Applied Energy 156 (2015) 374–389

Contents lists available at ScienceDirect

Applied Energy

journal homepage: www.elsevier .com/ locate/apenergy

Heat recovery using heat pumps in non-energy intensive industry: AreEnergy Saving Certificates a solution for the food and drink industryin France?

http://dx.doi.org/10.1016/j.apenergy.2015.07.0480306-2619/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Fax: +33 4 97 15 70 71.E-mail address: gondia.seck@mines-paris.org (G.S. Seck).

1 NACE: Statistical Classification of Economic Activities in the Europeannity. The 4-digit level is the most disaggregated level.

Gondia Sokhna Seck ⇑, Gilles Guerassimoff, Nadia MaïziCentre of Applied Mathematics, Mines ParisTech, Rue Claude Daunesse CS 10207, 06904 Sophia Antipolis Cedex, France

h i g h l i g h t s

� First non-energy intensive bottom-up analysis for heat recovery with heat pump.� Decrease of 12% and 9% of energy consumption and emissions in 2020 compared to 1990.� Almost 40% additional energy savings could be achieved with incentive policies.� A drop of 15% of the gas price could stop the heat pump deployment.� With higher gas prices we could observe heat pump deployment or inter-energy substitutions.

a r t i c l e i n f o

Article history:Received 3 February 2015Received in revised form 25 June 2015Accepted 15 July 2015Available online 25 July 2015

Keywords:Non-energy intensive industryEnergy Savings CertificateHeat recoveryHeat pumpsBottom-up modelingFood & drink industry

a b s t r a c t

Saving energy is crucial for all sectors following the new framework presented by the EuropeanCommission to drive continued progress towards a low-carbon economy. Many studies focus on theresidential sector, transport and energy-intensive industries, but there is a lack of tools to help decisionmakers in non-energy intensive industry (NEI). This paper presents the first bottom-up energy modeldeveloped for this sector. This prospective modeling enables us to analyze the impact of heat recoveryusing heat pumps (HP) in industrial processes up to 2020 in the French food and drink industry (F&D),the biggest NEI sector. The technology has high potential in this sector and may be eligible for EnergySaving Certificates. Our model determines the differentiated cost for energy savings in response to incen-tive policies under the ESC mechanism at a 4-digit level of NACE classification. Sensitivity analyses alsoshow how gas prices and electricity carbon footprints impact on HP penetration. Our study of this par-ticular sector shows that the model could be a useful decision-making tool for assessing potential energysavings and could be extended to other sectors of NEI industry for more efficient subsectoral screening.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

This paper focuses on the analysis of non energy intensive (NEI)industries, where, contrary to Energy Intensive (EI) industries,there is a lack of scientific energy analysis work despite the signif-icant number of publications and books about energy.

A survey of the topic highlights that the share of NEIs hasincreased in final energy consumption and industrial added value(80% of the industrial added value) [1,2], making them a prioritytarget for energy suppliers in France. Energy suppliers like EDF(Electricité de France) have an obligation to achieve energy savingsover a given period in France to avoid penalties according to the

Energy Savings Certificates (ESC) or ‘‘white certificates’’ mecha-nism established by the program law of July 13th 2005, which laysdown policy guidelines on energy. Since the ESC cannot be appliedto sectors subject to the national quota allocation (PNAQ inFrench), which are generally EIs, NEIs are expected to be moreimportant in the drop of the industrial energy intensity toavoid double counting with the European Emission TradingScheme (EU ETS). Thus, the promotion of a bottom-up energymodel (technology rich base) for non-energy intensive industriesat a very detailed disaggregation (4-digit level of NACE1 classifica-tion) could be a useful decision-making tool for energy operators.The heterogeneity and the disparity of the NEI explained the scarcity

Commu-

http://crossmark.crossref.org/dialog/?doi=10.1016/j.apenergy.2015.07.048&domain=pdfhttp://dx.doi.org/10.1016/j.apenergy.2015.07.048mailto:gondia.seck@mines-paris.orghttp://dx.doi.org/10.1016/j.apenergy.2015.07.048http://www.sciencedirect.com/science/journal/03062619http://www.elsevier.com/locate/apenergy

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 375

of models in this group of sectors. Then, the development of this firstbottom-up energy model, by using TIMES model (MARKAL familymodel), will allow assessing the energy efficiency potential of thenon-energy intensive industry. In this article, to follow on a previouswork on heat recovery by HP [3], we focus on an economic analysisby applying incentive policies to achieve the maximal potentialdepending on the penalties of the ESC mechanism. This analysiscould be helpful for the strategic department of energy operatorsto know at which level to invest in the industry and promote effi-cient technologies. Sensitivity analysis will also be done on theimpact of gas prices or electricity carbon footprint. So, this paper isbroken down as follows: Section 2 succinctly describes, as a remin-der to set the context, the structure and some assumptions of the NEIbottom-up model within TIMES with a study case assessment of heatrecovery using heat pumps in the food & drink industry up to 2020[3]; Section 3 is subdivided into two parts, the first of which analyzesthe impact on energy savings of incentive policies using ESCmechanisms, and the second providing a sensitivity analysis on theevolution of HP penetration in food & drink, with the natural gasprice and the electricity sector’s carbon footprint.

2. Methods

2.1. Non-energy intensive ‘‘bottom-up’’ model

We have developed the first bottom-up model for an energyprospective exercise for non-energy intensive industry at the4-digit level of NACE classification for future energy planning,subject to different constraints (energy price evolution, energyefficiency policies, environmental constraints, etc.) up to 2020using the TIMES framework [4–15] (Fig. 1). However, contrary toour previous article [3] based on energy prospective with technol-ogy spreads; we analyze these results in an economic point of viewwith incentives for policy makers. This is one of the strength ofprospective models based on an optimization paradigm.

TIMES2 is a ‘‘Bottom up’’ techno-economic model which providesa technology rich basis for estimating energy dynamics over amulti-period time horizon. It is based on a Reference EnergySystem (RES) which is a network describing the flow of commoditiesthrough various and numerous processes. The objective function isthe criterion that is minimized by the TIMES model. It representsthe total discounted cost of the system over the selected planninghorizon. The components of the cost of the system are expressedin each year of the study horizon (and even for some years off hori-zon) in contrast to the constraints and variables that are related toperiod. This choice allows a more realistic representation ofpayments flows performed in the energy system.

Each year, the total cost includes the following elements:

– The investment costs incurred for investing into processes.– Fixed and variable annual costs.– Costs incurred for exogenous imports.– Revenues from exogenous exports.– Delivery costs for required commodities consumed by

processes.– Taxes and subsidies associated with commodity flows and pro-

cess activities or investments.

The representation of the technologies calls for precise knowl-edge of the industrial installations in each sector. The existing

2 The Integrated Markal-Efom System was developed by the Energy TechnologySystems Analysis Programme (ETSAP), an implementing agreement under the aegis ofthe International Energy Agency (IEA), in 1997 as the successor of the formergenerators MARKAL and EFOM with new features for understanding and greaterflexibility.

and future technologies in the sectors over a given time horizonare considered with techno-economic parameters (capacity,energy intensity, efficiency, availability factor, investment costs,fixed and variable costs, economic and technical life, etc.) and theirrelated strategic orientation parameters (taxes, subsidies, etc.).

In the NEI model, the CO2 emissions of a fossil fuel f could bealso calculated using the following equation

Emi fCO2 ¼ ECf � EFf

where ECf: the quantity of fossil fuel f consumed by the sources ofcombustion, EFf: The emission factor of the fossil fuel f.

The total CO2 emissions could be deduced from the addition ofall fossil fuel emissions. Emission factors (EF) used in this modelare the specific emissions given by the ADEME (FrenchEnvironment and Energy Management Agency) on the verificationand quantification of emissions reported under the exchangesystem quotas of emissions of greenhouse gases in its guide ofemission factors [16].

Fossil fuel

EF (tCO2/MW h)

Coal

0.343

Heavy fuel oil (FOL)

0.282

Light fuel oil (FOD)

0.271

Natural gas

0.206

LPG

0.231

For an impact more adapted to global environmental con-straints, CO2 emissions allocated to the production of electricityhave been taken into account in the model in order to observethe impact of the power generation energy mix. In France, accord-ing to ADEME/EDF R&D [16], it is estimated at around55g CO2/kW h in the industry. These emissions are not added inthe global calculation of the emissions from industry, they willonly account for the effect of a tax on CO2.

The RES will allow us to obtain the optimal technology pathaccording to an energy/environmental scenario, such as thepotential for industrial energy efficiency and CO2 emissionsreduction in relation to France’s commitments to meet the EUclimate change targets across sectors, based on least cost criteria.It is a partial equilibrium model because it provides no feedbackon sector changes in other economies. However these impactsare of secondary importance in most developed economies likeFrance [17].

Several methods/tools were indeed developed for the energyefficiency of industrial sector. However, these methodologies aredone in an isolated depiction of specific processes in a specificindustrial manufacture which do not facilitate the implementationof a coherent global frame that takes into account all interactionsbetween the decisions of economic, public and private agents[18]. Moreover, in the case of the NEI industry, it is more relevantto generalize then pass the decision to individuals, in which casethis bottom-up model for an energy prospective analysis couldprovide individual decision-making support. The question is there-fore not to ‘‘think individually to act globally’’ but rather to take theinitiative ‘‘to think globally with a view to acting individually’’.Then, unlike individualized studies on specific processes in a speci-fic manufacture, it would represent the best way to motivateindustrialists with very low energy costs in the case ofnon-energy intensive industries (one of their principle characteris-tics) to make energy savings [3].

Furthermore, the methodologies of energy modeling used forenergy intensive sectors are not suitable for NEI. The energydescription of Energy intensive industry (e.g. Iron and steel, pulpand paper, etc.) is done through the main manufactured product,

Fig. 1. The TIMES model reference energy system (IER).

