1 Introduction

Electricity wholesale prices are the basis for many analyses in energy economics, where cost of electricity impacts investment decisions in generation and demand technologies. Market results also determine the emissions of the energy sector as well as the degree to which renewable energy systems can economically feasibly be integrated into the energy system. For many detailed studies with reduced spatial scope, prices and emissions of “electricity from the grid” play a crucial role, be it in the context of energy autonomous municipalities, energy concepts for districts, diffusion scenarios of PV battery storage systems or electric mobility. The price and generation technology for the majority of procured electricity in Europe is determined on the Day-Ahead market, by the Single Day-Ahead Coupling (SDAC). After Bulgaria joined on 9 March 2021, SDAC covers about 95% (1,500 TW h/year) of the EU’s consumption (ENTSO‑E (2021a)).

Market Coupling and the respective determination of available transfer capacities have a large impact on the outcome of electricity markets. It allows for the efficient utilization of physical transport capacities and increases liquidity in markets (Next Kraftwerke (2021)), impacting energy mix, fuel consumption and CO\({}_{2}\) emissions in the coupled market zones. In large parts, the market coupling between price zones is based on bilateral exchange capacities, so-called Net Transfer Capacities (NTC), which are also widely applied scenario-based energy system analyses. While easy to apply, NTCs cannot account for the physical reality in meshed electricity networks, because they do not reflect how exchanges affect each other. This is due to the fact, that power flows follow the physical parameters of the network and only partly match commercial exchanges. To overcome these limitations, the Flow Based Market Coupling (FBMC) methodology was proposed and implemented in the Central Western Europe (CWE) region in May 2015, to explicitly consider grid constraints in the process. FBMC is the target model for the internal electricity market and is planned to be introduced in the larger Core region in 2021, then covering the borders of the bidding zones Austria, Belgium, Croatia, Czech Republic, France, Germany/Luxembourg, Hungary, the Netherlands, Poland, Romania, Slovenia and Slovakia (ENTSO‑E (2020a)). Apart from desirably allowing for larger exchanges compared to NTC market coupling, the flow based (FB) methodology also allows for a more transparent representation of market coupling constraints, which makes it possible to attribute exchange limitations to single grid elements (Marjanovic et al (2018)).

Unfortunately, these advantages come at a price and result in largely increased input data sets covering the transmission grid topology and parameters, spatially resolved generation units, the ability to produce regionalized capacity allocations and output for new-built power plants (especially decentralized renewable energy sources (RES)), which are necessary for the adequate simulation and analysis of present and future energy systems. For market modellers and practitioners alike, it is important to develop an understanding for the implications FBMC has on the European Electricity Market. In particular, to assess the difference between the application of FBMC and the well-known NTC constraints, with respect to the most relevant scenario outcomes like electricity prices, exchange positions and emissions resulting from the energy mix.

In this article, a framework, which enables this evaluation, is presented and the effects of different market coupling approaches are quantified in real-world scenarios for the inter-connected European electricity market. Section 2.1 revisits the main concepts of FBMC to facilitate the reader’s appreciation of the following literature review on FBMC in Sect. 2.2. The developed model and the analysed scenarios are presented in Sect. 3. Section 4, covers the main findings and assesses the magnitude of change for market outcomes for the year 2025 using NTC and FB constraints. Moreover, the impact of minimum exchange capacities (minRAM) introduced in 2020 on the outcome of FBMC are investigated. In Sect. 5, the limitations of the presented approach and wider implications of the results are discussed. Section 6 concludes and gives an outlook on future research.

2 Flow based Market Coupling

FBMC was introduced in the CWE region in 2015 to better account for the physical electricity network constraints in the market clearing procedure. The goal is to increase cross-border capacities, promote supplier competition, increase grid security and minimize prices across in the market zones (Next Kraftwerke (2021)). This Section revisits methodology and important concepts of FBMC to provide the reader with context for the following literature review and the remainder of the article.

2.1 Methodology

For the calculation of FBMC the information gap, which arises from the zonal market design in the European electricity markets has to be closed. Due to the separation of electricity suppliers and grid operators accompanied by the clearing of anonymous bids at the energy exchange, the grid operators cannot know which power plants participate in additional commercial exchange and thus how much the grid infrastructure is loaded. For the necessary calculation of available transmission capacities, the grid operators have to approximate the market outcome to forecast the grid utilization. This approximation is called the base case, which consists of the forecasted demand, supply, exchange position and grid topology.

