Process Integration : Application in the food industry · PDF fileProcess Integration : Application in the food industry h ... –Heat integration –Energy conversion integration

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Prof. François Marechal

Industrial Process and Energy Systems EngineeringEcole Polytechnique Fédérale de Lausanne

Switzerland

Process Integration : Application in the food industry

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Goals

• What is the benefit of applying process integration techniques in the food industry ?– Heat integration– Energy conversion integration– Water integration– Waste integration

• How do we have to adapt the methodology ?– Data collection– Combined Water, Heat & Power – Restricted matches– Multi-period problems

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The food processing system analysis

Energy Conversion systemHeatElectricityCoolingRefrigeration

Production processProducts from raw materialsGeneric process unitsUses distributed energyProduces waste

Waste managementNew products/servicesEnergy ResourcesRecycling

massenergy

Production supportWater

CIPPackqging

Storage

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Example : Brewing process

LENI Systems

Energy perspective Energy bill (electricity, gas and water)Waste handlingMaintenance and operation costEnvironmental regulation

Boiler

RefrigerationWashing

Hot section Cold section

Natural gas (steam boiler) 3133 kW

Steam 2819.7 kW

Cooling water 1578.7 kW

Refrigeration 465 kWe

Steam vented 455 kW

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The cogeneration temptation

• Replace the boiler (2800 kW) by an engine– GE engine size 2*1250 kWe

• eff el = 41%• eff th = 46 %

• Recover the vented steam• Export electricity

present cogeneration

Natural gas kW 3133 6130

Steam kW 2820

Cooling water kW 2220 2220

Electricity kWe 465 -2048

Steam vented kW 455 0

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

What about process integration

• Defining the process requirement– Distributed Energy– Water

• Defining the heat recovery target• Integrating the energy conversion• Integrate the waste management• Synthesize the energy saving process

configurations• Evaluate and compare the options

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The present system

Raw materialsProducts& by-products

Heat losses

Food or agro process

WasteFossil resources

ABC

ABC

CO2 Exergy

Key performance indicators

Conversion

CIPPackaging

ConditioningProcessing

Coo

ling

& r

efri

gera

tion

Hea

ting

Heat

Electricity

ABC

Costs

Biomass

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The integrated process

Products and by-products

Heat losses

Waste

Raw materials

ElectricityHeat recovery

Heat pumps and refrigeration

CogenerationConversion

Waste management

Waste

Fossil resources

Biomass + solar

ABC

ABC

CO2 Exergy

ABC

Costs

Industrial food and agro symbiosis system

Key performance indicators

CIPPackaging

ConditioningProcessing

Coo

ling

& r

efri

gera

tion

Hea

ting

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Analysing the process requirement

Malt Water

Mashing

Masche

Filtration

Water

Wort

Cooking

Hop

Cooling

Fermentation

Chilling Pasteurization et Packaging

Beer

Wort

WortYeast

Steam

Cleaning in Place

Water

Husk

Water

System boundaries

? Bio methane ?

? Recover ?

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

• A top-down approach– What is the function of energy in the process– What are the most important units ?

• 80/20 Rule (80 % of consumption explained in 20% of time)• Apply on distributed energy

– Characterize the units that are the most important consumers/producers

Identify important unit operations

Muller, D, and F Marechal. “Energy Management Method in the Food Industry.” In Handbook of Water and Energy Management in Food Processing, Woodhead Publishing Ltd, 2008.

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Defining the process requirements

• Target the energy bill– Time average approach

• specific consumptions of the operations– Multi-period and storage solved in 2nd stage– Water is the process support

• intermediate stream for heat transfer• Define the process requirements

– Generic models (drying, evaporation, etc...)• modules

– Process integration models– combined water and heat requirement

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Process unit models

• Analysis of the heat transfer requirement• Generic modules

– Data base– Web based

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Maximum heat recovery in the system

• Define the heat recovery potential : – 2700 kW out of 4000 kW

!"#$%&'()*+,-&.%/+0%

Estimated !

Utility MER

[kW]

Current

[kW]

Hot utility 1386 2220

Cold utility - 16

Refrigeration utility 837 1200

Heat recovery leads to 37 % energy savings

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Energy conversion needs

• Heat-temperature profile for the utility system

2nd E uropean Conference on Polygenerat ion - 30th M arch-1st A pril, 2011 - Tarragona, Spain

In the following example, we are discussing the integrat ion of a trigenerat ion energy con-version system in a brewing process.

2 Process integration and trigeneration

T he fi rst step of the methodology is the defi nit ion of the energy requirement . In an industrialprocess, the energy requirement is defi ned by the set of streams to be heated up and cooled down.T he defi nit ion of the requirement is obtained from a process model in which the process unitsare calculated in order to defi ne the hot and cold streams enthalpy-temperature profi les. T hedetails of the analysis are presented in [13], the focuss being here to comment on the integrat ionof the trigenerat ion system. T his analysis results in the defi nit ion of the hot and cold compositecurve of the process ( F igure 1) that allows one to calculate the possible heat recovery by heatexchange between process streams. R esult ing from the heat balance of the process requirement ,the hot and cold composite defi ne also the heat ing and cooling requirement of the process. T hecalculat ion of the G rand composite curve ( F igure 2) defi nes the enthalpy-temperature profi leof the heat ing, cooling and refrigerat ion requirement . R esult ing from the pinch analysis, theheat recovery potent ial corresponds to 114 3 k W i.e. 4 5 % of the actual consumpt ion. T his alsocorresponds to more or less doubling the present heat exchange recovery.

