Introduction and overview

The effects of exploitation of fossil raw materials are inevitable. Consequently, the structural transformation towards a bio-based economy is essential. To master the challenges of finite fossil resources, increasing energy demand, and greenhouse gas (GHG) mitigation and to reduce dependence on imports as well as to secure well-being of future generations, policy makers promote application of bioenergy [1] by defining political targets, such as the EU 20-20-20 strategy (EC Renewables Directive) [2]. Bioenergy as the largest renewable energy source and on-demand power provider [3] is expected to be a main driver towards a bio-based economy, or bioeconomy [4].

While ensuring security of food and feed as well as the conservation of natural resources, bioeconomy aims at using sustainable renewable resources for the production of bioenergy [5, 6]. However, various pathways exist for bioenergy generation, as is illustrated in Fig. 1. A general biomass value chain (BVC), also referred to as biomass supply chain (BSC) in literature, is characterized by the valorization of biomass feedstock, such as organic, wood, and crop material or residues, as well as of municipal organic wastes and manure for the production of bioenergy and innovative bio-based materials [7–10]. Within the BVC, strategic, tactical, and operational decisions are made based on different decision variables [11]. These interrelated decisions cover the type, quantity, condition, and location of biomass supply; the shipment size, the mode and route of transport as well as the biomass transshipment, allocation and scheduling; the storage type, duration, and location; and the type of conversion technology, the capacity and size as well as the location and the output product type [12–14].

Fig. 1
figure 1

Illustration of a biomass-based value chain (BVC)

Full size image

Thermo-, physico-, and biochemical conversion processes transform biomass into electricity, heat, and gaseous or liquid products. Another biomass valorization pathway, the biorefinery concept, recently gained increasing attention. Biorefineries aim at a complete utilization of input material fractions in an integrated plant providing multiple outputs of material products and energy [15]. Full-scale biorefineries, e.g., sugar/starch biorefineries or a green biorefinery, exist. However, due to the technology readiness levels of the relevant biomass feedstock applications and the immediate implementation of technologies, the biorefinery concept is not part of this research paper.

Having various biomass pathways and conversion options, selection of technologies is critical affecting the BVC design [16]. Considering all these aspects and interdependencies, such as economies of scale (EoS), the techno-economic assessment of biomass valorization pathways is difficult and has to be supported by optimization models [14, 17]. Numerous BVC optimization models exist [11, 12, 18–22]. Some of the models are case-specific and analyze single biomass valorization pathways [20–22]. For the techno-economic assessment of biomass pathways and the selection of biomass feedstocks and technologies, however, one model formulation is required, which takes into account technology and capacity planning as well as the trade-off between transportation costs and technology investments [19, 23].

Unlike fossil resources, biomass is spatially distributed and its valorization is restricted by low-energy densities and high water contents [24, 25]. Long-distance transportation is not cost-effective, as a result of which small- or medium-scale conversion facilities are required [26]. Large-scale facilities, on the other hand, benefit from economies of scale and lower specific investments. For this reason, a trade-off between transportation and investment-related cost exists, which affects the structure of the BVC [12, 27]. When assuming a fixed system boundary with an explicit amount of biomass, the network structure may result in a few (medium- and) large-scale conversion facilities forming a centralized network or in plenty of small-scale facilities in a decentralized or mixed structure [26, 28], as is illustrated in Fig. 2. Existing modeling approaches do not cover the multiplicity of processes and decisions within a BVC and the trade-off between those processes [14].

Fig. 2
figure 2

Network structure illustration

Full size image

We aim to develop a model which integrates those crucial aspects of biomass pathway assessment by formulating a mixed-integer linear program (MILP) combining location, technology, and capacity planning of several biomass valorization pathways. In contrast to common BVC MILP modeling approaches, we incorporate multiple biomass feedstocks, technologies, and end products, together with the trade-off assumption of economies of scale and technological capacity ranges. As technologies are restricted by capacity ranges (e.g., the size of a combined heat and power station restricts its capacity), selection of the optimal capacity in comparison to the associated investments is difficult. By considering these multiplicity levels, significant total system cost reductions can be achieved [13, 29, 30].