In the European market design, bidding zones are regarded as (almost) free of congestion. This limits the grid elements, which are considered in the market clearing to constrain commercial exchanges. The relevant elements are identified with linearized sensitivity factors called Power Transfer Distribution Factors (PTDF). Only lines and transformers, which are affected by cross-border exchanges above a certain threshold (often 5 percent, e.g. CREG (2017)) are included. This selection threshold is referred to as PTDF threshold for the remainder of the article. Because the market outcome has to result in a secure grid operation, possible grid outages, which reduce the available capacity on the lines, are considered. Only a limited number of contingencies have a critical impact on the grid, threatening security. Contingency screening is performed to identify the most relevant outages, which are then included as additional constraints. The relevant grid elements are called Critical Network Elements under a Contingency (CNEC). In this article, outage constraints are included in the FBMC if the line loading in the outage simulation lies above a certain threshold referred to as CNEC threshold.

Without the information which bid at the energy exchange is linked to which generator and thus unable to exactly determine the effect of commercial exchanges on the network elements, the grid operators have to develop strategies to approximate the impact of trade results on the grid. This is accomplished by assigning linear participation factors to generators, which distribute the delta in net export position (NP) of a market zone to the generators most likely participating in this change. These factors are called Generation Shift Keys (GSK).

Having determined the relevant grid elements to incorporate in the market coupling and having forecasted the effects of a change in net position on the power flows, the grid operators have to determine the Remaining Available Margin (RAM) on these grid elements, which can be used for commercial exchanges. Due to the meshed nature of the European transmission grid, even in situations with balanced net positions, CNECs are loaded to a certain degree. The RAM is the remaining capacity, deducting the physical transmission capacity by loop and transit flows and a security factor to account for uncertainties in the FB calculation process (Flow Reliability Margin).

Within the Clean Energy Package (European Commission (2019)), the regulator has introduced a minimum share of physical capacities, which need to be made available for commercial exchange, the so-called minRAM. These minimum capacities were introduced in 2020, but allowed for derogations so they only come into full use in the year 2025.

2.2 State of Research

The body of literature on FBMC has been growing in recent years and is summarized in Table 1. It can be classified into three groups with regard to the scope of the analysis all of which provide important insights into different aspects of FBMC, which are presented in the following.

Table 1 Literature overview on articles covering FBMC
Full size table

The first group incorporates mainly conceptual work regarding FBMC (Kurzidem (2010), Mekonnen and Belmans (2012), Marien et al (2013), Schavemaker and Beune (2013), Sores et al (2013), Hagspiel et al (2014), Lang et al (2020), Poplavskaya et al (2020)) and fundamental research of parameter choice like base case calculation (Byers and Hug (2020), Felten et al (2021), minimum capacities (Henneaux et al (2021)) or Flow Reliability Margin (Weinhold (2021)) in the FBMC methodology. The research is mainly performed on stylized systems or with reduced spatial or temporal resolution, but provides important insights on what to keep in mind when applying FBMC to real life scenarios. While qualitative conclusion can be derived from the results, quantitative interpretation remains difficult due to the difference to real-world systems.

The second group covers analyses of FBMC on real-world systems but with reduced temporal or spatial scope. The majority of research in this group covers the analysis (van den Bergh et al (2016), Kristiansen (2020)) or simulation of (historic) FBMC results in the CWE region (Dierstein (2017), Felling et al (2019), Matthes et al (2019), Schönheit et al (2021a), Schönheit et al (2021b), Weinhold (2021)). Some articles look at different sub-regions, where the effects of FBMC are analysed, e.g. Finck et al (2018), who analyse different GSK strategies in the Central Eastern Europe region, Makrygiorgou et al (2020), who compare FBMC to NTC market coupling in several zonal combinations in South Eastern Europe and Bo et al (2020), who compare FBMC and NTC market coupling in the Nordic region. Zad et al (2021) propose a clustering approach for the flow based domain in the CWE region.

A key insight from groups one and two is that base case and Generation Shift Keys need to match realisation as close as possible for FBMC to have a cost-efficient outcome (e.g. Marien et al (2013). For GSK strategy “pro rate” seems to be a common understanding for good fit (Felten et al (2021), Schönheit et al (2020)). This follows the same idea of a close match of GSK and base case with realisation, hence GSKs based on the specific base case seem reasonable. Some authors raise awareness, that the analysis of cost-effectiveness and welfare impacts should not be limited to the analysis of market results, but need to incorporate the congestion management, which is necessary to ensure secure grid operation (Schönheit et al (2021b), Felten et al (2021)).

