This article was prepared with assistance from Iustinian Captariu and Catalin Graure of Kinstellar.

Romanian authorities unceremoniously ended the partnership with the previously selected state-owned Chinese company, China General Nuclear Power Corporation, to develop Units 3 and 4 of the Cernavodă Nuclear Power Plant ("C 3&4"). The decision was formalised on 12 June 2020 at the shareholders' general meeting of Romania's state-owned nuclear company.

This development comes at a time of important amendments to Romania's energy legislative framework. After a prolonged hiatus, there is momentum for long term bilateral power purchase agreements, as electricity producers can now conclude such agreements for electricity generated by production capacities commissioned after 1 June 2020. This may assist with the financing potential of new capacities, provided other obstacles are also overcome.

Further, Romania is contemplating a Contracts for Difference support scheme meant to encourage new investments in power generation capacities. The scheme is intended to set the stage for a wide range of new investments, including nuclear, but raises concerns in the latter context.

In this article, we review the history and status of C 3&4, and discuss the nuclear project financing challenges that, although not unique to the Romanian context, have proven insurmountable to date.

Background on the power sector in Romania

Romania is no stranger to the desire to achieve energy independence, aligning itself with policy objectives of the European Union; a desire intensified by Romania's geographical location and geopolitical considerations. For this and other reasons, the energy sector is more important than ever and is viewed by Romanian authorities as a catalyst for economic recovery in the aftermath of the COVID-19 pandemic. To this end, the Romanian government is taking measures aimed at attracting investment in new electricity generation capacity.

Romania benefits from a balanced electricity mix, combining traditional fuels (oil, natural gas, coal and nuclear) with renewable energy sources (mainly hydro, wind and solar). According to official data as of 28 May 2020,[1] Romania has an installed power generation capacity of over 20,000 MW (mostly dispatchable) with hydro (6,693 MW, 32.4% of total), coal (4,787 MW, 23.2% of total), hydrocarbons (3,210 MW, 15.5% of total), wind (3,023 MW, 14.6% of total), nuclear (1,413 MW, 6.8% of total) and solar (1,398 MW, 6.8% of total) accounting for more than 99% of the generation capacity.

The largest power producers in Romania are state-owned, generally not diversified in terms of production sources. They include juggernaut hydropower producer Hidroelectrica SA, nuclear power producer Societatea Națională Nuclearelectrica SA ("SNN")[2] and coal power producer CE Oltenia SA. Private investment generally has been directed toward power generation capacity from renewable energy sources (mainly wind and solar during the 2010–2016 period, as a result of a green certificates support scheme implemented by the state). This stream eventually dried up as the state drastically cut – what some have called an over-generous support scheme – which pushed power bills higher for end consumers, especially businesses in electricity intensive sectors.

There are significant opportunities for investment in the sector, as current infrastructure is a relic: the 20,000 MW installed capacity includes multiple coal-fired power plants that generate very expensive electricity because of environmental compliance costs, as well as other capacities that have either exceeded or are due to reach end-of-life in the near future.

As the days of coal seem to be numbered, nuclear is a logical choice going forward. Nonetheless, capital-intensive investments are also necessary to modernise the power transmission and distribution systems and enhance the flexibility and interconnection capacity of the national electricity grid.

Background on the Cernavodă Nuclear Power Plant

Romania's nuclear program was launched in 1968 during the communist era in what was to become the Cernavodă Nuclear Power Plant. In 1978, the Romanian and Canadian Governments agreed to cooperate in the field of peaceful uses of nuclear energy. A joint team prepared a feasibility study that supported the implementation in Romania of Atomic Energy of Canada Limited's ("AECL") CANDU 6 nuclear reactor design. Subsequently, ROMENERGO, the Romanian company dedicated to international trade in the power sector, concluded a CANDU 6 intellectual property license agreement with AECL as well as other contracts for the provision of engineering and technical services, equipment and materials procurement for the nuclear island for Cernavodă Unit 1 ("C 1"). In 1981, the contract for the conventional part was signed with ANSALDO. On April 16, 1996, C 1 achieved first criticality and began supplying power to the grid on December 2, 1996, operating at a nominal power rating of 706 MW(e).[3]

Although designed to incorporate five units, the ordeals of a slow transition to a market economy means that today only C 1 (commissioned in 1996)[4] and Cernavodă Unit 2 ("C 2") (commissioned in 2007) are generating power. Currently operated by SNN, the two units — each with an installed capacity of about 700 MW — account for roughly 20% of the country's electricity production, although representing a smaller percentage of installed capacity. Work on the other units ceased in 1992, leaving C 3&4 at around 15% completion and Cernavodă Unit 5 ("C 5") at below 3%.

In 2007, Romania decided to move forward with the construction of C 3&4 and managed, by 2009, to attract a series of investors (i.e. ArcelorMittal, CEZ, GDF SUEZ, ENEL, Iberdrola and RWE Power). A project company, EnergoNuclear was established with the Romanian state through SNN, in which all of the investors were shareholders. However, in the ensuing years all foreign investors eventually withdrew from the project.