The types of biomass feedstock taken into account consist of household waste (incl. a fossil fraction), forest residues, straw, and vineyard residues (prunings). Biomass is converted by combustion (Comb), waste-to-energy (WtE), and three gasification technologies, i.e., downdraft gasifier (DG), fluidized-bed gasifier (FBG), and biomass integrated gasification combined cycle (BIGCC). In accordance with the equilibrium of biomass and bioenergy flow, the conversion process is modeled as a black box with biomass input, including the specific lower heating value, yield, and bioenergy output, which depend on the type of biomass and the specific technology efficiency, which again is dependent on the type of biomass and the technology [11]. The generated output is bioenergy in thermal and electric form. The model specifications are illustrated in Fig. 3.

Fig. 3
figure 3

Modeled conversion process

Full size image

The presented optimization model is in line with current research approaches [12, 17]. In addition, it includes multiple biomass feedstocks in combination with multiple technologies and multiple output products. According to literature on BVC modeling, this has hardly been done before [18]. Diverse biomass conversion routes can be compared in detail using one model formulation as the basis. The model can be adapted easily to real-life case studies.

Our approach is tailored to the techno-economic assessment of different biomass conversion routes to provide decision support in fostering potential biomass valorization pathways on a regional scope. The regional scope of model application is the Upper Rhine Region (URR) encompassing France, Switzerland, and Germany. The biomass potentials in the URR and further model input data were obtained in the course of the tri-national OUI Biomasse project on “Innovations for a Sustainable Biomass Utilization in the Upper Rhine Region” (http://www.oui-biomasse.info/). Given a pre-selected set of biomass supply locations and feasible conversion locations in the URR, most economic biomass pathways are investigated and locations of conversion facilities determined. The geographical expansion of the URR covers the entire area of the Alsace region in France, the northwest of Switzerland with five cantons, and most of Baden and the southernmost part of Rhineland-Palatinate in Germany. The URR has an area of about 21,500 km2, 43% of which are covered by forest and 37% are used for agriculture. Hence, a large area is cultivated extensively due to political and financial support in the tri-national region. In particular, wood is the most prominent source of bioenergy for heat production, while agricultural products, residues, and waste are primarily used in fermentation processes for biogas production. Arable land is predominantly found in the Rhine Valley, while permanent grassland as well as the economically important vineyards are located in hilly areas and along rivers. Due to the geographic and climatic location of the URR, favorable conditions for biomass production are given.

In the next section a brief overview of existing biomass value/supply chain models is provided and research gaps are identified. Considering these research gaps, an approach to the design of BVC is developed and implemented in the subsequent section. The model is applied in the following section. Relevant input data of the URR application are described and the model results are shown and discussed. The concluding section summarizes the insights gained from model application and identifies areas needing further research.

Literature Overview

A great number of reviews of biomass supply chain modeling approaches exist. Most reviewed models apply similar methods and techniques to cover the complexity of biomass valorization pathways. In this section a few models and reviews as well as most recent modeling approaches are listed and discussed for their suitability, the objective being to identify potential research gaps.

An overview of requirements for the design of biomass supply systems is given by Friedler et al. [26]. Additionally, the authors present principles and a planning approach for the case study of biomass-to-liquid production in North-Eastern Germany. Their GIS-based (geographic information system) modeling approach optimizes cost-efficient design of industrialized biomass supply systems by combining logistics and location planning. Kerdoncuff [31] defines a two-stage binary capacitated warehouse location problem for the allocation of biomass-to-liquid facilities in the southwest of Germany. Forest residues and crop residues are utilized for the production of Fischer–Tropsch fuels taking into consideration investments, biomass supply, and labor expenses as well as transport and processing costs.

In the doctoral thesis of Schwaderer [32] a mixed-integer linear program (MILP) is developed to minimize total costs of one biomass-based value chain on a regional scope of the federal state of Baden-Wuerttemberg in Germany. Schwaderer focuses on a biomass-specific process design, respects regional infrastructure restrictions, and takes into account different capacity levels as well as mass and energy flows for the technology selection process. The author applies the model in a case study to assess the production of second-generation biofuel from both the technological and economic perspective. Although the model by Schwaderer is the basis of the model presented in this contribution, multiple biomass feedstocks and output products are not analyzed.

