Ecos 2008 - Process Integration Of Yeast Fermentation Plant Raskovic Anastasovski Markovska Mesko

ECOS 2008 - Poland

PROCESS INTEGRATION OF YEAST FERMENTATION PLANT Rašković P.1, Anastasovski A.2, Markovska Lj.2,Meško V.2 1

Faculty of Technology, University of Nis, Serbia, [email protected]

2

Faculty of Technology and Metallurgy, Ss Cyril and Methodius University, Skopje, Republic of Macedonia ABSTRACT: Process integration (PI) is a system oriented approach to process design of new or retrofitting of existing industrial plants, which applications are focused on resource conservation, pollution prevention and waste management. Some of research institutions reported that: “PI is probably the best approach that can be used to obtain significant energy and water savings as well as pollution reductions for different kind of industries”. The potential of PI exceeded the results obtained by traditional audits, and approaches based on separate optimization of individual process units. This paper provides a part of process integration study in the plant for yeast and alcohol production, situated in the town Bitola, Republic of Macedonia. Research roadmap is divided in four phases and the final solutions, based on three starting scenarios, are obtained by Pinch analysis. Evaluation of results revealed the parameters of Heat Exchanger Network with significance energy saving potential and short pay-back potential. Keywords: process integration, yeast fermentation plant, pinch technology.

NOMENCLATURE A -area (m2) specific heat (J/kgK) cp CC -capital cost ($/year) CO -operating cost ($/year) CT -total annual cost ($/year) CU -utility cost ($/year) d -diameter (m) Fl -fuel (heavy oil) G -mass flow rate (kg/s) h -specific enthalpy (J/kg) HEN(S)-heat exchanger network (synthesis) HRAT-min. temperature difference (0C) Hd -lower heating value (kJ/kg) m -mass flow rate (kg/s) Nu -Nusselt dimensionless number(-) Pr -Prandt dimensionless number(-) Re -Reynolds dimensionless number(-) Q -heat flow rate (W)

T -temperature (0C), (K) v -velocity (m/s) I={i} -streams Greek symbols α -heat transffer coefficient (W/m2K) λ -conductivity (W/mK) η -boiler efficiency μ -viscosity (Pa s) ρ -density (kg/m3) Subscripts a - water inlet in boiler C - cold g - gaseous f - liquid H - hot min- minimum S - steam outlet from boiler

1. INTRODUCTION In this paper a part Process Integration (PI) study of yeast fermentation plant is presented. The primal goal of the study is to validate the improvements made by energy audits, and to identify and estimate new opportunities to reduce the energy and raw materials consumption in the plant.

based on the fact that PI task involved only the material and energy flows between the main production subsystems. Research roadmap is divided in four phases. In the first phase, the physical model of yeast fermentation plant is described and defined as the network of inter-connected subsystems. Mathematical model of the plant was based on mass end energy balance equations of subsystems. In the next phase, numerical model of plant is implemented in the spreadsheet software tool named BtY [2] developed on Microsoft Excel programming platform. Thermodynamic properties of some streams are calculated by the use of water/steam properties simulator. For the phase of PI (Pinch analysis) three scenario are defined. Data extraction task is improved with spreadsheet software which enabled the calculation of thermodynamic and economic parameters for HENS purpose. Pinch analysis is realized by NiPinch[3] software, and limited on the results of targeting task. The simplified roadmap is presented in Fig. 1. 2. YEAST AND ALC. PRODUCTION

Figure 1: The research

simplified

roadmap

of

Generally, yeast manufacture can be classified as semi-continuous processes, remarking that during some operating period the processes have continuous character. In order to create the feasible and practical solution with short-payback period, PI task is addressed to continuous operations, during that period. Another assumption is

The word "yeast" comes from the Sanskrit 'yas' meaning "to seethe or boil". Today the name yeast is applied specifically to a certain group of microscopic fungi and to commercial products consisting of masses of dried yeast cells or of yeast with a starchy material and pressed into yeast cakes. Louis Pasteur, in the 1850's, first discovered and described the fermentation process, which led to the development and cultivation of the yeast we use today. Yeast belong to the group of unicellular fungi, a few species of which are commonly used to leaven bread and ferment alcoholic beverages. The most number of yeasts belong to the division Ascomycota, but the well-known and commercially significant yeasts are Saccharomyces cerevisiae. These organisms is commonly used as baker's yeast and for some types of fermentation.

