Experimenting with co-ownership of energy storage facilities - A case study of the Netherlands

The objectives of energy policy in the Netherlands and the need for flexibility in the energy grid

The Netherlands has set an ambitious agenda for the decarbonization of the Dutch electricity system. Under the European Green Deal – and the subsequent Fit-for-55 and REPowerEU packages – Member States are to ensure a 55% reduction in CO2 emissions from the 1990-benchmark by 2030. ¹⁸ In the Dutch Climate Agreement (Klimaatakkoord) the Netherlands has set its own target of 60% reductions by 2030 and a fully decarbonized energy system by 2035. ¹⁹ The end goals is for the Netherlands to be completely carbon neutral by 2050. This article distinguishes between the short term objectives (2030) and the long term objectives (2050) and focuses in particular on the short-term objectives. ²⁰

While the ambitions of the Netherlands are in line with obligations imposed by EU legislation, the plans have fallen behind in execution. ²¹ The first obstacle exists in the extreme congestion issues in the Netherlands. In the provinces of Limburg and Noord-Brabant, it is not possible to realize any new connections to the energy grid for large users in the foreseeable future. ²² It is likely that these problems will follow for other regions. ²³ This significantly limits the ability of Dutch enterprises to be founded or expanded, thereby harming the economy. Moreover, congestion issues and negative prices are harming incentives to invest in sources of renewable energy. During peak hours of production there is a structural risk of overloading the energy grid, which leads to low or negative electricity prices throughout large parts of summer. ²⁴ With the rapid expansion of wind energy, a likelihood of low or negative prices continues to exist in other periods of the year and times of the day as long as this energy cannot be efficiently transported and used. This leads to reduced investments at both the commercial and individual level. For instance, the construction of a proposed wind park by Vattenfall has been cancelled due to a combination of high construction costs, geopolitical uncertainty and low electricity prices. ²⁵ It is also becoming more likely that the net metering arrangement ²⁶ will be retracted for households and that extra tariffs are charged to households with solar panels. ²⁷ This may dissuade households from investing in Photovoltaic (PV) installations, but may lead to other efficiencies, for instance investments in energy storage solutions or smarter use of energy by households. ²⁸ In short, obstacles to the emergence of flexible capacity (including storage) may delay the energy transition as a whole.

Aside from technical problems, the Netherlands is also facing legal obstacles to establish novel constructions for the production or use of energy such as energy communities, peer-to-peer trading or other types of energy hubs. The Dutch government has not yet implemented Directive 2019/944 in national law. The proposed Energy Act has been delayed significantly, and current legislation does not facilitate these legal constructs. ²⁹ In response to these issues, the Netherlands has developed a National Action Program for Net Congestion where it attempts to develop strategies as to resolve congestion issues. This program is based on three pillars: building faster, steering more, and increasing flexibility. ³⁰

In the pillar ‘building faster’, the Dutch government proposes that decentralized governments such as provinces and municipalities should provide more direction to grid managers on how to expand and ease procedures for permitting, that there should be more collaboration between grid managers, and that the available resources should be used more efficiently. ³¹ It has however been observed in expert reports, that there is insufficient personnel to fortify the energy grids in a timely manner. If the strategy would be to solely rely on grid fortification, the congestion issues cannot be resolved in a timely manner. ³²

The pillar ‘steering more’ would allow grid managers to connect more – or certain – parties even in the presence of congestion issues. Here, there are several concrete initiatives. Firstly, the Dutch National Regulatory Authority, the Autoriteit Consument & Markt (ACM), has allowed for ‘societal prioritization’. ³³ This allows grid managers to prioritize the connection of certain parties over others, rather than relying on the first come first served principle. In principle, the grid manager – using guidelines provided by the ACM - would determine whether a requestee has an important societal function (for instance if it concerns a school or a hospital) or if a party is able to help in resolving congestion issues (for instance through the inclusion of integrated battery storage). ³⁴ It is then up to the grid manager to determine which connection has more societal benefit than another, which it then has to justify towards the ACM. ³⁵ This can improve the societal benefits of the created connections, but it does not necessarily resolve congestion issues over time. Moreover, it presupposes that any large connections are still possible at all, which is not the case in Noord-Brabant or Limburg. ³⁶ There is also an initiative to change the tariff structures so that firms which are not connected for peak-load capacity can be compensated. This means that a firm will receive a smaller connection and that it may have to lower its energy usage upon request. The firm will be compensated for this type of connection. ³⁷ These initiatives are helpful in mitigating the crisis and increase energy efficiency, but may not be sufficient to resolve congestion in the long-term without heavily fortifying the energy grid or increasing (non-fossil) flexibility, or using energy storage solutions as a buffer for the energy grid. ³⁸

