Dubrovnik 2009 - Energy Analysis Of Distillation Units In The Process Of Yeast And Alcohol Produc - Anastasovski, Markovska, Meshko, Raskovic

Energy analysis of distillation units in the process of yeast and alcohol production Aleksandar Anastasovski* Ss Cyril and Methodius University Faculty of Technology and Metallurgy, Skopje, Macedonia email: [email protected] Vera Meško Ss Cyril and Methodius University Faculty of Technology and Metallurgy, Skopje, Macedonia email: [email protected] Predrag Rašković University of Nis Faculty of Technology, Leskovac, Serbia email: [email protected] Liljana Markovska Ss Cyril and Methodius University Faculty of Technology and Metallurgy, Skopje, Macedonia email: [email protected]

ABSTRACT This paper reports a simulation analysis of energy consumption in steam distillation units, which are used as a part of manufacture in the factory for yeast and alcohol production. Distillation units are stored in the final stage of alcohol manufacture route following the anaerobic fermentation processes. The main products of distillation process are: 96% Vol. ethyl alcohol and technical alcohol, which presents the mixture of ethanol, aldehydes and some other less volatile components than ethanol. Distillery is consisting of columns in series. First of them is distillation column, but other are rectification columns. Their work parameters are determined by inlet feed composition. When composition is not changing a lot, whole distillation system is close to steady state. Flow, temperature and pressure are parameters for monitoring. The simulation analysis has been made by simulation software Aspentech HYSYS. The validity of the model in terms of mass flow rate, temperature and pressure was examined using actual plant data at steady-state conditions. Comparison between model results and plant data shows good consistency and enabled the retrofit design with improved efficiency of the unit. Keywords: Aspentech Hysys, process simulation, ethanol, distillation 1. INTRODUCTION Last few decades increase software using for simulation of production facilities with improving their work [1]. There are many types of software that are in use these days. Some of them are specialized for some specific processes in chemical technology, but some of them *

Corresponding author: Aleksandar Anastasovski, e-mail: [email protected]

have wide field of possibilities. Using of simulation software becomes very useful in many critical areas. Software can predict some problems that could become, finding solutions for any case, upgrading of present process with choosing the right project proposed solution, predict of process dynamics in case of inlet disturbance and many other features. So, simulation software become great solution in process complexity, such as uncertainly, dynamic behaviour and its feedback. There are developed algorithms which could be used for exploring and optimizing the real process with this kind of software [2]. 2. BASE CASE: THE SEPARATION PROCESS This work is based on previous edited works. That works [3], [4], describe distillery - real production system as subsystem of yeast and alcohol factory production process. 2.1 Description of the distillation system Alcoholic fermentation is natural process where sucrose and other simple sugars are transforming in ethanol. Enzymes that are produced by different kind of industrial yeasts (such as Saccharomyces cerevisiae, Saccharomyces vini, etc.) are initiators for starting of alcoholic fermentation process. Alcoholic wort is the final product of that fermentation. It is mixture of ethanol, water and components that are contained in raw material as well as byproducts of the fermentation process. The main task of the production is separation of purest potable ethyl alcohol. Distillation columns are used for its separation. The purpose of this work is to simulate this real system for distillation using simulation software and comparing simulation result data with the same parameters in real system. That system is consisting of 3 distillation columns in series. Alcoholic wort (1), steam (3) and cooling water (22) are inlet streams for this distillation system. Every column separate group of components with similar physical and chemical characteristics at atmospheric pressure of work. The first component group represent solid particles and components with high point of evaporation, such as sugars, proteins and water. First column (DC-100, figure 2) is designed for working conditions with higher temperatures. There is crude separation of heavy and light phase. Inlet stream (feed inlet) as alcoholic wort is heated directly with steam. Heating cause forming of vapour (light phase) and liquid (heavy phase). Light phase contain our product, that is interesting for production. Heavy phase is going out throw “Silage” stream on the bottom of the column. Light components are condensed in series of heat exchangers. They go out throw the stream 6 (Figure 1) on the column top. First heat exchanger is using inlet alcoholic wort (1) for cooling of light components. That process use heat to prepare alcoholic wort to appropriate temperature (2) for feeding the column. Three heat exchangers in series (HE-100, HE-101, and HE-102) are shell and tube heat exchangers, in vertical position with connection for outlet of gaseous and liquid phase (phase separation). Specific construction of heat exchangers to separate gaseous and liquid phase determine connections between exchangers. Gaseous phase (7) is going to the next heat exchanger and liquid phase (38) is mixing with other liquid phases condensed in other heat exchangers at atmospheric pressure (open pipe). The same is for second and third heat exchanger with exception for third, where connection for gaseous phase is vent to the atmosphere. Vent stream (44) contain gases that are present in alcoholic wort. They are formed in process of fermentation, such as CO2. All summarized condensates from all heat exchangers are mixed in “MIX – 100” and form new stream (9) that is inlet for second column in series. All three columns have similar cooling system as first one. Cooling in second and third heat exchanger as well as other cooling systems for other columns is done with cooling water.