A review of methods to optimize the design and management of biomass-for-bioenergy supply chains is offered by De Meyer et al. [11]. Relevant decisions on different levels are outlined and used as a basis of a broad literature study to mark modeling developments relevant to bioenergy production. Their study points out that MILP models are a highly suitable method for the integration of multiple decisions into one model formulation. Moreover, the authors highlight possible methodological extensions by incorporating GIS tools and multiple objectives as well as recommend the development of holistic approaches. Still, holistic modeling approaches tend to require long computation times, which often results in a limited and case-specific application. These approaches frequently cannot be transferred to other problems.

The review paper by Yue et al. [12] illustrates the complete framework behind the idea of biofuel supply chain optimization. Biomass-to-biofuels pathways are described, general structures and main components of the supply chain are identified, and key issues in multi-scale optimization are explained. Yue, You, and Snyder provide a useful outline of biofuel supply chain literature at the end. As they focus biofuel supply chains, however, research into other bioenergy pathways is lacking. This gap is filled by Mafakheri and Nasiri [28], who review applications, challenges, and research directions of generic biomass supply chain operations. The authors apply common supply chain management practices to identify research gaps in recent approaches and conclude that case study applications need to be employed for model validations. With a key focus on model application to the URR, we aim for an extensive validation of our approach in next but one section.

Garcia and You [20] summarize opportunities in designing supply chains and identify multi-scale (coordination through the supply chain), multi-objective (economy vs ecology vs sociology), and sustainability (life cycle optimization) as being the major challenges. With reference to Yue and You [33], Garcia and You provide an overview of sustainable supply chain approaches, including the biomass-to-electricity routes. Moreover, the authors emphasize the combination of simulation and optimization methods for a generic modeling approach of BVC, but disregard the multiplicity of processes, technologies, and products.

Zhang et al. [30] pick up on this approach and combine the methods simulation and MILP optimization with integrated GIS in a decision support system for a biofuel supply chain. The approach first preselects potential biofuel facility locations for converting forest biomass into biofuel and subsequently optimizes to obtain a cost-efficient network configuration taking into account simulation-based data, such as costs and material flows. By way of example, the model is applied to the northern part of Michigan (USA) and yields insights regarding the optimal facility size and the assigned quantity of biomass supply, but disregards economies of scale and process interrelations contrary to the approach presented in the following section.

A very recent study on sustainable biomass supply chain concepts is introduced by Hong et al. [21]. Following the advice of How and Lam, we integrate the technology selection process as a key element into our research goal. The authors investigate various modeling approaches for sustainable supply chains and conclude that technology selection is one of the central components in biomass supply chain modeling.

An operations research perspective of models for optimization and performance evaluation of biomass supply chains is offered by Ba et al. [18]. Starting from the basics of biomass supply chains, continuing through various types of optimization methods, and ending with a comprehensive list of contemporary models, the authors provide an excellent overview of the state of the art of research relating to biomass supply chain design. According to this overview, most models include multiple biomass feedstocks, many select among multiple technologies, and some generate multiple output products, but only few take the whole multiplicity of processes, technologies, and products into account [34, 35].

The result of the literature review is the identification of the necessity of an approach which integrates multiple biomass feedstocks, technologies, and output products and enables a plain model application for an assessment of various biomass conversion routes in one model formulation. Such a model is presented in the next section.

Modeling Methodology

Various modeling approaches exist. However, to consider the multiplicity of processes, conversion technologies, and products, use of linear models is mandatory, as they are easy to apply, reduce computation time, and ensure optimality [11, 21]. Hence, a mixed-integer linear model is formulated. Whereas the model notation is summarized separately below, its main characteristics are described briefly by referring to the relevant model equations. The biomass flow of feedstock u is represented by continuous variables (\({{x}_{uijk}}\in \mathbb{R^+}\)) and starts at supply locations, referred to as sources i, and ends at the sinks j at conversion facilities k. The system boundaries are defined by the source and sink locations and can encompass regional, national or continental levels depending on the scope of application.

