Egieya et al. BMC Chemical Engineering (2020) 2:3 BMC Chemical Engineering https://doi.org/10.1186/s42480-019-0025-5
RESEARCH ARTICLE Open Access Optimization of biogas supply networks considering multiple objectives and auction trading prices of electricity Jafaru Musa Egieya1,2, Lidija Čuček3* , Klavdija Zirngast3, Adeniyi Jide Isafiade2 and Zdravko Kravanja3
Abstract This contribution presents an hourly-based optimization of a biogas supply network to generate electricity, heat and organic fertilizer while considering multiple objectives and auction trading prices of electricity. The optimization model is formulated as a mixed-integer linear programming (MILP) utilizing a four-layer biogas supply chain. The model accounts for biogas plants based on two capacity levels of methane to produce on average 1 ± 0.1 MW and 5 ± 0.2 MW electricity. Three objectives are put forward: i) maximization of economic profit, ii) maximization of economic profit while considering cost/benefits from greenhouse gas (GHG) emissions (economic+GHG profit) and iii) maximization of sustainability profit. The results show that the economic profit accrued on hourly-based auction trading prices is negative (loss), hence, four additional scenarios are put forward: i) a scenario whereby carbon prices are steadily increased to the prevalent eco-costs/eco-benefits of global warming; ii) a scenario whereby all the electricity auction trading prices are multiplied by certain factors to find the profitability breakeven factor, iii) a scenario whereby shorter time periods are applied, and investment cost of biogas storage is reduced showing a relationship between cost, volume of biogas stored and the variations in electricity production and (iv) a scenario whereby the capacity of the biogas plant is varied from 1 MW and 5 MW as it affects economics of the process. The models are applied to an illustrative case study of agricultural biogas plants in Slovenia where a maximum of three biogas plants could be selected. The results hence present the effects of the simultaneous relationship of economic profit, economic+GHG profit and sustainability profit on the supply and its benefit to decision-making. Keywords: Biogas production, Auction trading prices of electricity, Supply network optimization, Multiple objectives, Economic profitability
Introduction more recent study in Hegnsholt et al. [3] shows that on In December 2015, over 190 countries across the globe average 33% of total annual worldwide food production acceded to employing activities and technologies that (1.6∙109 t/y) costing 1.2∙1012 $/y goes to waste. This minimize the effects of global climate change [1]. massive loss of food is a subject of concern; it is un- Among the technologies considered is to increase the acceptable, and severely hampers the United Nations’ utilization of biomass-derived energy sources (also called Sustainable Development Goals target to cut the food bioenergy). Furthermore, in Yue et al. [2], utilising bioe- loss and waste by 50% in 2030 [4]. nergy has the potential to: bolster energy security in Consequently, against this backdrop, some countries have economies not having fossil energy sources; mitigate the laid down policies to accelerate the increased integration of effects of variable fossil energy prices and availability; bioenergy in their economy. For example, the national gov- improve waste management concerning exploiting food ernment of India in 2009 adopted a policy to produce about wastes to produce bioenergy thereby creating wealth. A 14∙105 t/y of biofuels to meet 20% blending of biofuels used in the transportation fuels by the year 2020 [5]. It is note- * Correspondence: [email protected] worthy that this policy only considers the use of non-edible 3Faculty of Chemistry and Chemical Engineering, University of Maribor, feedstock retrieved from lands unsuitable for agriculture to Smetanova ulica 17, 2000 Maribor, Slovenia prevent food versus fuel conflicts. In another case, the Full list of author information is available at the end of the article
© The Author(s). 2020 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 2 of 23
government of Ghana enacted a biofuels policy to substi- of food crops. However, the 3G feedstocks usually tute petroleum fuels with 10% biofuels by 2020 and a fur- referred to as algae, on average produce more en- ther increase to 20% biofuels incorporation by 2030 [6]. ergy per area than any other generation of feed- The Ghana government included additional policies di- stocks but today, 3G feedstock are only used on a rected to exploiting energy from wastes such as municipal, laboratory scale. industrial and agricultural wastes [6].IndiaandGhanaare ii. Production/Conversion Technologies – There generally described as developing countries but in other sit- are three groups of biomass conversion uations, developed countries have also instituted policies to technologies currently in use today and these are accelerate bioenergy exploitation. The USA, through the thermo-chemical conversion (direct combustion, li- Energy Independence and Security Act (EISA) of 2007, sets quefaction, gasification, pyrolysis), physico-chemical an annual production target of 116∙106 t/y of biofuels by conversion (transesterification) and biochemical 2022 [2]. Moreover, the EU in another case aims to substi- conversion (anaerobic digestion, fermentation, tute 10% of the transportation fuel in all EU countries with composting). biofuels by 2020 [7]. Alternatively, it is interesting to note iii. Demand centres – These centres are usually the that China places bold targets to harness biomass energy locations whereby the bioenergy products are either from disparate sources implementable on five-year plans. blended with conventional fossil fuels for onward For example, its current policy (13th five-year spanning use or directly consumed. 2016–2020) indicates that on or before the year 2020, the rate of utilization of energy derived from biomass should As stated earlier, these bioenergy products must be de- exceed 58∙106 t/y of standard coal while biogas employed livered to meet certain demands in a timely fashion. The for cooking should reach 80∙106 m3 and electricity gener- demand constraints placed on bioenergy may be set on ated from the same biogas should be at least 500 MWe [8]. economic, environmental and/or social objectives or With these policies in place, implementing them has each of the individual objectives [2]. Hence, as a back- presented some challenges. In Hegnsholt [3], designing, drop of the above context, the following section intro- modelling and effecting robust supply chains is sug- duces some of the recent works carried out in the gested as a veritable tool required to bolster bioenergy modelling and optimization of bioenergy or biogas pro- integration in any economy. In effect, all bioenergy pro- duction supply chains which is the focus of this work. duction supply chains consist of several actors (farmers/ waste collection and acquisition centres, production/ Review of literature on biogas/bioenergy supply conversion facilities and different demand zones) which chain optimization are constantly interacting [9]. These actors are usually In recent years, there has been a considerable increase present in different geographical locations which neces- in research on bioenergy supply chain optimization. sitate producing bioenergy products in a timely fashion These studies have generally been geared towards indi- to meet certain demands. A closer examination of these vidually meeting economic, environmental, social ob- actors shows the following: jectives or a combination of the objectives on certain timescales. For example, El-Halwagi et al. [12]simul- i. Farmers/waste collection centres – Usually taneously modelled the minimization of risk (as a contain feedstocks that are described as first- metric of the social objective) and total annual cost generation (1G), second-generation (2G) and third- (TAC) in the supply chain of biorefining system applied generation (3G) [10]. The 1G feedstocks (starch-, to bio-hydrogen production. Zirngast et al. [13]pro- sugar- and oil-based) are edible foods which have posed four-step methodology for flexible supply net- led to the rise in prices of food due to competition work synthesis under uncertainty applied to biogas in accessing limited resources (like land) to produce production where economic, eco- and viability profits bioenergy products [1]. Today, majority of bioe- were maximized. Emara et al. [14] developed a MILP nergy investments are based on 1G feedstocks (such model, using C#, MATLAB and Excel Solver, to as sugarcane, corn and palm oil) whereby most minimize the TAC in the supply chain of biofuel and commercialized bioenergy production technologies chemicals from waste cooking oil. Ivanov et al. [15] also utilize the 1G feedstocks [11]. On the other researched the supply chain production of bioethanol hand, 2G feedstocks are those which contain ligno- from 1G and 2G feedstocks whereby TAC and green- cellulosic and waste materials (like manure, munici- house gas (GHG) emissions are minimized and the pal wastes, straw or even bagasse) and unlike the number of jobs created maximized. The economic and 1G feedstocks, these feedstocks are not edible. They environmental optimization of biogas supply chain have enough potential to produce bioenergy while (maximization of annual profit and GHG emission sav- simultaneously not affecting the cost and availability ings) has been performed using MILP optimization Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 3 of 23
