The Electricity Generation Landscape of Bioenergy in Germany

Abstract

Disaggregated data on electricity generation from bioenergy are very helpful for investigating the economic and technical effects of this form of renewable energy on the German power sector with a high temporal and spatial resolution. But the lack of high-resolution feed-in data for Germany makes it necessary to apply numerical simulations to determine the electricity generation from biomass power plants for a time period and geographic region of interest. This article presents how such a simulation model can be developed using public power plant data as well as open information from German TSOs as input data. The physical model is applied to an ensemble of 20,863 biomass power plants, most of which are in continuous operation, to simulate their electricity generation in Germany for the year 2020. For this period, the spatially aggregated simulation results correlate well with the official electricity feed-in from bioenergy. The disaggregated time series can be used to analyze the electricity generation at any spatial scale, as each power plant is simulated with its technical parameters and geographical location. Furthermore, this article introduces the electricity generation landscape of bioenergy as a high-resolution map and at the federal state level with meaningful energy figures, enabling comprehensive assessments of this form of renewable energy for different regions of Germany.

1. Introduction

Over the past decades, energy from biomass has already proven to be a very versatile renewable energy, and due to its storage option, the flexible electricity generation from bioenergy can also play a key role in future energy systems, which will be increasingly dominated by variable renewables such as wind power and photovoltaics. Worldwide, bioenergy has further grown to an installed electricity capacity of almost 149 GW and in the European Union to almost 35 GW by the end of 2023 [1]. In Germany, this installed capacity increased to over 9 GW at the same time, showing that the German bioenergy sector is relevant and still growing, albeit with slower rates in recent years [1,2].

Currently, different types of biomass power plants are in operation for electricity generation in Germany, ranging from power plants for solid or liquid biomass to gaseous biomass such as biogas power plants. As the future integration of renewable energies, also addressing bioenergy with its flexibility option, becomes increasingly crucial for grid stability and sustainability of supply, understanding and optimizing their penetration in the German energy system is essential. For this purpose, disaggregated data on electricity generation from biomass power plants are also required. In order to determine their electricity generation with a high temporal and spatial resolution, numerical simulations have to be applied using public power plant data and open information from German TSOs, as presented in this article. In particular, both the current continuous operation of the vast majority of existing biomass power plants as a relevant contribution to the base load and a future option for more flexible electricity generation from bioenergy can become essential aspects for the rapid and sustainable transformation of the German power sector, also strongly influencing the power grid expansion [3,4,5]. But the lack of disaggregated electricity generation data from biomass power plants, because of strict data protection regulations, especially for the huge number of smaller biomass power plants of private operators in Germany, makes it very difficult to study such aspects at local and regional scales, for instance, to be able to analyze the impact of bioenergy on grid stability and reliability with a high temporal and spatial resolution in order to better assess the future role of bioenergy in the transition to a low-carbon, circular, and bio-based economy [6,7,8].

To date, energy studies and models for investigating electricity generation from bioenergy [9] often focus on the use of specific types of biomasses, e.g., forest residues [10], or on optimizing their supply chain, e.g., for the coproduction of forest biomass electricity and bioethanol [11]. Other studies primarily refer to heat generation from bioenergy plants [12] or develop hybrid models for the integrated planning of bioenergy and biofuel supply chains [13], which are also relevant for electricity generation from bioenergy. But all these energy studies and models do not consider disaggregated electricity generation from huge plant ensembles with many different power plant sites and sizes. Moreover, there are currently no public information platforms that provide such data with a high temporal and spatial resolution for Germany.

To make such disaggregated electricity generation data from biomass power plants also available for detailed analyses of the German energy system [14,15], this article presents a new simulation model that only uses public power plant data as well as open information from the four TSOs in Germany. This bioenergy model also belongs to the ReSTEP models, which form a collection of similar physical models to determine the electricity generation from renewable energy power plants with a high temporal and spatial resolution. For this purpose, these models use public power plant data and extensive weather information for variable renewables, such as wind power and photovoltaics [16,17,18,19], or open information from German TSOs for run-of-river and biomass power plants [20].