The third pillar, increased flexibility, explores possibilities to use energy storage, conversion, or peak-shaving products to increase transport capacity and alleviate congestion issues. Here, it discusses increased oversight and evaluation by public authorities, increased public-private cooperation to create flexibility and identify congestion issues, utilizing smart solutions, and facilitating the creation of energy hubs to alleviate the grid. ³⁹ Increasing the flexible capacity of the energy grid will be a necessity in both the short term and the long term. In the short term it is necessary to complement efforts to fortify the energy grid in a timely manner, for the long-term objectives flexibility is necessary to achieve the 2050 objective of a carbon-neutral society. Koolen et al. have noted that the Dutch ambitions with respect to producing energy from wind at sea would lead the Netherlands to have the highest relative (per capita) share of wind energy for the total demand in 2030, but that as a result of the Dutch wind-at-sea and renewable plans, the Netherlands would also have to realize the highest relative growth in flexibility capacity in the timelines 2021 to 2030 and 2031 to 2050. ⁴⁰ Thus, rapidly expanding flexible capacity is not only important to achieve the short term objectives of resolving congestion issues, but also to secure the long term objective of a carbon neutral society and the Dutch ambitions related to wind parks at sea. The upcoming section discusses how energy storage can be used to introduce this flexibility, and the issues that arise.

Exploring the obstacles in an emerging market for energy storage in the Netherlands

Energy storage and the conversion of electricity (from renewable sources) into alternative energy carriers is one of the methods to introduce more flexibility into the energy grid. The Dutch government has published its Roadmap Energy Storage (Routekaart Energieopslag) where it explores options. ⁴¹ Here, it distinguishes between electricity storage in batteries, conversion to molecules, and storing energy as heat. The Dutch government expresses faith in energy conversion and heat networks as important technologies in creating an emissions neutral society. ⁴² However, these technologies lend themselves poorly to resolving congestion issues in the short term. The market for molecule storage – predominantly in hydrogen – will likely be scaling up towards 2030. However, its impact on the energy system is likely limited as electrification remains the main strategy in the energy transition. ⁴³ Possibilities for heat storage depend largely on available local heat sources where it concerns heat sources with low temperatures or developing infrastructure for storages at high temperatures. These are important alternatives to study in light of the 2050 objectives, but the large-scale implementation of battery storage is the most likely candidate to provide energy storage en route to 2030. ⁴⁴

Electricity storage through batteries can happen in different modalities and for different purposes. The Roadmap distinguishes between home batteries (installed behind the meter with small users, mainly for self-consumption of produced energy), collective batteries (installed in front of the meter by groups of (small) users, mainly for self-consumption of produced energy), large scale integrated batteries (installed behind the meter with large users, mainly for self-consumption of produced energy), and stand-alone battery services (provided commercially). Each of these types of battery storage has their own business case – or lack thereof – and effects on grid congestion. ⁴⁵ This article focuses predominantly on the possibility to operate stand-alone batteries, as the use of battery storage is most discussed as a short-term solution which can be brought to scale and shared ownership is mostly relevant to shared ownership between grid managers and commercial parties active in large scale energy storage. ⁴⁶ However, the findings of the article are technology-neutral, and shared ownership of energy storage solutions can be utilised for any other type of storage facilities.

The Roadmap notes that there is currently around 185 MW of energy storage capacity ⁴⁷ in operation in the Netherlands. TenneT – the sole TSO in the Netherlands – has noted that it will need around 10 GW of battery storage capacity in 2030 to support the high voltage energy grid. This does not yet provide insights into the required battery storage to resolve local congestion issues. ⁴⁸ It is therefore vital that the expansion of energy storage facilities happens in a short timeframe. There is significant interest from third parties in supplying and operating energy storage solutions in the form of battery storage. ⁴⁹ A report by Stratergy (2023), which is referred to in the Roadmap, estimates that there is around 34 GW in battery storage in the pipeline in terms of requests. 28 GW of this capacity would be requested with Tennet, while 6 GW is requested for connection at DSOs. ⁵⁰ These requests are however in different stages, ranging from pre-investigations to actual requests, the Roadmap and the expert report by CE Delft note that only a small number of these requests is likely to be finalized. CE Delft notes that only around .5–2 GW can be profitably operated. ⁵¹

The tension between what is needed, what is profitable, and what is legally allowed, raises complex questions about the integration of electricity storage into the energy grid. ⁵² There is a need for flexibility and energy storage solutions to increase significantly within a matter of years, but relying purely on market mechanisms is unlikely to be viable. It is uncertain when cost-efficient technologies will be on scalable, and there is no legal framework that allows for the exploration of alternative energy solutions such as peer-to-peer trading yet. The Roadmap Energy Storage sets out several technical and financial obstacles and several legal obstacles.