approach and applied for a region in Mexico by Díaz- Problem statement Trujillo et al. [16]. A holistic supply network management comprising several Several optimization studies have been carried out agricultural feedstocks (with different harvesting periods, over hourly, daily, monthly and yearly timeframes. For availability and prices), transport modes, conversion tech- instance, Egieya et al. [17] modelled a multi-month and nologies, and products with various prices including multi-year MILP model for bioelectricity supply chain hourly-based electricity prices, was considered. Egieya production in Slovenia. Mousavi Ahranjani et al. [18], in et al. [28] introduced the single objective function of eco- a more recent study, developed a fuzzy programming nomic profit maximization which has been opined in sev- model for bioethanol supply chain network design over eral quarters to be unsuitable for realistic complete a 10 years’ planning horizon in Iran. Čuček et al. [19] analysis and synthesis of bioenergy supply chains/net- presented an optimally integrated supply chain network works. In this respect, for the optimal design of biogas to produce bioenergy from 1G, 2G, and 3G feedstocks supply networks, other optimization criteria such as maxi- monthly using a MILP model. Besides, Egieya et al. [20], mizing unburdening from GHG emissions (maximizing in another study, employed a MILP model which pro- the profit from GHG emission unburdening) and includ- poses hourly, daily, and monthly generation of bioelec- ing the sustainability profit maximization are therefore ad- tricity from biogas in Slovenia. dressed, see Fig. 1. Some economic objective-based studies involve the Hence, the design problem entails the problem minimization of TAC [15], others emphasize maximizing statement given in Egieya et al. [28] and the following profit [21], while a few others maximized the net present additions: value(NPV)ofasupplychain[22]. Concerning environ- mental objectives, minimizing global warming potential, ex- eco-costs [29] of feedstocks during harvesting and emplified by limiting GHG emissions, is receiving the most collection; attention [23]. Several supply chain optimization studies eco-benefits [30] of feedstocks use; performed optimization of economic-environmental parts eco-costs/eco-benefits of intermediate and final while social sustainability is less commonly addressed. It is products; the least understood sustainability pillar, thus also called eco-costs of transport modes; “missing pillar” [24]. Social pillar is often qualitative by na- GHG emissions of feedstocks during harvesting and ture and it is challenging to build a single metric for social collection; sustainability and to incorporate it into mathematical GHG emissions of intermediate and final products models. Some work has been done recently, e.g. by You generation; et al. [25] who maximized the number of jobs created in a GHG emissions due to transport; bioenergy production supply chain, El-Halwagi et al. [12] avoided GHG emissions due to harmful feedstocks who modelled safety as a metric of the social objective and use and substitution of products [30]; Zore et al. [26] who optimized social profit from various social costs and profits [26]. micro- and macroeconomic perspectives [27]. From the previous studies, it is found that only a The objectives of the upgraded model are to few researchers considered shorter time-periods, such maximize economic profit, maximize economic profit as hourly time periods, while the bulk of bioenergy while including costs and benefits due to released supply chain optimization studies have concentrated and avoided GHG emissions in the biogas supply on bioethanol and biodiesel production, with a limited network on the other hand, and to maximize the few on biogas supply chains. In this work, beyond sustainability profit. Four scenarios are performed to what has previously been done to the best of our improve the applicability of the biogas supply net- knowledge, two biogas plant capacities (on average of work, such as: about 1 MW and 5 MW capacity of electricity produc- tion) are considered, and an optimization is per- Increasing the price for GHG emissions from formed based on different economic and sustainability approximately 20 €/t (as of 24 February 2019 [31]) objectives simultaneously accommodating hourly, daily or 26.6 $/t by using the conversion rate of 1.33 $/€ and monthly optimization basis and auction trading (as used in [17, 20, 28] to the value of eco-costs / prices of electricity. An additional input to this study benefits of global warming which is 116 €/t [32]or is the integration of biogas storage to enable simul- 154.28 $/t (based on considered conversion); taneous electricity production at higher prices with Multiplying the values of auction trading prices by biogas storage at low electricity prices. Moreover, certain factors to obtain the prices where biogas model size reduction techniques are implemented to production becomes economically profitable. This shorten the computational time of each model. scenario could provide answers to how much Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 4 of 23
Fig. 1 Biogas supply network considering multiple objectives (modified from [28])
subsidies may be needed to make the biogas slightly extended to include new economic, environmen- production plant profitable in a case where GHG tal and social sustainability objectives. emission unburdening is not considered; Shortening the length of the time period which will enable higher differences in electricity prices. Methodology Additionally, this scenario shows the relation This work follows the concept put forward by Egieya between the investment cost of biogas storage, the et al. [28] while considering the following additions and volume of biogas stored and electricity production. extensions: Varying the capacity of the biogas plant capacity from 1 MW to 5 MW while observing its effects on The model is formulated on an hourly basis economic profits. (previously on monthly basis in Egieya et al. [28]), where the year is divided into monthly (mp), daily It follows that the variables to be optimized are: (dp) and hourly (hp) time periods. Consequently, all the equations which were based on monthly periods, quantity, geographical location and total acquisition are now delineated to monthly, daily and hourly cost of feedstocks and/or raw materials; periods. To implement this, certain model reduction cost incurred in the supply chain and types of techniques are therefore introduced to reduce transport modes selected; computational time. other supply network management costs Instead of subsidized prices of electricity (fixed), (depreciation, maintenance, operating, storage…); hourly-based auction trading prices of electricity are primary and secondary conversion facilities location considered based on 2017 prices, ranging from − and capacities; 42.93 to 199.00 €/MWh (between − 57.1 and 264.67 the sustainability profits effect on the supply $/MWh) [33]. The highest electricity price was in network; August, while the lowest price was in December global warming (GHG emissions) effect on the 2017. The hourly-based electricity price variations supply network; are illustrated in Figs. 2 and 3 for the months of impact of the solution on the profit maximization; August and December 2017. All the data related trade-offs when choosing different objectives; to electricity prices (in €/MWh) as obtained from subsidies required to obtain economic break-even BSP South Pool Energy Exchange [33]are point for biogas production. presented in Additional file 1: Tables S1-S12). Furthermore, the average electricity prices (in The general model (MILP problem) discussed in $/MWh) for each of the considered period based Egieya et al. [28] is also applicable to this study which is on model reduction techniques and implemented Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 5 of 23
Fig. 2 Hourly-based electricity prices for August 2017 (data obtained from [33])
in the model are also given in the Additional file benefits attributed to GHG burdening and 1:TablesS13-S24). unburdening. Hence, the model is upgraded to Biogas storage is incorporated to account for include environmental (GHG emissions) and possible variations in electricity production, i.e. to sustainability (eco-cost and benefit and social cost enable storing biogas instead of electricity and benefit) objectives. production at low electricity prices. However, in such situations, heat is also not produced, and thus Description of biogas supply network a backup is required to generate heat from other The biogas supply network utilized (see Fig. 4) consists sources. of four layers: Instead of considering only one agricultural biogas production plant as the optimal plant and with the i) First layer (L1): harvesting and collection. This capacity of up to 999 kW, a maximum of three layer consists of a set pb of biomass feedstocks biogas plants could be selected. Despite the (corn, wheat and triticale grains, straw, silage, and variations in electricity production, biogas grass silage) and different manure types (cattle, pig production should be constant with slight variations and poultry manure, poultry bedding and poultry allowed. Thus two scenarios are performed based on slurry). For the feedstocks, characteristics such as the demand for methane, i) between 1.95∙106 and dry matter and methane contents and biogas yields 2.38∙106 m3/y (average 0.9–1.1 MW of electricity [34] are considered in the study. produced) and ii) between 9.76∙106 and 11.93∙106 ii) Second layer (L2): primary processing technology m3/y (average 4.8–5.2 MW of electricity produced). which is anaerobic digestion. In L2, the primary Two additional objectives are considered besides an conversion product pi (a combination of biomass economic one in the form of maximizing and waste feedstocks pb, recycled products poutpim sustainability profit [26] and the simultaneous and purchased products pbuy) is generated. These maximization of profit with the costs and the are later converted to intermediate products pm
Fig. 3 Hourly-based electricity prices for December 2017 (data obtained from [33]) Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 6 of 23
Fig. 4 Four-layer biogas supply network applied in this study (after [14, 28])