!"#$%&'()*+,-&.%/+0%

F igure 1: H ot and cold composite curves of the process

F igure 2: G rand composite curve of the process

T he analysis of the energy requirement leads to the following conclusion

2

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Energy conversion needs analysis

• What are the options ?

2nd E uropean Conference on Polygenerat ion - 30th M arch-1st A pril, 2011 - Tarragona, Spain

In the following example, we are discussing the integrat ion of a trigenerat ion energy con-version system in a brewing process.

2 Process integration and trigeneration

T he fi rst step of the methodology is the defi nit ion of the energy requirement . In an industrialprocess, the energy requirement is defi ned by the set of streams to be heated up and cooled down.T he defi nit ion of the requirement is obtained from a process model in which the process unitsare calculated in order to defi ne the hot and cold streams enthalpy-temperature profi les. T hedetails of the analysis are presented in [13], the focuss being here to comment on the integrat ionof the trigenerat ion system. T his analysis results in the defi nit ion of the hot and cold compositecurve of the process ( F igure 1) that allows one to calculate the possible heat recovery by heatexchange between process streams. R esult ing from the heat balance of the process requirement ,the hot and cold composite defi ne also the heat ing and cooling requirement of the process. T hecalculat ion of the G rand composite curve ( F igure 2) defi nes the enthalpy-temperature profi leof the heat ing, cooling and refrigerat ion requirement . R esult ing from the pinch analysis, theheat recovery potent ial corresponds to 114 3 k W i.e. 4 5 % of the actual consumpt ion. T his alsocorresponds to more or less doubling the present heat exchange recovery.

!"#$%&'()*+,-&.%/+0%

F igure 1: H ot and cold composite curves of the process

F igure 2: G rand composite curve of the process

T he analysis of the energy requirement leads to the following conclusion

2

Cogeneration with enginecheck compatibility of temperature for cooling water

Mechanical vapor recompression from steam recovery ?

Pinch analysis says NO !

It is not needed to refrigerate at the lowest temperature

multiple levels

What about heat pumping ?with refrigeration cyclePinch analysis says YES

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Energy conversion unit models

Ecole Nationale des Ponts et Chaussées – Projet de fin d’Etudes

Monika Dumbliauskaite – Département Génie Civil et Construction 62

273 10 273 10 273 10 273 101118 550 160

373.5 432 337.5 373.5tL kW kW kW+ + + +

! " # + " # !

The use of steam at 123°C to supply heat to the process generates therefore around

170kW of exergetic losses. This corresponds to 160kW of mechanical work which could be

generated through the use of reversible Rankine cycles operating between T*steam and T*

process.

Therefore, it is necessary to reduce as much as possible the temperature difference between

the process and the utilities in order to lessen the exergy losses resulting from the heat transfer

between the utility streams and the corresponding process streams.

Solutions allowing the improvement of the present configuration of the utilities are

studied in the following paragraphs.

3.1.4 Integration of a Cogeneration Engine

The integration of a cogeneration engine is a sustainable solution known to reduce the

operating costs, as the combined heat and power system produces both mechanical power and

heat by taking advantage of fuel combustion.

A reciprocating engine fed with natural gas is considered in this study (see Figure 38).

It appears to be the most relevant technology, as it is possible to recover heat from both

exhaust gases and cooling water, which can be used in low temperature processes like

breweries.

Figure 38: Cogeneration Installation (Internal Combustion Engine) Source: Model GE-Jenbacher type 3, www.gejenbacher.com



Specification Symbol Value Unit Fuel Natural Gas Nominal speed N 1500 min-1 Effective power Pe 1063 kWe Mechanical efficiency mech 0.408 - Thermal efficiency therm 0.456 - Exhaust gas temperature (default value) Tgas,out 470.5 °C Stack temperature (default value) Tstack 120 °C Cooling water inlet temperature (default value) Twat,in 87.0 °C Cooling water outlet temperature (default value) Twat,out 79.9 °C Exhaust gas heat gases,thQ 537 kW

Cooling water heat water,thQ 653 kW Fuel cost cfuel 0.01961 !/s

Table 28: Implemented Specifications of the Cogeneration Engine

For engine sizes close to 1000kW, a linear approximation of the heat loads and

mechanical power delivered by the engine can be accepted, based on the product described in

Table 26.

The computation was performed using the same hypotheses as in the previous case for

the estimation of costs and emissions (see Table 17 and Table 19). The maintenance fees are

not taken into account in the expression of the operating costs resulting from the purchase of a

new utility. This is due to the fact that the increase in maintenance fees compared with the

current setup can not be evaluated, as it would imply the complete characterisation of the

current installation and the associated maintenance costs. This criterion will not be taken into

account, since it would unfairly penalise the purchase of new installations.

As no information was provided concerning the electrical consumption of the different

production units, the electricity produced can either be sold or directly used on site. It is

assumed that in both cases, the production of 1kWhe corresponds to a saving of 0.0541 ! (see

Table 17).

The integration of the cogeneration engine described previously leads to the results

presented in Figure 40.



Figure 32: NH3 Refrigeration Cycle with Two Evaporation Levels (Belsim-Vali® model)

Figure 33: Example of a (h,log(P)) Diagram for a Two-Level Evaporation Refrigeration Cycle Where e1 and e2 [kJ/kg] are the specific compressor works

The advantage of this installation over single-stage refrigeration cycle is the saving in

mechanical consumption. A comparison between both solutions is shown in Table 22.