2. Input Data

2.1. Plant Dataset

The presented bioenergy model requires, for the numerical simulations, detailed information about the existing biomass power plants in Germany. For each individual power plant, the plant dataset comprises the plant site in geographical coordinates and the corresponding TSO region, the installed electricity capacity, and the date of (de)commissioning, as shown in

Table 1. Plant data for each biomass power plant used in the bioenergy model.

Plant Data	Bioenergy Model
Latitude	used
Longitude	used
TSO region	used
Installed electricity capacity	used
Commission date	used
Decommission date	used (if existing)

.

The applied power plant data were retrieved from the research repository Zenodo [21], which is a public online repository hosted by CERN. According to the release notes, the information about the geographical location and technical parameters of the biomass power plants was taken from the CEMDR [22,23]. The release notes and [24] also indicate that the CEMDR raw data have been extensively completed and cleaned, i.e., the missing information has been added to the plant data, and duplicate or erroneous entries have been removed. After the preparation, cross-checking, and selection of the downloaded plant data for the investigated time period, the compiled dataset comprises 20,863 biomass power plants for the year 2020. The plant dataset includes various types of biomass power plants, such as power plants using forest residues, wood pellets, vegetable oils, or biogas for electricity generation. The individual plant sizes of this dataset range from 1.0 kW to maximal 94.5 MW. The installed electricity capacity of the whole compiled plant dataset is 8.57 GW, which is only slightly lower than the total value of 8.77 GW provided by official German institutions for the year 2020 [2]. This small deviation of less than 3% indicates that most of the existing biomass power plants are contained in the plant dataset applied for the simulations.

2.2. Load Factor

In recent years, most of the biomass power plants generate electricity at a constant rate and, hence, mainly cover the base load. Due to fluctuations of the biomass supply chain, e.g., temporal changes in the microbial biogas production, or extended long-term weather extremes, which can, for instance, cause higher and, therefore, cheaper amounts of deadwood with a certain temporal delay, the electricity generation from biomass power plants can vary over the time. Furthermore, larger power plants especially can often adjust their electricity generation to the actual demand, i.e., they can already provide flexible operation. To consider such effects also in the bioenergy model, so-called plant load factors have to be calculated for the considered time period using the installed electricity capacity and electricity feed-in profiles of the biomass power plants located in a certain region. For the calculation of the spatiotemporal load factor time series LF_st, the electricity feed-in from bioenergy of a TSO region was related to the average installed electricity capacity of all biomass power plants in this region using the following relationship:

$${LF}_{\mathrm{s}\mathrm{t}}={\displaystyle \frac{2·F_{\mathrm{b}\mathrm{i}\mathrm{o}}}{C_{\mathrm{s}\mathrm{t}\mathrm{a}\mathrm{r}\mathrm{t}}+C_{\mathrm{e}\mathrm{n}\mathrm{d}}}}$$

(1)

In this relationship, F_bio stands for the electricity feed-in profile from bioenergy of the considered TSO region, and C_start and C_end are the installed electricity capacities of all biomass power plants in this region at the start and at the end of the investigated time period, respectively. The calculated entries of this load factor time series can vary between 0 and 1, where 0 means that the biomass power plants cannot deliver any power at this time, and 1 means that the maximal power can be delivered into the grids equal to their installed electricity capacity. For a leap year, such as 2020, the spatiotemporal load factor time series comprises 8784 hourly entries for this period. The measured electricity feed-in profiles and the corresponding installed capacity of the biomass power plants are available on the public online platform SMARD of the BNetzA [25].