376 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

where each sector is then described by a process flow sheet asso-ciated with energy consumption ratios by process step (iron andsteel, paper and allied products, glass and glass products, etc.)[18,19]. However, in the case of the non-energy intensive industrysuch as food and drink industry, a different approach is requiredbecause of the diversity and the large number of operations and,end products that necessarily have no physical links [3]. Thisexplains our choice to develop a modeling approach by energyend-uses (e.g. drying, heat treatment, etc).

The advantages of this modeling approach are, firstly, its perfectdepiction of the considerable number of operations to enable adetailed, reliable analysis of non-energy-intensive industry pro-cesses. Secondly, unlike energy-intensive models, the NEI modelthat we have developed is generic and therefore common to allnon-energy intensive industries, bringing a clear advantage interms of time cost modeling (conciliation of data availability,disaggregation level and modeling effort).

As we already know, the proportion of thermal purposes (boil-ers, thermal end-uses) is around 60% of the final energy consump-tion in NEI industry (75% in industries) due to mechanicalprocesses (motive power, compressed air, pumping, ventilation)and cooling production. According to the literature, around onethird of the final energy used for thermal purposes is then wastedthrough losses [20,21]. Then, a rationalization of the thermalend-uses should be searched out first by saving, recovering andusing this waste heat released in the form of liquid/gas effluentsbefore optimizing heat production processes. A possibility torecover the lowest temperature could be done technically throughheat pumps (Fig. 2), even if scarcely ever is the case today.

We therefore applied this model to analyze a case study involv-ing heat recovery using new heat pump (HP) technologies in a keyarea of the NEI industry, i.e. the food and drink sector (F&D). Thiscase allowed us to explore some of the multiple potentialities thatour model offers. The choice of food and drink was imposedbecause of its economic and energy weight in non-energy intensiveindustry (nearly a third of the total energy consumption for aroundone fifth of the added value) as well as in industry in general (15%energy for 14% added value) [22–26].

The technological progress of the heat pump could allow thesubstitution of the boilers which consume fossil fuel for the

production of useful heat at higher temperatures [27,28]. Basedon the detailed survey on the French food & drink, where theCEREN subdivided the heat consumption for each end-use processinto eight temperature ranges between 60 �C and 200 �C, i.e. four10 �C ranges between 60 �C and 100 �C, two 20 �C ranges between100 and 140 �C, the 140 and 200 �C range and the last range called‘‘other’’ [22]. This disaggregation relies on a data exploitation of4 years of French industrial survey by the CEREN (Centred’Etudes et de Recherches Economiques sur l’Energie). Theydefined a sampling mode which was realized per subsector andper energy consumption level to get the best representativenessof each industrial sector. In Fig. 3, the heat consumption in theFrench food and drink is depicted by temperature range and it isnoticed that around 55% is beyond 100 �C. According to theADEME (French Environment and Energy Management Agency),in France, it is around 15% of the final energy consumption in thissector which is wasted through heat losses if we only consideredbeyond 100 �C [29]. In the case of Germany, IHPs (output temper-ature 100 �C) could provide around 16% of the total energy con-sumption of German industry, however up to now just a verysmall number of this potential is reached by them [30]. For devel-opment of sustainable energy, three important technologicalchanges have been required: energy economies on the demandside, efficiency improvements in the energy production, andrenewing of fossil fuels by various sources of renewable energy.In this regard, industrial heat pump (IHP) systems improve energyefficiency and cause less fossil fuel consumption. Thus, the heatpump technology could contribute significantly in the energydemand management and CO2 emissions in France. This is espe-cially true in the case of France due to the large share of nuclearin the power generation sector. Moreover, the energy suppliersare interested in the white certificates in the industry, in the NEIin particular to avoid double counting with the PNAQ.

However, despite these benefits, according to the market over-view based on 9 countries (Austria, Canada, Denmark, France,Germany, Japan, Korea, The Netherlands and Sweden), the penetra-tion of IHP is limited or non-existent due to their low economicadvantage, lack of knowledge (industry doesn’t know HP as wellas boilers) and experience [31,32]. The growing costs of fossil fueland more restricted environment regulations lead industry to

Pâte et PapierDrying

Ciment, ChauxPlâtre

Coolingproduction

Tuiles, Briques,CéramiqueMechanicaloperations

SidérurgieHeating Liquid

& Gas

VerreOthers End-uses

End-uses

Demand

ChaudièreBoiler

TurbineTurbine

CogénérationCHP

Transformationprocesses

HTH

ELCHTH

ELC

Heat PumpElectricity

HTHImp: ImportationHTH: HeatELC: ElectricityFOD: Light fuel oilFOL: Heavy fuel oil

Natural Gas

ImpCharbonImp Coal

ImpGaz NatImp Nat Gas

ImpÉlectricitéImp ELC

ImpFioulImp FOL

Electricity

Heavy fuel oil

Coal

Resources

ImpBiomasseImp FODLight fuel oil

Industrial Sector:

Food &Drink(35 sectors)

Recovered Heat(low temperature)

Fig. 2. Reference energy system for French food and drink in the case of heat recovery analysis using heat pump systems.

0

1

2

3

4

5

6

Food & Drink

Hea

t Dem

and

(TW

h/yr

)

60-70°C 70-80°C 80-90°C 90-100°C 100-120°C 120-140°C 140-200°C Other

0

1

2

3

4

5

660-70°C 70-80°C 80-90°C 90-100°C 100-120°C 120-140°C 140-200°C Other

0%10%20%30%40%50%60%70%80%90%

100%

Hea

t dem

and

60-70°C 70-80°C 80-90°C 90-100°C 100-120°C 120-140°C 140-200°C Other

Hea

t Dem

and

(TW

h/yr

)

Fig. 3. Heat needs in food and drink.

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 377

discover again the energy efficiency potential of the heat pumps.The possibility of replacing boilers, which are a conventional heatsource, with heat pumps has increased in the industrial field.This interest of IHPs with their integration into process and exist-ing installations has been documented in the literature [33–45].The application of IHP for industrial waste heat utilization in thefood and drink is reasonable up to 100 �C or even above depending

on the heat pump technology applied [46]. In order to determinethe realizable potential for the application of IHPs, it should bementioned, that each industrial sector and each production processitself have to be regarded in detail.

This article will give us at which level they could be adoptedaccording to their economic competitiveness and the response toincentive policies.

3 AGRESTE (French Agricultural Statistics Agency of the Ministry of Agriculture,Agrifood and Forestry).

378 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

2.2. Scenarios and sectorial assumptions

We derived our results from this two following scenarios:

– BAU, ‘‘Business As Usual’’, It is the reference scenario, in whichno new policy or measure is considered necessary or adopted.The past trends will keep going in the future and moreover, thisscenario will be used as a baseline to assess the impact of a newpolicy or measure [47].

– Sc_HP is characterized by the implementation of heat pumps inthe food-processing industry until 2020. This scenario willallow us to monitor and estimate the potential of heat recoveryby using a heat pump in non-energy intensive industry.

For the sectoral demands in our analysis, we considered theassumptions below for their evolution in the model (Fig. 4).

Due to a decrease in the meat consumption, we assume anincrease by only 0.15% per year for the processing and preservingof meat and poultry meat. This assumption could be explainedby the decrease in the number of slaughterhouses and eatinghabits changes (sensitivity to hygiene and food safety guarantees,price, French origin, animal welfare conditions). In the case of theproduction of meat and poultry meat products, a decline isassumed by 0.25%/yr for the same reason (drop of the companies’number) [48,49].

With stringent international regulations and Europeanregulatory due to the limitation of natural resources, the fish sectorshould expect an increase by 0.6% per year, slower than in the lastten years which was around the double [50,51].

The growth rate in the sector of fruits and vegetables isassumed to be +2% per year if we consider it should be driven bynutritional reasons. In a like manner, the manufacture ofvegetables and animal oils and fats should follow the sametrend (+1.5% per year) as in the past due to the promotion ofbiodiesel in France [51]. At the contrary, the manufacture ofdairy products is expected to be stable with +0.3% per year until2020.

The evolution of the starch products sector will follow the sametrend as the paper industry. The growth rate should be almost+1%/yr after a sharp decline by 11% because of the lower demandfrom pulp and paper manufacturers with the economic crisis[19,52,53].

In the manufacture of sugar, the evolution should be stable dueto the significant trend for restructuring in recent years, and thenaccelerate particularly since 2006 in the framework of the reformpromoted by the Community institutions to reduce the Europeansugar production and achieve market equilibrium from 2010[54–56]. The chocolate sector should recover after the crisis if weassume that the creativity and innovation of the manufacturersin a highly competitive industry should greatly boost sales. Weexpect a growth rate for wine production of 3%/yr and alcoholicbeverages of 1.2%/yr in the beverage sector according to theincrease of champagne and cognac exports to the United States,or distilled alcoholic beverages and wines to China. Other sectorssuch as malting and brewing are expected to continue their declineof recent years due to a general downward trend in beer produc-tion and consumption [57]. Other industries, including health foodproducts, animal feed sector, manufacture of bread, pastry goodsand cakes should expect a growth rate by almost 1% per yearreflecting respectively general concern for health, decrease in thecanine population or according to last years.

2.3. Energy price and technology assumptions

Our energy price scenarios (Table 1 and Fig. 5) are based onprices observed historically in the non-energy intensive industry

from 1993 to 2010 from AGRESTE3 and price projections fromIEA’s energy price projections [58]. In the case of electricity weexpect a growth rate of +1.3% per year until 2020, as advised byEDF subsector experts. These prices are given all taxes included i.e.the amount paid per unit of energy by a manufacturer in a givensector.

