, the estimated cost of the induced draft fan is US$13,500 and forced draft fan is US$52,000.

Production of Hydrogen via Syngas Route

3-31

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.6.1.4 Burners

There are 3072 small premix burners used for this operation. Assuming each of the premix burner cost USD 100, the total cost is USD 307200.

3.6.2 Utility Cost

Electricity (440V, 3-phase, 50Hz) – USD 100/MWh Natural gas feed is not considered for furnace operation as the off gas from the PSA is sufficient to supply enough heat energy required.

3.6.2.1 Electricity cost

The units of the furnace that runs on electricity are the induced and forced draft fans. Hence, the electricity consumption will be based on the Horsepower rating of the individual models that were selected according to the required capacity of volumetric gas flow rate. Fan

Horsepower

Power (KW)

Induced Draft

25

19

Forced Draft

30

22

Given the power rating, the amount of electricity to operate both fans is US$14,300/year.

3.6.3 Total Annual cost

Assuming 15 years of operation: Total annual cost = total bare module cost / 15 years + operating cost per year Total annual cost = USD 1.55 x 105 million per year

Production of Hydrogen via Syngas Route

3-32

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.7 SUMMARY & CONCLUSION

In this project, a fired heater design for the steam methane reforming process has been developed. The design methodology was presented. Design of the fired heater began with consideration of the reaction heat energy requirement. The radiation section is then designed with heuristics and maximum temperature specifications in mind. After the completion of the radiation zone design, the convection design was explored to increase the efficiency of the fired heater. The streams to be heated in the convection zone were identified and the stream data and tube dimensions were obtained from the heat exchanger person-in-charge. With the information available, the finned tube arrangements were determined and the sizing of the convection section was obtained. Both design process paid attention to heuristics and chances for optimisation. Finally, stack design was carried out to meet specifications by governmental bodies. Costing was then performed to determine the total annual cost of the fired heater constructed.

Production of Hydrogen via Syngas Route

3-33

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.8 SPECIFICATION OF FIRED-HEATER

Service: Steam Methane Reforming Design Duty: 5.6 x 108 BTU/h

Unit: Vacuum Unit heater

No. of heaters 1

Type: Box

Design Service Heat absorption (mmBTU/h) Fluid

Radiant Section SMR 561

SMR Process Fluid

Allowable pressure drop (psi) Allowable average heat flux (BTU/h-ft2) Fouling factor Residence time

Convection Section (Total) Preheating Streams 108.6 15.9 24.7

Superheated steam 150-200

Air

SMR feed

25000 0 N/A

Inlet Conditions Temperature (°C) Pressure (kPa)

Liquid flow (kg/h) Vapour flow (kg/h) Liquid density (kg/m3) Vapour density(kg/m3) Viscosity(cST) Specific heats (KJ/KJmole-C) Thermal conductivity (W/m-K) Design

539.4 2679

253.3 4200

25 100

25 4000

N/A 1.835 x 105 N/A 7.059 3.497 46.29

1.743 x 105 N/A 790.7 N/A 0.1336 99.98

N/A 2.997 x 105 N/A 1.167 16.12 29.24

N/A 4.3 x 104 N/A 28.99 0.4180 40.72

0.08

0.6121

0.02586

0.0369

Radiant Section

Convection Section (Total)

Outlet Temperature (°C) Pressure (kPa) Liquid flow (kg/h) Vapour flow (kg/h) Liquid density (kg/m3) Vapour density(kg/m3) Viscosity(cST) Specific heats (KJ/KJmole-C) Thermal conductivity (W/m-K)

851.9 2630 N/A 1.835 x 105 N/A 3.55 8.662 38.47

254.3 4179 N/A 1.743 x 105 N/A 20.27 0.8705 43.58

80.30 79.32 N/A 2.997 x 105 N/A 0.7804 27.33 29.57

250 3979 N/A 4.3 x 105 N/A 15.10 1.192 49.05

0.2003

0.05

0.02972

0.07

Production of Hydrogen via Syngas Route

3-34

CN 4120: Design II Team 32: Ng Su Peng (U046929L)

Furnace Unit Design Report

3.9 REFERENCES

1. Furnace/ Fired Heater Design and Control Lecture notes 2. Robin Smith. (2005). Chemical Process Design and Integration. John Wiley & Sons. 3. W. L. Nelson. (1985). Petroleum Refinery Engineering, Auckland. McGraw-Hill 4. V. Ganapathy. (1982). Applied heat transfer. PennWell. 5. Role of Alloying Elements. Retrieved from World Wide Web on 16 Mar 2008: http://www.valve-world.net/pdf/11022.pdf 6. Retrieved from World Wide Web on 30 Mar 2008: http://www.fitness4service.com/publications/pdf_downloads/Jaske-Shannon%20Paper.PDF 7. Inspection and Remaining Life Evaluation of Process Plant Equipment. Retrieved from World Wide Web on 16 Mar 2008: http://www.kubotametal.com/alloys/heat_resistant/HK40.pdf 8. Dilton, C.P. (1992). Materials selection for the chemical process industries. 9. Role of Alloying Elements. Retrieved from World Wide Web on 16 Mar 2008: http://www.valve-world.net/pdf/11022.pdf 10. Rostrup-Nielsen, J. (1993). Steam Reforming Opportunities and Limits of the Technology, Catalysis Today, Vol. 18, P305-324. 11. James R. Cooper, W. Roy Penney, James R. Fair. (2005). Chemical Process Equipment, Second Edition: Selection and Design. Elsevier. 12. R.K. Sinnott, Coulson & Richardson's chemical engineering - Volume 6: Chemical engineering design, Elsevier Butterworth-Heinemann (2005) 13. D.S.J. Jones. (1996). Elements of Chemical Process Engineering. John Wiley & Sons. 14. Melting Point of Iron – Jefferson Lab. Retrieved from World Wide Web on 20 Mar 2008: http://72.14.235.104/search?q=cache:3kYX9gKVDVEJ:education.jlab.org/qa/meltingpoint_0 1.html+melting+point+of+steel&hl=en&ct=clnk&cd=1&gl=sg 15. James R. Welty, Charles E. Wicks, Robert E. Wilson & Geogory Rorrer. (2001). Fundamentals of Momentum, Heat, and Mass Transfer – 4th Edition. John Wiley & Sons, Inc. 16. S. Singh, S Goyal. (2002). Fired Heaters in Chemical Process Industries CPECNews:P2-6 17. Retrieved from World Wide Web on 30 Mar 2008: http://www.vulcanfinnedtubes.com/ 18. R.K. Sinnott. (2005). Coulson & Richardson's Chemical Engineering - Vol 6: Chemical engineering design. Elsevier Butterworth-Heinemann. 19. S. E. Haaland. (1983). Trans. ASME, JFE: Vol. 105, P89

Production of Hydrogen via Syngas Route

3-35

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Chapter 4 : HIGH TEMPERATURE SHIFT REACTOR 4.1 INTRODUCTION

Hydrogen is an extremely important compound required in our lives. It is used to provide food, fuel and chemical resources for us. The largest processes using hydrogen are ammonia synthesis, methanol synthesis, and hydrogenation. Without it, it would cause us to live very differently in the world today. The production of hydrogen can be carried out using steam reforming of any hydrocarbon source such as coal, methane, petroleum naphtha or biomass. Methane is usually used due to its cheaper costs than the other hydrocarbon sources.

4.1.1 Water gas shift

During the steam reforming stage, side reactions happening in the steam reformer would cause carbon monoxide to be formed, and this limits the production of hydrogen. Thus, water gas shift reaction was developed to obtain more hydrogen from carbon monoxide. This is the water gas shift reaction: CO + H 2 O ⇔ CO2 + H 2

which involves the reaction of carbon monoxide and water in the presence of a suitable catalyst to form carbon dioxide and hydrogen. There are three alternatives for carrying out the reduction of CO. [1] 1. Remove part of CO with iron catalyst in one bed. Then absorb CO2 and go to a second bed of the same catalyst with a more favourable equilibrium since the product CO2 is absent. 2. Conduct the entire reaction in a single bed on copper-zinc catalyst. 3. Remove part of the CO in a bed with iron catalyst and complete the removal in a second bed of the more expensive copper-zinc catalyst. The second and third alternatives are more attractive as the additional absorption equipment in the first alternative creates added maintenance problems, particularly due to the corrosive character of monoethanolamine, which is the usual absorbent used.

Production of Hydrogen via Syngas Route

4-1

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

For the purpose of this design, I have decided to use the third alternative. This is because the second alternative would be much more expensive. The cost of copper-zinc catalyst is 3 times the cost of iron catalyst. Therefore, it is justified to use alternative 3, to achieve the required CO reduction.

4.1.2 High temperature shift CO + H 2 O ⇔ CO2 + H 2

∆H rxn = −44.447 kJ / mol

As seen from the enthalpy of the water gas shift reaction, it is an exothermic reaction. Thermodynamically, the conversion of the reaction is favored at low temperature. The lower the temperature is, the higher the conversion will be. However, at low temperatures, the rate of reaction is slow. Though conversion is high, it might take a very long time for it to reach that conversion equilibrium. Therefore, to ensure a high rate of reaction and a high overall conversion, it is necessary to use a High Temperature Shift (HTS) followed by a Low Temperature Shift (LTS). This mechanism is needed so that in the HTS reactor, the reaction occurs at a reasonably high rate. Then the reaction is completed in the LTS, which would ensure a reasonable overall conversion. The higher temperature in the HTS reactor also allows recovery of the heat of reaction at a sufficient temperature level to generate high pressure steam. The HTS is usually conducted at a range of 315oC-480 oC. [1]

Production of Hydrogen via Syngas Route

4-2

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.2 PROBLEM DESCRIPTION

After Steam Methane Reforming (SMR), the products would be thrown into the HTS reactor. The objective of the HTS reactor is to reduce carbon monoxide composition to 3% (dry basis). This was justified by literature data. [1] After which, the products would be thrown into the LTS reactor for further conversion so that the exit carbon monoxide composition would be 0.7% (dry basis).

Feed specification

The relevant data of the outlet stream from SMR are as follows: Table 4.2.1: SMR outlet stream data Flow rate (kmol/hr)

14470

Pressure (kPa)

2610

Temperature (K)

1125

Table 4.2.2: SMR outlet stream composition Component

Mol fraction

Molar flow (kmol/hr)

CH4

0.03493

505.4

H2O

0.34518

4994.8

CO

0.08737

1264.3

CO2

0.05451

788.8

H2

0.47513

6875.2

N2

0.00072

10.4

C2H6

0.00216

31.2

Steam-to-CO Ratio

According to literature, steam to carbon monoxide ratio must surely be more than 4:1.[1] The optimum amount of steam to be used is based on economic considerations, such as the cost of steam. Furthermore, using more steam requires equipment with a larger diameter due to a greater flow rate.

Production of Hydrogen via Syngas Route

4-3

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

From the above table, the steam to carbon monoxide molar ratio is 4:1. However, it was still unable to achieve a reduction of CO to 3% (dry basis) at this steam ratio. Therefore, steam was added to make the steam to carbon monoxide molar ratio 5:1. This was justified by literature data, as they also used a steam to carbon monoxide ratio of 5:1. [1] Doing this would reduce the amount of catalyst needed, as well as make it possible to achieve a reduction of CO to 3% (dry basis). Amount of steam added = [(1264.3 × 5) − 4994.8] kmol/hr = 1326.6 kmol/hr After adding 1326.6kmol/hr of steam to make the steam to carbon monoxide molar ratio (5:1), also cooling the inlet stream down to 627 K, and assuming a pressure drop of 20.88kPa across the heat exchanger, the HTS inlet stream data are as follows:

Table 4.2.3: HTS inlet stream data Flow rate (kmol/hr)

15796.6

Pressure (kPa)

2589.12

Temperature (K)

627

Table 4.2.4: HTS inlet stream composition Component

Mol fraction

% composition

Molar flow

(dry basis)

(kmol/hr)

CH4

0.03200

5.33

505.4

H2O

0.40017

-

6321.4

CO

0.08004

13.34

1264.3

CO2

0.04993

8.32

788.8

H2

0.43523

72.56

6875.2

N2

0.00066

0.11

10.4

C2H6

0.00198

0.33

31.2

Production of Hydrogen via Syngas Route

4-4

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Product specification

Therefore, the problem is defined to reducing carbon monoxide from 13.34% (dry basis) to 3% (dry basis). This is the table of the compositions of the HTS outlet stream after the single fixed bed catalytic reactor was designed: Table. 4.2.5 HTS outlet stream composition Component

Mol fraction

% composition

Molar flow

(dry basis)

(kmol/hr)

CH4

0.03200

4.85

505.4

H2O

0.33994

-

5369.9

CO

0.01980

3.00

312.8

CO2

0.11017

16.69

1740.3

H2

0.49546

75.06

7826.6

N2

0.00066

0.10

10.4

C2H6

0.00198

0.30

31.2

Conversion of CO in designed HTS reactor = (1- 0.01980/0.08004) x 100% = 75.27%

Production of Hydrogen via Syngas Route

4-5

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.3 REACTION THERMODYNAMICS 4.3.1 Criteria for Chemical Reaction Equilibrium

The fundamental property relation for single-phase systems, provides an expression for the total differential of the Gibbs energy: `

a `

a

`

a

d nG = nV dP @ nS dT + Σ ui dni

(4.3.1.1)

i

If changes in the mole numbers ni occur as the result of a single chemical reaction in a closed system, then by substituting dni = v i dε , equation (2.3.1.1) gives: `

a `

a

`

a

d nG = nV dP @ nS dT + Σ v i ui dε

(4.3.1.2)

i

Because nG is a state function, the right side of this equation is an exact differential expression; thus, ` a F∂fffffffffffffffffffff nG G

Σ v i ui =

∂ε

i

t F ∂G ffffffffffff G

=

∂ε

T,P

T,P

Thus the quantity Σ v i ui represents the rate of change of total Gibbs energy of the system i

with respect to the reaction coordinate at constant T and P. This quantity is zero at the equilibrium state. A criterion of chemical-reaction equilibrium is therefore: Σ v i ui = 0

(4.3.1.3)

i

The definition of the fugacity of a species in solution is as such: `

a

^

µ i = Γ i T + RTln f i

In addition, the following equation may be written for pure species i in its standard state at the same temperature: `

o

a

o

G i = Γ i T + RTln f i

The difference between these two equations is: ^

o i

µ i @G = RTln

fi ffffffff o

fi

(4.3.1.4)

Combining equation (4.3.1.3) with equation (4.3.1.4) gives for the equilibrium state of a H

f

J o Σ v i G i + RTln f i ^

chemical reaction:

)

g vi o

fi

i

Production of Hydrogen via Syngas Route

I K

=0

4-6

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report f ^

Σ v i G + RTΣ ln f i

or

o i

i

f ^

Π fi

or

g vi

)

f

o i

=0

i

g vi

)

o

fi

o

=

i

@Σ vi G i fffffffffffffffffffffffffff i

RT

where Π signifies the product over all species i. In exponential form, this equation becomes: i

f

)

^

Π fi

g vi o

fi

=K

i

4.3.2 Effects of Pressure on Reaction Equilibrium

The equilibrium state of a chemical reaction is given as: f ^

Π fi

g vi

)

f

o i

i

=K

(4.3.2.1)

Where Π represents the product over all species i, f is the fugacity of species i in solution, o

f i is the fugacity of species i at standard state, and the equilibrium constant K is a function

of temperature only and is defined by: f

og

@∆G K = exp ffffffffffffffffffff RT

(4.3.2.1a)

The standard state for a gas is the ideal-gas state of the pure gas at the standard state pressure Po of 1 bar. Because the fugacity of an ideal gas is equal to its pressure, fio = Po for each ^

species i. Thus for gas-phase reactions

fi

)

o

^

fi = fi

)

o

P , and equation

(4.3.2.1) becomes: h

^

i

vi

fi k =K Πj ffffffff o i P

(4.3.2.2)

Equation (4.3.2.2) relates K to fugacities of the reacting species as they exist in the real equilibrium mixture and these fugacities reflect the non-idealities of the equilibrium mixture. The fugacity is related to the fugacity coefficient by: ^

f i = Φ^ i y i P

Production of Hydrogen via Syngas Route

4-7

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Substituting this equation into (4.3.2.2) yields an equation which relates the pressure to the composition: d

e vi

^

Π yi Φi

f

g @v

P ffffffff = o P

i

(4.3.2.3)

K

Where, v a P v i . Assuming that the equilibrium mixture is an ideal solution, then each Φ^i becomes Φ i . Thus, equation (4.3.2.3) becomes: b

Π yi Φi

c vi

i

f

g @v P ffffffff

=

P

o

(4.3.2.4)

K

Each Φ i for a pure species can be calculated from a generalized correlation once the equilibrium temperature and pressure is specified. For low pressures or high temperatures, the equilibrium mixture behaves as an ideal gas where Φ^i = 1. Thus, assuming that the equilibrium mixture is an ideal gas, equation (4.3.2.4) reduces to: `

Π yi

av i

i

f

=

g @v P ffffffff

P

o

(4.3.2.5)

K

In the WGS shift reaction, the stoichiometric coefficients of the reactants and products are all `

a

1 which means that v = P v i = 1 + 1 @1 @1 = 0 . Therefore, equation (4.3.2.5) reduces to: `

Π yi

av i

i

=K

(4.3.2.6)

From equation (4.3.2.6), it can thus be seen that the equilibrium constant of the WGS reaction, K is independent of pressure. Thus, the pressure conditions within the HTS reactor will not affect the equilibrium of the reaction.

4.3.3 Effects of Temperature on Reaction Equilibrium

From the first law of thermodynamics for a closed system of n moles, is as such for the special case of a reversible process: `

a

d nU = dQ + dW `

(4.3.3.1) a

`

a

As applied to this process, dW = @Pd nV and dQ = Td nS . Combining these three equations gives: `

a

`

a

`

d nU = Td nS @Pd nV

Production of Hydrogen via Syngas Route

a

(4.3.3.2)

4-8

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

The enthalpy and the Gibbs energy are defined as: H a U + PV

(4.3.3.3)

G a H @TS

(4.3.3.4)

Upon multiplication by n, equation (4.3.3.3) may be differentiated to give: `

a

`

a

`

a

`

d nH = d nU + Pd nV + Vd nP

a

(4.3.3.5)

Substituting equation (4.3.3.2) into (4.3.3.5), `

a

`

a `

a

d nH = Td nS + nV dP

(4.3.3.6)

In the same way, equation (4.3.3.4) may be multiplied by n and differentiated to give: `

a

`

a

`

a

`

d nG = d nH @Td nS @Sd nT

a

(4.3.3.7)

Equation (4.3.3.6) and equation (4.3.3.7) combine to yield: `

a `

a

`

a

d nG = nV dP @ nS dT

(4.3.3.8)

In the application of equation (4.3.3.8) to a one mole of homogeneous fluid of constant composition, equation (4.3.3.8) simplifies to: dG = VdP @SdT

(4.3.3.9)

An alternative form of equation (4.3.3.9) which is a fundamental property relation that follows from the mathematical identity is: f

g

G 1 G d ffffffffff a ffffffffffdG @ fffffffffffff dT 2 RT RT RT

(4.3.3.10)

Substituting equations (4.3.3.9) and (4.3.3.4) into (4.3.3.10): f

g G ffffffffff

V H a ffffffffffdP @ fffffffffffff dT 2 RT RT RT

d

(4.3.3.11)

All terms in this equation are dimensionless. When applied in restricted forms, H b cI G+ RT ∂ H J ffffffffffffffffffffffffffffffK fffffffffffff

@ 2 = RT

∂T

(4.3.3.12)

P

The relation between the standard heat of reaction and the standard Gibbs energy change of reaction may be developed from equation (4.3.3.12) written for each species i in its standard state: h

o

H i = @RT

b ci o * G RT i 2j dffffffffffffffffffffffffffffffffffk

dT

Production of Hydrogen via Syngas Route

(4.3.3.13)

4-9

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

By multiplying both sides with ν i and summation over all species yields: h

b

o

ci

d P vi Gi * RT o 2j ffffffffffffffffffffffffffffffffffffffffffffffffffk P v i H = @RT dT o

o

o

(4.3.3.14)

o

By definition, ∆G a P v i G i and ∆H a P v i H i . Thus equation (4.3.3.14) can be i

i

expressed as: h

ci

b

∆G * RT 2j dfffffffffffffffffffffffffffffffffffffk o

o

∆H = @RT

dT

(4.3.3.15)

Substituting equation (4.3.2.1a), equation (4.3.3.15) becomes: o

dln K ∆H ffffffffffffffffff ffffffffffffff = 2 dT RT

(4.3.3.16)

Equation (4.3.3.16) gives the effect of temperature on the equilibrium constant, and hence on o

the equilibrium conversion. If ∆H is negative, i.e. the reaction is exothermic, the equilibrium constant decreases as the temperature increases. Conversely, K increases with T for an endothermic reaction. Since the water-gas shift reaction is slightly exothermic with ∆H = 41.1kJmol-1, thus the equilibrium constant increases with decreasing temperature. Thus, it is desirable to operate at the lowest possible reactor inlet temperature to obtain maximum removal of carbon monoxide.

Production of Hydrogen via Syngas Route

4-10

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.4 REACTION KINETICS

The overall reaction is CO + H 2 O ⇔ CO2 + H 2 Using literature data [1], we have chosen Chromia-promoted iron oxide as our catalyst. This catalyst has been used for many years for the shift reaction.

Rate Equation

The rate equation for this catalyst is shown below, and is assumed to represent midlife activity:

(−rCO ) =

ψk ( y CO y H O − y CO y H / K ) 379 ρ b 2

2

2

Where, k = rate constant = exp(15.95 −

4900 ) T

K = equilibrium constant = exp(−4.33 +

4578 ) T

(-rCO) = rate, lb moles CO converted / (lb catalyst) (hr) T = temperature, K yj = mole fraction of component indicated ρb = catalyst bulk density, lb/cu ft

ψ = 4.0 for P > 20.0 atm

The manufacturer has subjected the rate equation to many tests, as well as observations on full-scale plants. The rate constants are expressed on the basis of a reasonable “lined-out” activity that the catalyst would maintain for a considerable time, if operating errors which cause deactivation do not happen. The ψ term is the product of the total pressure (atm) and ratio of the first-order constant at pressure P to that at atmospheric pressure and is a function of pressure and Thiele modulus. Thus, it is considered that the effectiveness factor of the catalyst has already been taken into account in the ψ term.

Production of Hydrogen via Syngas Route

4-11

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.5 CATALYST

As mentioned, the catalyst used is chromia-promoted iron oxide. [1] The specifications of the catalyst are as follow: Table 4.5.1 Maximum operating temperature (oF)

890

Tablet size (inch)

0.25 x 0.25

Bulk density (lb/cu ft)

70

Particle density (lb/cu ft)

126

Catalyst poisons

Inorganic salts, boron, oils, or phosphorous compounds, liquid H2O is a temporary poison. Sulfur compounds in an amount greater than 50ppm

Catalyst life

3 years and above, depends on care in startup and operation (Use times up to 15 years have been reported)

Production of Hydrogen via Syngas Route

4-12

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.6 REACTOR

An adiabatic single fixed bed catalytic reactor is used for the HTS reactor.

4.6.1 Type of reactor

The reason why the reactor was chosen as a single bed was because the heat of reaction for CO + H 2 O ⇔ CO2 + H 2 was not that high. Therefore, temperatures in the reactor would not

rise by too much. A single bed would suffice to convert CO to its desired composition without raising temperature too high such that conversion would be affected. An adiabatic reactor was chosen because it is cheap and easy to maintain. It is not only the lowest cost and simplest type of reactor, but its performance can be predicted reliably for single phase reactions.

This is also justified by literature data as shown in the case study, where the author also used an adiabatic single fixed bed catalytic reactor.

4.6.2 Reactor design Vessel Design

Vessel costs are an important element in reactor design decisions. In the U.S.A. the American Society of Mechanical Engineers has established a code for the design and fabrication of pressure vessels. Similar organizations in Europe also have established codes. All such codes give the minimum standards. Normally vessels as important as reactors are designed to comply not only with a code but also with supplement specifications considered important for a particular service. These can include special impact test requirements to assure against brittle fracture, heat-treating specifications for steel in severe service such as high hydrogen partial pressures. Below is a picture of the design of the vessel.

Production of Hydrogen via Syngas Route

4-13

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Fig. 4.6.2.1 Design of vessel

Corrosion Allowance

Although practices vary, on the average a material is selected that will not corrode more than 0.010 to 0.015 in. /yr. For a vessel of life 10 years, this approximates a corrosion allowance of 1/8 in. Because of the many variables and unknowns associated with corrosion, a minimum allowance of 1/8 in is specified for carbon steel and low alloy steel even if no corrosion or erosion problems exist. For higher alloys, such as stainless steel, a lower minimum of 1/32 in is often used.

Production of Hydrogen via Syngas Route

4-14

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Material Selection

In the case of hydrogen services which cause pitting, corrosion allowance may not be that useful. Hydrogen destroys metal strength by producing cracks or blisters, but the thickness of the metal is not reduced. At low temperature atomic hydrogen produce by thermal or catalytic dissociation diffuses into the metal along imperfections, ultimately recombining to form molecular hydrogen. The hydrogen pressure can increase to a point where it causes internal and surface blistering. [4] At high temperatures, hydrogen diffuses even more rapidly and forms methane by reacting with the carbon content of the steels. The larger methane molecule builds up pressure that produces high internal pressure and ultimately cracks [5]. Neither of these processes reduces the metal thickness. Thus one selects for high temperature service a metal that will not be subjected to attack, containing a carbide stabilizing element such as molydenum.

Thus the material ASTM A 387 Grade 22, Class 1 (2 ¼ Cr-1 Mo) was chosen for its resistance to hydrogen attack.

Production of Hydrogen via Syngas Route

4-15

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.7 METHODOLOGY AND CALCULATIONS

4.7.1 Weight of catalyst Methodology

The following equations were used in the calculations for weight of catalyst used Mass balance equation: ∆W (−rCO ) = (−∆FCO )

(Eq. 4.7.1.1)

Heat balance equation: ∑ F j c p , j (T j +1 − T j ) = (− rCO )( −∆H CO ) T ∆W

(Eq. 4.7.1.2)

= (−∆FCO )(− ∆H CO )

The heat capacities of gases were taken from literature text. [2]

Based on these equations and the rate equation, a MATLAB program was written based on the following algorithm to find the mass of catalyst needed. Algorithm 1. Input the inlet temperature of HTS in K. 2. Assume ∆W of 200lbs 3. Calculate (-rCO) at inlet conditions to increment, i. 4. Calculate (-rCO)avg = (-rCO)i + [(-rCO)i – (-rCO)i-1]/2 (skip for i=0). 5. Calculate new flow rates: Fi+1 = Fi ± (-rCO)∆W 6. 4. Calculate cp and (-∆HCO) @ Ti 7. Calculate ∆T from Eq. 4.7.1.2. 8. Ti+1 = Ti + ∆T 9. yi+1 = Fi+1/(FT)i 10. Mole fraction CO in dry gas = [yCO/(1-yH2O)]i+1 11. If mole fraction CO in dry gas is more than 3%, go back to step 1. 12. If mole fraction CO in dry gas is 3%, mass of catalyst is found as number of increments multiplied by 200lbs. The MATLAB program can be found in Appendix 4.12.1.

Production of Hydrogen via Syngas Route

4-16

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Calculations

The temperature of inlet stream to the HTS reactor was varied to get the corresponding mass of catalyst required to reduce CO to 3% (dry basis). This was done to optimize the mass of catalyst used with its optimum inlet temperature of the inlet stream. Then, a graph was plotted to show the relationship between mass of catalyst and inlet temperature. Fig. 4.7.1.1 Graph of mass of catalyst against inlet temperature

From the graph, the minimum mass of catalyst needed was 121909 kg. However, 5% more catalyst was added to allow for any degrading of catalyst. Therefore mass of catalyst used = 1.05×121909kg = 128000kg This occurred at the inlet temperature of 627 K. Thus, for the HTS reactor, the inlet stream was fixed at 627 K.

Production of Hydrogen via Syngas Route

4-17

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

A temperature profile graph was also plotted in the figure below. The temperature range was from the inlet stream of 627 K to that of the outlet stream was which calculated to be 693.2 K.

Fig. 4.7.1.2 Temperature profile graph

Production of Hydrogen via Syngas Route

4-18

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

A conversion profile graph was also plotted as shown below. The inlet CO was at 13.34% dry basis and the outlet CO was 3% dry basis. Fig. 4.7.1.3 Conversion profile graph

4.7.2 Pressure drop

Pressure Drop, though negligible in some reactors, can be a major concern in others. It is an important variable in the rate equations for gaseous reactions. Since compressors and compressor operating costs often dominate the economic structure of a reactor system, pressure drop is not only important but must be predicted with good accuracy. The resulting force must not exceed the crushing strength of the particles. In homogeneous clean beds, one would expect the maximum stress to occur at the bottom of the bed, where the weight of the catalyst combines with the stress created by the ∆P across the bed. In down flow, this force created by the ∆P is transmitted by the contacting solids to the bottom of the bed. Some catalysts are quite fragile and this issue demands close attention with sufficient safety factor applied.

Production of Hydrogen via Syngas Route

4-19

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Mass velocities through the bed must be high enough to minimize inter-phase gradients and assure good distribution. Incremental increases in ∆P, however, should not cause pumping or compressing costs to exceed savings realized from improved reactor performance. In many packed bed systems, the maximum economical ∆P is in the range of 3-15 % of the total pressure.

In gaseous systems, higher pressure drop and thus higher velocity, also means smaller diameter reactors, which can be important in reducing costs of high pressure reactors; but this advantage can be offset by higher energy costs. The given fraction of plant pressure drop allotted to drop across the bed is directly proportional to the fraction of power consumed, which is essentially a function of energy costs and independent of total pressure. Thus economic allowable ∆P will be a fixed fraction of total pressure and can vary from a few inches of water for reactors operating near atmospheric pressure to several atmospheres for reactors operating at higher pressure. A unique value of particle density does not even exist for a given catalyst. Generally, smaller sizes will have higher particle densities than larger sizes, which can be rationalized by considering the limit of a catalyst approaching the size of an average pore. Dense packing in a full-size bed is preferred for uniform flow distribution and is obtained by raking or spreading the catalyst between each load. Although a rapidly dumped bed will result in looser arrangement and lower pressure drop, it is more likely to cause channelling. Although small catalyst particles have higher effectiveness factors, it is not wise to specify sizes below 1/8 in. unless some means is provided for removing fines, dirt and scale from the feed stream. The greatest care should be exercised in packing a bed to eliminate fines and dirt and the reactor should be protected by suitable filters whenever plugging by scale or polymer formation in upstream equipment is anticipated. These materials can be carried to the reactor and deposited on the top part of the bed and limit the throughput drastically. Plugging of a catalyst bed is a serious problem that can ultimately lead to shutdown and dumping of the bed as pressure drop becomes excessive. Prior to this event, serious malfunction of the reacting fluid can occur, resulting in poor yields and reduced production. The ability to predict cleanbed ∆P is often foreshadowed by our inability to predict the rate and character of plugging that may occur.

Production of Hydrogen via Syngas Route

4-20

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Methodology

For this HTS reactor, the inlet pressure is 2610kPa. Since maximum economical pressure drop is in the range of 3-15% of the total pressure. [3] Taking pressure drop to be 4% of total pressure. ∆P = 0.04×2610 kPa

= 104.4 kPa ≈ 1 bar

Therefore, a pressure drop of 1 bar is to be obtained. These are the equations required to find the pressure drop and aspect ratio. N Re =

DpG

µ

,

Where NRe = Reynolds’ number Dp =

6d c = 0.25 dc 4+2 hc

µ = Average of inlet and outlet viscosity f k = 1.75 + 150

1− ε , N Re

Where fk = friction factor

ε = voidage ∆P =

fkG2 Dp ρ f gc

1 − ε   3 L ,  ε 

Where G = mass flux = mass flowrate per cross sectional area

ρ f = density of feed L = length of reactor

Production of Hydrogen via Syngas Route

4-21

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

To find the aspect ratio of the reactor which corresponds to a pressure drop of 1 bar, the following algorithm was used.

1. The volume of the total catalyst was calculated based on mass of catalyst used and density. This volume is multiplied by 1.2 to give an extra 20% volume for the allowance of inert support as well as poor packing of catalyst. 2. Calculate average µ (viscosity) based on inlet and outlet. 3. Assume a value of L

V  4. Calculate the corresponding value of diameter, D = 2    πL  5. Calculate aspect ratio, AR=L/D 6. Calculate G 7. Calculate ε 8. Calculate N Re 9. Calculate f k 10. Calculate ∆P 11. Go back to step 3 and assume another value of L to get corresponding ∆P . 12. Plot graph of ∆P against AR. 13. Identify the AR where ∆P =1 bar The MATLAB program can be found in Appendix 4.12.2.

Production of Hydrogen via Syngas Route

4-22

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Calculations

The pressure drop against aspect ratio figure is plotted as shown below:

Fig. 4.7.2.1 Graph of pressure drop against Aspect ratio

Therefore, a pressure drop of 1 bar corresponds to an Aspect ratio of 3.497. L = 3.497 -------------- (1) D

Solving simultaneous equations

V= Mass of catalyst/ bulk density V= 113.92m3 2

D  L = 113.92 --------(2) 2

π L, length of reactor = 12.11m

D, diameter of reactor = 3.46m

Production of Hydrogen via Syngas Route

4-23

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.7.3 Thickness of vessel

With an internal pressure of 2610 kPa and diameter of 3.46m=11.35ft,

Design Pressure: 2610 x 1.2 x 0.145 = 454.2 psi, taking into account an allowance of 20% for increased operating pressure Design Temperature: 890 0F (maximum catalyst use temperature) S = 13100 psi E, the joint efficiency = 1.0, for double butt welded and fully radio-graphed welds Minimum Corrosion allowance = 1/8 in.

t=

PR 454.2 × 5.68 × 12 = = 2.413 in SE − 0.6 P (13100 × 1.0) − (0.6 × 454.2)

t actual = 2.413 + 0.125 = 2.538 in

The MATLAB program used to solve this can be found in Appendix 4.12.3.

4.7.4 Reactor size and cost Catalyst cost

Mass of catalyst = 128000 kg Cost of catalyst = US$20/cu ft in 1977 Particle density of catalyst = 126 lb/cu ft Volume of catalyst = 2234.9 cu ft CEPCI in 1977 = 204.1 CEPCI in 2006 = 499.6 Cost of catalytic bed = 2234.9 x $20 x 499.6/204.1 = US$109,414

Production of Hydrogen via Syngas Route

4-24

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Vessel Costs

From the literature data [1], the vessel cost can be estimated as follows: Using 2:1 elliptical heads of same thickness Material Density: ( ρ = 490 lb/ cu ft) Shell, π × 11.35 × (

2.538 ) × 39.72 × 490 = 146,778 lb 12

Heads 490 ×

π 4

× [(1.23 × 11.35) +

2.538 2 2.538 ] × × 2 = 32,696 lb 12 12

Total Weight = 146,778 + 32,696 = 179,474 lb For this size and type vessel, a fabricated cost of 73 cents/lb without nozzles was suggested as an estimating figure (1971 cost) by a fabricator. CEPCI in 1971 = 132.3 CEPCI in 2006 = 499.6 Cost of vessel= 179474 × 0.73 ×

499.6 = US $494,750 132.3

Production of Hydrogen via Syngas Route

4-25

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.8 HEAT EXCHANGER

With a conversion of 75.27% CO, a conversion reactor is simulated in Hysys. This is the figure that is extracted from Hysys to get the duty needed to cool down the HTS outlet before it goes into the LTS inlet. The LTS inlet is to be cooled to 493.1K.

Fig. 4.8.1 Hysys diagram of HTS reactor

The duty needed to cool the HTS outlet stream from 692 K to 493.1K is 1.099e+008 kJ/hr according to Hysys. However, the actual HTS outlet temperature as calculated from MATLAB is 693.2K. This is quite close to the calculated value from Hysys. The actual duty would be further discussed in the next part, Chapter 4.8.1.

Production of Hydrogen via Syngas Route

4-26

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.8.1 Heat Exchanger Design Considerations

Shell and tube heat exchanger which is the most commonly used basic heat exchanger configuration in the process industries is selected because it provides a comparatively large ratio of heat transfer area to volume and weight and it is mechanically rugged enough to withstand normal shop fabrication stresses and normal operating conditions. Also, it can be easily cleaned and components susceptible to failure (gaskets and tubes) can be easily replaced.

According to the heat exchanger network design, there are 3 heat exchangers designed to cool down the HTS outlet stream to the cooled LTS feed stream. I will be designing the heat exchanger which is used to cool HTS outlet and to heat up SMR feed. Split ring internal floating head heat exchanger is selected for this heat exchanger. It can be used for liquids that foul as the tubes and bundle can be removed from shell for cleaning or repairing without removing the floating head cover. Since the HTS outlet is the stream that causes more fouling than the SMR feed, I have chosen to use HTS outlet in the tube side, and SMR feed in the shell side.

4.8.1.1 Physical properties extraction

The physical properties of the two streams are extracted from Hysys. The average values are used for the design. Where duty, Q = 4.622 MW.

HTS outlet

inlet

outlet

mean

temperature (o C)

263.3

234.3

248.8

specific heat(kJ/kg-C)

2.637

2.622

2.629

thermal conductivity (W/m-C)

0.1123

0.1072

0.1097

density(kg/m3)

7.444

7.893

7.661

viscosity(cp)

0.01759

0.01700

0.01735

Production of Hydrogen via Syngas Route

4-27

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

SMR feed

temperature (o C)

218

253.3

235.7

specific heat(kj/kg-C)

2.326

2.342

2.334

thermal conductivity (W/m-C)

0.04511

0.0491

0.04712

density(kg/m3)

12.60

11.57

12.09

viscosity(cp)

0.01424

0.01535

0.01480

4.8.1.2 Determination of overall heat transfer coefficient

After iterations, the overall heat transfer coefficient was U=540.1 W/m2oC. For an exchanger of this type with light gases as hot gas and methane and water vapor as cold gas, the overall heat transfer coefficient according to Table 12.1 of Coulson and Robertson’s Chemical Engineering Design textbook falls in the acceptable region.

4.8.1.3 Exchanger type and dimensions

∆Tm = Shell can be carbon steel. Tube can be stainless steel due to H2 pitting. The HTS outlet is dirtier than the SMR feed, therefore put the HTS outlet through the tubes and the SMR feed through the shell. ∆TLMTD =

R=

(Th,in − Tc,out ) − (Th ,out − Tc ,in )  Th,in − Tc ,out  ln    Th,out − Tc ,in 

263.3 − 234.3 = 0.8215 , and 253.3 − 218

=

(263.3 − 253.3) − (234.3 − 218) = 12.89 Ο C  263.3 − 253.3  ln    234.3 − 218 

S=

253.3 − 218 = 0.7792 263.3 − 218

From Fig 12.19, Ft = 0.70 , which is acceptable. ∆Tm = 0.70 × 12.89 = 9.02 Ο C

Production of Hydrogen via Syngas Route

4-28

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.8.1.4 Heat transfer area

AΟ =

Q 4.4622 × 10^ 6 = 915.3m 2 = U × ∆Tm 540.1 × 9.02

4.8.1.5 Layout and tube size

A split-ring floating head exchanger is used for efficiency and ease of cleaning. Use 19.00mm outside diameter, 15.00mm inside diameter, 5m long tubes on a triangular 23.80mm pitch.

4.8.1.6 Number of tubes

Area of one tube(neglecting tube sheets thickness) = π × 19.00 × 10 −3 × 5 = 0.2985m2 Number of tubes = 915.3 / 0.2985 = 3066 So, for 2 passes, tubes per pass = 3066 / 2 = 1533 (Check for tube-side velocity to see if reasonable) Tube cross-sectional area = ( π /4)(15 × 10-3)2 = 0.0001767 m2 Thus, area per pass = 1533 × 0.0001767 = 0.2709m2 Volumetric flow = (2.074 × 10^5/3600) × (1/7.661)=7.52 m3/s Tube side velocity, u t = 7.52 / 0.2709=27.76 m/s

4.8.1.7 Bundle and shell diameter

For 2 tube passes, K1= 0.249, n1= 2.207, So, Db = 19.0 × ( 1533 / 0.249 )1/ 2.207 = 1.36 m For a split-ring floating head exchanger the typical clearance is 20 mm, so the inside shell diameter, Ds= 1.36 + 0.02 = 1.38 m

Production of Hydrogen via Syngas Route

4-29

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.8.1.8 Tube-side heat transfer coefficient

Re =

7.661 × 27.76 × 15 × 10 −3 = 183,801 0.01735 × 10 −3

Pr =

2.629 × 10 3 × 0.01735 × 10 −3 = 0.42 0.1097

5000 L = 333 = Di 15.00

From figure 12.23, jh = 0.045 Nu = 0.023 × (183801) 0.8 (0.42) 0.33 = 281  0.1097  hi = 281 ×  = 2049 W/m2C −3   15.00 × 10 

4.8.1.9 Shell-side heat transfer coefficient

Take baffle spacing to be Ds /5 = 1.38/5 = 0.276 m = 276mm. This spacing should give good heat transfer.  23.80 − 19.00  3 2 2 As =  1380 × 0.4 × 1380 = 151 × 10 mm = 0.151m 23.80    1.27  2 2 D e=  (23.80 − 0.785 × 19.00 ) = 18.76mm  19.00 

Volumetric flow rate = 1.835 × 105/3600/12.085 = 4.217m3/s Shell-side velocity, us= 4.217/0.151 = 27.9 m/s

1.835 × 10 5 × 0.0188 Re = 3600 × 0.151 −5 = 4.274 × 105 1.48 × 10 2.334 × 10 3 × 0.01480 × 10 −3 Pr = = 0.733 0.04712 From Fig 12.29, jh=0.45 hs = (

0.04712 ) × 0.45 × 4.274 × 10 5 × 0.733 0.33 = 4.36 × 10 5 W/m2C 0.01876

Production of Hydrogen via Syngas Route

4-30

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.8.1.10 Overall coefficient  19  19 × 10 −3 × ln  1  1 1  19 1  15  + 1 + = +  + U 0  2049 1249  15 2 × 55 5678.6 436000

U 0 =540.1W/m2 oC.

4.8.1.11 Pressure drop

∆Pt = 2 × (8 × 0.045 ×

5 + 2.5) × (0.5 × 7.661 × 27.7512 ) = 722739 Pa = 722kPa 0.015

From Fig 12.30, jf = 0.028

∆Ps = 8 × 0.028 ×

1.37 4.83 12.085 × 27.9 2 × × = 185418 Pa = 185kPa 0.0188 0.5 2

Production of Hydrogen via Syngas Route

4-31

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.9 CONCLUSION

The HTS reactor designed has determined the parameters as shown: 1. Conversion of CO from 13.34% dry basis to a composition of 3% dry basis 2.

The mole fraction of the outlet of HTS

3.

The weight of catalyst needed for the reaction.

4. Pressure drop of the reactor. 5. The dimensions of the reactor were also calculated, namely length of reactor, diameter of reactor and thickness of reactor. 6.

The cost of the catalyst needed was calculated, as well as the cost of the vessel.

These give a good idea on the design of the HTS reactor as well as the cost of building the reactor.

