Application of Three Independent Consequential LCA Approaches to the Agricultural Sector

electronic supplementary material

Uncertainties in LCA

Application of three independent consequential LCA approaches to the agricultural sector in Luxembourg

Ian Vázquez-Rowe • SameerRege • AntoninoMarvuglia • JulienThénie • Alain Haurie • EnricoBenetto

7 March 2013 / Accepted: 16 May 2013

Responsibleeditor: Matthias Finkbeiner

I. Vázquez-Rowe() • S. Rege • A. Marvuglia • E. Benetto

Public Research Centre Henri Tudor (CRPHT) / ResourceCentrefor Environmental Technologies (CRTE) – 6A, avenue des Hauts-Fourneaux, 4362, Esch-sur-Alzette, Luxembourg

e-mail:

J. Thénie • A. Haurie

ORDECSYS. Operations Research Decisions and Systems, rue de Gothard 5, 1225, Chêne-Bourg, Switzerland

() Corresponding author:

Ian Vázquez-Rowe

Tel. +352-425 991-4926

e-mail:

Contents:

Section S1 – Description of the energy scenario

Table S1 – Reference scenarios for final energy demand forecast in Luxembourg in the period 2010-2020 (Ministry of Economy, 2010). Data reported per 1000 t oil-eq

Table S2 – Reference scenario for final energy demand forecast in Luxembourg in the period 2010-2020 with additional energy saving measures (Ministry of Economy, 2010). Data reported per 1000 t oil-eq

Figure S1 –Renewable energy share (%) forecast in Luxembourg according to LUREAP up to 2020 (Ministry of Economy, 2010)

Figure S2 – Development of energy production from biogas and installed biogas power until 2020 according to LUREAP (Ministry of Economy, 2010)

Figure S3 – Summary of calculations and assumptions leading from the additional electricity production from biogas (generated from maize feedstock) foreseen for 2020 to the necessary agricultural land

Table S3 – Development of the biogas sector until 2020 according to LUREAP (Ministry of Economy, 2010)

Section S2 – Description of the PE model

Table S4 – Crops set details

Table S5 –Metabolic requirements for animals and animal product prices in 2009

Table S6 –Demand, supply and excess demand for metabolic requirements for animals

Section S3 – Decision tree for expert opinion 1

Section S4 – Decision tree for expert opinion 2

Section S5 – System boundaries and Life Cycle Inventory

Figure S4 –Schematic representation of the consequential-LCA system boundaries for the production system assessed

Table S7 – Life Cycle Inventory table for maize cultivation in Luxembourg (data per ha)

Table S8–List of main crops and other activities inventoried for each of the developed C-LCA approaches

Section S6– Results. Environmental consequences with the maximization of revenues PE model (Approach A)

Table S9 – Characterization endpoint values for the selected scenarios in Approach A as compared to the baseline scenario. Data per FU= 1 MJ

Section S7 – Results. Environmental consequences with the opportunity cost PE model (Approach B)

Table S10 – Characterization endpoint values for the selected scenarios in Approach B as compared to the baseline scenario. Data per FU= 1 MJ

Section S8 – Results. Environmental consequences with the consequential system delimitation for agricultural LCAapproach (Approach C)

Table S11 – Characterization endpoint values for the selected scenarios in Approach C as compared to the baseline scenario. Data per FU= 1 MJ

References

Section S1

Description of the energy scenario

Luxembourg’s Renewable Energy Action Plan (LUREAP) describes, in accordance with the National Energy Efficiency Action Plan (NEEAP), the expected development of the end energy demand, considering the energy saving measures reported in the efficiency action plan (Table S1). Beyond this, the LUREAP foresees a second scenario, considering additional energy efficiency measures (Table S2).

Table S1 – Reference scenarios for final energy demand forecast in Luxembourg in the period 2010-2020 (Ministry of Economy, 2010). Data reported per 1000 t oil-eq.