Refrigeration cycle Single-level of evaporation

Two-levels of evaporation

Ammonia mass flow [kg/s] 0.2 0.1/0.1 Evaporation temperature [°C] -8 -4/-8 Condensation temperature [°C] 30 30 Total cooling load [kW] 225.83 223.85 Compressor power [kW] 52.71 49.78 Energetic efficiency (COP) 4.28 4.65 +9% Exergetic efficiency (Tamb=25°C) 0.53 0.54 +1%

Table 22: Comparison between Single and Two-Levels of Evaporation NH3 Refrigeration Cycles

3.1.3.2 Unit Costs

The evaluation of the operating costs by Easy2 requires the determination of the unit

cost of the utilities [!/s].

Utility Reference flow [kg/s] Heat load [kW] Cost [!/s] Steam 1 2297.2 0.02034

Cooling water 1 4.19 0.00000657 Table 23: Unit Costs of Steam and Cooling Water

The detailed calculation of the values presented in Table 23 can be found in Appendix I.

3.1.3.3 Results for the Current Utility Setup

Using the assumptions formulated previously, the integration of the existing utilities

was performed so as to fulfil the MER.

The streams of the energy conversion technologies are added to the process hot and

cold streams. The resulting composite curves are called “integrated composite curves”, as

they take into account the utility streams. When the utilities are well integrated, there is no

additional energy requirement.

In Figure 34 and Figure 35 are shown the integrated composite curves of the steam

cycle and the refrigeration unit defined previously.

GE engine type 3

Module : cogeneration engine

Module : refrigeration cycle

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Energy conversion system integration

• Utility system made of a list of optional sub-systems “w”– Mechanical vapor recompression– Steam boiler– Cogeneration engine– Refrigeration cycle (multi levels)– Cooling water

• For each subsystem “w”– Calculate hot and cold streams

• qw,r : contribution of a stream to the heat cascade interval r if the stream is used– Calculate power consumption/production

• ew : electricity – Calculate fuel consumption => operating cost C2w

– Investment cost : piecewize linearized function : CI1w,CI2w

• Unknowns are :– is the sub-system “w” used : integer variable yw ={0,1}– flow in utility sub-system w : continuous variable fw : fminw ≤ fw ≤ fmaxw

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

MILP formulation

minRr,yw,fw,E+,E!

(nw!

w=1

C2wfw + Cel+E+! Cel!E!) " t

+nw!

w=1

C1wyw +1

!(

nw!

w=1

(CI1wyw + CI2wfw))

nw!

w=1

fwqw,r +

ns!

s=1

Qs,r + Rr+1 ! Rr = 0 "r = 1, ..., nr

Rr ! 0 "r = 1, ..., nr; Rnr+1= 0;R1 = 0

nw!

w=1

fwew + E+ ! Ec " 0

nw!

w=1

fwew + E+! Ec ! E!

= 0

fminwyw ! fw ! fmaxwyw yw ! {0, 1}

E+ ! 0;E! ! 0

Subject to : Heat cascade constraints

Electricity consumption Electricity production

Feasibility

Energy conversion Technology selection

Operating cost

Fixed maintenance

Investment

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Calculation with the boiler and refrigeration

2nd European Conference on Polygeneration - 30th March-1st April, 2011 - Tarragona, Spain

The present system

In the present system, a boiler fed with natural gas is currently generating steam at highpressure (8.5bar) that is distributed to the process at a lower pressure (2.2bar/123.3°C). Therefrigeration requirement is satisfied by a NH3-refrigeration cycle with two evaporation levels,at -4°C and -8°C. The cooling requirement is satisfied by cooling water. The integration resultsare presented on figure 3 using the integrated composite curve of the utility system. The utilitystreams are represented by the line “brewery_utility” and the process requirements correspondto the grand composite curve “Others”. The mechanical work supplied to compressors (heatpump and refrigeration cycle) is represented by the line “Mech. Power”. It can be observedthat this situation in addition of realizing the process heat recovery already realizes heat re-covery form the refrigeration system and therefore corresponds to an attractive energy saving.However, the major part of the refrigeration cycle hot streams is removed by the cooling waterand evacuated to the environment. In the integrated solution, the refrigeration cycles consumes184 kWe. This corresponds to a reduction of 225 kWe (56%) of the present mechanical powerconsumption of the refrigeration cycle. This is mainly explained by the fact that in the presentsituation, the refrigeration cycle is used in penalizing heat exchangers that use the refrigerationcycle to cool down stream above the cooling water temperature. Reaching the minimum cycleconsumption requires therefore to identify the penalizing heat exchangers through the coolingwater temperature.

Figure 3: Current Utility Setup: Boiler & Refrigeration Cycle (RC)

Improvement of the utility system integration

The analysis of Figure 3 reveals that the current utility configuration could be improved byreplacing the high temperature steam used in the process by a cogeneration unit that couldsupply heat at lower temperature while producing electricity. The analysis of the refrigerationcycle integration suggests that the temperatures at which the heat is removed could be opti-mized by better staging the refrigeration requirement and by increasing the temperature of the-4 °C level to be closer from the 5 °C temperature of the requirement. In addition, applying therules for the proper integration of heat pumps, it can be suggested to increase the condensationtemperature of the refrigeration cycles in order to create a heat pumping e�ect. As the COPof the refrigeration cycle depends on the compression ratio and therefore of the temperaturelift in the cycle, several condensation levels will be assessed. For each combination of conden-sation/evaporation levels, the NH3 cycle is calculated and a collection of cycles is added in theutility sub-systems list.