(biogas and wet digestate) or final products pd electricity generated from the CHP and water from the using given conversion factors. dewatering plants to be reused within the supply network. iii) Third layer (L3): secondary conversion For sustainable supply of all materials within the supply technologies involve cogeneration (CHP) combining network, four storage facilities are also modelled at the lo- heat and power production and physical dewatering cations of biomass and waste collection centres and pri- as in [28]. It should be noted that there are other mary and secondary conversion facilities, where all possible conversion technologies, such as biogas feedstocks and products could be stored. Additionally, it is upgrading to biomethane [35], ammonium sulfate assumed that water, electricity, and heat are excluded from recovery from digestate [36] and several other, storage and that the purchased materials should not be however they have not been considered in this stored. Note that from the previous work of Egieya et al. study. The products pz (a sum of intermediate [28] biogas could additionally be stored. products pm, recycled product poutpin, and Similarly, as in Egieya et al. [28], certain characteristics purchased products pbuy) are converted (using of biomass and waste feedstocks are considered, such as conversion factors) to the desired products pp different dry matter contents, methane contents and bio- (electricity, heat and dry digestate). gas yields [34]. Also, other parameters as presented in iv) Fourth layer (L4): demand locations. Egieya et al. [28] are considered, except instead of guar- anteed purchase prices which are fixed, auction trading The model considers three optional distribution modes prices which vary hourly are considered. between the layers to convey feedstocks, intermediate and For more details on the biogas production supply net- final products, in the form of road, pipeline transport, and work methodology, the reader is referred to the paper by transmission lines. Besides, the model allows heat and Egieya et al. [28]. Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 7 of 23
Description of mathematical model monthly periods mp (PRi, pb, mp, dp, hp in kt/period), see The mathematical model includes material and energy also Eq. (7) in Egieya et al. [28]: balances, primary and secondary conversion constraints X X and cost correlations. However, as the model now con- ∧ PRi;pb;mp;dp;hp ðÞdp;mp ∈DPM siders hourly production, all the variables, and equations dp⊆DP hp⊆HP which were based on monthly periods, are now based on ¼ HYi;pb;mp ; ∀ ∈ ; ∈ ; ∈ monthly, daily and hourly periods. Á Ai;pb;mp i I pb PB mp MP ð3Þ As the hourly-based model is computationally expen- sive, certain model reduction techniques have been im- where HYi, pb, mp is the yield of feedstocks pb in month 2∙ plemented based on the work by Lam et al. [37]to period mp at harvesting zone i (in kt/(km month)) and reduce computational time. Hence, instead of 24 h a day, Ai, pb, mp is the available area for growing biomass pb at 2 three “hourly periods” or shift periods (morning, after- harvesting zone i in month period mp (in km ). “ noon and night) are considered and are thereby defined The equations for storages additionally consider cir- ” as H1 (7 am – 2 pm), H2 (3 pm – 10 pm) and H3 (11 pm cular operations . The equation for storage at the inlet – 6 am). Furthermore, instead of 28–31 days a month, of primary conversion facilities is for example defined as seven “daily periods” are applied based on the days of shown in Eq. (4), see also Eq. (9) in Egieya et al. [28]. – the week (Monday Sunday) and are defined as D1: L2 ∪ L2 Ainm;pi;mp;dp;hp ¼ Ainm;pi;mp−−1;dp−−1;hp−−1þ ðÞmp ;k¼1∧ðÞdp ;k¼1∧ðÞhp ;k¼1 {d1, d8, d15, d22, d29}, D2: {d2, d9, d16, d23, d30}, D3: {d3, k k∈K k k∈K k k∈K L2 L2 ∪ Ain ; ; ; ; − þ ∪ Ain ; ; ; − ; −− þ ; > m pi mp dp hp 1 ; > ∧ ; m pi mp dp 1 hp 1 d10, d17, d24, d31}, D4: {d4, d11, d18, d25}, D5: {d5, d12, d19, ðÞhpk k∈K k 1 ðÞdpk k∈K k 1 ðÞhpk k∈K k¼1 L2 ∪ Ain ; ; − ; −− ; −− þ d26}, D6: {d6, d13, d20, d27} and D7: {d7, d14, d21, d28}, see ; > ∧ ; ∧ ; m pi mp 1 dp 1 hp 1 ðÞXmpk kX∈K k 1 ðÞdpk k∈K k¼1 ðÞhpk Xk∈K k¼1 X L1;L2;net L3;L2;net also Egieya et al. [20]. This is due to different electricity Fi;m;pb;mp;dp;hp þ Fn;m;poutpim;mp;dp;hpþ i∈XI pb⊆PI n∈XN poutpim⊆PI X X consumption patterns of the weekdays and weekends. buy;L2 − L2;T − L1;L4;net Fm;pbuy;mp;dp;hp Fm;pi;t;mp;dp;hp Fm; j;pn;mp;dp;hpþ ⊆ ∈ ⊆ All 12 months of a calendar year are on the other hand pbuy PI ðÞpi;t ∈PIT∧t2∈T j J pn PI L2 L2 ðAin ; ; ; ; þ ∪ Ain ; ; −− ; −− ; −− þ fully considered in order to preserve the variability of m pi mp dp hp ; ∧ ; ∧ ; m pi mp 1 dp 1 hp 1 ðÞmpk k∈K k¼1 ðÞdpk k∈K k¼1 ðÞhpk k∈K k¼1 L2 L2 the model as much as practicable. Merging of time pe- ∪ Ain ; ; ; ; − þ ∪ Ain ; ; ; − ; −− þ ; > m pi mp dp hp 1 ; > ∧ ; m pi mp dp 1 hp 1 ðÞhpk k∈K k 1 ðÞdpk k∈K k 1 ðÞhpk k∈K k¼1 L2 riods is done by defining the sets MPOM, DPOD and ∪ Ain ; ; − ; −− ; −− þ ; > ∧ ; ∧ ; m pi mp 1 dp 1 hp 1 ðÞmpk k∈K k 1 ðÞdpk k∈K k¼1 ðÞhpk k∈K k¼1 HPOH which convert the maximal number of time pe- ∪ L2 = ψ Ainm;pi;mp−1;dp−−1;hp−−1Þ 2 Á pi;mp;dp;hp ðÞmp ;k>1∧ðÞdp ;k¼1∧ðÞhp ;k¼1 riods (mpo, dpo and hpo) to merged time periods (mp, k k∈K k k∈K k k∈K ∀m∈M; pi∈PI∧pi∉NOSTOR; mp∈MP; dp∈DP; hp∈HP; ðÞdp; mp ∈DPM dp and hp). All the prices except electricity prices are considered ð4Þ at merged hourly basis as shown in Eq. (1): L2 X In Eq. (4) Ainm;pi;mp;dp;hp represents the storage quantity ∧ Pp;mpo of material pi in each monthly mp,dailydp and hourly time ðÞmpo;mp ∈MPOM mpo∈MP period hp at the location of primary conversion facility m, ; ; ; X ; Pp mp dp hp ¼ L2 ∧ j mpo j Ainm;pi;mp−−1;dp−−1;hp−−1 refers to the quantity of material pi ðÞmpo;mp ∈MPOM mpo∈MP in the storage tank at the beginning of January (first hour, ∀ ∈ ∧ ∉ ; ⊆ ; ⊆ ; ⊆ ; ; ∈ p P p fgelectricity mp MP dp DP hp HP ðÞdp mp DPM first day and first month) which equals the quantity of ma- ð1Þ terial pi in the storage tank at the last hour of December where ∧ stands for logical condition (dollar operator in (last hour, last day, last month) of the previous year. Simi- L2 GAMS [38]). larly, Ainm;pi;mp;dp;hp−1 refers to quantity of material pi in As electricity prices are provided on hourly basis, they the storage tank for each month and day where the hour L2 are averaged in order to more properly account for their shouldnotbethefirsthouroftheday,Ainm;pi;mp;dp−1;hp−−1 variations. Averaging electricity prices is illustrated in refers to the quantity of material pi in the storage tank for Eq. (2): each first hour of the day and if the day is not the first day of the month and AinL2 refers to quantity Pelectricity;mp;dpX;hp ¼ X X m;pi;mp−1;dp−−1;hp−−1 ∧ Pelectricity;mpo;dpo;hpo ; ∈ ; ; ∈ ; ; ∈ ; ; ∈ ∈ ðÞmpo mp MPOM ðÞhpo hp HPOH ðÞdpo dp DPOD ðÞdpo mpo DPM Xmpo MPO dpo∈DPO hpo∈HPO X X X ; of material pi in the storage tank for each first hour in a ∧ j mpo jÁ ∧ j dpo jÁ ∧ j hpo j ðÞmpo;mp ∈MPOM ðÞdpo;dp ∈DPOD ðÞhpo;hp ∈HPOH day and for each first day in a month of any given month mpo∈MPO dpo∈DPO ðÞdpo;mp ∈DPM hpo∈HPO ∀mp⊆MPO; dp⊆DPO; hp⊆HPO except January (first month). ð2Þ L1;L2;net Additional terms in Eq. (4)are: Fi;m;pb;mp;dp;hp represents The hourly-based variations in the model have been the net quantity of biomass and waste feedstocks pb shipped introduced with the production rate of feedstocks pb at to the primary conversion location m from the harvesting the harvesting zone i, which is now defined based on location i in each considered time period (mp, dp, hp), L3;L2;net “ ” merged hourly periods hp, merged daily periods dp and Fn;m;poutpim;mp;dp;hp is the net flow of recycled material in Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 8 of 23
thesupplynetworkpoutpim between the secondary n and selected at mth location while the AD is not selected at primary conversion location m, also for each considered that location when the binary variable equals to 0. time period. Such products are electricity, heat and water, as buy;L2 showninFig.4. Fm;pbuy;mp;dp;hp stands for the quantity of j mpo j j dpo j j hpo j Dem ; ; ; ¼ cap Á f Á Á Á ; purchased resources pbuy to be used at L2 (primary conver- electricity mp dp hp time j mp j j dp j j hp j sion) on the location of m in each of the considered time ∀mp∈MP; dp∈DP; hp∈HP; ðÞdp; mp ∈DPM L2;T period. Fm;pi;t;mp;dp;hp is the flow of intermediate product ð8Þ pi ∈ (pb, poutpim, pbuy) from storage to technology t2 ∈ T at primary conversion location m in each time period and L1;L4 where cap is capacity of electricity production (1 or 5 F ; ; ; ; ; quantifies the flow of unprocessed feedstocks m j pn mp dp hp MW) and ftime is the fraction of time when biogas pro- pn ⊆ PI to the demand location j.ThelasttermofEq.(4) duction is operating and is defined as the number of op- represents the losses of stored intermediate materials pi dur- erating hours in a year divided by the total number of ing the storage. Similarly as in Egieya et al. [28], it is as- hours in a calendar year. In this study, value of 0.935 sumed that the amount of stored intermediate products jmpoj jdpoj jhpoj (8192 h/y) is assumed for ftime. The part jmpj Á jdpj Á jhpj (pi ∈ PI ∧ pi ∉ NOSTOR) available in any considered time relate to the number of all periods divided by the total period mp, dp, hp is the average of two consecutive time pe- number of considered periods. ψ riods. Parameter pi, mp, dp, hp represents the deterioration Moreover, various new equations, data and variables ψ rate in storage which is defined on monthly basis pi, mp, have been included in the model to account for the and is then divided by the length of the daily and hourly two additional objectives included in this study com- period (cardinality of sets DP and HP),asshowninEq.