Production of Hydrogen via Syngas Route

4-32

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.10 NOTATIONS

νi

:

stoichiometric coefficient of species i

µi

:

chemical potential of species i

n

:

number of moles

G

:

specific gibbs energy

ε

:

reaction coordinate

T

:

temperature of system

P

:

pressure of system

Gt

:

total gibbs free energy

∏

:

product over all species i

ν

:

total stoichiometric number

i

(− rCO ) :

reaction rate in lb moles CO converted/(lb catalyst/hr)

ψ

:

activity factor

k

:

rate constant

K

:

equilibrium constant

yi

:

mole fraction of species i

ρb

:

bulk density of catalyst (lb/ft3)

T

:

temperature in K

F

:

component molar flow rate

W

:

weight of catalyst

cp,j

:

heat capacity of component j

t

:

minimum thickness of wall without corrosion

P

:

design pressure of the reactor vessel

R

:

internal radius of shell without corrosion

S

:

maximum allowable stress value

E

:

joint efficiency (assume = 1)

Dp

:

characteristic length of pellet (ft)

G

:

mass flux (lb/s)

µ

:

average viscosity of fluid (cP)

Production of Hydrogen via Syngas Route

4-33

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

NRe

:

reynolds number

L

:

length of reactor

D

:

diameter of reactor

dc

:

diameter of cylindrical catalyst pellet (ft)

hc

:

height of cylindrical catalyst pellet (ft)

ρb

:

bulk density of catalyst pellet (70 lb/ft3)

ε

:

voidage

fk

:

friction factor

ρf

:

density of fluid (lb/ft3)

Production of Hydrogen via Syngas Route

HTS Unit Design Report

4-34

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.11 REFERENCES

1. H.F. Rase, Chemical Reactor Design for Process Plants, Vol 2, New York Wiley, 1977 2. J.M. Smith, H.C. Van Ness, M.M. Abbott, Introduction to Chemical Engineering

Thermodynamics 7th ed, 2005 3. H.F. Rase, Chemical Reactor Design for Process Plants, Vol 1, New York Wiley, 1977 4. R. Q. Barr, A Review of Factors Affecting the Section of Steels for Refining and

Petrochemical Applications, Climax Molydenum Co., Greenwich, Conn., 1971 5. C.H. Samans, Hydrocarbon Process., 42(10), 169 and (11) 241, 1963 6. H. M. Spencer, Industrial Engineering Chemistry, Vol 40, pg 2152-2154, 1948 7. K. K. Kelley, U.S Bur.Mines Bull. 584, 1960 8. L. B. Pankratz, U.S. Mines Bull. 672, 1982

Production of Hydrogen via Syngas Route

4-35

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

4.12 APPENDICES

Appendix 4.12.1 T= input('Please enter initial temperature for reactor in K:'); F= 2.2*15797; %flowrate in lbmol/hr yco= 0.080035; yh2o= 0.400173; yco2= 0.049934; yh2= 0.43523; yc2h6= 0.001975; ych4= 0.031995; yn2= 0.000658; Fco= F*yco; Fh2o= F*yh2o; Fco2= F*yco2; Fh2= F*yh2; Fc2h6= F*yc2h6; Fch4= F*ych4; Fn2=F*yn2; k=exp(15.95-4900/T); K=exp(-4.33+4578/T); rate=4*k*(yco*yh2o-yco2*yh2/K)/(379*70);

Fco1= Fco -rate*200; Fh2o1= Fh2o-rate*200; Fco21= Fco2 + rate*200; Fh21= Fh2 +rate*200; cpco= (3.376+(0.557/10^3)*T -(0.031*10^5)*T^(-2))*8.314; cpc2h6=(1.131+(19.225/10^3)*T-(5.561/10^6)*T^2)*8.314; cph2o=(3.47+(1.45/10^3)*T +(0.121*10^5)*T^(-2))*8.314; cpco2=(5.457+(1.045/10^3)*T -(1.157*10^5)*T^(-2))*8.314; cph2=(3.249+(0.422/10^3)*T +(0.083*10^5)*T^(-2))*8.314; cpn2=(3.28+(0.593/10^3)*T +(0.04*10^5)*T^(-2))*8.314; cpch4=(1.702+(9.081/10^3)*T -(2.164/10^6)*T^2)*8.314; deltaa = 5.457 + 3.249 - 3.376 - 3.470; deltab = (1.045 + 0.422 - 0.457 - 1.450) * 10^-3; deltad = (-1.157 + 0.083 - (-0.031) - 0.121) * 10^5; integral =(deltaa * 298.15 * ((T/298.15) - 1) + (deltab/2)*(298.15^2)*(((T/298.15)^2)-1) + (deltad/298.15)*(((T/298.15) 1)/(T/298.15))); dHco = -41166 + 8.314 * integral; dT = ((rate * 200) * (dHco))/(Fco*cpco+Fh2o*cph2o+Fco2*cpco2+Fh2*cph2+Fc2h6*cpc2h6+Fn2*cpn2+Fch4* cpch4);

Production of Hydrogen via Syngas Route

4-36

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

T1 = T + dT; yco1 = Fco1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh2o1= Fh2o1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yco21= Fco21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh21= Fh21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); ydryco= (yco1/(1-yh2o1)); i=1; while ydryco>0.03,

k=exp(15.95-4900/T); K=exp(-4.33+4578/T); rate=4*k*(yco*yh2o-yco2*yh2/K)/(379*70); k=exp(15.95-4900/T1); K=exp(-4.33+4578/T1); rate1=4*k*(yco1*yh2o1-yco21*yh21/K)/(379*70);

rateavg= (rate+rate1)/2; Fco = Fco1; Fh2o = Fh2o1; Fco2 = Fco21; Fh2= Fh21; Fco1= Fco -rateavg*200; Fh2o1= Fh2o-rateavg*200; Fco21= Fco2 + rateavg*200; Fh21= Fh2 +rateavg*200; cpco= (3.376+(0.557/10^3)*T1 -(0.031*10^5)*T1^(-2))*8.314; cpc2h6=(1.131+(19.225/10^3)*T1-(5.561/10^6)*T1^2)*8.314; cph2o=(3.47+(1.45/10^3)*T1 +(0.121*10^5)*T1^(-2))*8.314; cpco2=(5.457+(1.045/10^3)*T1 -(1.157*10^5)*T1^(-2))*8.314; cph2=(3.249+(0.422/10^3)*T1 +(0.083*10^5)*T1^(-2))*8.314; cpn2=(3.28+(0.593/10^3)*T1 +(0.04*10^5)*T1^(-2))*8.314; cpch4=(1.702+(9.081/10^3)*T1 -(2.164/10^6)*T1^2)*8.314;

integral = (deltaa * 298.15 * ((T1/298.15) - 1) + deltab/2*298.15^2*((T1/298.15)^2-1) + deltad/298.15*(((T1/298.15) 1)/(T1/298.15)));

dT = ((rateavg * 200) * (dHco))/(Fco*cpco+Fh2o*cph2o+Fco2*cpco2+Fh2*cph2+Fc2h6*cpc2h6+Fn2*cpn2+Fch4* cpch4);

Production of Hydrogen via Syngas Route

4-37

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

yco = yco1; yh2o = yh2o1; yco2 = yco21; yh2 = yh21; yco1 = Fco1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh2o1= Fh2o1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yco21= Fco21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh21= Fh21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); ydryco= yco1/(1-yh2o1);

x(i) = i*200/2.2; y(i) = T; a(i) = ydryco; T = T1; T1 = T + dT; i= i + 1; end Wt=i*200/2.2; fprintf('mass of catalyst is %f kg.\n',Wt); fprintf('outlet temperature of HTS is %f K.\n',T1); fprintf('mol fraction of CO is %f .\n',yco1); fprintf('mol fraction of H2O is %f .\n',yh2o1); fprintf('mol fraction of CO2 is %f .\n',yco21); fprintf('mol fraction of H2 is %f .\n',yh21); fprintf('mol fraction of N2 is %f .\n',yn2); fprintf('mol fraction of CH4 is %f .\n',ych4); fprintf('mol fraction of C2H6 is %f .\n',yc2h6); fprintf('mass flowrate in lb/hr is %f .\n',F); plot (x,y); plot (x,a);

Appendix 4.12.2 mass = 281609.79; %mass of catalyst in lbs flowrate = 456300; %flowrate in lbs per hour u1 = 0.01999; %inlet viscosity in cp u2 = 0.02118; %outlet viscosity in cp pf = 0.4125; %density of feed in lbs per feet3 u = (u1+u2)/2; V = 1.2*mass/70; for i=1:600; L = i*0.1; D = 2*((V/(pi*L))^0.5); AR = L/D; area = pi*(D/2)^2; G = flowrate/area;

Production of Hydrogen via Syngas Route

4-38

CN 4120: Design II Team 32: Ong Song Kun (U046829M)

HTS Unit Design Report

Dp = 0.25/12; Nre = Dp*G/(2.42*u); fk = 1.75+150*(0.555)/Nre; dPft = L*((fk*G^2/(Dp*pf*32.17*(3600^2)))*0.555/(0.445^3)); dPsi = dPft/(12^2); dPbar = dPsi/14.7; x(i,1) x(i,2) x(i,3) x(i,4) x(i,5)

= = = = =

AR; dPft; dPsi; dPbar; G;

end plot(x(:,1),x(:,4));

Appendix 4.12.3 mass = 281609.79; % mass of catalyst in lbs AR= 3.497; %AR P = 454.2; %pressure in psi S = 13.1; %maximum allowable stress value in kips per inch square

V E D L

= = = =

mass*144*12/70; 1; (4*V/(pi*AR))^(1/3); D*AR;

t=P*(D/2)/(S*1000*E-0.6*P); D=D*0.0254; L=L*0.0254; fprintf ('t= %f inches \n',t); fprintf ('D= %f metre \n',D); fprintf ('L= %f metre \n',L);

Production of Hydrogen via Syngas Route

4-39

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Chapter 5 : LOW TEMPERATURE SHIFT REACTOR 5.1 INTRODUCTION

The water shift reaction usually occurs in an fixed bed adiabatic system with the presence of a catalyst to speed up the reaction rate. In an adiabatic system, CO slip is determined by the exit temperature of the shift converters, because low temperatures results in low equilibrium levels of CO, as the following exothermic process is taking place:

CO + H2O ⇌ CO2 + H2

∆H = -41.2kJ/mol

(5-1)

On the other hand, favorable kinetics occurs at higher temperatures. Either a high steam-togas ratio or low temperature can be used to improve CO conversion percentage, but that also correspondingly contribute to higher capital and operation cost. Hence there is a tradeoff between CO conversion percentage and costs.

Fig.1

Typical CO variation in high temperature and low temperature shift catalyst beds [Frank, 2003a]

Conversion in a single high-temperature-shift(HTS) converter is equilibrium limited. Since this reaction is exothermic, the rise in temperature as reaction proceeds will eventually not favor further reaction. This limitation can be overcome by employing a second converter, the low-temperature-shift (LTS) converter after the HTS converter. Usually an inter-bed cooling process is employed between the two converters to keep the reaction occurring at low temperature in the second converter. A knock-out drum is then employed to condense and remove all water prior to feeding into the pressure swing adsorption (PSA). This part of the design project presents detailed chemical engineering design of a LTS converter and the knock-out drum.

Production of Hydrogen via Syngas Route

5-1

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Fig.2 LTS converter in the HYSYS environment

5.2 LTS DESIGN CONSIDERATIONS 5.2.1 Current Status

Carbon monoxide exits the HTS converter with a molar fraction of 0.03(dry basis) at 420oC. The stream, with a molar flow of 11354kmol/h, is then cooled to bring the temperature down to 220oC before feeding into the LTS converter. Molar composition of the feed is illustrated in the following diagram.

Fig.3 Molar compositions of feed into LTS converter

The outlet composition was automatically generated using Hysys, using the rate equation associated with this reaction (and catalyst type). However, it should be noted due to its iterative nature, Hysys could not obtain a value closer to that of the exact situation than Matlab. Hence, there is still a need to carry out interations (based on the same rate equation) in the Matlab environment. A comparison between results calculated from both programs will be made in latter section.

Production of Hydrogen via Syngas Route

5-2

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.2.2 Kinetics of Low-Temperature Water-Gas-Shift (LTWGS)

Equation (5-1) may be represented in the following form

A+B ⇌ C+D

(5-2)

where A, B,C and D are CO, H2O, CO2 and H2 respectively Rase (1977) has come up with the following equation for application to the shift conversion:

(5-3)

Where Xi k

= the dimensionless concentration of component i (Ci/Cref) = rate constant = exp (12.88 -1855.6/T) for copper-zinc catalyst

K

= equilibrium constant = exp (-4.72 + 8640/T) for 760 ≤ T ≤ 1060

P

= pressure, atm

(-rco)

= rate, lb moles CO converted/(lb catalyst)(hr)

T

= temperature, K

yj

= mole fraction of component indicated

ρb

= catalyst bulk density, lb/cu ft

ψ

= activity factor for the copper-zinc catalyst

Copper-zinc catalyst ψ

= 0.86 + 0.14P for P ≤ 24.8 = 4.33 for P ≥ 24.8

Production of Hydrogen via Syngas Route

5-3

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.2.2.1 Assumption made for equation (5-3) :

This equation represents the activity level characteristic of mid-life of the catalyst. These rate constants have been expressed on the basis of a reasonable “lined-out” activity that the catalyst would maintain for a considerable time provided operating errors which cause deactivation do not occur. Multiplying the rate equation by ρb , we obtain the rate of reaction in units of moles of CO converted per unit volume of catalyst per second, and converting the units to S.I units, we obtain the following equation

(5-4) Where (5-5) The pre-exponential factor ko includes the diffusion effect as given by the catalyst manufacturer (Rase, 1977).

Production of Hydrogen via Syngas Route

5-4

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.2.3 LTS Catalyst

From (5.3), it is important to decide the catalyst which we are using for the LTS shift. Copper-Zinc Oxide supported on alumina will be taken as the catalyst for our design and its specifications will be used for the calculations.

Copper-Zinc oxide offers the thermodynamic advantage of a lower operating temperature for the exothermic reaction in eq. (5-1) Characteristics of the catalyst assumed for this design are as follows: Catalyst Type

Copper-Zinc Oxide supported on alumina o

Maximum Operating Temperature ( C) Tablet Size (in.)

260 - 288 ¼ x 1/8

Bulk Density (lb/cu ft)

90

Particle Density ($/cu ft)

155

Cost ($/cu ft)

75

Catalyst Poison

Sulfur and halogen compounds, as well as unsaturated carbons

Catalyst life

2-3 yr

Fig.4

Production of Hydrogen via Syngas Route

5-5

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.2.3.1 Characteristics of the industrial LTS catalyst

The low-temperature shift catalyst is usually a mixture of copper oxide and zinc oxide in a ratio between 1:1  1:2, with alumina added in place of some of the zinc oxide. In addition, promoters such as Cr2O3, MnO, or some metal oxides have been used. Chromium oxide has been used in place of alumina. Preparative procedures on the whole are more critical for the LTS catalyst than for the HTS catalyst.

5.2.3.2 Preparation

Preparative procedures are much more critical for the LTS catalyst as compared to the HTS catalyst. Coprecipitation of the metals as metal nitrates are carried out via pH adjustment with ammonium bicarbonate. The oxides formed in this way are intimately intermixed by this procedure, which is essential for high activity and stability. It has been suggested that ZnO in excess can protect the copper content from inadvertent sulfur poisoning. Aluminum oxide also serves as a stabilizer for the copper, preventing it from being sintered easily. Thus we can see that the manufacture of these LTS catalysts involve great skill and refined proprietary techniques.

Since the LTS catalyst is pyrophoric, it must be sequested during system shutdown when only air flows through the system.

5.2.3.3 Supply

Catalyst suppliers usually offer thorough instructions for the start-up, catalyst reduction, operation and shutdown for the particular catalysts purchased. Instructions for catalyst reduction and start-up are particularly critical, since excessive temperatures must be avoided.

Production of Hydrogen via Syngas Route

5-6

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

One of the many suppliers for the LTS catalyst is Haldor Topsoe. Listed below are properties of an example of the LTS catalyst produced from this company.

http://www.topsoe.com/ 5.2.3.4 Deactivation of LTS Catalyst (I) Poisons

Common poison of the LTS catalyst are sulfur and chlorine compounds. Sulfur compounds such as H2S are removed in the ZnO adsorber beds prior to feeding into the steam-methane reformer. However there could exist times of upset such as short-periods of high-sulfur feed. In such instance, break-through sulfur will occur and pass to the HTS converter. There is a high possibility the HTS catalyst will be able to safely adsorb the H2S and protect the LTS bed. In some circumstances, sulfur may still get into the LTS. It is for this reason that LTS catalysts contain excess ZnO so that upper portion of the bed can serve as a sulfur guard. Zinc sulfide forms for the early part of the bed but further down the bed, sulfur in the form of H2S is chemisorbed, the extent of this happening depends on the operating operations. Consequently, deactivation of actives sites will take place. Chloride compounds are a major and permanent poison of LTS catalyst, and worth a mention despite the fact that no chloride compounds are involved in this design project.

Production of Hydrogen via Syngas Route

5-7

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

CuCl and ZnCl2 are formed and they can cause copper crystal growth (sintering) and significant loss of catalyst activity. Ways to tackle the problem of choride poisoning involve installing a bed of chlorine adsorbent (e.g. CaO/ZnO or alkalized alumina) upstream of the ZnO adsorbent bed prior to the reformer, another bed is placed above the LTS catalyst, composed of a chlorine adsorbent as well. (II) Sintering

Excessive temperature can result in sintering. Very small crystallites of copper are thermodynamically favored to coalesce into large crystals and thus produce a less active catalyst due to low porosity overall. Despite the fact that these crystallites are stabilized by the associated ZnO and also alumina, this protection is destroyed at elevated temperatures. Inlet operating temperatures for LTS between 175-275oC have been suggested, but it is always encouraged to operate at the lowest temperature possible, since sintering is a phenomenon related to both time and temperature. However, there is a lower temperature limit for the operating condition in the LTS converter, to avoid the any condensation of the steam we use in the low temperature shift reaction, as any condensation in the pores can result in catalyst damage. It is often suggested that the lowest temperature should be no lower than 20oC above the dewpoint.

5.2.3.5 LTS catalyst in operation

A common practice of some hydrogen-producing companies is to increase temperature during a LTS operating cycle to overcome deactivation of the catalyst. This will however, increase the growth of the crystals and shorten the life of the catalyst. In the lower regions of temperature, sintering rate is very low but this increases as temperature is raised. Ultimately, deactivation rate becomes significant and the catalyst activity will suffer. The higher the bed temperature reached, the more critical temperature control becomes, particularly if the process gas is introduced after the bed reaches operating temperature. Sudden rapid rise in temperature can damage the catalyst.

Reducing gas of H2 mixed with N2 or with natural gas is usually recommended for LTS catalyst to keep the catalyst in the reduced form.

Production of Hydrogen via Syngas Route

5-8

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.2.3.6 Assumptions made for LTS Catalyst

The following assumptions have been made for the derivations of the differential equations which characterize the pellet mass and heat balance 1. The Copper Zinc catalyst pellet particles have a homogenous porous structure 2. Mass transfer within the catalyst particles occurs by diffusion only which may be expressed by a constant effective diffusion coefficient D, and rate of intraparticle diffusion is described by Fick’s Law 3. Conduction is responsible for the thermal transfer within catalyst particles and effective thermal conductivity λe is used with the Fourier’s law, to describe the intraparticle heat conduction 4. Both mass and heat transfer within the catalyst pellets only take place in the radial direction

5.2.3.7 Mass balance on the Copper-Zinc catalyst pellet

Dimensionless steady state material balance for component I over a shell of dimensionless thickness dw ( where w is the dimensionless radial coordinate, z/Rp) is given by:

(5-6) where (5-7)

φi

is Thiele’s modulus of the pellet, for component i and is defined as

(5-8) and (5-9)

Production of Hydrogen via Syngas Route

5-9

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

γ is the dimensionless activation energy, defined as (5-10)

Equation (5-10) is a second order differential equation of the boundary value type having two split boundary conditions: At ω = 0 (5-11) At ω = 1

(5-12) where i = A,B,C,D

5.2.3.8 Heat balance on the Copper-Zinc catalyst pellet

The dimensionless steady state enthalpy balance over the shell of dimensionless thickness dω is given by (5-13) Where βi is the thermicity factor of the pellet based on component I, defined as

(5-14)

The boundary conditions being: At ω = 0 (5-15) At ω = 1 (5-16)

Production of Hydrogen via Syngas Route

5-10

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

The non-isothermal effectiveness factor for a spherical particle the non-isothermal effectiveness factor η is defined as:

(5-17)

In dimensionless form the above equation becomes

(5-18)

It can also be written as

(5-19)

Production of Hydrogen via Syngas Route

5-11

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

5.2.4

LTS Unit Design Report

Modeling the converter

5.2.4.1 Assumptions made for the converter

The heterogeneous model is developed in terms of bulk variables with the effectiveness factor introduced to account for the diffusional limitations. The assumptions made for the overall reactor model are as follows: 1. There is uniform distribution of gas flow velocity inside the converter 2. The reactor is studied under steady state conditions 3. The radial distribution of the temperature and concentration of the different

components inside the converter is uniform, i.e. the model is one-dimensional 4. Heating and mass diffusion in the longitudinal direction are negligible considering the

very high gas velocities at which the reactor is operated, i.e. axial dispersion is negligible 5. The pressure drop across the reactor is negligible compared with the total pressure of

the reactor

5.2.4.2 Reactor mass balance

For the bulk gas phase, the rate of reaction is formulated in terms of the mole fractions Yi instead of the dimensionless concentrations Xi (as in the catalyst pellets equations). This is a more convenient approach as the total number of moles is constant, while the volumetric gas flowrate is changing due to the change of temperature.

At steady state, a component mass balance on CO over an element of catalyst bed of thickness dl and a cross sectional area Ai , with a constant total molar flow rate nT , gives

(5-20) where nA is the molar flow rate of component A, and the rate of reaction is given by (Rase, 1977) (5-21)

Production of Hydrogen via Syngas Route

5-12

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

In a dimensionless form, the equation becomes (5-22) Where z’ = I/L and

KB is the temperature dependent equilibrium constant which is defined as (Borgars and Campbell, 1974):

(5-23)

where yB = TB/Tref and (CO2), (H2), (CO), (H2O) are the partial pressures or fugacities for the different species in equilibrium. The boundary condition at the inlet of the reactor is at z’=0 (5-24) 5.2.4.3 Reactor mass balance At steady state, the heat balance equation is

(5-25) In a dimensionless form, the equation will be (5-26) where (5-27) The boundary condition is at z’= 0 (5-28) The bulk phase temperature can also be computed from the bulk phase concentration by making a cumulative heat balance over any reactor length (5-29)

Production of Hydrogen via Syngas Route

5-13

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Therefore, it is sufficient to integrate (5-22). There is no need to integrate equation (5-27) along the length of the reactor, equation (5-29) can be used instead. From stoichiometry, at any depth of the reactor, the concentration of CO2, H2O, H2 can be expressed in terms of the bulk concentration as follows:

(5-30) Where I = B, C, D and a =-1 for reactants and a = +1 for products.

5.2.4.4 Transport parameters Viscosity and thermal conductivity

The viscosity µ of the fraction, designated by subscript x at a density ρ and temperature T is given in terms of a reference fluid, designated by subscript o. The equation is (5-31) where (5-32) With M as the molecular weight and To, ρο defined by the ratios (5-33) and (5-34)

fx,o and hx,o are scaling ratios, which are in general (5-35) and (5-36)

Production of Hydrogen via Syngas Route

5-14

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

The subscript c denotes the critical value and the superscript * denotes reduction of the variables by the critical value. The functions θ and φ are the shape factors expressed in terms of Pitzer acentric factor, ω , via functions of the form (5-37) and (5-38) where F and G are universal functions reported for example, by Leach et al. (1968) and Ely and Hanley (1981). The thermal conductivity λ is also evaluated through the same procedure (Ely and Hanley, 1981)

Prandtl number

The prandtl unmber Pr is computed as

(5-39) Diffusion coefficients

The binary diffusion coefficient of each component is computed by the relation

(5-40) where vi are the values of the atomic and structural diffusion-volume coefficients (Perry et al,1984) The value of the diffusion of each component in the mixture is calculated by the relation (Bird et al. 1960) (5-41) where Yi = mole fraction of each component, Dij = the binary diffusion coefficient, and Di,mix = diffusion coefficient for each component in the mixture.

Production of Hydrogen via Syngas Route

5-15

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

External mass and heat transfer coefficients

Correlations for both mass and heat transfer coefficients kg and h are found empirically from the mass and heat transfer J-factor (JD and JH) correlations, which are defined as (5-42) and

(5-43) The values for JD and JH are almost equal and are computed as a function of the Reynolds number:

JD = JH = 0.989 Re-0.41 for Re>350 = 1.820 Re-0.51

for Re<350

(5-44)

where, Re = Reynolds number = Gdp / µ

The external mass transfer in the pellet equation takes the form (5-45) and the external mass transfer coefficient is calculated by the relation

(5-46)

where RG= the universal gas constant and T = average temperature for the catalyst bed.

Production of Hydrogen via Syngas Route

5-16

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3 DESIGN CONDITIONS 5.3.1 Temperature

Inlet temperature has been set at 220oC, a temperature which is low enough to prevent the sintering of the LTS catalyst, and high enough to prevent any condensation of steam taking place, damaging the catalyst as a result. Referring to (5-3), as pointed out by Rase, it can shown that rate doubles for a rise of 200oC for the LTS catalyst. By using around 232oC as an approximate maximum for design, we can have the opportunity to raise the temperature to compensate for a 50% loss in activity for the LTS catalyst.

The rate data in (5-3) is based on activities of the catalyst at mid life, so it is expected that the unit will perform better than the design at the outset. Designed outlet CO values can be kept by adopting the strategy of increasing temperature over a long period of time. This is done in view of the fact that activity of catalyst will decline towards the last half of its life. Increasing temperature further however, can result in sintering of the catalyst sintering, leading to lower activity of the catalyst. Hence, increasing temperature in such case does not remedy the situation at all.

5.3.2 Pressure

The LTS converter will be operating at the pressure of 24.86 bar. Much care has to be taken to prevent a high pressure drop across the LTS converter. This is to facilitate a high enough pressure for the Pressure Swing Adsorption unit that follows the knock-out drum.

5.3.3 Steam to CO ratio

As suggested by Rase (1977), Steam to CO ratio should be in the range of 4:1 and above. Economic analysis on design studies has to be done before the optimum value can be obtained. The disadvantages with using a higher steam rate will be that of higher flow rates and larger diameter equipment.

Production of Hydrogen via Syngas Route

5-17

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

For the HTS unit, inlet’s molar flow of H2O and CO is as follows: Species

Molar flow (kmol/hr)

H 2O

6326.9

CO

1265.6

The Steam to CO ratio employed for the HTS case is (6326.9/1265.6) = 5

For the LTS unit, inlet’s molar flow of H2O and CO is as follows: Species

Molar flow (kmol/hr)

H 2O

5374.4

CO

313.1

The Steam to CO ratio for this case is (5374.4/313.1) = 17.2

Explanation and justification for the high ratio used for LTS:

Both CO and H2O react to form CO2 and H2. So it is understood that both the molar flow of CO and H2O will decrease once they go through the first water-gas-shift reaction in the LTS converter. However, CO gets converted to a much greater extent (from 1265.6kmol/h to 313.1kmol/h) as compared to H2O. Given a much greater change in the denominator of (H2O/CO), we can hence expect an inevitable high Steam to CO ratio for the inlet feed of the LTS converter. Drawing steam from the feed to the LTS converter just to meet a H2O:CO ratio of around 4:1 will be an unwise move as this will give rise to greater need for extra units prior to feeding into the LTS converter to condense the steam and then remove the liquid phase. Higher steam does have the advantage of the driving the water-gas-shift equation to the right, producing more of the products. However as mentioned, the drawback of higher steam content will be bigger equipment needed to accommodate a higher volumetric flow.

Production of Hydrogen via Syngas Route

5-18

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3.4 Design Procedure for LTS outlet compositions and Mass of Catalyst used

The design procedure suggested by Rase(1977) has been adopted and is as follows:

Component Mole Balance

∆W(-rCO) = (-∆FCO)

(5-47)

FCOi+1 = FCOi – (-∆FCO)

(5-48)

FCO2i+1 = FCO2i + ∆FCO2

(5-49)

Where i designates increment number

Heat Balance

Heat of reaction is based on the known inlet temperature of the increment ∑FjCpj(Ti+1 – Ti) = (-rCO)(- ∆HCO)Ti∆W = (-∆FCO)(- ∆HCO)Ti

(5-50)

Algorithm principle Basis: An ∆W of 200lb can never cause a ∆T greater than 1o, as pointed out by Rase(1977).

An increment size of 1o is then selected and the change in molar quantity (∆X) of a specie i.e CO can be calculated for each increment. Algorithm is run with using n steps(cycles), the number of steps required to reach target amount of CO in the outlet. With every increment step involving 200lb of catalyst, total amount of catalyst can then calculated using n x 200lbs The following algorithm has been set up: 1. Calculate (-rCO) at inlet conditions to increment, i. 2. Calculate (-rCO)avg = (-rCO)i + [(-rCO)i – (-rCO)i-1]/2 (skip for i = 0) 3. Calculate new flow rates: Fi+1 = Fi ± ( −ro) ∆W 4. Calculate Cp and (-∆HCO) @ Ti 5. Calculate ∆T from (5-50) 6. Ti+1 = Ti + ∆T 7. yi+1 = Fi+1/(FT)i The amount of catalyst, including outlet compositions have been worked out with this algorithm in MATLAB. Relevant codes for this portion have been attached in Appendix A.

Production of Hydrogen via Syngas Route

5-19

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3.5 LTS outlet compositions and Mass of Catalysts used

Results of our previous algorithm are as follows:

CO H2O CO2 H2 C2H6 CH4 N2

Mole Fraction

Outlet kmol/hr

0.00338 0.32347 0.12660 0.51192 0.00066 0.03200 0.00198

53.41 5114.12 2001.53 8093.44 10.40 505.87 31.22

Mass flowrate in kmol/hr Mass of LTS catalyst used(kg) Outlet T (celsius)

15810 34720 239.3

The water shift reaction occurring in the low temperature reactor changes the composition of the syngas species and temperature of the syngas. The CO conversion efficiency ξcan be used to show how much CO is converted into CO2 in the LTS converter.

(4-51)

ξfor our case = (313.10 – 53.41)/313.10 = 82.9% Mol fraction of CO in dry basis = 0.5%

Fig.5 Molar compositions of inlet and outlet feed

Production of Hydrogen via Syngas Route

5-20

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Comparison of HYSYS and MATLAB results for outlet feed molar compositions Species

HYSYS

MATLAB

CO

0.003390

0.003378

H 2O

0.323482

0.323474

CO2

0.126590

0.126599

H2

0.511910

0.511919

N2

0.000658

0.000658

CH4

0.031999

0.031997

C2H6

0.001980

0.001975

As we can see from the table above, the molar compositions for HYSYS agree well with

MATLAB‘s. Conversion value of 82.9% for CO, previously calculated in the MATLAB environment, is inputted as part of the specifications in the Conversion Reactor module under

HYSYS.

5.3.6 Design Procedure for Aspect Ratio

As an initial approach to calculating the dimension of the catalyst bed, the aspect ratio is first derived. The aspect ratio is defined as follows: Aspect Ratio (AR) = (Length of Bed/Diameter of bed)

(4-52)

The following equations are used:

(4-53)

(4-54)

(4-55)

(4-56)

Production of Hydrogen via Syngas Route

5-21

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

(4-57) The MATLAB codes for this portion are located under Appendix A2.

Algorithm principle

For every loop in the algorithm, there is a small increment of bed diameter accompanied by a very small increment in bed height involved. Both variables are linked based on the volume of the catalyst used for the design. Corresponding pressure drop, followed by the aspect ratio are generated for every cycle. In this case, the controlled term is pressure drop, which we have fixed to be at 1 bar max. Generating a graph displaying Aspect Ratio against Pressure-Drop, we read off the aspect ratio which corresponds to a pressure of 1 bar-drop.

Justification for controlling pressure drop at 1 bar:

Industrially the pressure drop across the LTS converter is no more than 1 bar. A large pressure drop is uneconomical process-wise due to the usage of a pressure-swing adsorption unit in the downstream. Furthermore, too high a pressure drop can damage the expensive catalyst used for LTS.

5.3.7 Results for Aspect Ratio

Fig.6 Graphs of Aspect Ratio against Pressure Drop

Production of Hydrogen via Syngas Route

5-22

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

The aspect ratio which corresponds to a pressure drop of 1 bar is 1.2287. Using this aspect ratio, the length and height of the catalyst bed can be determined in the next section

5.3.8 Design Procedure for the dimensions of bed and thickness of vessel wall

The MATLAB codes for this portion are located under Appendix A3. Formulas used are as follows: Volume of bed = (Mass of catalyst used * /90)

(4-58)

By using ‘90’, the bulk density of the LTS catalyst, Equation (4-58) has already taken in account space occupied by the voids in the catalyst packing Diameter of bed = (4 * Volume/( π * Aspect ratio))1/3

(4-59)

Length of bed = Diameter of bed * Aspect ratio

(4-60)

Thickness of vessel wall = Pressure(psi) * (Diameter/2)/(Allowable stress value*1000 – 0.6*Pressure) (4-61)

Allowable stress value is based on the material that has been used for the construction of vessel. For this design, low-alloy steel A387 Gr.12 has been used for the construction of the LTS converter given its resistance to hydrogen which is existent in the process stream. The allowable stress value for this material is given as 13.8 Kips/in2

Production of Hydrogen via Syngas Route

5-23

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3.9 Results for bed dimensions and wall thickness Bulk Volume of Catalyst (m3)

848.71

Diameter of Catalyst Bed (m)

2.92

Length of Catalyst Bed (m)

3.59

Wall thickness (in)

0.13

For this design, a guard bed will not be considered in view of the fact there is no sulfur content at all in our process streams. Even if there is, I am making the assumption in this case that all sulfur poisons will be adsorbed on the catalyst in the HTS converter, hence there is no chance for any sulfur contents to come into contact with the LTS converter.

In reality however, due to the high sensitivity of the low temperature catalyst to sulfur poisoning, a guard bed containing zinc oxide or other guard solids is usually positioned on top of the catalyst bed. The guideline given by Rase(1977) is that the height for a good distribution of the catalyst will be at 100 times the diameter of the catalyst particle.

Production of Hydrogen via Syngas Route

5-24

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3.10 Allowances set for design

•

Temperature

As highlighted in a previous section, I have set a design temperature of 232oC suggested by Rase(1977), approximately 12oC higher than the intended value of 220oC, with the assumption that this will be able to compensate for 50% loss in activity for the catalyst •

Amount of catalyst used

The amount of catalyst used will be designed with 10% extra. No references have been found addressing this portion on the ‘extra amount of catalyst’ to use. I have chosen a reasonable figure of 10%-extra, with the priority of bringing outlet CO concentration down to the desired level of less than 0.5% (dry basis) in mind. Using excessive catalyst give rise to higher catalyst cost as well as a larger reactor to accommodate the larger volume. New catalyst amount = 34720 * 110% = 38192kg •

Size of catalyst bed and wall thickness

Incorporating the new catalyst amount into the calculations, corresponding dimensions for the new catalyst bed are 3.01m(diameter) * 3.70m(length) Again, a 10% extra allowance is made for the size of bed and wall thickness. Before

After

Bed Diameter (m)

3.01

3.31

Bed Length

3.70

4.07

0.13

0.14

(m)

Wall Thickness (in)

Any space that is not taken up by the catalyst, in case of an over-design, can be easily filled up by the inert ceramic balls which are used to support the catalyst.

Production of Hydrogen via Syngas Route

5-25

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.3.11 Study of controlling parameters Pressure drop vs bed length

It is observed pressure drop increases exponentially with an increase in bed length. As mentioned in a previous section, maximum pressure drop has been controlled at a value of 1 bar since industrially pressure drop across the LTS converters are usually not more than 1 bar. It is evident from this graph that increasing bed length can result in an increasing increment of pressure drop each time. This means that an infinitely long catalytic reactor can result in a plunge of pressure drop to atmospheric pressure. Hence there is a need to control the total amount of catalyst use as more catalyst will call for the need of larger beds.

Production of Hydrogen via Syngas Route

5-26

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.4 CHOICE OF A REACTOR BED

Fig.7 Single Adiabatic bed incorporating measurements for bed size

Since the water-gas-shift is only moderately exothermic, the single adiabatic bed will be a suitable reactor for performing the low temperature shift. The feed will flow downwards from the top through a bed of the LTS catalyst packed in between layers of inert ceramic support balls, which serve in creating even flow distribution over the entire cross section and for separating solid contaminants that may have entered the feed. The ceramic balls in the lower region act as a support while the ones at the top region prevent movement of catalyst particles by high-velocity gas. Alternatively, a metal grid with holes sized somewhat smaller than this catalyst can be used in place of the ceramic balls. Rase has pointed out that in cases where flow rates are large and pressure drop must be minimized, large diameter short beds will come into the picture. This seems to be the case for our design.

Production of Hydrogen via Syngas Route

5-27

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.4.1 Cost estimation for the LTS converter Material Cost for Converter

Calculation of the amount of material i.e. low alloy steel A387 has been simplified to obtain an approximate value. Assuming two half-spherical ends (at both sides) with diameter 3.31m, total surface area of the reactor can be calculated as follows: Surface area of reactor

= π D * L + 4π(D/2)2 = π ∗ 3.31 * 4.07 + 4π(3.31/2)2 = 76.74 m2

Given a design thickness of 0.14in (0.0035m), total volume of material required = 76.74 * 0.0035 = 0.26859 m3 Density of the steel is around 8g/cm3 = 8000kg/m3 Hence, mass of the steel required

= 0.26859 * 8000 = 2148.72kg

Cost of low alloy steel (Cr-Mo) is quoted as around USD$700 – 850/T, as stated in the book by Coulson and Richardson. For design purpose, we will take the higher value of USD$850/T. Hence cost of the steel required = USD$850 * (2148.72/1000) = USD$1826.41 This calculation fails to take into account material required for additional features of the reactor, e.g. the flanges. Hence the amount of steel, and consequently the cost of steel used could have been much more. An important assumption made in the calculation of this material cost is an uniform wall thickness of 0.14inch for the reactor. This may not hold in reality as thickness can be slightly larger in areas where a greater stress is experienced i.e. bottom of reactor to handle the weight of the catalyst and inert ceramic balls. Cost of the inert ceramic balls will not be included in the cost calculation. Due to the inavailability of its cost info online, another assumption is made: Cost of the ceramic balls is assumed a total cost of not more than USD$500. In my view, this is a fair assumption as it is well known that the LTS catalyst accounts for a significantly large portion of the LTS converter’s total cost.

Production of Hydrogen via Syngas Route

5-28

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Catalyst cost for converter

Copper-Zinc oxide has been quoted by Rase(1977) to be approximately USD$75/ft3. Since total bulk volume of the catalyst in this case is (84022/90) = 936ft3, Cost of the catalyst

= 936 * 75 = USD$70,200

Factoring in the inflation rate

It ought to be noted that the cost figure quoted by Rase for Copper-Zinc Oxide was 30 years. Over these 30 years, an increase in price has in fact taken place. The same goes for the price of the low alloy steel A387 which was last quoted back in 2005. Hence the cost for both the catalyst and the steel required has to be adjusted to get us a better approximation to the actual situation. The Consumer Price Index could be a good reference to account for the amount inflation to be factored in.

Fig.8 CPI for USA from 1913-2006

An index of 100 (base case) has been assigned to the year of 1980. From year 1977 to 2008, it is approximately a jump from index 65 to 210 in 2008 (Value for 2008 has been approximated from the trend in the graph). From 2005 to 2008, it is probably an increment of 10 index, from 190 to 210. The adjusted total cost of the reactor is as follows: Cost of LTS Converter(adjusted)

= USD$70,200 * (210/65) + 1826.41 * (210/190) + 500 = USD$229,318

Production of Hydrogen via Syngas Route

5-29

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

5.5 DESIGN OF THE KNOCK-OUT DRUM

Fig.9 Knock-out drum in the HYSYS environment

5.5.1 Working principle of the knock-out drum The knock-out drum is basically a vapor-liquid separator drum which is in the form of a

vertical vessel into which a liquid and vapor mixture (or a flashing liquid) is fed. Within the drum, the liquid is falls to the bottom of the vessel because of gravity and hence a separation is achieved, the liquid phase(water for this case) is then withdrawn. The vapor, on the other hand, travels upward at a design velocity which minimizes the entrainment of any liquid droplets in the vapor as it exits the top of the vessel.

5.5.2 Sizing of the knock-out drum The size of the knock-out drum will be dictated by the anticipated flow rate of vapor and

liquid from the drum. Based on the assumption that those flow rates are known, the following sizing methodology is proposed: (1) A vertical pressure vessel with a length-to-diameter ratio of about 3 to 4 is used, vessel is sized to provide about 5 minutes of liquid inventory between the normal liquid level and the bottom of the vessel (with the normal liquid level being at about the vessel's half-full level)

Production of Hydrogen via Syngas Route

5-30

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

(2) The vessel diameter is calculated using the Souders-Brown equation to determine the maximum allowable vapor velocity:

(5-62) Where V=maximum allowable vapor velocity, m/s ρL= liquid density, kg/m³ ρV= vapor density, kg/m³ k= 0.107 m/s (when the drum includes a de-entraining mesh pad

Then the cross-sectional area of the drum (A) is obtained from: A, in m² = (vapor flow rate, in m³/s) ÷ (vapor velocity V, in m/s)

(5-63)

And the drum diameter (D) is: D, in m = [(4) (A) ÷ (3.1416) ] 0. 5

(5-64)

5.5.3 Results and cost estimation

The following values have been obtained from HYSYS ρL = 993.3 kg/m3 ρV = 9.92kg/m3

Vapor mass rate = 32.17kg/s Vapor volumetric rate = 32.17/9.92 = 3.24m3/s V

= (0.107) ((993.3-9.92)/9.92))1/2 = 1.07m/s

Cross sectional area (A) of drum

= 3.24/1.07 = 3.03

Production of Hydrogen via Syngas Route

5-31

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

Diameter (D) of drum

Height of drum

=

[(4 * 3.03)/3.1416]0.5

=

1.96m

=

1.96 * 4

=

7.84m

LTS Unit Design Report

Treating shape the knock-out drum as a cylinder with a thickness similar to that of LTS converter, Surface area of the knock out drum = 2 * π (1.96/2)2 + π*1.96*7.84 = 54.49m2 Low-alloy steel A387 Gr.12 has been used for the construction of the knock-out drum

Assuming a thickness of 0.14in, Volume of A387 used = 54.49 * 0.025 * 0.14 = 0.19 m3 Mass of A387 required = 0.19 * 8000 = 1520kg Cost of the A387 = USD$850 * (1520/1000) = USD$1292 Cost of the A387(adjusted by CPI) = USD$1292 * (210/190) = USD$1428 Cost of the De-entrainment mesh pad, inlet diffuser and liquid level control valve have to be included as well. Due to the inavailability of their price, a value not more than US$2000 is assumed in this case for the total cost of these equipment. Hence a total of USD$(1428+2000) = USD$3428 is approximated for the knockout drum.