The third group of research consists of more comprehensive frameworks, which are capable of analysing the effects of FBMC in full-scale, consistent energy system scenarios, including an explicit grid representation with expansion path, power plant and RES (de-) commissioning with hourly resolution as well as the relevant methodological variety necessary for quantitative analysis of FBMC in the wider European context. Carlini et al (2020) present a detailed FBMC model for the Italy North capacity calculation region and while the real common grid model for this region is used, the framework considers a reduced flow based region not covering the Core region and is limited to historic data. Matthes et al (2017) present a framework to include security constraints into the FBMC methodology efficiently in large-scale power systems but are limited in scope to the CWE region. Marjanovic et al (2018) analyse the effect of an increased flow based region, which includes Italy North and Denmark West but misses Croatia and Romania not resembling the Core capacity region. Wyrwoll et al (2018) cover the Core region but neglect security constraints. Wyrwoll et al (2019) analyses the affect of base case methods in a focus region covering Germany and the electric neighbours. To the best of the author’s knowledge, no framework has been presented, which is capable of covering FBMC with an hourly resolution and accounting for security constraints, with a nodal representation of the Core region including consistent scenario information regarding grid expansion, power plant fleet and demand development. Such a framework is necessary to adequately assess quantitative implications of FBMC on the European Electricity Markets in the near future after the extension of the flow based region to the Core capacity calculation region. The article at hand presents a framework capable of these requirements and quantifies the implications of FBMC for several scenarios in the CWE and Core region.

3 Modelling Framework

This Section presents the modelling framework, which is used for the analysis. It introduces the modelling scope in Sect. 3.1 and provides information on the input data for the analysed scenarios in Sect. 3.2.

3.1 Model Description

The presented model framework consists of a Day-Ahead wholesale market model and a transmission grid model. Both models use the same input data, which are aggregated to bidding zone level for the zonal market simulation. The market model is used for NTC market simulations, which also represent the base case for FBMC, and to solve the market coupling with the derived FB constraints. The model structure is shown in Fig. 1.

Fig. 1
figure 1

Overview of input parameters and model flow

Full size image

The market model covers a unit-wise representation of the European thermal power plant fleet, which is based on Platts (2016), including the member states of the EU-27, except for Malta, Cyprus and Ireland, extended for Albania, Bosnia and Herzegovina, Great Britain, Kosovo, North Macedonia, Montenegro, Norway, Serbia and Switzerland. For the regionalization of renewable energy sources and demand, the methodology and weather data presented in Slednev et al (2017) and Slednev et al (2018) are applied. The NTC results of the model are used as base case from which power flows on the relevant grid elements are determined in normal operation and contingency cases for relevant outages. Additionally, the GSKs are determined for each grid node according to the generators energy output in the base case, as described in Sect. 2.1.

The grid model is the detailed transmission grid model presented in Ruppert et al (2020), which is extended for the countries to complete the Core region (Croatia, Hungary, Romania, Slovenia, Slovakia) and the high voltage direct current (HVDC) inter-connectors to neighbouring bidding zones. Grid data for these countries are based on ENTSO‑E (2021b) and ENTSO‑E (2016) including expansion measures from the Ten Year Network Development Plan (TYNDP) 2018 (ENTSO‑E (2018)) and national expansion plans, where applicable. The grid model is used for power flow studies, contingency screening and to derive the sensitivity matrix (PTDF), which are calculated using MATPOWER (Zimmerman et al (2011)).

The market coupling problem for each market time unit (e.g. hour) can be formulated as a linear optimization problem with the objective of maximizing social welfare given by Eqs. 1–13 in Appendix 1.