(5): pared to the work of Egieya et al. [28]. These equa- ψ tions and variables related to the additional objectives pi;mp ψ ; ; ; ¼ ; ∀p∈P; mp∈MP; dp∈DP; hp∈HP; ðÞdp; mp ∈DPM pi mp dp hp j dp‖hp j are hereby presented in the next section and the data ð5Þ assumed are presented in Additional file 1.Thedata related to GHG emissions are shown in Additional As it was stated above, all the potential biogas plants file 1:TablesS25– S27 and the data related to sus- could be selected. Since only slight variations in capacity tainability profit maximization are given in Additional of anaerobic digesters are allowed, two scenarios are per- file 1:TablesS28– S31. formed based on the demand for methane, i) between 1.95∙106 and 2.38∙106 m3/y (average 0.9–1.1 MW of elec- tricity produced) and ii) between 9.76∙106 and 11.93∙106 Objectives in the study m3/y (average 4.8–5.2 MW of electricity produced). The The goal is to synthesize an optimal biogas supply network capacities of methane between their upper and lower under different objective functions: (i) economic objective bounds are shown in Eq. (6) for lower bound and in Eq. defined with maximizing economic profit (similar to Egieya (7) for upper bound. et al. [20, 28], while excluding the tax on the profit ac- X crued); (ii) economic and environmental objectives by ∧ L2;P ≥ : Demelectricity;mp;dp;hp L2;T ; Fm;pi;methane;AD;mp;dp;hp 0 9 Á ; ; Á ym;AD maximizing economic profit while including costs and ben- ðÞpi;methane ∈PIPM;ðÞpi;AD ∈PIT conv T L3 pi∈PI f methane;electricity;CHP ∀m∈M; mp∈MP; dp∈DP; hp∈HP; ðÞdp; mp ∈DPM efits due to GHG emissions (price for GHG emissions), and ð6Þ iii) economic, environmental and social objectives by maxi- mizing sustainability profit [26] which includes all three X ∧ L2;P ≤ : Demelectricity;mp;dp;hp L2;T ; Fm;pi;methane;AD;mp;dp;hp 1 1 Á ; ; Á ym;AD sustainability objectives, economic, environmental and so- ðÞpi;methane ∈PIPM;ðÞpi;AD ∈PIT conv T L3 pi∈PI f methane;electricity;CHP ∀m∈M; mp∈MP; dp∈DP; hp∈HP; ðÞdp; mp ∈DPM cial (similar as in Bogataj et al. [39]). Further, four add- itional scenarios are performed to improve the profitability ð7Þ of the biogas supply network: i) price for GHG emissions is L2;P increased from the price of carbon allowances in the In Eq. (6)andEq.(7) Fm;pi;methane;AD;mp;dp;hp represents the flowrate of methane produced from material pi using European Union Emissions Trading System (EU ETS) [40] anaerobic digestion AD technology at L2 within time pe- up to the value of eco-costs / benefits of global warming [32], ii) the auction trading prices are increased by multiply- riods mp, dp, hp. Demelectricity, mp, dp, hp stands for the de- mand for electricity in each considered time period (see ing them with various factors, as explained above iii) the ; ; length of time period is decreased and the relation is ex- Eq. (8)). f conv T L3 is conversion factor of me- methane;electricity;CHP plored between the price of biogas storage and capacity of thane to electricity using technology CHP, and the binary biogas storage and electricity production and (iv) increasing L2;T variable ym;AD represents the selection of technology AD the biogas plant capacity from 1 MW to 5 MW and observ- at location m. If the binary variable equals 1, AD is ing its effects on economic profit. Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 9 of 23
Economic objective Economic and environmental objectives (economic+GHG The economic objective is defined with maximizing eco- profit) nomic profit (PEconomic) from the generation of electri- The second objective includes economic objective and city, heat and digestate within the biogas supply chain priceforGHGemissions.Theeconomicobjectiveand network: equations describing it in detail are shown above (Eqs. (9)–(11)). The environmental objective is defined as a PEconomic ¼ RTotal−CTotal ð9Þ maximization of GHG unburdening and it is based
Total Total and extended from the work of Bogataj et al. [39]. where R is total revenue accrued ($/y) and C is Avoided and released GHG emissions (unburdening total cost incurred in the supply chain ($/y). Total and burdening) are multiplied by the price of GHG Thetotalrevenue(R ) is calculated as shown in emissions (also called carbon price [41]) and included Eq. (10): in the economic objective. Both burdening and unbur- X X X X X X Total ∧ L2;L4;net dening are considered, whereby burdening is related R ¼ Fm; j;pd;mp;dp;hp Á Ppd;mp:dp;hpþ ðÞdp;mp ∈DPM Xm∈MXj∈ J Xpd∈PDXmp∈MPXdp∈DPXhp∈HP to the negative impacts on the environment due to re- ∧ L3;L4;net Fn; j;pp;mp;dp;hp Á Ppp;mp;dp;hpþ ðÞdp;mp ∈DPM source use, production and use of products, while Xn∈N Xj∈ J ppX∈PP mpX∈MP dpX∈DP hpX∈HP ∧ L1;L4;net unburdening is due to the direct utilization of harmful Fm; j;pn;mp;dp;hp Á Ppn;mp;dp;hp ðÞdp;mp ∈DPM m∈M j∈ J pn∈PN mp∈MP dp∈DP hp∈HP (waste) materials and due to substitution of environ- ð10Þ mentally more harmful products with less harmful ones [42]. L2;L4;net where Fm; j;pd;mp;dp;hp represents the net flowrate of direct The environmental objective follows the same product pd (wet digestate) produced from anaerobic di- principle of evaluation as the eco-profit calculation gestion at site m and sold as a fertilizer in site j to [30]. First, avoided and released GHG emissions are L3;L4;net farmers at each considered time period, Fn; j;pp;mp;dp;hp calculated with units based on t CO2 equivalent emit- stands for the net flow of produced products pp (electri- ted per t of raw material or product except for electri- city, heat and dewatered digestate) from the plant n to city and heat which are in t CO2 equivalent emitted L1;L4;net per MWh. The Life Cycle Assessment (LCA) principle demand j. F ; ; ; ; ; represents materials that do not m j pn mp dp hp is applied to the biogas supply network from the har- undergo any treatment (pn) shipped directly to the de- vesting and collection zones to the demand zones. mand zone in site j. P , P and P pd, mp. dp, hp pp, mp, dp, hp pn, GHG emissions include those emissions that originate are prices of direct products (pd), produced mp, dp, hp from the whole life cycle of product, from extraction products (pp) and products that do not undergo treat- of raw materials, through pre-processing and process- ment (pn). ing to disposal of harmful products, including emis- Total costs accrued (CTotal) in the biogas supply chain sions due to transportation and distribution within the network are a sum of costs for feedstocks, purchase of supply network (similarly as in [30]). additional materials needed in L2 and L3, shipment (T Hence, the amounts of GHG emitted or preserved (see Total Cp ), storage (SCp), labour (LC), depreciation (DCC), Eq. (12)) is a measure of the difference between GHG maintenance (MC), and miscellaneous cost (MSC)as unburdening (GHGUB) as shown in Eq. (13) and GHG displayed in Eq. (11): burdening (GHGB) in the supply network, presented by X X X X X Eq. (14). Total C ¼ ∧ PRi;pb;mp;dp;hp Á cpb;mpþ ðÞdp;mp ∈DPM Xi∈I pbX∈PB mp∈MPXdp∈DPXhp∈HPX UB− B buy;L2 GHG ¼ GHG GHG ð12Þ ∧ F ; ; ; ; Á cpbuy;mpþ ; ∈ m pbuy mp dp hp ∈ ∈ ∈ ∈ ∈ ðÞdp mp DPM X X X X X mXM pbuyXPBUY mpXMP dpXDP hpXHP UB L1;L2;net GHG;UB buy;L3 GHGp ¼ Fm;p;t;mp;dp;hp Á cp þ ∧ F ; ; ; ; Á cpbuy;mpþ ∈ ∈ ∈ ∈ ∈ ; ∈ n pbuy mp dp hp XmXM t TXmp MPXdp DPXhp HP ∈ ∈ ∈ ∈ ∈ ðÞdp mp DPM ; ; ; Xn N pbuy PBUY mpXMP dp DP hp HP FL2 L4 net Á cGHG UB Á f Sþ Total m; j;p;mp;dp;hp p p TCp þ SCp þ LC þ DCC þ MC þ MSC mX∈MXj∈ J mpX∈MP dpX∈DP hpX∈HP p∈P p∈P L3;L4;net GHG;UB S; ∀ ∈ ; ; Fn; j;p;mp;dp;hp Á cp Á f p p fgPB PD PP ð11Þ n∈N j∈ J mp∈MP dp∈DP hp∈HP ð13Þ where, cpb, mp and cpbuy, mp are cost for feedstocks ac- where, cGHG;UB is the GHG emission coefficient related quired (pb) and purchased materials (pbuy). PRi, pb, mp, p to unburdening or avoided GHG emissions (see Add- dp, hp is total quantity of feedstocks harvested at site i S and shipped to storage at primary conversion location, itional file 1: Table S25) for material p and f p is substi- buy;L2 buy;L3 while Fm;pbuy;mp;dp;hp and Fn;pbuy;mp;dp;hp are quantities of tution factor defined as the amount of produced product additional raw materials purchased in L2 and L3 within divided by the amount of substituted product [30]. GHG a given monthly, daily and hourly period. emission coefficients have been obtained from the Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 10 of 23
þGHG website of Delft University of Technology, The Model of PEconomic ¼ PEconomic þ GHG Á pGHG ð15Þ the Eco-costs / Value Ratio (EVR) [32] and checked with GHG OpenLCA software [43] using ecoinvent 3.1 database where p stands for the prices of GHG emissions. [44] and the ecoinvent 3.1 Life Cycle Impact Assessment (LCIA) method [45] IPCC 2007. Economic, environmental and social objectives The following substitution factors are assumed in the (sustainability profit) study: 0.9 for electricity, 0.04 for dry digestate and 0.029 The third objective considers economic, environmental for wet digestate [46]. and social parts which implements the concept of Sus- tainability profit (PSustainability), first proposed in Zore X X X X X Sustainability B L1;L2;net GHG;B GHGp ¼ Fm;p;t;mp;dp;hp Á cp þ et al. [26]. P combines economic, environmen- Xm∈MXt∈T Xmp∈MPXdp∈DPXhp∈HP ; ; ; F L2 L4 net Á cGHG Bþ tal and social indicators into monetary values ($/y). m; j;p;mp;dp;hp p Sustainability mX∈MXj∈ J mpX∈MP dpX∈DP hpX∈HP L3;L4;net GHG;B P is stipulated mathematically (see Eq. (16)) Fn; j;p;mp;dp;hp Á cp þ Economic nX∈N j∈XJ mp∈MPXdp∈XDP hp∈HP X X X X as the sum of Economic profit (P , see Eq. (9)), buy;L2 GHG;B buy;L3 GHG;B Fm;p;mp;dp;hp Á cp þ Fn;p;mp;dp;hp Á cp þ Eco Social m∈M0mp∈MP dp∈DP hp∈HP n∈N mp∈MP dp∈DP hp∈HP 1 Eco-profit (P , see Eq. (17)) and Social Profit (P , X X X X X X @ La;Lb;tropt La;Lb GHG;BA see Eq. (20)): 2 Á F x;y;p;mp;dp;hp;tropt Á Dx;y Á cp;tropt þ x∈fgI;M;N y∈fgM;N; J tropt∈fgroad mp∈MP dp∈DP hp∈HP Sustainability Economic Eco Social L2 L2 L3 L3 GHG;B : ; Ainm;pi;mp;dp;hp þ Aoutm;pm;mp;dp;hp þ Ainn;pz;mp;dp;hp þ Aoutn;pp;mp;dp;hp Á cp Á 0 05 P ¼ P þ P þ P ð16Þ ∀p∈fgPB; PN; PD; PP; POUTPIM; POUTPIN Eco ð14Þ The eco-profit (P )[30] derives from the difference between the sum of all the eco-benefits (EB) and the GHG;B eco-costs (EC) within the biogas supply network: where cp refers to the GHG emission coefficient of material p related to the released GHG emissions (bur- PEco ¼ EB−EC ð17Þ GHG;B dening) (see Additional file 1: Table S26), cp;tropt is GHG Eco-benefit (EB) (see Eq. (18)) is described in monet- emission coefficient related to transport (see Additional ary terms ($/y) as the sum of all positive impacts of ac- file 1: Table S27). In addition, it is worth stating that the tivities/materials which unburden the environment while last section of the Eq. (14) illustrates the GHG emissions the eco-cost (EC) (see Eq. (19)) shows the sum of all