2005 / 2010 / 2011 / 2012 / 2013 / 2014 / 2015 / 2016 / 2017 / 2018 / 2019 / 2020
Heat and cold / 1225 / 1293 / 1303 / 1313 / 1324 / 1334 / 1344 / 1363 / 1381 / 1399 / 1417 / 1436
Electricity / 524 / 553 / 556 / 559 / 562 / 565 / 568 / 574 / 581 / 588 / 595 / 602
Traffic / 2160 / 2309 / 2337 / 2365 / 2393 / 2421 / 2450 / 2477 / 2505 / 2532 / 2560 / 2587
Gross and energy demand / 4315 / 4558 / 4599 / 4639 / 4680 / 4720 / 4760 / 4812 / 4864 / 4504 / 4967 / 5019
Air traffic / 414 / 414 / 414 / 413 / 413 / 413 / 412 / 412 / 412 / 411 / 411 / 411
Reduction due to limit value concerning air traffic / 147 / 132 / 129 / 127 / 124 / 121 / 118 / 115 / 111 / 133 / 104 / 100
TOTAL demand / 4168 / 4426 / 4469 / 4513 / 4556 / 4599 / 4642 / 4698 / 4753 / 4371 / 4863 / 4919

2005 / 2010 / 2011 / 2012 / 2013 / 2014 / 2015 / 2016 / 2017 / 2018 / 2019 / 2020
Heat and cold / 1225 / 1235 / 1235 / 1235 / 1234 / 1234 / 1234 / 1241 / 1248 / 1255 / 1261 / 1268
Electricity / 524 / 549 / 548 / 547 / 546 / 545 / 544 / 549 / 554 / 559 / 564 / 569
Traffic / 2160 / 2086 / 2111 / 2137 / 2162 / 2187 / 2213 / 2238 / 2263 / 2288 / 2312 / 2337
Gross and energy demand / 4315 / 4283 / 4306 / 4329 / 4352 / 4376 / 4399 / 4428 / 4458 / 4487 / 4516 / 4546
Air traffic / 414 / 414 / 414 / 413 / 413 / 413 / 412 / 412 / 412 / 411 / 411 / 411
Reduction due to limit value concerning air traffic / 147 / 149 / 147 / 146 / 144 / 142 / 140 / 138 / 136 / 134 / 132 / 130
TOTAL demand / 4168 / 4134 / 4159 / 4183 / 4208 / 4234 / 4259 / 4290 / 4321 / 4353 / 4385 / 4417

In addition, following the 20/20/20 European Union (EU) targets, the government of Luxembourg predicts that by 2020 11% of the domestic energy demand should be provided by renewable sources, as shown in Figure S1. More specifically, the Luxembourgish policy regarding the development of renewable energies is based on three main pillars. In the first place, the use of national potentials is a top priority. Therefore, an intensive development of electricity and heat generation is desired. Regarding electricity production, biomass and wind power are intended to become major production carriers. The second component of the national strategy involves energy from renewable sources in the transport sector. In this case, a 10% target of renewable sources of energy should be achieved. Finally, the third main pillar is represented by the cooperation mechanisms. Based on limited national potentials and according to the current situation, Luxembourg must rely on this possibility in order to be able to reach its 11% overall target (see Figure S1).

Figure S1 – Renewable energy share (%) forecast in Luxembourg according to LUREAP up to 2020 (Ministry of Economy, 2010).

According to the 20/20/20 directive, the additional production of energy from biogas in Luxembourg in 2020 (as compared to 2010) should amount to 294.12 GWh, of which the amount relating to electricity is 144 GWh (see Figure S2 and Table S3). The assumption made in this study, therefore, is that this amount of biogas would be supplied domestically through an augmentation in the production of energy maize to produce biomethane.

Figure S2 – Development of energy production from biogas and installed biogas power until 2020 according to LUREAP (Ministry of Economy, 2010).

Table S3 – Development of the biogas sector until 2020 according to LUREAP (Ministry of Economy, 2010).

2005 / 2010 / 2011 / 2012 / 2013 / 2014 / 2015 / 2016 / 2017 / 2018 / 2019 / 2020
[MW] / 5 / 8 / 9 / 12 / 16 / 20 / 23 / 26 / 28 / 29 / 29 / 29
[GWh] / 27.2 / 44 / 49 / 67 / 88 / 108 / 123 / 133 / 139 / 143 / 143 / 144

The necessary additional quantity of maize fresh matter (FM) is estimated in about 80,000 tonnes and the area of agricultural land necessary to produce it is estimated in about 1600 ha, based on the following assumptions (see Fig. S3):

the electricity is produced at cogeneration plants, which generate 60% heat and 40% electricity;
the electricity production plant fed by biogas has an efficiency of 85% and the feedstock used to produce biogas is made by 50%of energy crops (of which 50% maize, 40% other cereals and 10% other crops), 40% of agricultural residues (mainly manure) and 10% of biogenic waste (household waste, industrial biological waste);
maize yield is equal to 50 tonnes of fresh matter per hectare(tFM/ha);
storage losses of maize in silos are equal to 12%;
the dry matter (OM) content in maize is 35%, of which 96% is organic dry matter (ODM);
biogas yield is 0.6 m3 per kg of ODM;
methane content in biogas is 52%