3.2 Scenarios

For the investigations, several scenarios for the year 2025 are used. Hence, the grid expansion state is a snapshot of 2025 including various expansion measures from the expansion plans. Figure 2 shows the transmission grid used in the study, where 380 kV lines are shown in red, 220 kV lines in green and HVDC lines in blue. The grid consists of 3,384 busses at 1,686 substations, with 6,680 AC lines and transformers, 24 HVDC lines (including inter-connectors) and 344 reactive power compensation elements. The installed capacity for thermal power plants is chosen according to Platts (2016) following assumption on the economic lifetime of power plants and already known decommissioning decisions. The installed capacity of renewable energy sources is chosen according to the “Best Estimate” scenario for 2025 from ENTSO‑E (2018). For the regional allocation, the generation profiles of renewable energies and temperature dependent load characteristics, weather data from the year 2016 have been used. To build a consistent scenario, the non-availability of conventional power plants correspond to the outages reported on the ENTSO‑E transparency platform for the year 2016 (ENTSO‑E (2021c)). The NTC values for the market simulations are chosen according to the ENTSO‑E Mid-term Adequacy Forecast 2020 (ENTSO‑E (2020b)), including the explicit profiles where given, but neglecting the stochastic unavailability. Also included are the external constraints for the Polish bidding zone. The market simulation is performed for 8,760 hours in weekly blocks. Storage optimization for pumped hydro storage power plants is performed for a weekly time horizon, where the state-of-charge is forced to 50% in the first and last time step, allowing for decoupling and calculation in parallel of the weekly blocks. Apart from Italy, which is modelled in two zones (Italy North and others), the market zones correspond to the bidding zones as they are present in the Day-Ahead market coupling. Reserves and strategic bids are omitted.

The different FB parameters for the experiments are given in Table 2. The scenario “NTC” represents the base case for the FB simulations and consists of a market simulation with NTC coupling constraints for the entire region. In scenario “FB1”, FBMC is applied in the CWE region, with a minRAM criterion of 20% as it was in use in CWE until 2020. In scenario “FB2”, FBMC is also applied in the CWE region, but the minRAM level is increased to 70% as intended by the Clean Energy Package. The scenario “FB3” also applies a minRAM level of 70%, but extends the geographic coverage of FBMC to the Core region. For all scenarios the PTDF threshold is chosen to 5% and the CNEC threshold to 80% to ensure comparability.

Fig. 2
figure 2

Electricity network model for the Core region for the scenario year 2025

Full size image
Table 2 Flow based parameter selection for the scenario calculations
Full size table

4 Results

For the three FB scenarios, the change in electricity prices, the net export position and the binding line constraints are analyzed. The average electricity price is shown in Fig. 3. As stated in Sect. 3.2, Italy is modelled in two price zones. Nevertheless, all zones are depicted but the southern zones have the same price. The aggregated net export positions for the bidding zones in the three scenarios are shown in Fig. 4. The results for the scenario “NTC” are shown in Fig. 6 in Appendix 3.

Fig. 3
figure 3

Average Price in €/MW h for each bidding zone in the scenarios FB1 (a), FB2 (b) and FB3 (c)

Full size image
Fig. 4
figure 4

Net export position for each bidding zone in TW h in the scenarios FB1 (a), FB2 (b) and FB3 (c)

Full size image

The first comparison regards the impact of increasing the minimum capacity threshold (minRAM) for the CWE region. Keeping everything else constant, the net export positions, mean prices and binding grid constraints are compared for an increased minRAM level from 20% (FB1) to 70% in (FB2).

Most prominent is the change in the French bidding zone. The average electricity price increases by up to 11%, due to an increased export position, which grows by 9%. Prices in Belgium, Spain, the Netherlands and Portugal also show increased levels. In the southern Nordic zones prices decrease, which is induced by lower prices in Germany, without a significant change of net positions. The Baltic countries, Central and Eastern Europe and the Balkan zones also show reduced price levels compared to the 20% minRAM scenario.

The increased minRAM reduces the number of binding grid elements. This mostly affects intra-zonal constraints in Belgium and Germany, which are no longer in any hour restricting the market. This demonstrates the effectiveness of minimum trading capacities to reduce the importance for internal lines for the market clearing. Figure 5 shows that the majority of binding lines in the scenario with a minRAM of 70% are inter-connectors or adjacent elements. Another important result is the share of constraints being replaced by the minRAM condition. The minRAM share in Table 3 is indicating the number of CNECs that are subject to minRAM, i.e. where the available capacity for trading would be lower than allowed by the regulator. As anticipated, the number of replaced constraints increases from 16% in “FB1” to more than 50% with a higher minRAM in “FB2”.