For the year 2020, the information for F_bio, C_start, and C_end was provided via SMARD as CSVs for the four German TSOs, i.e., for 50Hertz Transmission, TransnetBW, TenneT TSO, and Amprion. Due to similar feed-in profiles of the TSOs over the investigated year 2020, the sum of these four profiles and the total installed electricity capacity from all biomass power plants were taken into account for the calculation of the plant load factors for the German region. Figure 1a depicts a data extract of this SMARD information, which is provided as an hourly resolved time series for F_bio and as yearly values for C_start and C_end, while in Figure 1b, the actual generation profile of F_bio is shown for the first 15 days of the year 2020.

2.3. Verification Data

In order to verify the introduced bioenergy model and benchmark its accuracy, the obtained simulation results, which show the electricity generation in hourly and daily resolutions, have to be compared with the measured data from biomass power plants. As such data are not available for individual biomass power plants due to strict German data protection regulations, it is not possible to evaluate the simulation results at any spatial scale. But if the simulated electricity generation data are spatially aggregated over Germany, the simulation results can be checked against the official electricity feed-in from bioenergy in Germany for 2020. Such measured data can also be retrieved from the public online platform SMARD [25].

3. Simulation Model

From a general point of view, electricity generation data from various renewable energy power plants can be obtained applying either physical [26,27], statistical [28,29,30], or hybrid models [31]. The simulation results produced using physical models are often based on existing power plant information, extensive meteorological data, or detailed measurements of energy systems. Hence, a typical advantage of physical models is the ability to produce disaggregated electricity generation data with a high temporal and spatial resolution for the time period and geographic region of interest.

The main simulation steps of the developed bioenergy model are shown in Figure 2, in which the power plant data and the calculated load factor time series are the input data of this physical model. All simulation steps are depicted as grey rectangles, and the input and output data are shown as yellow and green containers, respectively. The data flow with its direction is indicated by black arrows between the elements.

As shown in Figure 2, the bioenergy model is subdivided in three main simulation steps, which are executed for each biomass power plant of the applied plant dataset: Firstly, we have bioenergy-to-power conversion using the installed electricity capacity and the plant load factors (LF_st), as calculated according to Section 2.2, for each individual power plant. Secondly, we have the simulation of the generated electricity with an hourly resolution, also considering the (de)commissioning data of the biomass power plants. Both simulation steps can be summarized by the following relationship, using the hourly resolved time t_h and the date of commissioning t_cd and decommissioning t_dd for the investigated time period:

$$E_{\mathrm{h}}=\left\{\begin{array}{l}\mathrm{ 0}·{LF}_{\mathrm{s}\mathrm{t}}·t_{\mathrm{h}}{t}_{\mathrm{h}}<t_{\mathrm{c}\mathrm{d}}\\ C_{\mathrm{e}\mathrm{l}}·{LF}_{\mathrm{s}\mathrm{t}}·t_{\mathrm{h}}{t}_{\mathrm{c}\mathrm{d}}\le t_{\mathrm{h}}<t_{\mathrm{d}\mathrm{d}}\\ \mathrm{ 0}·{LF}_{\mathrm{s}\mathrm{t}}·t_{\mathrm{h}}{t}_{\mathrm{h}}\ge t_{\mathrm{d}\mathrm{d}}\end{array}\right.$$

(2)

In this relationship, E_h denotes the hourly generated electricity, and C_el stands for the installed electricity capacity of the biomass power plant. For a leap year, such as 2020, each simulated time series for an individual power plant comprises 8784 electricity generation values in an hourly resolution. Thirdly and lastly, each simulated time series is also converted from an hourly into a daily resolution, corresponding to a temporal aggregation of the time series with 8784 hourly entries to an additional time series with 366 daily entries for each biomass power plant of the compiled plant dataset. Finally, all these times series are stored in CSV format, e.g., for subsequent data analyses and GIS-based applications.

Based on the bioenergy model and the public data sources introduced herein, the electricity generation from biomass power plants can be realistically and easily simulated using, e.g., R or Python for the development of the model code.