occurring during storage of material p over the consid- negative impacts of activities/materials which burden the ered time periods. Hence, in the storage, it is assumed environment [30]. that the burdening equals 5% of the burden of product X X X X X L1;L2;net EB stored. EBp ¼ Fm;p;t;mp;dp;tp Á cp þ La;Lb;tropt mX∈M t∈XT mpX∈MP dp∈XDP hp∈XHP Additional terms in Eq. (14) are: Fx;y;p;mp;dp;hp;tropt L2;L4;net EB S Fm; j;p;mp;dp;tp Á cp Á f pþ shows the quantity of materials p transported from the mX∈MXj∈ J mpX∈MP dpX∈DP hpX∈HP L3;L4;net EB S; ∀ ∈ ; ; Fn; j;p;mp;dp;tp Á cp Á f p p fgPB PD PP location x in layer La to location y in layer Lb with n∈N j∈ J mp∈MP dp∈DP hp∈HP transportation mode tropt at the considered time period ð18Þ La;Lb mp, dp, hp, Dx;y is the distance between object x in EB L2 where c is the eco-benefit coefficient (see Additional layer La and object y in layer Lb, and Ain ; ; ; ; , p m pi mp dp hp file 1: Table S28) of material or energy p ($/kg, $/kWh). AoutL2 , AinL3 and AoutL3 m;pm;mp;dp;hp n;pz;mp;dp;hp n;pp;mp;dp;hp It is worth stating that Eq. (18) is formulated in a similar stand for the quantity of pi stored in the inlet of L2, the way as Eq. (13) but considers eco-benefit coefficients in- quantity of pm stored in the outlet of L2, the quantity of stead of GHG emission coefficients related to avoided pz stored in the inlet of L3 and for the quantity of pp GHG emissions. The same substitution factors are as- stored in the outlet of L3 in the assessed time periods. sumed as previously mentioned in the case of avoided It should be noted that for simplicity of this study, the GHG emissions due to substitution of products. emissions given off during constructing the biogas pro- X X X X X L1;L2;net EC duction plants and pipelines are omitted and as such, ECp ¼ Fm;p;t;mp;dp;tp Á cp þ Xm∈MXt∈T Xmp∈MPXdp∈DPXhp∈HP the equipment are assumed to be used over the entire L2;L4;net EC Fm; j;p;mp;dp;tp Á cp þ Xm∈MXj∈ J mpX∈MP dpX∈DP hpX∈HP lifetime of the plant. This assumption therefore suggests L3;L4;net EC Fn; j;p;mp;dp;tp Á cp þ relatively small contribution of GHG emissions during nX∈N j∈XJ mp∈MPXdp∈XDP hp∈HP X X X X ; ; Fbuy L2 Á cEC þ Fbuy L3 Á cECþ construction over the plant’s lifetime. m;p;mp;dp;tp p n;p;mp;dp;tp p m∈M0mp∈MP dp∈DP hp∈HP n∈N mp∈MP dp∈DP hp∈HP 1 The objective which considers economic and environ- X X X X X X @ La;Lb;tropt La;Lb EC A 2 Á Fx;y;p;mp;dp;tp Á Dx;y Á cp;tropt þ mental parts is defined as maximizing the economic x∈fgI;M;N y∈fgM;N; J tropt∈fgroad mp∈MP dp∈DP hp∈HP L2 L2 L3 L3 EC : ; profit while including the multiplication of the released Ainm;pi;mp;dp;hp þ Aoutm;pm;mp;dp;hp þ Ainn;pz;mp;dp;hp þ Aoutn;pp;mp;dp;hp Á cp Á 0 05 and avoided GHG emissions with the prices of GHG ∀p∈fgPB; PN; PD; PP; POUTPIM; POUTPIN þGHG emissions, PEconomic : ð19Þ Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 11 of 23
whereby cEC is the eco-cost coefficient for p ($/kg) (see PSocial ¼XSS þXBJobs−cSocialhi¼ p Jobs Gross− Net s;UNE− s;Country s;Organisation EC N t;mp Á St;mp St;mp þ cmp cmp þ cmp Additional file 1: Table S29) and cp;tropt is eco-cost coef- mp∈MP t∈T ficient related to transport of material or energy p ð20Þ ($/(kg·km), $/(kWh·km)) (see Additional file 1: Table S30). More details on calculation of eco-profit could be The number of new jobs created is given by Eq. (21) found in Čuček et al. [30]. where, LC is labour cost. Eco-cost and eco-benefit coefficients have also been obtained from the website
Fig. 5 Region in case study (after [28])
The optimizations were performed for two capacities of processor with 8 GB installed RAM. The average time methane production and three different objectives (maxi- for completion of each model is about 31 min within mizing economic, economic+GHG and sustainability profits) 532,726 iterations. while considering hourly, daily and monthly time periods. To determine the Economic+GHG profit (see Eq. (15)), three Maximization of economic, economic+GHG and new equations (see Eq. (12) – Eq. (14)) and the data as sustainability profits showninAdditionalfile1:TablesS25– S27 have been Firstly, the models are optimized on the different objectives, incorporated in the model. Concerning optimizing the Sus- maximization of economic, economic+GHG and sustainabil- tainability profit (see Eq. (16)) of the biogas supply network, ity profits. As stated earlier, two capacity levels with speci- five new equations (see Eq. (17) – Eq. (21)andthedatapre- fied demand for methane are considered: i) production of sented in Additional file 1:TablesS28– S31 has been used methane between 1.95 × 106 and 2.38 × 106 m3/y (average in the model. 0.9–1.1 MW of electricity produced), and ii) production of Sincethepricesofelectricityvaryhourlywhile methane between 9.76 × 106 and 11.93 × 106 m3/y (average showing significant differences in prices, additional 4.8–5.2 MW of electricity produced). Table 1 shows the biogas storage is thereby introduced as opposed to main results from the smaller capacity level and Table 2 the work of Egieya et al. [28] which excluded the use shows results from the larger capacity level. of biogas storage. The biogas storage facility suggests Considering the auction trading prices (averaged for 3 an increase in the storage of biogas during periods of periods a day for all days in the week of each month in a low electricity prices while enhancing more electricity year), economic losses are accrued in the objective sce- production at periods whereby the electricity prices narios as shown in Table 1. Furthermore, maximizing are higher. The capital cost of the biogas storage is sustainability profit presents eco-profit values (2,802,824 thus included, and its relationship with capacity is $/y) about 8.7 times more than that accrued when the shown in Additional file 1:FigureS1. economic+GHG profit is maximized (321,269 $/y) but an However, like in Egieya et al. [28], agricultural res- eco-profit loss is incurred when the economic profit is idues could be used as potential feedstocks for bio- maximized. Moreover, the economic+GHG profit is nega- gas production while all the parameters related to tive in all scenarios while the social profit obtained dur- grains (except the agricultural area) are now set to ing the maximization of sustainability (230,344 $/t) is 3 zero. times the social profit value obtained in other scenarios The upgraded models, also called BIOSOM models (77,456 $/y) mainly due to three biogas plants and three [28] containing 12 monthly, 7 daily and 3 hourly time times the number of employees required. periods comprise 327,042 single equations, 516,641 con- Alternatively, the total feedstock used to meet the sus- tinuous variables, and 684 binary variables. The models tainable production of electricity when sustainability profit formulated as MILP are solved using GAMS modelling is maximized (85,200.3 t/y) is about four times the feed- environment and GUROBI solver with 0% optimality stocks used when economic and economic+GHG profit is gap on Intel® Core™ i7–8750 H CPU at 2.20 GHz each maximized (21,948.6 t/y). It is worth stating that Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 13 of 23
Table 1 Main results when maximizing different profits for biogas supply network with smaller capacity (1 ± 0.1 MW in average electricity production) GHG Max PEconomic Max PEconomicþ Max PSustainability b Feedstock (t/y) Corn grain 7488 7488 11,820 Corn silage / / 14,354 Corn stover 5105 5105 8059 Poultry bedding 9355 9355 / Poultry manure 2236 2236 25,505 Poultry slurry / / 25,463 Total feedstock 21,949 21,949 85,200 Products Electricity (MWh/y) 8386 8368 23,968 Heat (MWh/y) 5955 5955 17,378 Digestate (23% dry solids, t/y) 13,353 13,399 48,749 Corn grain (direct product) 7239 7239 11,424 Water (t/y) Purchased 3230 3230 213 Recycled 11,503 11,503 46,438 Utilities required (MWh/y) Electricity “recycled” 531 548 2092 Electricity purchased 185 168 / Heat “recycled” 1700 1700 4968 Heat source purchased // / Sustainability item ($/y) Economic profit − 548,346 −548,452 −2,323,475 Eco-profit − 113,681 321,269 2,802,824 Social profit 77,456 77,456 230,344 Sustainability profit − 584,572 − 149,728 709,692 Economic+GHG profit −500,291 − 498,191 −2,157,453 a CO2 eq. emissions (t/y) − 1807 − 1890 − 6241 Selected Site I I I, II, III Dry matter content (%) 13 13 12.06|11.97|11.91 Methane content (%) 56.18 56.18 55.87|56.36|56.92 aavoided GHG emissions b if not specified for each biogas supply network separated, results present the sum from both networks poultry bedding (9355.3 t/y) is the dominant feedstock obtainable in the economic and economic+GHG profits exploited in the maximized economic and economic+GHG situations (13,398.5 t/y). Similarly, the heat generated in profits scenarios while poultry manure and poultry slurry the maximized economic and economic+GHG profit cases both with an average quantity of 25,500 t/y are the domin- are each about 34% (5954.8 MWh/y) the amount gener- ant feedstocks used in the maximized sustainability profit ated when the sustainability profit is maximized (17, situation. 377.6 MWh/y). In the same manner, the electricity generated in the Considering the emissions given off or avoided, all maximized sustainability case (23,967.7 MWh/y) is about economic, economic+GHG and sustainability profit ob- 3 times that produced in the maximized economic and jective cases present positive (unburdening) effects on +GHG economic profits cases (8368.4 MWh/y). Further- the supply chain and are between 1807 t CO2 eq./y for Sus- more, the dry digestate produced in the maximized P the economic objective case, and 6241 t CO2 eq./y in the tainability scenario (48,748.7 t/y) is 3.63 times the quantity case of maximal sustainability profit. On the other hand, Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 14 of 23