Production of Hydrogen via Syngas Route

5-32

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

LITERATURE REVIEW Developments in Water Gas Shift

Catalysis using Fe-Cr and CuO/ZnO for the HTS and LTS converters has not changed much for the past 40 years, but continue to be improved marginally. The efforts for improving the water gas shift reactor performance have been focused on a wide spectrum of subjects. This involves the modification of the conventionally used catalysts improve overall activity and stability. In particular, researchers have been making attempts at in developing catalysts which are more tolerant to sulfur contents. In comparison with the conventional copper-zinc water-gas-shift catalyst, researchers1 have reported that noble metal (such as gold, silver, platinum, palladium and rhodium) catalysts have the advantage of high activity and eliminating the self-heating issue. LangmuirHinshelwood (LH) kinetics have been used to derive the rate equations involving the use of these newly-developed metal catalysts. A good fit with the experimental data has been obtained and it is suggested that the LH kinetic model could be a suitable one for application in catalysis of WGS by other metals. In other papers2, theoretical studies have been performed to formulate new kinetic expressions for the novel catalysts, with the goal of a better control over the water-gas-shift’s reaction rate and conversions. Still, up to now researchers have been trying to optimize the industrial water-gas-shift reaction by tuning the various parameters involved in the process. Model-based reactor optimization has been employed for various reactor configurations such as microreactors3, monolith reactors4 and membrane reactors5 with the objective to achieve an overall reduced volume for the reactors or a better energy-integration within the system. Furthermore, various reactor configurations are analyzed in order to find out the limiting values of the main design variables. In a work by Javier & Co-workers6, it is found that insulating material of the reactor plays a major role in the shift converters in the sense there exists an optimal thickness of the insulator that affects the final volume of the reactor as well as other design variables. Such results from this study will be useful for estimating the minimum and relative sizes that allows conventional reactor technology. 1 Jian Sun, Joel DesJardins, John Buglass, Ke Liu “Noble metalwater gas shift catalysis: Kinetics study and reactor design 2 M. Levent, Int. J. Hydrogen Energy 26 (2001) 551–558 3. G. Kim, J.R. Mayor, J. Ni, Chem. Eng. J. 110 (2005) 1–10. 4. A.S. Quiney, G. Germani, Y. Schuurman, J. Power Sources 160 (2006) 5. A. Brunetti, Barbieri, E. Drioli, K.-H. Lee, B. Sea, D.-W. Lee, Chem. Eng.Process. 46 (2007) 119–126

6 Javier A. Francesconi, Miguel C. Mussati, Pio A. Aguirre “Analysis of design variables for water-gas-shift reactors

Production of Hydrogen via Syngas Route

5-33

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

Recently a new application for the water gas shift is in the reforming systems for fuel cells. Fuel cell development has seen remarkable progress in the past decades because of an increasing need for enhanced energy conversion efficiency and because of serious concerns about the environmental consequences of using fossil fuels for electricity production. The water gas shift reaction in fuel cell application has been studied extensively to obtain highly accurate kinetics expressions in order to create a tool for an integrated and optimized simulation of a whole fuel processing system. In addition, the new application of hydrogen gas as a raw material for fuel cells for mobile power sources (PEM fuel cells) requires that the anode inlet gas have a CO concentration lower than 10-20ppm. Otherwise, the anode is poisoned and the cell efficiency will drop abruptly. This explains why a water gas shift has to be employed not just to produce the hydrogen fuel but also to reduce the CO concentration. In a work by Zalc7, simulation of a fixed bed reactor was carried out, and the water gas shift reaction forms part of a purification train for a 10kW PEM fuel cell. In that work, a commercial Cu/Zn/Al2O3 catalyst doped with Ba was used because it showed a higher activity than the traditional one. A one-dimensional heterogeneous model was applied in that simulation, and a parametric sensitivity analysis was carried out for some of the process variables, with the purpose of finding criteria to minimize the reactor volume.

As we can see from the examples above, extensive research has been made into the water gas shift reaction, with a significant movement towards developing the fuel cell technology. The water gas shift reaction has played a significant role for the last 60 years at least, conventionally employed for the production of hydrogen gas for ammonia synthesis. And it can very well be one of the most important reactions for the 21st century, especially when fuel cell technology has overtaken all other forms of energy production to claim leader in fuelling our future. At present, all research made into the water gas shift is definitely a worthwhile investment.

7 J. Zalc, V. Sokolovskii, D. Loffler, J. Catal. 206 (2002) 169–171

Production of Hydrogen via Syngas Route

5-34

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

CONCLUSION LTS converter specifications and cost summary CO H2O CO2 H2 N2 CH4 C2H6

Dimensions Mass of Catalyst used Inlet Temperature Outlet Temperature Total Cost

Inlet Composition Outlet Composition 0.0198 0.0034 0.3399 0.3235 0.1102 0.1266 0.4955 0.5119 0.0007 0.0007 0.0320 0.0320 0.0020 0.0020 Diameter

Height

3.31

4.07

34720kg 220 [C] 239.3 [C] USD$229,318

H2 yield at the exit of the LTS : 75.6mol%(Dry Basis) Knock-out drum specifications and cost summary Dimensions Total Cost

Diameter

Height

3.31

4.07

USD$3,428

In this report, the detailed design of a fixed bed catalytic reactor for the low temperature shift was presented. Based on the energy and mass balances, the fixed bed catalytic reactor was modeled in MATLAB. All specifications and costing are presented in the tables above. An study made on the pressure drop along the bed length shows the exponential relationship of increasing pressure drop with increasing bed length. Due to the inavailability of the price data for equipments such as the de-entrainment mesh pad, inlet diffuser, liquid level control valve and the inert ceramic balls, conservative price values have been assumed for these equipment and incorporated in the cost mode l of our design. It is important to take into account the assumptions that have been established for our modeling. Overall, design and operational optimizations have been completed for this work.

Production of Hydrogen via Syngas Route

5-35

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

BIBLIOGRAPHY 1. Chemical reactor design for process plants / Howard F. Rase 2. Fixed-bed reactor design and diagnostics : gas-phase reactions / Howard F. Rase 3. Handbook of commercial catalysts : heterogeneous catalysts / Howard F. Rase. 4. Perry's chemical engineers' handbook 5. Modelling, simulation, and optimization of industrial fixed bed catalytic reactors / S.S.E. H. Elnashaie and S.S. Elshishini. 6. http://en.wikipedia.org/wiki/Souders-Brown_equation

Production of Hydrogen via Syngas Route

5-36

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

APPENDIX A1

Matlab codes •

Calculation of Catalyst Weight and Outlet Compositions

T= input('Please enter initial temperature for reactor in K:'); F= input('Please enter flowrate for reactor in kmol per hour:'); yco= input('Please enter initial mol frac of CO:'); yh2o= input('Please enter initial mol frac of H2O:'); yco2= input('Please enter initial mol frac of CO2:'); yh2= input('Please enter initial mol frac of H2:'); yc2h6= input('Please enter initial mol frac of C2H6:'); ych4= input('Please enter initial mol frac of CH4:'); yn2= input('Please enter initial mol frac of N2:'); Fco= F*yco; Fh2o= F*yh2o; Fco2= F*yco2; Fh2= F*yh2; Fc2h6= F*yc2h6; Fch4= F*ych4; Fn2=F*yn2; k=exp(12.88-1855.6/T); K=exp(-4.72+4800/T); rate=4*k*(yco*yh2o-yco2*yh2/K)/(379*90); Fco1= Fco -rate*200; Fh2o1= Fh2o-rate*200; Fco21= Fco2 + rate*200; Fh21= Fh2 +rate*200; cpco= (3.376+(0.557/10^3)*T -(0.031*10^5)*T^(-2))*8.314; cpc2h6=(1.131+(19.225/10^3)*T-(5.561/10^6)*T^2)*8.314; cph2o=(3.47+(1.45/10^3)*T +(0.121*10^5)*T^(-2))*8.314; cpco2=(5.457+(1.045/10^3)*T -(1.157*10^5)*T^(-2))*8.314; cph2=(3.249+(0.422/10^3)*T +(0.083*10^5)*T^(-2))*8.314; cpn2=(3.28+(0.593/10^3)*T +(0.04*10^5)*T^(-2))*8.314; cpch4=(1.702+(9.081/10^3)*T -(2.164/10^6)*T^2)*8.314; deltaa = 5.457 + 3.249 - 3.376 - 3.470; deltab = (1.045 + 0.422 - 0.457 - 1.450) * 10^-3; deltad = (-1.157 + 0.083 - (-0.031) - 0.121) * 10^5; integral =(deltaa * 298.15 * ((T/298.15) - 1) + (deltab/2)*(298.15^2)*(((T/298.15)^2)-1) + (deltad/298.15)*(((T/298.15) - 1)/(T/298.15))); dHco = -41166 + 8.314 * integral; dT = ((rate * 200) * (dHco))/(Fco*cpco+Fh2o*cph2o+Fco2*cpco2+Fh2*cph2+Fc2h6*cpc2h6+Fn2*cpn2+Fch4*cpch4);

T1 = T + dT; yco1 = Fco1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh2o1= Fh2o1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yco21= Fco21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh21= Fh21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); ydryco= (yco1/(1-yh2o1)); i=1; while ydryco>0.005,

k=exp(12.88-1855.6/T); K=exp(-4.72+4800/T); rate=4*k*(yco*yh2o-yco2*yh2/K)/(379*90); k=exp(12.88-1855.6/T1);

Production of Hydrogen via Syngas Route

5-37

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

K=exp(-4.72+4800/T1); rate1=4*k*(yco1*yh2o1-yco21*yh21/K)/(379*90);

rateavg= (rate+rate1)/2; Fco = Fco1; Fh2o = Fh2o1; Fco2 = Fco21; Fh2= Fh21; Fco1= Fco -rateavg*200; Fh2o1= Fh2o-rateavg*200; Fco21= Fco2 + rateavg*200; Fh21= Fh2 +rateavg*200; cpco= (3.376+(0.557/10^3)*T1 -(0.031*10^5)*T1^(-2))*8.314; cpc2h6=(1.131+(19.225/10^3)*T1-(5.561/10^6)*T1^2)*8.314; cph2o=(3.47+(1.45/10^3)*T1 +(0.121*10^5)*T1^(-2))*8.314; cpco2=(5.457+(1.045/10^3)*T1 -(1.157*10^5)*T1^(-2))*8.314; cph2=(3.249+(0.422/10^3)*T1 +(0.083*10^5)*T1^(-2))*8.314; cpn2=(3.28+(0.593/10^3)*T1 +(0.04*10^5)*T1^(-2))*8.314; cpch4=(1.702+(9.081/10^3)*T1 -(2.164/10^6)*T1^2)*8.314; integral = (deltaa * 298.15 * ((T1/298.15) - 1) + deltab/2*298.15^2*((T1/298.15)^2-1) + deltad/298.15*(((T1/298.15) - 1)/(T1/298.15))); dT = ((rateavg * 200) * (dHco))/(Fco*cpco+Fh2o*cph2o+Fco2*cpco2+Fh2*cph2+Fc2h6*cpc2h6+Fn2*cpn2+Fch4*cpch4); yco = yco1; yh2o = yh2o1; yco2 = yco21; yh2 = yh21; yco1 = Fco1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh2o1= Fh2o1/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yco21= Fco21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yh21= Fh21/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); ydryco= yco1/(1-yh2o1);

x(i) = i*200/2.2; y(i) = T; a(i) = ydryco; T = T1; T1 = T + dT; i= i + 1; end yc2h6 = Fc2h6/(Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); ych4 = Fch4/ (Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); yn2 = Fn2/ (Fco+Fh2o+Fco2+Fh2+Fn2+Fch4+Fc2h6); Wt=i*200/2.2; fprintf('mass of catalyst is %f kg.\n',Wt); fprintf('outlet temperature of LTS is %f K.\n',T1); fprintf('Outlet CO mole ratio is %f \n',yco1); fprintf('Outlet H2O mole ratio is %f \n',yh2o1); fprintf('Outlet CO2 mole ratio is %f \n',yco21); fprintf('Outlet H2 mole ratio is %f \n',yh21); fprintf('Outlet C2H6 mole ratio is %f \n',yc2h6); fprintf('Outlet CH4 mole ratio is %f \n',ych4); fprintf('Outlet N2 mole ratio is %f \n',yn2); fprintf('Outlet CO amount is %f \n',Fco1); fprintf('Outlet H2O amount is %f mol.\n',Fh2o1); fprintf('Outlet CO2 amount is %f mol.\n',Fco21); fprintf('Outlet H2 amount is %f mol.\n',Fh21); fprintf('Outlet C2H6 amount is %f mol.\n',Fc2h6); fprintf('Outlet CH4 amount is %f mol.\n',Fch4); fprintf('Outlet N2 amount is %f mol.\n',Fn2); plot (x,y); plot (x,a);

Production of Hydrogen via Syngas Route

5-38

CN 4120: Design II Team 32: Tham Zhi Yong, Andrew (U046754W)

LTS Unit Design Report

APPENDIX A2 • Determination of Aspect Ratio mass = input('Please enter mass of catalyst in lbs:'); flowrate = input('Please enter flowrate in lbs per hour:'); u1 = input('Please enter input viscosity in cp:'); u2 = input('Please enter output viscosity in cp:'); pf = input('Please enter density of feed in lbs per feet3:'); u = (u1+u2)/2; V = mass/90; for i=1:600; hc = i*0.1; dc = 2*((V/(pi*hc))^0.5); AR = hc/dc; area = pi*(dc/2)^2; G = flowrate/area; Dp = 0.15/12; Nre = Dp*G/(2.42*u); fk = 1.75+150*(0.555)/Nre; dPft = hc*((fk*G^2/(Dp*pf*32.17*(3600^2)))*0.555/(0.445^3)); dPsi = dPft/(12^2); dPbar = dPsi/14.7;

% % % end

x(i,1) = AR; x(i,2) = dPft; x(i,3) = dPsi; x(i,4) = dPbar; x(i,5) = G; if dPsi == 14.7, fprintf ('AR = %f.\n',x(i,1)) end plot(x(:,4),x(:,1));

APPENDIX A3 • Determination of dimensions of catalyst bed and wall thickness mass = input('Please enter mass of catalyst in lbs:'); AR= input('Please enter AR:'); P = input('Please enter pressure in psi:'); S = input('Please enter maximum allowable stress value in kips per inch square:');

V E D L

= = = =

mass*144*12/90; 1; (4*V/(pi*AR))^(1/3); D*AR;

t=P*(D/2)/(S*1000*E-0.6*P); fprintf fprintf fprintf fprintf

('t= %f inches \n',t); ('Bulk Volume of catalyst = %f cubic feet \n',V); ('Diameter of Catalyst Bed = % f ft \n',D); ('Length of Catalyst Bed = %f ft \n',L);

Production of Hydrogen via Syngas Route

5-39

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Chapter 6 : PRESSURE SWING ABSORPTION 6.1 INTRODUCTION

Conventionally, the industrial purification of hydrogen from steam-methane reforming will involve the use of a CO2 scrubber with aqueous monoethanolamine (MEA) and a methanator to convert the unreacted CO to CH4 before recycling it back to the Steam-methane reformer as feed. However, this purification method gives rise to the problems of personnel safety and solvent disposal due to the toxicity of MEA [1].

Pressure Swing Adsorption (PSA) has been increasingly adopted as the preferred mode of purification in the production of oxygen and hydrogen due to the advantages it possessed over its rivals in terms of selectivity, throughput and efficiency. Compared to the CO2 scrubber (95 - 97% in product purity), the attainable product purity for a typical PSA system is 99.99% [2]. This can result in a possible improvement in the refinery downstream operating margin, since a purer treat gas to the hydrotreaters can minimize the effect of catalyst coking and deactivation, which translates to a lower operating cost. Furthermore, PSA would have eradicated the problem of solvent disposal since it is a solvent-free process. However, due to its high operating pressures of up to 20 bar, additional reinforcement of the process equipments is required which might lead to a higher initial capital cost.

In this report, the PSA configuration is incorporated in the design of a hydrogen plant that is situated in Singapore. The PSA’s feed is directed from the knockout drum that is downstream of the low temperature shift reactor (LTS) and the stream composition is approximately in the range of 75% H2, with the remainder comprising of CO2, CO, CH4 trace amounts of N2 and H2O.

Production of Hydrogen via Syngas Route

6-1

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.2 PROBLEM STATEMENT

In the interim report that was submitted earlier, a polybed PSA system of 7 – 10 beds was proposed to recover up to 85% of the hydrogen product at a purity of 99.9%. According to the Hysys simulation generated for the interim report, table 1 is a summary of the composition of the incoming stream from the knockout drum to the PSA:

Table 1: Hysys screenshot of the composition of feed entering PSA

Thus the objective of this report is to design a PSA system which is able to achieve a production of 1.25 x 109 m3 (STP)/yr at a product purity specification of 99.9%.

6.3 THEORETICAL BACKGROUND

6.3.1 Separation via adsorption

Adsorption is a process in which there is a selective transfer of solutes (adsorbates) in the fluid phase to the surface of solid particles (adsorbents) through the formation chemical bonds or electrostatic attractions [3]. Figure 1 illustrates the accumulation of adsorbates such as CO2 on the internal pores surfaces of adsorbents, such as activated carbon for which a highly porous structure is required to achieve a large surface area for adsorption per unit volume of adsorbents used.

Production of Hydrogen via Syngas Route

6-2

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Figure 1: Diagram on adsorption [4] 6.3.2 Pressure-Swing Adsorption (PSA)

A pressure-swing adsorption is a process that selectively separates certain gaseous components from a gas mixture by effecting a change in the system pressure. The gas components can be separated either via their molecular characteristics or affinity for an adsorbent material. Under high pressure, the adsorbates are selectively adsorbed onto the adsorbents, which is the crux of the whole separation process. The system pressure is subsequently lowered so as to effect the desorption of the adsorbates, thereby allowing the regeneration of the bed.

6.3.3 Skarstrom Cycle

The design of this report’s PSA system is based on the basic form of Skarstrom Cycle [5] which comprises of 2 beds and 4 basic steps:

•

Pressurization o Pressurization of the PSA unit increases the affinity of the adsorbate with the

absorbent bed which causes enrichment of the less selectively adsorbed species. This is done using the feed.

Production of Hydrogen via Syngas Route

6-3

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

•

PSA Unit Design Report

High Pressure Adsorption o The raffinate is recovered during this process until the absorbent bed is relatively

saturated. •

Blow-down o Depressurization of the PSA unit occurs at this step, which allows the adsorbed

species to desorb from the bed and be removed. •

Purge o The bed is purged with part of the raffinate from the HPA step of another bed.

This is to regenerate the bed and allow it to be used for another cycle.

6.3.4 Adsorbents

Figure 2: Picture of activated carbon

Activated carbon, is a general term that includes carbon material derived mostly from charcoal. Possessing an exceptionally high surface area due to a high degree of microporosity, it is used in the PSA process for the adsorption of CO2 and CH4.

Production of Hydrogen via Syngas Route

6-4

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Zeolites are hydrated aluminosilicate minerals commonly referred to as "molecular sieves". The term molecular sieve refers to a particular property of these materials with the ability to selectively sort molecules based primarily on a size exclusion process. This is due to a very regular pore structure of molecular dimensions. Zeolites are considered crystalline while activated carbon is amorphous. Zelites are used in PSA design to remove N2 and CO from the effluent gas.

6.4 DESIGN CONSIDERATIONS

Being a dynamic process, the model for PSA involves the evaluation of nonlinear partial differential equations. The high level coupling of the model parameters further adds to the complexity of the whole simulation process. Thus before the commencement of the simulation process, an appropriate model should be selected and certain assumptions would have to be adopted so as to reduce the modeling complexity.

The assumptions made were as follows [5]: 1. An isothermal system is assumed. 2. There is negligible frictional pressure drop along the bed length. 3. Mass transfer between the gas and the adsorbed phases is accounted for all the steps and is sufficiently represented by the Linear Driving Force (LDF) model. During the adsorption and the purge steps, the total pressure in the bed remains constant while it varies exponentially with time during the blow-down and pressurization. 4. Fluid velocity in the bed varies along the length of the column, as determined by the overall mass balance. 5. The flow pattern is described by the axial dispersed plug flow model. 6. Equilibrium relationships for the components are represented by extended Langmuir isotherms. 7. The ideal gas law applies. 8. The PSA process is assumed to be equilibrium controlled.

Production of Hydrogen via Syngas Route

6-5

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

In the LDF model, the mass transfer rate equation was represented as:

∂qi = k i q i* − q i ∂t

(

ki = Ω where Ω where D M

τ

)

ε p D p  c0 

   q0 i D Dp = M , rp

2

i = A ( or ) B

τ

LDF parameter ( = 15 ) molecular diffusivit y(chapman − Enskog Equation)

c0

tortuosity ( = 3) gas concentrat ion in feed gas

q0

solid phase concentrat ion in equilibriu m with c o

In order to simplify the model, the mass transfer coefficient, ki for macropore control was adopted by assuming that the adsorbents were spherical in shape with no micropore diffusion of the adsorbates.

By neglecting the interactions between the adsorbed components and assuming that the reduction of the vacant surface area for the adsorption of A is solely due to the adsorption of the other components, the extended Langmuir model was applied as follows:

qi =

K i (q i )m p i 1+ ∑ K jpj j

where (qi )m is the maximum amount of adsorption of species i for coverage of the entire surface.

Production of Hydrogen via Syngas Route

6-6

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.5 ACTUAL MODELING OF PSA

Even after simplifying the LDF model, a moderate degree of difficulty still exist in the simulation of PSA via the use of COMSOL. Coupled with the issue of time constraint, the complexity of the model was further reduced to as follows: 1. All the water was assumed to be removed via the use of a layer of silica gel before the gas stream entered the PSA. 2. The subsequent six component system was reduced to a binary system of hydrogen (carrier gas) and carbon dioxide (adsorbate). 3. The molecular diffusivity of carbon dioxide in hydrogen, DM was derived using Hirschfelder Equation for which the molecules were assumed to be non-polar and non-reacting. 4. The pure component isotherms and the respective constants on a common adsorbent were applied. 5. The Peclet number was assumed to be a constant due to its small numerical value. 6. The product of time constants, a1 and a2 and time, t for both pressurization and blowdown step were assumed to be 6.

The selection of hydrogen and carbon dioxide as the two components was in accordance with the Hysys simulation whereby they accounted for more than 93% of the total number of moles entering PSA (CO2 ≈ 19%, H2 ≈ 75%). Thus for conservative design, the mole ratio of H2 to CO2 was assumed to be 1:3. Activated carbon [7] was employed as the adsorbent in this simulation due to its higher affinity for CO2 as compared to H2.

Production of Hydrogen via Syngas Route

6-7

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

The following partial differential equations were employed:

6.5.1 Component Mass Balance

For which A = CO2 and B = H2,

∂ 2 cA ∂c ∂v ∂c 1− ε ∂qA − DL 2 + v A + cA + A + =0 ∂z ∂z ∂t ε ∂t ∂z

6.5.2 Overall Mass Balance

∂2C ∂C ∂v ∂C 1− ε ∂qA ∂qB + ) =0 − DL 2 + v + C + + ( ∂z ∂z ∂t ε ∂t ∂t ∂z

6.5.3 Pressure terms

∂P = ( PH − PL )( a1 ) e − a1t ( For Pressuriza tion step ) ∂t ∂P = − ( PH − PL )( a 2 ) e − a 2t ( For Blow - down step ) ∂t ∂P =0 ( For Adsorption and Purging steps ) ∂t

6.5.4 Adsorption rates

∂qB e = kB (qB − qB ) ∂t

∂qA e = kA (qA − qA ) ∂t

Production of Hydrogen via Syngas Route

6-8

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

The above equations were subsequently converted to dimensionless form so as to firstly, reduce the number of variables and complexity in the modeling and secondly, to facilitate in the subsequent scale-up to the actual plant capacity.

6.5.5 Overall Mass Balance in Dimensionless Form

∂v ε Rg T0 ∂xA ∂xB ∂P 1 ( )− = −qSA + γs 1− ε P ∂χ ∂τ ∂τ ∂τ P

6.5.6 Component Mass Balance in Dimensionless Form

∂yA 1 ∂2 yA ∂yA ∂x ∂x  1− ε RgT0  = − v + q (yA −1) A + γs yA B  = 0 SA 2 ∂τ Pe ∂χ ∂χ ε P  ∂τ ∂τ 

6.5.7 Dimensionless Pressure terms Lτ  a1 L  −a1 Vo ∂ P  PL   e = 1 − (For Pressurization step) PH  v  ∂τ   o 

Lτ ∂ P  PL  a2 L  −a2 Vo  e = −1 − (For Blow-down step) PH  v   ∂τ  o 

∂P = 0 (For Adsorption and Purging steps) ∂τ

6.5.8 Dimensionless Langmuir Adsorption Isotherms

  yB PPH   yA PPH   b   bA B R T   Rg To   ∂xB g o ∂x A = αA  − xB  = αA  − xA  ∂τ ∂τ 1+ b yA PPH + b (1− yA )PPH  1+ b yA PPH + b (1− yA )PPH  A B A B     RgTo RgTo Rg To Rg To    

Production of Hydrogen via Syngas Route

6-9

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

The 5 dimensionless equations were inputted into COMSOL for simulation with the set of boundary conditions.

6.5.9 Boundary Conditions

Pressurization

Adsorption

1 dy A χ =0 = v Pe dχ dy A χ =1 = 0 dχ

−

v

χ =1

χ =0

(y

A χ =0 −

− yA χ =0+

)

=0

v

Blowdown dy A d χ dy A d χ v

χ =1

1 dy A χ=0 = v Pe d χ dy A χ =1 = 0 dχ

−

χ = 0

χ =1

χ =0

χ =0

(y

A χ=0−

− y A χ =0+

)

=1

Purge = 0 = 0

= 0

Production of Hydrogen via Syngas Route

dy A χ =0 = 0 dχ 1 dy A − χ =1 = v Pe dχ v

χ =1

χ =1

(y

A χ =1+

− y A χ =1−

)

= purge to feed velocity ratio (G )

6-10

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.6 MODEL OPTIMIZATION

The aim of the optimization process is to maximize the recovery of the H2 product whilst attaining the specified purity of 99.9%. For a reasonable and simplified design of the PSA system, the following heuristics would be adhered to in this report [8]:

•

A two bed system was assumed for the COMSOL-Matlab simulation.

•

Bed Porosity should be maintained in the range of 0.3 – 0.5.

•

Adsorption time should be close to the breakthrough time.

•

Adsorption and Desorption time should be kept constant for 2 bed processes.

•

Purge to feed volume ratio, G should in the range of 1.0-2.0.

•

The ratio of pressurization time to adsorption time should be capped at 0.2.

•

The superficial velocity entering the bed should not exceed 75% of the minimum fluidizing velocity [5].

There are multiple decision parameters that govern the operations of PSA and this includes the bed length, bed diameter, superficial velocity of the feed and purge ratio, G. A systematic approach was thus adopted wherein a variation of one of the parameters would be performed with the others held constant. For this varied variable, the optimum point would be that at which the required purity of 99.9% was achieved with the maximum allowable recovery. By maintaining the constant value for this variable, this optimization process was subsequently repeated for all the other parameters.

Production of Hydrogen via Syngas Route

6-11

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.6.1. Process Methodology

In this report, the following optimization sequence was adopted: 1. Initial approximation of the adsorption time from the breakthrough curve. 2. Determination of the cyclic steady state. 3. Refinement of the pressurization time. 4. Optimization of feed superficial velocity and diameter of the bed.

Production of Hydrogen via Syngas Route

6-12

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.6.2 Initial approximation of the adsorption time from the breakthrough curve

According to A. Varnia [9], an industry expert, a 5-bed PSA system with the height and diameter of approximately 8.5 m and 3.2 m respectively, can achieve a hydrogen production of (STP) 1.25 x109 m3/yr at 99.9% product purity. From this capacity and diameter, an inlet flow-rate of 3.33 m3/s and subsequently, a velocity of 0.41m/s, were approximated with the Hysys simulation. Thus this initial bed height, diameter and velocity were adopted in COMSOL to estimate the time period for breakthrough to occur (the time duration before which the maximum allowable adsorbate concentration (CO2) in the effluent gas (H2) is exceeded). The purge to feed volume ratio, G was also assumed to be at 2 [8].

Breakthrough occur at T = 17 τ

Figure 3: Breakthrough curve of yco2 with time

Based on this initial breakthrough time, an adsorption time could be subsequently approximated to initiate the optimization process, since the duration of the high pressure adsorption should be theoretically equal to that in order for breakthrough to occur [8].

Production of Hydrogen via Syngas Route

6-13

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

From the COMSOL simulation for the adsorption step, it could be observed that the time duration for breakthrough was about 17 τ in dimensionless time and this translates to:

t= =

T * BL Vo 17 * 8.5 0.4 / 0.36

= 130 s

Thus according to figure 7.1, 130 s was the upper limit for the varying of the subsequent adsorption times. The purity of the product started to decline once the breakthrough time was exceeded and in order to attain a high level of product purity, it is necessary to set the adsorption time to as near as possible to the breakthrough time.

6.6.3 Determination of Cyclic steady state

The dynamic steps in PSA meant that no steady state would be attained. However, after a sufficiently large number of cycles, there will be a point in time whereby the profiles achieved by the bed at the end of a cycle will be the same at the start of the next and that is when cyclic steady state is achieved.

In order to determine the cyclic steady state, the number of cycles, h required for each run of simulation was varied as shown below in figure 3. Four values of h (3, 4, 10 and 14) were used. For each run of simulation, a plot of the yco2 wave front versus bed length was generated and subsequently, they were compared on the same graph. There was a shift in the wave front to the right with an increase in the number of cycles. However, beyond the run for 10 cycles, a constant wave front was observed, which led to the conclusion that 10 cycles was required for cyclic steady state to be attained.

Production of Hydrogen via Syngas Route

6-14

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Run for 10 and 14 cycles Run for 3 cycles

Run for 5 cycles

Figure 4: Plot of varying yco2 wave front versus bed length

6.6.4 Refinement of the pressurization time

The durations of the adsorption and pressurization time are two important parameters that affect the performances of a PSA system. According to the simulations performed by S. Jain [8], the ratio of the pressurization time and adsorption time should be capped at a limit of 0.2 so as to achieve the best possible result for both purity and recovery. According to the earlier discussion, the optimum adsorption time should be set near to the breakthrough time and a value of 120 s was assumed for conservative reasons in order not to exceed the breakthrough time. The pressuration time was subsequently varied to obtain the best result. As seen from figure 7.3, the change in pressurization time was inversely proportional to recovery. (amount of H 2 obtained during adsorption − Re cov ery =

amount of H 2 used in purge step) (amount of H 2 used during adsorption step in feed + amount of H 2 fed during pressurization step)

Production of Hydrogen via Syngas Route

6-15

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

According to the above equation used in the calculation of recovery, with an increase in the pressurization time, more of the H2 will be fed during the pressurization step and this will result in a decrease in recovery, which was consistent with the trend displayed in figure 7.3. The values of Tads = 120 s and Tpr = 24 s were selected for further optimization. Purity vs Re cove ry for varying Tpr at cons tant Tads = 120s 0.9992 0.999 0.9988 0.9986

Tads = 120 s Tpr = 24 s

0.9984 0.9982 0.998

Tads = 120 s Tpr = 22 s

0.9978 0.9976 0.817

0.819

0.821

0.823

0.825

0.827

0.829

0.831

Figure 5: Plot of purity versus recovery at varying Tpr (Tads = 120 s)

6.6.5 Possible optimization of feed superficial velocity and diameter of the bed

If the purity specification of 99.9% was not attained or if there was potential for further improvement in product recovery, the last parameter for optimization would be the feed superficial velocity. In this section, the feed velocity was adjusted while maintaining a constant Tads and Tpr. An optimum value of 0.4 m/s for feed superficial velocity was subsequently attained from figure 7.4. It should be noted that the calculated feed superficial velocity should not exceed 75% of the minimum fluidizing velocity. This was because fluidization of the fixed bed could result in the possible loss of adsorbents with the exiting product stream. The minimum fluidizing velocity was calculated to check for the validity of the superficial feed velocity:

Re =

D p vs ρ f (1 − ε ) µ

Production of Hydrogen via Syngas Route

=

(0.003)(0.4)(9.634) = 1619 (1 − 0.36)(1.116 x10 −5 )

6-16

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Turbulent flow occurs where 500
1 1 2 (3)(850 − 9.634)(9.81)0.003 2  = ( ) = 2.78 (9.634) 

Particle Density, ρp = 850 kg/m3

Fluid Density, ρf = 9.634 kg/m3

g = 9.81 m/s2

Particle Diameter, dp = 0.003 m

ε = 0.36

Kinematic Viscosity, µ = 1.180 x 10-5 Pa s

Based on the above mentioned heuristics, the calculated feed superficial velocity of 0.4 m/s was safely within the range of less than 75% of the minimum fluidizing velocity, which was calculated to be 2.8 m/s.

Constant Tads , Tpr (120 s, 22 s), varying V 1.00000 0.99950

Purity

0.99900 0.99850 0.99800

V = 0.41 m/s

V = 0.40 m/s

0.99750 0.99700 0.80000

0.81000

0.82000

0.83000

0.84000

Re cove ry

Figure 6: Plot of purity versus recovery for different feed superficial velocity

Production of Hydrogen via Syngas Route

6-17

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.7 FINAL RESULTS AND DISCUSSIONS

Adsorption Time, Tads = 120 s Pressurization Time, Tpr = 24 s Superficial feed velocity, V = 0.4 m/s Recovery = 82.0% Purity = 99.9%

Pr oductivity =

Total output − Total purge = 1.15 x 109 m3 (STP)/yr Total time

where total time is time taken for both pressurization and high pressure adsorption to occur

Required capacity = 1.25 x 109 m3 (STP)/yr

It can be observed that the productivity achieved in this simulation was lower than the required plant capacity. One probable reason could have been due to the higher recovery assumed in the Hysys interim report (85%), which had subsequently resulted in a lower quantity of feed required. It should also noted that the separation performed in this report is on a binary system, which fails to account for the presence of the other components, such as CO, CH4 and N2 that are present in the actual feed stream to PSA. Thus the actual purity might have in fact been lower.

Production of Hydrogen via Syngas Route

6-18

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.8 COST ESTIMATIONS

The cost estimation is done by assuming a diameter of 3.0 m and bed length of 8.5 m.

Volume of vessel, V = π

D2 L 4

32 (8.5) 4 = 60 m 3

=π

From Figure A.7, purchased cost of vessel using carbon steel, C po = 850 × 60 = $ 51,000

Production of Hydrogen via Syngas Route

6-19

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

Pressure factor, FP ,vessel

PSA Unit Design Report

( P + 1) D + 0.00315 2[850 − 0.6( P + 1)] = for t vessel > 0.0063m pp925 [1] 0.0063

Where P = pressure in barg and D = diameter of vessel

P = 25 bar = 24 barg

FP ,vessel

(24 + 1)3 + 0.00315 2[850 − 0.6(24 + 1)] = = 7.63 0.0063

Purchased cost of vessel after pressure correction and adjusted for inflation

C p = C op × Fp, vessel ×

CEPCI in Nov 2007 CEPCI in Sept 2001

CEPCI in Nov 2007 = 593.6 [Chemical Engineering February 2008] (latest data available)

C p = $51,000 × 7.63 ×

593.6 = $ 581,833 397

Assuming auxiliary equipment, which include expander, control valves and safety valves piping,

Total Cp = $ 581,833 × 1.3 = $ 756,383

Production of Hydrogen via Syngas Route

6-20

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Calculation of cost of activated carbon

Mass of activated carbon used = V(1- ε b) ε p = 60 × (1-0.36) × 850 = 32,640 kg Cost of Calgon activated carbon ≈ $5/kg [2] Cost of activated carbon = 32,640 × 5 = $ 163,200

The real PSA bed for H2 purification is packed by not only activated carbon, but also Zeolite which is of much higher price than activated carbon.

Total cost for a PSA adsorber = $ 581,833 + $ 163,200 = $ 745,033

This estimated cost is expected to be less than the actual price because the actual PSA bed will use not only activated carbon but also Zeolite

For 2-PSA adsorber = $ 745,033 × 2 = $ 1,490,066

Assuming installation of PSA is 10% of total cost Therefore, total estimated cost of installed 2 PSA bed system = $ 1,490,066 × 1.1 = $ 1,639,073

Production of Hydrogen via Syngas Route

6-21

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.9 CONCLUSION

For this report, varying degrees of success has been achieved in the simulation of the PSA system. A productivity of 1.15 x 109 (STP) m3/yr at 99.9% product purity and 82% recovery has been obtained from the COMSOL Matlab simulation of a two bed, binary PSA system through the variation of pressurization time, Tpr and feed superficial velocity, V. However, this productivity still falls short of the required capacity of the H2 plant which was fixed at 1.25 x 109 (STP) m3/yr. A reasonable explanation for this shortfall could have been attributed to the lower feed required arising from a higher product recovery assumed initially (85%).

Industrially, the attainable recovery is within 80 – 90% at a purity of up to 99.999% with a 7 10 bed configuration and the use of more steps such as pressure equalization, co-current pressurization to increase the recovery of the system. Thus it might not have been feasible to achieve a recovery of 85% at a purity of 99.9% with a complete multi-component simulation of a two bed system.

Due to severe time constraints, only the binary system was considered in this simulation and this could have been expanded further into a multi-component system if more time was allowed. Furthermore, the simulation was performed for one adsorbent (activated carbon), which is not applicable for a multi-component system where at least two adsorbents are required for the near complete sorption of the various components so as to achieve a product with a very high level of purity.

Finally, a buffer tank could have been considered in the overall design of the system, so as to deal with the possible occurences of irregular flow from the upstream units. Through the installation of a buffer tank, irregular flow rates can be eliminated and this will ensure the smooth operation of the PSA system. Furthermore, in times of an upstream unit upset, the avaliability of the buffer tank can continue to sustain the PSA operations until the upstream unit is up and running again.

Production of Hydrogen via Syngas Route

6-22

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.10 NOTATIONS

Symbol

Meaning

Symbol

Meaning

c

Concentration

v

Superficial Velocity

q

Adsorbed Phase

Vo

Interstitial Velocity

L

Bed Length (m)

D

Bed Diameter (m)

qas

Saturated Adsorbed

Concentration (mol/kg) K

Mass transfer coefficient

DL

Molecular Diffusivity

Pe

Peclet Number

phase concentration (CO2) T

Time (s)

qbs

Saturated Adsorbed phase concentration (H2)

T

Temperature (K)

b

Langmuir constant

Ε

Void Fraction

Y

Gas phase concentration

z

Axial position in

G

Purge to feed ratio

g

Gravitational

adsorption bed (m) Vf

Minimum

acceleration (m/s2)

fluidization velocity Dp

Particle Diameter

µ

Dynamic Viscosity (Ns/m)

Production of Hydrogen via Syngas Route

6-23

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

Subscripts Symbol

Meaning

Symbol

Meaning

A

CO2

s

Saturated

B

H2

H, L

High, low

Production of Hydrogen via Syngas Route

6-24

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.11 APPENDIX

The molecular diffisivity:

Assuming 1) negligible viscosity change over temperature 2) Feed at 50ºC

A: Carbon Dioxide B: Hydrogen

Using Equation 24 -33 [6],

 1 1  0.001858T  +  MA MB   = 2 Pσ AB ΩD

1/ 2

3/ 2

D AB

Where σ AB =

σ A +σB 2

ε AB = ε Aε B κ = 1.38 x10 −16 ergs / K

From Appendix K, Table K.2

εA = 190 κ εB = 33.3 κ

σ A = 3.996 σ B = 2.968

Ω D = 0.88216(int erpolate) D AB = 3.645E-2 cm2/s = 3.645E-6 m2/s

Production of Hydrogen via Syngas Route

6-25

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

The original axial dispersed plug flow model [] is given by − DL

∂ 2ci ∂c ∂v ∂c i 1 − ε ∂ q i + v i + ci + + =0 2 ∂z ∂z ∂t ε ∂t ∂z

− DL

∂ 2 ci = axial dispersion term ∂z 2

where

D L = 0.7 D M + 0.5vd p = 0.7 D M + rp v v

∂ci ∂v + ci = convective mass transfer term ∂z ∂z

∂ci = accumulation term ∂t 1 − ε ∂q i = adsorption to solid phase term ε ∂t

for i = A and B, A = carbon dioxide and B = hydrogen − DL

∂ 2c A ∂c ∂v ∂c A 1 − ε ∂ q A + v A + cA + + =0 2 ∂z ∂z ∂t ε ∂t ∂z

We will first derive the overall mass balance equation which will be used to find the velocity profile along the bed.