4. Investigation Results

4.1. Simulation Results

The generation of electricity from biomass power plants in Germany for 2020 was simulated using the bioenergy model and input data presented in Section 2 and Section 3. After performing the numerical simulations for the plant dataset compiled in Section 2.1, the disaggregated electricity generation data of the whole plant ensemble were aggregated into a single time series. This was carried out to check the simulation results against the official electricity feed-in from bioenergy in Germany for the year 2020. The electricity generation patterns of the measured and simulated time series are depicted with a daily resolution in Figure 3.

As is clearly shown in Figure 3a,b, the simulated electricity generation pattern (green line) agrees very well with the official, i.e., measured, electricity feed-in (grey line) from all biomass power plants in Germany over the whole year of 2020. In addition, the total values for the simulated and measured annual electricity generation are nearly the same, with 39.2 TWh and 40.3 TWh [25], respectively, exhibiting a small deviation of less than 3%. The differences of both patterns, especially the frequently lower values of the simulated electricity generation, are mainly caused by a smaller value of installed electricity capacity of the applied plant dataset and unavoidable uncertainties of the calculated plant load factors.

In addition, Figure 3a also shows that the electricity feed-in from bioenergy is relatively constant over the whole period, i.e., for 2020, the vast majority of existing biomass power plants were in continuous operation. Furthermore, an MAE of 3.1 GWh and an RMSE of 3.4 GWh for the aggregated daily values also indicate a high correlation of both patterns from the statistical perspective. Regarding the measured annual electricity feed-in of 40.3 TWh from bioenergy in Germany, the relative MAE and RMSE also reach a small value of 0.01%. For a further assessment of the temporal variability of electricity generation from bioenergy, Figure 3b shows, for comparison, the total electricity feed-in from photovoltaic systems in Germany for 2020. It becomes clear from these depicted electricity feed-in patterns that most of the biomass power plants mainly cover the base load in Germany, and there is only a small amount of flexibility that is primarily delivered by larger power plants.

A subsequent bias correction using a constant factor can further reduce the deviations of the simulation results, as shown in Figure 4. For this purpose, a correction factor of 1.03 was applied to the results obtained from the simulations, where this constant factor was calculated using the quotient of the measured and simulated annual electricity generation, i.e., 40.3 TWh/39.2 TWh = 1.03, for the year 2020. This bias correction also reduces the MAE to 1.2 GWh and the RMSE to 1.4 GWh.

4.2. Bioenergy Landscape

To date, bioenergy belongs to one of the most important renewable energies in the German energy system [2]. In this energy system, various technologies are implemented to produce heat, transport fuels, and electricity from many different types of biomasses, also including residual and waste materials [4]. Since biomass is a limited resource, it has to be used more sustainable in the future, especially in the transition to a low-carbon, circular, and bio-based economy [7]. In order to also find an optimized path for this transition at local and regional scales, a temporal and/or spatial highly resolved distribution of the electricity generation from bioenergy can be very helpful for comprehensive assessments of this form of renewable energy. For this purpose, Figure 5 shows the annual electricity generation from biomass power plants at the municipal level in Germany for the year 2020 by applying the simulation results performed for this article.

Figure 5 indicates that the electricity generation from bioenergy is mainly concentrated in federal states from North Rhine-Westphalia over Lower Saxony and Saxony-Anhalt to Brandenburg and, furthermore, in Southern Germany, i.e., in many municipalities of Bavaria and Baden-Württemberg. This map also shows that in other federal states, like Thuringia and, in particular, in Rhineland-Palatinate and the Saarland, there are, by far, smaller shares of electricity generation from biomass power plants for 2020.

As a temporally resolved example of the bioenergy landscape in Germany, Figure 6 depicts the electricity generation with a daily resolution for the federal states of Baden-Württemberg, Schleswig-Holstein, and Rhineland-Palatinate.

As another spatially resolved example of the German bioenergy landscape, Figure 7 depicts the electricity generation at the municipal level with a monthly resolution.