The Roadmap set out by the Dutch government recognizes the stalemate concerning the installation of additional battery capacity, noting the lack of a business case for market participants and obstacles related to grid congestion, which in turn negatively affect investment incentives. ⁵³ The report observes several legal and policy barriers which contribute to the stalemate. ⁵⁴ Firstly, tariff structures may further hinder the business case of operating battery storage. As battery storage is not a recognized activity, tariffs are charged for loading the battery as if it were a user and for unloading the battery as if it were a producer. With these double tariffs, batteries become less likely to operate profitably, but the ACM also does not (yet) want to exempt operators of battery storage from paying these tariffs altogether. ⁵⁵ Secondly, there is little experience with decentral governments in approving battery storage projects. Here, more guidance is necessary from the central government, as well as possibly the ACM and grid managers. Thirdly, relying on subsidies to make batteries profitable raises the societal cost of batteries and may lead to over subsidization. ⁵⁶ For instance, CE Delft (2021) published a study that shows that with targeted stimulus for 5.5 GW of battery storage, it would be possible to install an additional 7.5 GW of solar energy in solar parks. This would prevent curtailment and without leading to congestion issues. However, the costs of batteries for this purpose would cost €2.500 per ton of CO2 emissions prevented per year, while the law dictates the maximum costs would be €300 per ton of CO2 emissions prevented per year. ⁵⁷

The Roadmap concludes on stand-alone battery storage by arguing that – based on the report by CE Delft (2023) – these batteries could add to congestion, act congestion neutral, or resolve congestion. Under the current market conditions, it would be highly likely that they add to congestion if operators of battery storage are to operate profitably. The government and ACM could steer use and use subsidies to ensure that they operate congestion neutral, which would increase transport capacity but not resolve congestion. Moreover, subsidies would add to the societal costs of increasing energy storage. ⁵⁸ The report concludes by noting that it is unlikely that stand-alone battery services can resolve congestion issues. ⁵⁹ Here, shared ownership between grid managers and third-party market players may however provide outcomes. This is explored in section 3.

Conclusions

The Netherlands faces significant congestion issues which delay efforts in creating the energy transition. Increasing the capacity for energy storage may help to mitigate congestion issues in the short term and help to achieve the long-term objectives set out in EU and national policies. However, the emergence of a market for energy storage is hindered because of a misalignment between the interests of grid managers and third-party operators of energy storage services. The Dutch Ministry of Economic Affairs has investigated how it could break the stalemate but finds that the use of batteries is either not economically viable, raises societal costs to an unacceptable level, or leads to additional congestion of the grid. The upcoming section explores how the under investigated possibility for shared ownership of energy storage services between grid managers and third-party market players may help to break this stalemate.

The changing role of grid managers in the energy transition: from facilitator to coordinator

Grid managers have been bound to several obligations and principles under EU (and subsequent national) laws. Grid managers operate a regulated monopoly on distribution activities within their area of operations. This involves rules on tariff setting, unbundling, and non-discrimination in the connection of users. These rules facilitate the emergence of competition and the operation of the market in other activities related to energy such as the production, storage, or retail-selling of energy. ⁶⁰ The non-discrimination obligation also limits the possibility for grid managers to place their own interests over societal interests in their decision-making, and ensures that they generally do not have to make political choices. ⁶¹ The main objective of this regulatory system is to introduce market mechanisms in the energy sector which are intended to result in efficiencies such as cost efficiencies or robustness of the energy sector. However, one may wonder if with the energy transition, the role of grid managers is also changing. The energy transition is more than a transition to a new technology, but rather a complex socio-technical transition that impacts the role and functioning of public and private bodies as well as consumers in the production and use of energy. ⁶² Grid managers are – especially in crisis situations such as in the Netherlands – required to coordinate the production of energy on one side and the use of energy on the other. This may require a reconceptualization of the regulatory framework that governs them.

This shift towards the grid manager as coordinator is already happening gradually in the Netherlands and at the EU level. One example from the Netherlands is the ACM’s announcement that grid managers are allowed to engage in societal prioritization. Following article 6 (1) of Directive 2019/944, grid managers must facilitate access to all third parties that require access to the grid without discrimination. Article 6 (2) of Directive 2019/944 provides an exemption that allows grid managers to refuse new requests for connections if they lack the necessary capacity. Such a refusal must however be substantiated on the basis of technical or economically justifiable criteria, and the grid manager must – where possible – provide information to the requestee about when the connection could be realized. In the Netherlands, where congestion issues are prevalent, the ACM has provided a policy for a structural exemption in the form of ‘societal prioritization’, where grid managers are encouraged to actively use the exemption to prioritize some connections over others. This is a significant shift from the ‘first come, first served’ principle which would normally govern such rejections, allowing grid managers to not only refuse a connection but also create a waiting list based on the societal impact of new connections. ⁶³

There are also several provisions at the EU level that grant more discretion to grid managers to vertically integrate or to influence national energy policies. Article 36 of Directive 2019/944 contains the unbundling obligations for grid managers. While grid managers are in principle not allowed to vertically integrate into storage facilities, they are provided with the possibility to open tendering procedures for energy storage facilities. Thus, grid managers are not only responsible to connect new third-party operators, but may also act pro-actively to request storage facilities in certain locations and for specific capacities. Important for this article, and discussed in-depth hereunder, is the exemption to vertical integration. If the tender fails to attract interested market parties, the grid manager may integrate the storage facility and operate it for a certain period of time before re-evaluation. The ability to open tendering procedures under the supervision of national regulatory authorities highlights the coordinating role of grid managers and divergence from the idea of the grid manager as a neutral market facilitator.