The OVERALL MASS BALANCE, which is the sum of component mass balances − DL

∂2 ∂ ∂v ∂ (c A + c B ) 1 − ε ∂ q A ∂ q B ( c A + c B ) + v (c A + c B ) + (c A + c B ) + + ( + )=0 2 ∂z ∂z ∂t ε ∂t ∂t ∂z

cA + cB = C − DL

∂ 2C ∂C ∂v ∂C 1 − ε ∂ q A ∂ q B +v +C + + ( + )=0 2 ∂z ∂z ∂t ε ∂t ∂t ∂z

Production of Hydrogen via Syngas Route

6-26

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

The total concentration is not a function of the distance of bed c A + c B = C ≠ f (z )

and

∂ 2 C ∂C = =0 ∂z ∂z 2

The overall material balance becomes C

∂v ∂C 1 − ε ∂ q A ∂ q B ( + + + )=0 ∂z ∂t ε ∂t ∂t

∂v 1  ∂C 1 − ε  ∂ q A ∂ q B   =−  + + ∂z ε  ∂t ∂t   C  ∂t C=

P R g T0

∂C 1 ∂P = ∂t R g T0 ∂t R g T0 1 − ε ∂ q A ∂ q B ∂v ∂P 1 ( )− =− + OVERALL MASS BALANCE ∂z P ε ∂t ∂t ∂t P

This overall mass balance equation is used to find the velocity profile along the bed

Dimensionless form ∂v ε R g T0 ∂ x A ∂x B ∂P 1 D’LESS OVERALL MASS BALANCE = −q SA ( + γs )− ∂χ 1− ε P ∂τ ∂τ ∂τ P

Production of Hydrogen via Syngas Route

6-27

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

We will then proceed to find component mass balance ∂c ∂ 2c A ∂v ∂c A 1 − ε ∂ q A + v A + cA + + =0 − DL 2 ∂z ∂z ∂t ε ∂t ∂z cA =

yAP RT

− DL

∂y P ∂ 2 yA P y P ∂v ∂y A P 1 − ε ∂ q A +v A + A + + =0 2 ∂z RT RT ∂z ∂t RT ε ∂t ∂z RT

− DL

∂ 2 yA ∂y ∂v ∂y A 1 − ε RT ∂q A + v A + yA + + =0 2 ∂z ∂z ∂t ε P ∂t ∂z

Substitute OVERALL MASS BALANCE INTO COMPONENT MASS BALANCE R g T0 1 − ε ∂ q A ∂q B ∂v ∂P 1 ( =− + )− ∂z P ε ∂t ∂t ∂t P − DL

 R g T0 1 − ε ∂ q A ∂ q B ∂ 2 yA ∂y ∂P 1  ∂y A 1 − ε RT ∂ q A + v A + y A  − ( + )− 2  + ∂t + ε P ∂t = 0 ∂z P ε ∂ t ∂ t ∂ t P ∂z  

Simplify the above equation − DL

 R g T0 1 − ε ∂ q A ∂ q B  1 − ε RT ∂ q A ∂ 2 yA ∂y A − + v + y ( + )  + =0 A  ∂z P ε ∂ t ∂ t ε P ∂t ∂z 2  

Rearrange it ∂y A ∂ 2 yA ∂y 1 − ε R g T0 = DL −v A + 2 ∂t ∂z ε P ∂z

 ∂q ∂q   ( y A − 1) A + y A B  = 0 COMP. MASS BAL  ∂t ∂t  

∂y A ∂y ∂x ∂x  1 ∂2 yA 1 − ε R g T0  = − v A + q SA  ( y A − 1) A + γ s y A B  = 0 2 ∂τ Pe ∂χ ∂χ ε P  ∂τ ∂τ  D’LESS COMPONENT MASS BALANCE

Production of Hydrogen via Syngas Route

6-28

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.12 CONSTANTS APPLIED IN COMSOL SIMULATION

Temperature

323K

Bed voidage

0.36

Particle density

0.85 g/m3

qco2

7.12 mol/kg

Saturated adsorbed CO2

qH2

4.32 mol/kg

Saturated adsorbed H2

bco2

2.54e-6 Pa-1

CO2 Langmuir constant

bh2

7.02e-8 Pa-1

H2 Langmuir constant

kco2

0.1 s-1

Mass transfer coeff CO2

kh2

1.0 s-1

Mass transfer coeff H2

High P

2.5e6 Pa

Low P

1e5 Pa

Production of Hydrogen via Syngas Route

[7]

6-29

CN 4120: Design II Team 32: Zhang Zihong (U046816H)

PSA Unit Design Report

6.13 REFERENCES

[1] Retrieved from ‘www.astrochemicals.com/10136.htm’, 24/03/2008 [2] Tinall, B.M and Crews, M.A, ‘Economics of export steam for hydrogen plants,’ Hydrogen Engineering, (2003) 39 [3] J D Seader and E J Henley, “Separation Process Prinicples”, John Wiley & Sons, pp 798 (1998) [4] Retrieved from ‘http://www.norit-americas.com/images/adsorption-image.gif’ [5] ‘Pressure Swing Adsorption’, D.M. Ruthven et al, VCH Publishers, 1994, Pg 225 [6] J.R Welty, C.E.Wicks, R.E. Wilson, G.Rorrer, Fundamentals of Momentum, Heat, and Mass Transfer, John Wiley & Sons, Inc. (2001) 4th ed [7] J.H Park, J.N Kim, S.H Cho, Performance Analysis of Four-Bed H2 PSA Process Using Layered Beds, AIChE Journal (2000) Vol 46 [8] S.Jain, A.S.Moharir, P.Li, G.Wozny, Heuristic Design of pressure swing adsorption: A preliminary study, Separation and Purification Technology 22 (2003) 25-43. [9] V.Aspi (2008) Singapore Refinery Company, Jurong Island, Singapore, Personal communication [10] Retrieved from ‘http://faculty.washington.edu/finlayso/Fluidized_Bed/FBR_Fluid_Mech/fluid_bed_scroll.ht mcosting’ [11] Turton et al, Analysis, Synthesis, and Design of Chemical Processes, 2nd edition, Prentice Hall, 2007 [12] Retrieved from ‘http://www.apswater.com/shopdisplayproducts.asp?id=157&cat=Activated+Carbon’

Production of Hydrogen via Syngas Route

6-30

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Chapter 7 : HEAT EXCHANGER NETWORK EXECUTIVE SUMMARY Energy integration is of utmost importance in a chemical plant, where huge demands are being placed on plant personnel to fully utilize available utilities, in order to obtain the most economical and practical solution. To achieve optimal heat integration, it would involve the systematic development of a heat exchange network (HEN), together with detailed designs of each and every heat exchanger. The primary targets set out in this problem statement would be:  To design a heat exchanger network for a hydrogen plant to be sited in Singapore  To look into possible optimization of the heat exchanger network, considering a multitude of network variations and possible network evolution  To design one of the heat exchangers in detail after the heat exchanger network has been finalized Pinch analysis was the method of choice in deriving a minimum energy recovery (MER) network. Though 2 loops were identified, various infeasibilities limited the evolution of the MER network. The final network was eventually chosen based on the minimum total annual cost. The total annual cost of the selected network was within 1% of the target set. Detailed thermal sizing of the SMR feed preheater was then performed. A TEMA style AES shell and tube exchanger was utilized for this service. Computation of the estimated area, heat transfer coefficient, pressure drops and cost were calculated. The final design for the heat exchanger is projected to be able to handle about 99% of its intended heat load.

ACKNOWLEDGEMENTS This section dedicates acknowledgements to all who have helped the author by offering their valuable insights and advices. In particular, the author would like to express gratitude to Prof. Rangaiah for his advice. Last but not least, this work would not have been possibly done without the coordination and teamwork in Team 32 (CN4120 – Sem 2 of AY2007/08), hence the author would like to thank all of them for their assistance and understanding.

Production of Hydrogen via Syngas Route

7-1

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.1 DESIGN METHODOLOGY OF A HEAT EXCHANGER NETWORK Pinch technology was systematically adopted for the design and evolution of this design problem. Substantial effort was first channeled into the collection of relevant stream data. Manual determination and verification of these stream data properties, coupled with cost optimization were then carried out before the actual synthesis of the heat exchanger network. Two software programs, Hysys and HX-Net, were extensively used in this design process to facilitate data extraction, data processing, pinch analysis and super-targeting. 7.1.1 Determination & Verification of Stream Data Properties Extracted from Hysys Stream data properties (mass flow rates, M, supply temperature, Ts, target temperature, Tt, heat capacity, Cp and convective heat transfer coefficient, h) extracted into HX-Net from Hysys have to be validated first before the actual design of the heat exchanger network can proceed. Data relevant to the design are validated using suitable equations and correlations. 7.1.1.1 Calculations of Maximum Design Velocities As HX-Net uses a default stream velocity of 1 m/s, which would not be an accurate reflection of the actual stream velocities, it is essential to compute manually the maximum allowable design velocities. Correlations detailing these calculations are shown belowError! Bookmark not defined.. •

Allowable velocity for given liquid = 1

 2 ( Allowable velocity for water ) Density of water   Density of given liquid  where the allowable velocity of water in low carbon steel tubes is taken to be 10 ft/s. Therefore, a sample calculation for HP steam generation is shown below: 1

•

 62.4 lb/ft 3  2 Allowable velocity = 10 ft / s  = 11.2 ft / s = 3.43 m / s 3  49.4 lb/ft  For gases and dry vapors in steel tubes,

  1800 Allowable velocity for gas (ft/s) =    ( Absolute pressure in psia )(Molecular weight )  A sample calculation for Natural Gas feed stream is shown below:  Allowable velocity for NG feed =  

  = 18.42 ft / s = 5.62 m / s (580.2 psia )(16.45) 

Production of Hydrogen via Syngas Route

1800

7-2

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

The computed values for each process and utility stream were tabulated as follows in Table 1 and Table 2.

It can be seen that the liquid stream velocities were in the range of 1 to 4 m/s and vapor

streams were also well within the stipulated standard process design guidelines. Hence, we can deduce that reasonable estimates have been obtained and these values were subsequently used in place of the default value in HX-Net. Table 1: Maximum design velocities for vapor streams.

Absolute Pressure (psia)

Molecular Weight

Allowable Velocity (ft/s)

Allowable Velocity (m/s)

Pressure Category

NG Feed SMR Feed SMR Outlet HTS Outlet LTS Outlet Combustible Air

580.2 394.6 381.5 363.5 360.5

16.45 17.63 12.67 13.12 13.12

18.42 21.58 25.89 26.06 26.17

5.62 6.58 7.89 7.94 7.98

High High High High High

14.5

28.91

87.92

26.8

Atm

Flue Gas

11.5

27.07

102.02

31.1

Atm

Vapor Streams

Guideline VelocitiesError!

Bookmark not defined.

Atmospheric pressure: 10 to 30 m/s High pressure: 5 to 10 m/s

Table 2: Maximum design velocities for liquid streams.

Liquid Streams

Density (lb/ft3)

HP Steam Generation Cooling Water

49.4 62.4

Allowable Velocity (ft/s) 11.2 10.0

Allowable Velocity (m/s) 3.43 3.05

Guideline VelocitiesError!

Bookmark not defined.

1 to 4 m/s

7.1.1.2 Determination of Flow Area Diameter The flow area diameter in HX-Net was defaulted as 0.0254 m or equivalent to 1 inch. According to heuristics1, 3 4 in. (19 mm) is a recommended trial diameter to commence design calculations.

Selection of a suitable tube thickness also depends on internal pressure and corrosion issues. For plain carbon steels where severe corrosion is not expected, a minimum allowance of 2.0 mm should be used.Error! Bookmark not defined. Hence a tube thickness of 14 BWG (wall thickness = 2.11 mm) was used. Therefore, considering the above factors, an internal flow area diameter of 0.01483 m was adopted for HX-Net computations.

Production of Hydrogen via Syngas Route

7-3

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.1.1.3 Calculations of Convective Heat Transfer Coefficients (HTC)

Of paramount importance is the verification of extracted HTC values against manually calculated values computed from suitable correlations. This is because the magnitude of HTC values of each component stream are used to compute the overall heat transfer coefficient, U, which is then used to calculate the heat transfer area, an essential component of overall costing. Therefore, it is essential to ensure the accuracy of the HTC values in order to obtain an accurate cost estimation of the heat exchangers. Heat transfer that occurs in a heat exchanger is classified under forced convection in a closed conduit. For this type of flow, several correlations are applicable, dependent on whether the flow is laminar or turbulent. After computing the Reynolds number for all the streams using the maximum allowable velocity calculated in the previous section, it was found that the flow was mostly in the turbulent region. For single phase turbulent flow either in gases or liquids, the correlation proposed by Dittus and Boelter in 1930 can be readily applied2: k Re 0.8 Pr n d where n = 0.30 if fluid is being cooled, and n = 0.40 if fluids is being heated. hi = 0.023

To obtain the necessary dimensionless numbers in the above correlation, the arithmetic-mean bulk temperature is used as the basis for evaluating stream properties. However, for the heat exchanger involved in steam generation in the convection section of the furnace, there arises a need to compute the convective HTC value for a 2-phase flow system where in-tube boiling occurs. The Boyko-Kruzhilin equation3 was utilized to compute the HTC value as shown below as it is generally conservative and adequately accurate for most purposes. Sample calculation for steam generation in convection section of furnace: hi = 0.024

 ϑ − ϑv where (ϑ ϑm )i = 1 +  l  ϑv

 (ϑ ϑm )i + kl Re i0,.l8 Pri0,l.43  di 2 

  ϑ − ϑv  xi , (ϑ ϑm )o = 1 +  l   ϑv

Production of Hydrogen via Syngas Route

(ϑ ϑm )o   

  xo 

7-4

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

kl = thermal conductivity of the fluids, W/m °C; d i = inner diameter of tube, m

ϑl = density of liquid phase, kg/m3, ϑv = density of vapor phase, kg/m3 xo = inlet vapor fraction, xi = outlet vapor fraction hi = 0.024

0.58 1.0 + 2.2  240407 0.8 0.94 0.43  = 29474 W/m 2 .°C  0.01483 2  

(

)(

)

It was also noted that the heat exchanger cooling the LTS Outlet stream would result in the condensation of that process stream into a 2-phase system too. The Boyko-Kruzhilin equation was similarly applied to obtain its HTC value.

For the flue gas stream in the convection section, initial estimates of its Reynolds number seem to indicate that the fluid is in transition flow. In this case, the HTC value would be bounded by the laminar and turbulent conditions. Firstly, the HTC assuming a laminar flow regime was calculated using the Hausen correlationError! Bookmark not defined.: k f  0.0668 Re Pr (d i L )  µ  hi = 3.65 + 23 d i  1 + 0.04 + [Re Pr (d i L )]  µ w

  

0.14

where hi is the mean coefficient for the entire length L of a single tube, which was designated to be 16 ft (4.88 m) long according to heuristicsError! Bookmark not defined. available in literature.  µ Neglecting the viscosity correction term  i  µ w,i

 ,  

hi =

ki di

 0.0668 Re Pr (d i L )  3.65 + 23  1 + 0.04 + [Re Pr (d i L )] 

Another HTC value was then calculated as if the flow was turbulent using the Dittus-Boetler correlation. The final estimated HTC value for transition flow would then be calculated as follows:

(hi )T

 Re − 2000  = hi + hi − hi    8000 

[

]

7.1.1.4 Fouling Factors When process and service fluids flow through a heat exchanger, fouling will usually occur, the

extent dependent on the nature of the fluid. The deposited material on the heat transfer surfaces will normally have a low thermal conductivity and this will decrease the overall heat coefficient. Hence the effect of fouling should be incorporated into the preliminary design, so as to oversize

Production of Hydrogen via Syngas Route

7-5

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

the exchanger to allow for a performance reduction in actual operations. The list of fouling factors used is included in the following Table 3. Table 3: Typical fouling factors for process and utility streams.

Natural Gas

Typical Fouling Factors (ft2.h.°F./Btu) 0.001

Typical Fouling Factors (m2.°C/W) 0.000176

Light Hydrocarbon Vapors (clean)

0.001

0.000176

Hydrogen (saturated with H2O)

0.002

0.000352

Flue Gas Air (atmosphere) Steam (non-oil bearing) Cooling Water (treated makeup)

0.001-0.003. 0.002 was used 0.0005-0.001. 0.00075 was used 0.0005 0.001

0.000352 0.000132 0.0000881 0.000176

Type of Stream

7.2 TARGETING

Besides achieving an integrated heat exchanger network which satisfies the heating and cooling requirements of the various process streams, an important consideration is always an economic one; the network design should be one that is a balance between cost and efficiency. Different scenarios would warrant different approaches towards network synthesis. If utility costs are high, a network which maximizes energy recovery within the plant would then be the optimal choice. Conversely, if fuel costs are low, it would be an appropriate approach to opt for a network with fewer heat exchangers so as to lower annual costs. Hence, an estimation of capital (purchased equipment) cost and operating (utility) costs would have to be factored into the total annual cost of the network, and subsequently, it would provide the direction and set the criteria for evaluating each possible network design.

7.2.1 Cost Considerations

The total annual cost (TAC) of a network is calculated from the following formula: TAC = OC + AF × CC

where OC is the operating cost corresponding to the total utilities cost, AF is the annualization factor and CC is the capital cost. The annualization factor is given by:

Production of Hydrogen via Syngas Route

7-6

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

 Rate of return(% )  1 +  100  AF =  Plant life

Production of Hydrogen via Syngas Route

7-7

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

The rate of return is 10% as stated in the problem statement. However, although the plant life was stated to be 15 years, a plant life of 6 years was used instead after consultation with Professor Rangaiah. This is due to the typical of the lifespan of heat exchangers. These values were then keyed into HX-Net, where it will be used to automatically estimate the TAC with appropriate OC and CC inputs. 7.2.2 Utility Cost Calculations As the cost data given by the problem statement were in units incompatible with the input format

in HX-Net, there was a need to convert them to suitable units. Using high pressure (HP) steam for sample calculations, Cost of saturated HP steam (42 bar) = US$33 / metric ton = US$0.033 / kg Temperature of Saturated HP Steam (42 bar) = 254.3°C Mass heat of vaporization = 1726 kJ/kg Plant operation time = 8000 h/year = 28800000 s/year Cost index of HP Steam =

(Cost of

HP steam )(Plant operation time ) (US $ 0.033 kg )(28800000 s yr ) = mass heat of vaporization 1726 kJ kg

= US $550.6 kW . yr Similarly, the cost of cooling water used was calculated as follows: Cost of cooling water = US$0.067 / m3 Temperature range of cooling water = 90 to 120 °F = 32.22 to 48.89 °C At 32.22 °C, Mass heat capacity = 4.314 kJ/kg °C, Mass density = 1002 kg/m3 At 48.89 °C, Mass heat capacity = 4.320 kJ/kg °C, Mass density = 989.1 kg/m3 Taking the arithmetic mean, Mass heat capacity = 4.317 kJ/kg °C, Mass density = 995.55 kg/m3 Cost index of cooling water =

(Cost of cooling water )(Plant operation time ) (Mass density )(mass heat capacity )(Temperature difference )

=

(US $0.067 m )(28800000 s yr ) (995.55 kg m )(4.317 kJ kg − C )(48.89 C − 32.22 C ) = US $26.9 kW . yr 3

3

o

o

Production of Hydrogen via Syngas Route

o

7-8

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.2.3 Heat Exchanger Capital Cost Estimations Capital cost estimations have to be inputted into HX-Net so as to facilitate cost target

calculations. The capital cost index is represented in HX-Net as: CC = a + bA c where CC = capital cost and a, b and c are constants. Based on 1986 costs, the capital cost coefficients were given as a = 30800, b = 750 and c = 0.81.4 Any cost data must take into account the effect of time on purchased equipment cost, and this was carried out by using the Chemical Engineering Plant Cost Index (CEPCI) to incorporate inflationary effects over time5. CEPCI for 1986Error! Bookmark not defined. = 318, CEPCI for Nov 20076 = 526 Therefore, correcting for changing economic conditions over time:  526  0.81 CC =  = 50946 + 1241A 0.81  30800 + 750 A  318 

(

)

7.2.4 Supertargeting Generally, as the minimum approach temperature, ∆Tmin, increases, a dual effect is seen. Firstly,

the amount of heat recovered from hot process streams decreases, leading to an increased need for utilities, thus raising associated operating costs. However, having an increased ∆Tmin would also mean more effective heat transfer, as the temperature gradient, acting as the driving force, would be greater. Therefore, a smaller heat transfer area for the same target load would be achieved, giving cost savings in terms of reduced capital investment in smaller heat exchangers.

Figure 1: (Left) Total cost and operating cost as a function of ∆Tmin. Figure 2: (Right) TAC and OC as a function of ∆Tmin.

Production of Hydrogen via Syngas Route

7-9

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Supertargeting using the cost indices that were computed earlier (capital and operating costs) was carried out. It allows us to select the optimal ∆Tmin corresponding to the minimum TAC for a network. As seen from Figure 1, the optimal ∆Tmin is in the region of 4°C. However, this was on the low side and could be attributed to the faster rate at which the operating cost is increasing as compared to the savings in the capital cost as ∆Tmin increases. Hence, to provide a sufficient driving force, a ∆Tmin value of 10°C was chosen instead, since this would conform more to industrial standards and also lead to only a marginal increase of TAC (less than 1%). 7.2.5 Comparison between the usage of HP and LP Steam Generation It was found that TAC was actually a negative quantity, indicating that the designed network was

generating profits instead, due to the sale of high-value HP steam. Hence, it was decided that a choice should be made to either generate HP or LP steam. After making the necessary computations and keying the updated LP steam properties into HX-Net, a range target graph was generated, as seen from Figure 2. It can be seen that the TAC has dropped significantly as compared to Figure 1 where HP steam was used. Furthermore, ∆Tmin decreased to an even lower value. Hence, it was decided that the usage of HP steam would be a better economical choice. 7.2.6 Calculation of Utility Targets Once the heating and cooling requirements of the process streams have been determined together

with a specified ∆Tmin, the minimum amount of utilities can then be estimated. The temperatureinterval analysis, also known as the problem table algorithm, can be used for such a purpose.Error! Bookmark not defined.

Alternatively, the composite curve method first proposed by Umeda et al.7, can

be applied. Figure 3: (Left) Hot and cold composite curves.

Production of Hydrogen via Syngas Route

Utility Pinch

7-10

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 4: (Right) Utility composite curve showing the utility pinch created by HP steam generation.

Production of Hydrogen via Syngas Route

7-11

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

As seen from Figure 3, the graph is reminiscent of a threshold problem which only requires a cold utility, unaffected by ∆Tmin. Figure 4 shows the utility composite curve where HP steam generation and cooling water were used as the cold utilities. It is evident that the usage of HP steam generation actually introduces a utility pinch at around 253.3 to 254.3°C, the temperature of saturated HP steam at 42 bar. Hence the initial threshold problem with no pinch has now evolved into one with a utility pinch, where pinch analysis can be applied by treating the utility streams as “dummy” process streams8.

Figure 5: Network targets for ∆Tmin = 10°C.

7.3 MER NETWORK DESIGN

Although there are no process pinches in this design problem, a utility pinch has been introduced with the usage of HP steam generation. A method, first introduced by Linnhoff and Hindmarsh9, would be used to design two networks of heat exchangers, one on the hot side (above pinch) and one on the cold side of the pinch (below pinch). Certain rules would have to be followed according to the abovementioned method1: 1. The heat exchanger network is designed from the most constrained point where the approach temperature difference is at its minimum, i.e. at ∆Tmin. 2. At the pinch, streams are paired such that above the pinch, Cp, cold ≥ Cp, hot and below the pinch, Cp, hot ≥ Cp, cold. 3. The heat duty of each interior heat exchanger is selected to be as large as possible, so as to reduce the total number of exchangers.

Production of Hydrogen via Syngas Route

7-12

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

4. Hot utilities are added up to meet the heating energy targets, and no cold utilities are used above pinch.

Production of Hydrogen via Syngas Route

7-13

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

5. Cold utilities are added up to meet the cooling energy targets, and no hot utilities are used above pinch. The conditions for feasible pinch matches are as follows:  Number of streams going out of pinch ≥ Number of streams going into pinch. This is to

ensure that there are adequate streams for stream matching as certain utilities are not permitted above and below the pinch.  MCp of the out streams ≥ MCp of the in stream. This is to ensure no minimum driving

force violations. 7.3.1 Stream matching above pinch The number criterion is fulfilled here as there are 2 out streams (including the HP steam

generation utility) and 2 in streams. Although the only process cold stream (SMR Feed, circled red in Figure 6) above the pinch should have been matched with the Flue Gas stream according to the MCp criterion, this match was purposely left out as it was decided that the all the heat energy above the pinch should be used for HP steam generation in the convection section. This would not only give rise to more profits obtained from exporting the HP steam credit, but also, the steam generation tubes in the convection section can act as shield tubes in the first few rows of the convection section. This is because these tubes are exposed directly to radiation from the radiant section. During actual furnace operations, there will be fluctuations in the temperature of the flue gas; hence these tubes can thus act as a heat sink to regulate the temperature of the exiting flue gas.

Excluding the above exception, the rest of the streams were matched so long as the head loads were comparable and no driving force violations exist. MCp criterion need not may obeyed in this non-pinch matches. The pinch point has been calculated by HX-Net to be 263.3 / 253.3°C. 7.3.2 Stream matching below pinch The number criterion is also fulfilled here with 2 out streams versus 1 in stream. A pinch match

was made between the HTS outlet stream and the SMR feed stream (circled yellow in Figure 6),

Production of Hydrogen via Syngas Route

7-14

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

following the MCp criterion. Similar to the above pinch scenario, the rest of the matches were then matched accordingly using similar heat loads and no driving force violations as the basis. Stream splitting was only carried out for the HP steam generation stream. However, it was represented in this way, rather than as separate utility streams, for network simplicity and easy summing up of the total cooling load needed above the pinch. 7.3.3 Number of units in MER network After stream matching has been completed for both sides of the pinch, the total number of heat

exchangers in the MER network was compared with the calculated minimum number of units using the following equation: u min = n p + nu − 1

where n p represents the number of process streams and nu the number of utilities. Minimum number of units above utility pinch = 4 + 1 − 1 = 4 Minimum number of units below utility pinch = 6 + 1 − 1 = 6 Total number of heat exchangers in the HEN network = 6 + 4 = 10 Therefore, MER condition has been met in the designed heat exchanger network as seen from Figure 6. 7.3.4 Alternative MER Network Designs for Consideration Two other MER networks were also designed to see if alternative MER networks would give a

lower TAC. 7.3.4.1 Network 1a As seen from Figure 7, the heat exchanger network above the pinch remained the same. However,

a variation below the pinch was adopted. The combustible air to the furnace was solely preheated by the convection section, instead of being heated by the LTS outlet stream, HTS outlet stream and the convection section. This was because of considerations over the physical limitations to the placement of the heat exchangers as increased pipe lengths and insulation would then have to be factored into the cost of the network. However, it can be seen from the network performance and cost indices table in Figure 7 that this network would not offer a lower TAC. Instead, capital cost and total area would be higher than that of the original network design (Network 1). Since it cannot be ascertained at this preliminary design stage that piping cost would be a major factor in

Production of Hydrogen via Syngas Route

7-15

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

TAC and the units that make up a hydrogen plant are usually located close by, it has been advised that Network 1 would suffice at this juncture. 7.3.4.2 Network 1b As seen from Figure 8, again the heat exchanger network above the pinch remained in a similar

configuration as the previous 2 networks discussed. The combustible air was now preheated solely by the LTS outlet stream. The TAC and total area calculated by HX-Net seemed to indicate that this network was at a similar performance level as that of Network 1. However it was apparent that cooling water would have to be used to cool the hot flue gases, which would not be feasible as the primary use of the convection section should be to heat up the other process cold streams. Cooling water should only be used as a last resort unless stack gas temperatures are above environmental regulations. Hence this alternative was also discarded. The network performance and cost indices of each MER network are summarized in Table 4. The total areas for the 3 networks were satisfactorily within 20% of the target area. Network 1 is thus selected for network evolution. Table 4: Network performance and cost indices of each MER network.

Network 1 % of HEN Target Capital (US$) Total Cost (US$) Total Area (m2)

Network 1a % of HEN Target

Network 1b HEN

% of Target

7.540e+6

114.1

7.717e+6

116.8

7.533e+6

114.0

-4.819e+7

100.6

-4.813e+7

100.7

-4.819e+7

100.6

1.774e+4

82.67

1.836e+4

85.56

1.769e+4

82.41

Production of Hydrogen via Syngas Route

7-16

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 6: Network 1, original MER network design

Production of Hydrogen via Syngas Route

7-17

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 7: Network 1a, air preheated solely by furnace convection section

Production of Hydrogen via Syngas Route

7-18

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 8: Network 1b, air preheated solely by LTS outlet stream

Production of Hydrogen via Syngas Route

7-19

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.4 NETWORK EVOLUTION

Evolution of the selected MER network (Network 1) was carried out, so as to reduce the number of heat exchangers required, which would then result in a lowering of capital costs. However, it is to be noted that the operating costs of the network would then increase consequently as some energy recovery would have to be sacrificed and more utilities would be needed. As a result, the TAC of an evolved network may be lower than the MER network. Hence evolution of the MER network was done to facilitate a comparison of the TAC of both networks.

7.4.1 Steps involved in network evolution  Determination of the scope of improvement.  Identification of loops present in exchanger network. An even number of units in

a loop have to be ensured  Breaking of loop, calculate resultant stream temperatures and check driving forces  Restore ∆Tmin if there is any violation by forcing heat along the path  Continue breaking other loops and restoring ∆Tmin one at a time if possible and

attractive The number of minimum units assuming no pinch is u min = n p + nu − 1 = 7 + 2 − 1 = 8 , which was found to be the same as the minimum number of units computed by HX-Net. Hence the scope of improvement would be 10 − 8 = 2 loops. These 2 loops were identified by HX-Net and shown respectively in Figure 9 and Figure 11.

7.4.2 Evolution of 1st loop The heat exchanger with the smallest heat load (circled green) was identified and its heat

load transferred to the next heat exchanger in the loop, as indicated by the green arrow in Figure 9. The smallest heat load was eliminated because the cost savings in removing a small exchanger would be much higher than the cost incurred in increasing the area of another larger exchanger already in place. The temperatures of the affected streams were then re-calculated with the new heat loads placed on them.

Production of Hydrogen via Syngas Route

7-20

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

The evolved loop is shown in Figure 10. It can clearly be seen that there are now 2 heat exchangers which have a temperature cross situation (circled white). Efforts were made to restore ∆Tmin, however as there were no cold utilities already in place at the ends of the 2 process hot streams below the pinch, there was no viable way to prevent ∆Tmin from being violated and a temperature cross situation to ensue. It would not be a good solution to have to add a cooling water as a cold utility at the ends of either of the 2 streams, as it defeats the purpose of evolving a network and reducing the number of exchangers. Hence this loop was not considered for evolution.

7.4.3 Evolution of 2nd loop

Similarly, the heat exchanger with the smallest heat load (circled magenta) was identified and its heat load transferred to the next heat exchanger in the loop, as indicated by the magenta arrow in Figure 11. Again, as seen in Figure 12, a temperature cross was encountered in one of the heat exchangers (circled white) due to the lack of cold utility below the pinch, at the other end of the process stream. Hence, similar reasons prevented the restoration of ∆Tmin, hence this loop was also not considered for evolution.

Production of Hydrogen via Syngas Route

7-21

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 9: Network 1 showing 1st loop.

Production of Hydrogen via Syngas Route

7-22

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 10: Evolution of 1st loop.

Production of Hydrogen via Syngas Route

7-23

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 11: Network 1 showing 2nd loop

Production of Hydrogen via Syngas Route

7-24

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Figure 12: Evolution of 2nd loop.

Production of Hydrogen via Syngas Route

7-25

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.5 HEAT EXCHANGER DESIGN Shell and tube exchangers are the most commonly used type of heat transfer equipment in

the process industry, as they offer a large range of operating conditions, good mechanical layout, large surface area in a compact setup, flexibility in choice of construction materials, ease of fabrication and cleaning, and most importantly, well-established design procedures. Standards of the American Tubular Heat Exchanger Manufacturers Association (TEMA) are universally adapted. For petroleum and related industries with generally severe duties, heat exchangers are typically fabricated in accordance with class “R” TEMA specifications.10 7.5.1 Stream Data With reference to Figure 6, the heat exchanger chosen for thermal sizing would be the

one using the SMR outlet stream to preheat the SMR feed stream (circled yellow). The details of the unit are shown below: Table 5: Heat exchanger stream properties.

Hot Stream Temperature Effective Cp in kJ/kg°C Thermal Conductivity, kf, in W/m°C Density, ρ, in kg/m3 Viscosity, µ, in cP Mass flow rate, M, in kg/hr Fouling Factor in hr.sq. ft.°F /BTU 11 Fouling Factor in m°C/W

Inlet 851.9

Cold Stream Temperature Effective Cp in kJ/kg°C Thermal Conductivity, kf, in W/m°C Density, ρ, in kg/m3 Viscosity, µ, in cP Mass flow rate, M, in kg/hr Fouling Factor in hr.sq. ft.°F /BTU Error! Bookmark

Inlet 253.3

not defined.

Fouling Factor in m°C/W

Production of Hydrogen via Syngas Route

Outlet 606

2.964 0.183 4.031 0.02797 183500 0.002 0.000352 Outlet 539.4

2.483 0.067 9.258 0.02002 183500 0.001 0.000176

7-26

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.5.2 Material of Construction

The choice of a suitable material of construction is more often that not determined by corrosion-related issues. Factors which influence the amount of corrosion include oxidizing agents, solution pH, temperature, fluid velocity, films etc.12 For the heat exchanger chosen for further thermal design, both the hot and cold streams are not of a corrosive nature as hydrogen levels still has not made up a significant proportion of the streams, and are also of low enough temperature such that temperature-related corrosion is negligible. Even if fluid velocities are high, it can be overcome with design features such as tube inserts or impingement baffles within the exchanger unit. After considering the above factors, it was decided that carbon steel, the most common material for heat exchangers, would be a suitable material for both the tube and shell sides. 7.5.3 Shell and Tube-Side Fluid Allocation The following are considerations for fluid allocation in a heat exchanger with no phase

change10: Table 6. Analysis for shell and tube-side fluid allocation

Factor Corrosion

Fouling

Fluid Temperatures

Operating Pressures Pressure Drop

Viscosity

Stream Flow Rates

General Guidelines More corrosive fluid to tube-side to reduce cost of expensive alloy or clad components More fouling fluid to tube-side to increase fluid velocity, reduce fouling and facilitate cleaning Hotter fluid to tube-side to reduce heat loss, cost and for safety reasons Higher pressure stream to tube-side to reduce material cost Fluid with lower allowable pressure drop to tube-side to obtain higher heat-transfer coefficients More viscous fluid to shell-side, provided turbulent flow is achieved

Fluid with lower flow rate to shellside

Production of Hydrogen via Syngas Route

SMR Outlet / SMR Feed Both are non-corrosive

SMR outlet stream is more fouling due to the presence of H2 SMR outlet stream is hotter than SMR inlet Both streams are of similar pressures Both streams have similar allowable pressure drops SMR outlet stream slightly more viscous, but difference not significant Both streams have similar flow rates

7-27

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

From Table 6, it was decided that the SMR outlet stream would be assigned to the tubeside, while the SMR feed stream would be the shell-side fluid. 7.5.4 Exchanger Type Designation of TEMA shell and tube heat exchangers follows a three-letter code to

specify the front end stationary head type, shell type and rear end head type. It was decided that the AES type exchanger would be employed. Front head type A has a removable channel and cover plate and is the least expensive option. Shell type E is the most common shell construction but higher pressure drop may ensue. This can be controlled by varying the tube layout and baffle pitch. Rear end head type S is most commonly used for internal floating head designs, and allows for the cheapest straight tube removal bundle. 7.5.5 Baffles Baffles are used in the directing of fluid stream across the tubes so as to increase the fluid

velocity and thus improve the rate of transfer. For single segmental baffles, the maximal baffle cut is about 45%. Optimum baffle cuts range from 20 to 25%, while the optimum spacing is usually 0.3 to 0.5 times the shell diameter.10 7.5.6 Tube Dimensions Typical tube diameters range from 5 8 in. (16 mm) to 2 in. (50 mm). A smaller diameter

( 5 8 to 1 in.) is usually used for most duties, as they give smaller and cheaper exchangers. According to heuristics13, 3 4 in. (19 mm) is a recommended trial diameter to commence design calculations. However, considering the large volume of cooling water needed, a 1 in. tube diameter (do = 25.4 mm, di = 21.18 mm) was employed. Selection of a suitable tube thickness depends on internal pressure and corrosion issues. For plain carbon steels where severe corrosion is not expected, a minimum allowance of 2.0 mm should be used.12 Hence a tube thickness of 14 BWG (wall thickness = 2.11 mm) was used. Tube length L was designated to be 16 ft (4.88 m) long according to heuristics4 available in literature.

Production of Hydrogen via Syngas Route

7-28

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.5.7 Tube Arrangements Several tube arrangements exist: equilateral triangular, square or rotated square pattern.

An equilateral triangular pattern, as seen in Figure 13, was used as recommended by heuristics, as it provides higher heat transfer rates. Furthermore, heavy fouling requiring mechanical cleaning is not expected in this particular heat exchanger since shell-side fluid is relatively clean. A tube pitch Pt (distance between tube centres) of 1.25do was also utilized in accordance with common practices10.

Figure 13. Equilateral triangular tube arrangement.

7.5.8 Calculations The log mean temperature difference ∆Tlm is calculated as follows: ∆Tlm =

where ∆Tlm

(T

1

)

− t2 − (T2 − t1 ) (851.9 − 539.4 ) − (606.0 − 253.3) = = 332.2oC 851.9 − 539.4) ( T1 − t2 ln ln (606.0 − 253.3) (T2 − t1 )

(

)

= log mean temperature difference

T1

= hot fluid temperature, inlet

T2

= hot fluid temperature, outlet

t1

= cold fluid temperature, inlet

t2

= cold fluid temperature, outlet

To account for non-ideal counter-current flow within the exchanger, a correction factor Ft is applied. The correction factor is a function of shell and tube fluid temperatures, and the number of shell and tube passes, and is correlated as follows: R=

T1 − T2 851.9 − 606.0 = = 0.859 t2 − t1 539.4 − 253.3

S=

t2 − t1 539.4 − 253.3 = = 0.478 T1 − t1 859.9 − 253.3

From the temperature correction factor plots for one shell pass, two or more tube passes10 or using a correction factor equation derived by Kern, the corresponding Ft = 0.883.

Production of Hydrogen via Syngas Route

7-29

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Since this value is above the benchmark value of 0.75 for an economic exchanger design, it confirms that the single pass configuration is sufficient. The true temperature difference

∆Tm can then be computed: ∆Tm = Ft ∆Tlm = 0.883 × 332.2 = 293.3o C

Since both hot and cold streams are gases, the overall heat transfer coefficient U should be 10-50 W/m2°C. A value of 10 W/m2°C was used as the starting value for iteration. The provisional required heat transfer area A would then be: A=

36158.9 kW × 1000 Q = 12300 m 2 = U∆Tm 10 W / m 2 ° C × 293.3° C

Tubes with the following properties are chosen for use:

do

= 25.4 mm

di

= 21.18 mm

L

= 4.88 m

As a first approximation, an allowance of 50 mm for tube-sheet thickness (2 tubes sheets) was included, take L = 4.83 m . External surface area of one tube, At = πd o L = π (25.4 × 10 −3 )(4.83) = 0.385 m 2 Number of tubes needed, N t =

A 12300 = = 31948 ≈ 31900 At 0.385

Number of tubes per pass, assuming 2 passes, Np = 31900 ÷ 2 = 15950 Tube internal cross-sectional area, Ai =

π

(21.18 × 10 ) 4

−3 2

= 0.0003523 m 2

Area per pass = N p × Ai = 15950 × 0.0003523 m 2 = 5.62 m 2

Volumetric flow = =

mass flow rate per hr 1 × 3600 density

183500 kg / h 1 × = 12.65 m 3 / s 3600 s 4.031 kg / m 3

volumetric flow 12.65 m 3 / s = = 2.25 m / s Tube-side velocity, u t = area per pass 5.62 m 2

Production of Hydrogen via Syngas Route

7-30

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Check the tube-side velocity to ascertain reasonable value. The velocity is a bit too low for high pressure vapor streams, where velocities should range from 5 to 10 m/s10. This will be checked on later together with the pressure drop specifications. 7.5.8.1 Tube-Side Heat Transfer Coefficient Calculations ρu d 4.031 × 2.25 × 0.02118 Reynolds number, Re = t i = = 6870 µ 0.02797 × 10 −3 C p µ 2.964 × 10 3 × 0.02797 × 10 −3 Prandtl number, Pr = = = 0.453 kf 0.183

L 4.83 × 10 3 mm = = 228 di 21.18 mm From Figure 12.23, the tube side heat transfer factor, j h = 4.1 × 10 −3 . Tube side heat transfer coefficient then can be computed by 10: hi = j h Re Pr

0.33

 µ   µw

  

−0.14

kf   di

  

 µ  Neglecting the viscosity correction factor   due to non-viscous nature of gases,  µw   0.183  2° hi = 4.1 × 10 −3 × 6870 × 0.4530.33 ×   = 187 W / m C 0 . 02118  

7.5.8.2 Shell-Side Heat Transfer Coefficient Calculations For two tube passes in triangular pitch, the tube bundle diameter Db is estimated based on

the following empirical equation10: N  Db = d o  t   K1 

1

n1

 31900  = 25.4   0.249 

1

2.207

= 5237 mm = 5.235 m

Since a split-ring floating head was used, the bundle diametrical clearance is approximated as 78 mm10. Therefore, minimum shell inside diameter Ds = Db + 78 mm = 5315 mm = 5.315 m. This could be too large a value and would be corrected in the next iteration by increasing the number of tube passes. Kern’s “bulk-flow” method was used to estimate the shell-side heat transfer coefficient and pressure drop. Although Kern’s method does not take into consideration the bypass

Production of Hydrogen via Syngas Route

7-31

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

and leakage streams unlike the more vigorous Bell’s method, it was still adopted as it is simple to apply and provides a satisfactory preliminary approximation. An initial baffle spacing, lB equal to 0.4Ds, and a baffle cut of 25 percent was used to give a balance of good heat transfer rates and minimal pressure drop. Baffle spacing, l B = 0.4 × 5315 = 2126 mm = 2.126 m tube pitch, pt = 1.25 × 25.4 = 31.75 mm = 0.03175 m

( pt − d o )Ds l B

Cross-flow area, As =

=

pt

Mass velocity, G s =

(31.75 − 25.4) × 5237 × 2126 × 10 −6 31.75

= 2.23 m 2

Ws (183500 kg / h × 1 h / 3600 s ) = = 22.9 kg / s m 2 As 2.23

Linear fluid velocity, u s =

Gs

ρ

=

22.9 kg / s m 2 = 2.47 m / s 9.258 kg / m 3

The shell-side fluid velocity is lower than the recommended 5-10 m/s and would be taken note in the next iteration.

Shell-side equivalent diameter, de (hydraulic diameter) for an equilateral triangular pitch arrangement is computed as follows: de =

(

[

)

]

1.10 2 1.10 2 pt − 0.917 d o2 = 31.75 2 − 0.917(25.4 mm ) = 18.0 mm = 0.018 m d0 25.4 mm

Re = Pr =

Gs d e

Cpµ kf

µ =

22.9 × 18.0 × 10 −3 = = 20600 0.02002 × 10 −3

2.483 × 10 3 × 0.02002 × 10 −3 = 0.74 0.067

 µ Shell-side Nusselt number, Nu = j H Re Pr   µw 13

  

0.14

where jH is the dimensionless heat transfer factor. For the calculated Reynolds number, the corresponding value of jH is 0.0042. 10  µ   , Neglecting the viscosity correction term  µ  w

Production of Hydrogen via Syngas Route

7-32

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Nu = j H Re Pr 1 / 3 = 0.0042 × 20600 × 0.741 / 3 = 78.3 hs =

Nu × k f de

78.3 × 0.067 W / m 2 ° C = = 291 W / m 2 ° C 0.0180 m

7.5.8.3 Overall Heat Transfer Coefficient Calculations The overall heat transfer coefficient is given by d  d o ln 0  1 1 1  di  + do × 1 + d0 × 1 = + + U h0 hod 2k w d i hid d i hi

Where U

= overall heat transfer coefficient, W/m2°C

ho

= outside fluid film coefficient, W/m2°C

hi

= inside fluid film coefficient, W/m2°C

hod

= outside dirt coefficient (fouling factor), W/m2°C

hid

= inside dirt coefficient, W/m2°C

kw

= thermal conductivity of the tube wall material, W/m2°C

From literature, thermal conductivity of carbon steel = 55 W/m2°C 10  0.0254  0.0254 ln  1 1 1 1 0.0254 1 0.02118  0.0254  = + + + × + × U 291 5679 2 × 55 0.02118 2839 0.02118 187 o

U = 95.3 W / m 2 C % error from estimated U =

95.3 − 10 × 100% = 853% 10

Since the overall heat transfer coefficient is not within 30% deviation from the initial estimate, the design is not satisfactory and the calculated U would be used as the initial value for the next iteration.10 7.5.8.4 Tube-Side Pressure Drop Calculations Tube-side pressure drop is computed as:

 L ∆Pt = N p 8 j f    di

 µ     µ w 

−m

 ρu 2 + 2.5 t  2

Where jf is the dimensionless friction factor.

Production of Hydrogen via Syngas Route

7-33

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

Ignoring the viscosity correction term, 4.031 × 2.25 2 ∆Pt = 2[8 × 0.0052(228) + 2.5] = 245 N / m 2 2 The calculated tube side pressure drop is below the specifications (below 3 psi = 20.68kPa, the designated pressure drop for heat exchangers with vapor service). The number of tube passes should be increased in the next round of iteration. 7.5.8.5 Shell-Side Pressure Drop Calculations Shell-side pressure drop is computed as: D ∆Pt = 8 j f  s  de

 L  ρu s2    l B  2

 µ   µw

  

−0.14

Ignoring the viscosity correction term,  5315 mm  4.83 m  9.258 × 2.47 2   ∆Pt = 8 × 0.045 = 6820 N / m 2 2  18 mm  2.126 m  Both pressure drops for the tube side and shell side are too low. The calculations would

have to be iterated again. 7.5.9 Modification of Design As the overall heat transfer coefficient is not within 30% deviation from the initial

estimate, some modifications of the design parameters need to be implemented.

