Ellen Beckstedde, Leonardo Meeus
The support of decentralized energy resources under the Fit for 55 package and the REPowerEU plan places distribution grid users and distribution system operators (DSOs) at the center of the future European energy system. Also, the interaction between both types of agents is gaining importance for two reasons. First, DSOs face challenges connecting these new grid users to their network, leading to an increased need for grid investments and congestion management measures. Second, engaging these new grid users can bring opportunities for DSOs to manage their network and its possible congestion more efficiently.
In this article, we describe the need, organization, and open issues of congestion management in distribution grids. We focus on Europe, and we will address the following questions. Do we already have congestion in distribution grids, and how did that happen? Do we plan to have more congestion in distribution grids, or will we avoid congestion with investment planning? How do DSOs procure grid services to solve congestion, and what are the main differences? What are some of the open issues?
In Europe, DSOs are increasingly faced with congestion in their grids. It started in countries like Germany with injection peaks caused by wind and solar farms that created situations with more generation than load in some areas, sometimes congesting the local lines or transformers. It then spread to countries like The Netherlands, where DSOs started to experience congestion due to generation peaks from renewables and load peaks from new data centers. The next wave of grid congestion is expected to come from electric vehicles (EVs). Leading countries in EV penetration, like Norway, already have distribution grid congestion caused by EV charging. The United Kingdom is also experiencing congestion in distribution grids, which is mainly driven by EVs or renewable generation, depending on the area.
As illustrated by Figures 1 and 2, heatmaps or hosting capacity maps are often used by DSOs to report congestion, but there are different practices.
Figure 1 is an image from The Netherlands for new grid connections of load [Figure 1(a)] and generation [Figure 1(b)]. Red means that all network capacity has been reserved for other grid users, and you cannot connect anymore in that area. Orange indicates that you cannot connect unless certain congestion management measures are taken. Depending on the case, these measures can be limited capacity contracts or market-based redispatch. The shaded areas indicate where congestion management measures are already in place. Yellow means that the connection is uncertain; there is an application procedure to follow that will tell if you can connect. Only in transparent zones can you connect without capacity limitations.
The ongoing debate in The Netherlands is to what extent the DSOs should continue connecting new grid users. The more they overbook the network, the more they will need to resort to congestion management in peak hours. If they do not overbook, the connection queues start to be awkwardly long. As a result, they have entered into discussions on who should get priority to connect. Should it be a local housing project or a data center of a multinational? Should it be first come, first served, or should there be auctions for distribution grid connection capacities?
Figure 2 is an illustration from the DSO Schleswig-Holstein Netz in Germany. The colors represent the number of hours renewable generators have been curtailed in a selected period. Red means that you can still connect new renewable generation projects, but you have a higher risk of being curtailed if you do. For example, renewable generators located in the dark red zones of Figure 2 were curtailed for more than 16,000 h in 2022. Green means that no curtailment was necessary for that area.
German DSOs used to curtail renewables in real time at the cost of the lost revenues of these generators under the so-called feed-in management regime. In October 2021, this regime was replaced by Redispatch 2.0. In more detail, congestion is now managed by a schedule-based process that contains several validation loops between DSOs and transmission system operators (TSOs) and ends 15 min before real time. All generators, such as renewable energy, combined heat and power plants, conventional units, and storage facilities, with an installed capacity above 100 kW must provide congestion management services. To ensure cost efficiency, system operators select generators based on their imputed costs, which consider the technical impact of the generator on the congestion issue and the feed-in priority for renewables. A final difference is that system operators have to compensate for the lost revenues of the generators and the imbalance costs of its balancing responsible party.
In many other European countries, congestion in distribution grids is still far from being a concern for stakeholders. However, the lessons learned from countries like the ones mentioned reveal that congestion can rapidly become an issue in certain zones, catching DSOs unprepared. The decisions for grid users to invest in renewable generation, build a new data center, or switch to an EV are quicker than the typical grid expansion planning and execution processes. This issue is already well known in transmission grids, and the same is now happening at the distribution level. The main difference is that including network constraints in market pricing algorithms is more challenging for distribution than transmission. For instance, the IEEE community has already worked on theoretical models for distribution locational marginal pricing. However, these approaches are not yet considered an actual solution to manage congestion in distribution grids.
For more than a decade, European transmission investment plans have been publicly discussed. These national plans are developed with standardized methodologies and coordinated by a pan-European strategy. This exercise, led by the European Network of TSOs for Electricity (ENTSO-E), is referred to as the Ten-Year Network Development Plan. The plan, which is updated and improved every other year, has been an impressive achievement of harmonization and collaboration across many countries.
In the first two decades of electricity market reforms, congestion in distribution grids has not been an issue. But recently, it became evident that distribution grids can turn into a bottleneck for the functioning of the European electricity market and the transition toward a more sustainable energy system. Article 32 of Electricity Directive 2019/944 of the European Union (EU) Clean Energy Package introduced several new regulations for distribution network planning. The legislation uses the terminology “network investment plans for distribution systems,” but some are already talking about Ten-Year Network Development Plans for distribution. DSOs have promoted the EU DSO Entity, aimed at replicating the role of ENTSO-E, to develop a new methodology for the future investment plans of distribution grids that all DSOs will apply. In the meantime, different approaches to designing these network investment plans are emerging.