Figure 1. Graphical presentation of distillation system for ethanol production Great parts of component that have temperature of evaporation lower then ethanol are separated in second column. All of them are contained in top outlet stream (10) in column DC-101. The main components of that stream are acetaldehyde and ethanol. Ethanol is present because of component molecules interaction. Condensate of first heat exchanger is consisting partially of heavy components. It return back to the column and is known as reflux (11). All other condensates (18) are collected in tank for technical alcohol storage. Bottom outlet stream from column DC-101 (13) is feeding stream for column DC-102. Column DC101 is positioned higher then column DC-102, so there is not need pump for enforces the flow. That stream contains different kind of alcohols and organic acids present in alcoholic wort. Their separation is the aim of third rectification column DC-102. Heavy components like water and acetic acid from column DC-102 are separated and transported with outlet stream from the bottom of the column named “Luter” (21). It has the highest operating temperature and pressure in this column. Separated components are more evaporative than ethanol at the top of column. That stream (16) is going to technical alcohol storage tank. Better purification of ethanol depends by fusel oil separation as well as separation of other present impurities. Fusel oils (32) are mixture of higher alcohols. They are taken from trays where its concentration is highest. This alcohol mixture has specific properties when this mixture is dissolved in water at appropriate concentration. In that case, layer of non soluble components are formed on the top. So, decanting is appropriate method for its separation. Decanted stream (34) takes them and transport to place for them. Soluble part (35, 50, and 51) is going back to the column DC-102. Potable alcohol (19) is taken from the tray with little bit lower temperature then temperature of ethanol boiling. In that tray concentration of ethanol is maximal, because it is azeotropic mixture ethanol – water (96% vol. ethanol). That stream is cooled and transported to the potable alcohol storage tank.

3. MODEL DEVELOPMENT ON ASPEN HYSYS PLATFORM ASPENTECH HYSYS software is CAD/CAM type of software. It interactively changes whole simulated system in case to change some of inputted parameters. Simulation method is consisting of few steps. That is described in further text for distillation system simulation model development. Hyprotech (Aspentech now) is a leader in the area of process simulation. This technology is possible to have tremendous impact on not only the way that engineers approach control strategy development, but on how we approach modelling in general [5]. 3.1 Method of distillery simulation Simulation model of distillery using software is done in few steps. As first step is input of basic data for the system. That means input of components (software data base) and choosing of thermodynamic model that will be used for simulation of component`s physical and chemical characteristics. Components that are present in distillation system are: ethanol, propanol-2, buthanol-2, glycerol, acetic acid, lactic acid, pyruvic acid, ethyl acetate, acetaldehyde, H2S, CO2, water and air. NRTL (Non Random Two Liquid) is thermodynamic model that was chosen for this simulation. This simulation case is without chemical reactions. All processes are based on physical characteristics of present components. Equitation (1) is basic NRTL for binary system. Every of parameters are determined with equations (2) and (3) [6]. GE G21 21 G1212   (1) x1 x2 RT x1  x2G21 x2  x1G12