Section S2

Description of the PE model

Part of the model dealing with crops

The set of 21 crops included in the Luxembourgish agricultural system (Table S4) were divided into those that can undergo land use change (CG) and those that cannot (SC). Those crops that are classified as SC cannot undergo land use changes (LUCs) on account of regulatory requirements, or based on the fact that they are grown in extreme weather and land specific conditions that are not economically feasible to grow other crops. In order to implement the dynamic sets in GAMS, an additional set NS(CG) is defined that can be set to a binary “yes” or “no”. It is initialised to “yes”, thus allowing the crops under this set to undergo LUCs. Finally, a dummy crop is also introduced in the decision set so as to reproduce the base case. This dummy crop has 0 hectares under cropping and remains at zero throughout all scenarios.

Table S4 – Crops set details.

Crop / Crop name / LUC
Cr1 / Wheat humans / CG(C)
Cr2 / Wheat animals / CG(C)
Cr3 / Spelt / CG(C)
Cr4 / Rye / CG(C)
Cr5 / Barley winter / CG(C)
Cr6 / Barley spring / CG(C)
Cr7 / Oats C / CG(C)
Cr8 / Mixed-grain / CG(C)
Cr9 / Grain-maize / CG(C)
Cr10 / Triticale / CG(C)
Cr11 / Other-forage-crops / CG(C)
Cr12 / Maize-dry-matter-BG / CG(C)
Cr13 / Dried pulses / SC(C)
Cr14 / Beans / SC(C)
Cr15 / Potatoes / SC(C)
Cr16 / Rapeseed / CG(C)
Cr17 / Other-crops / SC(C)
Cr18 / Meadows / SC(C)
Cr19 / Pastures / SC(C)
Cr20 / Vineyards / SC(C)
Cr21 / Crops-NES* / SC(C)
Cr22 / Dummy crop / CG(C)
*Crops Not Elsewhere Specified

As mentioned above, the crops were divided into those that are legally permitted to undergo LUCs and those for which any LUCs are forbidden by the regulations in force. Direct cost (v_cost_direct) is the sum of seed (dc_seed), protection (dc_prot) and other miscellaneous costs (dc_other), as expressed in Eq. (1):

(1)

Variable cost (v_cost_variable) is the sum of rental (vc_rent) and variable machine costs (vc_vcmc):

(2)

Fixed cost (v_cost_fixed) is the sum of labor costs (fc_plab) and costs incurred on maintenance of farmland (fc_area) and buildings (fc_bldg):

(3)

Fertilizer cost () is the cost incurred on use of fertilizers. The model treats fertilizer costs separately from other variable costs, due to the response of crop yield to fertilizer use. The logic followed in the model was to endogenize the decision making linked to fertilizer use, based on two variables: yield and cost. A higher use of fertilizers generates a higher yield, albeit with diminishing returns despite the increased cost. There is an optimal use of fertilizers that trades-off the use with the return from incremental crop yield and revenue. Moreover, the fertilizer use by type (NPK) is within lower (lb=0.8) and upper (ub=1.8) bounds (Eqs. 5-7). For crops that do not undergo LUCs there are no changes in the intensity of fertilizer use.Therefore, the variable cost of fertilizer equals the direct cost of fertilizer in the base case:

(4)

(5)

(6)

(7)

wherekg_Nin, kg_Pin, and kg_Kin represent the base case application levels of N, P and K fertilizers respectively.

The functional relationship between the quantity of NPK (kg_NPK) and the expected yield (v_yield_NPK) is expressed by the following equation:

(8)

The benefit per hectare of crop (v_benefit) is given by the difference between the revenue per hectare and the costs:

(9)

The gain (net_gain) for each crop is the benefit per hectare of crop multiplied by the area under cultivation under the crop (new_area), while the total gain (total_gain) from the system is the sum of the gains from all crops.

(10)

(11)

We calculate the new area under crop allocation after the farmers undertake an optimization exercise and impose the constraint that the summation of the new area cannot exceed the total agriculture area of Luxembourg (total_area). The area under each crop in the base case is denoted by acreage(C).