Problematic nuclear project finance models

The trend in the West over the past two decades has been to attempt to attract private financing and project risk absorption for nuclear new build projects, in an effort to distance governments from nuclear project risk. However, in Romania and elsewhere it has proven virtually impossible to successfully build privately financed nuclear power plants without government support. The nuclear project risk profile involves elements of systemic risk that are beyond most proponents' control and risk tolerance.

In one form or another, successful nuclear programs have engaged public institutional backing for project risk. Historically, in the build-out phase of major Western programs, risk absorption and financial backstops have been provided through the rate base in regulated markets. Other major programs have achieved the same effect by direct government backing for risk and finance.

The recent Western trend has entailed the search for substitutes for the approach used to successfully build out the fleets in the first instance.

Those substitutes have included the attempted use of the Makala model in Finland — a model that is based on the premise that nuclear project risk and price volatility can be effectively addressed through joint ownership of the projects by energy-intensive companies who would buy power at cost price to shareholders. While Makala can offer a viable approach to market price risk stabilization, it has not proven to be a model fit to anchor nuclear project risk during the construction phase. In fact, as we have seen in the Olkiluoto 3 and Hanhikivi projects, it becomes necessary for project risk to be anchored by governments other than that of the host nation; namely, the vendor nations' governments.

EnergoNuclear initially sought to adopt a structure similar in principle to Mankala but with some ownership participation by the Government of Romania through SNN. It is perhaps not surprising that, in the absence of a vendor nation anchoring the risk (the Canadian government having stepped off nuclear project risk at the time), the EnergoNuclear joint venture could not be sustained.

The Contract for Difference ("CfD") is another substitute model that has been under discussion since the time of the collapse of the joint venture. Under this model, private capital is expected to invest in the development of the nuclear assets and to recoup its investment through a mechanism that stabilizes long-term electricity prices to provide certainty of returns. One of the early uses of the CfD in nuclear projects was the British Energy takeover of the Bruce Nuclear Generating Station in Canada, through Bruce Power. In effect, the investor group would fund the refurbishment and operating cost of the nuclear reactors, and enjoy a stable model of return through a contractually regulated pricing mechanism. Refurbishment does not carry the same risk profile as new build projects (whether in probability or magnitude) and for that reason stabilization of market risk facilitated the success of the Bruce Power CfD model.

By contrast, the UK has struggled with its attempts to adapt the CfD mechanism, together with the use of government debt supports, to new build nuclear project risk. Put simply, the UK approach to Hinkley Point C was to use a contractual and regulatory market price risk adjustment mechanism (the CfD) as a surrogate for nuclear project construction uncertainty, while using government guarantees to reduce the cost of capital (driven by project risk and traditionally regarded as one the greatest challenges for nuclear projects). This indirect engagement of public risk absorption capacity has been the subject of significant criticism. On the one hand, anchoring project risk uncertainty through market price risk adjustments, necessarily puts upward pressure on pricing. On the other hand, indirectly addressing the issue of nuclear project risk, left the European Commission ("EC") musing about what, if any, market failure the UK Government was looking to address through the CfDs and credit enhancements. Ultimately, the EC required the imposition of market pricing on the government guarantees, effectively negating their usefulness to reduce the cost of capital.[5]

The central 'market failure' for nuclear new build projects is the means of addressing risk absorption and (related) cost of capital. Nuclear project risk is multi-dimensional (involving several types of risk including political, regulatory, technology, supply chain, construction, and market risks among others). Particularly in Western nations, where political and regulatory risks have had notable impacts, there is a very wide margin of uncertainty around project cost and schedule, making both the cost of capital and availability of private investment challenges that are very difficult to resolve without government backing. There is no known instance of these problems being addressed successfully without some manner of government backing.

Enter China General Nuclear Power Corporation

Confronted with the investors' withdrawal from the project, in 2014 SNN adopted a strategy to continue construction of C 3&4 through the selection of a private investor, while any prospect of commissioning C 5 was abandoned. China General Nuclear Power Corporation ("CGN") was the selected investor, and the parties signed a joint letter of intent in October 2014 to accomplish the project. The following year, in November 2015, SNN and CGN signed a memorandum of understanding for the construction of C 3&4 and negotiations began on the terms of the Investors' Agreement and articles of association of the new project company. Agreement on these two documents by the parties was a prerequisite before taking a final investment decision.

In October 2016, SNN approved continuing negotiations with CGN until 20 December 2016, but the parties did not manage to find common ground. As the project had reached a standstill, in July 2017 the Romanian government adopted a memorandum on continuing negotiations with CGN that included a revised mandate for the Romanian commission handling the negotiations.

In spring 2019, the parties signed the preliminary form of the Investors' Agreement for C 3&4. Thereafter, discussions dragged along with no real progress having been made with the investment plans.

Now Romanian authorities have announced that the development plans of C 3&4 with CGN would be dropped. During SNN's general meeting of shareholders held on 12 June 2020, the Romanian Ministry of Economy (as majority shareholder of SSN) together with other minority shareholders passed a motion to cancel the previous strategies on continuing C 3&4 and mandated SNN's Board to terminate negotiations and previous agreements concluded with CGN. SNN's Board was also entrusted with analysing strategic options for the development of new electricity production capacities from nuclear sources.

