study of a cogeneration plant for agro-food industry

Study of a cogeneration plant for agro-food industry

Francesco Fantozzi*, Sandro Diaconi Ferico, Umberto Desideri

Dipartimento di Ingegneria Industriale, UniversitaÁ di Perugia, Via G. Duranti 1A/4, 06125 Perugia, Italy

Received 20 March 1999; accepted 5 August 1999

Abstract

A technical and economic feasibility study for a natural gas fueled cogeneration plant was conductedin an important Italian pasta and animal feed factory. The layout analysis pointed out three maindivisions; in each division electric and thermal users were pointed out and their e�ective energyconsumption and power demand rate was monitored. A technical feasibility analysis was then carriedout to determine the type and scale of the possible Combined Heat and Power (CHP) plants focusing onInternal Combustion Engines (ICEs) and gas turbine based power plants. The actual energy costs wereevaluated on the base of the energy bills for the biennium 1996±97 while the detailed economicfeasibility analysis was conducted on the base of the o�ers received from manufacturers on the market.The results obtained show the possibility to have low payback periods and appealing internal rate ofreturns when investing on ICEs based CHP plants covering the entire electric demand and partiallyful®lling the thermal needs of the factory. 7 2000 Elsevier Science Ltd. All rights reserved.

Keywords: CHP; Agro-food industry; Internal combustion engine; Techno-economic assessment

1. Introduction

The food industry is present throughout Italy and contributes considerably to its economy.A typical food factory consumes a considerable amount of heat to process raw materials andto preserve food from the action of bacteria. Electric energy is required as well to run all theprocess machinery. Therefore energy costs contribute considerably to the cash ¯ow of the ®rm.

Applied Thermal Engineering 20 (2000) 993±1017

1359-4311/00/$ - see front matter 7 2000 Elsevier Science Ltd. All rights reserved.PII: S1359-4311(99)00074-5

www.elsevier.com/locate/apthermeng

* Corresponding author. Tel.: +39-0755-852-738; fax: +39-0755-852-736.E-mail addresses: [email protected] (F. Fantozzi), [email protected] (U. Desideri).

Nomenclature

CF Cash FlowCHP Combined Heat and PowerDCF Discounted Cash FlowENEL Italian electricity providerGT Gas TurbineHRSG Heat Recovery Steam Generatori Interest rateI InvestmentICE Internal Combustion EngineIRR Internal Rate of ReturnLHV Lower Heating Value (kJ/Sm3)MARR Minimum Attractive Rate of ReturnMV Medium VoltageNPV Net Present ValuePP Payback PeriodQ Heat (kW)RR Rate ReturnSH SuperheatedSNAM Italian natural gas providerT Temperature (8C)VAT Value added taxW Power (kW)f Relative Humidityx Speci®c Humidity (kg/kg)

SubscriptsA AmbientE ExtrusionS SievingW WrappingD Dryinga Aire Electricf Fuelp Pastar Recoveredt Thermal0 Zero (year zero of an investment)

F. Fantozzi et al. / Applied Thermal Engineering 20 (2000) 993±1017994

In a very competitive sector of the market, such as the food one, a reduction in energy costsmay turn in an important factor for the competitiveness of the ®rm [1,2].Thanks to a positive economic conjuncture on the natural gas market, to the high e�ciency

of the latest natural gas fueled internal combustion engines and to the growing awareness inenergy savings, CHP plants are becoming very popular in Italy even without the governmentaleconomic contributions that characterized the ®rst phase of their spread [3]. These plant arenow mature to become a competitive heat and electricity resources for the new born Italianfree market of energy.Pasta is the Italian food for excellence therefore an important testbench for such a

technology in the food industry. An important Italian food and animal feed factory was thenconsidered in a technical and economic feasibility study for a cogeneration plant.

2. Manufacturing cycle analysis

The plant under study is a pasta and animal feed manufacturing site which is the productivenucleus of an important Italian food ®rm. The production can be divided into three mainplants that can be analyzed separately because they are independent in terms of energy,materials requirements and management. Namely the production plants are: the mill, the pastaand the animal feed factory. Obviously the divisions exchange products so the mill grindscereals both for pasta and animal feed while pasta byproducts are mixed as animal feedintegrators.A brief description of the three divisions is now presented [4±6].

2.1. The mill

The machinery of a grinding site is usually distributed in di�erent ¯oors of a verticalbuilding. Wheat, ¯our and by-products are moved from machine to machine by a pneumaticsystem or by gravity. The main operation that are accomplished in a mill are: cleaning of thewheat, setting the right degree of humidity (conditioning), grinding and selection of theproducts. Following the scheme of Fig. 1, the wheat arrives at the reception where a sample is

Fig. 1. Layout scheme of the mill.

F. Fantozzi et al. / Applied Thermal Engineering 20 (2000) 993±1017 995

taken in order to check if the moisture is acceptable. Once that the wheat has been acceptedand unloaded it passes through a fanning mill and a series of screen in order to provide a ®rstrough pre-cleaning of the product. The byproducts of this phase (mainly hay, seeds, insects anddust) are selected and will be ground together with other residuals and the resulting mealadded to animal feed. The partially cleaned wheat is then stocked into vertical silos waiting toenter the mill. The ®rst operation carried out in the mill is the cleaning of the cereal byscreening and bolting. The machines are of the recirculating air type and the mixture is blownby fans through riddles, screens and vibrating separators. The cleaned wheat is then mixedwith water in order to increase the presence of moisture and facilitate the separation of thebran in the following grinding phase. The residuals of the cleaning phase are mixed with thepreviously gathered ones.The grinding phase is carried out by cluster mills with di�erent numbers of cylinders and

di�erent surface ®nish depending on the type of wheat (hard or tender) and on the stage ofgrinding. After each phase of grinding the products are pneumatically carried to the separatorsplaced in the upper ¯oors where the bran is sorted from the ¯our. The bran is then added tothe residuals of the cleaning phases and the whole mixture is ground and directed to theanimal feed factory.The ®ne ¯our obtained from the tender wheat is packaged and sold while the hard corn meal

is sent to the pasta factory.The energy requirements of this site are only electric and correspond to the power required

to run the mills, the sorting and cleaning machines and the fans for the transportation of theproducts.