Within these boundaries, the objective function maximizes the profit (1), which consists of the revenue of selling the output product p subtracted by the cost of biomass provision, transportation, processing, and the related investments for a set planning period. Although biomass is spatially distributed, we assume that biomass is supplied at source locations (2) and allocated to sink locations without any transshipment or storage activity to satisfy a pre-determined demand (3). In case of low output prices, no biomass is transported and converted or rather no optimal solution is found by the model. Hence, a demand range or a particular demand must be provided to enable model solutions for analysis purposes although not profitable (5, 6). The source and sink locations can be identical, which means that biomass is utilized directly. At sink locations j, up to ten biomass facilities k can be built. Depending on the technology t the conversion facilities are restricted by minimal capacities (10) while generating a specific output product (\({e}_{utpjk}\)). In order to reach a minimal output capacity, a binary output flow (bundle constraint) is formulated (9). The mass balance constraint (11) approximates the equilibrium between ingoing biomass and outgoing bioenergy. Depending on the biomass and the applied conversion technology, the equilibrium constraint defines the scaling of the capacity with two modeling variables. In order to achieve equilibrium, one continuous variable estimates the approximated output and another dummy variable the irrelevant output values (11–13). For the alignment of the biomass type and the technology as well as the output product, only relevant biomass and technology combinations are taken into account (14).

The trade-off between transportation costs and economies of scale of technology capacity investments might have an impact on the design of biomass value chains. Economies of scale are subject to the technical capacities of conversion facilities. These capacities affect the size of the conversion facilities and the associated investments. In general, numerous biomass conversion technologies with different capacity level exist. However, technology investments commonly decrease with increasing capacities, yielding a (nonlinear) degressive function. Modeling of nonlinear functions has a major impact on the performance of the model as well as on the calculation time. For this reason, the degressive function has to be linearized. By splitting this function into segments and by introducing capacity intervals, each segment can be linearized separately. As a result, a linear function exists for every segment, which can be modeled. Figure 4 shows an exemplary linearization into four segments.

Fig. 4
figure 4

Economies of scale approximation

Full size image

The approach of linearizing the investment function and defining the capacity intervals of the conversion technology in accordance to the multiple-choice methodology is based on Schwaderer [32]. Schwaderer approximates the step-wise linear function as depicted in Fig. 4 with a SOS2-implementation (special-order-sets of type 2). Here, the degression function applies a specific technology scaling factor. Depending on the chosen capacity level of the technology a certain amount of output product is generated (15, 16). This amount of output is applied in the objective function (1) in order to estimate the final investments. The required constraints for the interval approximation consist of one continuous and one binary variable (17–22). The model notation is summarized below.

figure b

Case Study

Unused biomass potentials exist in the Upper Rhine Region (URR) [36]. However, not all biomass potentials can be utilized due to technology capacity restrictions as well as for economic and environmental reasons. For instance, the current potential amount of green waste for anaerobic digestion in the URR is not sufficient to be converted into biogas for thermal or upgrading purposes. The same holds for vineyard residues, such as pomace and yeast as well as manure. Even a combined utilization of all three types of biomass is not sufficient for profitable biogas production in accordance to modelling results. For this reason, the case study covers the above types of biomass feedstock with the highest potentials (i.e., household waste, forest residues, straw, and vineyard residues). Anaerobic digestion is integrated into the model, but is not presented due to the limited biomass potentials.

A description of model application and pre-selected source and sink locations in the URR are illustrated in Fig. 5. In accordance with the biomass potentials of the year 2010 estimated by the OUI Biomasse project assuming a water content of 20% for lignocellulosic matter and without taking seasonal availability into account by assuming a continuous biomass flow throughout the year, the biomass input data are: Household waste: 874,806 t/a; forest residues: 163,551 tbd/a; straw: 930,363 t/a; vineyard residues (prunings): 151,665 t/a.

Fig. 5
figure 5

Description of model application with biomass supply and potential conversion locations in the Upper Rhine Region

Full size image

Whereas research of the OUI Biomass project focused on future scenarios of the year 2030 for one single type of biomass, the current approach integrates multiple types of biomass into one scenario and assumes that one technology can utilize different biomass feedstocks without capacity-contingent efficiency losses. This allows for a techno-economic assessment by comparing transport costs (TC), biomass process and supply costs (SC), as well as conversion- and investment-related costs (IC) for a planning period of 20 years. In the first scenario, the model considers simultaneous valorization of three types of biomass feedstock, i.e., forest residues, straw, and vineyard prunings, for the production of heat and electricity by applying combustion and gasification technologies in accordance with the biomass-to-bioenergy mapping matrix (see Table 1). Household waste is applied separately due to WtE conversion restrictions in the second scenario. The third scenario focuses on the price of the output product and determines its effect on the model results.