Table 2 Main results when maximizing different profits for biogas supply network with larger capacity GHG Max PEconomic Max PEconomicþ Max PSustainability b Feedstock (t/y) Corn grain 18,838 18,838 18,251 Corn silage / / 1200 Corn stover 12,844 12,844 12,444 Grass silage 11,080 11,244 55,158 Poultry bedding 4808 5635 1026 Poultry manure 49,895 49,656 125,313 Poultry slurry / / 32,413 Total feedstock 97,464 98,216 245,804 Products Electricity (MWh/y) 34,730 34,925 43,376 | 43,437 Heat (MWh/y) 24,662 24,849 51,366 | 51,439 Digestate (23% dry solids, t/y) 76,493 77,106 90,468 | 91,332 Corn grain (direct product) 18,206 18,206 17,639 Water (t/y) Purchased 26,902 26,892 63,624 Recycled 71,981 75,306 738,432 Utilities required (MWh/y) Electricity “recycled” 2197 2283 6968 Electricity purchased 767 703 / Heat “recycled” 7039 7093 16,549 Heat source purchased // / Sustainability items ($/y) Economic profit − 659,523 − 660,736 −3,675,154 Eco-profit − 993,356 1,608,143 6,302,895 Social Profit 147,697 148,441 334,676 Sustainability profit −1,505,181 1,095,848 2,962,417 Economic+GHG profit − 454,124 −441,056 −3,214,145 a CO2 eq. emissions (t/y) − 7721 − 8259 −17,331 Selected Site I I I, III Dry matter content (%) 12.54 12.32 5.90 | 5.93 Methane content (%) 57.55 57.53 55.85 | 55.93 aavoided GHG emissions b if not specified for each biogas supply network separated, results present the sum from both networks
Site I is the optimal location for the construction and obtained in the three anaerobic digestion plants selected operation of the anaerobic digestion plants, dewatering in the sustainability profit objective scenario. plants and CHP in the economic and economic+GHG It is worth noting that the model in all scenarios, opti- profits scenarios while all three locations are selected in mally selects the transportation mode that supports the the sustainability profit case. three objectives. From L1 to L2, road transport by truck Additionally, it could be seen from Table 1 that a 13% is selected. From L2 to L3, both biogas and wet diges- dry matter content is obtained in the economic and eco- tates produced are shipped using pipelines. From L3 to nomic+GHG profits scenario while an average dry matter L4, dry digestate is transported using trucks while elec- content of 11.98% is gotten when the sustainability profit tricity and heat are transported through transmission objective is considered. Moreover, a 56.18% methane con- lines and pipelines. Recycling of water from dewatering tent is accrued in the economic and economic+GHG profits plants and heat from L3 to L2 occurs via pipelines, and objective while an average of 56.38% methane content is electricity through transmission lines. Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 15 of 23
Fig. 6 Breakdown of costs incurred in the three objectives
A breakdown of the costs incurred in the three ob- that water may be reused in the supply network. On jectives (see Fig. 6) shows the dominance of depreci- the other hand, while all cost values in the sustain- ation cost in the three scenarios. The depreciation ability profit objective are substantially higher than cost accounts for a little over a third of the total the other two objectives, the purchased material cost cost incurred in the three supply chains. Investment of the sustainability profit objective is the only cost cost in the maximum economic and economic+GHG attribute lower than those of the other two objectives. profit cases consists of the cost for anaerobic di- Thereasonforthelowvalueofthepurchasedmater- gester (3.14 × 106 $), CHP (1.40 × 106 $) and dewa- ial cost in the sustainability profit objective is because tering (0.23 × 106 $) while biogas holder is not only water is purchased in comparison to the water selected. In case of maximal sustainability profit bio- and electricity purchased in the economic and eco- gas holder is also not selected, while the cost for nomic+GHG profits objectives. threeanaerobicdigestersis9.25×106 $, for CHP A small portion of total expenditure is also due to plants it’s4.19∙106 $ and dewatering is 0.81∙106 $. additional cost in the supply network when the feed- Furthermore, it is seen from Fig. 6 that the cost of stocks are partly transported between the zones. In obtaining extra materials (i.e. purchased materials) this situation and for the three cases, corn stover is contributed a small portion to the total supply net- suggested to be transported between the zones. For work expenditure. Purchased materials in the maximal instance, in the case of maximal economic and eco- economic and economic+GHG cases are only water nomic+GHG profits corn stover is partly transported and electricity used in anaerobic digesters, while in from zone II to zone I as seen in Fig. 7 which illus- maximal sustainability case only water is purchased, trates the areas intended for agricultural feedstocks and this indicates that electricity needed is recycled when maximizing different objectives. It should also inthesupplynetwork(seealsoTable1). One of the be noted that from the total area (250 km2)amax- reasons for the low purchased material costs is the imum of 5% of the actual area (50% in zones I and II relatively low price of water and the pre-condition and 37% in zone III) are set aside for feedstocks
Fig. 7 Use of area for agricultural feedstocks in biogas production considering smaller capacity Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 16 of 23
Fig. 8 Use of area for agricultural feedstocks in biogas production considering larger capacity growing to meet the biogas generation, as previously The Site I is the optimally selected location for an- stated. Hence, in the cases of maximal economic and aerobic digestion plants, dewatering plants and CHP economic+GHG profits about 63.7% of the area is still for the economic and economic+GHG profits objec- unused, while in case of maximal sustainability profit, tives while Sites I and III are now chosen to meet all the available area is suggested to be used for maximal sustainability profit objective. Besides the growing corn grain (and stover) and silage. Corn sto- selection of corn stover and silage as agricultural ver is used as a feedstock for biogas production, while feedstocks, grass silage is also selected because it grains are used for food. may be grown on marginal lands (in parks, river- In all the results obtained as shown in Table 1, biogas banks, road verges, and other lands). For grass silage, storage is not selected, and the electricity was produced itisassumedthatitcouldbegrownonupto50% at constant capacity. The main reasons are the relatively of area present in zones I and II and on up to 37% high investment cost of biogas holders (base case invest- oftheareainzoneIII.Theuseofareaintendedfor ment cost of 701,600 $ and base case capacity of 3000 agricultural feedstocks in biogas supply network for m3 of biogas stored is considered), higher investment the case of larger capacity and three objectives is cost for CHP plant in the case of higher capacity and showninFig.8. The unused area in Fig. 8 is attrib- backup for heat consumed in the anaerobic digestors are uted to the area that may be additionally used for required. grass silage growing while the area that may be used Table 2 further illustrates the results obtained when for crops is fully utilised by corn grains and silage. maximizing different profits in the situation whereby the As in the case of smaller capacity biogas produc- average capacity of electricity production is within the tion, most of the water used for anaerobic digesters range of 4.8 MW and 5.2 MW. come from a recycled stream. For instance, in the In a similar vein as the 1 MW demand capacity sce- case of maximal sustainability profit, the quantity of nario (see Table 1), the total feedstocks in the economic recycled water is 738,431.7 t/y, mainly due to lower and economic+GHG profit objectives each exploit ap- dry matter content (5.9%) in fermenters as opposed proximately the same quantity of feedstocks (98,000 t/y) to approximately 12% dry matter content in the case which is about 40% of the quantity of feedstocks utilized of the smaller plant’s capacity. Most of the electricity in the sustainability profit case (245,804 t/y). used for the plant comes from “recycled” produced
Fig. 9 Digestate storage when maximizing economic profit at higher biogas capacity level Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 17 of 23
Fig. 10 Digestate storage when maximizing economic+GHG profit at higher biogas capacity level electricity of which the maximal sustainability profit Sensitivity analysis for improved profitability of biogas case indicates all the electricity utilized being of re- production newable origin. Also, the heat required for the di- Since the results from all economic, economic+GHG gesters is “recycled” in all the cases considered. and sustainability profits present economic loss (see Furthermore, from Table 2, it is noticed that in Tables 1 and 2), sensitivity analysis is additionally the three cases, economic and economic+GHG profits carried out to investigate possible improvements re- present negative values while on the other hand, so- garding the profitability of biogas production. Four cial profits are positive in all the cases. The eco- and additional scenarios are hereby put forward: i) the sustainability profits are negative when economic scenario whereby in calculating the economic+GHG profit is maximized but gives positive values in the profit, the prices of GHG emissions are steadily in- cases of maximal economic+GHG and sustainability creased from 26.6 $/t (or 20 €/t) (allowances in EU profits. In the maximal economic profit scenario, it ETS [40]) up to 154.28 $/t (or 116 €/t) (eco-costs of is suggested that much more digestate is stored in global warming [32]), ii) the scenario whereby all the comparison to the digestate stored when econom- electricity auction trading prices are simultaneously ic+GHG profit is maximized (see Figs. 9 and 10), and multiplied by certain factors to know how much thus the burden related to the digestate storage is subsidies should be included on top of market prices higher in the case of maximal economic profit of electricity to make it profitable, iii) the scenario (3.719∙106 $/y vs 1.732∙106 $/y). Figure 9 hence whereby for a selected month where only the hours shows digestate storage when economic profit is are merged into 3 shift periods the cost of biogas maximized, while Fig. 10 shows digestate storage storage is reduced in order to study the relationship when economic+GHG profit is maximized, both at the between the biogas storage cost and biogas storage higher capacity of biogas supply networks. However, capacity and electricity production based on averaged in the case of maximal sustainability profit, digestate auction trading prices of electricity, and (iv) the sce- storage is comparably lower (see Fig. 11), since two nario in which biogas plant capacity is varied from plants are selected. 1 MW to 5 MW to study how capacity affects the