On the one hand, DSOs gathered via their industry associations and asked consultants to produce a first European plan as a dry run. On the other hand, DSOs have already published the first version of their local plans to comply with the new regulations of the Clean Energy Package. We now introduce both approaches.
There is not yet a consensus on the actual potential of flexibility as an alternative to distribution grid investments. Some argue that cost-reflective distribution network tariffs would bring enough incentives for grid users to reduce their peaks. We believe there is a potential for DSOs to do more than provide cost-reflective signals via their network tariffs. One reason to defend the need to explicitly procure flexibility is that tariffs will always depend on the grid users’ voluntary response and be imperfect as they compromise between cost-reflectiveness and other principles, such as fairness and simplicity. Another reason is that investment planning under uncertainty can result in unexpected congestion.
The European countries that currently experience congestion in distribution grids indeed did not plan for it, but they still have to deal with it. The experience has shown that DSOs cannot simply stop all requests to connect to distribution grids; they are subject to significant pressure to overbook and manage the congestion resulting from this overbooking. An additional concern is that grid users could start to create congestion, anticipating that they can get paid to solve it (i.e., inc-dec gaming). Gaming is a valid concern limiting the potential of market-based flexibility, but we believe it will not apply equally in all situations. When and how DSOs will contract flexibility also plays a role, which is what is discussed next.
Many DSOs in Europe have set up demonstration projects to test flexibility services to manage (potential) congestion in their grids. Figure 5 illustrates some of the biggest projects financed by the EU’s Horizon 2020 research and innovation program and the countries that have hosted the demonstrations.
DSOs with a lot of congestion in their networks evolved from demonstration projects to full-scale flexibility markets. Some DSOs, such as Enedis and Enel, have developed their own platforms to tender flexibility services, but market platforms owned and operated by third-party companies also entered this space. In what follows, we will discuss three of these third-party platforms that are currently the most relevant ones in Europe in terms of procured volumes or capacities: NODES, Piclo Flex, and GOPACS. All initiatives started in countries that were among the first to experience congestion in distribution grids: Norway (and Germany) for NODES, the United Kingdom for Piclo Flex, and The Netherlands for GOPACS.
Table 1. Contracted capacity or traded flexibility volumes on third-party market platforms.
A fundamental difference in the approaches of the United Kingdom, The Netherlands, and Germany exists. In the United Kingdom, the DSOs really plan for flexibility. They make the tradeoff between distribution grid expansions and procuring flexibility. UKPN, for example, recently committed in their RIIO-ED2 Business Plan 2023–2028 to 410 million £ of deferred load-related investments through the use of low-voltage flexibility. They estimated the cost of the flexibility services based on their experience with flexibility tenders. The DSOs in The Netherlands did not plan to use flexibility. They are forced to overbook the grids as they cannot follow the demand for grid connections and then have to procure flexibility to solve the resulting congestion in their grids. This situation is not the result of a cost-benefit analysis.
The DSOs in Germany are also in a different situation. They have also been overbooking their grids because there was a bigger demand for grid connection than they could offer, leading to high curtailment rates in certain areas. However, after controlling the most severe capacity issues with network investments, German DSOs can do a cost-benefit analysis to compare the cost of curtailment with the investment cost to expand their grids. In more detail, they can consider a curtailment of 3% of the annual output of each connection point in their network planning. In this context, buying flexibility services can be an alternative to compensating grid users for curtailing them. In other words, the German situation nicely illustrates how we can avoid DSOs being at the mercy of flexible service providers to solve congestion in distribution grids (the biggest worry of some skeptics).
There are many open issues. In what follows, we illustrate a few.
Flexibility services are operating expenditures (OPEX), and DSOs typically have efficiency benchmarks for OPEX with rewards if they outperform their OPEX baseline and penalties if they underperform. Distribution grid investments, however, are treated differently as capital expenditures (CAPEX). Once approved, CAPEX enter into the regulated asset base, on which the DSO receives a regulated rate of return. When DSOs use flexibility as an alternative to distribution grid investments, OPEX (cost of flexibility services) increase and CAPEX (cost of investments) decrease, negatively impacting their efficiency benchmarks and return on investments.
The regulatory authority in the United Kingdom, Ofgem, has been one of the first to address this financial disincentive by introducing what it refers to as the total expenditures (TOTEX) approach. It implies that a fixed share of the TOTEX (OPEX and CAPEX) can enter into the regulated asset base, which gives DSOs incentives to consider flexibility as an alternative to grid investments. Today, there is an ongoing discussion on whether to address this disincentive with regulatory measures. The most advanced incentive regulation schemes developed to address this issue have reached an inadvisable level of complexity. Considering that DSOs are under pressure anyway to keep their network tariffs under control, maybe the current push for more transparent network investment plans can be sufficient to compensate for the financial disincentive.