According to expert of the European Centre & Laboratories forEnergy Efficiency Research (ECLEER), the temperature range ofthe heat recovered in thermal uses is between 30 �C and 60 �C inthe food industry [59]. We considered in Table 2 these assumptionsabout the IHP’s investment cost for different heat temperaturerange from EDF R&D’s experts.

We consider as future technology heat pump which couldoperate at high temperatures (100 �C) or very high temperatures(up to 140 �C). Indeed, these technologies are currently beingdeveloped and tested by EDF laboratories in partnership with othercompanies [60].

3. Results

The results show that the implementation of heat pumps in thefood & drink sector achieves a final energy consumption reductionof about 12% by 2020 compared to 1990 levels, which representsaround 8.1 TW h in total (13% compared to 2001), with a CO2 emis-sions reduction of about 9% (around 2 Mt of emissions) (Fig. 6).

In Fig. 7 we could track and target at a highly disaggregatedlevel, through the results of the model, the cost-effective tempera-ture ranges which allow achieving these energy savings. The intro-duction of heat pumps into French F&D is considered to be easywhen the temperature required in process is below 100 �C.Around 12% of the final energy consumption is achievable econom-ically by IHP by 2020 and the most representative energy end-usesin terms of heat recovery are the cooling production (45%), Liquidand gas heating (13.5%), mechanical processes (10.5%) and drying(9.6%). The results clearly demonstrate the relevance of the optimaltechnological paths available in the model according to environ-mental and/or energy constraints [3].

However, the implementation of these technologies in food anddrink did not achieve the maximum potential. Thus, we aim toillustrate the potential for achieving additional energy savingsthrough incentive policies using ESC mechanisms. Then we carryout a sensitivity analysis of our results relative to the natural gasprice and the electricity sector’s carbon footprint.

4. Discussion

4.1. Analysis of incentive policies for improving energy savings

We can distinguish two main types of opportunity to improveenergy efficiency in non-energy intensive sectors:

– Those that induce modest gains for a limited investment (quickwins) as mentioned above, using heat pumps below 100 �C inthe food & drink industry.

– And those that require greater investments to achieve higherenergy savings.

To induce the penetration of high temperature heat pumps inthe food & drink industry, we planned to apply an incentive policywithin the mechanism of the Energy Savings Certificate or ESC(also called White Certificate). This analysis highlights the probable

0

20

40

60

80

100

120

140

160

180

200

2000 2005 2010 2015 2020

Val

ue a

dded

(Ind

ex 2

000=

100)

Man. of meat and poultrymeat

Man. of meat products

Fish sector

Man. of Fruits & vegetables

Man. of Oils & fats

Man. of Dairy products

Man. of Grain mill products

0

20

40

60

80

100

120

140

160

180

200

2000 2005 2010 2015 2020

Val

ue a

dded

(Ind

ex 2

000=

100)

Man. of Starches

Man. of animals feed

Man. of Bread & pastry goods

Man. of Sugar

Man. of cocoa

Man. of farinaceous prod.

Man. of tea & coffee

Man. of other food prod.

0

20

40

60

80

100

120

140

160

180

200

2000 2005 2010 2015 2020

Val

ue a

dded

(Ind

ex 2

000=

100) Man. of other food prod.

Man. of alcoholic beverages

Man. of wines and watersdrinkMan. of cider

Man. of beer

Man. of malt

Fig. 4. The evolution of the demand (value added) in the 21 sub-sectors of food and drink [19,48–57].

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 379

additional savings using this mechanism and its inherent costs forthe food and drink sector.

The increase in energy prices due to depleting fossil fuels andenvironmental impacts inevitably leads to industrialists’ involve-ment in energy demand management policies [61].

It was in this context that the Energy Savings Certificate (ESC orWhite certificates) system was established in France by the law ofJuly 13, 2005 [62] which laid down policy and energy guidelinesbased on an obligation to achieve energy savings imposed onenergy suppliers (also called the ‘‘obligated actors’’) over a givenperiod. Non-energy intensive industries thus became a prioritytarget for energy operators because, to avoid double counting, sitessubject to the national quota allocation plan (PNAQ in French)cannot attain Energy Savings Certificates. A threshold was set at54 TW h for the first three-year period from July 1, 2006 to June30, 2009 and distributed among the energy suppliers. EDF andGDF (Gaz de France, now GDF Suez) (30 TW h and 13.4 TW hrespectively) represent nearly 80% of energy savings targetsbecause they are proportional to energy sales [63,64]. Failure toachieve these objectives incurs a penalty of € 20/MW h cumac4 atthe end of three years. For this first period, the goal was largelyachieved (about 65 TW h). A new target, six times higher than duringthe first period, was set for the second three-year period (January 1,2011 – December 31, 2013) [65]. However, unlike for the first period,the list of eligible agents and individual decrees required for energysavings will only be made known at the end of the three years.

Power suppliers are given the freedom and creativity to choosewhat action to undertake. They can encourage their clients to saveenergy by providing information including financial incentives inconnection with industrials like NEIs (where the energy costs are

4 MW h Cumac (MW h Cumulé Actualisé): cumulative MegaWatt-hours (MW h)updated.

insignificant in the production value, less than 3%). Thus, our anal-ysis relies on this framework by examining what impact an incen-tive to use heat pumps will have on energy savings in non-energyintensive sectors such as the food and drink industry (F&D). Thisanalysis allows us to observe the potential additional savings andthe related additional costs. This evaluation could be used as a ref-erence on the penalty level and thus, could be an importantdecision-making tool for eligible agents during this new triennialperiod of ESC measures.

Fig. 8 shows that we could make maximum additional energysavings of around 3.1 TW h in 2020 compared to a scenario with-out incentives in F&D. It establishes that this maximum is reachedwith an incentive representing 70% of the investment cost of HP. Itis worth noting that up to an incentive of 10%, the additionalenergy savings are not significant.

The evolution of this potential allows us to deduce from ourmodel the cost for each additional energy savings level.

Thus, as shown in Fig. 9, with incentives of 20% and 30%of the investment cost resulting in around 265 GW h and420 GW h respectively of additional energy savings, the costsper MW h cumac remain high and are in the range of€ 90–95/MW h cumac.

We have to reach an incentive of 40% to achieve a minimal costof € 42/MW h cumac, which corresponds to an additional potentialof 1.44 TW h in energy savings. This lower cost of additional energysavings is due to the steep slope from an incentive of 30%. At 50%,the cost increases (€ 49/MW h cumac) because of the reduction ofthe additional energy savings made compared to a 40% level.Thereafter, the cost of € 46.75/MW h cumac is reached for the max-imum additional potential of 3.1 TW h at an incentive of 70%. Thus,the cost increase for the maximum additional energy savings hasalready been reached. This vision of aggregate F&D costs thereforeshows that the minimum attainable cost would be about

Table 1Energy price projections by sector (constant price €2005).

2000 2003 2005 2008 2009 2010 2015 2020

Remaining food and drink sectors (€2005/kW h)Electricity 0.049 0.045 0.050 0.054 0.057 0.056 0.060 0.064Light fuel oil 0.037 0.029 0.045 0.062 0.042 0.044 0.053 0.058Heavy fuel oil 0.019 0.019 0.024 0.035 0.028 0.030 0.040 0.045Natural gas 0.018 0.017 0.021 0.031 0.029 0.030 0.041 0.047LPG 0.034 0.031 0.040 0.046 0.039 0.040 0.047 0.050Coal 0.009 0.008 0.011 0.014 0.012 0.012 0.012 0.013Steam 0.016 0.015 0.020 0.032 0.023 0.024 0.029 0.032

2000 2003 2005 2008 2009 2010 2015 2020

Manufacture of dairy products (€2005/kW h)Electricity 0.047 0.043 0.049 0.056 0.057 0.057 0.061 0.065Light fuel oil 0.036 0.029 0.045 0.061 0.042 0.044 0.053 0.058Heavy fuel oil 0.019 0.019 0.024 0.035 0.027 0.028 0.038 0.043Natural gas 0.018 0.018 0.021 0.030 0.028 0.030 0.041 0.047LPG 0.032 0.027 0.038 0.046 0.035 0.036 0.041 0.044Steam 0.015 0.015 0.021 0.031 0.025 0.027 0.039 0.046

2000 2003 2005 2008 2009 2010 2015 2020

Manufacture of sugar (€2005/kW h)Electricity 0.057 0.053 0.057 0.064 0.060 0.060 0.063 0.067Light fuel oil 0.038 0.029 0.047 0.059 0.043 0.045 0.054 0.058Heavy fuel oil 0.019 0.018 0.025 0.038 0.028 0.030 0.040 0.046Natural gas 0.016 0.017 0.023 0.032 0.026 0.028 0.038 0.043LPG 0.034 0.030 0.040 0.048 0.041 0.042 0.049 0.052Coal 0.009 0.009 0.013 0.018 0.016 0.016 0.016 0.017Coke 0.017 0.020 0.032 0.031 0.023 0.023 0.024 0.025

380 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

€ 42/MW h cumac more than twice the penalty, which correspondsto 1.5 TW h in energy savings. This implies that ultimately, theadditional proven potential is not financially worthwhile.