4

Total&fuel&consump/on

2088&kW&

533%U/lity&Electricity 184$kWe

Cooling&Water 17.1&kg/s

• Integrated curves representation

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The Carnot composite curves and the conversion system

The area between the 2 curves is the exergy destruction in the heat exchange system

(1-T0/T)

Room for improvement

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Integration of the cogeneration engine

• Natural gas CHP system integration• Increased fuel

consumption, but…

• Mechanical power generation

• Heat pumping to lessen heat losses ?

8

Total&fuel&consump/on

3279&kW

+57%

U/lity&Electricity ($863$kWe


francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Heat Pump Integration

• Closed cycle heat pump• Refrigeration cycle condensation heat

pumped to higher temperature levels

• Mechanical vapour recompression• Evaporating wort is directly compressed

and delivers heat to the process by condensing

• Process stream → Utility stream

• Cost optimisation design variables• Operating temperatures

9




fulfil part of the heating demand at high temperatures and the remaining heating requirements

(at lower temperatures) would be satisfied using the cooling water of the cogeneration engine.

Thus, the size of the MVR system is determined in conjunction with the heat provided by the

cogeneration engine to fulfil the process heating requirements.

!!!! A heat pump could offer the opportunity to make the condensation heat of the

refrigeration cycle available at higher temperature levels, so as to satisfy the heating

requirement of the bottle pasteurisation device (above 62°C). In this case, the most suitable

refrigerant is water.

!!!! The configuration of the refrigeration cycle can be improved by adapting the

temperature of the higher pressure level so as to minimise the compressor power and the

exergy losses.

3.1.5.1 MVR Model

As mentioned previously, an MVR system can be used to compress the steam

emanating from the boiler and thereby increase its temperature. The augmentation in

temperature of the steam must be sufficiently high so as to compensate the minimum

approach temperature between the evaporating wort and the condensing steam.

Figure 41: Illustration of the MVR System

MVR

E+

Q- Q+

(T1,P1)

(T1+!T,P1+!P)

Steam

Liquid

Boiler

wort

To heat exchange with process

Utility steam

Figure 5: Mechanical vapor on the wort evaporation

refrigeration cycle for process preheating.

As the heat of the vapor condensation is a priori also useful to satisfy process needs (selfsu⇡cient zone on figure 2), only the useful part of the MVR has to be calculated. This is doneby introducing a decision variable that represents the amount of recompressed vapor.

The choice of the heat pump operating conditions defines the temperature at which the heatwill be made available and therefore the amount of heat that will be useful for the process. Asa function of the selected level, the other utility flows will be updated by optimization.

Two heat pumps with an evaporation at 6°C(299K) and with respectively 66.5°C(340K) and77.5°C(351K) condensation temperature are proposed and compared.

The second refrigeration cycle produces cold at -6 °C(299K) and optional condensing tem-peratures at 45 °C(318K) and 50 °C(323K) are considered.

The results of the optimised configurations, including the integration of MVR and heatpumping systems, are presented in Figures 6 and 7.

It can be seen a clear reduction of exergy losses: utility temperatures are as close as pos-sible to the temperatures of the process energy requirements. One can also observe a drasticreduction in utility losses: for the case where the heat pump condenses at 77.5 °C (351K) :the external cooling water requirement is close to zero, indicating that the overall refrigerationheat is used as a source for satisfying the process heat.

Table 3 presents the results of the di⇠erent utility integration solutions. The economicalperformances are calculated considering the value of energy and the CO2 emissions for theelectricity data given on table 2.

Combined with the heat recovery, the advanced trigeneration system o⇠ers an energy savingof up to 60 %, while reducing the electricity import by the same amount. It is important torealize that the optimal solutions depends on the equivalent CO2 content of the electricity mix.In a country like Germany with heavy loaded electricity, the solutions with cogeneration only

6

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Heat pump conditions optimisation

• 2 heat pumping conditionsHP1 set up 1 (Tcond=340K)

Fuel 1677&kW

CHP (374$kWe

«&Heat&Pumps&» 295$kWe


Fuel 1140&kW

CHP (166$kWe

«&Heat&Pumps&» 379$kWe


11

Engine

HP 2 set up (Tcond=351K)

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Results (Maximum Heat Recovery)

1. Gas Boiler 2.Gas CHP 3.Gas CHP+MVR+HP (Tcond=66.5°C) 4.Gas CHP+MVR+HP (Tcond=77.5°C)

•

Unit 1. 2. 3. 4.

Natural Gas kW 2088 3279 1677 1140Electricity kWe 184 -863 -80 212

Water kg/s 17.1 17.1 3.2 0.2

Run. Costs FR k€/yr 332 210 205 212Run. Costs GER k€/yr 520 283 312 336

TOTAL Costs FR k€/yr 332 308 274 274

TOTAL Costs GER k€/yr 520 380 381 398

TOTAL CO2 FR* ton/yr 2459 3544 1912 1372

TOTAL CO2 GER* ton/yr 2987 1094 1686 1976

12

� � � � � ��

� � � ��

� � 2 � � ��

� �� e � � � CO2� � � � e

� � �� LHV � � � � CO2� � � � LHV

� � �� 3 �� e � � � � CO2� � � � e

� � �� LHV � � � � CO2� � � � LHV

� � � �� CO2 � � ��

� � � � ��

� � � �� cond� � � �� °�� cond� � � �� °��

�

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Bio-Methanation Opportunity

• Opportunity in breweries: organic waste (husk) bio-methanation– 75 Nm3 CH4/t husk

• However…– Extra investment (digester), increased electric

consumptions (blender, pumps)– Heating requirement (Cold stream @ 35 °C)

• Available : 1660 kW as LHV of CH4

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

MER & Bio-Methanation: Results

• Highly profitable: investment compensated by a drastic reduction in running costs

• France/Germany: similar conclusions

Unit 1. 2. 3. 4.