Fig. 11 Digestate storage when maximizing sustainability profit at higher biogas capacity level Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 18 of 23
Fig. 12 Effects of GHG emissions price variation on Fig. 13 Effects of electricity auction trading prices variation on economic+GHG profit economic profit economic profit (loss). In a situation of the relatively economic profit (lower biogas capacity level). For in- high cost of biogas storage, biogas storage was never stance, in the base case scenario where there is no selected, and the electricity was always produced at change in auction trading price, a negative economic constant capacity. profit (approximately of 548,346 $/y economic loss) is incurred as shown in Table 1. In the instance, Maximizing economic+GHG profit while increasing prices of when the auction trading price is doubled, a break- GHG emissions even value is obtained in the biogas supply network. Figure 12 illustrates the effects of steadily increasing Moreover, an increase beyond the doubled auction the GHG emissions prices on the economic+GHG trading price presents positive economic profit profit (see Eq. (15)) when the average capacity of 1 values. As a backdrop of this, it could be concluded MW of electricity production is considered. At 20 that for the biogas supply network to be profitable, €/t GHG emission price (26.6 $/t with the consid- either subsidy in the amount of auction trading ered conversion rate), economic+GHG profit of prices should be provided or the carbon benefit approximately − 498,191 $/y (see also Table 1)isob- should be at least 133 $/t. tained (economic loss). With a steady increase of GHG emissions price, it is seen that a breakeven Influence of shorter period and the investment cost of value is obtained at about 133 $/t. Furthermore, a biogas storage on electricity production hike of the GHG emissions price above 100 €/t As shown in Additional file 1: Tables S13-S23, aver- presents positive effects on the economic+GHG profit aging electricity prices based on 7 periods a month (economic profit including benefits from avoided alleviates the variability of electricity prices. For this GHG emissions). From this, it could be concluded reason, an additional scenario is performed whereby that biogas production could be competitive at auc- for a selected month, each day in the month is con- tion trading prices when the benefits of GHG unbur- sideredwhilethehoursaremergedinto3shiftpe- dening would be recognised and when the price of riods. Such a scenario is performed for the average GHG emissions would be higher than 100 €/t CO2 capacity of 1 ± 0.1 MW of electricity produced and by eq. maximizing the economic profit where only one plant at Site I is selected. In this scenario, the variations in Maximizing economic profit while increasing auction electricity prices are slightly better highlighted, as can trading prices of electricity be seen from Fig. 14 which shows an example of In this section, the effects of increasing the auction electricity prices for the three periods considered in trading prices of electricity as it affects the maxi- each day for the month of January (based on the data mized economic profit are put forward. It might not from BSP South Pool Energy Exchange (2008) and be expected that the activities which have net unbur- considered conversion rate). In this sense, January is dening effects on the environment would get benefits considered as an illustration. It should be noted that from it. Due to this, the auction trading prices are this scenario is performed for the specific month sep- multiplied by a certain factor, which could be inter- arately and yearly due to the computational time re- preted as how much subsidies may be required for quired to solve such a problem. the electricity generation from biogas to be profit- From Fig. 14, the prices in January are on average the able. Figure 13 showsthatanincreaseinauction highest within the afternoon/evening period (3 pm - 10 trading prices of electricity causes an increase in pm), and lowest during the night period (11 pm – 6 am). Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 19 of 23
Fig. 14 Electricity prices when considering 3 periods a day for January
On average the prices in the afternoon/evening are 1.71 no biogas is stored. On the other hand, the prices in the times higher than that of the night period and in the morning are “in-between”, and thus also the amounts of morning 1.49 times higher than during the night. Note stored biogas are “in-between” the afternoon and night that in Fig. 14, 1.1, 2.1, 3.1, represent day 1 in January, periods. day 2 in January, day 3 in January, etc. Figure 16 shows electricity production in the month of However, despite the relatively significant differences January, where the electricity is produced in the after- in the prices, again biogas storage was not selected, and noons/evenings at the highest capacity levels (9.54 MW), the electricity was produced at constant capacity. As it while in the morning hours electricity produced fluctuates was stated previously, the main reason is the high invest- between 9.6 MW and 2.8 MW. On the other hand, apart ment cost of biogas holders, but also higher investment from 6 days (14, 29, 21, 27, 28 and 31) in January, little or cost incurred in acquiring the CHP plant and a backup no electricity is produced. It is worth stating that the aver- is required for the heat consumed by the plant itself. age daily electricity generated in January is 19.6 MWh/day To study the relationship between the volume of bio- (accounting for 93.5% of operating time and between gas stored and electricity production, investment cost 1.95 × 106 and 2.38 × 106 m3/y of methane produced), of for biogas storage was thus reduced. It was found that which 48.5% on average is produced in the 3 pm – 10 pm the biogas storage is selected when the base case price period (9.51 MWh/day), 42.2% in the 7 am – 2pmperiod (with a considered base case capacity of 3000 m3 of bio- (8.32 MWh/day) and only 9.1% during the night period gas stored) is reduced from investment cost of 701,600 $ (1.78 MWh/day). Some variations from the “typical” trend to 300,000 $ or lower (see also Additional file 1: Figure could be seen during the days 1, 14–15, 20–22, 27–28 and S1). Figure 15, which is for the month of January, shows 31, when electricity production is either lower (morning how a reduction in the investment costs of the biogas hours) or greater (night hours). This scenario occurs storage facility affects the daily storage of biogas. Hence, mainly in the days when biogas is not stored (days 14, 20, a reduction of investment cost for the biogas holder pre- 21, 27, 31), as shown also from Fig. 15. sents maximum daily period storage of 3154 m3 occur- Similarly, Figs. 17 and 18 show the biogas storage and ring during the night period (11 pm – 6 am). The electricity production profiles for the month of December storage pattern follows the electricity prices in that in which is the month whereby the electricity prices also the night period when electricity prices are low, the have negative values in some hours. amount of biogas stored is the highest, and when electri- Figure 17 displays the maximum daily period stor- city prices are higher (in the afternoon period), almost age as 2877 m3, which occurs during the night
Fig. 15 Biogas storage in January when the biogas holder investment cost is reduced from 701,600 $ to 300,000 $ Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 20 of 23
Fig. 16 Electricity production in January when the biogas holder investment cost is reduced to 300,000 $ period (11 pm – 6 am), and on certain days in the less than that incurred in the 5 MW biogas plant cap- other two periods. Figure 18 also illustrates that acity. However, considering Fig. 19,itisevidentthat electricity is mostly produced during the day periods. when biogas capacity is 3 MW, the lowest economic However, significant “disturbances” in the patterns loss (349,825 $/y) is accrued in the supply chain. The could be seen in the days between 6 and 10 Decem- value of economic loss obtained in the 3 MW capacity ber and 23–26 December. During these days, electri- scenario may be as a result of the selection of relatively city prices are either the highest or the lowest, as cheaper feedstocks, such as poultry manure (28,719 t/y) showninFig.3. and corn stover (12,844 t/y) and due to economy of Considering the scenario where investment cost is scale. At capacities higher than 3 MW, the more expen- further reduced, more significant differences in bio- sive grass silage is selected to satisfy the consumption gas storage and electricity production patterns are and restrictions in terms of available area for biogas obtained and higher amounts of biogas are to be production. stored. For example, if the base case cost of biogas holder is reduced to 2000 $, the model solution shows that 3430 m3 of biogas is maximally stored in Conclusions January, while still the same amount (2877 m3)is This study presents a model showcasing the effects of +GHG stored in December. With the decrease of the invest- economic profit, sustainability profit and economic ment cost for biogas storage, more biogas is indi- profit on the supply chain network production of electri- cated to be stored. city based on two capacities (1 MW and 5 MW) over hourly, daily, and monthly timeframes. An illustrative case study of three sites in Slovenia is used and results Variation of biogas plant capacity with economic profit show negative economic values obtained for the three (loss) objectives considered. This scenario considers a situation whereby the biogas The effects of increasing the GHG emission prices plant capacity is varied from 1 MW to 5 MW with incre- when maximizing economic+GHG profits gives a break- ments of 500 kW as illustrated in Fig. 19. even GHG emissions price of 133 $/t. Moreover, a From Tables 1 and 2, it is noted that at 1 MW cap- sensitivity analysis involving the variation of the auc- acity of the biogas plant an economic loss of about 550, tion trading prices of electricity against maximizing 000 $/y is incurred in the supply chain which is 13% the economic profit shows that the breakeven value