The evolution in the cost of one avoided ton of CO2 subject tothe level of HP investment incentives is shown in Fig. 10. Thisallows a comparison with changes in CO2 taxes. Indeed, we findthat incentives between 35% and 40% correspond to a furtherreduction of costs per ton of CO2 in the range of € 90–110/tCO2.Industries that do not come under the EU ETS could be subject tospecific measures, whereby the EU would set goals with a choice

0

50

100

150

200

250

300

2000 2003 2005 2007 2009 2010 2015 2020

Ene

rgy

pric

e (in

dex

2000

=100

)

Manufacture of dairy products

Electricity

Natural gasLPGSteam

Light fuel oilHeavy fuel oil

0

50

100

150

200

250

300

2000 2003 2005 2007 2009

Ene

rgy

pric

e (in

dex

2000

=100

)

Remaining food and drink

Fig. 5. The evolution of the energ

of tools for each individual state [66]. These industries includealmost all sectors of NEI industry in France that are regularly asso-ciated with a ‘‘CO2 tax’’ quoted on European or national level. Thus,this analysis shows that higher levels of taxes will be needed toachieve more substantial energy savings. And very high taxes, evenif they encourage efforts, may lead to competition distortionsbetween countries that do and do not apply them, as well as com-petition within the NEI industry in France. The model could thusprovide an opportunity to highlight research that could lead toan incremental adjustment of the tax by the regulatory authorities

0

50

100

150

200

250

300

EN

ergy

pri

ce (i

ndex

200

0=10

0)

Manufacture of sugar

ElectricityDomestic fioul

Natural gasLPGCoalCOKEH

Heavy fuel oil

2010 2015 2020

sectors

Electricity

Natural gasLPGCoalSteam

Light fuel oilHeavy fuel oil

2000 2003 2005 2007 2009 2010 2015 2020

y prices (index 2000 = 100).

Table 2Heat pump assumptions for the NEI model [60].

Temperature range of heat demand 60–69 �C 70–79 �C 80–89 �C 90–99 �C 100–119 �C 120–139 �C

COP 6.29 4.85 3.99 3.42 2.85 2.36Investment costs assumed (€/kWth) 83.49 247.23 475.83 643.43 927.84 1270.78

7.8%

12.5%

13.6%

0%

2%

4%

6%

8%

10%

12%

14%

16%

0

10

20

30

40

50

60

70

80

BAU Sc_HP BAU Sc_HP BAU Sc_HP BAU Sc_HP BAU Sc_HP

2001 2005 2010 2015 2020

Ene

rgy

savi

ngs d

ue to

HP

(%)

Fina

l ene

rgy

cons

umpt

ion

(TW

h)

Coke Electricity Natural gasLPG Coal Special fuel Heat Energy savings (%)

Light fuel oil Heavy fuel oil

11.7%

16.6%

19.2%

20.9%

0%

5%

10%

15%

20%

25%

0

0.5

1

1.5

2

2.5

2010 2012 2015 2020

Perc

ent r

educ

tion

(%)

Red

uctio

n (M

tCO

2/yr

)

CO2 reduction by HP introduction in food & drink industry (MtCO2/yr)

Percent reduction in total CO2 emission in food & drink industry (%)

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

2010 2011 2012 2013 2014 2015 2017 2020

Ene

rgy

savi

ngs (

TW

h/yr

)

Man. of tea & coffeeMan. of alcoholic beveragesMan. of wines and waters drinkMan. of Bread & pastry goodsFish sectorMan. of farinaceous prod.Man. of ciderMan. of cocoaMan. of Grain mill productsMan. of Oils & fats Man. of beerMan. of maltMan. of animals feedMan. of Fruits & vegetablesMan. of meat and poultrymeatMan. of SugarMan. of meat productsMan. of other food prod.Man. of StarchesMan. of Dairy products

Fig. 6. Results for the French food and drink industry.

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 381

to achieve their goal of total emissions and energy efficiency in NEIindustry in the short, medium and long term. For example, author-ities would have to tax emissions by around € 112/tCO2 for anadditional potential of 1.5 TW h corresponding to an incentive of40% of the investment cost of HPs.

The structure of the NEI industry with TIMES allows us to ana-lyze costs up to a 4-digit level of nomenclature NACE and thencheck whether this result is related to all subsectors of the F&Dindustry. The French F&D industry is divided into 35 subsectors(grouped into 20 sectors) at 4-digit level. This flexibility meanswe can disaggregate the additional energy savings obtained in allF&D subsectors (Fig. 11).

Note that, unlike in Fig. 9, we have only represented theadditional potential to make the subsectors more legible. Inaddition, we have subdivided them into two graphs: from thehighest potential (top graph) to the lowest potential (bottomgraph) in additional subsectors with lower deposits.

Thus, we can see that the five subsectors: Dairy (NACE 155),Fruit & vegetable (NACE 153), Manufacture of prepared animalfeeds (NACE 157), Manufacture of meat products (NACE 1513),and Meat production (NACE 1511 & 1512), account for nearly80% of total additional energy savings at respectively 25%, 20.5%,14%, 10% and 9%.

Conversely, we note that there is no additional potentialobserved, whatever the level of incentive, in the starch productssector (NACE 1562) and other food industry sectors (NACE 1587,1588 & 1589). On the other hand, almost every remaining subsec-tor is below 3% of the total.

The three graphs in Fig. 12 below show the differentiation ofcosts by level of incentive and the additional energy savings forall subsectors of the F&D industry. Along with the starch productssector (NACE 1562) and other food industry sectors (NACE 1587,1588 & 1589), we have not represented the sugar sector becauseof its high costs.

57.0%

30.3%

24.2%

44.4%

38.1%

50.0%

18.5%

0%

10%

20%

30%

40%

50%

60%

0

1

2

3

4

5

6

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Man. of Dairy products

Man. of Starches

Man. of Sugar Man. of Fruits & vegetables

Man. of other food prod.

Man. of meat products

Man. of Oils & fats

Hea

t rre

cove

red

by H

P re

lativ

e to

sect

oral

fina

l

Hea

t Dem

and

(TW

h/yr

)

60-70°C 70-80°C 80-90°C 90-100°C 100-120°C120-140°C 140-200°C Other Heat recovered (%)

36.9%

43.1%

12.7%

31.0%

50.7%

52.0%

35.1%

0%

10%

20%

30%

40%

50%

60%

0

0.2

0.4

0.6

0.8

1

1.2

1.4

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Man. of animals feed

Man. of meat and

poultrymeat

Man. of alcoholic beverages

Man. of cocoa Man. of malt Man. of beer Man. of Grain mill products

Hea

t rre

cove

red

by H

P re

lativ

e to

sect

oral

fina

l

Hea

t Dem

and

(TW

h/yr

)

60-70°C 70-80°C 80-90°C 90-100°C 100-120°C120-140°C 140-200°C Other Heat recovered (%)

75.3%

31.0%36.0%

67.4%

38.0%

36.8%

0%

10%

20%

30%

40%

50%

60%

70%

80%

0

0.05

0.1

0.15

0.2

0.25

0.3

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Hea

t dem

nd

Hea

t pro

duce

d by

HP

Man. of cider Man. of Bread & pastry goods

Fish sector Man. of farinaceous

prod.

Man. of tea & coffee

Man. of wines and waters drink

Hea

t rre

cove

red

by H

P re

lativ

e to

sect

oral

Hea

t Dem

and

(TW

h/yr

)

60-70°C 70-80°C 80-90°C 90-100°C 100-120°C120-140°C 140-200°C Other Heat recovered (%)

ener

gy c

onsu

mpt

ion

(%)

ene

rgy

cons

umpt

ion

(%)

fina

l ene

rgy

cons

umpt

ion

(%)

Fig. 7. Detailed results on the substitution HP-steam boilers in food and drink industry.

8

8.5

9

9.5

10

10.5

11

11.5

12

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Ene

rgy

savi

ngs (

TW

h)

HP investment incentive (% of investment cost)

Fig. 8. Impact of an HP investment incentive on additional energy savings in theF&D industry by 2020.

0

0.5

1

1.5

2

2.5

3

3.5

0 10 20 30 40 50 60 70 80 90 100

Add

ition

alen

ergy

savi

ngs(

TW

h)

Energy savings average cost (€/MWhcumac)

40%

70%

60%

50%

30%

20%

100%90%

80%

Fig. 9. Evolution of additional energy savings with their costs for each level of HPinvestment incentive (in% in the graph) in the F&D industry by 2020.

382 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

We can take the example of the dairy sector curve in the firstgraph in Fig. 12 (first blue5 curve with circle-shaped marks in thetop graph) to better understand the evolution of all other cost curves.

5 For interpretation of color in Fig. 12, the reader is referred to the web version ofthis article.

Each point (circle-shaped mark) on the curve of the dairy sector rep-resents, for a given level of HP investment, an incentive by steps of10% (beginning at 20%). The X axis represents the cost of the addi-tional energy savings (€/MW h cumac) and the Y axis the amount ofadditional energy savings (GW h). Thus it is a representation of a3-dimensional result (cost, additional energy savings and incentivelevel) in a 2-dimensional diagram. For an incentive of 20% of the HPinvestment cost, we achieve additional savings of 254.5 GW h at a

0

50

100

150

200

250

300

350

400

7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9 8 8.1 8.2

Ave

rage

Cos

t (€/

tCO

2)

Emissions (MtCO2 )

80%

70%

60%

50%

40%

30%

20% 10%

Fig. 10. Evolution of the average cost to reduce CO2 emissions with an HPinvestment incentive in the F&D industry by 2020.

0

500

1000

1500

2000

2500

3000

0% 10% 20% 30% 40% 5

Add

ition

alen

ergy

savi

ngs(

GW

h)

HPs investment incent

Dairy sector Man. of FruitsInd. Meat products Ind.of meat prSector alcohol prod. Fish sectorMan. of cocoa

0

50

100

150

200

250

Add

ition

alen

ergy

savi

ngs(

GW

h)

HPs investment incent

Processing tea and coffee Man. Of cidMan. of beer Ind. farinaceMan. Of malt Sugar sectorMan. Other food prod.