Biogas kW 1660 1660 1660 1660Natural Gas kW 664 (2088) 711 (3279) 480 (1677) 200 (1140)

Electricity kWe 264 (184) -924 (-863) -298 (-80) -219 (212)

Water kg/s 17.1 17.1 3.2 0.2

Run. Costs FR k€/yr 161 (332) -31 (210) -16 (205) -32 (212)

Run. Costs GER k€/yr 260 (520) -280 (283) -38 (312) -60 (336)

TOTAL Costs FR k€/yr 238 (332) 145 (308) 124 (274) 115 (274)

TOTAL Costs GER k€/yr 338 (520) -105 (380) 101 (381) 88 (398)

TOTAL CO2 FR* ton/yr 839 (2459) 566 (3544) 471 (1912) 170 (1372)

TOTAL CO2 GER* ton/yr 1588 (2987) -2060 (1094) -377 (1686) -452 (1976)

14

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

The tools needed

• Identification of optimal integrated heat pump(s) for a given process– Appropriate fluids – Operating conditions – Temperature levels (discontinuous optimization problem) – Size of installations and economic evaluation

• Realistic solutions: Heat pump data base– Compressor types --> Operating condition ranges (volumetric flow rate, pressure

ratio)– Refrigerants --> Operating condition ranges (temperature

levels) • Multi-stage / Multi-fluid heat pumps• Synergies with other utilities• Feasibility heuristic criteria

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Proposed solution : computer aided integration method

Cycles Simulation by flowsheeting tools

Process integration by linear programming optimisation

Thermo-economic Performance evaluation by sizing, cost estimation and environmental impact evaluation

Heat pump data base models

Multi-objective optimisation

algorithm

Fluids/configurationsTechno selection operati

ng conditio

ns

Fluids/configurationscompressor

Electricity Hot and cold streams

Selection Flows

InvestmentOperating cost

Environmental impact

Process requirements

Other energy conversion units

Becker H., Spinato G. and Marechal F., 2011b, A multi objective optimization method to integrate heat pumps in industrial processes, Computer Aided Chemical Engineering 29, 1673–1677.

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Results for the case study (dairy process)

Pareto front: optimal solutions in terms of operating and investment costs

After 1350 evaluations

HP refrigerants: R717 / water

InvC OpC Fuel Cooling water Electricity HP units PR[kEuro] [kEuro/year] [kW] [kW] [kW] [-] [year]

ref 199.3 1662 862 162 -point1 244.2 198.3 1760 882 94 4 244.2point2 312.3 165.9 1404 589 122 5 9.4point3 357.0 128.1 1000 241 147 3 5.0

Investment cost [kEuro]

Ope

ratin

g Co

sts

[kEu

ro/y

ear]

260 280 300 320 340 360

130

140

150

160

170

180

190

200

point 1

point 2

point 3

Becker H., Spinato G. and Marechal F., 2011b, A multi objective optimization method to integrate heat pumps in industrial processes, Computer Aided Chemical Engineering 29, 1673–1677.

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Heat transfer restriction

Sub-system1

Sub-system2

Sub-system3

Heat transfer system (HTS)

No direct heat exchange possibleDirect heat exchange possible

Commonunits (CU)

Heat transfer units (HTU)

Energy SupportElectricity Fuel Water Air Inert Gas

Heat losses Solids Water GasWaste

Rawmaterials

EnergyservicesProductsByproducts

1

Bibliography

Abou-Khalil B., 2008, Methodologie d’analyses energetique et exergetique des procedes de trans-formation de produits dans l’industrie, Thesis, Ecole de Mines Paris.

ADEME, 2005, Note de cadrage sur le contenu co2 du kwh pas usage en france, ADEME.URL http://www.ademe.fr/

Adonyi R., Romero J., Puigjaner L. and Friedler F., 2003, Incorporating heat integration in batchprocess scheduling, Applied thermal engineering 23, 1743–1762.

Assaf K., 2010, Integration d’une pompe a chaleur dans un procede agro-alimentaire - simulation,experimentation et integration, Thesis, Ecole de Mines Paris.

Bagajewicz M., 2000, A review of recent design procedures for water networks in refineries andprocess plants, Computers and Chemical Engineering 24, 2093–2113.

Bagajewicz M. and Barbaro A., 2003, On the use of heat pumps in total site heat integration,Computers and Chemical Engineering 27, 1707–1719.

Bagajewicz M. and Rodera H., 2000, Energy savings in the total site heat integration across manyplants, Computers and Chemical Engineering 24, 1237–1242.

Bagajewicz M. and Rodera H., 2001, On the use of heat belts for energy integration across manyplants in the total site, Canadian Journal of Chemical Engineering 79 (4), 633–642.URL http://www.scopus.com/inward/record.url?eid=2-s2.0-0036322550&partnerID=40

Bagajewicz M. and Rodera H., 2002, Multiple plant heat integration in a total site, Americaninstitute of chemical engineering journal 48 (10), 2255–2270.

Bandyopadhyay S., Varghese J. and Bansal V., 2010, Targeting for cogeneration potential throughtotal site integration, Applied Thermal Engineering 30 (1), 6–14.

Becker H. and Marechal F., 2011, Energy integration of industrial sites with heat exchange restric-tions, Computers and Chemical Engineering, doi:10.1016/j.compchemeng.2011.09.014.

Becker H. and Marechal F., 2012, Targeting industrial heat pump integration in multi-period prob-lems, Proceedings of 11th International Symposium on Process Systems Engineering.