2.2. The animal food factory

A composite animal food factory can be addressed as a `mixing' factory rather than a`producing' factory because its goal is to add the appropriate quantity of each compound inthe resulting feed. Nevertheless some processing steps are required to prepare the di�erent

Fig. 2. Layout scheme of the food division.


ingredients (Fig. 2). The raw materials that enter the animal feed factory can be divided intofour main categories: seeds (cereals and legumes), meals (from the food industry as the onesfrom oil extraction), liquids (molasses, fats, choline and beet protein concentrate) andintegrators (vitamins, antibiotics etc.). At the reception, before stockage, a quality control oningredients is made. From the stocking silos ingredients are then sent to the mixer while someof them experience a pre-treatment: seeds are ground by four hammer mills in order to renderit more digestible and cereals are cooked with steam, crushed by a rolling mill and steam driedto improve the attractiveness and the digestibility. The ground seeds and crushed cereals enterthe mixer together with liquids, integrators and molasses previously heated with hot water toimprove their capacity to be mixed. To improve the attractiveness and the preservability of theproduct, the meals obtained are often compressed into small blocks named pellets. Thisoperation consists in a preparation phase in which the meal is heated and humidi®ed withsteam and molasses are added. At this stage the binding capability of molasses is exploitedrather than their energetic value. The mixture is then compressed by six presses and cut to therequired length. The pellets obtained are then cooled down to ambient temperature.The energy requirements of this site are both thermal and electric. Heat is supplied from a

thermal unit to produce steam at 12 bar together with hot water at 808C. The pellettingoperation is the most onerous, in terms of electric power consumed (1 MW), of the wholefactory and it is thus carried out in the night hours.

2.3. The pasta factory

According to Italian regulations, pasta is de®ned as the product obtained by mixing,extruding or drawing and ®nally drying a dough of water and hard wheat ¯our. The procedureset up to obtain traditional pasta is slightly varying depending on the variety of pastaproduced (Fig. 3). Three main streams can be isolated: short pasta (macaroni, penne etc.), longpasta (spaghetti, fettuccine etc.) and nested pasta (tagliolini). The dough used is the same sothe processes can be described as unique pointing out the di�erences in the procedures (Fig. 3).Ingredients such as ¯our and water are directed to the mixer from where a uniform dough witha humidity over 30% is drawn out. A tangent screw compresses the dough forcing it to passthrough a drawplate whose holes determine the form of the pasta; ®nally a cutting-o� machinecuts the pasta to the right length. The resulting pasta is then directed to the drying section.Before passing through the dryers di�erent varieties of pasta are prepared in di�erent manners:long pasta is doubled and hung on canes, short pasta is laid on screens as well as the nestsobtained by tangling a thin uncut long pasta. Long and nested pasta could not resist theexsiccation without deforming or sticking together. To prevent this, the screens are put intovibration and hot air (0608C) is blown through for approximately 10 min. This forms a driedand strong super®cial ®lm that maintains the shape and prevents the single pieces of pasta tostick together when laid one on the other in the dryers. A similar result is obtained for longpasta simply keeping it in movement with hot air.The drying process, depending on the variety of the pasta, may last from 5±6 h (short) to

10±12 h (long) and it is made by passing the canes and screens through a tunnel were air atdi�erent temperature is blown. The process is always made in two steps and the di�erenttemperatures are reached gradually to prevent thermal stresses that may break down the


structure. In a ®rst pre-exsiccation phase (commonly addressed as the wrapping phase) thatlasts from 30 min to 1 h the surface is dried and humidity passes from 30% to 17±19% leavingthe inside still humid. The second phase is the real exsiccation in which humidity is broughtdown to 12,5%: the entire exsiccation processes eliminates the whole water added in the mixerplus 3% of the humidity contained in the meal. Finally pasta is cooled down to ambienttemperature with fresh air and sent to the packaging section ready to be shipped.Energy requirements of the pasta division are both electric and thermal. Electricity is

required to run the operating machines and transportation devices. Heat is required for thedrying processes: hot air is warmed in a heat exchanger by superheated water at 1208C,produced at 5 bar in a dedicated thermal unit.

3. Energy loads analysis

The manufacturing cycle analysis carried out in the previous section shows for each group

Fig. 3. Layout scheme of the pasta division.


the main energy demands to which we must add energy consumption for the lightning andconditioning of the administration buildings (Table 1).A detailed historic analysis of electro-thermal consumption in the biennium 1996±1997 was

made to obtain average values of power demand rate and needs. Data were collected fromenergy consumption registers and electric and gas bills and related to the e�ective workinghours per year provided from the administration.