Input Data Definition

The input data for model application are based on reports published by research partners during the OUI Biomasse project [37]. The project was completed in 2015 and aimed at evaluating biomass as a renewable energy source in the Upper Rhine Region (URR). Project results, relevant information, and data are available online: http://www.oui-biomasse.info/en/.

Extensive research was conducted to obtain specific data of conversion processes by project partners of research area 3, who investigated technological conversion routes. Industry data were compiled and literature was reviewed to determine the required input parameters, which are presented below.

The described BVC processes are assessed with technological factors (i.e., calorific value and conversion efficiency) and economic factors (i.e., supply costs, fixed and variable transportation costs, by-product disposal costs, and processing costs, as well as investments, and operations and maintenance costs). The economic factors are compared with the market prices of the generated output. The price of the output affects the total revenue and is one crucial parameter in the model. An excerpt of model parameters and required input data is presented in Fig. 6. Another crucial input parameter is the biomass potential of a certain type of biomass feedstock. Depending on the available biomass feedstock, the type of conversion technology is assigned [26]. The implemented biomass feedstock is: Forest residues; household waste; straw; and vineyard residues (Vre), such as prunings. The applied conversion technologies with the specific output product are: Waste-to-energy (WtE) for electric energy, combustion (Comb) for electric and thermal energy, gasification with a downdraft gasifier (DG), a fluidized-bed gasifier (FBG), and the biomass integrated gasification combined cycle (BIGCC) producing electric energy. The mapping of biomass feedstock and conversion technology as well as the corresponding output product is presented in Table 1.

Fig. 6
figure 6

Model input data

Full size image
Table 1 Biomass feedstock to conversion technology mapping matrix
Full size table

The lower heating value (LHV) of the biomass feedstock is an important input parameter affecting processing technology. The LHV depends on the biomass and is the main parameter next to the efficiency (η) of the applied conversion technology that influences the conversion process (see Table 2). The overall efficiency of combustion generating electricity and heat is based on a 15% electric and a 60% thermal efficiency range as well as a scalable energy ratio. Due to the lack of available data, a constant efficiency for biomass with different LHV is assumed.

Table 2 Lower heating values (LHV) and technology efficiency (η) according to [37, 38]
Full size table

By-product disposal costs (e.g., removing the ash residues of combustion and gasification processes) (\({c}_{ut}\), \({c}_{ut}^{f}\)), the cost of biomass feedstock (\({c}_{u}),\) and the fixed (\({c}_{u}^{tf}\)), and variable (\({c}_{u}^{tv}\)) transport cost parameters, which are dependent on the biomass density, are defined and presented in Table 3.

Table 3 Biomass-specific input parameters
Full size table

In accordance with the biomass feedstock and conversion technology mapping matrix, the technology-specific parameters are shown in Table 4.

Table 4 Technology parameters according to [37]
Full size table

The investment parameter (\(c_{tp}^{in{{v}^{p}}}\)) represents the technology investments (I 0) in Fig. 4. The scaling factor (n) describes the economies of scale of the technology capacities and the associated technology investments in accordance with the formula shown in Fig. 4 adjusted by the explicit Chemical Engineering Plant Cost Index (CEPCI) values. Those technology investments are based on an average cost factor (\(c_{t}^{inv}\)) defining the additional expenses for all technologies to be 0.155 (see [32]). The capacity ratio (cf., Fig. 4) is represented by the bioenergy output variable (\({e}_{utpjk}^{n}\)), which is based on the approximation interval of the linearized capacity function (\({n}_{utpjkn}^{e}\)), the selected capacity interval (\({n}_{tpn}^{t}\)), and the input of biomass flow (\({x}_{uijk}\)). The location factor (\({lf}_{j}\)) which enables comparison of international capital costs is assumed to be 0.98 for France and 0.92 for Germany as well as for Switzerland, and was obtained from research of the projects partners. The distances (\({d}_{ij}\)) between the locations are based on own calculations spanning a 36 × 36 matrix with a distance for \(i=j\) of 5 km. The locations \((i,j)\) were defined as biomass sources and biomass sinks. The assumptions for the preliminary selection of locations cover the proximity to transport infrastructure, such as roads, rails, and waterways, and surrounding urban areas as well as accessibility of water body.