Firstly, 5 more iterations were done using the same overall parameters, each time using the new calculated value of U. This generated a U that was within 30% of the previous value. However, the value is 515.1 W/m2°C, still way much higher than the recommended range of 10-50 W/m2°C. Furthermore, the shell side pressure drop was more than the specifications.

Therefore, it was decided that a thicker tube with tube diameter (do = 50.8 mm, di = 46.59 mm) should be used. This yielded better results. After another 3 iterations, the value of U decreased to 351 W/m2°C, however, the tube side and shell side velocities remained above the 5-10 m/s guideline. A final change to the baffle cut from 25% to 45% was used.

Production of Hydrogen via Syngas Route

7-34

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

After 2 iterations, it resulted in the following design specifications, which met most of the specifications and remained fairly constant after a few iterations. Table 7: Final heat exchanger specification.

Fluid velocity (m/s)

Tube Side 32.6 (5-10 m/s)

Shell Side 26.7 (5-10 m/s)

h (W/m2°C)

1650

576.3

∆P (kPa)

19.2 ( 20.68<∆P<258)

150 ( 20.68<∆P<258)

Overall 2

A(m ) U (W/m2°C) Q (MW)

351 351 36.1 (99.9% of target heat load)

This was taken to be the final heat exchanger specification as shown in Table 7. 7.5.10 Exchanger Cost The cost of the single heat exchanger that was designed was estimated using the

CAPCOST program5. The bare module cost of a heat exchanger is computed as follows: C BM = C p ° FBM = C p (B1 + B2 FM FP ) °

log10 C p ° = K 1 + K 2 log10 A + K 3 (log10 A) log10 F p = C1 + C 2 log10 P + C 3 (log10 P )

2

2

Where Cp° = purchased cost, FM = material factor (carbon steel = 1), FP = pressure factor, A = heat exchanger area (m2), P = design pressure (barg), B1, B2, K1, K2, K3, C1, C2, and C3 are constants. For a 1-2 shell and tube exchanger, made up of carbon steel tubes and shell, with a heat exchanger area = 351 m2 and a maximum design pressure of 30 barg (an estimated 10% safety factor was added on top of the normal operating pressure of 27 bar) , the bare module cost as computed by CAPCOST is: log10 C p ° = 4.8306 − 0.8509 log10 (351) + 0.3187[log10 (351)]

2

C p ° = US $ 53,644 log10 F p = −0.00164 − 0.00627 log10 (30) + 0.0123[log10 (30)]

2

F p = 1.04 Production of Hydrogen via Syngas Route

7-35

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

C BM = C p ° FBM = 53644(1.63 + 1.66 × 1 × 1) = US $ 176,488 Given that CEPCI in 3Q 2001 = 39755 and CEPCI for Nov 20076 = 526, correcting for inflation over year, taking into consideration that the above equations apply to the cost in 3rd quarter 2001,  526  C BM =  US $ 176,488 = US $ 233,835 ≈ US $ 234,000  397 

7.6 RECENT DEVELOPMENTS

In recent years, there has been a growing interest in the research on the aspect of heat exchanger fouling, with emphasis being placed on the methods of prevention and cleaning. This is due to the high costs and time incurred with such operations. In order to account for the effects of fouling in heat transfer, numerous models and simulations have been developed to predict the rate of fouling but their accuracies are often limited by the use of ideal fouling resistance that has few uses in real world applications. Several measures have been adopted to combat fouling, an especially prevalent problem in refineries as refineries seek to increase their profit margins, increasingly buying and processing heavier, high sulphur and cheaper crudes, leading to elevated deposition problems. Strategies to ease problems includes13:  Feed analysis to minimize concurrence of two dominant fouling mechanisms  Blending light and denser crude oils together while avoiding precipitation  Further study of surface characteristics so as to fundamentally understand fouling  Use of intermittent pulsed flow  Feed at varying temperatures and pH values

These methods have shown promise in reducing fouling and had also resulted in significantly improved HTCs that can allow small heat exchangers to be used instead. Also, maintenance costs would be mitigated with these implementations.

Production of Hydrogen via Syngas Route

7-36

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.7 HEAT EXCHANGER SPECIFICATION SHEET Equipment number (Tag) A-123 Description/Function Hydrogen Plant Type/Class AES / R No. of units 1 Shells per unit 1 Connected in (parallel or series) 1 1 Surface per unit (m2) 351 Surface per shell (m2) 351 Performance of one unit SHELL SIDE TUBE SIDE Fluid circulating SMR Feed SMR Outlet Total fluid entering (kg/hr) 183500 183500 IN OUT IN OUT Vapor flow rate (kg/hr) 183500 183500 183500 183500 Liquid flow rate (kg/hr) Non-condensables flow rate (kg/hr) Temperature (K) 526 813 1125 879 Density (kg/m3) 9.258 4.031 Molecular weight 17.63 12.67 Viscosity liquid (kg/m.s) 2.002e-5 2.797e-5 Latent heat (kJ/kg) Specific heat (J/kg.K) 2483 2964 Thermal conductivity (W/m.K) 0.067 0.183 Operating pressure (kPa) 2700 2655 Velocity (m/s) 26.7 32.6 Number of passes 2 2 Fouling factor (m2.K /W) 0.000176 0.000352 Pressure drop (kPa) 150 19.2 Heat transferred (kJ/hr) 1.30e+8 MTD (corrected) (K) 293 Overall U (W/m2.K) 351 Construction of one Shell Maximum operating pressure (kPa) 3000 3000 Maximum operating temperature (K) 813 1125 Type of unit Tube pitch 0.0635 m Joint Strength weld Tube material Carbon steel O.D (m) 0.0508 I.D (m) 0.04659 Length (m) 4.88 Shell material Carbon steel Diameter (Approx.) (m) 1.53 Tube Sheet material Carbon steel Baffle material Carbon steel Corrosion allowance (m) Tube side 0.00211 Shell side 0.00211 Baffle cross C.S. Type Segmental Spacing (m), % Cut 0.642, 45 Heat Exchanger data sheet

Baffle arrangement

Nozzle arrangement

Remarks: Flared nozzle to reduce high inlet gas velocities.

Production of Hydrogen via Syngas Route

7-37

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

7.8 INTEGRATED HEN WITH PFD OF PROPOSED HYDROGEN PLANT

Figure 14: Integrated heat exchanger network with PFD of proposed hydrogen plant.

Production of Hydrogen via Syngas Route

7-38

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

APPENDIX A – STREAM DATA SMR Feed To Heated SMR Feed SMR Outlet To Cooled HTS Feed

Cold

Inlet Temp (°C) 218.0

Hot

851.9

353.9

0.1470

73234.72

471

183497

2.88

NG Feed To Heated NG

Cold

25.0

250.0

0.0325

7307.54

764

42828

2.73

Flue Gas To Stack Gas

Hot

565.6

150.0

0.1062

44118.78

42

312845

1.22

Combustible Air To Preheated Air

Cold

25.0

150.0

0.0853

10667.62

102

299742

1.02

HTS Outlet To Cooled LTS Feed

Hot

418.3

220.0

0.1539

30506.41

546

207402

2.67

LTS Outlet To Cooled PSA Feed

Hot

238.8

50.0

0.1774

33501.16

2074

207402

3.08

Cold T In (°C)

Cold T Out (°C)

Hot Stream

Hot T In (°C)

Hot T Out (°C)

Load (kW)

Area (m2)

dT Min Hot (°C)

253.3

539.4

SMR Outlet To Cooled HTS Feed

851.9

606.0

36159

519

312.6

352.7

25.0

250.0

Flue Gas To Stack Gas

263.3

194.5

7308

3269

13.3

169.5

Cold Stream SMR Feed To Heated SMR Feed NG Feed To Heated NG HP Steam Generation

253.3

254.3

HP Steam Generation Combustible Air To Preheated Air SMR Feed To Heated SMR Feed Combustible Air To Preheated Air

253.3

254.3

25.0

80.3

218.0

253.3

80.3

106.1

Cooling Water

32.2

48.9

HP Steam Generation

253.3

254.3

Combustible Air To Preheated Air

106.1

150.0

Outlet Temp (°C) 539.4

MCp (MW/°C) 0.1264

Enthalpy (kW) 40621.1

HTC (W/m2°C ) 479

Flowrate (kg/h) 183495

Effective Cp (kJ/kg-°C) 2.48

dT Min Cold (°C)

HTS Outlet To Cooled LTS Feed Flue Gas To Stack Gas

418.3

263.3

23844

844

164.0

10.0

565.6

263.3

32091

8859

311.3

10.0

Flue Gas To Stack Gas

194.5

150.0

4721

1368

114.2

125.0

263.3

234.3

4462

1554

10.0

16.3

234.3

220.0

2200

192

128.2

139.7

217.7

50.0

29754

335

168.8

17.8

606.0

353.9

37076

410

351.7

100.6

238.8

217.7

3747

393

88.8

111.6

HTS Outlet To Cooled LTS Feed HTS Outlet To Cooled LTS Feed LTS Outlet To Cooled PSA Feed SMR Outlet To Cooled HTS Feed LTS Outlet To Cooled PSA Feed

Production of Hydrogen via Syngas Route

7-39

CN 4120: Design II Team 32: Sin Yew Leong (U046835M)

HEN Unit Design Report

REFERENCE

1. Seider, W.D., Seader, J.D., Lewin, D.R. (2004). Product and Process Design Principles: Synthesis, Analysis and Evaluation. 2nd Ed., John Wiley & Sons, Inc. 2. F.W. Dittus and L.M.K. Boelter, University of California, Publ. Eng., 2, 443 (1930 3. Bell, K.J., & Mueller, A.C. (2001). Wolverine Engineering Data Book II, Retrieved March 27, 2008, from Wolverine Tube Inc. Web site: http://www.wlv.com/products/databook/databook.pdf 4. Hall, S.G., Ahmad, S., & Smith, R. (1990). Capital Cost Targets for Heat Exchanger Networks Comprising Mixed Materials of Construction, Pressure Ratings and Exchanger Types. Computers & Chemical Engineering, 14, 3, p. 319-335 5. Turton, R. et al (1998). Analysis, Synthesis and Design of Chemical Process. 2nd Ed., Upper Saddle River, NJ: Prentice Hall. 6. Economic Indicators. Chemical Engineering. Retrieved March 28, 2008 from Chemical Engineering Web site: http://www.che.com/business_and_economics/economic_indicators.html 7. Umeda, T., J. Itoh, J., & Shiroko, K. (1978). Heat Exchange System Synthesis. Chem. Eng. Prog., 74, 70. 8. Shenoy, U.V. (1995). Heat Exchanger Network Synthesis : Process Optimization by Energy and Resource Analysis. Houston: Gulf Pub 9. Linnhoff, B., & Hindmarsh E. (1983). The Pinch Design Method for Heat Exchanger Networks. Chem. Eng. Sci., 38, 745 10. Chemical Engineering Design. 4th Ed., Oxford: Elsevier Butterworth-Heinemann 11. Branan, C. R. (2002). Rules of Thumb for Chemical Engineers : A Manual of Quick, Accurate Solutions to Everyday Process Engineering Problems. 3rd Ed., Amsterdam; New York: Gulf Professional Pub 12. Perry, R.H., & Green, D.W. (1997). Perry’s Chemical Engineers’ Handbook. 7th Ed., McGraw-Hill 13. Muller-Steinhagen, H. et al. (2007). Recent Advances in Heat Exchanger Fouling Research, Mitigation, and Cleaning Techniques. Heat Transfer Engineering, 28, 3: pg. 173– 176

Production of Hydrogen via Syngas Route

7-40

CN 4120: Design II Team 32

Economics & Profitability Report

Chapter 8 : COOLING TOWER 8.1 PROBLEM STATEMENT

The cooling tower in this hydrogen plant is designed to provide a continuous flow of cooling water required for the condensation and elimination of water vapour in the outlet stream of low temperature shift (LTS) reactor, before it is fed into the pressure swing adsorption (PSA) column for purification of hydrogen and removal of carbon dioxide. The cooling duty of the tower is found to be 2.975 ×104 KW . In order to meet this requirement, an induced draft cooling tower with counter-flow pattern is designed. The detailed design of this cooling tower would consist of the following sections: •

Physical dimensions of the cooling tower

•

Cooling tower internals

•

Material of construction

•

Optimization and cost analysis

Production of Hydrogen via Syngas Route

8-1

CN 4120: Design II Team 32

Economics & Profitability Report

8.2 WORKING PRINCIPLES OF COOLING TOWER

Heat transfer in cooling towers occurs by two major mechanisms [1]: the transfer of sensible heat from water to air by convection process and the transfer of latent heat by evaporation of water.

Although sensible heat transfer due to temperature difference between the air and water occurs, the extent of this heat transfer is much smaller as compared to the removal of heat from water via latent heat of vaporization.(20% due to sensible heat:80% due to latent heat) The governing equation for the heat transfer in cooling tower is the Merkel Equation [2], defined to be −

KaV −

T1

dT h − ha T2 sa

=∫

L Cp

where Ka = volumetric air mass transfer coefficient ( lb air / hr ft 3fill ) _

2 V = specific fill volume ( ft 3fill / ft Base Area )

_

2 L = loading factor ( lb H 2O / ft Base Area )

hsa = enthalpy of saturated air at water temperature (Btu/lb dry air) ha = enthalpy of air stream (Btu/lb dry air) T1 = inlet water temperature ( o F ) T2 = outlet water temperature ( o F )

The derivation of this equation ignores the mass transfer resistance from bulk water to the interface, the effect of evaporation, and the temperature differential between the bulk water and interface. It demonstrates that the driving force for the cooling process is the enthalpy potential difference between the interfacial film and surrounding air.[3]

Production of Hydrogen via Syngas Route

8-2

CN 4120: Design II Team 32

Economics & Profitability Report

8.3 Preliminary Design

Before we can proceed with the actual design of the cooling tower, the different configurations for cooling tower must be examined for its advantages and disadvantages in order to make a choice the configuration that is going to be used in this design report.

8.3.1 Selection of cooling tower

Cooling towers can be classified according to the means by which air is supplied to the towers, i.e., natural draft vs. induced or forced mechanical draft (fans) and according to relative movement of air and water, that is, counter-flow or cross flow.

8.3.1.1 Justification to reject the use of natural draft tower

Natural draft or hyperbolic cooling tower depends on the natural draft created by the difference in the density of the entering and leaving air for movement. Due to their large sizes, they are often used for water flow-rates above 200,000gal/min.[3] Though it does not incur any operating or maintenance cost for fans and experiences almost no recirculation of hot air that could affect tower performance, it is not used in this design due to the following reasons: 1. The construction of hyperbolic cooling tower requires large plot space which results in higher capital investment on land 2. Natural draft tower depends completely on atmospheric conditions. This implies that water temperature is difficult to control and maintained, which might affect downstream units that utilize the cooling water. In our case, the affected unit will be the heat exchanger that makes use of cooling water to eliminate water vapout [4]

Production of Hydrogen via Syngas Route

8-3

CN 4120: Design II Team 32

Economics & Profitability Report

8.3.1.2 Justification to use induced draft tower

Mechanical draft towers can be either forced or induced draft, depending on the position of the fans. Forced draft towers have the fans located at the base of the tower which push the air through the fill while induced draft units have fans located on the top of the tower which pull the air through the packing and discharge them vertically upward at high velocity. Induced draft cooling tower is preferred over forced draft towers due to following factors:[5] 1. Unlike forced draft tower which is subjected to recirculation of hot humid discharged air into the fan intake, the recirculation issue is completely avoided in induced draft since the air is discharged upwards at high velocity. 2. The induced draft allows a more uniform distribution of air inside the tower. 3. The power requirement of the fan system in induced draft tower is about half that of forced draft tower for the same capacity. 4. Compared to forced draft tower, induced draft tower require less initial cost to start up, take up less space and have the capability to cool over wide range.

8.3.2 Comparison between counter-flow and cross-flow Pattern

For cooling towers, counter-flow pattern is preferred over cross-flow pattern mainly because: 1. For a tower of similar capacity, 20-50% less pumping head is required for counterflow cooling tower as compared to cross-flow tower. This implies that counter-flow tower can operate at a lower cost. [6] 2. Unlike cross-flow tower, counter-flow tower does not experience recirculation, which greatly reduces tower performance due to higher wet-bulb temperature. [6] 3. Under the same design condition, a counter-flow tower produces more cooling per unit volume at a lower cost.[6] Based on the above factors, counter-flow pattern is chosen for my design of cooling tower.

Production of Hydrogen via Syngas Route

8-4

CN 4120: Design II Team 32

Economics & Profitability Report

8.4 DETAILED DESIGN OF COOLING TOWER

(All calculations in this section is based on Reference 2) 8.4.1 Specification of cooling tower design parameters

8.4.1.1 Wet bulb temperature

Wet bulb temperature is the temperature of the air entering the cooling tower and is the lowest temperature at which water can be cooled to theoretically. Since a counter-flow induced draft tower is designed, it is valid to assume that there is no recirculation. Hence, the wet bulb temperature is taken to be that at ambient condition. Taking a conservative approach, the ambient wet bulb temperature is determined using the maximum dry bulb temperature in Singapore As of 2007,[7] Average daily maximum dry bulb temperature = 31.1 oC Mean relative humidity at 2 pm = 74% From the psychrometric chart, Ambient wet bulb temperature (t1) = 27.2 oC (81.0 o F )

8.4.1.2 Range

The range is defined to be the difference between the cooling tower water inlet (T1) and outlet temperature (T2). Heuristic [8] assumes the maximum inlet temperature of cooling water to be 120 o F (48.9 oC ) and cooling water exit temperature to be 90 o F (32.2 oC ) .

Hence, Range = T1 – T2 = 120 – 90 = 30 o F = 16.7 oC

Production of Hydrogen via Syngas Route

8-5

CN 4120: Design II Team 32

Economics & Profitability Report

8.4.1.3 Cooling water requirement

From the Heat Exchanger Network Design, the cooling water duty is found from simulation using HX-Net to be 2.975 ×104 KW . Hence, the amount of cooling water (L) required to achieve this cooling load can be calculated by Q

L=

C p (T1 − T2 )

where Cp = Effective heat capacity of water at temperature L=

T1 + T2 . 2

2.975 ×104 × 3600 = 1.488 ×106 kg / hr 4.320(48.9 − 32.2)

8.4.1.4 Approach Approach is defined to be the difference in temperature between the cooling water leaving the

tower and the ambient wet bulb temperature. As a general rule, the closer the approach to the ambient wet bulb temperature, the more expensive the cooling tower due to increased size. Approach = T2 – t1 = 90 – 81 = 9 o F = 5 oC This approach is very close to the typical approach of 10 − 15 o F in most cooling towers.

8.4.2 Exit air temperature and water to air flow ratio (L/G) 8.4.2.1 Exit air temperature

For a given set of cooling tower design conditions, an optimum design of the outlet air wetbulb temperature exists. This is desired as it will result in minimum construction and operating costs. A good correlation exists between the optimum exiting air temperature (t2) and the inlet and outlet cooling water temperature. This correlation which is to be used as a rule of thumb for design is as followed: t2 =

T1 + T2 120 + 90 = = 105 o F = 58.3 oC 2 2

Production of Hydrogen via Syngas Route

8-6

CN 4120: Design II Team 32

Economics & Profitability Report

8.4.2.2 Water to air flow (L/G) ratio L The ratio of a cooling tower is the ratio between the cooling water and air mass flow rate. G

Thermodynamically, the heat removed from the cooling water must be absorbed by the surrounding air. Hence, the following energy balance can be used to evaluate the

L ratio: G

LC p (T1 − T2 ) = G (h2 − h1 ) (h2 − h1 ) L = G C p (T1 − T2 )

where h2 = Enthalpy of air-water mixture at the exit air temperature (kJ/kg dry air) h1 = Enthalpy of air-water mixture at inlet air temperature (kJ/kg dry air) Cp = Heat capacity of water at 25 oC (h2 − h1 ) L (189.1 − 104.1) = = G C p (T1 − T2 ) 4.18(48.9 − 32.2)

= 1.219 Assumptions made: 1) Cooling water mass flow is relatively constant (little evaporation) 2) Sensible heat transfer from water to air is negligible

8.4.3 Cooling tower characteristic

Tower characteristic can be determined by the Chebyshev method [1], whereby −

KaV −

L Cp

T1

=∫ T2

dT T −T  1 1 1 1  ≅ 1 2 + + +  hsa − ha 4  ∆h1 ∆h2 ∆h3 ∆h4 

where hsa = Enthalpy of air-water mixture at bulk water temperature (Btu/lb dry air) ha = Enthalpy of air-water mixture at wet bulb temperature (Btu/lb dry air) ∆h1 = value of (hsa – ha) at T2 + 0.1(T1 – T2) ∆h2 = value of (hsa – ha) at T2 + 0.4(T1 – T2) ∆h3 = value of (hsa – ha) at T1 – 0.4(T1 – T2) ∆h4 = value of (hsa – ha) at T1 – 0.1(T1 – T2)

Production of Hydrogen via Syngas Route

8-7

CN 4120: Design II Team 32

Economics & Profitability Report

1 ∆h

hsa − ha

T, o F

hsa ( Btu / lb dry air )

ha ( Btu / lb dry air )

90

55.93

44.80

93

60.28

48.46

∆h1 = 11.82 0.0846

102

75.42

59.43

∆h2 = 15.99 0.0625

108

87.76

66.74

∆h1 = 21.02 0.0476

117

110.60

77.71

∆h1 =32.89

120

119.54

81.37

0.0304

**Note that the enthalpy of air-water mixture increases 1 Btu multiplied by L ratio for every G −

KaV −

L Cp

≅

1

o

F of cooling.

120 − 90 ( 0.0846 + 0.0625 + 0.0476 + 0.0304 ) ≅ 1.688 4

At this point, it should be noted that mechanical-draft cooling towers are usually designed for −

L/G ratios ranging from 0.75 to 1.50 and the values of

KaV −

vary from 0.50 to 2.50

L Cp consequently. Hence, the results obtained so far for the design of the cooling tower have been satisfactory.

8.4.4 Loading factor _

Loading factor ( L ), also known as specific water flow rate, is the recommended cooling water flow rate per unit volume of tower cross-sectional area. A general rule for loading factor is that for difficult cooling jobs (large cooling range and/or close approach), a lower loading factor is used and vice versa. A graphical method is presented to determine the loading factor.

Production of Hydrogen via Syngas Route

8-8

CN 4120: Design II Team 32

Economics & Profitability Report

Method to obtain the loading factor from the sizing chart 1. A straight line is drawn first to connect the inlet and outlet water temperatures as illustrated in the figure 2. Another line is drawn to intersect the first line at the wet-bulb temperature. This line. would yield the water concentration or the loading factor L

** Note that the loading factor determined from this graph is lower than loading factor used with presently-used fills. However, method for determining modern loading factor is proprietary information and is not available.

From the graph, _

L = 2.35

gal lbs = 1176 2 2 hr. ftbase min . ft Base Area area = 5746

kg 2 hr.mbase area

Assumption made: Density of cooling water does not change with temperature (1000 kg/m3)

Production of Hydrogen via Syngas Route

8-9

CN 4120: Design II Team 32

Economics & Profitability Report

8.4.5 Dimensions of Tower

8.4.5.1 Fill Height _

The required fill or packing height Z is assumed to be equal to the specific fill volume V in the Merkel equation. It can be calculated from the tower characteristic and loading factor. The fill height (z) can be determined from the following equation: −

−

KaV

L z=( − )calc × Ka L Cp

Past literature studies show that Ka value varied from 49 to 152 with 100 ± 30 as the average value. Since Ka is proprietary information, Ka is assumed to be 100 based on previouslydesigned cooling tower. Hence, z=

1.688 × 1176 = 19.85 ft = 6.05m 100

8.4.5.2 Base area

The required base area or cross-sectional area (B) can be determined from the below equation: B=

L −

=

L

1.488 × 106 = 260m2 5746

8.4.5.3 Fill volume

The fill volume (V) can be calculated by V = B × z = 260 × 6.05 = 1570 m3

Production of Hydrogen via Syngas Route

8-10

CN 4120: Design II Team 32

Economics & Profitability Report

8.4.6 Make-up Water Requirement

In the cooling tower system, water makeup is required to replace the cooling water which is lost through evaporation, drift and blow-down.

8.4.6.1 Evaporation loss (E)

The evaporation rate is assumed to be 1.0% of the water flow rate for each 10 o F temperature drop through the tower. Hence, evaporation loss is calculated as shown below:  T −T  E =  1 2  × 0.01× L( m3 / hr )  10 

1.488 × 106  120 − 90  = ) = 44.64 m3 / hr  × 0.01× ( 3 10  10 

8.4.6.2 Drift loss (D)

Drift is the entrained water in the tower discharge air. Drift loss is usually a function of the drift eliminator design and typically varies from 0.1 to 0.2% of the water supplied to the cooling water. Adopting a conservative approach, the drift loss is assumed to be 0.2% of the circulating cooling water. D = 0.002 × L(m3 / hr ) 1.488 ×106 = 0.002 × ( ) = 2.98 m3 / hr 103

8.4.6.3 Blow-down (B)

Blow-down is the continuous or intermittent discharge of a small amount of the circulating cooling water. Its purpose is to limit the increase in the solids concentration in the water due to evaporation. Since chlorides remained soluble in the cooling water, the blow-down rate can be determined from the cycles of concentration, which is the ratio of chloride content in the circulating water to the chloride content in the makeup water as shown in the equation.. B = E(

1 )− D Cycle − 1

Production of Hydrogen via Syngas Route

8-11

CN 4120: Design II Team 32

Economics & Profitability Report

Cycles of concentration in cooling tower operations typically range from 3 to 5 but for conservative design, it is assumed to be 3. Hence, B = 44.64(

1 ) − 2.98 = 19.34 3 −1

m3 / hr

8.4.6.4 Makeup water requirement (M) The makeup water requirement is the summation of evaporation loss, drift loss and blow-

down. M = E+D+B

= 66.96 m3 / hr

8.4.7 Power Requirement

Fans are essential for the function of induced-draft counter-flow cooling towers, so that air can be forced to flow vertically upwards to be in contact and cool the process water. As this circulating water is sprinkled down the tower by using nozzles, a pump is also required to pump the water to the top of the tower for cooling. These two auxiliary units are the main usage of energy for the operation of the cooling tower, and hence must be considered as a factor in the design.

8.4.7.1 Pump power (Pp)

Pump power is determined from the following equation

Pp =

L× Hp 1.98 × 106 ×η

where L = Water flow rate (lbs H2O/hr Hp = Pump head (ft) η = Fan Efficiency (dimensionless, assumed to be 0.80) H p = z + 10 = 19.85 + 10 = 29.85 ft = 9.10 m

Production of Hydrogen via Syngas Route

8-12

CN 4120: Design II Team 32

Economics & Profitability Report

Hence, Pp =

1.488 ×106 × 2.20462262 × 29.86 1.98 × 106 × 0.80

= 61.84 hp

8.4.7.2 Fan Power (PF)

An estimate of fan power requirement is obtained from the volume of moist air moved by the fan. For induced draft towers, this estimate is based on the exit air temperature which is 58.3 oC (105 o F )

At the air exit temperature, Saturated absolute humidity of the air-water mixture: H2 = 0.0507 lbs H2O/lb dry air Density of dry air:

ρ dry air =

3 42.6439 42.6439 = = 0.0755 lbs / ft t2 + 460 105 + 460

Density of water vapour:

ρ water =

3 26.6525 26.6525 = = 0.9304 lbs / ft ( t2 + 460 ) × H 2 (105 + 460 ) × 0.0507

Density of air-water mixture:

ρ mixture =

(1 + H 2 ) ( ρdry air × ρ water ) (1 + 0.0507 )( 0.0755 × 0.9304 ) = = 0.0734 lbs / ft ( 0.0755 + 0.9304 ) ( ρdry air + ρwater )

3

Air flow rate: L G = L( ) −1 = 1.488 ×106 × 2.20462262 × (1.219)-1 = 2.69 × 106 lbs / hr G Air flow rate (actual cubic feet of air per minute) F=

(1 + H 2 ) G = (1 + 0.0507 ) (2.69 ×106 ) = 6.418 ×105 acfm 60 ( ρ mixture ) 60 ( 0.0734 )

Assuming one hp is required for each 8000 actual cubic feet of air per minute moved by the fan, Fan power can be calculated by: PF =

F 6.418 × 105 = = 80hp 8000 8000

Production of Hydrogen via Syngas Route

8-13

CN 4120: Design II Team 32

Economics & Profitability Report

8.5 COOLING TOWER INTERNALS

8.5.1 Liquid Distributor

The liquid distributor serves the purpose of ensuring good liquid distribution by maintaining uniform flow of liquid through the column. The type of liquid distributor used is very much dependent on the flow pattern of the cooling tower. A pressure type system of closed pipe and spray nozzles, like the one seen in figure 1, is usually necessary for the counter-flow configuration.

Counter-flow distribution system in operation

Pressure spray system is more susceptible to clogging and more difficult to clean, maintain or replace. However, it contributes significantly to overall heat transfer and does not require high pump head in larger tower. As for cross-flow tower configuration, the gravity flow distribution system is more commonly used whereby the supply water is elevated to the hot distribution basin above the fill. From this basin, the water flow over the fill (gravity-induced) through metering orifices located in the distribution basin floor. Although this type of distribution system can be easily maintained, it does not contribute to overall mass transfer and tends to require a higher pumping head. There is also a tendency for algae growth if the basin is not covered.

Production of Hydrogen via Syngas Route

8-14

CN 4120: Design II Team 32

Economics & Profitability Report

8.5.2 Fill

The fill or packing is actually heat transfer surface that maximize water/air contact area and increase the contact time between air and water for more effective heat transfer. Fill is available in two types of design – the splash mode or the film mode. Splash fill normally consists of horizontal slats in horizontal rows offset to one another to cause the water to break up into droplets as it falls downward through the cooling tower. As such, maximum exposure between the water surface and passing air is achieved. Splash fill is characterized by reduced air pressure losses and is less conducive to clogging and can be cleaned easily after a spill. However, it is very sensitive to inadequate support and must remain horizontal and level. If sagging occurs, the water and air will channel through the fill in separate flow paths, impairing the thermal efficiency greatly. [9] On the other hand, film fill causes the water to flow in films over large vertical surface, thus promoting maximum exposure to air. Film fill has the capability to provide more effective cooling capacity within a given amount of space than splash fill. Film sheets are usually spaced very close to each other. Due to the smaller passages, film fill is more sensitive to plugging and makes the cleaning difficult if plugging do occurs. Hence in operations where contamination by debris is possible, film fill should be avoided. [9] 8.5.3 Drift Eliminators

Drift Eliminators, as the name suggests, are used to remove entrained water droplets (also known as drift) in the discharge air stream so as to prevent unnecessary water losses. The separation is achieved by subjecting the discharge air to sudden change in flow direction. Through the sudden directional change, a centrifugal force is created which cause the entrained water droplet to deposit on the eliminator surface, from which it will flow back into the tower. Drift eliminator exists in many configurations but are usually classified according to the number of directional changes or passes. More commonly found eliminators are the “herringbone” type or the “honeycomb” type with labyrinth passage as shown in the figure below.

Production of Hydrogen via Syngas Route

8-15

CN 4120: Design II Team 32

Economics & Profitability Report

“Honeycomb” type drift eliminator

Tighter control for drift release is expected nowadays due to possible environmental impact associated with it. Drift can cause Legionnaires disease when mist droplets containing the bacteria are inhaled into the body.

8.5.4 Supports

Although framework of cooling tower is already supported by massive cross-section, it is not unbendable. Operations of large fans at high horsepower can result in large torsional forces which could affect the stability of the tower. Therefore, it is necessary to construct a support to maintain the proper positioning of the mechanical equipment used. These supports can be in the form of heavy wall torque tubes welded to the outskirts of steel framework.

8.5.5 Cooling tower basin

The primary function of the cooling tower basin is to collect the cooled water leaving the tower and to provide a reservoir for the cooling water pumps. In addition, it also serves as the primary foundation for the tower and is also the collecting point for foreign materials washed out of the air by water. Hence it must provide easy access for cleaning, have adequate drainage facilities and be equipped with screening to prevent entry of debris into the suction side piping.

Production of Hydrogen via Syngas Route

8-16

CN 4120: Design II Team 32

Economics & Profitability Report

To enhance the reliability of cooling tower, a minimum storage capacity should be provided in the cooling tower basin to obtain the necessary time for corrective action during emergency. Design practices worldwide recommend a minimum storage of 10 minutes.

8.6 MATERIAL OF CONSTRUCTION

A cooling tower must be able to withstand long duration dead loads imposed by the weight of tower components, circulating water, snow and ice; as well as short term loads caused by wind, maintenance and even seismic activities. Its design should be able to accommodate a wide range of temperatures, a variety of external atmospheric conditions and internal pressure. Corrosion caused by oxygenation and high humidity should also be taken into account. Typical materials used are wood and steel. However, due to the above requirements of the cooling tower, wood would not deem as a very suitable material. This is because although wood is cheap and can last up to 30 years if it is well maintained, the drawback is that it is susceptible to fungal and bacteria attack. Moreover, fungal and bacteria attack are more prone to happen under the wet operating conditions of the cooling tower in a humid and wet tropical country like Singapore. As a result of these factors, galvanized carbon steel was chosen as the choice of material.

8.6.1 Liquid Distributor

Distribution systems are subjected to a combination of high temperature (hot water) and oxygenation which are conditions favourable for corrosion. Hence, the material of construction should be highly resistant to corrosion and erosion. Materials that are popularly used include hot-dip galvanized steel, cast iron and redwood stave pipe. Because of the relatively low pressure experienced by cooling tower piping, plastic can also be used for pipe and nozzles construction. These plastic pipes are then reinforced with fibre for mechanical strength.

Production of Hydrogen via Syngas Route

8-17

CN 4120: Design II Team 32

Economics & Profitability Report

8.6.2 Fills Treated wood lath (primarily Douglas Fir) is considered as material of choice for splash type

fill due to its strength, durability, availability and relatively low cost. However, plastic such as PVC which naturally has low flame spread rate is fast gaining popularity and dominance due to safety consideration (fire-resistance properties). This is especially the case in steel framed cooling water where fire-proofing is compulsory

Splash Type Fill – Wood Splash Bar

Splash Type Fill – Plastic Splash Bars

Film fill, on the other hand, can be made of any material that is capable of being fabricated or molded into shaped sheets, with a surface suitable for channeling of air and water. Currently, the most popular material is PVC because of its chemical inertness, good strength and light weight properties, low flame spread rate and most importantly, it can be molded to different shape easily. 8.6.3 Drift eliminator Similar to the fill, the material used for eliminator should be corrosion-resistant. In the

industry, treated wood and various plastic (predominantly PVC) are material acceptable for drift eliminator. 8.6.4 Mechanical support Traditional material used for these supports include carbon steel, hot dip galvanized after

fabrication or stainless steel at a significant additional cost. It is important to note that stainless steel is not necessary as the combination of heavy construction and galvanization is enough to meet the requirement for support.

Production of Hydrogen via Syngas Route

8-18

CN 4120: Design II Team 32

Economics & Profitability Report

8.7 COST ANALYSIS

8.7.1 Construction cost of for cooling tower

The number of tower units (TU) in a given cooling tower can be calculated by the following equation: TU = L( gpm) × Rating factor

The rating factor is a measure of cooling job difficulty and a linear correlation exists between the rating factor and the tower characteristic: −

Rating factor = 0.9964(

KaV −

)calc − 0.3843

L Cp = 0.9964(1.688) − 0.3843 = 1.298

Hence, TU =

1.488 × 106 × 4.40286754 (1.298) = 8504 1000

Cost of each tower unit is assumed to be US$14.45 (in 1978 dollars) Construction Cost (US $1978 ) = 14.45 × 8504 = US $122,879.97

To correct 1978 dollars to 2007 dollars, we need to know the CECPI value for 1978 and 2007 CECPI (2007) = 528.2 CECPI (1978) = 218.8 Construction Cost (US $2007 ) = US $1978 ×

CECPI (2007) CECPI (1978)

= US $122,879.97 ×

Production of Hydrogen via Syngas Route

528.2 = US $296, 641.68 218.8

8-19

CN 4120: Design II Team 32

Economics & Profitability Report

8.7.2 Operating Cost

The operating cost in the plant consists of two main components: the cost of makeup water and the utility cost (electricity) that arises due to fans and pump operation.

8.7.2.1 Cost of makeup water

The cost of process water is assumed to be US$0.067 / m3 Plant operation time is taken to be 8000hours/year Hence, Cost of makeup water = 0.067 × 66.96 × 8000 = US $35,890.56 / yr

8.7.2.2 Cost of Electricity

The total power requirement is the summation of pump power and fan power PTotal = PF + Pp = 62 + 81 = 143hp

Since electricity is sold at MW-h, the horsepower must be converted to MW-h 1 hp = 746 W

Hence, 143 hp =

143 × 746 × 3600 = 384MWh 106

Since electricity cost is at US$100/MW-h Electricity Cost = US $38, 400

Production of Hydrogen via Syngas Route

8-20

CN 4120: Design II Team 32

Economics & Profitability Report

8.7.2 Optimization between the operating and construction cost

Optimization of the cooling tower is performed to minimize the operating and construction cost of the cooling tower. The variable that is changed is the exit air temperature. From the calculations shown above, it is clear that when exit temperature air temperature differ, the −

KaV L ratio, the tower characteristic − , the tower dimensions and the power consumption by G L Cp fan and pump will vary. The parameter that will not change during varying exit air temperature is the makeup water requirement. Hence, it is not used as a guideline for optimization. Steps to perform optimization 1. An excel file is set up that contains all the equation used in the design of cooling tower. 2. The exit air temperature is changed, and the construction cost and the operating cost which is also the cost of electricity is monitored. 3. Five data point was tested, including the optimum temperature used as the initial guess and the results were plotted in a graph

60000

700000 C o n s tr u c ti o n C o s t

Operating Cost (US$)

50000 40000 30000 20000

600000 500000 400000 300000 200000 100000

10000

0 95

0 95

100

105

110

115

Exit Air Temp(o F)

Chart of Operating Cost (US$) vs. Exit Air Temperature

Production of Hydrogen via Syngas Route

100

105

110

Exit Air Temp (o F)

Chart of Construction Cost (US$) vs. Exit Air Temperature

8-21

115

CN 4120: Design II Team 32

Economics & Profitability Report

From these two plots, two observations can be made: 1. The operating cost, which is also the cost of electricity, decreases with increasing exit temperature until around 43.3 oC (110oF) where it increases with increasing exit temperature 2. The construction cost increases exponentially with increasing exit air temperature. At the current moment, it is very difficult to consider whether 43.3 oC is the optimum operating temperature because a lot of cost has not been factored into this optimization investigation. These costs include the cost of auxiliary units such as fans and pumps which is not included in the construction cost of the cooling tower, the cost for water treatment in term of buying the additives. With all these information, only then can we calculate the payback period to determine whether it is worthwhile to switch to 43.3 oC. Hence, I shall adhere to the calculations in the previous section.

8.8 ADDITIONAL CONSIDERATIONS TO COOLING TOWER DESIGN

8.8.1 Water Treatment

Cooling tower water treatment is necessary to minimize or eliminate: corrosion, scale and biological fouling of heat transfer surface (in heat exchanger) which is caused by impurities and minerals in the water. The difficulties caused by these impurities and minerals are summarized in the table found in Appendix A [4].

8.8.1.1 Corrosion control

Corrosion is an electrochemical process that deteriorates metals exposed to water in the presence of corrosive agents such as acids, oxygen, or bacteria. A common form of corrosion is pitting.

Production of Hydrogen via Syngas Route

8-22

CN 4120: Design II Team 32

Economics & Profitability Report

Some of the possible causes of corrosion include: 1. Process leaks into the cooling water 2. water flow velocity that is too low (causes deposits and fouling lending to corrosion) or too high (causes erosion and corrosion) 3. low pH and high temperature To reduce corrosion to an acceptable level, chemical corrosion inhibitors which form protective films on heat transfer surface are most effective. Examples of corrosion inhibitor include phosphates, organics, zinc, nitrites, and molybdate salts. Unfortunately, the use of chromate, which is a reliable corrosion inhibitor, is prohibited due to environmental constraints.[4] 8.8.1.2 Scale control

Scaling is characterized by the formation of hard, dense deposits on material surfaces. These deposits impact heat transfer. Calcium carbonate, which is formed from the reaction between calcium ions and bicarbonate, is the main scaling constituent. The key to prevention of scale formation in a cooling system is to maintain a reasonable water velocity and to use chemical additives (dispersants) in combination with blow-down to keep impurities concentration below the level which causes deposits. Equipment such as sand pressure filter which requires minimal operating and maintenance cost can also be used.

8.8.1.3 Biological control

Operating conditions in cooling tower are ideal for growth of biological matters. These conditions that encourage microbiological growth include favourable water temperature (20 to 50°C) and pH, continuous supply of nutrients and sunlight. If biological growth becomes uncontrolled and form large sticky agglomerations, it may lead to operating problems as listed below: 1. Fouling of heat transfer surfaces by bacterial slimes, resulting in flow restrictions and high process temperatures. 2. Reduced cooling tower efficiency due to algae, fungi, and bacterial slime growth in the water distribution basin and fill area of the cooling tower.

Production of Hydrogen via Syngas Route

8-23

CN 4120: Design II Team 32

Economics & Profitability Report

3. Corrosion 4. Clogging of water distribution nozzles. Oxidizing biocides such as chlorine or sodium hypochlorite chlorine can be used to control biological activity to prevent these operating problems from happening.

8.8.2 Environmental Concerns

Some of the environmental concerns with regards to cooling tower include: 1. Cooling tower blow-down is normally bypassed around major wastewater treatment and discharged with treated wastewater. 2. Noise emission from fans and from the flow of cooling water over the tower may require suppression if located near a community. 3. Spills and overflow of toxic and hazardous chemicals used for treatment of cooling water must be contained.