3.1. Electric energy

Electricity is totally supplied from the Italian electricity dealer ENEL at a middle voltage (20kV). Appendix A shows how the cost of the kWh varies depending on the hour of the day andon the contractual power demands rate. For the pasta-feed factory the contractual powerdemands rate in each hour band are shown in Table 2.Figure 4 shows the average electric consumption of each group (kWh/yr) and of each

substation. Data are obtained by averaging the measures collected on a monthly base in thebiennium 1996±97. Some considerations can be made.The animal feed factory consumes 46% of the electric energy required in the whole ®rm and

most of it (59%) is used to run the pelletting machines. The pasta factory consumes 29% ofthe total energy and 93% of it is used to run the kneader, the extruders and the dryers. Themill contributes with a 25% to the total and 66% of it is used only in the milling operation.Less than 1% is consumed in the o�ces.Considering data from another point of view gave interesting information when focusing on

the monthly consumption and power demand in each group.Data referring to this analysis are not presented but it showed that:

Animal food factory. The energy consumption shows positive peaks, with respect to theaverage value of 762 � 103 kWh/month, in the months of January, May, July and October,negative peaks in June and November. The trend goes alongside with the pellettingoperation. The average power demand rate is about 1300 kW with a maximum monthlydeparture from this value of about 12%.Mill. The average monthly energy requirement is 416 � 103 kWh/month with a su�cientlyconstant trend that departs negatively in November and December. The power demand canbe considered approximately constant on a value of 750 kW with less than 10% ofmaximum departure.Pasta factory. Energy consumption shows negative peaks (avg. 483 � 103 kWh/month) inApril, August and December in correspondence with disinfestation, operational andunscheduled maintenance respectively. Power demand can be assumed 800 kW throughout

Table 1Summary of energy usage

Mill electricityAnimal food factory electricity saturated steam (12 bar), hot water (808C)Pasta factory electricity superheated water (5 bar, 1208C)O�ces electricity hot water (808C)


the year with less than 10% departures.O�ces. Energy consumption has a constant trend with an average value of about 6280kWh/month; power demand trend is slightly oscillating around 35 kW.

Figure 5 shows the average monthly energy and power consumption in each of the four hourbands. All the power demands are calculated dividing the monthly energy consumption by thenumber of working hours of the month. Working hours relating to August were not reliablebecause the maintenance hours were not reported. Twenty working days in August 1996 and19 in august 1997 were then assumed. In any case all the average values are calculated oneleven months.

Table 2Contractual power for the factory

Power demand rate F1 F2 F3 F4

Bounded (kW) 3100 3500 3500 3500Maximum available (kW) 3875 4375 4375 4375

Fig. 4. Average electric consumption for each division and main operations (kWh/a).


In conclusion it can be stated that:

. the power demand of the whole industrial site in the F1 to F3 varies slightly throughout theyear between 2800 kW and 3200 kW (except April);

. the power demand in band F4 presents signi®cant oscillations;

. the overall energy consumption throughout the year is of 20 GWh with a relatively constanttrend; sensible variations from the average value of 1667 MWh/month are present only inthe months dedicated to disinfestation and maintenance.

3.2. Thermal energy

The thermal energy requirements of the factory are satis®ed by dedicated heat productionunits. The fuel burnt in the di�erent boilers is always natural gas with the exception of the feedfactory emergency boiler which is ®red by fuel oil. The gas is entirely supplied by the nationalgas distributor SNAM with a continuous contract, high use and binomial tari�. The so called`binomial' tari� considers the price of gas made of two contributions: a constant part, linkedto the daily gas demand rate, varies each semester according to socio-economic parameters.The second part is proportional to the volumes of gas drawn and the tari� varies monthly inrelation to the LHV of the gas and to the prices of fuel oils on national and internationalmarkets. The average LHV of the gas supplied in 1996 and 1997 was 37,631 kJ/Sm3 and it canbe assumed as constant during the two years.

3.2.1. Animal food factoryThe animal food factory has two di�erent heat lines: one for the production of saturated

steam (12 bar), used to cook cereals, heat up liquids and in the pelletting operation. The otherheat line is for the production of hot water (808C) used in the heating of molasses. Two boilersof 20,000,000 kcal/h nominal power are used to produce steam: a diathermic oil natural gas

Fig. 5. Contribution to the average monthly electric energy consumption and electric power demand rate of eachhour band.


®red one is used in normal operation while a traditional forced circulation boiler ®red bybunker C fuel oil is used for emergencies. Hot water is produced by two natural gas ®redforced circulation boilers of respectively 135,000 kcal/h and 74,000 kcal/h nominal power. The®rst is run in winter while the second in summer; sometimes, the two are run together inwinter. Steam and hot water are produced by heating water from the local water supply andthey are both absorbed by the process.Figure 6 shows the average monthly gas consumption for the feed division in the two years

considered. Thermal requirements for hot water and steam production are as well shown inFig. 6 assuming a boiler e�ciency of 0.9 and an inlet water temperature of 408C (after thewater treatment section). The corresponding average ¯ow rates values obtained for hot waterand steam are respectively of 2273 kg/h and 1337 kg/h.The average trend of energy consumption varies considerably from 1996 to 1997 depending

on di�erent atmospheric conditions; this energy is mainly used to heat up liquids and molasseswhose initial thermal level is proportional to ambient temperature.

3.2.2. Pasta factoryTo run the dryers the pasta division uses superheated water (1208C, 5 bar) produced in two

natural gas fueled diathermic oil boilers of respectively 5,000,000 kcal/h and 3,000,000 kcal/hnominal power. Superheated water ¯ows in a closed circuit entering the boiler at a temperatureof approximately 1088C.The monthly average gas consumption is shown in Fig. 7. together with the average energy

consumption obtained as explained in Appendix B.O�ces are heated by a traditional natural gas ®red boiler of 400,000 kcal/h delivering hot

water at 808; energy consumption and power demand rate are presented in Fig. 6.Figure 8 shows how thermal energy is consumed in each section, with data expressed in kcal/

year. The pasta division is the biggest user, consuming 67% of the whole thermal needs. This

Fig. 6. Contribution to the food division's average monthly natural gas consumption and thermal energyconsumption of each thermal user.


heat is all used to diminish the pasta content of humidity and is distributed in the three phases:sieving (46%), wrapping (40%) and drying (14%). The feed division absorbs 31% of the heatproduced mainly for the production of saturated steam (90%). Finally o�ces consume only2% of the total and all concentrated (Fig. 7) in the cold season.