While the main focus of this article is on flexibility markets, there are also other ways to source flexibility. Generally, the provision of flexibility can be mandatory or voluntary, and flexibility contracts can be short or long termed. Table 2 illustrates both approaches by mapping different flexibility tools on these two dimensions. While each approach has its opportunities and disadvantages, the magnitude of these effects still needs to be determined. As a result, DSOs are examining different ways to contract flexibility in their networks. For example, the Dutch DSO Liander currently considers four congestion management alternatives to connect new grid users in congested network areas. Two types of short-term flexibility markets are tested using the GOPACS platform characterized by voluntary or mandatory participation of this new grid user in the market. Besides that, new grid users can also enter two kinds of long-term connection agreements, with and without day-ahead curtailment announcements by the system operator.
Table 2. Illustration of the two approaches to source flexibility using existing flexibility tools.
It will be interesting to learn more from theory and practice about the optimal approach to source flexibility and the interdependence of this choice on local network characteristics, such as the number of available flexible resources, grid topology (rural, urban, etc.), voltage level (low voltage, medium voltage, etc.), and congestion cause (renewables, EVs, data centers, etc.). Also, it will be important to better understand the pros and cons of combining different flexibility tools. While incompatibilities among the different approaches might exist, we also see opportunities for combining them, for instance, long-term flexibility contracts (voluntary or mandatory) with shorter term flexibility markets.
We have discussed the challenges and opportunities of procuring flexibility from a DSO perspective. However, the DSO’s activation of flexibility might also impact other energy stakeholders, such as the TSO. There are at least two interactions between TSOs and DSOs to consider. First, TSOs and DSOs might want to access the same flexible resources for different grid services, such as congestion management and balancing. This competitive interaction among system operators might create a need for cooperation or sequence in selecting flexible units. Second, TSOs and DSOs might impact each other’s networks when activating flexible resources for their own purposes. When the activation of flexibility moves closer to real time, there might be a need for coordination or validation mechanisms between the system operators to avoid network issues.
Many stakeholders and academics have already recognized the importance of TSO–DSO coordination, which led to the development of different coordination schemes for the TSO’s balancing and the TSO’s and DSO’s congestion management services. However, translating these coordination schemes into practice is often difficult because of the complexity of the problem and the required information sharing among the stakeholders. Therefore, new regulations to manage the described interactions among system operators might arise in the meantime. An example is the European System Operation Guideline, which allows DSOs to refuse the participation of flexible resources to the TSO balancing market based on technical reasons. It is only remains to see how these rules and coordination schemes will evolve in the coming years.
This article has four key takeaways, each relating to one of the discussed questions on congestion management in distribution grids.
In other words, when “fit and forget” is not an option anymore, we will have DSOs that proactively engage in flexibility and DSOs that might regret they did not; hence the title.
E. Beckstedde et al., “EUniversal Deliverable 10.3: Regulatory recommendations for flexibility options and markets,” 2022. [Online] . Available: https://euniversal.eu/wp-content/uploads/2022/01/EUniversal_D10.3_Regulatory-recommendations-for-flexibility-options-and-markets-2.pdf
“TSO-DSO report: An integrated approach to active system management,” CEDEC, EDSO, ENTSO-E, Eurelectric, GEODE, Brussels, Belgium, TSO-DSO Rep. 2019. [Online] . Available: https://docstore.entsoe.eu/Documents/Publications/Position%20papers%20and%20reports/TSO-DSO_ASM_2019_190416.pdf
S. Chondrogiannis, J. Vasiljevska, A. Marinopoulos, I. Papaioannou, and G. Flego, “Local electricity flexibility markets in Europe,” European Union, Luxembourg City, Luxembourg, Tech. Rep. JRC130070, 2022.
“Active power services implementation plan,” Energy Networks Association, London, U.K., 2020. [Online] . Available: https://www.energynetworks.org/industry-hub/resource-library/open-networks-2020-ws1a-p3-final-implementation-plan.pdf
“Connecting the dots: Distribution grid investment to power the energy transition,” Eurelectric, Brussels, Belgium, 2021. [Online] . Available: https://www.eurelectric.org/connecting-the-dots/
L. Meeus, E. Beckstedde, and A. Nouicer, “Towards a regulatory framework for the use of flexibility in distribution grids,” Oxford Energy Forum, Future of Energy Networks Issue, Oxford, U.K., 2022. [Online] . Available: https://www.oxfordenergy.org/publications/the-future-of-energy-networks-in-a-decarbonized-world-issue-134/
“Investeringsplan 2023-2032,” Fluvius, Melle, Belgium, version 8, Jun. 2022. [Online] . Available: https://over.fluvius.be/sites/fluvius/files/2022-06/investeringsplan-2023-2032.pdf#page=6
Ellen Beckstedde is with Vlerick Business School, 9000 Ghent, Belgium and KU Leuven, 3000 Leuven, Belgium.
Leonardo Meeus is with the European University Institute, 50133 Florence, Italy; KU Leuven, 3000 Leuven, Belgium; and the Florence School of Regulation, 50133 Florence, Italy.
Digital Object Identifier 10.1109/MPE.2023.3269545
Date of current version: 21 June 2023
1540-7977/23©2023IEEE