0% 10% 20% 30% 40% 5

Fig. 11. Disaggregation by subsector of the impact of an HP in

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 383

cost of about € 24.9/MW h cumac. Up to 40% (third point on the dairysector curve), there are practically no additional savings, whichexplains the increase of the cost to € 54/MW h cumac. However, ata level of 50%, we almost double the potential. The calculation givesus a cost of € 38.4/MW h cumac. This potential remains constant,even at a 60% incentive level, which again only explains the increasein the cost. And finally at 70%, the dairy sector attains the maximumadditional savings possible, i.e. 759.2 GW h at a cost of€ 42.3/MW h cumac. Thereafter, as the maximum is reached, thecurve remains horizontal regardless of the incentive increase. Thisanalysis allows us to better understand the ‘‘inverted Z’’ shape of costcurves. It gives decision-makers greater visibility of the level of costsusing an energy policy with an incentive on a given technology in agiven sector compared to a penalty.

The curve for the meat production sector (NACE 1511 & 1512) isanother example. Indeed, it would be interesting to establish a 50%incentive level, since for practically the same marginal energy sav-ings cost, additional energy savings are almost doubled. Further

0% 60% 70% 80% 90% 100%

ive (% of investment cost)

& vegetables Man. prepared animal feedsoduction Oils and fats sector

Grain mill products

ive (% of investment cost)

er Ind. spirits and wine prod.ous prod. Ind. of bread, pastry prod.

Man. of Starches

0% 60% 70% 80% 90% 100%

vestment incentive on additional energy savings by 2020.

0

20

40

60

80

100

120

140

160

0 10 20 30 40 50 60 70 80 90

Add

ition

alen

ergy

savi

ngs(

GW

h)

Oils and fats sector Fish sector Grain mill productsMan. of cocoa Man. Of cider Ind. spirits and wine prod.

0

10

20

30

40

50

60

0 20 40 60 80 100 120 140

Add

ition

alen

ergy

savi

ngs(

GW

h)

Ind. farinaceous prod. Man. of beer Ind. of bread, pastry prod.Processing tea and coffee Man. Of malt

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50 60 70 80

Add

ition

alen

ergy

savi

ngs(

GW

h)

Energy savings average cost (€/MWhcumac)

Sector alcohol prod. Dairy sector Man. of Fruits & vegetablesMan. prepared animal feeds Ind. Meat products Ind.of meat production

Energy savings average cost (€/MWhcumac)

Energy savings average cost (€/MWhcumac)

Fig. 12. Evolution of additional energy savings with their costs by subsector by2020.

Table 3Additional energy savings with a penalty of 20€/MW h cumac.

F&D sub-sector NACE code Energy savings(GW h cumac)

Average costs(€/MW h cumac)

Alcoholproduction

NACE 159A &1592

135 13.6

Fish NACE 152 80 20Oils and fats NACE 154 68.5 18Tea and coffee

processingNACE 1586 50 17.4

Spirits and wineproduction

NACE 159B,1593 & 1595

25 17.3

384 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

analysis could center on potential additional savings in the F&Dindustry depending on the penalty level.

Therefore, we deduce that the F&D sector could achieve addi-tional savings. Regarding energy savings costs compared to thecurrent penalty of € 20/MW h cumac, a total of 358.5 GW h couldbe attained in the following F&D sub-sectors (Table 3).

However, we could assume a harsher penalty of about€ 30/MW h cumac, in which case the F&D industry could poten-tially achieve additional savings of around 1.11 TW h respectivelydisaggregated in the following sub-sectors (Table 4).

These results strongly underline the strategic importance of thisdisaggregated modeling using TIMES. The results of differentiatedenergy savings’ average costs in the most disaggregated industry(4-digit level) could be a good decision support tool for energyoperators using the ESC mechanism; they encourage investmentin new technologies for energy demand management policies inany industrial sector and better scrutiny of its subsectors.

However, all of these results were obtained with a given sce-nario for all exogenous variables of our model. It is now importantto make a sensitivity analysis of these scenarios. Regarding thepossibility that HPs could replace or help energy efficiency policy,in the following part of our analysis we focus on the natural gasprice and the carbon footprint. Indeed, these two parameters cansignificantly affect the results.

4.2. Sensitivity analysis of natural gas prices and carbon footprint ofthe electricity sector

4.2.1. Results and discussion of the impact of gas pricesThe use of heat pumps in the food industry mainly involves sub-

stituting natural gas for electricity. This is because the main fuelconsumed by boilers in the NEI industry is natural gas, at around65% of total energy consumption. Thus, we studied the impact ofchanges in gas prices on the implementation of HP systems inthe food and drink industry.

The international market for natural gas is not a unified market,but a market segmented into three distinct geographical areas[67,68] in which price formation obeys different rationales: theNorth American market, the European market and the Asianmarket.

The European gas market is currently in a dual situation: withthe UK operating with spot market prices and the continentdirectly indexed to oil products. The European Union is highlydependent on gas imports (nearly 60% of its gas), mostly throughlong-term contracts because of the high investment cost oflong-distance transport (about 80% of imported gas) [69]. Thisdependence is expected to increase in the future [70]. Thus, theprojected natural gas price scenario of the World Energy Outlook2010 is based on the assumption that the ratio of gas prices andoil prices remains unchanged until 2035 whereas it remainedbelow average between 1980 and 2009.

Nevertheless, spot market prices are likely to gain ground in thefuture because they tend to be lower than the prices of recentlong-term contracts. However, some reservations can beexpressed. The rise in LNG supplies will increase transactions onmarket places, overcoming frontiers and reducing the price gapbetween these three major places. The new LNG liquefaction ter-minal project in the USA will make it possible to export cheap shalegas, mostly for the Asian market [71]. In this new context inEurope, higher gas prices could be possible in the future becauseof potential tensions on the supply side. Therefore, we analyzedhow different natural gas prices starting from 2015 would

Table 4Additional energy savings with a penalty of 30€/MW h cumac.

F&D sub-sector NACE code Energy savings(GW h cumac)

Average costs(€/MW h cumac)

Dairy NACE 155 254.5 24.9Meat products NACE 1513 177 29.7Manufacture of

prepared animalfeeds

NACE 157 166.4 22.8

Alcohol production NACE 159A &1592

135 13.6

Meat production NACE 1511 &1512

98 26.6

Fish NACE 152 80 20Oils and fats NACE 154 68.5 18Tea and coffee

processingNACE 1586 50 17.4

Grain mill products NACE 1561 42 21Spirits and wine

productionNACE 159B,1593 & 1595

25 17.3

Industrial bread,pastry products

NACE 1581 &1582

12 28.3

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 385

influence the impact of heat pump penetration in the food anddrink industry.

We consider six scenarios with natural gas price increases of 5%(Sc_HP + 5%), 10% (Sc_HP + 10%), and 15% (Sc_HP + 15%) anddecreases of �5% (Sc_HP-5%), �10% (Sc_HP-10%) and �15%(Sc_HP-15%) compared to the HP scenario (Sc_HP). Sc_HP ischaracterized by the implementation of heat pumps in thefood-processing industry until 2020. This scenario makes itpossible to represent and measure the potential for heat recoveryusing HP at the energy end-uses. We quantify the gains of thesescenarios regarding the ‘‘Business As Usual’’ (BAU) reference sce-nario, in which no new policy or measure is considered necessaryor adopted.

For the sake of better visualization and comprehension, wedivide the F&D industry (33 sub-sectors) into three groups: thedairy sector (NACE 155), the sugar sector (NACE 1583) and theremaining F&D sub-sectors. Figs. 13–15 below show the evolutionof their energy mix in final energy consumption depending on thefluctuating price of natural gas. These figures show that variations

6.7%

9.0%9.2%9.7%

11

0

10

20

30

40

50

60

70

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

20152010

Fina

l ene

rgy

cons

umpt

ion

(TW

h)

Electricity Natural gas LLight fuel oil

Heavy fuel oil

Fig. 13. Changes in the mix of energy purc

in natural gas prices influence energy savings, as we expected, inall sub-sectors starting from 2015. Indeed, a decrease in naturalgas prices is unfavorable to HP deployment, but the oppositedoes not necessarily encourage HP because there may be aninter-energy substitute in boiler energy consumption, as we shallsee later. The reactivity of each F&D subsector in relation to gasprice fluctuations is variable, as are the inter-energy substitutionsobserved.

Starting with the remaining F&D sub-sectors (Fig. 13), we onlyobserve a response from this group when the natural gas priceincreases up to 10% compared to the HP scenario.

With a 10% increase in the gas price, there is a substitution of0.22 TW h of natural gas for 0.06 TW h of electricity correspondingto a partial deployment of the high temperature HP (100–120 �C)up to 2020 in some subsectors, such as Manufacture of oils and fatsNACE 154 (23% of the total heat production), Manufacture of grainmill products NACE 1561 (18%) and Manufacture of cocoa NACE1584 (15%), and Alcohol production (5%). The increase in gas pricesboosts profitability, but results in limited penetration of very hightemperature HPs in these sectors. This provides 0.4% of additionalenergy savings in final energy consumption compared to the HPreference scenario. It is worth noting that with an increase of15%, the substitution of natural gas (4.7 TW h) affects not onlyelectricity (0.41 TW h) but also heavy fuel oil (3.27 TW h) up to2020, generating additional savings of around 1 TW h. The prof-itability of high temperature HP remains very limited, and the evo-lution of gas prices at this level promotes inter-energy substitutionin the boiler energy mix because boilers using heavy fuel oilbecome more competitive than natural gas boilers (substitutionof around 90% in heat production). Thus, we could assume that thisgroup will need an increase in gas prices of over 15% to promoteadditional high temperature HP deployment.