4. Technical feasibility analysis

Cogeneration means the simultaneous generation of electric power and heat through thesame combustion process. It is a wise way of exploiting the chemical energy of a fuel consisting

Fig. 7. Pasta division and o�ces average natural gas consumption and energy requirements.

Fig. 8. Average thermal energy consumption for each division and main thermal vectors.


in a ®rst high temperature conversion to mechanical and then electric energy. The residualenergy content of combustion gases and refrigerating ¯uids is used as a thermal source for heatutilities thus recovered.Appendix C describes the most common cogenerative modules as well as the characteristic

indexes usually utilized to evaluate their performances.A CHP plant is technically feasible, thus probably economically convenient as well, when

four main constraints are satis®ed:

. suitable ratio between electric and thermal energy requirements;

. simultaneous demand of heat and electricity;

. low temperature heat demand;

. vicinity of electric and thermal users.

Once the technical feasibility has been veri®ed, the possible plant should be identi®ed andsized. This phase may be carried out following three main alternative procedures:

. coverage of the total thermal thus concentrating on heat demand. The eventual excess ofelectricity can be sold (At the moment Italian regulations on the matter are varyingaccording to the liberalization of the energy market);

. coverage of the total electric thus concentrating on electricity demand. It is preferable towaste as little heat as possible because in this case heat in excess is not marketable. In anycase the loss would be of less importance because of the reduced quality of energy;

. coverage of the base thermal and electric loads thus still depending on the grid and ontraditional boilers for the coverage of peak electric and thermal loads respectively.

The pasta and animal feed factory satis®es the four conditions mentioned above, havingenergy loads evenly distributed, as it results from Fig. 5 and a low temperature heat demand(hot water). Moreover the coverage of base thermal and electric loads appears the mostpracticable solution for the sizing of the plant. The plant should be located in order tominimize thermal losses and piping costs in the delivery of the recovered heat: a possible layoutwould then consider the thermal requirements of the pasta division in alternative to the feeddivision ones, being satis®ed the whole electric requirement. The biggest thermal user is thepasta division which is next to the o�ces and opposite to the feed division. Therefore whenconsidering the pasta division as thermal user, the low temperature recovery would considerthe hot water of the o�ces which has also a bigger heat request with respect to the hot waterneeded in the feed division to heat up liquids. The main counterpart is that o�ces have verydiscontinuous thermal needs during the day and during a year and often the low temperatureheat recovery will be of no use.Di�erent solution have been considered with an electric power output of about 3 MW

therefore focusing on ICEs and GTs.

4.1. ICE based CHP

ICEs recover heat at two di�erent temperatures: at low temperature (90±1208C) from thejacket (25% of the heat available for recovery), lubricating oil (12%) and aftercooler (13%)cooling systems and at high temperature (450±5008C) from the exhaust gases (50%).


Considering the thermal needs of the factory, the high temperature recovery will be used forthe production of steam or SH water while the low temperature recovery will be dedicated tothe production of hot water.A generic ICE of 3 MWe with 40% of electric e�ciency would provide approximately the

same amount of thermal power, cooling down the gases to 1208C, equally distributed betweenhigh and low temperature. Therefore, considering the power requirements of Figs. 6 and 7 itappears impossible to recover the whole heat required for SH water. Two solutions areenergetically feasible for the recovery:

Solution (A) Partial heating of SH Water using the hot gases enthalpy and o�cesconditioning with low temperature recovery;Solution (B) Production of the whole steam required by the animal feed factory with hotgases enthalpy and o�ces conditioning with low temperature recovery.

A four stroke, natural gas fueled ICE was considered among the ones available on themarket. The main operational parameters are reported in Appendix D.Figure 9 shows the solutions considered while in Fig. 10 the electric power production is

compared to the factory needs in each hour band. Positive values represent the surplus ofpower needed with respect to the one auto-produced while negative values mean that energywas overproduced and therefore conveyed to the grid. These values are calculated consideringan availability of the engine of 6600 h/yr and machine stops for maintenance both as stated bythe manufacturer. Engine maintenance was considered, when possible, in parallel with the

Fig. 9. Possible solutions for the ICE and GT based power plant.


production stops for disinfestation and maintenance to maximize working hours. The biggestcolumns refer to the solution with one engine which is not always su�cient for the coverage ofthe whole electric needs therefore a so called integrative contract was considered with ENEL.With this solution it is possible to cover the power peaks (the plant is slightly undersized) andto run the factory during the power plant maintenance periods and in case of damage, thoughconsidering a higher price, with respect to a normal contract, for kWh drawn. For both thesolutions, the characteristics indexes were calculated splitting the cold season (Jan, Feb, Mar,Oct, Nov, Dec) when hot water is produced, from the hot season when there is no recovery atlow temperature. They are shown in Table 3.To optimize the low temperature recovery a further solution was assumed in which the hot

water produced is used to heat the pasta division storehouse in the solution A and the feeddivision storehouse for the solution B. The thermal load of the two was assumed considering avolume of approximately 15000 m3 and a speci®c thermal dispersion of 65±70 W/m3 equal to 1MWt. Obviously this further recovery is to be considered only in the cold season.The analysis of the characteristic indexes shows clearly that:

. solution A is always more performing than solution B thus it will be the one considered inthe subsequent cases;

. the optimization of the low temperature recovery improves the performances siugni®cately.

The solution A was then evaluated for a di�erent arrangement in which two engines of halfpower were considered instead of one. This was made mainly to reduce the complete stops formaintenance that can now be made alternatively on one of the two engines and, as shown in

Fig. 10. Electric energy balance for ICEs and GT based power plants: overproduction and underproduction. Widecolumns represent the 1 ICE solution, medium columns represent the 2 ICEs solution and thin columns represent

the GT solution. The contribution to the global production of each hour band is shown with di�erent colours.Negative values represent overproduction while positive power underproduction.