G12  exp   12 

G21  exp   21 

(2)

b12 RT (3) b21  21  RT There are three specific parameters ∝, b12 and b21 that are not depend of composition and temperature. They are specific to a particular pair of species. Multi component systems use this equitation, but it is more complex. It depends on number of components. Next step is starting simulation environment with adding system`s inlet stream (1). After that is starting with adding of equipment step by step as process is going to be done.

 12 

Figure 2. The first column with its cooling system, part of distillation system

Heat exchangers – (Figure 2) are determined with inlet and outlet streams, temperatures, pressures and flows of those streams, heat exchanger type, number of tubes, diameter, tube`s position, shell diameter, pipe connection in fluid flow spaces, material of construction, fouling factor, horizontal or vertical position of HE (all that for tube and shell type of heat exchanger), pressure drop in shell side and tube side. These are parameters that specify all heat exchangers in this system. There are three heat exchangers in series specified for the first column that are cooling top outlet stream. Cooling fluid for the first heat exchanger is alcoholic wort as the main feed of the system. Cooling water with specific characteristics is used for cooling in other two heat exchangers. These heat exchangers are positioned vertically and have duty to cool the top outlet stream as well as phase separation. Simulation model need phase separation. There is added phase separator to every heat exchanger outlet. So, phase separators separate gases and send them to the other following heat exchanger of the series. All liquid phases, separated after heat exchangers are sent to mixer with aim to form new stream that will feed second column. Mixer is specified to take the lowest pressure of streams (atmospheric pressure). Phase separators are specified to separated phases in conditions that they leave heat exchangers. All heat exchangers are positioned on higher level of columns, so streams not need pumps, but in simulation model pump must be added to simulate that pressure by higher level position. The last heat exchanger in series needs to have outlet to the atmosphere, where separated CO2 and air is going. That gases are coming from alcoholic wort, where are formed in process of fermentation. They are separated in correct operation of distillery. Mixing of three streams with different temperatures provoke small evaporation. That is simulated with phase separator, where (47) is stream of evaporation. Quantities are very small. Distillation column 1 (DC-100) – (Figure 2) the basic specifications that are need for column specification are: number of trays, high and diameter of trays, high of tray`s overflow, type of trays, temperature and pressure on the top and on the bottom of column, feeding stream inlet tray, specified trays for outlet streams, as well as the tray of steam inlet. This is case with directly injection of steam in column, so pressure and temperature of steam are very important. Quantity of steam that is need for column is interactively determined. Tray efficiency coefficient is needed to be input. It is used to be determined by the results of simulation. Physical and chemical characteristics of outlet streams (6) and (5) are simulation values. Those values are used in further simulation of process. Distillation column 2 (DC-101) – This column (Figure 3) need all specifications given for determination of the first column in series, but also need some new specifications. This column has reflux, so that kind of equipment is chosen. The first heat exchanger is contained into the column as one equipment unit. That means simulation of two heat exchangers in series. Reflux ratio is important specification for this column.