(12)

(13)

The output of crops (output_crop) depends on the yield (t/ha) and the area under cultivation (ha). The output is computed in metric tonnes for the two sets of crops: those that undergo land use change (NS) and those that do not (SC):

(14)

(15)

In order to compute the LUCs the model needs to be shocked by an additional demand for maize for biofuel (Cr12). This additional demand is called toncorn, which is set exogenously to 0 in the base case and 80,000 tonnes in the counter factual:

(16)

In order to compute the change (output_change) in production we compute the output in the base case (output_base) and in the counter factual (output_crop) and take the difference by crop:

(17)

(18)

(19)

The LUCs (change_landuse) by crop are the difference in the land use before and after the shock:

(20)

Part of the model dealing with livestock

The feed for animals (SS_GS), consists of grain, straw, silage and cake (feed), and is a fixed proportion POP(C; feed) of the output of the crop and given by:

(21)

The weight of the straw roughly equals that of grain in many crops. However, only 20% of the straw is used for feeding purposes. Feed from each crop has a certain percentage of dry matter (DM) and this dry matter has proteins (XP) [g/kg of dry matter], metabolic energy (ME) [MJ/kg of dry matter] and net lacto energy (NEL) [MJ/kg of dry matter]. Animal growth is largely a function of the amount of these inputs. The supply of dry matter [], protein [], metabolic energy [] and net lacto energy [] by feed for each crop are given by:

(22)

(23)

(24)

(25)

wherePDM, PXP, PME and PNEL are respectively proportion of dry matter, proteins, metabolic energy and net lacto energy in feed of crop C.

The total supply of dry matter (TS_DM), protein (TS_XP), metabolic energy (TS_ME) and net lacto energy (TS_NEL) by crop is the sum over all feeds for each crop:

(26)

(27)

(28)

(29)

In many cases, there is some additional effort in producing the different feed from the crops.Hence, the final cost of feed by crop [] is computed based on cost escalation [] of producing feed over the normal cost [] of producing the crop.

(30)

The farm operations aim to maximize the profits from milk and meat production. The meat and milk production are a function of the metabolic inputs given via feed of different crops. Feed eventually determines the cost of production of each animal in addition to other miscellaneous costs like veterinary, housing, etc.

The main objective is to determine Q_AFCM(anml,c,feed,MB), where MB ≡{DM,XP,ME,NEL}, is the least cost of feed (feed) from crop (C) with metabolic requirement (MB) for each type of animal (anml) such that it fulfills the minimum metabolic requirements of the animals. If ton_C_feed(anml,C,feed) is the animal feed in tonnes per year of feed from crop C, the dry matter content of this diet and the metabolic contents are given by the following equations:

(31)

(32)

(33)

(34)

The metabolic content of each crop Qty_AC_MB(anml,C,MB) fed to each type of animal is computed as follows:

(35)

The total metabolic content of each crop Qty_C_MB(C;MB) is given by the following equation:

(36)

The total metabolic input MB_anml(anml,MB) to each animal is given by the sum across all feed types and crops of the individual feed in tonnes per year of various crops Q_AFCM(anml,C,feed,MB). The total metabolic inputs of DM, XP, ME, NEL per animal type in a year are given by:

(37)

(38)

(39)

(40)

The total demand by crop for metabolic inputs (DD_DM, DD_XP, DD_ME, DD_NEL) is the metabolic input (DM, XP, ME, NEL) per animal times the number of animals present in the base case BASE_anml(anml). These are given by the following equations:

(41)

(42)

(43)

(44)

In equilibrium the demand should be less than or equal to the total supply of metabolic input (TS_DM, TS_XP, TS_ME, TS_NEL) by crop. These conditions are satisfied in following equations:

(45)

(46)

(47)

(48)

The condition that each animal gets the exogenously specified minimum metabolic requirement Min_Req(anml,MB) was imposed through the following equation:

(49)

This ensures that each animal gets the stipulated metabolic inputs obtained by different feeds from crops that maximize the gains from animal operations. Moreover, it was imposed that the total demand Feed_C_Lim(C, feed) should equal the total supply SS_GS(C, feed) of feed by crops.