Table 3, to improve the power production thus diminishing the need for a power integration.The plant is organized in the same way of the previous one with the only di�erence that theICE is split in two ICEs working in parallel. Referring to solution A then, a slightly higherexhaust mass ¯ow allows a SH water outlet temperature of 1168C, one degree higher than theprevious solution. This better performance at high temperature is opposed by a higher amountof heat lost at low temperature being the hot water mass ¯ow rate available of 132,500 kg/h.The result is anyway positive as it shown by the better thermal e�ciency with respect to thesolution with one engine.The production is actually organized in order to concentrate the power use during low cost

hour bands. This results (Fig. 10) in an excess of power in F1 and over requests in F2 and F3.Excess in F4 is due to the limited possibilities of production in these hour bands. When auto-producing there is no di�erence between hours and it is possible to reorganize the productionin order to equally distribute the operations throughout the day. This can balance over andunder power production thus avoiding the need for an integration contract. An assistancecontract in case of damage is as well not necessary because with two engines a simultaneousdamage has a very low probability.

4.2. GT based CHP

GTs are very suitable engines for cogenerative use. Heat recovery is always from exhaustgases at high temperature (05508C) generally by means of a Heat Recovery Steam Generator(HRSG) or Hot Water generator. There is no decay of the electric e�ciency, when introducingheat recovery, if we neglect the pressure drops in the generator. GT based cogeneration plantare convenient when there is a strong heat demand at high temperature because, as it shown in

Table 3

Values assumed by the characteristic indexes for the di�erent solutions

Season E.I. U Ue Ut

Solution A 1 engine cold 1.70 0.67 0.42 0.25hot 2.18 0.61 0.42 0.19

Solution A+storehouse heating cold 1.00 0.84 0.42 0.42

hot 2.18 0.61 0.42 0.19Solution B 1 engine cold 2.17 0.61 0.42 0.19

hot 3.00 0.56 0.42 0.14

Solution B+storehouse heating cold 1.14 0.79 0.42 0.37hot 3.00 0.56 0.42 0.14

Solution A 2 engines cold 1.58 0.68 0.41 0.26

hot 1.94 0.63 0.41 0.21Solution A+storehouse heating cold 0.97 0.84 0.41 0.43

hot 1.94 0.63 0.41 0.21Solution C 1 GT-1P HRSG cold 0.76 0.56 0.24 0.32

hot 0.82 0.54 0.24 0.29Solution C 1 GT-2P HRSG cold 0.42 0.82 0.24 0.58

hot 0.63 0.63 0.24 0.38


Table D1, the heat available, with respect to an ICE, is about the double while the electrice�ciency is considerably reduced. Technical and economic data on the GT are taken from amodel amongst the ones available on the market and brie¯y reported in Appendix D.Figure 10. shows the balance between electric over and sub-production considering, as stated

by the manufacturer, an availability of the engine of 7000 h/yr and 180 h/yr for maintenance,these latter considered to be made during the factory stops.Due to the low electric e�ciency, the GT based power plant has a high need for integration

in F1 and F2 versus a low overproduction in F4. A contract with ENEL providing emergencyaid and integration is then considered as well. Therefore the way to make convenient such aninvestment is to recover the maximum heat possible from exhaust gases. The thermal powerfrom exhaust is then considered to cover the whole thermal needs of the entire factory, namelySH water for pasta factory, steam for the feed division and hot water for both the feeddivision and the o�ces, by means of a single pressure level HRSG (Solution C). Figure 9shows the arrangement of the di�erent heat exchanger in the HRSG according to the di�erenttemperature level of the heat required.As expected, Table 3 shows the poor values assumed by the characteristic indexes due to the

still considerably high amount of unrecovered heat: exhaust gases are cooled only to 3008C. Tocool up hot gases to approximately 1208C, a two pressure levels HRSG with a secondary lowpressure (4 bar) steam production circuit was considered. This solution has also the advantageof approaching, in a temperature-heat exchanged diagram, the steam heating curve to theexhaust gases cooling curve thus improving the recovery e�ciency. Unfortunately at themoment there is no low pressure steam request in the factory therefore a feasibility study onpossible uses of low pressure steam pointed out two main necessities of the factory:

. heating of the storehouse area, as seen before for the ICEs cases; In this case there is anexcess of heat available so it is possible to heat the two stockage area for pasta and feeddivision for a approximately 50,000 m3 and an average thermal power of about 3000 kW;

. conditioning of the mill to avoid the thermal di�erences caused by the big quantities of freshair entering the structure required by the compressed air transportation system. The systemconsiders steam at 4 bar for winter heating, the same is used as heat input to an absorptionchiller for summer air conditioning. The chiller would guarantee as well the summer coolingof the ¯ours actually entrusted to traditional compression chillers.

The following section will carry on the economic analysis of the investments for Solution Awith one and two engines and solution C with a 1P level HRSG. For the three cases the lowtemperature optimized recovery was not considered because not explicitly required by thefactory. These solutions are then to be considered as possible improvement and the economicfeasibility may be taken into exam when directly approaching the air conditioning problem.