The Sugar sector NACE 1583 (Fig. 14) evolves completely differ-ently in comparison to the previous group. A +5% increase in gasprices immediately leads to 0.86 TW h of natural gas beingsubstituted by 0.74 TW h of electricity because it becomes moreprofitable to buy a certain amount of electricity than to produceit using natural gas. Even if the increase in gas prices reaches15%, this does not promote high temperature HP in comparisonto heat production from boilers because of the highly seasonal

.3%

11.1%

8.9%

11.2%11.6%

13.4%

0%

2%

4%

6%

8%

10%

12%

14%

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

2020

Ene

rgy

savi

ngs i

n co

mpa

riso

n to

the

BA

U sc

enar

io (%

)

PG Coal Steam Others Energy savings (%)

hases in the remaining F&D industry.

6.7% 7.2%6.9% 7.2%

11.7%12.1%

13.4%14.3%

16.3%

15.9%

11.8%

12.2%

13.6%14.4%

16.4% 16.2%

0%

2%

4%

6%

8%

10%

12%

14%

16%

18%

0

1

2

3

4

5

6

7

8

9

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

202020152010

Ene

rgy

savi

ngs i

n co

mpa

riso

n to

the

BA

U sc

enar

io (%

)

Fina

l ene

rgy

cons

umpt

ion

(TW

h)

Coke Electricity Natural gas LPG Coal Steam Energy savings (%)Light fuel oil

Heavy fuel oil

Fig. 14. Changes in the mix of energy purchases in the sugar sector (NACE 1583).

5.3%

12.9%

13.4% 13.3%

8.8%

16.6%

23.4%

24.8%

24.3%24.3%

8.8%

16.6%

23.4%

24.8%

24.4%

0%

5%

10%

15%

20%

25%

30%

0

2

4

6

8

10

12

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

Sc_H

P-15

%

Sc_H

P-10

%

Sc_H

P-5%

Sc_H

P

Sc_H

P+5%

Sc_H

P+10

%

Sc_H

P+15

%

202020152010

Ener

gy sa

ving

s in

com

pari

son

to th

e B

AU

scen

ario

(%)

Fina

l ene

rgy

cons

umpt

ion

(TW

h)

Electricity Natural gas LPG Coal Steam Others Energy savings (%)Light fuel oil

Heavy fuel oil

Fig. 15. Changes in the mix of energy purchases in the dairy sector (NACE 155).

386 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

character of production. Thus, it will influence only inter-energysubstitution in the boiler energy mix between heavy fuel oil andnatural gas in heat production. The importance of this substitutiondepends on the level of the increase in natural gas prices. Increasesof 10% and 15% in natural gas prices result in respectively0.88 TW h and 0.97 TW h of natural gas being substituted by0.9 TW h and 1.03 TW h of heavy fuel oil (respectively 50% and55% of gas are substituted). These increases in heavy fuel oilpurchases for scenarios of +10% and +5% of the price of naturalgas slightly reduce the energy savings obtained by 0.2%.

In the Dairy sector NACE 155 (Fig. 15), a variation in natural gasprices only results in inter-energy substitution between heavy fueloil and natural gas in the boiler energy mix. Indeed, as we haveshown in previous paragraphs, the first four temperature heatslices (heat requirements are subdivided into eight temperatureranges from 60 �C to 200 �C, i.e. four 10 �C ranges between 60 �C

and 100 �C, two 20 �C ranges between 100 and 140 �C, a 140 and200 �C range, and a final range called ‘‘other’’) are mainly producedby HPs because of their high profitability in this sector [2].However, at these price levels, high-temperature HPs are still notprofitable. We should consider natural gas price increases above15% to anticipate the profitability of high-temperature HPs in thedairy sector. It is worth noting that the substitution observedbetween natural gas and heavy fuel oil is already almost at its max-imum at a +5% gas price increase (1.89 TW h of natural gas for1.92 TW h of Heavy Fuel Oil).

In conclusion, the remaining F&D sectors (sugar and dairy notincluded) seem to be the most responsive to changes in naturalgas prices for additional HP deployment when consideringincreases up to 15% compared to the reference price. Conversely,in the dairy and sugar sectors, only inter-energy substitutionsbetween natural gas and heavy fuel oil in the boiler energy mix

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 387

are possible with rising natural gas prices. Higher natural gasprices should be considered to observe the deployment of heatpumps at higher temperatures in these two sectors.

On the other hand, it is also worth noting that a decrease innatural gas prices does not promote the penetration of heat pumpsystems in any F&D sector due to the profitability of producingheat directly from boilers. As a result, energy savings are reduced.

This kind of sensitivity analysis could be helpful to assess therange of flexibility in an investment choice regarding such avolatile parameter.

Many parameters taken into account in our model can influencethe results. One is particularly interesting when we are dealingwith electricity and climate change, i.e. the level of carbon thatcan be included in electricity production in terms of technologiesthat have been used to produce it (the carbon footprint). We there-fore present below the influence of the carbon content of the elec-tricity production used in heat pump penetration for heat recovery.

4.2.2. Results and discussion of the electricity sector’s carbon footprintAfter analyzing the impact of fluctuations in natural gas prices

on the deployment of heat pumps in the F&D industry, we studiedthe model’s response at different levels of electricity carbon con-tent to put the potential of this NEI model into perspective.Indeed, as HPs are used to substitute fossil fuels with the aim ofincreasing the total energy efficiency of the process, it is importantto look at the actual carbon footprint of these technologies depend-ing on the country (different electricity mixes) in which they couldbe deployed.

In this context, we assumed and implemented an environmen-tal policy with a CO2 tax. Industrial sectors that do not come underthe EU ETS may be subject to specific measures, whereby Europewould set goals with a free choice of tools for states [66]. Theseindustries include all NEI sectors in France, and the possibility ofa European or national ‘‘CO2 tax’’ has regularly been suggestedfor this environmental policy. We choose the carbon value scenariorecommended by the CAS (Strategic Analysis Centre which is adecision-making and expertise institution under the authority ofthe French Prime Minister) based on reasonable use of economictheory and models (GEMINI-E3, POLES and IMACLIM-R) [72]. TheCO2 tax is set at 200530 €/tCO2 in 2010 (200832 €/tCO2) with anannual growth of 5.8% until 2020 and reaching 200553 €/tCO2

0

2

4

6

8

10

12

14

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

Sc_H

P_PL

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

Sc_H

P_PL

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

2010 2012 2015

HPs

hea

t pro

duct

ion

(TW

h)

60-70°C 70-80°C 80-90°C 90-100°C

Fig. 16. Evolution of the penetration level of HPs by tem

(200593 €/tCO2 in 2030 corresponding to 2008100 €/tCO2). Thisscenario is based on the idea of exploiting the priority fields oflow-cost abatement available today and does not weigh on growthor help manage economic, social and professional transitions. Weconsider three additional scenarios [73,74]:

� A scenario in which we highlight the impact of electricity car-bon content at around zero gCO2/kW h (Sc_HP_NO) assumingthat electricity is produced with almost 98% hydraulic power,as in Norway.� A second scenario, where electricity does indeed have a carbon

content of 483 gCO2/kW h (Sc_HP_IT) corresponding to Italy, totake the example of electricity production using mainly naturalgas (about 55%).� The third scenario represents the case of electricity mainly

produced from coal (90%), like in Poland, with a carbon contentof 1191 gCO2/kW h (Sc_HP_PL).

These three scenarios are compared to France’s HP scenario(Sc_HP_FR) assuming a carbon content of 55 gCO2/kW h [75].

Fig. 16 shows the evolution of the deployment of the heat pumpby temperature range in the food and drink industry for each levelof electricity carbon content subject to CO2 emissions taxation. Ithighlights the fact that the deployment of HPs is almost the sameup to 2030 for carbon content (Sc_HP_FR and Sc_HP_NO), and up to55 gCO2/kW h regardless of taxation. At 483 gCO2/kW h, a differ-ence emerges by 2020, when a CO2 tax of 200553 €/tCO2 is reached,reducing production by about 2% in 2020, 5% in 2025 and 10.5% in2030.

When electricity production is mainly from coal sources(Sc_HP_PL), HP penetration is reduced by about 6% from 2012,when taxation reaches 200534 €/tCO2. This reduction increases to12.6% in 2015 because gas boilers become much more profitablethan high-temperature HPs (100–120 �C) due to the emission tax-ation and the electricity carbon content. Afterwards, the steadyincrease in natural gas prices compared to electricity pricesfacilitates their penetration until 2020. Thus, a 6% reduction in heatpump penetration is achieved by 2020. From 2020, natural gasprices have stabilized relative to electricity prices and induce fur-ther penetration of gas boilers in heat production. This reductionin the deployment of heat pumps increases with tax to around8.5% in 2025 and 13.5% in 2030 (Fig. 17).

Sc_H

P_PL

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

Sc_H

P_PL

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

Sc_H

P_PL

Sc_H

P_N

O

Sc_H

P_FR

Sc_H

P_IT

Sc_H

P_PL

2020 2025 2030

100-120°C 120-140°C 140-200°C

perature versus CO2 content per kW h of electricity.

0%

2%

4%

6%

8%

10%

12%

14%

16%

0 10 20 30 40 50 60 70 80 90 100

Red

uctio

n of

HP

pene

trat

ion

rega

rdin

g to

the

scen

ario

Sc_

HP_

FR (%

)

CO2 Tax (€/tCO2)

Sc_HP_PL (1191 gCO2/kWh)Sc_HP_IT(483 gCO2/kWh)

Fig. 17. Evolution of HP penetration regarding scenario Sc_Hp_FR depending on thetaxation level of CO2 emissions.

388 G.S. Seck et al. / Applied Energy 156 (2015) 374–389

As we might have expected, these results highlight the fact thatthanks to nuclear power and hydropower, the level of low-carboncontent electricity in France is a great advantage to promote heatpumps. Moreover, it also illustrates, as mentioned above, that veryhigh levels of taxation are needed to attempt to significantly influ-ence the penetration of HPs in non-energy intensive industry.However, our model can show that even in cases of high CO2 con-tent per kW h (as in Poland), limited heat pump penetration isalways possible. Such a result would not have been attainablewithout a detailed model of the non-energy intensive industry’sstructure, such as the one we have developed.