Fig. 5
figure 5

Number of hours when CNECs are binding for different levels of minRAM; 20% in scenario FB1 (a) and 70% in scenario FB2 (b) for FBMC applied in the CWE region

Full size image
Table 3 Numerical Scenario data
Full size table

The difference between scenarios “FB2” and “FB3” is the extension of FBMC from CWE to the Core region, then covering a major part of the coupled European electricity markets in scenario “FB3”. For bidding zones in Western Europe the effects are similar to the higher level of minRAM in CWE, with further increasing prices in France, Great Britain and Spain. Consumers in Romania and Hungary profit the most with regard to price levels. Despite not being part of the FB region, the southeastern European zones also show reduced price levels. Polish prices and net positions are mainly determined by the external import constraint, which makes the results comparably stable. The Nordic zones are almost not affected. Prices in the Baltic zones and some Central European zones increase slightly, making it difficult to identify a general trend. Interestingly, countries connected through NTCs to the FB region show relatively constant results across the investigated scenarios. This indicates that, while the distribution among members of the FB region changes, the fundamental exchange patterns from and to the FB region remain unchanged in the calculations. For the binding grid elements, the effect is also similar to the observations in the CWE scenarios. Increased levels of minRAM reduce the number of binding constraints and push considered grid elements towards the inter-connectors, which are heavily impacted by cross-border flows. The total number of constraints is highly sensitive to the extension of the flow based region from CWE to Core as shown in the third column of Table 3. As each line generates two constraints (one for the flow in each direction), the number of RAM is twice the number of relevant network elements. The total number of RAMs in the model varies between 16.7 million for CWE and 28.3 million for the Core region. The number of CNECs also varies widely between hours, indicating the different utilization of the grid in different situations. The minimum number of CNEC in one hour over all scenarios is 996, while the maximum occurs in the Core scenario consisting of almost 15,600 constraints in a single hour.

The highly volatile number of constraints indicates that the possible exchange might be limited differently in changing grid conditions, i.e. they depend on the supply and load distribution in the grid. This is increasingly important in energy systems with high shares of RES, where supply depends upon volatile weather conditions and demand is more and more flexible due to new and flexible consumers. This makes the suitability of static NTCs, as they are often applied in scenario-based analysis of the energy system, highly questionable. Besides the aggregated implications presented here, this might result in more frequently changing operational patterns for power plants, with implications for life-time and flexibility needs, which should be addressed in further research.

Compared to the most restrictive scenario “FB1”, which resembles FBMC in CWE until 2020 and is already found to increase transmission levels compared with the NTC market coupling, which was in operation beforehand ( CWE TSOs (2015)), the target scenario for the European internal market (FB3) with FBMC introduced in the Core region and a minRAM of 70% shows decisive differences with regard to prices and net export positions, both most notably visible in France. The difference in average price is up to 13%, indicating increased export of relatively cheap nuclear energy. Indeed this impression is confirmed by a difference in net position of 10% between the two scenarios.

A last remark regards the difference between the FB results and the NTC market outcome, where a surprisingly large discrepancy is visible. French price levels and net positions in the scenario “NTC” are most closely resembled by the Core scenario with a minRAM of 70% (FB3). For possible reasons and implications, see Sect. 5.

5 Critical review and wider implications

Although the model accounts for the most important aspects of the market coupling process to assess the differences in parameter choice, the presented approach is not without limitations. The biggest limitation stems from the transmission network model. The model is a static representation for the entire calculated year. Moreover, the topology is reduced to one bus bar per voltage level and substation. This reduces the ability to account for topology adjustments carried out by the grid operators.

Furthermore, the considered contingencies only consist of line and transformer outages. Generator outages or failures of HVDC lines as well as contingencies of multiple elements are not considered. However, comparing the FBMC results with the NTC outcome, the results of the power flow studies seem to be rather too restrictive in the FB cases. Additional constraints from more sophisticated outage consideration would restrict the FB domain even further.

The power plant outages are modelled according to the historic 2016 outages to be consistent with the weather year. While this helps to account for e.g. low river levels and resulting non-availability, no data is included for power plants commissioned after 2016. In the considered scenario year, this leads to underproduction for units, which showed above average outages in 2016, while overestimating the production of plants without any modelled outages. On the aggregated bidding zone level, these effects can assumed to be minor.

Uncertainties, with regard to load, wind and solar generation as well as outages are neglected in the model, which uses perfect foresight. Also not accounted for are the intraday and reserve markets. While the the impact of reserves is neglected, the approach partly accounts for the intraday price effects, by clearing the entire electricity demand in the Day-Ahead time frame.