Under proposed changes to Directive 2019/944 and Regulation 2019/943, grid managers will gain further abilities to influence and shape the emergence of energy storage facilities. The proposal introduces article 19c of Regulation 2019/943, which requires Member States to develop a national energy strategy and to assess the need for flexibility capacity on a yearly basis. This assessment happens on the basis of data and analyses provided by grid managers. The European Network of Transmission System Operators for Electricity (ENTSO-E) and the European Distribution Systems Operators Entity (EU DSO) are to (a) define the type of data and format that transmission and distribution system operators shall provide to the regulatory authorities; and (b) develop a methodology for the analysis by transmission and distribution system operators of the flexibility needs, taking into account at least all existing sources of flexibility and planned investments at interconnection, transmission and distribution level as well as the need to decarbonise the electricity system. Thus, determining the types of data and methodology happens by the representative organizations of grid managers rather than the European Union Agency for the Cooperation of Energy Regulators (ACER) or National Regulatory Authorities (NRAs). ⁶⁴ At the national level, grid managers themselves conduct the analyses based on this data. This provides them with significant capabilities to directly and indirectly influence the policy regarding energy storage.

This trend in legislative recognition of the grid managers as having a function of coordinator should be recognized in both EU and national law. Firstly, it may be the case that this shift is necessary to guide an ever decentralizing and complex energy system. The production and distribution of energy no longer happens in a centralised manner, where a small group of energy producers sells their energy and consumers use it. Instead, new types of producers and active consumers are gaining an important role in producing energy and may require coordination through the grid managers. ⁶⁵ Secondly, recognizing this trend allows the regulator to better steer this shift in how grid managers operate. By viewing grid managers as neutral facilitators – rather than coordinators – a mismatch may come to exist between the perception of what grid managers are supposed and allowed to do and what is expected of them in practice. Thirdly, and relevant for this article, viewing the grid managers as transitioning from a passive to an active role in governing the energy system may help to frame experimentation policies that explore what roles are desirable for grid managers in future legislation and study the benefits and drawbacks of certain legislative changes. Recognizing the idea that a shift in the functions fulfilled by grid managers is changing with the energy transition may require the development of new principles that underlie the regulatory framework that governs them. The next section discusses how allowing more space for grid managers to divert from unbundling obligations may help to resolve congestion issues, under which conditions it is attractive, which problems it may resolve, and how possible risks can be identified.

Utilizing the unbundling exemption – vertical integration and shared ownership of storage facilities

Allowing grid managers to vertically integrate storage capacity (temporarily) may help to resolve issues related to grid congestion. Directive 2019/944 has introduced an exemption in its unbundling obligations to allow for such constructions legally. The exemption laid down in article 36 (2) of Directive 2019/944 allows grid managers to vertically integrate and operate storage facilities if certain conditions are met. The exemption can be used if several conditions are fulfilled. First, the grid manager must publish a tender and must follow an open, transparent, and non-discriminatory tendering procedure that is subjected to review and approval of the regulatory authority. Second, a third party has not been awarded the right to own, develop, manage, or operate such facilities, or could not deliver those services at a reasonable cost and in a timely manner following the conditions set out in the tender. Third, the facilities are necessary for the distribution system operators to fulfil their obligations under this Directive for the efficient, reliable, and secure operation of the distribution system and the facilities are not used to buy or sell electricity in the electricity markets. Fourth, the regulatory authority has assessed the necessity of such a derogation and has carried out an assessment of the tendering procedure, including the conditions of the tendering procedure, and has granted its approval. Fifth, the NRA must periodically (at least once every five years) conduct a public consultation to assess whether the market is interested in taking over these activities from the grid manager. ⁶⁶

In practice, the assessment by the NRA in approving or rejecting the DSO’s request to use the exemption consists of several steps. ⁶⁷ First, a tender must be designed and offered to the market. Here, the DSO can request the required storage service and capacity and potentially set criteria on what is considers to be a reasonable (maximum) price for this service. Thus, the DSO has the first responsibility for setting the conditions of rewarding the tender. Second, the NRA conducts an ex-ante review of the tender as to determine whether it is reasonable. For this, the NRA may develop guidelines so that the DSO knows what to expect. This review process ensures that the tender is fair and non-discriminatory. Third, the outcome of the tendering procedure is determined. This may result in three outcomes: a) the storage facility if fully operated by a third-party market participant; b) the exemption is used so that the DSO can operate the storage facility; or c) there is shared ownership between the DSO and the third-party market participant. ⁶⁸ While full third-party ownership is the main rule, the exemption can be applied if there are no bids or if there are no bids that match the conditions set by the DSO. If there are no satisfactory bids, the DSO can apply with the NRA to make use of the exemption and the NRA will then conduct an ex-post assessment. If it is decided that the tender has been offered properly by the DSO, they can be granted the exemption. Finally, the NRA will conduct the periodical public consultation to see if the DSO can retain ownership rights. ⁶⁹