Figure 3. The second and third columns in series, part of distillation system Distillation column 3 (DC-102) – This is the most complicated part of simulation of this system (Figure 3). There are few inlet, outlet and recycled streams. Basically, this column is determined as other two, but there are added new specifications such as outlet and inlet stream`s tray number, reflux ratio and flows of outlet streams. Number of specifications is equal to number of freedom degree of the column. That must satisfy to be zero degree. One of streams is going out to be purified. After its purification the pure components are going back to the column. That is (32) stream. It is solving with water (33) in online mixer for two streams. Component splitter X-100 is used for decanting simulation. Component splitter is specified to split defined components and with determined efficiency. One outlet stream is going to drainage with fusel oils (34); other (50) is going back to the column DC-102. Recycling stream is simulating with logic connectors. That kind is connector for recycling RCY-1 (Figure 3). There are specified returning conditions of stream, such as temperature, pressure and flow. Potable alcohol stream, technical alcohol stream and fusel oils coolers are determined as all other heat exchangers. Cooling water is used as cooling medium. There are also made comparation of steam and cooling needs for changing of flow. This simulation model is used for this comparation and simulated situation of variables is used as referent values presented as 100% (100% of variable value). This comparation is made to be determined the optimal energy need for this plant. The main preferences are: - Column temperature and pressure profile is not changing - Steam and cooling fluid characteristics are the same - Ethanol concentrations are the same in every part of plant - Feeding fluid, steam mass flow and energy taken from coolers are changed The basic idea of this analysis is determination of optimal work of distillation plant.

4. RESULTS AND DISCUSSION Results that are calculated with software and the values of real distillation system are compared. Error analysis is done. Deviations are represented in Table 1. According to represented data, values of simulated parameters and parameters in real system are evidently close. This prepared simulation model of distillery could be used for further estimation of system. That means estimation of energy that would be needed for any change in distillery operating and estimation of equipment before doing correction in some part of system. Alcoholic wort composition deviation could be one of disturbances of the system. Changing of alcoholic wort composition could be happened as a result of fermentation problems. This model could be also used for upgrade of present system with new equipment and its design and efficiency estimation. Table 1. Deviation between simulated and real values of distillery system parameters (error analysis)

Stream

Pressure deviation (%)

Temperature deviation (%)

Flow deviation (%)

4

2,000

0,310

2,315

14

1,960

1,163

26,666

15

4,761

0,661

7,894

5

10,775

0,588

0,689

1

0,542

0,000

0,184

2

0,732

0,000

0,184

6

2,316

0,795

20,463

7

2,316

13,841

11,583

9

4,735

4,666

11,196

10

2,316

1,273

13,647

11

2,316

7,246

8,080

16

2,316

0,000

6,944

17

2,316

0,000

13,157

18

2,316

1,550

24,000

19

0,886

7,977

31,654

20

0,886

0,000

31,654

21

4,166

0,582

36,398

13

13,846

5,654

24,769

With this analysis, we approved the simulation model. So, we can try theoretically to determine optimal working state of plant. There are rewieved 5 cases. Case no. 3 is simulation model. Other cases are simulation of 50%, 75%, 125% and 150% of feeding flow value in case No.3. So, values of steam in every column, as well as total steam needed for every case are relative value of steam needs in case No.3. The same is used for cooling needs. The results are shown in table 2 and 3. The graphical presentation of results are given on Figures 4, 5, 6 and 7.

Table 2. Simulation results for every column separately

Case

1 2 3 4 5

% of feed flow

% of steam needs (DC-100)

150 125 100 75 50

152 127 100 75 50

% of cooling (DC-100)

168,67 142,79 100,00 75,21 50,15

% of steam needs (DC101) 200,00 166,67 100,00 93,33 73,33

% of steam needs (DC102) 136,73 114,29 100,00 67,35 44,90

% of cooling (DC-101, DC-102)

% of Technical Alc.flow

% of Potable Alc. flow

142,76 119,16 100,00 69,78 47,45

155,17 130,20 100,00 74,73 50,54

150 125 100 75 50

Table 3. Total steam and cooling need of distillation plant determined with simulation

Case

1 2 3 4 5

% of feed flow 150 125 100 75 50

% Total steam

149,06 124,53 100,00 73,33 49,31

Figure 4. Presentation of percentage of feeding flow, steam and cooling needs for every case compared with case No. 3