Becker H., Marechal F. and Vuillermoz A., 2011a, Process integration and opportunity for heatpumps in industrial processes, International Journal of Thermodynamics 14 (2), 59–70.

Becker H., Spinato G. and Marechal F., 2011b, A multi objective optimization method to integrateheat pumps in industrial processes, Computer Aided Chemical Engineering 29, 1673–1677.

Becker H., Vuillermoz A. and Marechal F., 2011c, Heat pump integration in a cheese factory,Chemical Engineering Transactions 25, 195–200.

153

Yes but : Sub-systems can not exchange heat (directly)

X X

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Modifying the heat cascade model

• Modification of the heat cascade

• Heat transfer constraints ?

Fobj = min(d · (nf�

f=1

(c+f

nu�

u=1

fuE+f,u) + c+

elE+el � c�elE

�el +

nu�

u=1

fucu))

nu�

u=1

fuE+el,u + E+

el �nu�

u=1

fuE�el,u ⇥ 0

nu�

u=1

fuE+el,u + E+

el � E�el �

nu�

u=1

fuE�el,u = 0

E+el � 0 E�

el � 0

nsh,k�

hk=1

fuQh,k,u �nsc,k�

ck=1

fuQc,k,u + Rk+1 � Rk = 0 ⇥k = 1..., nk

R1 = 0 Rnk+1 = 0 Rk � 0 ⇥k = 2..., nk

yu · fminu ⇥ fu ⇥ yu · fmax

u

Objective function

Heat cascade

Electricity consumption / production

Multiplication factors

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Heat cascade modifications

Sub-system Ain interval k

Heat of hot streams of sub-system A

Heat of cold streams of sub-system A

Cascaded heat fromhigher interval k+1

Cascaded heat tolower interval k

Heat supplied to sub-system Afrom the heat transfer system

Heat removed from sub-systemA to the heat transfer system

Sub-system Bin interval k

Heat of hot streams of sub-system B

Heat of cold streams of sub-system B


Heat supplied to sub-system Bfrom the heat transfer system

Heat removed from sub-systemB to the heat transfer system


Heat of hot streamsof heat transfer system

Heat of cold streamsof heat transfer system

Heat transfer system

in interval k



nsc,s=A,k!

cs=A,k=1

fuQc,s=A,k,u

Rs=A,k+1

Q!hts,s=A,k

Q+hts,s=A,k

nsh,s=A,k!

hs=A,k=1

fuQh,s=A,k,u

Rs=A,k

Q+hts,s=B,k

Rs=B,k+1

Rs=B,k

Q!hts,s=B,k

nsh,s=B,k!

hs=B,k=1

fuQh,s=B,k,u

nsc,s=B,k!

cs=B,k=1

fuQc,s=B,k,u

Rhts,k+1

Rhts,k

nsh,hts,k!

hhts,k=1

fuQh,hts,k,u

nsc,hts,k!

chts,k=1

fuQc,hts,k,u

MILP model => minimize the penalty of the constraints

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

−6000 −4000 −2000 0 2000 4000 6000−0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Heat Load [kW]

Car

not F

acto

r 1−T

a/T[

−]

Process

Utility without restrictions

boilerwithout constraints

cooling waterwithoutconstraints

Energy penalty due to restricted matches

Hot utility 6014 kW

Cold utility 1651 kW

Without constraints

Hot utility 9868 kW

Cold utility 5505 kW

Penalty 3854 kW

With constraints

−6000 −4000 −2000 0 2000 4000 6000−0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Heat Load [kW]

Car

not F

acto

r 1−T

a/T[

−]

ProcessUtility with restrictionsUtility without restrictions

boiler

cooling water

penalty

boilerwithout constraints

cooling waterwithoutconstraints

Do you accept the penalty ?

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Reduce the penalty of the constraints

Introduce heat transfer fluid

H

C

T

Q

2 Heat exchangersHigher ∆TminFlexibility

Temperature levels ?

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Intermediate heat transfer units - envelope curves

Sub-system Ain interval k

Heat of hot streams of sub-system A

Heat of cold streams of sub-system A



Heat supplied to sub-system Afrom the heat transfer system

Heat removed from sub-systemA to the heat transfer system

Sub-system Bin interval k

Heat of hot streams of sub-system B

Heat of cold streams of sub-system B


Heat supplied to sub-system Bfrom the heat transfer system

Heat removed from sub-systemB to the heat transfer system


Heat of hot streamsof heat transfer system

Heat of cold streamsof heat transfer system

Heat transfer system

in interval k



nsc,s=A,k!

cs=A,k=1

fuQc,s=A,k,u

Rs=A,k+1

Q!hts,s=A,k

Q+hts,s=A,k

nsh,s=A,k!

hs=A,k=1

fuQh,s=A,k,u

Rs=A,k

Q+hts,s=B,k

Rs=B,k+1

Rs=B,k

Q!hts,s=B,k

nsh,s=B,k!

hs=B,k=1

fuQh,s=B,k,u

nsc,s=B,k!

cs=B,k=1

fuQc,s=B,k,u

Rhts,k+1

Rhts,k

nsh,hts,k!

hhts,k=1

fuQh,hts,k,u

nsc,hts,k!

chts,k=1

fuQc,hts,k,u

Fictive hot stream in interval k

Fictive cold stream in interval k

Qh,env,k

Qc,env,k

Becker H. and Marechal F., 2012a, Energy integration of industrial sites with heat exchange restrictions, Computers and Chemical Engineering 37, 104–118.