5. Economic feasibility analysis

The evaluation of an investment in the energy ®eld is usually made in two steps. A ®rstphase using present or not present value methods is made considering absolute indexes thatgive an information regarding the feasibility of the investment. Once the feasibility has been


demonstrated, the second step will arrange the possible investments in terms of theirpro®tability thus considering relative indexes. The methods exposed in Appendix E will beconsidered.The technical solutions seen before will be now examined in terms of Net Present Value

(NPV) and those who have a negative value will be excluded. Finally the pro®tability ofpositive NPV investment will be examined in terms of Internal Rate of Return (IRR).The average electric energy expense of the factory in the biennium 1996±97 is calculated for

the High Use contract and the average monthly power demand rate and energy consumptionin each hour band shown in Fig. 5; regional taxes (UTIF) are as well calculated considering1.35 cent/kWh for the ®rst 200,000 kWh and 0.331 cent/kWh for the subsequent kWhs.Similarly the average natural gas expense was determined considering the monthly average fuelconsumption and demand rate shown in Tables B2 and D1. As stated in a previous section, thecosts for natural gas are varying on a monthly rate in function of statistical and economicalparameters: therefore the average value assumed in the biennium 1996±97 will be used. Costsare divided in proportional costs, ®xed costs, taxes, additional regional taxes plus 258.23 Eur/month for the monthly subscription.VAT is considered throughout the whole economic analysis even if it can be reclaimed

because it does not contribute equally to di�erent categories of costs and therefore it was notalways possible to unbundle it.Table 4 summarizes the energy costs for the factory.

5.1. Solution A, 1 ICE

The initial investment I0 for this solution is determined from the manufacturer o�erconsidering a forfeit amount of 130,000 Euro for the exclusions (mainly civil works, electric,

Table 4Average annual energy expenses

Annual expense (Euro) Average monthly expense (Euro)

Electricity

Energy 1,056,818 88,068Power demand rate 308,666 25,722UTIF 90,994 7583

VAT 10% 14,5648 12,137Total 1,602,125 133,510

Natural gasProportional cost 301,221 25,102Fixed cost 67,324 5610

Taxes 26,093 2174Regional taxes 13,047 1087Subscription 3099 258VAT 10% 41,078 3423

Total 451,862 37,655


mechanical and hydraulic connections, MV auxiliaries), 194,000 (15% of the o�er) for laborand 20% VAT. For the evaluation of the CF a distinction between bene®ts and additionalcosts has to be made. Bene®ts considered are: the reduction in the electricity expenses, theavoided expenses for thermal energy and the facilities provided by SNAM for natural gas usein electricity auto-production (0.31 cent/kWh). Additional costs are due to the natural gassupply for the cogeneration plant (natural gas is di�erently taxed when used for thermal orelectric purposes: the latter is tax free and VAT is 19%. As cogeneration produces the two, thenatural gas consumption has to be split and the part related to electric power is obtainedmultiplying the fuel consumption for a thermodynamic constant depending on the fuel and theeconomic conjuncture (at the moment its value is 0.250 m3/kWh).Costs related to auto-production are: taxes (UTIF with a VAT of 10% for auto-producers

has a slightly di�erent value than for buyers and it is equal to 1.193 cent/kWh for the ®rst200,000 kWhs and 0.24 cent/kWh for the subsequent), maintenance expenses (considering 17.51Euro/h+0.155 cent/kWh for the lubricating oil change every 1000 h) and integration-emergency contract expenses with ENEL (for the integration the best solution was a mediumuse contract with 500 kW demand rate in F1, 800 kW in F2 and F3 and 1000 kW in F4. Forthe assistance or emergency contract the expense is of 1.446 Euro/kW-month for the demandrate plus the eventual consumption 11.58 cent/kWh up to 100 kW and 9.85 cent/kWh over 100kW). Table 5 shows the CF and PP for this caseAssuming that the CF will be the same for the years following the investment, Fig. 11 shows

the NPV of the investment in the short term (7 yr) and long term (12 yr) when the interest ratevaries in the 0±40% range.The investment is convenient for both the long and short term. The IRR is particularly high

(20±26%) allowing besides the remuneration of the initial investment also an interesting pro®t.The economic feasibility is evident and con®rmed by a low PP that guarantees an amortizationin less than 4 yr thus compatible with the economic politics of the ®rm.

Table 5Cash ¯ow and payback period determination for the di�erent solutions. All values are expressed in Euro with the

exception of PP which is expressed in year

Solution No cogen. (A) 1 ICE (A) 2 ICEs (C) 1 GT, 1P (C) 1 GT, 2P (C) 1 GT, 2P+cond.

I0 1,940,000 2,293,000 3,582,000 3,796,000 4,187,700Electricity 1,602,125 206,627 ± 214,924 214,924 214,924

UTIF ± 73,786 77,471 73,604 73,604 73,604Maintenance ± 170,714 190,700 161,023 161,023 161,023SOS contract ± 47,200 25,422 46,332 46,332 46,332

Natural gas 451,862 1,025,996 1,069,517 1,343,712 1,343,712 1,343,712Total 2,053,987 1,524,323 1,363,110 1,839,595 1,839,595 1,839,595Flour refr.� 13,719 ± ± ± ± ±Total 2,067,706 ± ± ± ± 1,839,595

CF ± 529,664 690,877 214,392 214,392 228,111

PP ± 3.66 3.32 16.71 17.71 18.36


5.2. Solution A, 2 ICEs

In this case the evaluation of the I0 is made with the same assumption as above aboutexclusion and labor considering that the manufacturer is the same. Costs and bene®ts are aswell evaluated as before but obviously related to the varied maintenance and assistancecontract necessities. In this case there is no need for an integration therefore there will be noadditional costs for electricity. Table 5 shows the CF and PP while Fig. 11 shows the NPV inthe short and long period.Two engines permit a lower demand rate for an assistance contract thus halving the related

costs with respect to the 1 ICE solution. This positive e�ect added to the savings on electricitycosts, is bigger than the higher expenses for taxes, maintenance and natural gases therefore theCF is higher diminishing the PP of about 12% that is now lower than 3.5 yr. Considering thatthe I0 is about 18% higher, the investment is surely convenient. The NPV in Fig. 11 is higherthan before and the resulting IRR (24±29%) is very appealing.