% Total cooling 145,60 121,75 100,00 70,38 47,75

% Total steam per produced Potable alcohol 99,37 99,62 100,00 97,78 98,62

% Total cooling per produced Potable alcohol 97,07 97,40 100,00 93,84 95,49

Figure 5. Comparation of total steam and total cooling per unit of produced potable alcohol

Figure 6. Funcion of total steam needs per product per case (better minimum determination)

Figure 7. Funcion of total cooling needs per product per case (better minimum determination)

Results shown on table 4 are expected and proportional to feeding flow. Otherwise, total results and especially total steam and total cooling per product give us good data for optimization. In comparation of all cases, case No. 4 needs minimum energy and cooling. That is optimal case of plant working. The values are close, so there are 3% lower steam needs and 6% lower cooling needs.

CONCLUSION Preparing of simulation model for some production plant could predict way of plant working. Theoretically could be predicted optimal needs of energy. The next step of this theoretically prediction is practically aprovement.

REFERENCES 1. Ruiz Mercedes, Ramos Isabel, Toro Miguel, “A Dynamic Integrated Framework for Software Process Improvement”; Software Quality Journal 10, 181-194, 2002 2. Nissen Mark E., “An Intelligent Tool for Process Redesign: Manufacturing SupplyChain Applications”; The International Journal of Flexible Manufacturing Systems, 12 (2000): 321 – 339 3. A.Anastasovski, L. Markovska, V. Meshko, Heat integration of ethanol and yeast manufacture, Macedonian Journal of Chemistry and Chemical Engineering, 26(2) 2007 4. Rašković P., Anastasovski A., Markovska Lj., Meško V. “Process integration of yeast fermentation plant” , ECOS 2008 – Krakow, Poland 5. Mahoney, Donald P., Fruehauf, Paul S., “An Integrated approach for distillation column control design using steady state and dynamic simulation” (2009) URL: http://www.aspentech.com/publication_files/An_Integrated_Approach_for_Distillatio n_Column_Control.pdf 6. Perry Robert H, Green Don W., Perry`s Chemical Engineers Handbook McGraw Hill Companies Inc. 1999

APENDIX 1 Table A1. Description of streams shown on figures 1, 2 and 3 Stream number 1 2 3 4 5 6

Stream description Distillery feed Feed of DC-100 Main steam distillery feed Steam for DC-100

Stream number 28 28` 29 29`

Silage DC-100 top outlet

30 30`

7 8 9 10 11 12 13 14 15 16 16` 17 18 19 20 21 22 23 24 25 26

HE-101 inlet DC-100 Reflux DC-101 feed DC-101 top outlet DC-101 reflux Technical alcohol part 1 DC-103 feed Steam for DC-101 Steam for DC-102 Technical alcohol part 2 (hot) Technical alcohol part 2 (cooled) DC-103 reflux Technical alcohol (total) Potable alcohol (hot) Potable alcohol (cooled) Luter Cooling water – distillery feed Cooling water HE- 106 inlet Cooling water HE- 104 inlet Cooling water HE- 103 inlet Cooling water HE- 101/102 inlet

31 32 32` 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

26`

Cooling water HE- 102 inlet

51

27

Cooling water HE- 106 outlet

Stream description Cooling water HE- 104 outlet Cooling water DC-102 outlet Cooling water HE- 103 outlet Cooling water HE- 103 outlet (simulation) Cooling water HE- 102 outlet Cooling water HE- 102 outlet (simulation) Total outlet of cooling water Fusel oil (hot) Fusel oil (cooled) Water for fusel oil dissolve Decanted Fusel oil Recycled stream – fusel oil free Cooling water HE- 105 inlet Cooling water HE- 105 outlet Condensate HE-100 Separated gas in HE-100 HE-101 outlet Condensate HE-101 Separated gas in HE-101 HE-102 outlet Gas exit Condensate HE-102 Mixed condensates Gas exit from mixer Total liquid condensated from DC-100 Solved fusel oils Separated fusel oil free stream – simulation Pressurized recycled stream simulation