Shared ownership between DSOs and third parties has been discussed in the (non-binding) ACER Framework Guidelines on Demand Response. ⁷⁰ This option may be more attractive than full DSO ownership and may be possible if no third-party wants to carry the full responsibility for operating the storage service. When there is shared ownership, the grid manager is subjected to the same rules as it were to be for full DSO ownership while the third-party operator can operate their part of the storage freely. ⁷¹ By sharing the risk, the market participant does not have to take the full financial risk of operating the storage facility while the DSO can realize the desired storage facility while being aided by the market. Moreover, as this exemption is apparently considered a ‘second best’ solution by ACER, it may be easier to justify these types of constructions than it is to justify full DSO ownership. ⁷² In section 3.3 it is discussed in-depth why relying on such a construction may also be desirable from the perspective of grid managers, market players and society as a whole. First however, it is discussed hereunder how Member States can set conditions for the use of this exemption and how it would apply in the Netherlands following the proposed Energy Act.

The Council of European Energy Regulators (CEER) discusses the margin of discretion held by Member States in implementing a strict or lenient policy towards exemptions. Here, it looks at Croatia, Germany, and the United Kingdom. This comparison shows that there is significant divergence in the policies and considerations by NRAs in granting the exemption. In Croatia, the DSO must first perform an analysis to specify the need for reliability and security of supply. The DSO then submits the proposal for the tendering procedure to the NRA with the description of the product, technical requirements and the valuation of services etcetera. The NRA conducts the ex-ante review and opens the tender. If there are no bids, the DSO must submit a cost benefit analysis where it shows that integrating electricity storage is more cost effective than grid reinforcements. The NRA may then approve the request. ⁷³

In Germany, the NRA shall grant approval if the DSO has demonstrated that the energy storage facility is necessary to meet its legal obligations as a DSO in an efficient manner. The storage facility may only be used for the intended purpose and not to buy or sell power, even if it is operated by the third party. The DSO will open a tender procedure under supervision of the NRA, and if they are unable to award the bid to a third party or if the third party turns out to not be able to fulfil the service it can be integrated. Furthermore, it is stated that the DSO cannot award the contract to a third party if they offer this service is not available at a reasonable cost, where the costs are deemed reasonable if they do not exceed the costs for the construction, management and operation of a comparable energy storage system owned by a network operator. ⁷⁴ This is significantly less strict than the Croatian implementation: instead of having to determine whether electricity storage is more attractive than grid reinforcement, the DSO simply has to prove that it is more cost efficient to integrate the electricity storage than it is to procure it from a third party.

The United Kingdom already formulated its exemption policy while still (technically) bound by EU law. Here, they have chosen to introduce a condition in the electricity distribution license that DSOs cannot own and operate generation stations, including storage. However, a licensee may own or operate generation assets through three exemption grounds: category A, Island-based networks, category B, generation for specific authorised activities, or category C: generation pursuant to a direction by the Authority’. Category A is very situation specific and category B is similar to the FINC exemption. The exemption discussed previously mostly fits in category C. To use a category C exemption, the DSO must have taken reasonable steps to obtain a market-based solution; it must be justified that a licensee-operated asset provides the most economic and efficient solution; and arrangements are put in place that minimise the risk of discrimination or distortion of current and future markets. ⁷⁵ This again leaves a lot of space to rely on the exemption. The United Kingdom has also ensured that it can take appropriate measures to prevent discrimination or distortion of future markets, allowing the NRA to impose remedies so that there is a balance between the objectives set for electricity storage on the one hand and free competition on the other. Ofgem provides virtually no guidance on what those measures could or should be in their Guidance document. ⁷⁶ Tying the use of the exemption to a license does however introduce the possibility for regulators to increase their oversight by allowing them to develop terms for issuing licenses and retractions of the license in cases of non-compliance.

The question that then remains is how the policy regarding exemptions can be formulated in the Netherlands, which is the case study for this article and the Energy Act has not yet been adopted. ⁷⁷ The prohibition for DSOs and TSOs to integrate electricity storage facilities is laid down in art. 3.31 of the Energy Act. Art. 3.32 offers the FINC exemption and art. 3.33 Energy Act codifies the exemption by approval of the NRA. According to art. 3.33 Energy Act, the DSO can be exempted if: a) integration is necessary to meet its legal obligations under art. 3.25 Energy Act; b) The DSO or TSO does not use the storage facility to trade energy; c) the grid manager has demonstrated that market participants are not able to realise – or realise cost efficiently – the facility in a timely manner.