Yeast is often taken as a vitamin supplement because it content proteins, B vitamins, niacin, and folic acid. The process of fermentation in yeast production can be classified according to the presence of air either on Aerobic (with air) or Anaerobic (without air). Aerobic fermentation represents yeast biomass growth using energy from biochemical oxidation of glucose via energetic component as ATP (Adenosine Tri Phosphate). Final product of Aerobic fermentation commonly exist in two forms: compressed cakes (also called fresh yeast) and dehydrated granules (dry yeast). Fresh yeast (contain about 70% moisture is ivory coloured with a yellowish hue, soft, moist and easily crumbled. Dry yeast is fresh compressed yeast that has been dried until the moisture content of about 8%. The granules of dry yeast become active again only by mixing with a warm liquid. The main advantage of dry yeast longer shelf life and easy storage. The process of anaerobic fermentation (known as Alcoholic fermentation) has slower activity, focused on converting sugar to alcohol rather that increasing the number of yeast cells. Alcoholic fermentation make transformation of glucose to ethyl alcohol with byproducts obtained by yeast cells metabolism. Biochemical oxido-reduction reaction of this process is known as Embden-Meyerhof-Parnas pathway, or Glycolisys. Final product of anaerobic fermentation is Ethyl alcohol, also known as Ethanol, a flammable, colorless chemical compound that is used for beverage or industrial purposes. It is usually distributed in the form of pure ethyl alcohol, completely denatured or specifically denatured ethanol. 3. PLANT DESCRIPTION The reference plant for process integration research in this paper, is one of the Balkan’s oldest and biggest factory for yeast and alcohol production named AD” F-

ka za Kvasec i Alkohol–Bitola”. The factory is placed in the city of Bitola, southwest region of Republic of Macedonia, close to Macedonian – Greece border. The main product of the manufacture are: • refined and denaturized ethyl alcohol, • dry and fresh baker`s yeast, • wine and selenium yeast, • baker`s additives, bee food,.... The average annual production of these product are: 1200 t of dry yeast, 6000 t of fresh yeast and approx. 5x106 m3 of ethyl alcohol (as refined ethanol-potable alcohol). The physical model of the plant (Fig. 1) can be presented as the network of interconnected subsystems according to the technology operations of manufacture roadmap. Production processes are framed inside two main routes: • First baker`s yeast route, which comprises: o Preparation of feeding material (subsystem A), o Aerobic fermentation (subsystem B), o Centrifugal separation (subsystem E), o Filtration (subsystem F), o Drying (subsystem G), and o Final yeast storage and packaging stage. • The second alcohol route comprises: o Preparation of feeding material (subsystem A), o Anaerobic fermentation (subsist. H), o Separation of alcoholic wort and yeast cream (subsystem C), o Distillation of alcoholic wort (subsystem D) and o Final, alcohol storage stage. Production process (for both products) start in the subsystem A, where raw molasses (RML) (stored in the raw molasses tank) is treated in order to obtained the prepared molasses (PML). RML contain 45– 50% of sucrose and other components like: glucose, fructose and some vitamins(such as biotin). RML are first mixed with hot water(HW1), in order to create the molasses

solution (MS). Next, MS is leading in the first centrifugation stage, where solids and voluminous sediments (discharged streams CSD) are removed. One part of solids coagulates due to sulphuric acid presence. After centrifugation, clarified MS are sterilized by law pressure steam (LP1) and then lead to the second separation stage phase separator. Phase separation enabled the remove of volatile organic acids (lactic acid, butyric acid, sulphuric dioxide etc. which inhibit the process of yeast fermentation) from MS stream. Final product from this stage, PML is first cooled in plate heat exchanger (by cold water ,CW1) and stored in Prepared molasses tank. In the next stage of manufacture – Fermentation stage (subsystem B), production process is divided on two separate routes: yeast production route and alcohol route. In the case of yeast production, PML are transferred in the aerobic fermentation unit. Fermentation process is improved by the use of hot water (HW2). Another input in subsystem are: • feeding salts (FS), • cold water (CW2) stream-needful for vessels surface cleaning, • water from regional system (RWS)-used for control of fermentation temperature • air stream (AS)-used as source of oxigen for yeast biomass growth. Final product of fermentation process is yeast wort (YW). In the next phase YW is purified crossing the three stage centrifuge in subsystem E. After separating process, yeast cream (YC) is washed by cold water (CW3) and then stored in Yeast cream storage tank. Impurities obtained as the result of treatment are sending into drainage systems (ImS). Hot water stream (HW3) is used for washing of centrifugation system. The first form of final product -Fresh yeast (yeast with content of 30 – 34% dry matter) is produced in subsystem F, through the filtration process and by the use of