5.3. Solution C. 1 GT

The I0 for the case study of the GT was considered assuming as exclusion only civil worksthat contribute for about 15,500 Euro to the total; labour and VAT are considered as in theprevious cases. For the 2P HRSG the I0 considers as well an absorption machine and an airtreatment station for the summer conditioning of the mill that contributes, for about 391,700Euro to the total (exclusions and labour included). The integration contract required is of thesame type of case A, 1 ICE and a typical assistance contract as well was considered. Table 5shows the CF and PP.In this case a very I0 combined with the necessity for an integration contract is responsible

of the poor CF and the extremely high PP. Therefore the investment is not convenient, this

Fig. 11. NPV for the three solutions considered.


fact is con®rmed by the negative NPV in both the long and short period as it is shown inFig. 11.

6. Conclusions

The feasibility of a CHP plant for an Italian pasta and animal seed factory was conductedfrom both the technical and economic point of view. The layout analysis focused the threemain division, namely the mill, the pasta and the seed division. For each of them the e�ectiveaverage yearly energy consumption and power demand rate was analyzed in terms of electricityand thermal needs.Following this analysis two possible solutions were pointed out: an ICE based power plant

and a GT based power plant. The ®rst solution gave better characteristic indexes whenevaluated with two ICEs covering the whole electric demand and partially heating the SHwater with the recovered heat.The GT solution penalizes the electric e�ciency for a better thermal availability, not

necessary to the actual needs of the factory. Nevertheless possible solutions to use exceedingheat were proposed. The economic analysis con®rmed the technical results privileging the 2ICEs solutions with an optimal PP and NPV.The GT solution as expected, has a negative NPV and a PP longer than 15 years.Therefore a CHP plant powered by a couple of ICEs providing the energy needs of the

factory and recovering waste heat to integrate traditional boilers duty is not only feasible forthe pasta-seed factory but also very appealing.The project is actually under the study of the factory management.

Acknowledgements

For the help provided in the collection of data the authors would like to thank Mr.Casagrande and Mr. Belia of the feed division, Mr. Morbidini of the pasta division and Mr.Selvatico and Mr. Sportolari for the mill. Finally Mr. De Iuliis collaboration is acknowledgedfor useful discussion and advice.

Appendix A. Electrical costs

Costs are calculated depending on the hour of the day and on the power demand rate of thesupplying contract. Electricity production costs are lower during night hours were the demandis low and it can be covered with high e�ciency base load plants. During peak hours thesurplus in the demand is covered by peak power plants at a lower e�ciency thus rising thecosts. Moreover to a higher power demand rate corresponds a higher ®xed cost for the kWrequired but a smaller variable cost for the kWh consumed. Four categories of contract arepossible depending on the hours per year of consumption and there are as well four hour


bands that regulate the daily drawing of power. They are shown respectively in Tables A1 andA2.Referring to these categories, the price list of energy at 1 January 1998 is shown in Table A3

expressed in Euro (1 Euro=US$0.8)

Appendix B. Energy balance

The energy requirements of the drying section in the pasta division were determined throughan energy and mass balance of the three main thermal users, namely sifter, wrapper and dryer.The data utilized in this analysis are shown in Table B1. Further assumptions are a DT of128C for superheated water in the boiler (1088C in, 1208C out), 15% losses and a boilere�ciency of 0.9. This rather high e�ciency is justi®ed because the boiler is relatively new andthe fuel is clean being natural gas.A speci®c heat consumption and superheated water necessity for kg of pasta was then

obtained and the results are presented in Table B2.

Appendix C. CHP and characteristic indexes

The most common cogenerative modules are based on:

. Internal Combustion Engines for power output from 10 kW to 10 MWe with electrice�ciency ranging from 30% up to 40% and a global e�ciency (thermal+electric) reaching80±90% of input energy; thermal energy is available at two di�erent temperatures: about

Table A1Possible contracts with ENEL depending on consumption

Low use < 1043 h/yrMedium use 1043±3513 h/yr

High use 3513±5946 h/yrVery high use > 5946 h/yr

Table A2Hour bands

October to MarchF1 (peak hours) 8.30±10.30 16.30±18.30F2 (high load hours) 6.30±8.30 10.30±16.30 18.30±21.30

F3 (medium load hours) ±F4 (empty hours) 21.30±6.30 (Mon to Fri) 1±24 (Sat and Sun)

April to SeptemberF1 (peak hours) ±

F2 (high load hours) 8.30±12F3 (medium load hours) 6.30±8.30 12±21.30F4 (empty hours) 21.30±6.30 (Mon to Fri) 1±24 (Sat and Sun and August)


Table A3Price list of energy from ENEL at January 1st 1998

Hour bands Low use Medium use High use Very high use

Power Euro/kW-month Euro/kWh Euro/kW-month Euro/kWh Euro/kW-month Euro/kWh Euro/kW-month Euro/kWh

F1 < 3000 kW 3.248 0.144 5.433 0.107 7.835 0.092 12.452 0.0693000±10.000 kW 1.895 0.144 4.080 0.107 6.481 0.092 11.104 0.069

> 10.000 kW 1.518 0.144 2.691 0.107 5.097 0.092 9.709 0.069F2 1.141 0.126 2.071 0.085 3.925 0.072 7.964 0.050F3 0.392 0.078 0.728 0.058 1.425 0.050 2.913 0.041

F4 0.108 0.041 0.186 0.039 0.382 0.037 0.873 0.035

F.Fantozzi

etal./

Applied

Therm

alEngineerin

g20(2000)993±1017

1014

4008C from the exhaust gases and about 908C from the engine jacket and lubricating oilcooling systems;

. Gas Turbines (GT) for power output from 2 to 200 MW with electric e�ciency ranging from15 to 40%. Heat is mainly recovered from exhaust gases thus all available at hightemperature, about 5008C.