Production of Hydrogen via Syngas Route

8-24

CN 4120: Design II Team 32

Economics & Profitability Report

8.9 CONCLUSION In the sections above, the design of the cooling tower is carried out together with the cost

analysis and optimization. The specification obtained from the detailed calculations are summarized in the table below Cooling Tower Data Sheet

Tower Model

Induced Draft

Flow Pattern

Counter-Flow

Cooling Water Mass Flow Rate

1.488 x 106 kg/hr

Ambient Wet Bulb Temperature

27.2ºC

Exit Air Temperature

58.3ºC

On-Tower Cooling Water Temperature (Inlet)

48.9ºC (120 ºF)

Off-Tower Cooling Water Temperature (Outlet)

32.2ºC (90ºF)

Approach

5ºC

Range

16.7ºC

L Ratio G

1.219 −

Tower Characteristic

KaV

1.688

−

L Cp Loading Factor

5746

kg 2 hr.mbase area

Tower Dimension Fill Height, z

6.05m

Base Area, B

260 m2

Fill Volume, V

1570 m3

Make-up Water Requirement Evaporation (1% for every 10 ºF)

44.64 m3/hr

Drift Loss (0.02%)

2.98 m3/hr

Blow-down

19.34 m3/hr

Production of Hydrogen via Syngas Route

8-25

CN 4120: Design II Team 32

Economics & Profitability Report

Power Requirement Pump Head

9.10 m

Pump Power

61.84 hp

Fan Power

80 hp

Material of Construction Cooling Tower

Stainless Steel

Hot-Dip Galvanized Steel

Liquid Distributor Fill

PVC

Drift Eliminator

PVC

Mechanical Support

Hot-Dip Galvanized Steel

Cooling Basin

Concrete Cooling Tower Data Sheet

Production of Hydrogen via Syngas Route

8-26

CN 4120: Design II Team 32

Economics & Profitability Report

REFERENCES

1. Perry's Chemical Engineers' Handbook, Green, Don W. et al, McGraw-Hill, New York, 6th Edition, 1999, Chapter 12 2. Stephen A. Leeper, Wet Cooling Towers: ‘Rule-of-thumb’ Design and Simulation, U.S. Department of Energy; Idaho National Engineering Laboratory, EGG-GTH-5775, July 1981. 3. Cooling Towers: Design and Operation Considerations, retrieved on 28 March 2008 from Chemical Engineering Tools and Information website: http://www.cheresources.com/ctowerszz.shtml 4. G.B. Hill, E.J. Pring, Peter D. Osborn, Cooling Towers : Principles and Practice, London; Boston: Butterworth-Heinemann, 3rd edition, 1990 5. J. D. Palmer, P.E., C.E.M. Evaporative Cooling Design Guidelines Manual. 2002 [cited 2008 March 10]; Available from: http://www.emnrd.state.nm.us/ECMD/Multimedia/documents/EvapCoolingDesignManua l.pdf. 6. Donald R. Baker, Howard A. Shryock, A Comprehensive Approach to the Analysis of Cooling Tower Performance, Technical Bulletin R-61 P-13, reprinted from the Journal of Heat Transfer, August 1961. 7. Climate and Air Quality, in Yearbook of Statistics Singapore, National Environment Agency. p. 4 8. Seider W.D., Seader J.D., Lewin D.R. Product & Process Design Principles. Edition 2 9. John C. Hensley, Cooling Tower Fundamentals, SPX Cooling Technologies, Inc., Overland Park, Kansas USA, 2nd edition, 2006, retrieved on 1 April 2008 from SPX Cooling Technologies website: http://spxcooling.com/en/library/detail/cooling-tower-fundamentals/

Production of Hydrogen via Syngas Route

8-27

CN 4120: Design II Team 32

Economics & Profitability Report

APPENDIX A IMPURITIES FOUND IN COOLING WATER

CONSTITUENT

Hardness

CHEMICAL COMPOSITION Ca 2+ and Mg 2+ salts

DIFFICULTIES CAUSED

Form scale deposit on heat transfer surface

expressed as CaCO3 Alkalinity

Bicarbonate salts

Form calcium carbonate scales; attack

expressed as CaCO3

materials made of wood React with calcium in the water, forming

Sulphate

Sulphate ions SO4−

calcium sulphate deposits on condensers and coolers

Chlorides

Chloride ions Cl −

Silica

Reactive Silica SiO2

Add to dissolved solids content and increase corrosion potential of cooling water React with calcium, magnesium and iron that is in water to form silicate deposits Corrosion of copper and zinc alloys; Form

Ammonia

Ammonium ion NH 4+

complex ions with zinc component in corrosion inhibitors, rendering them ineffective; High concentration of dissolved solids causes

Dissolved solids

---

corrosion and increases precipitation of salts which form scale deposits on heat transfer surface

Suspended solids (undissolved

Settling occurs when velocity decreases; ---

matter) Oxygen and carbon dioxide Algae, bacteria, fungi etc

causes plugging, deposition in heat exchangers and enhance biological growth

O2, CO2 ---

Production of Hydrogen via Syngas Route

General corrosion and local pitting of all metal surfaces Organic growth and slime deposits

8-28

CN 4120: Design II Team 32

Economics & Profitability Report

Chapter 9 : ECONOMICS & PROFITABILITY 9.1 INTRODUCTION

It is important to carry out a cost analysis on the design and selection of equipment present in the chemical plant. Estimation of the initial cost of setting up the plant is done following the development of the process using simulation software such as HYSYS and MATLAB. Based on correlations and heuristics, we are then able to determine dimensions of our equipment and consequently an estimated cost. In addition to the cost of equipment, utilities cost can be obtained from HYSYS. A preliminary gauge of the plant start-up cost can then be obtained. Following the finalization of the plant design, we are able to approach a more realistic figure for the total cost, given the availability of price information regarding the major units and auxiliary equipment from the vendors. An economic analysis follows next. Such an analysis is crucial because it enables the management as well as the investors to assess the feasibility and profitability of the plan before deciding whether or not construction of the plant should take place. Potential returns of the plant are weighed against the risks that are involved. For this part of the design project, total capital investments and total operating cost are determined based on the correlation given by Turton [R1]. Revenues derived from the sales of hydrogen are computed.

9.2 ASSUMPTIONS

1. The plant is to be located on Jurong Island, Singapore 2. The plant land is rented instead of being purchased. Rent rate (based on per annum) is obtained from Jurong Town Council(JTC). 3. Operation time of the plant is 8,000 hours/year 4. Lifetime of the plant is 15 years, including construction time of 2 years.

Production of Hydrogen via Syngas Route

9-1

CN 4120: Design II Team 32

Economics & Profitability Report

5. Growth is at the same rate as the total cost of production 6. Salvage value for the plant is 10% of FCI. 7. All products of the steam methane reforming process are sold in the market. 8. The company is using its own profits to fund the investment cost of this project and no bank loan is taken. It is assumed that the company has sufficient cash flow to bear the investment cost and no interest would be paid to the bank. Hence hydrogen that is produced can be sold at a more competitive price in the market. In the event that the company wants to free up cash flow to sustain operations or make other investments, a bank loan can be considered. 9. Corporate tax in Singapore is given as 18%.

9.3 CAPITAL COSTS

The total capital investment in the chemical process plant is made up of two main components: the fixed capital and working capital [R1], i.e. Total Capital Investment = Fixed Capital + Working Capital

The fixed capital represents all costs associated with the construction of the plant. All fixed capital components are depreciable (except for land). The working capital represents initial investment required to finance the initial phase of the operation before revenues from the project starts. The working capital is usually used to pay wages, raw materials and contingencies. As the working capital must be recovered at the end of the project, it is a non-depreciable item on the cash flow statement.

9.3.1 Computations for Fixed Capital

A list of the fixed capital costs is shown in Table 9-1. The fixed capital investments include direct and indirect costs, costs for contingency and fee, as well as auxiliary facilities costs.

Production of Hydrogen via Syngas Route

9-2

CN 4120: Design II Team 32

Economics & Profitability Report

Fixed Capital costs Direct expenses

•

Equipment purchase cost

•

Materials used for installation cost

•

Labour cost resulting from equipment installation

Indirect expenses

•

Transportation, Insurance and taxes expenses

•

Construction Overhead

•

Contractor engineering expenses

Contingency and fee

•

Contingency – for use in unpredictable circumstances

•

Contractor fee

Auxiliary facilities

•

Land Purchase

•

Yard improvement

•

Auxiliary development

•

Offsite facilities and Utilities

Table 9-1: Items under Fixed Capital Costs The direct and indirect expenses can be expressed in terms of the Bare Module Cost (CBM). The CBM is the sum of all direct and indirect expenses incurred. To compute CBM, the following equation is used. CBM = C po FBM

(9-1)

where FBM is the Bare Module Cost Factor which accounts for the operating condition and the material of construction. Cop is the purchased cost for base conditions (equipment made of carbon steel and at atmospheric pressure)

Production of Hydrogen via Syngas Route

9-3

CN 4120: Design II Team 32

Economics & Profitability Report

Cop is given by Equation 9-2. log10 C op = K1 + K 2 log10 A + K 3 (log10 A) 2

(9-2)

where A is the capacity parameter of the equipment. When the capacity of the equipment lies outside the effective range of correlation, the smallest possible capacity is used for cost calculations. For towers that have larger volume than allowed, the costs are modeled as multiple columns in sequence. To account for inflation, the fixed capital costs are inflated using the following formula: C I = Cb I b

(9-3)

where Cb is the known cost in the base year when the index was (=397 in 2001) I (= 595.1 in December 2007) is taken to be the cost index in the year where the cost is C

To calculate the fixed capital costs, CAPCOST is employed. The costs for each equipment is present in Table 9.2

Production of Hydrogen via Syngas Route

9-4

CN 4120: Design II Team 32

Economics & Profitability

Table 9.2 Equipment Schedule

Section

Tag No. in PFD

Equipment Name

Function

R-103

D-101

Dimensions required for costing

LTS vessel

To further convert carbon monoxide and steam into hydrogen and carbon dioxide

1

Materials of construction Pressure Temperature Diameter Height

Low-Alloy Steel A387 25bar o 220 C 3.31m 5.07m

1,370,000

LTS knock-out drum

To remove condensate so as to prevent contamination of downstream catalyst

1

Materials of construction

Low-Alloy Steel A387

542,000

LTS

PSA

Cooling tower

V-101

V-102

Cost, CBM (USD)

Quantity

PSA columns

Cooling tower

To purify hydrogen gas

To provide cooling water as a source of cold utility

8

1

Subtotal for LTS section SS Clad Materials of construction 1bar-25bar Pressure o Temperature 50 C Diameter 3.00m Height 8.5m Subtotal for PSA section Materials of construction Carbon steel Pressure

1bar

Effective mass transfer area

733m

25,676,516

2

940,844

Heat exchangers

4,895,966

Expanders

4,476,000 Subtotal for Auxiliary units

Production of Hydrogen via Syngas Route

25,676,516

940,844

Subtotal for HTS section

Auxiliary units*

1,912,000

9-5

9,371,966

CN 4120: Design II Team 32

Economics & Profitability

*Auxiliary units details are on following page: Heat exchangers Tag No. to PFD

HX-101

HX-102 HX-103 HX-104 HX-105 HX-106 HX-107

Cold streams Expanded SMR Feed To Heated SMR Feed 1 Heated SMR Feed 1 To Heated SMR Feed 2 HP Steam 1 Generation HP Steam 2 Generation Preheated Air 1 To Preheated Air 2 Preheated Air 2 To Preheated Air 3 Cooling Water

Tube/shell designation

Hot streams

2

Tube/shell designation

Area (m )

Cost, CBM (USD)

Shell

Cooled LTS Feed 1 To Cooled LTS Feed 2

Tube

1553.7

1,629,379

Shell

SMR Outlet To Cooled HTS Feed 1

Tube

519.2

526,634

Shell

409.7

581,212

Shell

843.5

1,124,770

Tube

191.9

251,244

Tube

392.5

415,586

Tube

335

367,140

Cooled HTS Feed 1 to Cooled HTS Feed 2 HTS Outlet To Cooled LTS Feed 1 Cooled LTS Feed 2 to Cooled LTS Feed 3 LTS Outlet To Cooled Knockout Drum Feed 1 Cooled Knockout Drum Feed 1 to Cooled Knockout Drum Feed 2

Tube Tube Shell Shell Shell

Subtotal

Expanders Item E-101 E-101 E-105

Item SMR Feed Expander Steam Expander H2 Expander

Quantity 1

Material of construction SS

Cost, CBM (USD) 3,300,000

1

SS

360,000

1

SS

816,000 Sub Total

4,476,000

Total Bare Module Cost = USD 78,680,946

Production of Hydrogen via Syngas Route

9-6

4,895,966

CN 4120: Design II Team 32

Economic & Profitability

9.3.2 Computations for Total Module Costs

Besides accounting for the total bare module costs, it is also necessary to compute the contingency and fee costs so as to account for the total module costs. The contingency and contractor fee costs are assumed to be 15% and 3% of the bare module cost respectively. Adding these two costs to the total bare module cost will give the total module cost. Alternatively, the total module cost can be calculated from [R1]: n

CTM = 1.18∑ CBm , i

(9-4)

i =1

The contingency and fee costs are tabulated in the following table. Cost Item

Cost(USD)

Contingency

11,893,090

Contractor Fee

2,378,618

Total Bare Module Cost

93,558,977

Total Module Cost

107,830,685

Table 9-10: Table for contingency and fee costs

9.3.3 Computations for Grassroots Costs (FCI)

The grassroots cost of the plant is calculated by adding the auxiliary facilities costs to the previously calculated total module cost. The various auxiliary facilities costs is shown in Table 9-1. Since the plant designed is a new start-up, the grassroots cost is also equal to the fixed capital investment (FCI). As information on the various cost items is limited,grassroots cost is evaluated from [R1]: n

CGR = CTM + 0.50∑ C oBM ,i

(9-5)

i =1

= US$107,830,685 + 0.50 (79,287,268) = US$147,474,319

Production of Hydrogen via Syngas Route

9-7

CN 4120: Design II Team 32

Economic & Profitability

9.3.4 Computations for Working Capital

The typical amount spent on working capital ranges between 15% and 20% of the fixed capital investments, FCI. For conservative estimate, a value of 20% of fixed capital investment is assumed for the working capital. Therefore Working Capital

=

US$147,474,319 x 20%

=

US$29,494,864

9.4 MANUFACTURING COSTS

To estimate the manufacturing costs involved in this chemical plant, there are 3 categories of costs that are included. They are as follows: 1. Direct manufacturing costs: These costs represent operating expenses that vary with production rate. When product demand decreases, production rate is also dropped below the design capacity and there would be a decrease in the factors making up the direct manufacturing costs. 2. Fixed manufacturing costs: These costs are independent of changes in production rate. They include property taxes, insurance and depreciation. These costs are charged at constant rates even when the plant is not in operation. 3. General expenses: These costs represent an overhead burden that is necessary to carry out business functions. These include management, sales, financing and research functions. 4. Land lease cost: Since the land is rented, it will not be included in the fixed capital investment, but in the operating cost of the plant.

Production of Hydrogen via Syngas Route

9-8

CN 4120: Design II Team 32

Economic & Profitability

Cost item 1. Direct manufacturing costs, DMC

Value

Raw materials

CRM

Waste treatment

CWT

Utilities

CUT

Operating labor

COL

Direct supervisory and clerical labour

0.18COL

Maintenance and repairs

0.06FCI

Operating supplies

0.009FCI

Laboratory charges

0.15COL

Patents and royalties

0.03COM

Total DMC = CRM + CWT + CUT + 1.33COL + 0.03COM + 0.069FCI 2.Fixed Manufacturing Costs, FMC

Depreciation

0.1FCI

Local Taxes and Insurance

0.032FCI

Plant Overhead Costs

0.708COL + 0.036FCI

Total FMC = 0.708COL + 0.068FCI + Depreciation 3.General Expenses, GE

Administration Costs

0.177COL + 0.009FCI

Distribution and Selling Costs

0.11COM

Research and Development

0.05COM

Total GE = 0.177COL + 0.009FCI + 0.16COM 4. Land lease, CL Total Costs = CRM + CWT + CUT + 2.215COL + 0.190COM + 0.146FCI + CL + depreciation

Table 9-10: Components for Costs Of Manufacture

Production of Hydrogen via Syngas Route

9-9

CN 4120: Design II Team 32

Economic & Profitability

9.4.1 Operating labour costs, COL

The number of operators per shift, NOL, is calculated using the following equation: N OL = (6.29 + 31.7 P 2 + 0.23 N np )0.5

(9-6)

where P is the number of processing steps involving the handling of particulate solids and Nnp is the number of non-particulate processing steps handing steps and includes compression, heating and cooling, mixing, and reaction. In general Nnp is given by: N np = ∑ Equipment

(9-7)

where equipment comprises of compressors, towers, reactors, heaters and exchangers, and excludes pumps and vessels. No. of towers

9

No. of reactors

3

No. of heaters

1

No. of exchangers

10

Nnp

23

Table 9-11: Number of equipment Since the plant does not handle particulate solids, P = 0. Therefore, NOL= (6.29 + 31.7(0)2 + 0.23(23))0.5 = 3.40 The following assumptions are made when calculating COL. •

The plant is operating 8000hrs/yr = 47.6weeks/yr

•

Each operator works 5 shifts per week, and each shift is 8 hours, thus an operator works 47.6 × 5 = 238 shifts per year

•

Assuming plant operates 24hrs/day, there are 3 shifts in a day. Total number of shifts per year = 8000/8 = 1000 shifts/yr

•

Number of operators needed to provide this number of shifts = 1000/238 = 4.2 operators

Production of Hydrogen via Syngas Route

9-10

CN 4120: Design II Team 32

Economic & Profitability

Hence, operating labour needed = 4.2 × NOL = 14.3 ≈ 15 COL (2001) = 15 × US$50,000 = US$750,000 /yr COL (2007) = US$1,004,691/yr

9.4.2 Utility costs, CUT 9.4.2.1 Electricity

Below is the calculation for the electricity cost for cooling tower. Cooling tower

Electricity (hp)

Pump power

173.95

Fan power

225.87

Power

399.82

Table 9-12: Total Power for Cooling Tower Below is the calculation for the electricity cost for furnace. Furnace

Electricity (hp)

Induced draft fan

25

Forced draft fan

30

Power

55

Table 9-13: Total Power for Furnace Total power (hp)

Total power (MW)

Cost of electricity

Total annual cost

454.82

0.339

US$100/MWh

US$ 271,200

Table 9-14: Total power for plant Total electricity cost per annum (2007) = US$ 271,200

Production of Hydrogen via Syngas Route

9-11

CN 4120: Design II Team 32

Economic & Profitability

9.4.2.2 Cooling water cost

The cost for cooling water used in the cooling tower and the heat exchangers are calculated as follows: Water makeup

Volume (m3/hr)

Evaporative Loss

124.99

Drift Loss

8.34

Blowdown

54.15

Total water makeup

187.49

Table 9-15: Process water used for cooling tower Given cost of water is 0.067 US$/m3, Cost of water used for cooling tower = $100,494 Total volume of water (m3/hr)

Cost of water US$/m3

Total annual cost (US$/yr)

4280.71

0.067

2,258,570

Table 9-16: Process water used for heat exchangers Volume of water used for heat exchangers = 4280.71m3/hr Cost of cooling water used for heat exchanger (2007) = US$2,258,570 CUT

=

US$(271,200 + 2,258,570 + 100,500)

=

US$2,630,270

9.4.2.3 Waste treatment costs, CWT

In this plant, the waste water coming out of the LTS knockout drum should be treated. The volume of wastewater flowing out of this knockout drum is 92.23m3/hr. The cost of treating this water is US$41/1000m3 of waste water [R1].

Cost of treating LTS knockout drum waste water (2001) = 92.23 × 8000 ÷ 1000 × 41 = US$30,250

Production of Hydrogen via Syngas Route

9-12

CN 4120: Design II Team 32

CWT

Economic & Profitability

=

US$(30,250 × (595.1 ÷ 394.3))

=

US$45,657

9.4.3 Raw materials costs, CRM

The raw materials used in this plant are natural gas and high pressure steam. Section of Raw Consumption plant material Natural gas 2603 kmol/h Feed to plant HP steam 23910 kg/h HTS reactor HP steam 140700 kg/h To SMR Table 9-17: Raw materials used for the plant

Unit

Type of catalyst

HTS reactor

Cr iron oxide catalyst Cu Zn oxide LTS reactor catalyst PSA reactor Activated carbon and zeolite 5A catalyst Table 9-18 Cost for catalysts CRM

Cost

US$7.59/kmol US$33/tonne US$33/tonne

Annual cost of material (2007) US$158,054,160 US$6,312,240 US$37,144,800

Cost of catalyst US$109414

Lifespan of catalyst 3 years

Annual cost of catalysts (2007) US$36470

US$226800

3 years

US$75600

US$372372

3 years

US$124124

=

US$(158,054,160 + 6,312,240 + 37,144,800 + 36,470 + 75,600 + 124,124)

=

US$ 203,751,994

9.4.4 Land lease, CL

The land area required per year = 29400m2 From Jurong Town Corporation, the price of land rental at Jurong Island is $11.87psm per year. CL

=

$29400 ×11.87 ÷1.3577

=

US$257,036

Production of Hydrogen via Syngas Route

9-13

CN 4120: Design II Team 32

Economic & Profitability

9.4.5 Computation of manufacturing costs

To calculate COM, cost of manufacture, the following equations were used: DMC = CRM + CWT + CUT +1.33COL + 0.069FCI + 0.03COM FMC = 0.708COL + 0.068FCI + depreciation GE = 0.177COL + 0.009FCI + 0.16COM Thus, adding all 4 components together, we can solve for COM: COM = 0.180FCI + 2.73COL + 1.23(CUT + CWT + CRM) + CL + depreciation Cost item 1. Direct manufacturing costs, DMC Raw materials Waste treatment Utilities Operating labour Direct supervisory and clerical labour Maintenance and repairs Operating supplies Laboratory charges Patents and royalties Total DMC = US$226,559,293

Cost (US$)

203,751,994 45,657 2,630,264 1,004,692 180,845 8,848,459 1,327,269 150,704 8,938,259

2.Fixed Manufacturing Costs, FMC

Local Taxes and Insurance Plant Overhead Costs Total FMC = US$10,662,888

4,719,178 6,020,397

3.General Expenses, GE

Administration Costs Distribution and Selling Costs Research and Development Total GE = US$49,071,189

1,505,099 32,773,615 14,897,098

4. Land lease, CL = US$257,036

Total Manufacturing Costs, COMd (without depreciation) = US$301,540,961 Table 9-19: Computed values for Costs Of Manufacture

Production of Hydrogen via Syngas Route

9-14

CN 4120: Design II Team 32

Economic & Profitability

9.4.6 Salvage value

Salvage value, S, represents the fixed capital investment of the plant minus the cost of the land, at the end of the plant life. Assuming the salvage value of the property at the end of service life is 10% of FCI. Salvage Value, S

=

10% of FCI

=

US$14,747,432

9.4.7 Depreciation

Depreciation is the reduction in value of equipment due to physical deterioration. Depreciation is calculated using the straight line depreciation method as follows: Depreciation , dk

=

[FCI-S]/n

=

(147,474,319 - 14,747,432)/15

=

US$8,848,459/yr

9.4.8 Revenues Steam revenue

Steam is generated from the heat exchangers in the plant. Steam generation revenue (2007) = US$67,987,243/yr Electricity revenue

Below is the table showing the electricity generated by the expanders in this plant. Assuming turbine efficiency = 70%, Component

Duty (kW)

Electricity generated (kW)

Revenue (US$)/yr

SMR feed expander

1787

1251

1,000,800

Steam expander

41.28

28.90

23,120

H2 expander

287.2

201.04

160,832

Total Revenue

1,184,752

Table 9-20: Total Electricity Generated Revenue Electricity generated revenue (2007) = US$1,184,752/yr

Production of Hydrogen via Syngas Route

9-15

CN 4120: Design II Team 32

Economic & Profitability

Hydrogen revenue Component

Output (m3/yr)

Output (kg/yr)

Unit Price (US$/kg)

Hydrogen

1250,000,000

110,925,000

2.43

Total Revenue

270,024,351

Table 9-21: Total Hydrogen Generated Revenue Total revenue from Hydrogen, Steam and Electricity =

US$ 67,987,243 + 1,184,752 + 270,024,351

=

US$ 339,196,346

9.5 PROFITABILITY ANALYSIS

To find out the profitability and feasibility of a designed plant, an analysis of the cash flow diagrams will be useful. Discrete and cumulative cash flow diagrams provide a clear insight to the investments and profits which are made for every year of the plant project. The time value of money is also important concept for assessing the profitability of a plant. The value of money differs with time due to the earning capability of the money. The difference in the value of money with time is not due to inflation and does not include the purchasing power of money. This is an important concept as the designed plant is to operate over a span of several years and it is hence more accurate for our profitability analysis if the time value of money is taken into consideration.

In this section, the designed plant is considered to be built over 2 years and a plant life of 13 years follows, making up a total of 15 years for economic analysis. The cumulative cash flow diagram for this project is studied. At the same time, the price at which the produced hydrogen is to be sold so as to obtain a 10% return on the investment on the plant over a total payback period of 15 years will be investigated.

Production of Hydrogen via Syngas Route

9-16

CN 4120: Design II Team 32

Economic & Profitability

9.5.1 Land Cost

The land cost is usually an investment that is made to the plant at the building stage of the plant. The piece of land may be rented for the purpose of building and running the plant over its construction + operation lifespan, which is taken to be 15 years for this project. In this case, the land is being rented for the plant to be located in Jurong Island, Singapore. Jurong Town Corporation, JTC has provided the rental cost which is US$8.74/m2 for each rental year.

Taking account of the dimensions of the various units as well as the amount of land required for auxiliary facilities, the total required land area of 29400m2 for our plant will cost US$257,036/yr. This amount of yearly rental is to paid from the construction of the plant till the end of the plant operating life.

9.5.2 After Tax Cash Flow

The sales revenue made from the produced hydrogen, electricity and steam is not the profit made to the plant due to expenses such as manufacturing costs and depreciation as well as payable income tax. The net cash inflow to the plant or cash profit to the plant is thus after tax cash flow here the cost of manufacture, cost of land rental, depreciation and income tax have been taken into account.

The price of the hydrogen would be determined in the profitability analysis to assure a rate of return of 10% to the plant. Since the plant is to be built and operated in Singapore, the taxes payable would follow the corporate tax regulations on profits in Singapore by the Inland Revenue Authority of Singapore, IRAS. The tax regulations state that corporate tax rate would be at 18% from 2008 onwards and tax exemption is only valid for companies with income lesser or equal to S$300,000 [2]. Since the positive income generated by the plant yearly is more than S$300,000, a tax rate of 18% will be imposed on the designed plant. t=18%

Production of Hydrogen via Syngas Route

(9-8)

9-17

CN 4120: Design II Team 32

Economic & Profitability

This tax scheme is used to calculate the amount of tax payable. As income refers to the net income to the organization, the percentage tax should be multiplied to the net value of revenue taking away expenses which are the cost of manufacture without depreciation. Income tax = (R - COMd –d)t

(9-9)

The after tax cash flow to a plant is therefore After tax cash flow = (R - COMd –d) (1- t) +d

(9-10)

9.5.2 Rate of Return on Investment (ROROI)

The rate of return on investment of a project can be determined from the ratio of the average net profit to the fixed capital investment excluding land cost. The average net profit can be obtained from averaging the cumulative cash flow at the end of the project (after 15 years). As the ROROI refers to the rate of return on investment in a project, a positive ROROI is expected for a feasible project, hence the ROROI should ideally be as high as possible. ROROI has been set to be 10% for this project.

9.5.3 Net Present Value (NPV)

The net present value is the cumulative discounted cash flow at the end of the project. A non-negative NPV is required for a feasible project and hence is assumed to be zero. A nonnegative NPV indicates that the plant would at least have a rate of return that is equal to the discount rate used in the calculation. For this work, we employ a NPV of 0 to back-calculate the price of hydrogen which we should be selling at in the market.

9.5.4 Discounted Cash Flows in Project

Keeping the above cash flows of the project in mind, a cash flow table of the project can be determined once the selling price of the hydrogen is set. To ensure a 10% return on the

Production of Hydrogen via Syngas Route

9-18

CN 4120: Design II Team 32

Economic & Profitability

investment for the discounted cash flow for a payback period of 15 years, a price of the hydrogen is set such that the return on investment for the discounted cash flow is 10% and taking the cumulative discounted cash flow at the end of the whole project to be zero. The ‘Goal-Seek’ function under Microsoft Excel is employed for this purpose. Table 9-22 shows the cumulative discounted cash flows for our designed plant.

End of year, k

Investment

Depreciation

Gross Profit

After tax cash cash flow

Discounted Cash Flow

Cumulative Discounted Cash Flow

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-260599 -88484591 -88484591 0 0 0 0 0 0 0 0 0 0 0 0 44242296

8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459 8848459

0 0 0 28209037 28209037 28209037 28209037 28209037 28209037 28209037 28209037 28209037 28209037 28209037 28209037 72451333

-260599 -88484591 -88484591 24724133 24724133 24724133 24724133 24724133 24724133 24724133 24724133 24724133 24724133 24724133 24724133 61002816

-260599 -80440538 -73127761 18575607 16886916 15351741 13956129 12687390 11533991 10485446 9532224 8665658 7877871 7161701 6510637 14603589

-260599 -80701137 -153828898 -135253291 -118366375 -103014634 -89058505 -76371116 -64837125 -54351679 -44819455 -36153797 -28275927 -21114226 -14603589 0

Table 9-22: Cummulative discounted cash flows By fixing cumulative discounted cash flow to be 0 at the end of 15 years, a selling price of US$2.43/kg is obtained via the ‘Goal-Seek’ function.

Production of Hydrogen via Syngas Route

9-19

CN 4120: Design II Team 32

Economic & Profitability

Fig 9-1: Cumulative Discounted Cash Flow Plot As we can see from the graph above, a selling price of $2.43/kg for the hydrogen product will enable us to get a cumulative discounted cash flow of zero after 15 years (2 Construction years + 13 Operation years). We next study the price of hydrogen which we have to sell at, if we are concerned with a shorter discounted payback period, as illustrated in Fig 9-2. As we can see from figure 9.2, if we would like to recover our capital investment just after 1 year of operation, we will need to sell hydrogen at US$4.19/kg. An important assumption made here is that all hydrogen that is produced will be sold and there will be no leftovers. Given more years for the returns, the price of hydrogen can be set at a lower rate. This is better overall so that hydrogen can be priced more competitively, in view of numerous other hydrogen providers in the market.

Production of Hydrogen via Syngas Route

9-20

CN 4120: Design II Team 32

Economic & Profitability

Fig 9-2: Price of Hydrogen over the number of years of operation

9.6 FEASIBILITY OF STORAGE FACILITIES FOR NATURAL GAS FEED

This portion of the report addresses the feasibility of an installation of storage tanks for the natural gas feed, as a recommendation for future consideration. For our design at present, the natural gas feed has been assumed to be provided by vendors via pipes and on top of this, there will be no interruption in the provision of natural gas by the vendors. Product hydrogen is not stored and exported immediately once produced, since we have earlier made the assumption that every unit quantity of hydrogen will be able to find its customer in the market.

Production of Hydrogen via Syngas Route

9-21

CN 4120: Design II Team 32

Economic & Profitability

Realistically speaking however, it is unlikely that there will be perfect provision of natural gas without any failures everyday around the year. 2 possible scenarios of failures in the provision of natural gas are identified: 1. Temporary complete-termination of natural gas provision due to errors at vendor’s end 2. Fluctuations (less/more than the contracted amount) in the natural gas provision Both scenarios result in potential losses for the plant. Hence installation of storage tanks for stand-by natural gas feed can be a solution around these problems. In making a cost analysis of the proposed storage tanks, the following assumptions are made: 1. Interruptions in natural gas provision do not last not more than 3 days, hence designs for 3 days’ worth of natural gas feed are made 2. Should there be any identified emergency in natural gas provisions, stand-by natural gas from the storage tanks will be utilized with immediate effect and there is negligible delay associated with the operation of control components 3. Heaters, coolers and piping constitute 20% of tank equipment cost 4. Tanks are of floating-roof type Given that 2603 kgmole/h gaseous natural gas feed is provided by the vendor at 25oC & 40 bar, the following has to take place in order to store liquid natural gas in the storage tanks:

Fig 9-3: Storage process of natural gas feed

Production of Hydrogen via Syngas Route

9-22

CN 4120: Design II Team 32

1.

Economic & Profitability

Natural gas feed at 25oC will be cooled to -87oC to transform the gaseous feed into liquid form to facilitate storage in the tank. Temperature at which the gaseous feed will liquidize, i.e. -87oC has been verified with HYSYS.

2.

In the event of an emergency such that the stand-by natural gas has to be employed, the liquidized form of natural gas will be heated up slowly and gradually, to avoid the potential hazard of rapid expansion due to vaporization, and rupturing of the pipes as a result. Natural gas in gaseous form can then be fed into the furnace.

The following calculations are made: Feed rate of Methane Feed(kg/h)

42830.00

Density of Liquid Methane (kg/m3)

422.62

Volume of Liquid Methane (m3/h)

101.34

Total volume stored (m3)

7296.77

Table 9-23: Calculations for storage volume Total volume of liquid methane (natural gas) to be stored = 7296.77m3 (3 days’ worth) 20% volume allowance has been made for the vaporization of liquid methane within tanks, as well as the innage/outage of the tanks. Total volume needed = =

7296.77 x 1.2 8756.12m3

Designing each tank to have a dimension of 55m (Diameter) by 35m (Height), volume of each tank = 3022.25m3 Hence 3 tanks are necessary.

Production of Hydrogen via Syngas Route

9-23

CN 4120: Design II Team 32

Economic & Profitability

A proposed plot plan of the storage tanks is as follows:

Fig 9-4: Plot plan of storage tanks for natural gas feed

9.6.1 Capital Costs

Initial capital cost will constitute the 3 tanks as well as piping, heater and cooler. Based on CAPCOST,

Cost for 1 tank

=

US$431,000

Cost for 3 tanks

=

US$431,000 x 3

=

US$1,293,000

Cost for piping, heater and cooler has been taken to be 20% of tank costs. Hence their cost

Total capital cost

=

US$1,293,000 x 20%

=

US$258,600

=

US$1,293,000 + 258,600

=

US$1,551,600

Production of Hydrogen via Syngas Route

9-24

CN 4120: Design II Team 32

Economic & Profitability

9.6.2. Operating Costs Land Area needed (m3)

Unit cost for land (US$/m3.yr)

Land Cost (US$/yr)

21,025

8.84

185,829

Table 9-24: Operating Costs Total duty (MWh)

Unit cost for electricity (US$/MWh)

Total Electricity cost (US$/yr)

885

100

88,520

Table 9-25: Electricity Costs

9.6.3 Overall Costs

Cost for the 1st year constitutes the initial capital cost in addition to land rental cost and utility (electricity) cost. The remaining years in the course of plant operation will only involve the operation costs, i.e. land rental costs and utility cost.

Assuming a storage facility is erected together with the construction of the plant, starting from Year 0 that is, the cost breakdown is as follows: Cost for 1st year Cost for 2nd year Cost for 3rd year Cost for 4th year Cost for 5th year Cost for 6th year Cost for 7th year Cost for 8th year Cost for 9th year Cost for 10th year Cost for 11th year Cost for 12th year Cost for 13th year Cost for 14th year Cost for 15th year Total Cost for all 15 years

1,825,949 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 274,349 5,666,836

US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$ US$

Table 9-26: Costs for storage facilities over 15 years

Production of Hydrogen via Syngas Route

9-25

CN 4120: Design II Team 32

Economic & Profitability

Making a conservative allowance of 40% to take care of any unforeseen expenses, total cost of such a storage facility

=

US$5,666,836 x 1.4

=

US$7,933,570

From Section 9.3.3, our fixed capital investment(FCI) for the main plant has been determined to be at US$147,474,319. If we were to build a storage facility to last for 15 years, it will constitute only (7,933, 570/147,474,319) x 100% =

5.4% of FCI

From an economic point of view, 5.4% is a fairly small percentage. From this cost analysis, it can be seen that erection of the storage facility can be done at a cost which is only around 1/20 times of the capital investment made into the parent steam-methane reforming plant, rendering the former feasible as a project. As mentioned, the presence of such a storage facility for stand-by feed will to some extent, provide the much-desired security if the plant management is concerned with a steady provision of natural gas.

Similar to the main plant, a SHE analysis on the installation of storage tanks will have to be conducted. This is especially necessary as the tank contents involve methane which is a highly flammable substance. In view of the potential hazards associated with the feed, it is recommended that the storage facility be sited at a safe distance away from the main steamreforming plant.

9.6.4 Economic Compensation

Despite the attractiveness of a feed-storage, for our design we have assumed provision of natural gas via piping with no failures on any day of the year.

Production of Hydrogen via Syngas Route

9-26

CN 4120: Design II Team 32

Economic & Profitability

To back the decision of employing no storage tanks in our design, we have made the following important assumption: Should there be any failure in feed-delivery, compensations can be sought from the vendor

This is in view of the potential losses that can result from the inavailability of the natural gas feed. Such compensations due to disruptions in service should have already been pre-agreed upon as part of the terms of contracts between the plant management and their vendors.

However, the economic losses that arise from such a undesirable scenario can be difficult to quantify sometimes. Plant management may consider employing the services of the financial specialists to come up with a closer estimation to the actual damages suffered from a disruption in feed supplies.

9.7 RECOMMENDATIONS

Based on the preliminary economic analysis, in order for the plant to break even, the plant has to sell its hydrogen at US$2.43/kg. This analysis assumes a required rate of return of 10%, 18% corporate tax on income, 15 years discounted payback period and that all products of this steam-reforming operation will be able to be sold in the market. A more detailed analysis will encompass the inflation factor, as it is expected that the company will have to increase the price of hydrogen over the years, following the expected increase in operating costs, i.e. cost of utilities as a result of increases in crude oil prices, increased rent cost etc. Pricing of the hydrogen heavily depends on the pricing strategies of other competitors in the market, and in the event the minimum price which hydrogen should be priced at (i.e. US$2.43/kg for our case) is significantly lower than the market average, say US$2.80/kg, the company may want to sell its hydrogen at a higher price. Every cent increase in hydrogen prices generates corresponding increase in profits as well as reduces the number of years which

Production of Hydrogen via Syngas Route

9-27

CN 4120: Design II Team 32

Economic & Profitability

the company needs to break even with the initial capital cost. Taking for example, if the company chooses to price its hydrogen at US$2.70/kg, the company may actually require only 7 years for cumulative discounted cash flow to reach zero, as opposed to the 15 years necessary if hydrogen is priced at US$2.43/kg. Hence, we can see that pricing of hydrogen depends a great deal on the strategy which the company would like to employ: 1. Does the company want to gain substantial market share in the market by selling at a rate way below market average? 2. Does the company want to recover all its investments in a shorter period by selling hydrogen at a higher price, while still maintaining below market price?

A way to reduce the cost of manufacture, so as to increase the profit margins, is to reduce the utility costs. Cogeneration facilities can be implemented so as to be self sufficient in the production of electricity. However, the feasibility of this recommendation has to be determined through detailed calculations of the fixed capital investment costs of these facilities. On computations of the equipment costs, different vendors may have different quotes for the required equipments. Therefore, rather than relying on simplified correlations (such as using CAPCOST) to find the costs, it may be better to refer to the available catalogue from the vendors for more current and realistic cost estimations. This is made possible with the advanced telecommunication and readily available information from vendors working with firm’s parent company. Assuming that the designed plant is already in operation, in the event that all the above recommendations do not lead to an accurate economic analysis that result in a positive NPV, the firm can consider forming strategic alliance with some other firms that have prior experiences in building similar plants, to reap the internal and external economies of scales possible with such an alliance. Alternatively, the company may consider approaching Chemical Engineering Consultancies for suggested improvements on the plant designs and operation procedures.

Production of Hydrogen via Syngas Route

9-28

CN 4120: Design II Team 32

Economic & Profitability

Financial or risk-management consultancies may be approached for a better estimation of present and future financial status for the company. Should there be a definite forecast of unhealthy financial status for the company in time to come, a re-assessment of the pricing strategies and operation contingencies may be crucial.

9.8 CONCLUSION

Based on the assumptions made, the preliminary economic analysis shows that it is possible to build a profitable steam-methane reforming plant. If we are concerned with a required rate of return of 10% where payback period is 15 years, we can price our hydrogen at US$2.43/kg, which is close to the suggested range ($1.90/kg to $2.30/kg) of hydrogen prices

provided by NETL (National Energy Technology Laboratory) [3]. Our price is obtained using the Goal-seek function under Microsoft Excel by equating Cumulated Discounted Cash Flow in the 15th year to zero. Any pricing of hydrogen above this rate increases our profit margins, reducing the number the years which full returns can be realized. Again, here we are making the assumption that any hydrogen produced will be able to find its customer in the market. Competitiveness of the price, as mentioned, depends largely on the pricing strategies of other hydrogen vendors in the market. Factors that will make this economic analysis a more realistic study have been mentioned under the recommendation section. As highlighted earlier, making an accurate assessment of these factors in our work is beyond our expertise and hence employing finance specialists to perform these complicated tasks will be the recommended option. In addition, chemical engineering consultants can be approached for suggestions and further optimization on the overall design and operations. In any event of a disruption in natural gas provision via piping, compensations can be sought from the vendors, and availability of a legal support may reduce the complications that arise from such a scenario.

Production of Hydrogen via Syngas Route

9-29

CN 4120: Design II Team 32

Economic & Profitability

REFERENCES

[R1]: Turton R., Bailie R.C., Whiting W.B. and Shaeiwitz J.A., “Analysis, Synthesis, and Design of Chemical Process”, Prentice Hall, 2003

[R2]: Inland Revenue Authority of Singapore, http://www.iras.gov.sg/, retrieved on 12 April 2008.

[R3]: National Energy Technology Laboratory , http://www.netl.doe.gov/technologies/hydrogen_clean_fuels/systems_studies.html, retrieved on 12 April 2008

[R4]: Guthrie, K. M., “Data and Techniques for Preliminary Capital Cost Estimating”, Chem. Eng., 1969

[R5]: Perry, R.H., and Green, D.W., “Perry’s Chemical Engineers’ Handbook”, 7th Edition, McGraw Hill, New York, 1997.

[R6]: Peters, M. S., Timmerhaus, K. D., “Plant Design and Economics for Chemical Engineers” McGraw Hill, USA, 1968.

[R7]: J.R. Couper et al., “Chemical Process Equipment: Selection and Design”, Elsevier, 2005.

[R8]: R.K. Sinot, “Coulson and Richardson’s Chemical Engineering”, Vol. 6, 4th Edition, Oxford 2005.

[R9]: Jurong Town Council. JTC's Land Rents and Prices, Retrieved on 2 April 2008 at http://www.jtc.gov.sg/products/land/industrialland/pages/index.aspx

Production of Hydrogen via Syngas Route

9-30

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Chapter 10 : SAFETY, HEALTH & ENVIRONMENT (S.H.E.) 10.1 INTRODUCTION

Safety is of paramount importance in the operation of any chemical facility. Many industrial accidents, such as the infamous Bhopal incident, serve as poignant reminders of the significance of safety measures in preventing the escalation of minor accidents into major catastrophes. A safe plant would make both economical sense, and uphold a positive image of the company as a responsible corporate citizen in this global new economy. To cover all safety aspects of plant operation, the following studies have been carried out: 1. Hazard and Operability review on the operations of the Furnace/SMR integrated system

2. Plant layout of the plant to ascertain preliminary site area and required safety clearances

3. Possible issues related to the occupational safety and health of plant personnel

Furthermore, the possible impact that the operation of a hydrogen plant has on the environment would also be investigated in this report. It is essential that possible pollutants are minimized and mitigation measures are duly mapped out in the preliminary design of a chemical facility. This would be effected with the implementation of a risk assessment matrix. A product life cycle assessment on hydrogen would also be used as a tool to further illustrate the environmental impact of the manufacture of hydrogen through the steam methane reforming route.

Production of Hydrogen via Syngas Route

10-1

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.2 HAZARDS AND OPERABILITY STUDIES (HAZOP) REVIEW

A HAZOP study is used to identify the potential hazards in a chemical facility and is generally an effective method to determine operational hazards in any facility [R1]. In this preliminary design of a hydrogen plant, it was identified that the safe and functional operation of the furnace and the SMR unit were crucial to the smooth running of subsequent downstream units. A key issue that involves the integrated furnace/SMR system would be the possibility of explosion. This could arise because the integrated furnace/SMR system operates at the highest temperature and pressure in the plant. Furthermore, the presence of volatile fuel gas/air mixtures, coupled with methane from the natural gas feed and hydrogen formed after reaction in the SMR unit could also be possible explosive hazards. Hence a HAZOP study was employed to further investigate and analyze any potential impediments to their safe operation. This was done with reference to the P&ID diagram presented in the Chapter 11 on Instrumentation and Control.