The Energy Act grants the ACM the powers to develop guidelines for a proper tendering procedure, to set conditions for the use of the exemption, but it has to notify the European Commission and ACER when it grants exemptions. It also has to organize the mandatory public consultation at least once every five years. The ACM can also remove the exemption if it finds that market participants have become able to cost effectively develop, exploit or manage the electricity storage facility. Thus far, the ACM has not developed a strategy specifically for the exemptions related to electricity storage under art. 3.32 (2) Energy Act.

The wording of the text in the Energy Act does however seem to be formulated in a relatively permissive manner. Firstly, the condition that the storage facility is necessary to meet their legal obligations under art. 3.25 may be interpreted broadly. Art. 3.25 (1) states that a TSO or DSO must have a system that is able to – in the short and long term – meets the demand for transport of electricity (and gas) and maintains a system that is safe, reliable and efficient and that takes into account the environment, digitalisation, energy efficiency and the transition to a carbon-free society. This allows the grid manager to base their argumentation on more grounds than for instance reliability of security of supply, as is the case in Croatia. ⁷⁸ Instead, it can take into account a host of considerations from the availability to meet demand to sustainability. The grid manager must also demonstrate that market participants cannot realise the facility, realise it for reasonable costs, or within a timely manner. It can be ambiguous what realising the facility means in this context, but it can be assumed that this includes situations where the market participant cannot operate the facility cost efficiently or where it does not serve the intended purpose behind the tender.

The combination of factors in the Dutch energy sector could grant significant room for DSOs to utilize the exemption. First, there is an obligation for DSOs to meet the short-term and long-term demand as long as it is reasonable under art. 3.25 of the proposed Energy Act. Currently, the energy grid in the Netherlands is sufficiently congested that several DSOs are not able to meet short-term demand, while other regions are at risk. Moreover, long term supply is endangered because of a shortage of professionals that can fortify the grid and because many storage facilities cannot be built as they may contribute to congestion issues. ⁷⁹ This relates to the second factor, if it is not desirable that energy facilities are used to trade in profitable markets as this may increase congestion issues, they may not be able to operate cost effectively. ⁸⁰ Consequently, the DSO can open tender procedures for electricity storage facilities that do not trade on these markets regularly, which will be more likely to attract few or no bids if the operation is not profitable (or if the time for returns to investment is too long). Thirdly, and considering these first two factors, it is likely possible to argue that the integration is necessary from the perspective of other goals laid down in art. 3.25 of the Energy Act such as affordability and sustainability.

This analysis shows that utilizing the exemption for the temporary integration of storage facilities is likely a legal possibility. However, this does still not necessarily mean that it is attractive from the perspective of cost efficiency or definitively serves societal interests. The upcoming section discusses why shared ownership between grid managers and third parties may be one method to stimulate (private) investments, provide security for grid managers in (co-)financing the initial investment and how it may limit the reliance on public funds and subsidies.

Identifying desirable conditions – what are the benefits and drawbacks of exemption and shared ownership?

Unbundling obligations are the result of restructuring policies aimed at the liberalisation of the energy sector. Prior to these restructuring policies introduced in the 1980s and 90s, different stages of the production of electricity - generation, transmission, distribution and supply – were undertaken by one vertically integrated entity. ⁸¹ The idea behind these restructuring policies was that introducing competition would lead to a more efficient energy sector, as market mechanisms would be better suited to drive down costs than economic regulation. Distribution- and transmission activities, which required the operation of large and expensive infrastructures, was considered a natural monopoly. As such, it was considered that introducing competition in the markets for these activities would like lead to cost-inefficient outcomes. ⁸² As such, obligations were introduced to separate transmission- and distribution activities from all other activities such as the generation, supply, and now storage of energy. ⁸³ Unbundling remains a central rule in the energy sector, and there are no exceptions for the production or supply of energy. With the exemption for the operation of storage facilities, the regulatory is allowing limited vertical integration. That the rule remains is related to concerns that if there is a threat of having to compete with a vertically integrated entity that controls the infrastructure, third parties become disincentivised to enter the market due to the threat of exclusion by this entity. This is supported by empirical observations from the 1990s and early 2000s, where it was shown that there was limited entry following the choice for account unbundling, and entry became more prevalent once the operation of distribution and transmissions infrastructures become subjected to legal or ownership unbundling. ⁸⁴