Further input data comprises the interval steps (\({n}_{n}\)) for the piecewise linear function approximation. Up to six interval steps can be applied. As the model calculation times increase with a growing number of intervals, the interval parameter is restricted to two. Hence, the concave investment function is represented by two lines. The interval capacity investment approximation (\({n}_{tpn}^{t}\)) in combination with the interval steps (\({n}_{tpn}^{t}\)) and the potentiated maximal technology capacity (\({cap}_{tp}^{{max}^{pot}}\)) is calculated according to the scaling factor (n) and the investment parameter (\({c}_{tp}^{i{nv}^{p}}\)).

Scenario 1

To valorize the complete biomass potentials by assuming a bioenergy price of 500 €/MWh, bioenergy is generated in electric form with the output being 160 MWel and in thermal form with the output amounting to 60 MWth. The gasification of forest residues yields an output of 18.5 MWel. For this purpose, downdraft gasifiers at twenty-eight facilities form a decentralized network, as is shown in Fig. 7. Medium-scale facilities are distributed on the right bank of the Rhine along the Black Forest region, whereas small-scale facilities are distributed on the left bank of the Rhine in the French and Swiss parts of the URR.

Fig. 7
figure 7

Results of scenario 1

Full size image

With total costs of 44.5 million €, gasification of forest residues seems to be less expensive than gasification of prunings from vineyards with 50.8 million €. However, conversion of prunings from vineyards in eight medium- to large-scale combustion facilities produces an output of 17 MWth resulting in a cost per energy output value of 2.9 million €/MWth. Those eight facilities are distributed across the main wine-growing areas along the Upper Rhine Valley, with relatively little transportation efforts being required in the decentralized network.

Straw is different from prunings and forest residues, as it can be burned by combustion processes to produce heat. With outputs of 17.3 MWel and 43 MWth from ten large-scale combustion facilities at six locations, 18.5 MWel from eight large-scale fluidized-bed gasifiers at six locations, and of 112 MWel from sixty-four large-scale downdraft gasifiers at 34 locations, the total heat and electric energy production results in approx. 185 MW and total costs of 389 million €. As the highest straw potentials are situated in the Alsace region in France, most medium- and large-scale facilities are found there.

Combustion and gasification is restricted due to the limited amount of biomass. Although technologies with greater capacity ranges are available for conversion (such as FBG and BIGCC), small-scale gasification facilities in decentralized network structures are used mostly. A comparison of gasification technologies is provided by Kerdoncuff [31], who points out that a large amount of biomass is required for high-capacity facilities. The operation of low-capacity facilities results in decentralized structures, by means of which enable a higher mobilization of the available biomass potentials [26].

When taking into account the main cost drivers of the biomass valorization options, the costs of biomass supply and provision are found to have the largest share. Those costs account for 68% for straw, 75% for vineyard prunings, and 78% for forest residues. They include biomass feedstock costs and harvesting costs. Only around two percent of the total costs are assigned to transportation in all three valorization options. The technology investment-related costs have a share of 21% for forest residues, around 30% for straw, and 24% for vineyard prunings. When comparing the applied conversion technologies, the downdraft gasification process is found to be the most frequently used technology generating approx. 60% of the energy, but often reaching capacity limits. This might be explained by the relatively high scale factor of 0.808 and a lower investment factor in contrast to the other technologies, as is shown in Table 4.

Scenario 2

The second scenario is characterized by the incineration of household waste using waste-to-energy conversion technologies. Household waste is a component of municipal solid waste comprising non-recyclable garbage, rubbish, and biodegradable waste, such as food, kitchen waste, and plant material [39]. The composition of household waste varies from region to region (and country to country) depending on municipal waste separation restrictions [36]. Figure 8 shows the model results for the valorization of 874,806 t/a of household waste, which mainly occurs in densely populated areas of the URR, such as in the metropolitan regions of Basel, Freiburg, and Strasbourg [40]. Due to divergent definitions of household waste and deviating waste management restrictions in the three countries, the potentials cannot be determined easily and vary in consistency, esp. between France and Germany/Switzerland [41]. A total of approx. 70 MWel can be generated with large-capacity facilities of around 27 MWel in the region of Appenweier in Germany, 22 MWel in the region of Sissach in Switzerland, and 21 MWel in the region of Denzlingen in Germany. All available biomass potentials of household waste are distributed to the three facilities forming a centralized network with a total distance of 3647 km. The arising transportation costs account for only 7% of the total costs of approx. 88 million €. In contrast to this, the supply and provision costs amount to 33%, and the investment-related costs to 60%. Consequently, the production of one MWel costs approx. 1.26 million €. The results of scenario 1 and scenario 2 are summarized in Table 5.