Production of Hydrogen via Syngas Route

10-2

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Project Name: Production of Hydrogen via Syngas Route Process: Steam Methane Reforming Section: Steam Methane Reformer Reference Diagram: Furnace & SMR P&ID

Table 10-1: HAZOP study for SMR preheated reactor feed Study Node Preheated Reactor Feed

Process Parameters Flow

Item

Deviations

Possible Causes

Possible Consequences

Safeguards

Actions Required

1A

No

1. No SMR reaction 2. Shutdown of downstream units 3. Increase in temperature of flue gas 4. Disrupt preheating of other process streams in convection section

1. Installation of low flow alarm 2. Installation of redundant expander E-102

1. Regular operators’ training 2. Review of operating procedures for proper pipeline isolation 3. Regular inspections and maintenance of expander E-101 and E-102, heat exchangers HX101 and HX-102, piping, valves and fittings

1B

High

1. Control valves, FCV101, FCV-102 fail close 2. Controllers, FC-101, FC-102 fail and close valves 3. Three-way valves fails close 4. Plugging of upstream piping 5. Plugging in expander E-101 and E-102 6. Complete plugging of heat exchangers HX101 and HX-102 7. Operators’ error in isolating pipes for maintenance 1. Control valve, FCV101, FCV-102 fail open 2. Controller, FC-101, FC-102 fail and open valves 3. Flow indicator fails, indicating low 4. Error in flow control ratio

1. Increase in reactor pressure 2. Rupture of SMR tubes, leading to furnace explosion 3. Lower conversion due to lowered residence time 4. Increased corrosion of inner SMR tube

1. Installation of safety valves PSV-001, 002 and 005 2. Flow indicators FI203 for flow monitoring

1. Review of operating procedures for proper pipeline isolation 2. Regular inspections and maintenance of pipings, valves and fittings

Production of Hydrogen via Syngas Route

10-3

CN 4120: Design II Team 32

Pressure

Safety, Health & Environment (S.H.E.)

1C

Low

1D

Reverse

1E

High

5. Operator error in valve control 6. Faulty safety valve, fails to open 1. Control valves FCV101, FCV-102 fail partially open 2. Controllers FC-101, FC-102 fail and open valve partially 3. Flow indicator FT101 fails, indicating high 4. Plugging of heat exchangers HX-101 and 102 5. Operator error in valve control 6. Safety valves PSV001,002 and 005 not tightly close 1. Pressure build-up in SMR

1. Coking of catalyst 2. See 1B

Production of Hydrogen via Syngas Route

surface due to increased flow velocity

3. Regular operators’ training

1. Higher conversion rate due to longer residence time 2. Lower throughput 3. Possible sedimentation of contaminants in pipeline

1. Installation of low flow alarm FLA203

1. See 1B

1. Hydrogen embrittlement in feed pipeline not designated to carry hydrogen 2. Damage to expanders E-101 and 102

1. Installation of check valves 2. Flow indicator FI203 for flow monitoring 3. Installation of safety valve PSV-005 at the entrance to SMR 1. Installation of safety valve 2. Installation of PI101 and PI-102 to check for catalyst abnormalities through higher than normal pressure

1. Decrease the furnace firing rate

1. Increase in reactor pressure 2. Rupture of SMR tubes, leading to furnace explosion

1. Regular inspections and maintenance of valves, pipings, fittings and catalyst 2. Review of current operating

10-4

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

drop

Temperature

procedures to check for contributory factors to coking 3. Consider using medium pressure steam to decoke catalyst for maintenance 1. See 1B

1F

Low

1. See 1C

1. See 1C

1. See 1C

1G

High

1. Control valve fails, leading to increased flow in the heating stream in HX-102 2. Inefficient expander E-101 and 102, leading to under expansion 3. Increased preheating of natural gas feed

1. Higher rate of catalyst coking 2. Shorter catalyst life, leading premature changeout

1. Installation of temperature indicator TT-101 2. Installation of flow control valve FCV103 on the heating stream

1. Regular inspections and maintenance of valve and expander 2. Increase in generation of HP steam, which effectively reduces preheating of natural gas feed

1H

Low

1. Control valve FCV103fails, leading to decreased flow in the heating stream in HX102 2. Poor piping insulation 3. Shell-side fouling of heat exchanger 4. Inadequate preheating of natural gas feed

1. Lower conversion in SMR tubes with the maintenance of furnace duty 2. Higher furnace duty required, which reduces the operating lifespan of the furnace 3. Increase usage of fuel gas feed

1. Installation of temperature indicator TT-101 2. Installation of flow control valve FCV103 on the heating stream

1. Regular inspections and maintenance of valve and heat exchangers, HX102 2. Decrease in generation of HP steam, which effectively increases preheating of natural gas feed

Production of Hydrogen via Syngas Route

10-5

CN 4120: Design II Team 32

Composition of natural gas

Safety, Health & Environment (S.H.E.)

1I

High

1. Wrong setting of flow control ratio due to operators’ error 2. Controllers, FC-101, FC-102 fail 3. Control valves FCV101, FCV-102 fail

1. Increase in catalyst coking due to reduced steam

1. Installation of controllers FC-101, FC-102 2. Installation of control valves FCV101, FCV-102

1. Regular inspection and maintenance of valves and controllers 2. Regular operators’ training

1J

Low

1. See 1I

1. Higher cost due to large amounts of steam being injected

1. See 1I

1. See 1I

Production of Hydrogen via Syngas Route

10-6

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-2: HAZOP study for SMR reactor effluent Study Node Reactor Effluent

Process Parameters Flow

Item

Deviations

2A

No

1. Plugging of reactor outlet 2. See 1A 3. Plugging of catalyst

1. Shutdown of downstream units

1. Installation of low flow alarm FLA204

2B

Low

1. Partial plugging of reactor outlet 2. See 1B

1. See 1C

1. Installation of low flow alarm FLA204

Pressure

2C

Low

1. High pressure drop along tubes 2. Coking of catalyst

1. See 1H

1. Installation of PI101 and PI-102 to check for catalyst abnormalities through higher than normal pressure drop

Temperature

2D

High

1. Increased furnace firing

1. Installation of temperature indicator, TI-106

2E

Low

1. Decreased furnace firing

2F

High

1. Low reaction

1. Coking of pipelines and downstream units 2. Corrosion of pipes due to high temperature 1. Low conversion of natural gas to hydrogen 1. Eventual product off

Composition

Possible Causes

Production of Hydrogen via Syngas Route

Possible Consequences

Safeguards

Actions Required 1. Regular operators’ training 2. Regular inspections and maintenance of catalyst, pipings, valves and fittings 1. Regular inspections and maintenance of pipings, valves and fittings 1. Review of current operating procedures to check for contributory factors to coking 2. Consider using medium pressure steam to decoke catalyst for maintenance 1. Reduce furnace firing

1. See 2D

1. Increase furnace firing

1. Installation of

1. Monitoring of

10-7

CN 4120: Design II Team 32

of natural gas

Safety, Health & Environment (S.H.E.)

temperature 2. High flow rate, leading to shorter residence time 3. Possible deactivation of catalyst

Production of Hydrogen via Syngas Route

specification

methane analyzer A-004 2. Installation of temperature indicator TI-106

temperature and adopting necessary rectifications 2. Consider using medium pressure steam to decoke catalyst as regular maintenance 3. Catalyst changeout to be carried out if necessary

10-8

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Project Name: Production of Hydrogen via Syngas Route Process: Combustion of Fuel Gas Section: Furnace (Radiant Section) Reference Diagram: Furnace & SMR P&ID

Table 10-3: HAZOP study for furnace (radiant section) fuel gas feed Study Node Fuel Gas Feed

Process Parameters Flow

Item

Deviations

Possible Causes

Possible Consequences

Safeguards

Actions Required

3A

No

1. Control valve fails close 2. Controller fails and closes valve 3. Complete plugging of fuel gas feed line 4. Operators’ error in isolating pipe for maintenance 5. Rupture of feed line 6. Opened safety valve PSV-012. All flow redirect to safety valve

1. Flame extinguished due to lack of fuel supply 2. Pressure build-up upstream of blockage / closure 3. Little or no conversion attained in SMR 4. FCV-108 will close and lead to pressure buildup on preheat air line

1. Availability of make-up fuel gas to ensure continuous supply 2. Installation of low flow alarm FLA201

3B

High

1. Control valve (Main Fuel gas line) fails open 2. Controller FIC-05 fails and opens valve 3. Error in flow meter FT-105 readings; indicating low 4. Control valve fails

1. Increase in temperature of flue gas 2. Thermal creeping of furnace tubes and walls 3. Unneeded usage of additional fuel gas and air

1. PSV-012 to relieve any excess gas to flare 2. Control valve

1. Regular inspections and maintenance on valves, pipings and fittings 2. Regular operator training 3. Review standard operating procedure for pipeline isolation 4. Shut down of air blower B-101 in the event of pressure build-up 5. Adopt evacuation procedures for fuel gas release 1. Regular inspections and maintenance of valves, indicators, pipings and fittings 2. Shutdown of fuel gas to prevent pipe rupture

Production of Hydrogen via Syngas Route

10-9

CN 4120: Design II Team 32

Pressure

Safety, Health & Environment (S.H.E.)

3C

Low

3D

Reverse

3E

High

(Make-up CH4) open 5. Faulty safety valve PSV-012, fails to open

4. Rupture of pipeline due to high pressure 5. Flame impingement of tubes due to unstable flame

1. Control valve fails partially open / close 2. Controller FIC- 05 fails and opens / closes partially 3. Partial blockage / leakage of fuel gas feed line 4. Error in flow meter FT-105 readings; indicating high 5. Safety valve PSV-012 not tightly closed 1. Pressure build-up in furnace

1. Insufficient flame height due to low fuel supply 2. Optimal SMR conversion not achieved 3. Pressure build-up upstream of blockage / closure 4. Sedimentation of contaminants along the piping due to low fuel gas flow 1. Occurrences of backfiring

1. Availability of make-up fuel gas to ensure continuous supply 2. Installation of low flow alarm FLA201

1. Regular inspections and maintenance on valves, pipings, indicators and fittings 2. Adopt evacuation procedures for fuel gas release

1. Installation of check valve 2. Installation of flame arrestor

1. Control valve fails open 2. Controller FIC-05 fails and opens valve 3. Pressure indicator fails, indicating low 4. Faulty safety valve PSV-012, fails to open

1. Increase in temperature of flue gas 2. Thermal creeping of furnace tubes and walls 3. Unneeded usage of additional fuel gas and air 4. Rupture of feed line due to overpressure 5. Excess flow to

1. PSV-012 to relief any excess gas to flare

1. Regular inspections and maintenance of check valve and flame arrestor 1. See 3B

Production of Hydrogen via Syngas Route

3. Adopt emergency shutdown procedures 4. Adopt evacuation procedures for fuel gas release

10-10

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

3F

Low

1. Control valve fails partially open / close 2. Controller FIC-05 fails and opens / closes partially 3. Partial blockage / leakage of fuel gas feed line 4. Error in flow meter FT-105 readings; indicating high 5. Safety valve PSV012not tightly closed 1. PSA malfunction

Temperature

3G

Low

Composition of H2

3H

High

1. Deactivation of catalyst in PSA

3I

Low

1. PSA purge too low

Production of Hydrogen via Syngas Route

furnace resulting in unstable flame and flame impingement 1. See 3C 2. Occurrences of backfiring

1. Lower furnace efficiency 1. Unstable flame leading to possible flame impingement 1. Insufficient flame height due to low fuel supply 2. Optimal SMR conversion not achieved

1. See 3C

1. Installation of safety valve PSV-012

1. See 3C

1. Install temperature indicator 1. Changeout of catalyst in PSA if necessary 1. Introduction of makeup fuel gas

10-11

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-4: HAZOP study for furnace (radiant section) preheated air feed Study Node Preheate d Air Feed

Process Parameters Flow

Item

Deviations

Possible Causes

Possible Consequences

Safeguards

4A

No

1. Control valves fails close 2. Controller FIC-107 fails and closes valves 3. Loss of air supply due to damaged air blowers, B-101, B102 4. Complete blockage of air feed line 5. Operators’ error in isolating pipe for maintenance 6. Rupture of air feed line

1. Flame extinguished due to lack of oxygen supply 2. Pressure build-up upstream of blockage / closure 3. Unneeded usage of fuel gas 4. Little or no SMR conversion attained 5. Fuel gas released to the environment due to incomplete combustion

1. FI-108 to indicate flow 2. FCV-108 to control flow 3. Installation of low flow alarm FLA-202

1. Regular inspections and maintenance of pipings, valves and fittings 2. Regular inspections and maintenance of blower 3. Regular operator training 4. Review of operating procedures for proper pipeline isolation

4B

High

1. Excessive flame height 2. Pressure build-up in furnace

1. FI-108 to indicate flow 2. FCV-108 to control flow

1. Regular inspections and maintenance of pipings, valves and fittings 2. Regular operator training

4C

Low

1. Control valve fails open 2. Controller FIC-107 fails and opens control valves 3. Error in flow control ratio 4. Operators’ error 5. Error in flow meter FT-107 and 202 readings; indicating low 1. Partial plugging of control valves 2. Partial loss of air

1. Fuel gas released to the environment due to incomplete

1. FI-108 to indicate flow 2. FCV-108 to control

1. Regular inspections and maintenance of

Production of Hydrogen via Syngas Route

Actions Required

10-12

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

supply due to damaged blower Control valve FCV107 fails to respond Controller FIC-107 fails to respond Piping or fitting leakage / partially blocked Error in flow control ratio Operators’ error Error in flow meter FT-107 and 202 readings; indicating high Pressure build-up in furnace

combustion 2. Optimal SMR conversion not achieved 3. Pressure build-up upstream of blockage / closure

flow 3. Installation of low flow alarm FLA-201

pipings, valves and fittings 2. Regular operator training

1. Occurrences of backfire

1. Installation of check valve 2. Installation of flame arrestor

1. Regular inspections and maintenance of check valve and flame arrestor

1. CV fails open 2. Controller FIC-107 fails and opens control valves 3. Pressure indicator fails 4. Blower E-101 and 102 operating higher than normal 1. See 4C 2. See 4E.2

1. Pressure buildup within piping, leading to rupture 2. Excess flow to furnace leading to unstable flame

1. Installation of relief valve 2. FI-108 to indicate flow 3. FCV-108 to control flow

1. Regular inspections and maintenance of valves, indicators, and blower

1. Occurrences of backfire 2. Insufficient flame height or extinguished flame

1. Install low flow alarm FLA-202 2. FI-108 to indicate flow 3. FCV-108 to control flow

1. Regular inspections and maintenance of pipings, valves and fittings

3. 4. 5.

6. 7. 8.

Pressure

4D

Reverse

4E

High

4F

Low

1.

Production of Hydrogen via Syngas Route

10-13

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-5: HAZOP study for furnace (radiant section) flue gas effluent Study Node Flue Gas Effluent

Process Parameters Flow

Pressure Temperature

Item

Deviations

Possible Causes

5A

No

1. Malfunction of all burners 2. Complete blockages of fuel gas and air feeds

1. Reduced steam generation 2. Reduced preheating of natural gas 3. 3. Reduced preheating of air

1. Installation of peek holes

5B

High

1. Higher than normal fuel gas feed 2. Higher than normal air feed

1. Control valves that control fuel gas and air feeds

5C

Low

1. Malfunction of burner(s) 2. Malfunction of ignition source

1. Rupture of convection section due to pressure buildup 2. Collapse of tubes in convection section 1. Reduced steam generation 2. Reduced preheating of natural gas 3. 3. Reduced preheating of air

5D 5E 5F

High Low High

1. See 5B 1. See 5C 1. High air / fuel gas flow rates 2. Inefficient heat transfer due to SMR tube degradation and fouling 3. High temperature of combustible air feed

1. See 5B 1. See 5C 1. Overheating and weakening of tubes in convection section 2. Cracking of furnace walls 3. Formation and emission of NOx due to higher temperature within furnace 4. Increased preheating

1. See 5B 1. See 5C 1. Installation of control valves that control fuel gas and air feeds

Production of Hydrogen via Syngas Route

Possible Consequences

Safeguard

1. Installation of peek holes

Actions Required

1. Regular checks via peek holes on burning 2. Regular inspection and maintenance of piping for fuel gas & air feeds 2. Lower feed rates fuel gas and air

1. Regular checks via peek holes on burning 2. Regular inspection and maintenance of piping for fuel gas and air feeds 1. See 5B 1. See 5C 1. Lower feed rates of fuel gas and air 2. Lower combustible air temperature

10-14

CN 4120: Design II Team 32

Composition of CO

Safety, Health & Environment (S.H.E.)

5G

Low

1. Low air / fuel gas flow rate 2. Inefficient furnace burning 3. Low temperature of combustible air

5H

High

1. Incomplete combustion of fuel gas

Production of Hydrogen via Syngas Route

of other process streams in convection section, leading to process disturbances in other units 1. Less steam generation and preheating of other process streams in convection section, leading to process disturbances in other units

1. Furnace flooding leading to afterburn in convection section and possible explosion. 2. Emission of CO and fuel gas to the atmosphere

1. Installation of control valves that control fuel gas and air feeds

1. Installation of CO and O2 analyzer in the convection section.

1. Regular checks via peek holes on burning 2. Regular inspection and maintenance of piping for fuel gas and air feeds 3. Increase temperature of combustible air 1. Lower fuel flow rate. 2. Slowly increase combustible air flow rate.

10-15

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Project Name: Production of Hydrogen via Syngas Route Process: Heat Exchange of Flue Gas Section: Furnace (Convection Section) Reference Diagram: Furnace and SMR P&ID

Table 10-6: HAZOP study for furnace (convection section) condensate Study Node Condens ate

Process Parameters Flow

Item

Deviations

Possible Causes

Possible Consequences

Safeguards

Actions Required

6A

No

1. Control valve FCV104 fails close 2. Controller FC-104 fails 3. Bypass valve fails close 4. Low temperature reading on temperature indicator TE-104 5. Operator error in isolation of piping for maintenance

1. Installation of bypass valve 2. Installation of flow controller FC-104 and valve FCV-104

1. Regular inspection and maintenance of valves, controllers and indicators 2. Regular operator training 3. Review of current operating procedures

6B

High

1. Control valve FCV104 fails open 2. Controller FC-104 fails 3. Bypass valve fails open 4. High temperature reading on temperature indicator TE-104

1. No steam generation to buffer the flue gas temperature fluctuations 2. Too much preheating of other process streams in convection section, leading to disturbances in the other process units 3. Possible creeping of steam generation tube in convection section 4. Stack gas temperature would be elevated 1. Rupture of steam generation tube in convection section 2. Too little preheating of other process streams in convection section, leading to disturbances in the other process units 3. Stack gas temperature would be

1. See 6A

1. See 6A

Production of Hydrogen via Syngas Route

10-16

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

6C

Low

1. Control valve FCV104 fails partially close 2. Controller FC-104 fails 3. Bypass valve fails partially close 4. Low temperature reading on temperature indicator TE-104 5. Operator error in isolation of piping for maintenance

1.

2.

3.

4.

Pressure

6E

6F

High

Low

1. Control valve FCV104 fails open 2. Controller FC-104 fails 3. Bypass valve fails open 4. High temperature reading on temperature indicator TE-104

1.

1. Control valve FCV104 fails partially close

1.

Production of Hydrogen via Syngas Route

2.

3.

decreased, leading to acid corrosion issue in the stack Little steam generation to buffer the flue gas temperature fluctuations Too much preheating of other process streams in convection section, leading to disturbances in the other process units Possible creeping of steam generation tube in convection section Stack gas temperature would be elevated Rupture of steam generation tube in convection section Too little preheating of other process streams in convection section, leading to disturbances in the other process units Stack gas temperature would be decreased, leading to acid corrosion issue in the stack Little steam generation to buffer the flue gas

1. See 6A

1. See 6A

1. See 6A

1. See 6A

1. See 6A

1. See 6A

10-17

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

2. Controller FC-104 fails 3. Bypass valve fails partially close 4. Low temperature reading on temperature indicator TE-104 5. Operator error in isolation of piping for maintenance

Production of Hydrogen via Syngas Route

temperature fluctuations 2. Too much preheating of other process streams in convection section, leading to disturbances in the other process units 3. Possible creeping of steam generation tube in convection section 4. Stack gas temperature would be elevated

10-18

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-7: HAZOP study for furnace (convection section) natural gas feed Study Node Natural Gas Feed

Process Parameters Flow

Pressure

Item

Deviations

7A

No

7B

Possible Causes

Possible Consequences

Safeguards

Actions Required

1. Disruption of supply 2. Pipeline rupture

1. Discharge of natural gas into environment 2. No reaction in SMR 3. Higher preheating of combustible air 4. Elevation of stack gas temperature

1. Ensure duplicity in supplies

High

1. Surges in supply

1. Rupture of natural gas tubes in convection section, leading to possible furnace explosion 2. Less preheating of combustible air

1. Installation of safety valve

7C

Low

1. See 7A

High

1. See 7A.1 2. Lower SMR throughput 1. See 7B

1. See 7A

7E

1. Fluctuations in natural gas feed flow 2. See 7A 1. See 7B

1. See 7B

1. See 7B

7F

Low

1. See 7C

1. See 7C

1. See 7C

1. See 7C

Production of Hydrogen via Syngas Route

1. Regular inspection and maintenance of supply pipelines 2. Increase the amount of steam generation to offset the increase in preheating of combustible air and stack gas temperature 1. Consider the possible installation of feed surge drum to regulate the flow of natural gas and eliminate any fluctuations

10-19

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-8: HAZOP study for furnace (convection section) combustible air feed Study Node Combust ible Air

Process Parameters Flow

Pressure

Item

Deviations

Possible Causes

Possible Consequences

8A

No

1. Loss of air supply due to damaged air blowers, B-101 2. Complete blockage of air feed line 3. Complete plugging of air filter

1. Elevation of stack gas temperature 2. Flame extinguished due to lack of oxygen supply 3. Little or no SMR conversion attained

1. Availability of redundant blower, B-102

1. Periodic cleaning or replacement of air filter 2. Regular inspections and maintenance of air blowers

8B

High

1. Accidental operation of two blowers B-101 and 102 2. Safety valve PSV010 fails to open

1. Installation of safety valve

1. Regular operator training 2. Review of blower operating procedures

8C

Low

1. Availability of redundant blower, B-102

1. See 8A

8E

High

1. Damaged blower, B101 resulting in reduced air flow 2. Partial blockage of air filter 3. Partial plugging of air feed line 1. See 8B

1. Decrease in stack temperature, leading to possible acid corrosion in the stack 2. Rupture of air pipelines in the convection section, leading to possible ignition of residual fuel gas 1. Optimal SMR conversion not achieved 2. Elevation of stack gas temperature

1. See 8B

1. See 8B

1. See 8B

8F

Low

1. See 8C

1. See 8C

1. See 8C

1. See 8C

Production of Hydrogen via Syngas Route

Safeguards

Actions Required

10-20

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-9: HAZOP study for furnace (convection section) stack gas Study Node Stack Gas

Process Parameters Temperature

Item

Deviations

Possible Causes

1A

Low

1. Too much heat transfer to process and utility streams in convection section 2. Low flue gas temperature from radiant section

Production of Hydrogen via Syngas Route

Possible Consequences

1. Condensation of acid gases, leading to corrosion in stack

Safeguards

1. Installation of temperature indicator, TI-105

Actions Required

1. Lower rate of steam generation 2. Increase flue gas temperature by increasing rate of burning

10-21

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Project Name: Production of Hydrogen via Syngas Route Process: Heat Exchange of Flue Gas Section: Furnace Reference Diagram: Furnace & SMR P&ID

Table 10-10: HAZOP study for furnace Study Node Furnace

Process Parameters Pressure

Item

Deviations

Possible Causes

10A

High

1. Faulty control system, causing stack damper to remain closed 2. Increase in rate of burning

Temperature

10B

High

1. Too much fuel gas and air feed 2. No reactant flow in SMR tubes

Production of Hydrogen via Syngas Route

Possible Consequences

Safeguards

Actions Required

1. Rupture of SMR tubes, leading to furnace explosion 2. Cracking of furnace walls 3. Occurrences of backfire

1. Installation of flame arrestor 2. Installation of high pressure alarm

1. Thermal creep of SMR tubes 2. Furnace walls would crack

1. Installation of temperature indicator TI-202, TI-203 2. Install control valve FCV-108

1. Regular inspections and maintenance of control systems, tubes and damper 2. Review of current operating procedures if high pressure occurrences are frequent 1. Regular inspections and maintenance of valve and control system

10-22

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.3 PLANT LAYOUT

Plant layout is the spatial arrangement of items of process vessels and equipment and their connection by pipes, ducts, conveyors or vehicular transportation, as defined by Mecklenburgh [R2]. It is essential that plant layout is carefully thought out to satisfy the following key considerations:  Cost-effective use of space  Risk-free and efficient construction  Reliable, efficient and safe operations  Ease of maintenance and repair of process and associated auxiliary units, carried out on

site or ex situ  Minimal hazard and nuisance caused to the public

Besides safety and economic concerns, other factors such as process requirements, fire-fighting and emergency capabilities, administrative and medical support infrastructure would also have to be incorporated into the overall plant layout. An efficient transportation network within the plant for the movement of materials, personnel and emergency services would also be needed. These key issues would be duly covered in the following paragraphs.

10.3.1 Segregation

A main process area was designated, housing the primary process units such as the furnace, SMR, HTS, LTS and the PSA unit. This facilitated the movement of material from one process unit to the other and allowed for lower piping cost. For safety and loss prevention, this main process area was deliberately sited away from the main administrative building, laboratory and other buildings which house work personnel. Furthermore, to control access to the main process area and also to the auxiliary units and waste treatment area, a security fence was erected and an ‘official access only’ system was implemented, in view of increasing security and intrusion concerns.

Production of Hydrogen via Syngas Route

10-23

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

A safety clearance of at least 60 m would also be needed between this main process area and the main buildings where plant personnel are housed [R2].

10.3.2 Transportation Considerations

A rectangular grid was employed in the preliminary plant layout as it is often the most cost-effective arrangement. This is because overhead piperacks, sewer systems, underground piping systems, trenches, electric cables and instrument lines very often follow the road layout [R2]

. Therefore, curved roads should generally be avoided to facilitate the laying of these lines and

systems. Roads should also be built in such a way so as to allow easy access for emergency and fire-fighting services. This can be accomplished with the use of a peripheral road system which allows for at least 2 approaches to all major fire risks, e.g. the main process area. Roads in the plant should also have access to the public road system at a minimum of 2 points. Moreover, they should be sited at a reasonable distance from any process unit or building in order to effectively carry out firefighting or administer emergency care. The actual distance is usually set at around 18 to 45 m [R2]. The main perimeter road was set to be 10 m wide. This is to facilitate the bulk of daily vehicular traffic. Primary and secondary access roads were designated to be 6 m and 3.5 m wide respectively. Adequate parking space was also allocated for work personnel, visitors and loading/unloading vehicles. These were sited away from wind-blown dusts and the main process area for safety and security purposes, and suitably sized to prevent congestion during shift changeover [R2].

10.3.3 Administration

Plant personnel with more general site responsibilities, e.g. human resource and IT departments, should be sited in an administrative building located in a non-hazardous area away from the main process units. This building should be sited upwind of possible fumes and

Production of Hydrogen via Syngas Route

10-24

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

emission from the process units, and near the main entrance to the plant to facilitate evacuation. An estimated 50 administrative employees was used as the benchmark in this preliminary design, hence according to Mecklenburgh [R2], an area of 500 m2 would suffice for preliminary layout planning.

10.3.4 Laboratory

An onsite laboratory is critical in the setup of any plant as it provides a readily-available source of analytical information on the quality of intermediates and products. This allows engineers and operators to effect any changes to the process parameters in order to correct any deviations. An estimated 20 laboratory staff would suffice for this preliminary design; hence an area of 400 m2 would be used [R2].

10.3.5 Workshop

A workshop is needed for mechanical repairs and maintenance to be carried out onsite. This will help to defray costs related to delays in repair and transport elsewhere. An estimate of 30 people would be used; hence an area of 600 m2 would be set aside for such a workshop [R2].

10.3.6 Control Room

The location of a control room is based on normal operating requirements and the need for protection during emergency situations. Operators would be more inclined to be on the plant should the control room be nearer to the plant. As a result, they would be more likely to observe any malfunctions and plant deviations, hence preventing the occurrence of a serious fault. Instrument cables would also be shortened if the control room is sited nearer the plant. It has also been reported that at distances greater than 35 m and particularly over 100 m, operators will tend not to go into the plant, especially in inclement weather [R2]. Nevertheless, the shorter the distance the control room is to the plant, the greater the need for the room to be reinforced for personnel protection, and hence a higher cost is needed.

Production of Hydrogen via Syngas Route

10-25

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Therefore, a suitable distance of about 30 m between the control room and the main process area was chosen in this design [R2]. An estimated area of 200 m2 was used for the control room.

10.3.7 Transformer Substation

The substation is usually sited at the edge of the plant, in areas of the lowest electrical hazard rating. This will ensure a continuous electrical supply to the entire plant, critical to its daily functioning.

10.3.8 Emergency Services

Emergency services such as ambulance and fire stations need to be given rapid access to the entire site, without causing any hazard to the existing plant traffic. Preferably, they should be housed outside the main fence but close to the main plant entrance. A 500 m2 area would be recommended for this hydrogen plant. An area adjacent to the main administrative building would also be set aside as an emergency assembly area. This would help in the accounting of plant personnel should an emergency incident arise.

10.3.9 Amenities (Medical Centre and Canteen)

The medical centre is essential in administering emergency and daily medical care to the plant employees. Since it is frequently used by staff from all over the site, it should be located in a central, non-hazardous area, preferably grouped together with other amenities such as the canteen. This will help to cut down on time spent traveling to these amenities. Also, it should be sited upwind from drifting fumes and noise from process units. It was hence decided that an area of 30 m2 would be used for the medical centre, able to cater to about 200 employees and contractors. An area of 200 m2 would be used in sizing the canteen, also with the capacity to accommodate 200 personnel at one time [R2].

Production of Hydrogen via Syngas Route

10-26

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.3.10 Process and Auxiliary Units 10.3.10.1 Furnace (Housing SMR)

The foremost considerations in the location of a furnace are safety issues since it is a constant source of ignition. Where practical, furnaces should be sited in upwind locations so that flammable gases or vapors are less likely to be blown towards the furnace, resulting in ignition. Furnace transfer lines should also be kept short, with a common stack policy being employed. Hence, process equipment which is directly connected to the furnace should be sited as close as possible, without compromising on the recommended safety distance of 30 m from any equipment which could be possible sources of ignition.2 Also, underground drain-points and manhole covers within 30 m of the furnace walls should be sealed, and pits or trenches should generally be avoided from extending under furnaces. With regards to the abovementioned considerations, the furnace housing the SMR would be located within the designated main process area housing the other major units. The recommended safety spacing would be a distance of 15 m between the furnace and the HTS and LTS reactors, and a distance of 30 m between the furnace and the PSA unit, which is considered to be a process unit with a low flash point due to the high concentration of H2 present [R2].

10.3.10.2 Reactors (HTS, LTS), PSA and Knockout Drum

HTS and LTS are typical fixed-bed reactors loaded with catalyst in bulk between supports within the reactor vessel. The PSA unit also faces similar considerations due to its usage of bulk catalyst as an absorbent; hence it was mentioned together in this section. As catalysts have a fixed effective shelf life of a few years, provisions have to be made for the removal and loading of catalysts. The units have to be sufficiently elevated from the ground so as to allow catalyst removal by mechanical transport, belt conveyors, fluid conveying or hand trucks. Clearance should be allowed for the usage of mechanical drills and other equipment should coking, sintering, or hard agglomeration of the catalyst [R2].

Production of Hydrogen via Syngas Route

10-27

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

According to Mecklenburgh [R2], an area of 4 m2 must also be available at the base of each reactor for the transport and temporary storage of fresh and spent catalyst. It is also important to note that the PSA unit is used to concentrate the amount of H2. With H2 being a vapor with a very low flash point of -273°C [R3], care would need to be exercised in the safety distances needed. Since the PSA unit consists of 8 separate columns, it has been recommended that such process equipment with low flash points should be spaced 2 m away from one another [R2]. Also, it would be housed together in the main process area, at a distance of at least 5 m from the HTS and LTS reactors, and a distance of at least 30 m from the furnace. The knockout drum would be sited just next to the LTS unit.

10.3.10.3 Cooling Tower

Problems associated with the location of a cooling tower are often related to the large volumes of very humid air which emanates from it, compounding the already high levels of humidity normally experienced in Singapore. The moisture can lead to fog, precipitation and corrosion issues in areas downwind of it [R2]. Towers should also be sited to mitigate the effects of wind drift on roads, rail, plant and the neighborhood of the site. It is important to check that any possible corrosive emissions from the vents of the HTS and LTS reactors, and the stack emissions from the furnace, would not be entrained within the cooling tower. Specifically for our project, a mechanical-draught tower was used. These require power and may generate associated fan noise. If the entire plant was sited near a residential area, a slower fan speed should be employed. Alternatively, buildings or sound screens could be erected between the tower and the residential area. Hence, the cooling tower has been separately sited far away from the main process area, i.e. as an offsite facility, so as to prevent any entrainment of corrosive vapors.

Production of Hydrogen via Syngas Route

10-28

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

It has been accorded a safety distance of at least 30 m from the process area, at least 60 m from administrative buildings, at least 30 m from the fire station and at last 45 m from the main plant substation [R4].

10.3.10.4 Heat Exchangers

Heat exchangers should be located within the conventional process unit plot area, in close proximity to the equipment which they are associated with. This would minimize the cost of pipe runs and also facilitate operator and maintenance access. Heat exchangers between two process equipment which are far away should be sited at optimal points in relationship to pipe tracks [R2]. However, due to the close proximity of our major process units to one another, with the exception of the cooling tower, it would not be a major issue for consideration in this preliminary design. If exchangers are located in pairs, or in larger groups, they can be stacked on top of one another. A spacing of 0.45 m (18 inch) should be provided to allow maintenance to be carried out on the flange bolts easily [R4]. By this means, there could be resultant savings on service pipework, pipebridge and structural work etc [R2]. However, process piping and access steel work may actually increase consequently; therefore a compromise has to be made. Generally, most exchangers would also be placed on a base about 1 m above ground level for the provision of drain connections. As there are no restrictions on minimum safety clearances between heat exchangers and the specific process units that are found in our preliminary design, it would be considered that they occupy the space in between the safety clearances of the process units.

10.3.10.5 Flares

Flares are used to burn away excess gases in an emergency situation and also to flare away off-specification gases such as H2. A sterile radius of at least 60 m should be maintained

Production of Hydrogen via Syngas Route

10-29

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

around the flare, with only associated flare equipment and access roads allowed within this radius. Also, the flare stacks should be located downwind from the main process areas and at least 100 m away to provide for the dispersion of vapor releases [R2].

10.3.10.6 Wastewater Treatment Plant

A wastewater treatment plant is also needed to treat the condensate from the knockout drum, the blowdown from the cooling water tower and any other liquid plant effluent before it is discharged. It is usually sited at the perimeter of the site. The estimated total effluent from the knockout drum and the cooling water tower is around 873760 m3/year, assuming 8000 operating hours in a year. Comparing this with a water treatment plant sited in Changi, Singapore5, which processes 292 million m3 of wastewater per year (calculated from a daily rate of 800000 m3/day) and occupies an area of 55 ha (550000 m2), the area needed for the onsite treatment plant would be estimated to be around 1650 m2.

Process & Auxiliary Units

Base Area (m2)

Plant Facilities

Base Area (m2)

Furnace (housing SMR) HTS LTS Knockout Drum PSA Cooling Water Tower

891 9.4 8.6 3.0 144 733

Administrative Building Laboratory Workshop Control Room Transformer Substation Fire Station

500 400 600 200 160 500

Flare

25

Medical Centre

30

Wastewater Treatment Plant

1650

Canteen

200

Table 10-11: Base areas of process, auxiliary and other plant facilities

From Figure 10-1, the plant area was estimated to be around 29400 m2.

Production of Hydrogen via Syngas Route

10-30

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Fig 10-1: Plant layout 10.4 OCCUPATIONAL SAFETY 10.4.1 Personal Protection Equipment (PPE)

The safety requirements for different tasks in different parts of the plant are different. Thus it is important to consult the Safety, Health and Environment department to ensure that the correct PPE are worn [R1] when performing each task.

Production of Hydrogen via Syngas Route

10-31

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Personal protection equipment is extremely important in preventing or reducing exposure by providing a barrier between the worker and the workplace environment. This is for the benefit of workers’ safety and health. However, by itself, it does not prevent accidents. It is only acting as a secondary measure for safety and health purposes.

One of the most important equipment for this plant would be self-contained breathing apparatus. This is because the dangers of this plant would mostly be caused by inhalation. Therefore, this equipment has to be strategically placed in areas where leaks might occur, especially in confined spaces. Other equipment needed would be safety glasses, safety gloves and safety shoes.

Employers have to identify and provide the appropriate PPE for their employees. They would also have to train employees in using and caring for the PPE. Periodically, they should also review and check the PPE, replace worn out items and make the PPE program more effective.

Employees should properly adhere to the regulations on wearing their PPE. They should also attend training sessions for the use of their PPE. They should have the responsibility to take care and maintain their PPE and inform their superiors if there is a need to replace these items. 10.4.2 Noise

Noise [R1] problems are commonplace in chemical plants. They are measured in decibels (dB). According to regulations, noise levels should not exceed 85dB for an 8 hour workday. Measures must be taken if the exposed noise level is higher than 85dB. Control of the noise can happen at either the source or the receiver of the noise. At the source of noise, one can either enclose the source using shields such as plywood or noise absorbing foams. It is also possible to employ sound barriers to reduce the noise level transmitted to workers. Otherwise, at the receiver level, ear plugs and ear muffs are the most commonly used PPE for regulating exposure to noise.

Production of Hydrogen via Syngas Route

10-32

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.4.3 Ventilation

Ventilation [R1] is extremely important for this plant for the following reasons:  It can remove dangerous concentrations of flammable and toxic materials  It can be highly localized, reducing the quantity of air moved  Ventilation equipment is readily available and can be easily installed  It can be added to an existing facility

However, the major disadvantage is operating cost, which is rather prohibitive. Ventilation systems comprise of fans and ducts that effectively dilute the contamination using dilution ventilation. Fresh air is flowed in large amounts to dilute the contamination. Workers would still be exposed to the contaminants, but in lesser concentrations.

Production of Hydrogen via Syngas Route

10-33

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.5 OCCUPATIONAL HEALTH HAZARD IDENTIFICATION

It is important to identify any safety hazards brought about by the chemical nature of the reactants and products present in the steam methane reformer. A good knowledge of the nature of these components is necessary for operators to know how to react to exposure to these components. Table 10-12: Summary of nature and associated hazards of chemicals [R6] Chemical Threshold Lower Personal Health Effects Limit Exposure Protective Acute Value Limit and Equipment (TLV) Upper (PPE) Exposure Limit (LEL – UEL) Methane N.A. 5 – 15% Safety glasses Simple asphyxiant- reduces the amount of oxygen in the and/or face air. Exposure to oxygen-deficient atmospheres (less than shields 19.5 %) may produce dizziness, nausea, vomiting, loss of consciousness, and death. At very low oxygen concentrations (less than 12 %) unconsciousness and death may occur without warning. It should be noted that before suffocation could occur, the lower flammable limit for Methane in air will be exceeded; causing both an oxygen deficient and an explosive atmosphere. CO 25 ppm 12.5 – 74% Safety glasses, Inhaled carbon monoxide binds with blood hemoglobin to TWA safety gloves, form carboxyhemoglobin. Carboxyhemoglobin cannot safety shoes, and take part in normal oxygen transport, greatly reducing the self-contained blood’s ability to transport oxygen. Depending on levels

Production of Hydrogen via Syngas Route

10-34

Chronic

N.A.

N.A.

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

breathing apparatus during emergency H2

N.A

4% - 74.5%

Wear chemical resistant gloves

CO2

5000 ppm TWA

Nonflammable

Safety glasses, safety gloves, safety shoes, and self-contained breathing apparatus during emergency

and duration of exposure, symptoms may include headache, dizziness, heart palpitations, weakness, confusion, nausea, and even convulsions, eventual unconsciousness and death. Nausea, vomiting, difficulty breathing, irregular heartbeat, N.A headache, fatigue, dizziness, disorientation, mood swings, tingling sensation, loss of coordination, suffocation, convulsions, unconsciousness, coma Carbon dioxide is a cerebral vasodilator. Inhaling large N.A quantities causes rapid circulatory insufficiency leading to coma and death. Asphyxiation is likely to occur before the effects of carbon dioxide overexposure. Low concentrations cause increased respiration and headache. Product is a simple asphyxiant. Effects of oxygen deficiency may include any, all or none of the following: rapid breathing, diminished mental alertness, impaired muscle coordination, blurred speech, and fatigue. As asphyxiation progresses; nausea, vomiting, and loss of consciousness may occur, eventually leading to convulsions, coma and death.

These are the procedures to be taken for spills, fire-fighting and medical aid. Table 10-13: Summary of safety, fire-fighting and medical aid measures [R6] Chemical Spill/leak measures Fire-fighting measures Methane Personal precautions: Wear self Extremely flammable. Exposure to fire contained breathing apparatus when may cause containers to entering area unless atmosphere is rupture/explode. If possible, stop flow proved to be safe. Evacuate area. Ensure of product. Move away from the adequate air ventilation. Eliminate container and cool with water from a

Production of Hydrogen via Syngas Route

Medical Aid measures Inhalation: Remove victim to uncontaminated area wearing selfcontained breathing apparatus. Keep victim warm and rested. Call a doctor. Apply artificial

10-35

CN 4120: Design II Team 32

CO

H2

CO2

Safety, Health & Environment (S.H.E.)

ignition sources. Environmental precautions: Try to stop release. Clean up methods: Ventilate the area.

protected position. Do not extinguish a leaking gas flame unless absolutely necessary. Spontaneous explosive reignition may occur. Extinguish any other fire. In confined space, fire fighters must use self-contained breathing apparatus.

respiration if breathing stopped. Ingestion: Ingestion is not considered a potential route of exposure.

Personal precautions: Evacuate area. Eliminate ignition sources. Use selfcontained breathing apparatus and chemically protective clothing. Ensure adequate air ventilation. Wear self contained breathing apparatus when entering area unless atmosphere is proved to be safe. Environmental precautions: Try to stop release. Clean up methods: Ventilate the area. Evacuate all personnel from affected area. Use appropriate protective equipment. If leak is in user’s equipment, be certain to purge piping with an inert gas before attempting repairs. If leak is in the container of container valve, contact closest supplier location.