While unbundling became the main rule in the liberalisation of EU energy markets, there was significant debate on the desirability of unbundling at the time of implementation. Proponents of vertical integration argued that vertical integration had various coordination benefits, and that unbundling could lead to increased transaction and information costs. It was also argued that the separation of generation and transmission could lead to a structure based on bilateral monopoly and cause ‘hold up’ problems that would act as disincentives for investments in generation and transmission. ⁸⁵ The innovation of introducing non-discrimination obligations on an unbundled and thereby independent ‘systems operator’ helps to resolve short-term coordination problems related to bilateral monopolies. However, as noted by Decker, the longer-term issue of efficient coordination of generation and transmission decisions remains a cause for concern. Especially in situations where there is the need for substantial investments in new generation (including storage) capacities and in light of shifts in the profile and location of generation facilities particularly a need for transmission to connect to often remotely located renewable generation facilities. ⁸⁶ Under these circumstances, Decker notes that there has been debate in some restructured markets as to whether there is a need for some form of centralised transmission planner, who can co-optimise decisions relating to generation and transmission, including taking a 'strategic' approach to the development of the transmission network. ⁸⁷ Decker refers to the obligation in the Third Energy Package where transmissions operators must develop a ten year investment plan. From a broader perspective however, it alludes to the idea that under these circumstances, grid managers may have to fulfil a role of coordinator rather than merely an independent system operator. ⁸⁸ Moreover, it highlights the risk of disincentives for new investments if these coordination problems are not resolved.

The stalemate described in previous sections is largely the result of coordination problems and in turn a lack of (perceived) possibilities to co-optimise decisions. This article argues that shared ownership may help to resolve these issues while striking a balance between the interest of grid managers and third-party market players, as well as the objectives of competition and a secure and reliable energy system. This section will theorize on why the benefits of shared ownership are likely to outweigh maintaining strict unbundling rules and under which conditions. The final section of this paper discusses how these theories can be tested in practice through experimentation.

First, this article argues that the costs of certain solutions must be viewed from the perspective of their broader societal costs. This means that in calculating whether integrating energy storage is attractive from a cost per prevented ton of CO2 emissions, one must not only look at the losses incurred by building and operating energy storage facilities. In situations with heavy grid congestion, there are costs incurred by multiple parties. Grid managers face costs related to curtailment and the mandatory purchasing of renewable energy. To exemplify, in Germany alone, grid operators have paid up to €761 million in 2020 to remunerate producers of renewable energy for curtailment measures. ⁸⁹ In the Netherlands, there have been 192 hours of negative energy prices by July 2020. In some instances, prices were minus €500 per megawatt hour. ⁹⁰ When looking at the estimated costs of €2.500 per prevented ton of CO2 emissions as estimated by CE Delft (2023) for the installation of energy storage at solar parks, the costs of not developing energy storage facilities must be taken into account when looking at the issues from a societal perspective. If energy storage facilities are purely operated by independent firms, these costs should not be taken into account. After all, the gain of the grid manager does not imply any gains for the third-party operator. The third-party operator would fully incur the losses associated with operating the cost-inefficient storage facilities. If energy storage facilities are subsidized through public funds, the costs are deferred to citizens and society. Thus raising societal costs without creating incentives for grid managers or third-party operators to act efficiently. Shared ownership of storage facilities may help to realign incentives and co-optimize decision making.

Shared ownership of a battery system allows for co-investment of private funds into energy storage systems. The shared investment works as a type of insurance for the investing parties, which mitigates the uncertainty associated with investments currently. In a scenario where there is heavy grid congestion, the grid manager may open a tender for the procurement of storage facilities. Using the storage facility to trade in profitable energy markets, such as the energy imbalance markets, could lead to additional grid congestion, but the storage facility is not profitable if one cannot trade in this market. The tender therefore prescribes that the storage facility cannot be used for buying or selling energy and can only be used for managing grid congestion. If there are no interested third parties, the grid manager may choose to integrate the storage facility temporarily. After five years, the national regulatory authority will evaluate whether the integration of the storage facility is still necessary and may find that it is not. This raises an issue of sunk costs for the grid manager: if the return-of-investment of the storage facility is not under five years, it may have to sell the battery to an interested third party following a legal obligation. This harms the position from which the grid manager negotiates the sale, as the buying party is aware that the grid manager is forced to sell. As such, grid managers face uncertainty and the risk of significant losses if they pursue full ownership of the storage facility. The losses incurred by the grid manager raise network costs, ultimately impacting consumers through price raises.

Shared ownership allows for co-optimization. This article puts forward that, for instance, the grid manager can agree with a third party that they share the ownership of the battery based on percentages, and that the grid manager reimburses the third-party operator for their share of the storage facility (and potentially vice versa). With such a construction, the grid manager could own and operate (for example) 90% of the storage facility, while the third-party owns and operates 10%. The third party may use 10% of the capacity of the storage facility to trade in energy markets (perhaps with contractual limitations) whilst the grid manager pays the third party a compensation for the operation of 90% of the storage facility used for congestion issues. If congestion issues are resolved over time, the parties can agree on further sales of the owned percentage from the grid manager to the third party for a (predetermined) price. Thus, over the period of five years, ownership may already shift from 90-10% to 50-50%, and ultimately to complete divestment by the grid manager.