rotation vacuum filter. A part of fresh yeast (FY), obtained in this stage, is transferred in the storage stage, while the rest is sent to subsystems G, in additional treatment. Filtrate (FL), obtained by water washing treatment, is discharged. Subsystem G (fluid bed dryer process unit) is supplied with LP steam (LP2), hot water (HW5) and dry air (DA). Final product from this subsystem, Dry yeast (DY), with content of 95 – 97% matter, is transported to the storage and packaging stage. Outgoing stream (HA), from subsystem E, represent the hot humid air. In the case of alcohol production route PML is first dosed in anaerobic fermentation unit. The process of anaerobic fermentation is improved by the presence of hot (HW6) and cold water (CW6). Alcohol yeast cream (AYC), as starter culture for that feed batch fermentation, is added from Alcohol yeast cream tank. One of the products of fermentation are exhaust gases (EG), mixture of carbon dioxide, vapour and ethanol. The product of anaerobic fermentation process, is alcoholic fermented media (AFM), with content of approx. 8-10 %Vol. of ethanol. In the next stage, subsystem C, AFM is separated into two substreams: alcoholic wort (AF) and alcoholic yeast cream (AYC), stream. Alcoholic wort (AF) is then feeding in steam distillation unit, subsystem D. The results of distillation are: 96% Vol. ethyl alcohol (EA) and technical alcohol (TA) stored in STE and STA tanks. Technical alcohol present a mixture of ethanol, aldehydes (represented with acetaldehyde) and some other less volatile components than ethanol. The distillation waste streams are: • molasses residues (MR), • organic acids water solution known as Luter (LT), and • cooling water from shall and tube heat exchangers (WHE).

Figure 2: Physical model of yeast and alcohol production plant

4. PROCESSES INTEGRATION TASK The PI task is accomplished by the use of thermodynamic method – Pinch analysis[4]. The result of research presented in this paper are limited on targeting task and realized on three different scenarios. In all scenarios only the streams with considerable heat capacity are involved (Table I). Scenario 1 and Scenario 2 comprised only

the liquid streams, as Scenario 3 had one more cold stream-air stream (DA) utilized in drying subsystem. In the case of Scenario 2 and Scenario 3 the supply temperatures of hot streams (CWS and WHS) are obtained after preheating in HW and Water Heating systems. In the case of Scenario 1 supply temperatures of this streams are the temperatures before entering these units.

Table I: Process integration task Stream

Scenario

Temp (0C) Tsup Ttar

mcp (kW/0C)

ΔQ (kW)

(kW/m2K)

CU (utility cost) ($/GJ)

α

Cold streams CWS

WHS DA

Scenario 1 Scenario 2 Scenario 3 Scenario 1 Scenario 2 Scenario 3 Scenario 3

5.0 40.0 40.0 20 40 40 40.0

90.0 90.0 90.0 60 60 60 90.0

5.815 5.815 5.815 9.304 9.304 9.304 3.154

494.3 290.8 290.8 372.2 186.1 186.1 157.7

1.505 1.505 1.505 1.505 1.505 1.505 0.0101

19.36 19.36 19.36 19.36 19.36 19.36 19.36

Scenario 1 Scenario 2 Scenario 3 Scenario 1 Scenario 2 Scenario 3 Scenario 1 Scenario 2 Scenario 3

106.0 106.0 106.0 105.0 105.0 105.0 70.0 70.0 70.0

25.0 40 40 40 40 40 43.3 43.3 43.3

2.874 2.874 2.874 0.887 0.8866 0.8866 13.93 13.93 13.96

232.8 189.7 189.7 57.63 57.63 57.63 371.9 371.9 371.9

1.507 1.507 1.507 1.447 1.447 1.447 1.843 1.843 1.843

3.89 3.89 3.89 3.89 3.89 3.89 3.89 3.89 3.89

Cold Hot

5 174

10 173

1.505 3.522

3.89 19.36

Hot streams MR

LT

WHE

Utility CW LP

Heat transfer coefficient (α) and Utility cost (CU) are obtained from spreadsheet software. Mathematical expressions of these calculation are presented in Appendix.

operating cost and capital cost of HEN solution (in this case ΔTmin=20 (oC) and operating period is 4000 (h/year) for all scenarios). Table II: Sensitivity analysis parameters

5. RESULT AND FUTURE WORK Evaluation of targeting results are presented in the form of sensitivity analysis Short review of parameters and research ranges are presented in Table II. The first evaluation criteria was the potential for HEN solution wich need shortpayback period. This kind of potential is presented as ratio between annual saving of

Input parametars

Range

Operating period /year

2000÷4000 (hour)

HRAT (ΔTmin)

12÷28(oC)

Output parametars Qmin Amin Nmin CC CO CT

(kW) (m2) (-) ($/year) ($/year) ($/year)

Graphical presentation of this potential is expressed by slope angle of scenario curve in Fig 3. It is obvious that solutions obtained by Scenario 1 and Scenario 2, are more appropriate than solution obtained by Scenario 3, according to the duration of payback period.