Figure 6 and Figure 7 clearly confirm that both the temporal and spatial distributions of the electricity generation from bioenergy in Germany only vary slightly over the considered time period, i.e., the vast majority of existing biomass power plants were in continuous operation for 2020.

In addition, meaningful energy figures, such as the spatiotemporal capacity factor of biomass power plants, can be easily determined using the simulated electricity generation data. As an example,

Table 2. Capacity factors and capacity densities of biomass power plants in Germany for 2020.

Federal State	Capacity Factor (%)	Capacity Density (kW/km2)
Hamburg	53.6	57
Berlin	53.6	46
Lower Saxony	51.7	37
Schleswig-Holstein	50.9	36
North Rhine-Westphalia	51.8	32
Bremen	53.6	30
Bavaria	52.0	25
Baden-Württemberg	51.9	24
Saxony-Anhalt	52.5	23
Thuringia	53.1	18
Mecklenburg-Western Pomerania	52.2	17
Brandenburg	52.6	16
Saxony	52.4	16
Hesse	52.5	14
Rhineland-Palatinate	52.1	10
Saarland	53.6	4

shows the (electricity) capacity factor and capacity density, as defined and already used in recent energy studies [17,20], at the federal state level in Germany for the year 2020.

According to Table 2, the federal states have capacity factors from 50.9% to a maximal 53.6%, whereas this maximum is reached in the city-states of Hamburg, Berlin, and Bremen, as well as in the Saarland. This table further indicates that, with exception of the three city-states, the federal states from the northwest to the far north of Germany, i.e., from North Rhine-Westphalia over Lower Saxony to Schleswig-Holstein, reach the highest values of the spatiotemporal capacity density of biomass power plants, with a maximal value of 37 kW/km² in Lower Saxony. Otherwise, federal states in the southwest of Germany, i.e., Rhineland-Palatinate and the Saarland, only exhibit a capacity density of 10 kW/km² and 4 kW/km², which is in agreement with the maps in Figure 5 and Figure 7.

The examples presented clearly demonstrate that the disaggregated electricity generation data produced by the simulation model in this article can contribute to comprehensive assessments of the electricity generation landscape of bioenergy in Germany. Moreover, to be able to successfully tap the huge potential of bioenergy with its flexibility option for a rapid and sustainable transformation of the German power sector, useful monitoring tools, such as the bioenergy model introduced herein, are very beneficial.

5. Conclusions

In addition to presenting a new bioenergy model, which only uses publicly available information as input data, a further objective of this article was to produce disaggregated electricity generation data on bioenergy for Germany. This was carried out to better assess the role of biomass power plants in the German energy system at local and regional scales. The simulations were performed for the year 2020, because the simulation results for that time period are necessary to extend the energy transition studies for the German region [3,5,32]. It could be clearly demonstrated that the simulated electricity generation agrees well with the official electricity feed-in from all biomass power plants in Germany. In summary, realistic numerical simulations can be performed with the bioenergy model using the presented public data sources. With the help of this physical model, it is possible to break down the measured feed-in data from the national level to different regions of Germany, e.g., to the municipal level. To the best of my knowledge, such disaggregated electricity generation data considering an ensemble of 20,863 biomass power plants with a total installed electricity capacity of 8.57 GW have never been presented for the German region.

Furthermore, with this clear modeling approach, the interested scientific community is able to develop its own bioenergy models based on the ideas and information given in this article. This approach has already been successfully used in a simulation model for electricity generation from run-of-river power plants [20]. The presented physical model can also be applied to other countries, provided that all required input data are available for this region. Many other studies, e.g., for energy system modeling [33,34,35] or spatially resolved assessments of the German power sector [36,37], can profit from the highly resolved electricity generation data produced by this bioenergy model. In combination with other ReSTEP models [16,17,18,19], a holistic view of the electricity generation landscape from renewable energies can be developed for the German energy system.