The grid manager will gain access to the battery capacity it requires for congestion management with less initial investments than when it purchases the storage capacity by itself, at the same time the third-party operator can start trading with its share of the battery. This type of construction realigns the incentives of both parties and thereby optimises investment decisions: neither of the parties incurs full risks and investment costs, there is certainty for the grid manager with respect to the divestment and sale of the storage facility, and there are securities in place for the co-investing third party. One benefit of such a construction is that the grid manager has incentives to sell its share of the operated battery to the third party. As it pays a periodical fee to the third-party for the share it operates and receives a pre-determined compensation for the share that it sells, it stands to gain by selling shares of its battery relatively quickly.

A simple calculation may exemplify these incentives. In this example, there is a battery worth €1.000.000 and the grid manager and third-party both invest €500.000 for 50% of the share of the battery. They then agree on a reimbursement (from grid operator to third party) and a pre-determined sales price (paid by third party to the grid operator upon sales of their share of the battery). In this scenario, the grid operator pays the third party €500 per 10% share each month, totaling €30.000 per year for its 50% share. For each 10% stake it sells, it receives €100.000. Hereunder three hypothetical scenarios to exemplify different cost distributions:

Scenario A - The grid manager operates its 50% of the battery for 10 years without selling any share, paying €300.000 to the third party to compensate it for using its share of the battery. The total costs for the grid manager are €800.000 allowing it to manage congestion for ten years. For the third party, the total costs are €200.000 (€500.000 investment minus €300.000 reimbursements). It can trade in profitable markets with 50% of the battery for 10 years.

Scenario B - The grid manager sells 10% of its battery on (the hypothetical) day 0, making its share of the battery 40%. The grid manager pays €240.000 over ten years, while receiving €100.000 for its 10% share, meaning that the total investment for the grid manager €640.000 after 10 years, including its initial €500.000 investment. The third party invests €600.000, but gains the €240.000 in reimbursements, making its total costs €360.000 to operate 60% of the battery to trade in profitable markets for 10 years.

Scenario C - The grid manager sells its complete 50% stake after 5 years, paying €150.000 in reimbursements. Its initial investment costs including reimbursements are €650.000, while it receives €500.000 from the sales of its shares to the third party. The total costs for the grid manager are €150.000 to connect the storage facility and manage congestion over a five-year time span. The third party initially invests €500.000 and pays €500.000 for the rest of the shares. The third party also receives €150.000 in reimbursements, making its total costs €850.000. For this investment, the third party gains the rights to trade in profitable markets with 50% of the battery for 5 years, and with 100% of the battery thereafter.

These scenarios indicate that both parties would benefit from the arrangement. In each scenario, each party receives their desired outcome, while neither has to carry the full €1.000.000 investment costs. If the grid manager sells its entire stake on day 0, it would break even, meaning that the third party pays the full price of the battery. If the grid manager never sells, it pays the monthly reimbursement indefinitely, ultimately raising the costs beyond €1.000.000. The NRA can also intervene and force the sale (possibly for a lowered price if the grid manager has acted in bad faith).

In each of the described scenarios, the costs for the grid manager are lower if they sell their stake faster, while the third party gains the possibility to enter into profitable trade sooner. The longer the grid manager uses its share of the battery to manage congestion, the higher the costs for the grid manager (and the longer the third party is delayed in using the full battery to trade in profitable markets). However, there are benefits for the parties from such an arrangement. The grid manager prevents additional costs of curtailment and fortifications of the grid, while increasing its flexible capacity. The third party receives more reimbursements while still using its share of the battery to trade, likely (depending on the conditions of the agreement) making its investment profitable. The parties can freely negotiate the conditions of their agreement under the supervision of the National Regulatory Authority. It is up to the NRA to determine whether the agreement is necessary, and if the conditions it has set for the use of the exemption are met.

The benefits are not limited to the parties involved in the agreement; society benefits from this approach as any losses that may result from the construction are not shifted towards the public. Liabilities, such as the degradation or malfunctioning of the storage facility, can be arranged contractually between the parties. While the practical execution of such arrangements may find difficulties, such constructions can be worth experimenting with as to investigate the effect on investment incentives and congestions issues. The upcoming section discusses how the ACM can govern such shared ownership agreements, and how experimentation can be used to gain experience with these shared ownership constructions and to engage in regulatory learning.

Before moving on to the next section, it should be noted that this article focuses primarily on shared ownership between operators of stand-alone batteries and grid managers. However, similar considerations may apply for ‘behind the meter’ batteries that are integrated with large users. Here, one could balance between the costs of the firm incurred when they cannot open or expand their facility on the one hand, and the benefit that additional storage facilities may provide the grid manager on the other. To exemplify: co-investments between the grid manager and (a group of) firm(s) for the installation of a battery with more capacity than needed for the requestees may allow the grid manager to utilize parts of the battery for congestion management and secure the creation of surplus storage capacity for the creation and expansion of energy hubs in the industrial area. These types of constructions are not the focus of this article, but their desirability can be investigated further.