Finally, the calculations show that the FB scenarios mostly show lower exchanges than the NTC reference case, especially with regard to the French net position. This is an effect already reported by Marjanovic et al (2018), who attribute it to a misfit of the NTCs and the grid scenario. Apart from this, there are two possible explanations for this supposed contradiction. One is the conservative operating voltage assumed, when calculating the flow limits of the lines. These are based on 380 and 220 kV respectively, while in reality many lines are operated above nominal voltage. An increase of 10% would equally increase line limits and allow for larger exchanges.

Looking at the CNEC loading graphs in Fig. 5, a relatively large number of lines in Belgium are binding. The grid operator in Belgium, is already applying Dynamic Line Rating (DLR) in the CWE FBMC and is expected to use this tool intensively in the future. DLR allows for dynamically allocating lower or higher flow limits on particular lines, depending on the current weather conditions (radiation, temperature, wind speed). This allows for potentially higher flows on the lines, especially in hours where large amounts of wind energy are fed into the grid, in contrast to the static ratings applied in the presented approach throughout the year. Finally, the assumptions under which the NTCs for the scenario were calculated are not known do the author. This might affect the eastern European regions, where the exchanges are systematically overestimated, which might be due to missing minRAM consideration in the NTC calculation for these regions.

The analysis results in three key insights:

  1. 1.

    Increased minimum trading capacities lead to more exchange, hence more base load capacity can be used. This usually consists of nuclear or lignite power plants. In the analysed scenarios, the strongest impact on the European energy mix is the increased French nuclear position, where larger trading capacities allow for additional output of up to 10 TW h. To put this into perspective, when comparing a scenario with or without minRAM, the difference in CO\({}_{2}\) free generation could amount to around 100 TW h over the course of 10 years (around 15% of annual electricity demand in Germany today), which might have significant implications when evaluating the scenario.

  2. 2.

    Increased minRAM also reduces the amount of binding grid elements, which are shifted towards the inter-connectors. In this context, it might be tempting to consider market coupling approaches sufficient, which exclusively use inter-connector capacities in whichever form for market coupling. However, the intra-zonal distribution of flows, which results from the geospatially referenced supply and demand sources has a decisive impact on the FB domain, as shown by the changing number of binding constraints in the different hours of the year. Due to the more and more prominent effect of RES in the system, dynamic market coupling constraints, which consider the specific grid situation, have to be taken into account to represent the energy system adequately.

  3. 3.

    When analysing the effect of extending the FBMC region from CWE to Core, or in general comparing FBMC against NTC market coupling, care has to be taken concerning the NTC values, which are applied in the comparison. Poorly chosen NTCs or values inconsistent with the grid expansion state, make the comparison difficult and could lead to unintuitive results. This is also confirmed by Marjanovic et al (2018).

6 Conclusion and Outlook

In this article, a framework to model Flow Based Market Coupling (FBMC) for the internal European Electricity Market is presented. The developed framework is used to study the impact of different minimum transfer capacities (minRAM) for FBMC and compare its application in the Central Western Europe and the extended Core region, where FBMC is going to be introduced in the near future. FBMC is the target method for market coupling in Europe, but not considered in many studies, which analyse future scenarios of the European energy system. When working with such system studies, practitioners should be aware of the differences the market coupling method brings along and the implications for the market results. The presented results show that larger minRAM increase price convergence in the flow based region, while also reducing the number of relevant constraints for the market coupling process. Larger minRAM reduce the number of binding constraints in the market outcome, especially the number of internal grid elements as it is intended by the introduction of minRAM. This allows for more base load generation across Europe, which also affects emissions depending on the composition of the power plant fleet. In particular, the difference in minRAM level and geographic scope of the area, where FBMC is applied can account for price differences of up to 13%. In the specific scenario analysed, the increased base load generation mainly consists of emission free nuclear output in France, which increases by up to 10 TW h. This clearly highlights the important role, which exchange capacities play in energy system scenarios. Due to changing utilization patterns in the grid because of weather dependant generation from renewable sources, dynamic market coupling constraints need to be considered instead of static values for net transfer capacities (NTC) in order to adequately represent the energy system. For the comparison of NTC and FBMC in future scenarios, care has to be taken with regard to NTC selection, as they have to represent a grid expansion state consistent with the scenario analysed. With regard to future work, the insights from the literature that welfare implications need to be considered across the entire congestion management process, should be investigated in the Core region, to quantify shifts in welfare distribution between FBMC and NTC and the implications of minRAM on the grid operation and congestion management. Finally, future research should focus on the integration of dynamic line rating into Flow Based Market Coupling and the economic effects, as this methodology promises enormous potential for increased flow limits in the majority of market situations.