Extremely flammable. Exposure to fire may cause containers to rupture/explode. If possible, stop flow of product. Move away from the container and cool with water from a protected position. Do not extinguish a leaking gas flame unless absolutely necessary. Spontaneous explosive reignition may occur. Extinguish any other fire. Fire fighters must use selfcontained breathing apparatus. If possible, stop the flow of hydrogen. Cool surrounding containers with water spray. Hydrogen burns with an almost invisible flame of relatively low thermal radiation. Hydrogen is very light and rises very rapidly in air. Should a hydrogen fire be extinguished and the flow of gas continue, increase ventilation to prevent an explosion hazard, particularly in the upper portions. Non-flammable. If possible stop flow

Inhalation: Remove victim to uncontaminated area wearing selfcontained breathing apparatus. Keep victim warm and rested. Call a doctor. Apply artificial respiration if breathing stopped. Ingestion: Ingestion is not considered a potential route of exposure.

Personal precautions: Evacuate area.

Production of Hydrogen via Syngas Route

Inhalation: Remove victim to uncontaminated area wearing selfcontained breathing apparatus. Keep victim warm and rested. Call a doctor. Apply artificial respiration if breathing stopped. Ingestion: Ingestion is not considered a potential route of exposure.

Inhalation: Remove victim to

10-36

CN 4120: Design II Team 32

Wear self contained breathing apparatus when entering the area unless atmosphere is proved to be safe. Ensure adequate air ventilation. Environmental precautions: Try to stop release. Prevent from entering sewers, basements and workpits, or any place where its accumulation can be dangerous. Clean up methods: Ventilate the area.

Production of Hydrogen via Syngas Route

Safety, Health & Environment (S.H.E.)

of product. Move away from the container and cool with water from a protected position. In confined space, fire-fighters must use self contained breathing apparatus.

uncontaminated area wearing selfcontained breathing apparatus. Keep victim warm and rested. Call a doctor. Apply artificial respiration if breathing stopped. Skin/eye contact: Immediately flush eyes thoroughly with water for at least 15 minutes. In case of frostbite, spray with water for at least 15 minutes. Apply a sterile dressing. Obtain medical assistance. Ingestion: Ingestion is not considered a potential route of exposure.

10-37

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.6 ENVIRONMENTAL IMPACT ASSESSMENT

Environmental impact assessment (EIA) refers to the need to identify and predict the impact on the environment and on man’s health and well-being of legislative proposals, policies, programmes, projects and operational procedures, and to interpret and communicate information about the impacts [R7]. Although there is no legal requirement for EIA to be carried out for projects in Singapore [R8], an EIA would be carried out in this preliminary design of a hydrogen plant so as to address the environmental aspects of undertaking the operation of a hydrogen plant.

10.6.1 Objectives

The primary objectives of an EIA are summarized as follows [R9]:  To determine methods to prevent or mitigate environmental damage  To mitigate environmental damage through the application of practical meditative actions  To make known to the public and the public or private bodies in charge of the project the

noteworthy environmental effects of such an undertaking  To make known to the public justifications of governmental approvals of undertaking

with substantial environmental effects  To encourage cross-agency interaction in the assessment of projects  To engage the public in the planning process

10.6.2 Risk Assessment Matrix

In order to assess the environmental impact of the daily activities of the hydrogen plant and its accompanying wastewater treatment plant, a risk assessment matrix as shown in Table 10-14 was utilized to summarize the offending activities and its associated impacts on the environment and work personnel. Criteria employed in this risk assessment matrix are reflected in the accompanying legend, with risk ranking (RR) having considered the overall consequences (OC) and the likelihood of occurrence (LH). The RR would serve as a clear indication of the risk level of a particular activity in the operation of the hydrogen plant.

Production of Hydrogen via Syngas Route

10-38

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Table 10-14: Risk assessment matrix detailing the environmental impact of hydrogen plant operations Activity/ Process/ Aspect/Associated N/A/E Impact/Effect P PD ENV REP OC LH Product/ Hazard Services Flaring 1. Complete N 1. Increased IV IV III YES IV A combustion yields greenhouse CO2, a major effect greenhouse gas 2. Incomplete combustion yields methane (from natural gas), H2 (off-spec product), soot and CO

N

Production of Hydrogen via Syngas Route

1. Increased greenhouse effect 2. Emission of H2, a highly flammable gas and an explosive hazard 3. Soot can worsen to regional haze problem 4. CO can indirectly raise methane and tropospheric ozone, which also contribute to

III

III

II

YES

III

B

RR

M

M

Mitigation Measures

1. Reduce frequency of excess waste gas emission, emergency flaring and off-spec flaring incidents 2. Try to achieve absolute combustion through use of excess air 3. Flare got opacity analyzer 4. Installation of air scrubbers 5. Periodic monitoring to prevent excessive emissions

10-39

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

the global warming

Emission of stack gases during furnace operation

1. Emission of greenhouse gases such as CO2

N

1. Increased greenhouse effect

IV

IV

III

YES

IV

A

M

2. Emission of NOx due to high temperature in furnace radiant section

N

1. Acid rain formation

III

III

II

YES

II

A

H

3. Emission of soot and CO due to incomplete combustion

N

1. Soot can contribute to regional haze problem 2. Emission of CO can indirectly

III

III

II

YES

III

B

M

Production of Hydrogen via Syngas Route

1. Ensure efficient furnace operation, using the least possible fuel gas 1. Install temperature sensor to monitor radiant section temperature 2. Install NOx controller 1. Install opacity analyzer to detect excessive soot formation 2. Ensure excess air is present

10-40

CN 4120: Design II Team 32

Vapor leakages from pipelines

4. If sulfur is found in natural gas feed, SO2 could be released 1. Emission of methane, hydrogen, CO and CO2

Safety, Health & Environment (S.H.E.)

A

A

Production of Hydrogen via Syngas Route

raise methane and tropospheric ozone, which also contribute to the global warming 1. Acid rain formation

1. Increased greenhouse effect due to methane, CO and CO2, leading to global warming 2. Emission of H2, a highly flammable gas and an explosive hazard 3. Exposure to CO, an asphyxiant, can result in dizziness,

to minimize incomplete combustion

III

III

II

NO

III

D

L

1. Use fuel gas feed with less sulfur

II

II

III

NO

II

C

M

1. Isolation and repair of leaking pipes 2. Conduct regular checks

10-41

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

nausea, vomiting, loss of consciousnes s and death

2. Hot steam emissions due to ruptured pipe

Storage of 1. Leakage of noncooling volatile chemicals water chemicals (biocides, corrosion inhibitor)

A

1. Injury to work personnel

III

IV

IV

YES

IV

C

L

A

1. Soil contaminatio n 2. Water contaminatio n

III

IV

I

YES

III

C

M

Production of Hydrogen via Syngas Route

1. Conduct steam leak test 2. Clamp and repair leaking pipes 1. Set storage limits 2. Ensure proper storage of chemicals 3. Containment of leaks 4. Neutralization or dilution methods 5. Proper disposal of expired chemicals 6. Regular checks by cooling tower unit

10-42

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Knockout drum operation

3. Condensate leakage from drum and associated pipelines, carrying dissolved hydrocarbons and carboxylic acids)

A

1. Soil contaminatio n 2. Water contaminatio n

III

IV

I

YES

III

C

M

Cooling tower operation

1. Water droplets or water mist aerosol generated as drift

A

1. Legionnaire ’s Disease can be transmitted, leading to pneumonia in serious cases [R10]

II

IV

III

YES

III

C

M

1. Raise temperature s of surrounding water

III

2. Discharge of heated cooling water into surrounding water bodies

A

Production of Hydrogen via Syngas Route

operators 7. Periodic soil quality tests 1. Containment of leaks 2. Regular pipeline checks 3. Periodic maintenance of drum and pipelines 1. Periodic disinfection of cooling tower using chlorine [R11]

IV

I

YES

III

D

L

2. Installation of drift eliminators 3. Usage of biocides to prevent accumulation of algae and scaling 1. Ensure cooling water is not directly discharged into surrounding

10-43

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Waste water treatment

1. Discharge of treated effluent

N

Catalyst change out in reactors and PSA unit

1. Metal content present in catalyst

N

2. Dust particles from catalyst fines

N

Production of Hydrogen via Syngas Route

bodies, decreasing amount of dissolved oxygen 1. Water pollution

water bodies 2. Installation of temperature sensors at discharge outlet 2. Installation of sensors for detection of elevated levels of chemical release 3. Monitor pH levels

III

IV

IV

YES

IV

A

M

1. Soil contaminatio n by metals 2. Water contaminatio n by metals

III

IV

II

NO

III

C

M

1. Proper disposal of spent catalyst

1. Affects respiratory system of work personnel

III

IV

III

NO

III

C

M

1. Don respiratory masks during catalyst change out operation

10-44

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Legend for Table 10-14 Symbol

Description

N/E/A

Normal Operation / Emergency / Abnormal Operation

P PD ENV REP OC LH

Injury to People Property Damage Environmental Impact Repetitive Overall Consequences Likelihood of Occurrence

RR

Risk Ranking

Probability Category A B C D E

Consequence Category I II III IV

Definition Possibility of repeated incidents Possibility of isolated incidents Possibility of occurring sometimes Not likely to occur Practically impossible

Health/Safety Fatalities / serious impact on public Serious injury to personnel / limited impact on public Medical treatment for personnel / No impact on public Minor impact on personnel

Production of Hydrogen via Syngas Route

Considerations Property Damage Large community Small community

A

Probability B C

D

I II III IV

Environmental Impact Major/Extended duration/Full scale response Serious/Significant resource commitment

Minor

Moderate/Limited response or short duration

Minimal to none

Minor/Little or no response needed

10-45

E

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

10.6.3 Elements of Environmental Impact Assessment

10.6.3.1 Gaseous emissions

The air pollution caused by the H2 plant may include greenhouse gases such as carbon dioxide, methane and NOx, hazardous gases such as carbon monoxide and sulphur dioxide, and also flammable gases such as hydrogen. These gases could cause much harm to the environment, for example, global warming due to greenhouse gases, acid rain due to acidic gases, ozone depletion, and even fires and explosion caused by flammable gases. Some of these gases also can cause health problems in people who are exposed to it without proper protection. Sulphur dioxide, nitrogen dioxide and ozone in the lower atmosphere can cause respiratory diseases in people. Emission limits of sulphur dioxide are 500mg/Nm3, and that of nitrogen dioxide are 700mg/Nm3. Carbon monoxide is also dangerous and can cause death through asphyxiation in excessive doses. Emission limits of carbon monoxide are 625mg/Nm3. During changeout of catalyst, it could cause the broken, fine metal catalysts to emit into the air as respirable suspended particles (PM10). PM10 refer to particulate matter of size 10mm and below. These particles have health implications as they are able to penetrate into the deeper regions of the respiratory tract. In very large amounts, the particles cause breathing and respiratory problems, and aggravate existing respiratory and cardiovascular diseases. Hence, to reduce the emissions of these vapours, mitigation measures as mentioned in Table 10-14 is essential to be adhered to.

10.6.3.2 Effluent discharge

The Pollution Control Department in Singapore regularly monitors water quality of various inland water bodies and coastal areas. For this hydrogen plant, the effluent would come from the cooling tower water and chemicals as well as the wastewater from the knockout drum which comprises of hydrocarbons. The table below shows the allowable limits for trade effluent discharged into a public sewer/watercourse/controlled watercourse.

Production of Hydrogen via Syngas Route

10-46

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

Public Sewer

Watercourse

Controlled watercourse Units in milligram per litre or otherwise stated 1. Temperature of discharge 45oC 45oC 45oC 2. BOD (5 days at 20oC) 400 50 20 3. COD 600 100 60 4. Hydrocarbon 60 10 5. Total Suspended solids 400 50 30 Table 10-15: Allowable limits for trade effluent discharge Items of Analysis

Where, Biochemical Oxygen Demand (BOD) and Chemical Oxygen Demand (COD) are used as wastewater quality indicators. There would be fees levied for trade effluent with BOD in excess of 400mg/l to 4000mg/l. For BOD exceeding 4000mg/l, the trade effluent would have to be treated to below 4000mg/l before discharging into public sewers. Effluent discharge has to be treated to meet these limits before discharge.

10.6.3.3 Waste management & minimization

Waste management to be considered for this hydrogen plant would be the disposal of the spent metal catalysts after catalyst changeout. Licensed general waste collectors will be employed for this task. It is an offence for any person or company to collect or transport waste as a business without a valid General Waste Collector License. The spent catalyst has to be carefully handled, especially because catalyst fines might get into the air and cause health problems to the people.

10.6.3.4 Energy efficiency

The National Energy Efficiency Committee (NEEC) is a committee with 3P (People, Private, and Public Sector) representation. It seeks to integrate the promotion of energy efficiency and the use of clean energy sources with the reduction of emissions of air pollutants and carbon dioxide from the production of energy. The key objectives of the NEEC are as follows:

Production of Hydrogen via Syngas Route

10-47

CN 4120: Design II Team 32

•

Safety, Health & Environment (S.H.E.)

Promotion of energy conservation through efficient use of energy in the industrial, building, and transportation sectors

•

Promotion of the use of cleaner energy sources such as natural gas and renewable energy sources

•

Promotion of Singapore as a location for the pilot test-bedding of pioneering energy technologies and as the hub for development and commercialization of clean energy technologies

As can be seen from our mitigation measures, we are actively fulfilling the above mentioned objectives of the NEEC.

10.6.4 Hydrogen Product Life Cycle Assessment

Fig 10-2: Hydrogen life cycle Life cycle assessment (LCA) is defined as a systematic analytical method that helps identify and evaluate the environmental impacts of a specific process or competing processes [R12]

. To quantitatively account for its impact on the environment, material and energy balances

Production of Hydrogen via Syngas Route

10-48

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

are performed in a “cradle-to-grave” manner on the processes required to turn raw feedstock into tangible products. Possible emissions, resource consumption and net energy consumption would serve as primary indicators for the efficiency of the hydrogen life cycle. Finally, this LCA will be used as the basis of comparison with other hydrogen generation methods to weigh the environmental benefits and disadvantages of these various methods. A study done by the National Renewable Energy Laboratory [R12], under the U.S. Department of Energy, indicated that CO2 was emitted in the greatest amount, making up 99% by weight of the total air emissions during steam methane reforming. This amount of CO2 also contributed for 89.3% of the system’s global warming potential (GWP), defined as a combination of CO2, CH4, and N2O emissions expressed as CO2-equivalence for a 100 year time frame. Moreover, methane accounted for 10.6% of the GWP. Overall, the hydrogen plant itself contributed 74.8% of the greenhouse gas emissions. Besides these gases, other hydrocarbons (C2+), NOx, SOx, CO, particulates and benzene make up the remainder of the emissions. These usually came about from natural gas production and distribution. Water was also consumed in copious amounts in the hydrogen plant in the SMR, HTS and LTS reactors. In terms of energy balance, it was determined that a major component of energy consumption was found contained in the natural gas feedstock. On a life cycle basis, for every MJ of fossil fuel consumed by the system, 0.66 MJ of hydrogen are produced on a LHV basis. This figure has also included the energy used in the production, distribution of natural gas, and in the generation of electricity to power the hydrogen plant itself [R12].

10.6.4.1 Ramifications of Hydrogen LCA

Hydrogen, an energy carrier, is perceived by engineers and scientists to be the energy system for the 21st century. This is due to its abundance in the universe. However, H2 does not exist naturally on Earth. It is mainly found on Earth as water and in organic compounds such as methane, coal, petroleum and biomass.

Production of Hydrogen via Syngas Route

10-49

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

At present, the majority of hydrogen is produced by the steam methane reforming method, since it is still the most economical choice. However, this process relies heavily on liquid or gas hydrocarbon fuels as the basic material for manufacture and as energy input for its production. Therefore, the associated environmental impact is significant, as greenhouse gases such as CO2 and NOx would be emitted. Moreover, another inherent problem in the product life cycle of hydrogen is related to the lower energy density per unit volume of hydrogen. Hence, for hydrogen to be used as transportation fuel, an energy-intensive liquefaction process is required. This not only incurs additional cost, but also contributes to the emission of greenhouse gases, leading to increased global warming. Hence, although hydrogen is indeed a promising candidate as a future energy carrier, with its reputation as a clean fuel with zero toxic emissions, the overall life cycle efficiency at present is negative (-39.6%), i.e. the energy in the natural gas is greater than the energy content of the hydrogen produced. It has also been mentioned above that for every MJ of fossil fuel consumed by the system, 0.66 MJ of hydrogen are produced (LHV basis). 12 Therefore, unless higher life cycle efficiencies are attained, the amount of resources, emissions, wastes and energy consumption would remain a stumbling block towards its widespread implementation. Most importantly, its adverse effect on the environment would still remain an issue to be resolved.

10.7 CONCLUSION

To conclude, it is an undeniable fact that safety issues should take precedence ahead of economic considerations in the area of plant design and operation. The safe running of the hydrogen plant will not only minimize human casualties and environmental harm, but can also work hand-in-hand to meet economic demands being placed on the plant.

Production of Hydrogen via Syngas Route

10-50

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

REFERENCES

[R1] : Daniel, A.C. & Louvar, J.F. (2002). Chemical Process Safety: Fundamentals with Applications. 2nd Ed., Upper Saddle River, NJ: Prentice Hall. [R2] : Mecklenburgh, J. C. (1985). Process Plant Layout, London: G. Godwin. [R3] : Hydrogen Properties. Retrieved April 12, 2008, from U.S. Department of Energy Web site: http://www1.eere.energy.gov/hydrogenandfuelcells/tech_validation/pdfs/fcm01r0.pdf [R4] : Bausbacher, E. & Hunt, R. (1993). Process Plant Layout and Piping Design, New Jersey: Prentice-Hall. [R5] : Changi Water Reclamation Plant. Retrieved April 13, 2008, from CPG Corporation Web site: http://www.cpgcorp.com.sg/portfolio/viewdetails.asp?Lang=EN&PCID=11&PDID=163 [R6] : Physical properties of gases, safety, MSDS, enthalpy, material compatibility, gas liquid equilibrium. Retrieved April 14, 2008 from Air Liquide Web site: http://encyclopedia.airliquide.com/encyclopedia.asp?CountryID=19&LanguageID=11 [R7] : Munn, R.E. (1979). Environmental Impact Assessment: Principles And Procedures. 2nd Edition. New York: Wiley. [R8] : Briffett, C. (1994). The Effectiveness of Environmental Impact Assessment in Southeast Asia. [R9] : Bass, R.E., Herson, A.I. and Bogdan, K.M. (1999) CEQA Deskbook: A Step-by-step Guide on how to Comply with the California Environmental Quality Act. 2nd edn., Point Arena, CA: Solano Press. [R10] : OSH Answers: Legionnaire’s Disease. Retrieved on April 16, 2008 from Canadian Centre for Occupational Health and Safety Web site: http://www.ccohs.ca/oshanswers/diseases/legion.html [R11] : Legionnaire’s Disease eTool : Source and Control – Cooling Towers, Evaporative Condensers and Fluids Coolers. Retrieved on April 16, 2008 from U.S. Department of Labor Occupational Safety & Health Administration Web site: http://www.osha.gov/dts/osta/otm/legionnaires/cool_evap.html#Treatment

Production of Hydrogen via Syngas Route

10-51

CN 4120: Design II Team 32

Safety, Health & Environment (S.H.E.)

[R12] : Life Cycle Assessment Of Hydrogen Production Via Natural Gas Steam Reforming. Retrieved on April 16, 2008 from National Renewable Energy Laboratory Web site: http://www.nrel.gov/docs/fy01osti/27637.pdf

Production of Hydrogen via Syngas Route

10-52

CN 4120: Design II Team 32

Instrumentation & Control

Chapter 11 : INSTRUMETNATION & CONTROL 11.1 INTRODUCTION

Instrumentation and control of the plant is critical to the plant’s operation and product quality. It helps to manage the product quality through proper control of the plant. The objectives of instrumentation and control are based on two major aspects [R1]: Safety, Health and Environment (SHE) – A safe plant operation prolongs the life of the

expensive equipment and protects the health of the operators. Safe and smooth plant operation is achieved with proper control that will detect abnormalities and effect the corrective actions to maintain the process variable within the permissible limits. In addition, safe operation keeps the surrounding environment in check by preventing unexpected harmful emissions to the atmosphere. Product quality – Through control and instrumentation, the plant would be able to respond to

changes in operating conditions quickly and effectively, thus ensuring minimal disturbances to the plant. Therefore, the production scheme remains unperturbed and generates products of constant quality and yield. The protection scheme of the safety design is shown:

Fig 11-1: Typical layers of protection in a modern chemical plant

Production of Hydrogen via Syngas Route

11-1

CN 4120: Design II Team 32

Instrumentation & Control

This chapter looks at the process control system of the steam methane reformer (SMR). The control and instrumentation of this section is fully developed by our team and will be explained in detail. The installation of critical alarms and automatic safety lock systems to protect property and employees during emergencies are also explored. The controls are employed bearing in mind that non-essential controls were reduced to minimize cost of the process control system, without compromising on safety and quality of products.

11.2 PROCESS CONSIDERATION AND DESCRIPTION Aims and objectives

The furnace and SMR reactor are two important units to the plant. The steam methane reformer generates hydrogen while the furnace controls the SMR process. Controlling the steam methane reforming process ensures constant yield and profitability of the plant. It also prevents any possible runaway reactions. As the SMR process is endothermic, the heat duty is supplied by the furnace and hence combustion is strictly controlled to maintain constant reaction temperature. In addition, the high temperature and pressure associated with furnace operations creates more necessity to impose proper control on the furnace. The lack of control of process variables within the furnace may pose environmental and safety issues. The important process parameters related to these two units are listed below:

Important process variables of Furnace

1. Temperature 2. Pressure

Important process variables of SMR reactor

1. Temperature 2. Steam methane ratio

Production of Hydrogen via Syngas Route

11-2

CN 4120: Design II Team 32

Instrumentation & Control

11.3 PROCESS CONTROL METHODOLOGY

The following steps will be used in the formulation of our instrumentation and control design. 1. Identify the variables to be controlled, measured and manipulated 2. Select the control strategy and structure 3. State the controller settings The general control strategies are feedback and feed-forward control. Feedback control provides an easy control of variables without extensive knowledge of the process. Feed-forward control provides a safer control as compared to feedback control because it allows corrective action to be taken before the process variables go out of hand. The advantages and disadvantages of feedback and feed-forward control are summarized in Table 11-1:

Feedback

Feed-forward

•

Advantages Little knowledge is required of the control process

•

Disadvantages Poor control occurs if time lags are significant

•

Disturbance need not be measured

•

Closed-loop instability may occur

•

Corrective action will be taken regardless of the source and type of disturbances

•

Process upset takes place before corrective action is taken

•

Corrective action is taken before process upset occurs

•

In depth knowledge of the process is required

•

“Perfect” control can be achieved.

•

Ideal controllers may not be present to effect perfect control

Table 11-1: Summary of advantages and disadvantages of feedback and feed-forward control

Production of Hydrogen via Syngas Route

11-3

CN 4120: Design II Team 32

Instrumentation & Control

11.4 SELECTION OF CONTROLLED, MANIPULATED AND MEASURED VARIABLE

A good control system can be achieved if the appropriate controlled and manipulated variables are chosen. These are some guidelines we followed when selecting the variables [R1]. For controlled variable:

1. All variables that are not self-regulating must be controlled 2. Choose output variables that must be kept within equipment and operating constraints 3. Select output variables that represent a direct measure of product quality or that strongly affect it 4. Choose output variables that interact with other controlled variables 5. Choose output variables that have favourable dynamic and static characteristics For manipulated variable:

1. Select inputs that have large effects on controlled variables 2. Choose inputs that rapidly affect the controlled variables 3. The manipulated variables should affect the controlled variables directly rather than indirectly 4. Avoid recycling of disturbances For measured variable:

1. Reliable, accurate measurements are essential for good control 2. Select measurement points that have an adequate degree of sensitivity 3. Select measurement points that minimize time delays and time constraints

Production of Hydrogen via Syngas Route

11-4

CN 4120: Design II Team 32

Instrumentation & Control

In our design, the major variables affecting the overall safety of the furnace are identified as follows: 1. Pressure

Pressure control within the furnace is essential because it can compromise the performance. Pressure drop within the reformer tubes has to be kept to a minimum for favourable reaction conversion. Backflow within pipelines has to be avoided as well. 2. Temperature

Temperature control is essential to ensure that the reaction within the reformer tubes proceed smoothly. It is also important to ensure that the materials of construction remain intact by ensuring that the temperature within the furnace is kept below the creep temperature of the materials. Flue gas temperature has to be controlled to prevent acid gas condensation and to comply with governmental regulation. 3. Composition and Flow

The steam/methane ratio has to be kept above a minimum to prevent coking and high pressure drop within the tubes. In addition, the air-to-fuel ratio into the furnace has to be kept constant to ensure absolute combustion and maximum efficiency.

11.5 DETAILED CONTROL DESIGN FOR REFORMER FEED

11.5.1 Steam-to-Methane Ratio Control

One important control parameter in steam-methane reforming is the steam-to-methane ratio. A low ratio is undesirable as it promotes the side reaction of coke formation on the catalyst, which deactivates it and requires expensive replacement. Nonetheless, a high steam-to-methane ratio will result in better conversion, but at the expense of elevated operating costs due to the high cost associated with superheated steam. Hence, a compromise between methane conversion and operating expense has to be made and in industries, this ratio is typically kept at 3:1.

Production of Hydrogen via Syngas Route

11-5

CN 4120: Design II Team 32

Instrumentation & Control

Fig 11-2: Control scheme for steam/methane ratio control To achieve this, a ratio control depicted in Fig 10-2 is employed to maintain the ratio between steam and methane at 3:1 as stated in our design problem. The flow rate of the natural gas stream is measured and transmitted by FT-101 to the ratio station FY-101. At the ratio station, this signal is multiplied by an adjustable gain whose value is the desired ratio. The output signal from the ratio station is then used as the set-point for flow controller FIC-102. This feedforward controller then adjusts the flow rate of the imported superheated steam by manipulating the opening of the diaphragm valve FCV-102 using pneumatic signals. In the preliminary design of this ratio control, it is assumed that molar flow rate is equal to the volumetric flow rate which implies that possible pressure and temperature fluctuations in process streams are not compensated.

Production of Hydrogen via Syngas Route

11-6

CN 4120: Design II Team 32

Instrumentation & Control

11.5.2 Pressure Control Loop for Expander

A feedback control loop is used for all the expanders in the plant. The controlled variable is the outlet pressure of the expanded vapour, while the manipulated variable is the inlet pressure to the expander.

Fig 11-3: Control scheme for pressure control of expander discharge The pressure transmitter (PT) would detect any deviations from the set point and send a signal to the pressure controller (PIC) so as to adjust the valve which changes the inlet pressure to the expander. This would then serve to bring the controlled variable back to its set point.

11.5.3 Temperature Control Loop to Preheat SMR Feed

The diagram below depicts a temperature control loop prior to the entry of the process fluid into SMR. It is important to control the inlet temperature to SMR because it will affect the methane conversion in the SMR. The inlet temperature to SMR can be controlled by varying the flow of the SMR effluent through heat exchanger, HX-102. In this case, the temperature of the preheated SMR feed is the controlled variable, while the flow of the SMR effluent is the manipulated variable. A cascade control loop is employed as follows: 1. The master controller is TIC-101, and the temperature of preheated SMR feed serves as the set point for the slave controller, which is 539oC. 2. The two control loops are nested, with the secondary control loop (for the slave controller, FIC-103) located inside the primary control loop (for the master controller, TIC-101)

Production of Hydrogen via Syngas Route

11-7

CN 4120: Design II Team 32

Instrumentation & Control

Fig 11-4: Control scheme for temperature control loop to preheat SMR feed

The advantage of cascade control is to ensure a swift detection of any flow rate deviation of the SMR effluent for which necessary adjustments will be made by FIC-103 before an upset in the temperature of the preheated SMR feed can be effected. This provides a fast response to deviations from set point values and a feed-forward control loop for the secondary control loop is subsequently adopted for a faster response time.

11.5.4 Composition Analyzer for SMR Effluent

The composition analyzer A-004 can be used to determine reaction completion by measuring the methane content of the effluent stream. The composition analyzer can be either an infra-red or a chromatographic analyzer [R2] which measures a range of 0 to 10% methane in the background of hydrogen and carbon monoxide. The advantage of using this composition analyzer in that a desired furnace temperature profile can be arrived at by manipulating the furnace temperature to achieve the desired degree of conversion.

Production of Hydrogen via Syngas Route

11-8

CN 4120: Design II Team 32

Instrumentation & Control

11.6 DETAILED CONTROL DESIGN FOR SMR FURNACE 11.6.1 Air-to-Fuel Ratio Control

Operating without sufficient air can lead to fuel wastage due to inefficient combustion of air and can even escalate to a hazard when the flammable products of incomplete combustion ignite in the convection section.[3] However, excess air can reduce the efficiency of the furnace due to large volumes of air heated to exit stack temperature without producing useful heat transfer. From the above mentioned points, it can be seen that having an air-to-fuel ratio control is crucial for the safe and efficient operation of furnace.

Fig 11-5: Control scheme for air-to-fuel ratio control The air-to-fuel control scheme shown above is designed to fix the excess air at 15% to ensure complete and stable combustion. In this control scheme, the disturbance variable is the flow rate of the furnace fuel while the manipulated variable is the combustible air flow. The flow rate of furnace fuel, which is a combination of PSA purge gas and natural gas, is measured and transmitted by FT-106 to the ratio station FY-106. At the ratio station, this signal is multiplied by an adjustable gain whose value is the desired ratio. The output signal from the ratio station will

Production of Hydrogen via Syngas Route

11-9

CN 4120: Design II Team 32

Instrumentation & Control

act as the set-point for feed-forward controller FIC-107 to control the valve opening of diaphragm valve FCV-107 to adjust the flow of combustion air to the burners of furnace so as to maintain the excess air at 15%.

11.6.2 Temperature Control Loop to Regulate Effluent Exit Temperature

Natural gas is burned in the furnace to supply the heat duty required for the endothermic conversion of methane to hydrogen. Therefore, in this control system, the manipulated variable is the flow rate of natural gas, while the controlled variable is the temperature of the SMR effluent being heated.

Fig 11-6: Control scheme for temperature loop to regulate effluent exit temperature A cascade control loop is employed as follows: 1. The master controller is TIC-105, and the temperature of SMR effluent serves as the set point for the slave controller 2. The two control loops are nested, with the secondary control loop (for the slave controller, FIC-105) located inside the primary control loop (for master controller, TIC-105)

Production of Hydrogen via Syngas Route

11-10

CN 4120: Design II Team 32

Instrumentation & Control

This cascade control will ensure that any deviations in the flow rate of natural gas would be detected and FIC-105 would make adjustments to the flow rate of natural gas even before it could cause an upset to the SMR effluent temperature. This is the advantage of cascade controls, whereby a second measured variable is located close to the potential disturbance and its associated controller reacts quickly, where it offers very fast response time to deviations from set point values. To improve the response time further, a feed-forward control loop for the secondary control loop was employed.

11.6.3 Pressure Control Loop to Regulate Furnace Draft

Fig 11-6: Control scheme for pressure control loop to regulate furnace draft A pressure control loop is designed to maintain a small negative pressure at the top of the radiant section just before the convection section. This is because if the pressure is positive in the radiant section, it will cause the hot flue gas to leak outward and damage the steel structure of furnaces, thus shortening the lifespan of furnace. The pressure can be maintained at negative pressure by adjusting the opening of the stack damper using a feedback control. When the pressure before the convection-section inlet deviates from the set-point value, the pressure transmitter PT-001 will send a signal to pressure controller PIC-001 in the feedback control loop which will then adjust the opening of the stack damper to

Production of Hydrogen via Syngas Route

11-11

CN 4120: Design II Team 32

Instrumentation & Control

either increase or decrease the draft. In general, when the stack damper closes, the draft decreases and vice versa. A position transmitter ZT-001 will measure the opening of the stack damper and reflect it via the indicator ZI-001.

11.6.4 Flue Gas Exit Temperature Control

Flue gas exit temperature control is necessary for two reasons: 1) to prevent corrosion attack caused by acid gas condensation and 2) to conform to governmental regulation for flue gas exit temperature. The temperature control at the stack exit is achieved by a cascade control configuration which consists of a primary control loop utilizing TT-104 and TIC-104 and a secondary control loop that controls the flow of condensate via FT-104 and FIC-104.

The exit flue gas temperature will be measured by the temperature transmitter TT-104 and will be used by the master controller TIC-104 to establish a set point for the secondary loop controller, FIC-104.

Fig 11-7: Control scheme for temperature control loop for flue gas exit

Production of Hydrogen via Syngas Route

11-12

CN 4120: Design II Team 32

Instrumentation & Control

The reason why cascade control is used here instead of conventional feedback control is because if feedback control is used and there is variation in the flow of the condensate steam which cause a change in flue gas temperature, this change can only be effected after the temperature controller takes corrective action to adjust the condensate flow. In our case where cascade control is used, the flow controller FIC-104 will respond very fast to hold the condensate flow at its set point without causing any disturbances to the flue gas temperature.

11.6.5 Tube Metal Temperature Indicator

Reformer tubes are designed to operate at a particular pressure and temperature. Conditions such as flame impingement, poor radiant-heat distribution, interior tube deposits (coke) can cause localized overheating of tube known as hot-spots, which could result in high temperature creep. Hence, it is mandatory to have a temperature monitoring device to monitor furnace tube skin temperature to prevent the plastic deformation of tubes.

Fig 11-8: Tube metal temperature indicators To achieve this, thermocouples can be welded onto reformer tubes. In addition, temperature high alarms should also be incorporated together with the thermocouples to alert operators of high tube skin temperature. This practice is adopted in our design of the steam methane reformer as a safeguard against tube failure. A common practice in industry requires the

Production of Hydrogen via Syngas Route

11-13

CN 4120: Design II Team 32

Instrumentation & Control

installation of thermocouples on 15% of the reformer tubes but due to space constraints on P&ID, only 4 thermocouples are reflected.

11.6.6 Analyzers for Furnace Control

Analytical devices such as oxygen analyzers and combustible detectors are used to control the operation of the furnace. The oxygen sensor A-002, which is a solid state heated zirconium oxide probe that is inserted directly in the stack, can be used to measure the oxygen content in the flue gas in order to compute the excess air that is being delivered. As such, the efficiency of furnace combustion can be maximized by throttling the air entering the furnace to maintain the oxygen content at a desirable value. The sensor can also be tied to a low oxygen alarm to warn operators if a hazardous furnace atmosphere is developing. Furthermore, a combustible analyzer A-003, which uses an infra-red beam directed across the stack to measure the amount of carbon monoxide content in the flue gas, is also utilized for furnace control. The combustible analyzer usually works in tandem with an oxygen analyzer to provide a correlation between excess air and unburned fuel going up the flue stream. For this reason, they can serve as a form of backup to check for the effectiveness of the air-to-fuel ratio control loop. Although combustible analyzers can sometimes be used to automatically adjust the air/fuel ratio, this control scheme was not adopted in our control of furnace due to reliability issue of the combustible analyzer.

11.7 Safety Devices 11.7.1 Pressure Relief Valves

Relief valves are required to provide an outlet for over-pressurised fluids to prevent rupture of pipelines so as to protect operators from possible hazards of over-pressurizing equipment. In addition, relief valves also prevent damage to adjoining property; reduce insurance premiums and chemical losses during pressure upset. With these in mind, it is in fact stipulated by governmental regulations that installations of pressure relief systems are compulsory.

Production of Hydrogen via Syngas Route

11-14

CN 4120: Design II Team 32

Instrumentation & Control

In this preliminary piping and instrumentation diagram (P&ID) design of the SMR reactor/furnace, the pressure relief devices are designed to be installed at every point identified as potentially hazardous. Since most of the pipelines are in gas service, safety valves are chosen for pressure relief. Safety valves pop open to release the excess pressure when the operating pressure exceeds the set pressure. The only exception is the pipeline with steam in service. Hence a safety relief valve, which functions as relief valve for liquid and safety valve for steam, may be required [R4].

11.7.2 Process Alarms

The function of a process alarm is to warn operators of impending dangers when process parameters such as temperature, pressure, flow or level exceed or fall below the permissible limits. In general, alarms can be categorised as follows [R5]: Type

Description

Function

Type 1 Alarm

Equipment status alarm

Type 2 Alarm

Abnormal measurement alarm

Type 3 Alarm

An alarm without an adjoining sensor

Indication to whether an equipment is switched on or off When activated, it acts as an indication that the reading taken by the sensor is outside acceptable limit Alarm that is directly triggered by process instead of sensor signal when the process parameter is out of specification. Knowledge of actual process value is not required Serve as a backup to the regular sensor in the event that it fails

Type 4 Alarm Type 5 Alarm

An alarm with an adjoining sensor Automatic shutdown or start- Typical type of alarm that is widely used up system Table 11-2: Summary of different categories of alarm

Production of Hydrogen via Syngas Route

11-15

CN 4120: Design II Team 32

Instrumentation & Control

Upon activation of the alarm, an annunciator, either in the form of visual displays (flashing light on control panels) or audible sounds (horns or bells), will be triggered. Unless acknowledged by plant operators, the alarm will remain in activation. If the abnormal situation that sets off the alarm is deemed to be potentially dangerous, an automated corrective action will be initiated by the safety interlock system to shut down the affected unit. Excessive number of unnecessary alarms should be avoided because frequent “nuisance alarms” make plant operators less responsive to crucial alarms and may obscure the root cause of the abnormal situation in the presence of many unimportant alarms. Hence, alarms should only be installed at locations deemed absolutely necessary. The alarms in the P&ID and their functions are summarized in the following table: Alarm Tag No. Description for Alarm Cause THA -104 Flue Gas High Temperature THA-201, THA-202, THA-203, THA-204 SMR Furnace High Tube Skin Temperature ALA-002 Low Flue Gas Oxygen Content PHA -001 SMR Furnace High Pressure FLA -201 Fuel Gas Low Flow FLA -202 Combustion Air Low Flow FLA -203 SMR Feed Low Flow FLA -204 SMR Effluent Low Flow Table 11-3: Summary of alarms in our P&ID and their causes 11.7.3 Safety Interlocks or Emergency Shutdown System (SIS or ESD)

The control loops that are designed in the P&ID form the basic process control system (BPCS) which acts as primary protection against deviations in process parameters. During normal operations, BPCS can provide acceptable control but this may not be the case during abnormal or emergency situations. For this reason, implementation of SIS or ESD is required to serve as a backup especially when BPSC components malfunction or when there is utility failure. As such, considerations must be made for SIS and ESD to function independently of the BCPS.

Production of Hydrogen via Syngas Route

11-16

CN 4120: Design II Team 32

Instrumentation & Control

In view of the drastic measure such as the complete shutdown of process unit, the SIS and ESD can only be considered as the last resort to protect the equipment and prevent injury to personnel and used only when critical process variables go beyond the specified allowable operating limit. Although SIS is essential for safe plant operation, unnecessary plant shutdown should be precluded as it can reduce throughput due to downtime and may cause products to go off-specification during subsequent plant start-up.

11.7.3.1 Implementation of SIS or ESD for the protection of nickel catalyst

Fig 11-9: SIS control scheme for protection of nickel catalyst As mentioned earlier, a low steam/methane ratio is undesirable because it can deactivate the nickel catalyst due to coke formation. To prevent this situation from arising, a SIS should be in place to tackle any abnormalities in the ratio. The SIS to be implemented is illustrated in the above control scheme. When the steam/methane ratio falls below the limit of approximately 3:1, a flow switch FSL-001 will trigger off an alarm. If the ratio continues to fall below the critical limit (approximately 2.7:1), the natural gas stream will be shut off via FSL-002, which trips a solenoid that control the transducer I/P-101 and causes FCV-101 to be closed.

Production of Hydrogen via Syngas Route

11-17

CN 4120: Design II Team 32

Instrumentation & Control

In designing such a SIS, several points must be taken into consideration. Firstly, FCV101 must be a quick-closing valve (4 to 5 seconds for full closure) so that the flow of natural gas can be almost instantly stopped when there is a large deviation in reforming steam, thus protecting the reformer catalyst. FCV must also be a single-seated tight shut-off valve, to prevent leakage during the time while the electric-operated shutoff valve is in the process of closing.

11.8 Additional Considerations in Process Control 11.8.1 Redundancy of Air Blowers and Expanders

The reliability of plant equipment is critical to successful plant operation. Hence, spare air blowers and expanders are provided in parallel to the commissioned equipment and are placed on hot standby to ensure the continuous operation of the hydrogen plant whenever any of the equipment undergo mechanical failure or are sent offline for maintenance purposes, which is a common phenomenon during normal operation.

Fig 11-10: Redundancies used in process control

Production of Hydrogen via Syngas Route

11-18

CN 4120: Design II Team 32

Instrumentation & Control

11.8.2 Isolation Valves and Bypass

Isolation valves, most commonly gate valves, are installed upstream and downstream of process equipment such as heat exchangers, air pre-heater, control valves and air blower to ensure the operability of the plant whenever there is maintenance or repair of the equipment. When isolated, a bypass as illustrated must be present to divert the process flow to the other side of the process equipment.

B/P

Fig 11-11: Bypass of pipelines

11.9 REFERENCES

[R1]: Seborg D. S., Edgar T. F., Mellichamp D. A. Process Dynamics and Control, John Wiley & Sons (2004) [R2]: Liptak B. Instrument Engineer’s Handbook: Process Control and Optimization, CRC Press (2005) [R3]: Lieberman N. P., Lieberman E. T., Working Guide to Process Equipment. 2nd Edition McGraw-Hill [R4]: Crowl D.A., Louvar J.F. Chemical Process Safety: Fundamentals with Applications. 2nd Edition, Prentice Hall PTR (2002) [R5]: Connell, B., “Process Instrumentation Applications Manual”, Mc-Graw-Hill, New York (1996)

Production of Hydrogen via Syngas Route

11-19

E-105

AIR FROM INTAKE FILTER

MAKEUP NATURAL GAS TO FURNACE

CONDENSATE

A

ZI

ZT 001

001

001

PSV-008

opacity

E-106

THA 104 ALA 002 I/P 001

B-101

A 002

PSV-009

TIC

TT 104

TE 104

O2

104 FIC

A O2

003

CO

AP-101

104

AP-102

I/P 104

PSV-007 COMBUSTION AIR

PSV-010 FT 104 TE 103

NATURAL GAS FROM PIPELINE

B-102

TI

FT 107

103 TI

FLA PI

FI

FIC

203

103

101

FT 203

PHA

PIC

001

001

I/P 107

I/P 101 TIC

42 BAR HP STEAM FOR EXPORT

FT 103

101

SMR EFFLUENT

FIC

TI

203

203

107

202 TE 203

THA

I/P 105

FIC FLA

PT 001

I/P 103

FIC 105

PSV-005

TT 101

FT 105

TE 102

102

203

PSA PURGE GAS 2 ( CO removed)

PSV-006

FI

FT 202

202 FY 106

101

HX-102 FT 101

TI 201

THA

HX-101

F-101

TE 201

FLA

FI

201

201

FT 106

TE 202

PSV-012

201 See Detail A TI

E-101

FY

202

TE 204

101 PSV-002

THA TI

202

204

E-102 FIC

THA

102

204 PSV-001

SMR EFFLUENT TO HX-102

PSV-011 I/P 102 FT 102

E-103

IMPORTED STEAM

PI

AT 004

PSV-003

102

A 004

Methane

TE 105

FT 204

TT 105

TIC 105

Detail A FI

FLA

204

204

PIC PT

E-104

PSV-004

Applicable to all expanders