5. Conclusion and policy implications

This paper has allowed us to present other potentialapplications of the first detailed techno-economic model for thenon-energy intensive industry by analyzing a study case of heatrecovery using heat pumps in one of its key sectors, the food anddrink industry. Our prospective modeling allowed us to observethe pattern of HP investments in response to incentive policieswithin the Energy Savings Certificate mechanism. This energymodel could appraise the possibility of improving energy savingspotential and establish the marginal cost in non-energy intensiveindustries. We went on to analyze what level of incentive policiescould be implemented to encourage broader HP adoption in theseindustries and improve energy savings. This NEI model couldprovide a useful decision-making tool by determining the differen-tiated costs of energy savings through investing in efficienttechnologies at the highest level of disaggregation for bettersectorial scrutiny. It also merits a sensitivity analysis to showdecision-makers the limits of this kind of model. To complete thisstudy, we could carry out an analysis to provide regulation author-ities with of the levels of tax on GHG emissions required to reachtheir environmental and energy objectives in NEI over a mediumor long-term horizon.

In the industrial sector, few measures are subject to EnergySavings Certificates (White certificates) in France. The new 2030policy framework for climate and energy proposed by theEuropean Commission aims to achieve a 30% reduction in green-house gas emissions from sectors outside the EU emissions trading

system (EU ETS), such as the non-energy intensive industry, andbring them down to below the 2005 level by 2030. Heat recoveryusing heat pumps should be developed and recognized as eligiblefor Energy Savings Certificates in order to achieve energy efficiencyand CO2 goals. The non-energy intensive industry could contributesignificantly to energy efficiency goals in France, and more widelyin the European Union, by promoting efficient energy technologieslike heat pumps.

Acknowledgements

This research was supported by the Chair Modeling for sustain-able development, driven by MINES ParisTech, Ecole des PontsParisTech, AgroParisTech, and ParisTech, supported by ADEME,EDF, RENAULT, SCHNEIDER ELECTRIC and TOTAL.

References

[1] Hita A, Seck GS, Guerassimoff G, Djemaa A. Assessment of the potential of heatrecovery in food and drink industry by the use of TIMES model. EuropeanCouncil for an Energy Efficient Economy (ECEEE); 2011.

[2] Seck GS. Prospective modeling of the non-energy intensive industry for theevaluation of the impact of energy demand management policies by using themodel TIMES: assessment of the potential of heat recovery with Heat Pumpssystems (HP) in French Food and Drink industry, Mines Paristech; 2012.

[3] Seck GS, Guerassimoff G, Maïzi N. Heat recovery with heat pumps in non-energy intensive industry: a detailed bottom-up model analysis in the FrenchFood and Drink industry. Appl Energy 2013;111:p489–504.

[4] Fishbone LG, Abilock H. MARKAL, a linear-programming model for energysystem analysis: technical description of the BNL version. Int J Energy Res1981;4(p353):375.

[5] Fishbone LG, Giesen G, Goldstein G, Hymmen HA, Stocks KJ, Vos H, et al. User’sguide for MARKAL (BNL-51701), Brookhaven National Laboratory, Upton, NewYork; 1983.

[6] Berger C, Dubois R, Haurie A, Lessard E, Loulou R, Waaub JP. Canadian Markal:an advanced linear programming system for energy and environmentalmodeling. INFOR 1992;30:222–39.

[7] United States Department Of Energy (DOE). Model documentation report:Industrial sector demand module of the National Energy Modelling System,DOE/EIA-M064 (2001), US Department Of Energy (DOE), Energy InformationAdministration (EIA), Washington (DC); 2000.

[8] Loulou R, Goldstein G, Noble K. Documentation for the MARKAL family models,ETSAP; 2004. .

[9] Loulou R, Remme U, Kanudia A, Lehtila A, Goldstein G. Documentation forTIMES model, ETSAP; 2005. .

[10] Assoumou E. Modélisation MARKAL pour la planification énergétique long-terme dans le contexte français, Mines ParisTech; 2006.

[11] Ichinohe M, Endo E. Analysis of the vehicle mix in the passenger-car sector inJapan for CO2 emissions reduction by a MARKAL model. Appl Energy2006;83:1047–61.

[12] Maurizio Gargiulo, Getting started with TIMES-VEDA (Version 2.7), ETSAP; May 2009.

[13] Cayla J-M. Les ménages sous la contrainte carbone: Exercice de modélisationprospective des secteurs résidentiel et transports avec TIMES, MinesParisTech; 2011.

[14] Drouineau M. Modélisation prospective et analyse spatio-temporelle:intégration de la dynamique du réseau électrique. Mines ParisTech; 2011

[15] Føyn THY, Karlsson K, Balyk O, Grohnheit PE. A global renewable energysystem: a modelling exercise in ETSAP/TIAM. Appl Energy 2011;88:526–34.

[16] ADEME (French Environment and Energy Management Agency). Guide desfacteurs d’émissions version 6.1: Calcul des facteurs d’émissions et sourcesbibliographiques utilisées. Bilan carbone; 2010.

[17] Gielen D, Taylor M. Modeling industrial energy use: the IEAs energytechnology perspectives. Energy Econ 2007;29:889–912.

[18] Djemaa A. Modélisation bottom-up, un outil d’aide à la décision long termepour les mesures politiques en matière d’énergie et d’environnement – Lemodèle TIMES appliqué aux industries grandes consommatrices d’énergie(IGCE). Mines ParisTech; 2009.

[19] Djemaa A, Hita A, Guerassimoff G, Maïzi N. Modelling energy saving potentialin industry. European Council for an Energy Efficient Economy (ECEEE); 2007.p. 1383–93.

[20] Al-Rabghi OM, Beirutty M, Akyurt M, Najjar Y, Alp T. Recovery and utilizationof waste heat. Heat Recov. Syst. CHP 1993;13:p463–70.

[21] Industrial Energy-related Technologies and Systems (IETS). Industrial excessheat recovery technologies & applications. International Energy Agency (IEA);2010.

[22] CEREN (Centre d’études et de recherches économiques sur l’énergie). Energyprofile of the Food & Drink sector and Pulp & paper sector – Enquêtes sur leMarché Industriel de l’Energie (EMIE), no 6301; 2005.

[23] JRC-IPTS (Joint Research Centre – Institut for Prospective TechnologicalStudies). Analysis of the European Food industry, European Commission; 2008.

http://refhub.elsevier.com/S0306-2619(15)00877-6/h0015http://refhub.elsevier.com/S0306-2619(15)00877-6/h0015http://refhub.elsevier.com/S0306-2619(15)00877-6/h0015http://refhub.elsevier.com/S0306-2619(15)00877-6/h0020http://refhub.elsevier.com/S0306-2619(15)00877-6/h0020http://refhub.elsevier.com/S0306-2619(15)00877-6/h0020http://refhub.elsevier.com/S0306-2619(15)00877-6/h0030http://refhub.elsevier.com/S0306-2619(15)00877-6/h0030http://refhub.elsevier.com/S0306-2619(15)00877-6/h0030http://www.etsap.orghttp://www.etsap.orghttp://refhub.elsevier.com/S0306-2619(15)00877-6/h0055http://refhub.elsevier.com/S0306-2619(15)00877-6/h0055http://refhub.elsevier.com/S0306-2619(15)00877-6/h0055http://refhub.elsevier.com/S0306-2619(15)00877-6/h0055http://www.etsap.orghttp://refhub.elsevier.com/S0306-2619(15)00877-6/h0075http://refhub.elsevier.com/S0306-2619(15)00877-6/h0075http://refhub.elsevier.com/S0306-2619(15)00877-6/h0085http://refhub.elsevier.com/S0306-2619(15)00877-6/h0085http://refhub.elsevier.com/S0306-2619(15)00877-6/h0100http://refhub.elsevier.com/S0306-2619(15)00877-6/h0100

G.S. Seck et al. / Applied Energy 156 (2015) 374–389 389

[24] Berthou M, Lauruol J-M, Mourlin Y. Enervision Industrie – cartographie dutissu industriel français: description technique. Electricité de France (EDF)R&D, Eco-efficiency and Industrial Processes Department (EPI); 2009.

[25] Service de l’Observation et des Statistiques (SOeS). CO2 et activitéséconomiques de la France: Tendances 1990–2007 et facteurs d’évolution.Commissariat Général du Développement Durable (CGDD); 2010.

[26] Service de l’Observation et des Statistiques (SOeS). Chiffres clés du climat:France et Monde. Edition 2011; 2011. .

[27] Spoelstra S, Haije WG, Dijkstra JW. Techno-economic feasibility of high-temperature high-lift chemical heat pumps for upgrading industrial wasteheat. Appl Therm Eng 2002;22:1619–30.

[28] Kondou C, Koyama S. Thermodynamic assessment of high-temperature heatpumps using Low-GWP HFO refrigerants for heat recovery. Int J Refrig2015;53:p126–41.

[29] ADEME (French Environment and Energy Management Agency). Chaleur fataleindustrielle. ref. 8445; 2015..

[30] Jakobs R, Levacher L, Laue H-J. Application of industrial heat pumps. In: 10thIEA heat pump conference; 2011.

[31] Industrial Energy-related Technologies and Systems (IETS). Application ofindustrial heat pumps. International Energy Agency (IEA), Industrial Energy-related Systems and Technologies Annex 13, Heat Pump Programme Annex35; 2014.

[32] Auzenet E, Clerc-Renaud M. Pompes à chaleur: applications et systèmesparticuliers, Techniques de l’ingénieur BE 9621; 2006.

[33] Kew PA. Heat pumps for industrial waste heat recovery – a summary ofrequired technical an economic criteria. Heat Recov Syst 1982;2:283–96.