Fig. 8
figure 8

Results of scenario 2

Full size image
Table 5 Summary of results of scenario 1 and scenario 2
Full size table

Comparison of the specific energy costs values reveals that the use of household waste for energy production is the most economic option, followed by straw, forest residues, and vineyard residues. From another perspective, converting one ton of household waste results in an energy output of 80 kW, while conversion of one ton of forest residues and vineyard prunings yields around 114 kW, and straw conversion produces 200 kW. These values are in line with the levelized cost of electricity published by the Fraunhofer Institute ISE [42].

Scenario 3

The third scenario is characterized by the determination of a critical bioenergy price for heat and electricity. After the execution of several model runs based on a sensitivity analysis, a critical price of 235 € per MWh is determined. This price describes the break-even profit point between the earned revenue and the costs plus investments of approx. 1.27 million €. Only 18% of the straw potential is valorized, generating approx. 20 MW of thermal and 8 MW of electric energy at four medium-scale combustion facilities. Those facilities are located on the right bank of the Rhine at Bruchsal, Appenweier, Bad Krozingen, and Kandel as is shown in Fig. 9. Most of the straw is valorized on-site with only little transportation of about 102 km being required in a decentralized network structure. Note the total transportation costs of approx. 1 million €, biomass supply and provision costs of approx. 12 million €, and investment-related costs of approx. 18.6 million €. With supply and biomass provision costs having a share of about 40% in the total costs, a cost-to-energy ratio of 1.6 million €/MW is achieved. Further reduction of output prices results in lower revenues compared to total costs plus investments and, hence, is not profitable.

Fig. 9
figure 9

Results of scenario 3

Full size image

Results and Discussion

Judging from the scenario results, the supply and biomass provision costs are the main factor in lignocellulose-based biomass value chains. With a share of 60–70%, the supply and biomass provision costs considerably influence the profitability of bioenergy production facilities [26] and largely affect the design of biomass-based value chains [43]. Waste materials, by contrast, result in lower supply costs and a lower cost ratio of about 30%. The main cost drivers for converting waste into bioenergy are the investments and investment-related costs [44]. Logistics costs in general comprise only up to 10% of the overall costs of the analyzed BVC. This value contradicts some literature values, indicating that the accuracy of results mainly depends on the available data and on the correctness of the underlying estimations as well as modeling assumptions. The trade-off between transportation costs and investments therefore is of no significance in the presented scenarios. Although extensive research into technological input data has been carried out by project partners, validation of a real application is required to provide valuable guidelines for the design of biomass-based value chains. The presented modeling methodology is valid for any biomass-to-bioenergy conversion route and any type of case study due to its multiple-level approach. However, to overcome the discrepancy between the presented model and the existing practices of BVC management, the following aspects need to be considered in future research. Next to the consideration of multiple resources, technologies (and capacity- as well as biomass-dependent efficiencies), and products, the integration of all relevant processes is required by formulating a holistic approach. This holistic approach is to integrate the various decision levels starting from the process of constant supply, including the seasonal availability of biomass and the variable chemical as well as physical composition to the comprehensive pre-treatment, scheduling, transportation, and storage concepts to secure a stable and homogeneous biomass flow to profitable conversion locations with great market proximity [12, 27]. A holistic approach is the perfect fit for the configuration of BVC and the optimization of a super structure. However, the computation time will be longer [28]. Numerous methods exist to reduce long computation times, such as decomposition techniques [17]. As a result, the model itself is difficult to apply and case-sensitive, with specific adaptations being required.

The source and sink locations in the current case study are based on a limited number of aspects. Future research is to pre-select potential locations taking into account infrastructure accessibility, dispersion, population census, and location of existing industrial plants. Integration of Geographic Information Systems (GIS) into a holistic approach is the most suitable application, as it has already been included in several modeling approaches [30].