. Steam Turbines for power outputs from 10 MW with electric e�ciency ranging from 20 to40% are used for big installations. Heat is recovered at high temperature (steam extraction)or low temperature (back pressure).

The performances of a cogeneration unit are usually described in terms of characteristicindexes [7±9]:

Electric Index E:I: � We

Qr

;

Ratio of electric power produced to heat power recovered.

Cogeneration Efficiency U � We �Qr

Qf

;

ratio of power produced to power introduced with the fuel. It is de®ned as a ®rst principlee�ciency although it sums di�erent forms of energy (thermal and electric). Distinguishing thetwo di�erent forms of energy:

Electric Efficiency Ue � We

Qf

Thermal Efficiency Ut � Qt

Qf

:

Appendix D. Engine operational parameters

The operational parameters of the ICEs are guaranteed by the manufacturer considering a

Table B1Data used in the determination of the pasta division's thermal loads

Ta,A Fa,A Xa,A Tp,E Fp,E Ta,S Fa,S Xa,S Ta,W Fa,W Xa,W Ta,D Fa,D Xa,D Fp,D

30 65% 0.017 40 30 65 25 0.045 80 50 0.2 70 70 0.185 12

Table B2Thermal loads of the pasta division

Sieving Wrapping Drying

Heat (W/kgp) 90 102 31SH water (kg/kgp) 6.5 7.5 2.5


full power run for an entering LHV of 9.5 kWh/Sm3 are shown in Table D1. Technical andeconomic data on the GT are taken from a model amongst the ones available on the market.

Appendix E. Economic analysis methodology

E.1. Payback method

It is a simple method that gives an immediate evaluation of the feasibility although it doesnot consider the savings in the years subsequent to the amortisement period penalizing longterms investment. It does not consider as well eventual changes in the prices of energy and inthe value of money with time. It may be used as a parameter to quantify the period ofexposition to risk but it is good practice to associate it to other indexes.It is possible to calculate the PP (Payback Period) as:

PP � I0CF

;

In this analysis, bene®ts are represented by the savings in energy costs and they are referredto the year of the investment as well as the costs are.

E.2. Net present value (NPV)

It is based on the knowledge of the Discounted Cash Flow (DCF) de®ned as the cash ¯owreferred to each year taking into account a discount factor that depends on the interest rate i.

NPV �Xnj�1

DCFj ÿ I0 DCFj � CFj�1� i�ÿj

I0 is as before the initial investment while n is the years of life of the investment. Aninvestment with a positive NPV means that a return of the investment is possible within the

Table D1Engines technical data

1 ICE 2 ICEs 1 GT

Inlet air temperature (8C) 15 15 15Electric power output (kW) 2720 2930 2670Electric e�ciency (%) 42.0 41.5 24.3

Speci®c consumption (kWh/kWh) 2.38 2.41 4.11Exhaust temperature (8C) 400 410 570Thermal power cooling exhaust gases to 1208C (kW) 1290 1570 6460

Jacket water temperature (8C) 95 95 ±Thermal power at low temperature (kW) 1484 1510 ±Thermal e�ciency (%) 42.8 43.6 58.7Cogeneration e�ciency 84.8% 85.1% 83.0%


life of the investment. A negative NPV means that the investment is not able to pay itself. Thismethod simply eliminates negative NPV investments. In case of two investments with the sameI0 the one with a higher NPV is preferred. In case of di�erent I0 the choice prefers theinvestments with a higher Rate of Return (RR) or ratioNPV/I0.

E.3. Internal rate of return (IRR)

It is de®ned as the rate of interest that makes incomes equal to outlays thus bringing to zerothe NPV. Therefore is a relative index typical of every investment It may be seen as the intereston the residual part of the capital that is still invested when the investment has been amortized.To decide if the investment is to be made, the IRR has to be bigger or at least equal to theMARR (Minimum Attractive Rate of Return) that is the rate of interest on an investment thatthe ®rm may always obtain from the market. Each time capital is used for an investment theopportunity to earn the MARR is lost.

References

[1] J.A. Perrella Balestieri, Cogeneration in the food/beverage Sector: Brazilian and American experiences, in: Proc.

Flowers '97, 1997, pp. 533±540.[2] A. Ficarella, A.D. Laforgia, Operating experiences, on site performances and thermoeconomic analysis of a 5

MW combined cycle plant in the agro-food industry, in: Proc. Ecos '96, 1996, pp. 105±112.

[3] G. Bidini, U. Desideri, S. Saetta, P. Proietti Bocchini, Internal combustion engine combined heat and powerplants: case study of the University of Perugia power plant, Applied Thermal Engineering 18 (6) (1998) 401±412.

[4] G. Nicolai, Conservazione e trasformazione degli alimenti, Hoepli, Milano, 1985 (In Italian).[5] C. Aghina, S. Maletto, Tecnica Mangimistica; SocietaÁ Editrice Esculapio, Bologna (In Italian).

[6] L. Milatovich, G. Mondelli, La tecnologia della pasta alimentare, Chiriotti Editori, Pinerolo (TO), 1990 (InItalian).

[7] F.F. Huang, Performance Assessment Parameters of a Cogeneration System, in: Proc. Ecos '96, 1996, pp. 225±

229.[8] J.H. Horlock, Cogeneration-Combined Heat and Power, Pergamon Press, Oxford, 1987.[9] S.S. Stecco, G. Manfrida. A rational e�ciency analysis of comparison and trends in gas turbines for

cogeneration, (1985), ASME Paper 85-GT-13.


study of a cogeneration plant for agro-food industry

Documents