(50)

(51)

In order to ascertain the costs and benefits of the animal operations, information is needed on the number of animals existing in the system. In addition, information is necessary regarding their weight, meat and milk production capacity and the prices of meat and milk. Although these prices fluctuate on a weekly basis and slaughtering and milking are not annual phenomena, an annual perspective was assumed, taking the average annual prices for the products. The cost of feed per type of animal Cost_Feed_Anml(anml) is the sum of the costs of various feed of different crops:

(52)

To compute the benefits from the animal operations the related gains should be computed. Milk and meat sales imply gains, while feed costs incur in expenses (which are indirectly lined to crop costs). Value of milk [Value_Milk(anml)] is the value of milk per animal (ValPAnml) times the number of animals producing milk (BASE_ANML):

(53)

A maximum value of meat Value_Meat(anml) is also computed, which is the value of the livestock if all were to be slaughtered or sold and the operations wound-up:

(54)

Similarly the cost of all animals Cost_All_Anml(anml), is the feed cost per animal times the number of animals of each type:

(55)

The net benefit per type of animal Net_Benefit(anml), from animal operations include the benefits from milk and meat minus the cost of feed:

(56)

Only a proportion [N_slaughter(anml)] of animals were assumed to be slaughtered (within lower (lb=0) and upper (ub=0.7) bounds) and the value of meat [V_slaughter)], is based on this number according to Eq. (57):

(57)

(58)

The total benefit [Tot_Net_benefit(anml)] is the net benefit from all animals by selling milk and meat and incurring the feed costs:

(59)

The total gain is the sum of gains from crops and animals. Eq. (62) maximizes the total gain (Total_Gain) subject to the various constraints outlined above:

(60)

(61)

(62)

The model has 9 types of animals, of which six are bovine and three are swine’s. The bovine animals are split according to sex and age with male younger than 1 year (M1), female younger than 1 year (F1), male between 1 and 2 years (M2), female between 1 and 2 years (F2), suckler cows (SCow) and dairy cows (Dcow). The categories of pigs are piglets (PLet), fattening pigs (PFat) and sows (PSow). An average body weight per animal to calculate the carcass weight as 60% of the body weight was assumed. The meat realized is assumed to be 60% of the carcass weight or 36% of the body weight of the animal. Suckler cows and dairy cows are not slaughtered for meat. The prices of meat and milk are average prices as observed in 2009. Table S5 shows the minimum metabolic requirements by each type of animal along with the price of meat and milk in 2009.

Table S5 – Metabolic requirements for animals and animal product prices in 2009.

Animals / DM [kg/day] / ME [MJ/day] / NEL [MJ/day] / XP [gr/day] / Price of meat [€/kg] / Price Milk [€/l] / Number of animals / Weight [kg]
M1 / 4.8 / 77.44 / 0 / 866 / 5.77 / 0 / 20005 / 327
F1 / 4.6 / 70 / 0 / 790 / 5.77 / 0 / 32406 / 294
M2 / 8.84 / 111.71 / 0 / 1237 / 5.41 / 0 / 47710 / 540
F2 / 8.71 / 91.07 / 0 / 984 / 5.58 / 0 / 19257 / 739
Scow / 8.71 / 88.58 / 0 / 951 / 0 / 0 / 32783 / 700
Dcow / 15.17 / 0 / 101.55 / 2281 / 0 / 0.31 / 44310 / 700
Plet / 0.37 / 6.82 / 0 / 40 / 2.09 / 0 / 7395 / 50
Pfat / 2.5 / 29 / 0 / 220 / 1.58 / 0 / 65448 / 120
Psow / 2.5 / 29 / 0 / 600 / 0 / 0 / 7374 / 120
M1= male cows younger than 1 year; F1= female cows younger than 1 year; M2= male cows between 1 and 2 years; F2= female cows between 1 and 2 years; Scow= suckler cows; Dcow= dairy cows; Plet= piglets; Pfat= fattening pigs; Psow= sows; DM= dry matter; ME= metabolic energy; NEL= net lacto energy; XP= proteins.

The animals play an important role in the entire farming system and hence their existence in the model impacts the nature of the results. From Table S6, one can observe that, except for NEL, all other metabolic requirements like DM, ME and XP are in short supply. This means that Luxembourg is falling short of animal feed and is dependent on imports for its animal feed. Since import of animal feed is not considered in the model, it is assumed that any shortfall in feed for any animal is made from import of feed by the farmers. In fact, the farmers are optimizing the total gains from animals and crops and any shortfall is covered by animal feed imports. In the absence of this assumption, the incorporation of the imports of feed by crop and type for each animal would have been needed. Anyhow, a fully endogenous system would still not have been achieved for the animal feed decision making. In the model, an integrated decision (which crops to sow) is taken based on the expected price of the crops and the metabolic characteristics (DM, ME, NEL, XP) of these crops for feed for animals.