As a BVC differs from industrial value chains in terms of risks and uncertainties, different approaches are needed to consider raw material price fluctuations, seasonal biomass availability, and quality inconsistencies as well as supplier reliability and environmental requirements [17, 27, 28]. In addition to economic risks, social and environmental risks of bioenergy production refer to the expansion of monocultures, the use of pesticides, the competition with food, a high water demand, and deforestation impacts. Social and environmental risks cannot be integrated independently. Hence, multiple objectives are applied simultaneously [45]. Since social objectives are not tangible, most modeling approaches solely include ecological objectives into their model formulations. Those objectives use life cycle assessment (LCA) practices and identify Pareto frontiers for the preferable selection among Pareto-optimal solutions [12]. However, to provide support in implementing a bioeconomy, social aspects need to be considered next to economic and ecologic objectives. Integrated Assessment Models (IAM) and input–output modeling frameworks can be applied to merge those objectives and to additionally include governmental regulations and socio-economic restrictions [46]. This macroeconomic approach enables analysis of impacts of bioeconomy on the economic sector, the labor market, and regional development, for instance. Potential biomass conversion routes, key indicators influencing the BVC, and bio-economic impact factors can be identified to evaluate governmental strategies [47] and to counsel decision-makers in implementing a sustainable bioeconomy.

Conclusion

Diminishing fossil raw materials must be substituted by renewable resources to provide bioenergy and to pave the way towards bioeconomy. However, many biomass-to-bioenergy pathways exist, which need to be reviewed in order to provide decision support. Mathematical modeling can assists in the decision-making process by modeling the whole biomass-based value chain from the sources to the sinks, finding optimal logistic structures and assessing these, and by validating the model within a case study application.

Contrary to other biomass-based value chain modeling approaches, the developed model integrates multiple biomass feedstocks, technologies, and output products, while taking into account process interrelations in the form of economies of scale. Whereas previous literature seldom combined multiple functional decision levels, such as the simultaneous selection of biomass feedstock supply, transport, and conversion issues, we aim at developing a holistic approach. This approach considers the complex multiplicity of biomass-based value chains. Furthermore, it can be applied easily and, hence, enables a techno-economic assessment of various biomass conversion routes in one formulation. This formulation is validated by an application to the Upper Rhine Region (URR). Here, biomass combustion of lignocellulosic matter for the production of heat and power is compared with three gasification technologies for electricity generation. In addition, waste-to-energy conversion of household is evaluated based on biomass potentials existing in the URR and including costs and investments as well as the generated output. Future research, however, shall focus on thermal energy production, since heat becomes more and more a relevant source of bioenergy, esp. for long-distance heating.

Feedstock for the production of biogas is already utilized extensively in the URR and, therefore, limited. The application results support the vaporization of household waste for generating energy in centralized large-scale conversion facilities with a cost-to-energy ratio of 1.3 million € per MW. Compared to the valorization of forest residues and vineyard prunings, combustion and gasification of straw in decentralized medium-scale facilities is more cost-efficient with a cost-to-energy ratio of 2.1 million € per MW for straw, 2.4 million € per MW for forest residues, and 2.9 million € per MW for vineyard prunings. The main cost driver in the techno-economic assessment is the supply and provision cost, which reaches a 60–70% proportion of the total costs. In accordance with the spatially distributed biomass potentials, the facilities are located along the Rhine Valley and the French, Swiss, and German agricultural and forestry regions as well as in urban areas for household waste.

Additionally, a critical bioenergy price of 235 € per MWh is determined, which is the minimal price for biomass valorization in the URR by straw conversion via combustion into heat and power. The application of gasification technologies requires a price of approx. 300 € per MWh for being profitable. The focus on the complete biomass valorization requires immense subsidies to justify a market price of 500 € per MWh, which cannot be long-term sustainable. In conclusion, a limited amount of biomass exists in the URR, which can be valorized by waste-to-energy, combustion, and gasification technologies. Its profitability, however, depends on the selling price of bioenergy, which is competitive to fossil-based energy prices. Bioenergy can already be a viable alternative, but requires further research to become a key driver towards bioeconomy.