[34] Resume of a report compiled of staff of NEI projects Ltd. Cuthbert house.Potential applications for heat pumps in the dairy and brewing industries.Heat Recov Syst 1983;3:231–43.

[35] Smith SG. Case study – heat recovery with heat pumps. Heat Recov Syst1983;3:53–6.

[36] Wright JR, Steward FR. Three industrial heat pump installations operating inCanada. Heat Recov Syst 1985;5:81–8.

[37] Lazzarin RM. Heat pumps in industry – II. Applications. Heat Recov Syst CHP1995;15:p305–17.

[38] Berntsson T. Heat sources—technology, economy and environment. Int J Refrig2002;25:428–38.

[39] De Monte M, Padoano E, Pozzetto D. Waste heat recovery in a coffee roastingplant. Appl Therm Eng 2003;23:1033–44.

[40] Soroka B. Pompes à chaleur industrielles. Leonardo ENERGY, Laborelec; 2009.[41] Dupont M, Sapora E. The heat recovery potential in the French industry: which

opportunities for heat pump systems? ECEEE 2009 summer study; 2009. p1115–23.

[42] Colak N, Hepbasli A. A review of heat pump drying: Part 1 – Systems, modelsand studies. Energy Convers Manage 2009;50:2180–6.

[43] Murr R. Gains énergétiques globaux par installation de pompes à chaleur dansun procédé agro-alimentaire, Mines ParisTech; 2010.

[44] Assaf K. Intégration d’une pompe à chaleur dans un procédé agroalimentaire –Simulation, expérimentation et intégration. Mines ParisTech; 2010.

[45] Becker HC. Methodology and Thermo-Economic Optimization for Integrationof Industrial Heat Pumps, École Polytechnique Féderale de Lausanne (EPFL);2012.

[46] Heat Pump & Thermal Storage Technology Center of Japan. Survey ofavailability of heat pumps in the food and beverage fields; 2010.

[47] Reichl J, Kollmann A. The baseline in bottom-up energy efficiency and savingcalculations – a concept for its formalisation and a discussion of relevantoptions. Appl Energy 2011;88:422–31.

[48] AGRESTE (French Agricultural Statistics Agency). 2010: la reprise de lademande mondiale soutient l’activité des IAA, Agreste Conjoncture,synthèses n� 2011/147; Avril 2011.

[49] Ministère de l’Agriculture. de l’Alimentation, de la Pêche et des Affaires rurales,Panorama des industries agroalimentaires, Édition 2010.

[50] Règlement n� 1005/2008/CE du Conseil de l’Union Européenne du 29Septembre 2008 établissant un système communautaire destiné à prévenir,à décourager et à éradiquer la pêche illicite, non déclarée et non réglementée,modifiant les règlements (CEE) n� 2847/93, (CE) n� 1936/2001 et (CE) n� 601/

2004 et abrogeant les règlements (CE) n� 1093/94 et (CE) n� 1447/1999,Journal Officiel n� L 286 du 29/10/2008.

[51] Règlement n� 1006/2008/CE du Conseil de l’Union Européenne du 29Septembre 2008 concernant les autorisations pour les activités de pêche desnavires de pêche communautaires en dehors des eaux communautaires etl’accès des navires de pays tiers aux eaux communautaires, modifiant lesrèglements (CEE) n� 2847/93 et (CE) n� 1627/94 et abrogeant le règlement (CE)n� 3317/94, Journal Officiel n� L 286 du 29/10/2008.

[52] AGRESTE (French Agricultural Statistics Agency). IAA 2008: repli de l’activité,Agreste Conjoncture, synthèses n� 2009/71; 2009.

[53] Escarnot B, Escudero M, Bethenod C, Hita A, Pham-Van-Cang C, Morisset V.Prévisions de consommation d’électricité et de gaz naturel des industriesgrandes consommatrices d’énergie en France. EDF R&D, Eco-Efficiency andIndustrial Processes (EPI) department, Groupe «Expertise Clients et SecteursIndustriels» (Internal report); 2007.

[54] Règlement n� 318/2006/CE du Conseil de l’Union Européenne du 20 Février2006 portant organisation commune des marchés dans le secteur du sucre,Journal Officiel n� L 58 du 28/02/2006.

[55] Règlement n� 1234/2007/CE du Conseil de l’Union Européenne du 22 Octobre2007 portant organisation commune des marchés dans le secteur agricole etdispositions spécifiques en ce qui concerne certains produits de ce secteur(règlement «OCM unique»), Journal Officiel n� L 299 du 16/11/2007.

[56] Ministère de l’Agriculture. de l’Alimentation, de la Pêche et des Affaires rurales,Panorama des industries agroalimentaires, Edition 2008.

[57] AGRESTE (French Agricultural Statistics Agency). Boissons alcooliquesdistillées: reprise de la demande états-unienne et essor de la demandechinoise. Agreste Conjoncture, synthèses n� 2010/136; 2010.

[58] International Energy Agency (IEA). World energy outlook; 2010.[59] Levacher L. Heat: Energy efficiency issues, EDF R&D, European Centre &

Laboratories for Energy Efficiency Research (ECLEER) Department, Leprintemps de la Recherche 2009, Morêt-Sur-Loing (France); 2009.

[60] Berail JF, Sapora E. High temperature industrial heat pump. Eco-Efficiency andIndustrial Processes (EPI) Department, Groupe «Optimisation énergétique &expertise des procédés industriels». Technical report. EDF R&D; 2010.

[61] APESA (Centre technologique en en environnement et maitrise des risques).Les certificats blancs, noirs et verts, Dossier Technique; 2008. .

[62] Gouv 2005-781. Loi n�2005-781 du 13 juillet 2005 de programme fixant lesorientations de la politique énergétique. JORF n�163 du 14 juillet; 2005. p.11570.

[63] Guyonnet P. Le dispositif français des Certificats d’Économies d’Énergie.Association Technique Energie Environnement (ATEE), MINEFI/DGEMP; 2005.

[64] ADEME (French Environment and Energy Management Agency). Les certificatsd’economies d’énergie expliqués par l’ADEME, March 2008 n�6347; 2008.

[65] Gouv 2010-1663. Décret n�2010-1663 du 29 décembre 2010relatif auxobligations déconomies dénergie dans le cadre du dispositif des certificatsd’économies dénergie, JORF n�0302 du 30 décembre 201; 2010. p. 23222.

[66] Hansen JP, Percebois J. Energie, économie et politiques. de boeck; 2010. 779 p.[67] IFP (Institut Français du Petrole). Les liens entre les prix du gaz naturel et du

petrole. Panorama 2006; 2005.[68] IFP (Institut Français du Petrole). L’industrie gazière à l’horizon 2020,

Panorama 2006; 2006.[69] International Energy Agency (IEA). World energy outlook 2009; 2009.[70] Percebois J. Prix internationaux du pétrole, du gaz naturel, de l’uranium et du

charbon: La théorie économique nous aide-t-elle à comprendre les évolutions?Centre de Recherche en Economie et Droit de l’Energie (CREDEN), Cahier N�09.02.81; Février 2009.

[71] International Energy Agency (IEA). World energy outlook 2012; 2012.[72] Centre d’Analyse Stratégique (CAS). La valeur tutélaire du carbone, La

documentation Française; 2009. 424 p.[73] ADEME (French Environment and Energy Management Agency). Guide des

facteurs d’émission Version 5.0; 2007.[74] JRC-IE (Joint Research Centre – Institute for Energy). Comment développer un

plan d’action en faveur de l’énergie durable (PAED), European Commission;2010.

[75] ADEME (French Environment and Energy Management Agency). Note decadrage sur le contenu CO2 du kW h par usage en France; 2005.

http://www.statistiques.developpement-durable.gouv.frhttp://www.statistiques.developpement-durable.gouv.frhttp://refhub.elsevier.com/S0306-2619(15)00877-6/h0135http://refhub.elsevier.com/S0306-2619(15)00877-6/h0135http://refhub.elsevier.com/S0306-2619(15)00877-6/h0135http://refhub.elsevier.com/S0306-2619(15)00877-6/h0140http://refhub.elsevier.com/S0306-2619(15)00877-6/h0140http://refhub.elsevier.com/S0306-2619(15)00877-6/h0140http://refhub.elsevier.com/S0306-2619(15)00877-6/h0165http://refhub.elsevier.com/S0306-2619(15)00877-6/h0165http://refhub.elsevier.com/S0306-2619(15)00877-6/h0385http://refhub.elsevier.com/S0306-2619(15)00877-6/h0385http://refhub.elsevier.com/S0306-2619(15)00877-6/h0385http://refhub.elsevier.com/S0306-2619(15)00877-6/h0175http://refhub.elsevier.com/S0306-2619(15)00877-6/h0175http://refhub.elsevier.com/S0306-2619(15)00877-6/h0180http://refhub.elsevier.com/S0306-2619(15)00877-6/h0180http://refhub.elsevier.com/S0306-2619(15)00877-6/h0185http://refhub.elsevier.com/S0306-2619(15)00877-6/h0185http://refhub.elsevier.com/S0306-2619(15)00877-6/h0190http://refhub.elsevier.com/S0306-2619(15)00877-6/h0190http://refhub.elsevier.com/S0306-2619(15)00877-6/h0195http://refhub.elsevier.com/S0306-2619(15)00877-6/h0195http://refhub.elsevier.com/S0306-2619(15)00877-6/h0200http://refhub.elsevier.com/S0306-2619(15)00877-6/h0210http://refhub.elsevier.com/S0306-2619(15)00877-6/h0210http://refhub.elsevier.com/S0306-2619(15)00877-6/h0235http://refhub.elsevier.com/S0306-2619(15)00877-6/h0235htt