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Envelope composite for heat transfer fluid selection

Heat Load [kW]

Tem

pera

ture

[°C

]

0 2000 4000 6000 8000 10000

50

100

150

200

250

25

75

125

175

225

275

300

Envelope cold composite curvesEnvelope hot composite curves

Steam production 80 bar

Steam utilization 7 bar

Steam utilization 2 barIntermediate hot water loop25°C - 80°C

Hot envelope

Cold envelope


francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Integration of intermediate heat transfer units

−3000 −2000 −1000 0 1000 2000 3000 4000 5000−0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Heat Load [kW]

Car

not F

acto

r 1−T

a/T[

−]

ProcessUtility



Steam production 80 bar

Boiler

Intermediatehot water loop

Heat transfer constraints

•Heat recovery•Heat transfer fluid flows•Cogeneration

•Steam network•Heat pumps as a heat transfer fluid

Heat transfer constraints

•LP formulation•Easier HEN design•Multi-level constraints


francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Résultats

Unit No With Constraints andconstraints constraints heat transfer system

Operating Costs [kEuro] 2353.6 3844.8 2180.2Fuel consumption [kW] 6073 9920 8026Cooling water [kW] 1668 5516 1602Electricity [kW] 2019

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Heat load distribution calculation

400

600

800

1000

1200

1400

pinch

pinchpinch

5409.34 kW

84.99 kW

2786.39 kW

555.86 kW

166.71 kW

46.92 kW156.41 kW20.15 kW5956.76 kW29.12 kW75.91 kW29.99 kW791.44 kW112 kW56.32 kW100.56 kW

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16280

300

320

340

360

380

Stream No.

Corre

cted

tem

pera

ture

[K]

pinch

pinch

pinch

46.92 kW156.41 kW20.15 kW5956.76 kW29.12 kW75.91 kW

29.99 kW

791.44 kW

112 kW

56.32 kW

100.56 kW

421.38 kW

167.25 kW

98.43 kW

3197.68 kW

836.14 kW

3030.43 kW

7297 kW

1602.38 kW

1: pulping ph h12: drying st h33: drying st h24: boiler boi h15: boiler boi h26: wloop waterhe7: D1 C Ds8: D2 C Ds9: drying air h110: pulping ph c111: drying air c112: water cw13: boiler boi c114: wloop waterco15: drying st c116: C H1 Cs

Streams

Hot stream Cold stream Heat load [kW]pulping ph h1 pulping ph c1 7297.0wloop eauhe pulping ph c1 3030.4D1 C Ds pulping ph c1 836.1D2 C Ds pulping ph c1 98.4drying air h1 drying air c1 421.4drying air h1 water cw 1602.4boiler boi h2 boiler boi c1 30.0wloop waterhe boiler boi c1 167.3drying air h1 wloop waterco 3197.7drying st h3 drying st c1 100.6drying st h3 C H1 Cs 791.4drying st h2 C H1 Cs 112.0drying air h1 C H1 Cs 56.3

Exemple pour zone 1

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Multi-period problem

• Non simultaneous processes

• => multi-period problem– Heat storage tanks for heat recovery

Period2

streams 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

other

pasto1

pasto2

pasto3

pasto4

pasto5

evapo

proc6

proc7

proc8

proc9

proc10

heat

CIP

Times T1 T2 T4 T5 T8 T9 T10 T11 T13 T14

Saturday

hours

T3 T6 T7 T12 T15

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Storage tank integration

Storage tank2

Tst,2

HEX h1(t)

heat demand

Storage tank1

Tst,1

...

Tst,...

Storage tankl

Tst,l

...

Tst,...

Storage tanknl

Tst,nl

HEX h2(t) HEX h..(t) HEX hl(t) HEX hnl-1(t)

heat demand heat demand heat demand heat demand

heat excess

HEX c1(t)

heat excess

HEX c2(t)

heat excess

HEX c..(t)

heat excess

HEX cl(t)

heat excess

HEX cnl-1(t)

Units for heat integration

HL HL HL HL HL HL

HEX: Heat exchangersHL: Heat losses

Bibliography

Abou-Khalil B., 2008, Methodologie d’analyses energetique et exergetique des procedes de trans-formation de produits dans l’industrie, Thesis, Ecole de Mines Paris.

ADEME, 2005, Note de cadrage sur le contenu co2 du kwh pas usage en france, ADEME.URL http://www.ademe.fr/

Adonyi R., Romero J., Puigjaner L. and Friedler F., 2003, Incorporating heat integration in batchprocess scheduling, Applied thermal engineering 23, 1743–1762.

Assaf K., 2010, Integration d’une pompe a chaleur dans un procede agro-alimentaire - simulation,experimentation et integration, Thesis, Ecole de Mines Paris.

Bagajewicz M., 2000, A review of recent design procedures for water networks in refineries andprocess plants, Computers and Chemical Engineering 24, 2093–2113.

Bagajewicz M. and Barbaro A., 2003, On the use of heat pumps in total site heat integration,Computers and Chemical Engineering 27, 1707–1719.

Bagajewicz M. and Rodera H., 2000, Energy savings in the total site heat integration across manyplants, Computers and Chemical Engineering 24, 1237–1242.

Bagajewicz M. and Rodera H., 2001, On the use of heat belts for energy integration across manyplants in the total site, Canadian Journal of Chemical Engineering 79 (4), 633–642.URL http://www.scopus.com/inward/record.url?eid=2-s2.0-0036322550&partnerID=40

Bagajewicz M. and Rodera H., 2002, Multiple plant heat integration in a total site, Americaninstitute of chemical engineering journal 48 (10), 2255–2270.