Fig. 17 Biogas storage in December when the biogas holder investment cost is decreased to 300,000 $ Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 21 of 23
Fig. 18 Electricity production in December when the biogas holder investment cost is decreased to 300,000 $ for economic profit may only be obtained beyond the Supplementary information doubling of the auction trading prices. Furthermore, Supplementary information accompanies this paper at https://doi.org/10. 1186/s42480-019-0025-5. the variation of biogas plant capacity with economic profit showed the lowest economic loss obtained at Additional file 1: PART A. Data: Table S1. Auction Trading Prices for the 3 MW capacity. January (in €/MWh). Table S2. Auction Trading Prices for February (in In future studies, the model will be extended to in- €/MWh). Table S3. Auction Trading Prices for March (in €/MWh). Table S4. Auction Trading Prices for April (in €/MWh). Table S5. Auction tegrate electricity generation from solar and wind Trading Prices for May (in €/MWh). Table S6. Auction Trading Prices for energy sources. Furthermore, other bioenergy prod- June (in €/MWh). Table S7. Auction Trading Prices for July (in €/MWh). ucts in the form of bioethanol, biodiesel and other Table S8. Auction Trading Prices for August (in €/MWh). Table S9. Auction Trading Prices for September (in €/MWh). Table S10. Auction may also be included which extends the variety of Trading Prices for October (in €/MWh). Table S11. Auction Trading Prices products, raw materials and conversion technologies for November (in €/MWh). Table S12. Auction Trading Prices for employed. Moreover, a comparative analysis of the December (in €/MWh). Table S13. Considered (Averaged) Auction Trading Prices for January (in $/kWh). Table S14. Considered (Averaged) results obtained in this study could also be made for Auction Trading Prices for February (in $/kWh). Table S15. Considered another region or country. Finally, the study regard- (Averaged) Auction Trading Prices for March (in $/kWh). Table S16. ing processing the lower quality by-product (diges- Considered (Averaged) Auction Trading Prices for April (in $/kWh). Table S17. Considered (Averaged) Auction Trading Prices for May (in $/kWh). tate) into more valuable products will be performed Table S18. Considered (Averaged) Auction Trading Prices for June (in in order to further investigate the possible opportun- $/kWh). Table S19. Considered (Averaged) Auction Trading Prices for July ities for improving profitability of biogas production (in $/kWh). Table S20. Considered (Averaged) Auction Trading Prices for August (in $/kWh). Table S21. Considered (Averaged) Auction Trading while considering auction trading prices of electri- Prices for September (in $/kWh). Table S22. Considered (Averaged) city. As there are many uncertainties involved in dif- Auction Trading Prices for October (in $/kWh). Table S23. Considered ferent decisions (such as raw material composition (Averaged) Auction Trading Prices for November (in $/kWh). Table S24. Considered (Averaged) Auction Trading Prices for December (in $/kWh). and its availability, biogas yield, prices and other), Table S25. Coefficient for avoided GHG emissions for material p. Table flexible synthesis of biogas supply network could be S26. GHG emission coefficient of acquisition and consumption of performed. Methodology for sustainable design of material p. Table S27. GHG emission coefficient for transport of material p. Table S28. Eco-benefit coefficient of material p. Table S29. Eco-cost supply networks with a larger number of uncertain coefficient of material p. Table S30. Eco-cost coefficient related to trans- parameters proposed recently [49]andappliedto port of material p. Table S31. Parameters used to calculate Social Profit. biogas supply network [13] could be used for the Figure S1. Correlation for capital cost for biogas holder. PART B. caseofbiogassupplynetworksconsideringmultiple Nomenclature. objectives and auction trading prices of electricity. Abbreviations 1G: First-generation; 2G: Second-generation; 3G: Third-generation; AD: Anaerobic Digestion; BIOSOM: Biogas Supply Optimization Model; CHP: Combined Heat and Power plants; EISA: Energy Independence and Security Act; ETS: Emission Trading Scheme; EU: European Union; EVR: Eco- costs / Value Ratio; GHG: Greenhouse gas emissions; kW: Kilowatt; L1: First layer; L2: Second layer; L3: Third layer; L4: Fourth layer; LCA: Life Cycle Assessment; LCIA: Life Cycle Impact Assessment; MILP: Mixed Integer Linear Programming; MW: Megawatt; MWe: Megawatt of electricity; NPV: Net Present Value; TAC: Total Annualized Cost
Acknowledgments Not applicable.
Authors’ contributions Fig. 19 Variation of biogas plant capacity with economic profit JE and KZ collected the data, prepared the optimization model and wrote the first draft of the manuscript. LČ supervised the work and analysed the Egieya et al. BMC Chemical Engineering (2020) 2:3 Page 22 of 23
data and the model. All authors discussed the findings, contributed to 14. Emara IA, Gadalla MA, Ashour FH. Supply chain design network model for revising the manuscript and read and approved the final manuscript. biofuels and chemicals from waste cooking oil. Chem Eng Trans. 2018;70: 433–8. 15. Ivanov BB, Dzhelil YR, Ganev EI, Dobruzhaliev DG. Multi-period model of Funding sustainable integrated hybrid first and second generation bioethanol supply The authors are grateful for the support of funds from the Research Office of chains. Chem Eng Trans. 2018;70:289–94. the University of Cape Town, the Engineering and the Built Environment 16. Díaz-Trujillo LA, Nápoles-Rivera F. Optimization of biogas supply chain in Faculty of the University of Cape Town (UCT), the National Research Mexico considering economic and environmental aspects. Renew Energy. Foundation of South Africa under Grant number 87744, the Nigeria Atomic 2019;139:1227–40. Energy Commission, Slovenian Research Agency (research core funding No. 17. Egieya J, Čuček L, Isafiade A, Kravanja Z. Synthesis of Supply Networks over P2–0412 and P2–0032, project J7–1816 and PhD research fellowship MR- Multiple Time Frames: A Case Study of Electricity Production from Biogas. 39209), Slovenia-Croatia bilateral project Integration of Renewable Energy Comput Aided Chem Eng. 2017;40:1447–52. within Energy Systems – INTEGRES and the Erasmus+ programme of the 18. Mousavi Ahranjani P, Ghaderi SF, Azadeh A, Babazadeh R. Hybrid multiobjective European Union. robust Possibilistic programming approach to a sustainable bioethanol supply chain network design. Ind Eng Chem Res. 2018;57(44):15066–83. Availability of data and materials 19. Čuček L, Martín M, Grossmann IE, Kravanja Z. Multi-period synthesis of All data are provided in the manuscript or cited in the references. optimally integrated biomass and bioenergy supply network. Comput Chem Eng. 2014;66:57–70. 20. Egieya JM, Čuček L, Zirngast K, Isafiade AJ, Pahor B, Kravanja Z. Biogas Competing interests supply chain optimization considering different multi-period scenarios. The authors declare that they have no competing interests. Chem Eng Trans. 2018;70:985–90. 21. Mutenure M, Čuček L, Egieya J, Isafiade AJ, Kravanja Z. Optimization of Author details bioethanol and sugar supply chain network: a south African case study. 1 Nuclear Power Plant Development (NPPD) Directorate, Nigeria Atomic Clean Techn Environ Policy. 2018;20(5):925–48. 2 Energy Commission, Abuja, FCT, Nigeria. Department of Chemical 22. Wang B, Gebreslassie BH, You F. Sustainable design and synthesis of Engineering, University of Cape Town, Private Bag X3, Rondebosch, Cape hydrocarbon biorefinery via gasification pathway: integrated life cycle 3 Town 7701, South Africa. Faculty of Chemistry and Chemical Engineering, assessment and technoeconomic analysis with multiobjective University of Maribor, Smetanova ulica 17, 2000 Maribor, Slovenia. superstructure optimization. Comput Chem Eng. 2013;52:55–76. 23. How BS, Lam HL. Integrated biomass supply chain in Malaysia: a sustainable Received: 16 September 2019 Accepted: 16 December 2019 strategy. Chem Eng Trans. 2017;61:1573–8. 24. Boström M. A missing pillar? Challenges in theorizing and practicing social sustainability: introduction to the special issue. Sustainability: Sci Pract – References Policy. 2012;8(1):3 14. 1. Acheampong M, Ertem FC, Kappler B, Neubauer P. In pursuit of sustainable 25. You F, Tao L, Graziano DJ, Snyder SW. Optimal design of sustainable cellulosic development goal (SDG) number 7: will biofuels be reliable? Renew Sust biofuel supply chains: multiobjective optimization coupled with life cycle – – Energ Rev. 2017;75:927–37. assessment and input output analysis. AICHE J. 2012;58(4):1157 80. Ž Č č 2. Yue D, You F, Snyder SW. Biomass-to-bioenergy and biofuel supply chain 26. Zore , u ek L, Kravanja Z. Syntheses of sustainable supply networks with – optimization: overview, key issues and challenges. Comput Chem Eng. 2014; a new composite criterion sustainability profit. Comput Chem Eng. 2017; – 66:36–56. 102:139 55. Ž Č č 3. Hegnsholt E, Unnikrishnan S, Pollmann-Larsen M, Askelsdottir B, Gerard M: 27. Zore , u ek L, Kravanja Z. Synthesis of sustainable production systems Tackling the 1.6-billion-ton food loss and waste crisis
37. Lam HL, Klemeš JJ, Kravanja Z. Model-size reduction techniques for large-scale biomass production and supply networks. Energy. 2011;36(8):4599–608. 38. GAMS Development Corporation: GAMS – Documentation
Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.