1 and Scenario 2 that contribution is considerably less than in Scenario 3; logical conclusion: these scenarios are more suitable for previously defined goal: quickwin solution.

Figure 3: Pay-back potential

Figure 4: Percentage contribution of HEN capital cost in Total annual cost Explanation of such results can be found in sensitivity analysis of HEN cost inside ΔTmin operating range (operating period is 4000 (h/year)). Analysis expressed the percentage contribution of HEN capital cost in Total annual cost. In the case of Scenario

Figure 5: Total cost analysis in relation with operating period during the year

Sensitivity analysis inside ΔTmin operating range exposed intersection point (ΔTmin≈ 16 (oC)) between Scenario 1 an Scenario 2 curve. This exposure was the initial for total cost analysis, but this time in relation with operating period during the year. Analysis (presented in Fig. 5) involved only Scenario 1 and Scenario 2, and the results are obtained for 3 values of HRAT (ΔTmin=12, 16 and 28 (oC)). Graphical presentations demonstrate percent increase of total annual cost in relation with working time period of HEN. Total cost analysis indicate the existence of two regions: • First – region which covered the percentage increase of CT inside the HRAT range 12-16(oC), • Second – region which covered the percentage increase of CT inside the HRAT range 16-28(oC), According to the position and slope of the curves, one general conclusion can be made: inside the first region HEN solution of Scenario 2 is preferred than solution of Scenario 1. The conclusion inside second region is quite opposite. The reason for that is threshold problem position of composite curve in the case of Scenario 1 in range ΔTmin=12÷-20 (oC). By rising the value of HRAT, operating cost increase but capital cost do not decrease till the presence of threshold position. So despite the preferences presented in diagram of Fig.1, final conclusion is the privilege of Scenario 2 in the case of lower value of HRAT. However, Scenario 1 confirm the better payback potential in the case of higher HRAT values. As the conclusion we can say that this part of PI study clarified the potentials for energy savings inside the yeast and alcohol manufacture plant and establish the correct direction of future research. Results obtained in targeting phase will be the starting position for HEN design and optimization in the following papers.

REFERENCES [1] Rašković, P., Step-by-Step Process Integration Method for the Improvements and Optimization of Sodium Tripolyphosphate Process Plant. Energy 2007;32(6):983-998. [2] A. Anastasovski, L.Markovska, V. Meshko, Heat integration of ethanol and yeast manufacture, Macedonian Journal of Chemistry and Chemical Engineering, 26(2) ; (2007):135-146 [3] Rašković, P., Ni Pinch-software tool for heat exchanger network synthesis, In Proceedings of the Computational engineering in fluid dynamic and energy technology, I professional seminar, Niš, Serbia, 2004 p.99-708.. [4] Smith R. Chemical Process Design, New York: McGraw-Hill, 1995. APPENDIX Heat transfer coefficient-α: α = f (Nu = 0.023 ⋅ Re 0.8 ⋅ Pr 0.33 , i

λ ,c i

i , ρ , μ , v , d ),i ∈ ( CW ,CWS ,WHS , DA ) p,i i i i

α i = f (Nui = 0.026 ⋅ Re0.6 ⋅ Pr 0.3 , λi , c p,i , ρi , μi , vi , d ), i ∈ ( MR , LT ,WHE )

[ (

) ]

⎧ Nu = 0.021 ⋅ Re 0.8 ⋅ Pr 0.43 1 + ρ /ρ 0.5 / 2 ,⎫ ⎪ i ⎪ f g αi = f ⎨ ⎬, ⎪⎩λi , c p,i , ρ i , μ i , vi , d ⎪⎭ i ∈ ( LP )

Utility cost-CU: CU H = C Fl ⋅

(

)

1 ⋅ hS − ha ⋅ 10 6 [$ / GJ ], η ⋅ Hd

CU C − obtained from literature [4]