Bandyopadhyay S., Varghese J. and Bansal V., 2010, Targeting for cogeneration potential throughtotal site integration, Applied Thermal Engineering 30 (1), 6–14.

Becker H. and Marechal F., 2011, Energy integration of industrial sites with heat exchange restric-tions, Computers and Chemical Engineering, doi:10.1016/j.compchemeng.2011.09.014.

Becker H. and Marechal F., 2012, Targeting industrial heat pump integration in multi-period prob-lems, Proceedings of 11th International Symposium on Process Systems Engineering.

Becker H., Marechal F. and Vuillermoz A., 2011a, Process integration and opportunity for heatpumps in industrial processes, International Journal of Thermodynamics 14 (2), 59–70.

Becker H., Spinato G. and Marechal F., 2011b, A multi objective optimization method to integrateheat pumps in industrial processes, Computer Aided Chemical Engineering 29, 1673–1677.

Becker H., Vuillermoz A. and Marechal F., 2011c, Heat pump integration in a cheese factory,Chemical Engineering Transactions 25, 195–200.

153

To cascade during time t

From heat cascade during time t

Multi period formulation of the MILP heat cascade problem•Heat transfer constraints in each period•Selection of heat pump system for all the periods•Heat storage tank model from t to t+1

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Flows defined for each time t by optimisation

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150

0.2

0.4

0.6

0.8

1

Rel

ativ

e us

age

rate

boilercooling waterrefrigeration1refrigeration2heat pump

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150

20

40

60

80

100

Time slice

Stor

ed v

olum

e [m

3] Storage distribution

tank1 (20°C)tank2 (30°C)tank3 (40°C)tank4 (50°C)tank5 (60°C)tank6 (70°C) tank7 (80°C)tank8 (90°C)

Simultaneous resolution for each time t

•Operating strategy : tank level management•Size of the tank : max for all period + Investment cost

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Smart grids configuration

Connected to the main grid

− Stochastic Renewable sources Virtual Power Plant (VPP)

Heat driven consumers/producers− Cogeneration− Heat pumps

Heat/cold storage Tanks

− Predictive control

External Grid

VPP* central

VPP east

VPPwest

*VPP: Virtual Power Plant**MPC: Model Predictive Control

VPPnorth

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Smart Grid : optimal control of the system

External Grid

Node

Heat Pump

Cogen

Electricity flowHeat flow

Solar PV

Solar thermal

El. storagePower conditioning

Heat storage

Gas

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Integration in smart grids


4 Perspectives of the integration of the trigeneration system

The approach presented above is based on the time averaging approach that allows to considerthat all the streams are simultaneous. Considering the batch operation dimension requires theadaptation of the approach to integrate in the analysis the calculation of the storage tanksthat are required to make the heat recovery feasible. When studying the trigeneration systemintegration, it will be necessary to size the tanks not only to allow the heat recovery but alsoto take opportunities from the electricity market. The trigeneration system is indeed a wayof storing electricity from the grid in the form of heat or cold. The heat or cold storage alsoallows the cogeneration unit to play the role of the peak shaving.

The final configuration is presented on figure 8. The optimization method based on amulti-objective optimization strategy presented by Weber et al. ([17]) allows to design thetrigeneration system and the storage tanks considering the use of a predictive optimal manage-ment strategy. In addition, methods like the one proposed by Collazos et al. ([11]) can be usedto implement the predictive optimal management strategy in a control system.

3

6

Process hot water

CIP system

CoolingGlycol

110 °C

50 °C

2 °C

EngineGas

Industrial process

Refrigeration/Heat Pump

Electricity

Figure 8: Storage tank system configuration

5 Conclusion

The optimal integration of trigeneration systems is realized in several steps. The first stepis the definition of the requirement followed by the definition of the heat recovery potentialbetween the hot and the cold streams of the process. This step is mandatory since it allows todefine the heating and cooling requirement to be satisfied by the trigeneration system. Otherapproaches based on the use of the present utility system would lead to bigger systems andunnecessary investment that would in addition prevent the future energy savings options. Thetrigeneration system is sized by first identifying the possible trigeneration options based on theanalysis of the Grand composite curve of the system. The configuration of the system is thendefined by applying an optimization model that calculates the best flows in the system. It hasbeen demonstrated that the proper analysis of the trigeneration system requires to accountfor the possible integration, not only at the level of the process, but also at the level of thepossible integration inside the trigeneration system. The example presented shows that thecombination of a refrigeration cycle where the condensation heat is used as a heat pump topreheat the process streams with a mechanical vapor recompression system that allows for

9

External Grid

francois

.marechal@

epfl

.ch ©

Industria

l Energy S

ystem

s L

aboratory-

LEN

I-IG

M-S

TI-

EPFL 2

012

Conclusions : Process integration in the food industry

• Present consumption does not define the energy conversion demand !– buy 1 x 500 kWe unit instead of 2x 1250 kWe unit !

• Analyze systems requirement : Process efficiency– Consider water usage– Consider waste streams that could be recovered– Consider heat recovery before energy conversion integration– Consider waste streams as resources

• Integrate the conversion system : Energy conversion efficiency– integrated system (interdependent flows)– Optimize the flows (MILP + heat cascade)

• Mathematical models helps integrate process constraints– Restricted Matches– Multi-period problems– Storage systems

• Design the system : create confidence in proposed solutions– Consider system operation– Opportunities from system control ?

15

Documents

Process Integration : Application in the food industry · PDF fileProcess Integration : Application in the food industry h ... –Heat integration –Energy conversion integration