High-ranking officials commenting on the matter have declared that the investment project would be pursued with strategic partners from NATO and the EU.

The Sovereign Guarantee Problem

We are not aware of the reasons why CGN and the Romanian Government did not reach a successful solution.

One of the challenges to new-build nuclear power projects has been the availability of financing from preferential Export Credit Agencies ("ECA"). The Organisation for Economic Co-operation and Development ("OECD") Arrangement on Officially Supported Export Credits[6] effectively put an end to member vendor governments providing financing for the majority of the project cost on preferential terms. For instance, OECD member ECA financing can be provided for up to 85% of the vendor nation's project scope and about 30% of multi-national scope. It must be provided on market driven terms and at rates that reflect the lending risk, which also means that the absence of a sovereign guarantee drives rates to debilitating levels.

At the same time, for decades, small and mid-sized economies have been under considerable pressure from international institutions (like the IMF) to maintain debt to GDP ratios within prescribed levels and to work within specified debt and sovereign guarantee limits for infrastructure projects. The vast cost of nuclear power projects rarely, if ever, fits comfortably within such constraints. The result has been a number of nations—Romania, Turkey, and Argentina, among others—seeking to proceed with nuclear power projects without offering a sovereign guarantee for the capital investment. This again puts considerable emphasis on the need for a high degree of assurance for either the vendor or potential investors. The absence of such assurance likely contributes to the inability to fund projects.

Non-OECD members (e.g. China or Russia) remain free to provide ECA financing for nuclear projects on favourable terms, including extended grace periods, long tenors, and subsidized interest rates. Such terms have a demonstrated track record of success in international nuclear projects. However, such arrangements historically have required sovereign debt guarantees for the ECA financing. In the absence of sovereign debt guarantees, the funding mechanisms need to provide government-backed pathways to mitigate the uncertainties of project risk and cost of capital; sometimes through some combination of Build-Own-Operate ("BOO") models or similar structures, long-term market pricing mechanisms, and government backed PPAs. Such structures may now be more feasible in Romania because of the recent legislative amendments permitting long term PPAs, but are considerably more complex and raise greater uncertainties than traditional ECA debt financing against sovereign guarantees. We have not yet seen a proven successful model capable of replication, although Rosatom's Akkuyu Project in Turkey may pave the way.

It would not be surprising if the constraints on government guarantees had some role in bringing about the end of CGN's involvement in C 3&4. It is noteworthy that Romania sought and obtained indulgence from the IMF to enable the completion of C 2. Given that government spending is widely seen (including by the IMF) as a significant contributor to future economic recovery from the COVID-19 crises (G20 having spent 9 trillion dollars on economic supports through the end of May 2020), it may be that governments may find it more feasible to now give the necessary guarantees to anchor nuclear project risk.

C 3&4 – Quo Vadis?

The most relevant framework currently governing Romania's energy development is the draft 2019–2030 Energy Strategy ("Energy Strategy") and the 2021–2030 National Energy and Climate Plan ("NECP"). The latter was developed according to EU rules and is meant to ensure that Romania meets the EU's energy and climate targets for 2030.

Both the Energy Strategy and the NECP identify the construction of new nuclear power generation capacities as strategically important for Romania's energy security. More specifically, the Energy Strategy stipulates that the project for C 3&4 is an optimal solution to cover the generation capacity deficit foreseen for the 2028–2035 period as a result of multiple existing capacities reaching the end of their operational life. Moreover, the project would use to a great extent the existing infrastructure and important heavy water reserves currently available.

According to both documents, C 3 is expected to become operational in 2030 and C 4 by 2035. According to the Energy Strategy, the refurbishment process of C 1 could justify postponing the decommissioning of certain coal and natural gas generation capacities.[7]

In terms of financing for C 3&4, the NECP acknowledges the possibility of establishing a CfD-type mechanism. According to the NECP, a CfD mechanism would support new investments in renewables but also in other power sources (e.g. nuclear) and in electricity storage projects in order to increase the mix in power sources and the flexibility of the national grid. In this regard, the Romanian authorities are working toward setting up a CfD regulatory framework for investments in low-carbon generation technologies with the assistance of the European Bank for Reconstruction and Development.

As part of its efforts to boost investments in new power generation capacities, in mid-May 2020, the Romanian government also amended, by government emergency ordinance, the Romanian Energy Law.[8] Under an exemption included in the law, electricity producers can now conclude bilateral power purchase agreements for electricity generated by production capacities commissioned after 1 June 2020. Prior to this amendment, trading all generated electricity on the centralised national market was the rule, irrespective of the date of commissioning of the generation capacity. This amendment has been long requested by sector players, ever since power trading limitations were introduced back in 2012, as such agreements can serve as guarantee instruments and help persuade financial institutions to support investments in new local power generation capacities.

This latest amendment to the Romanian Energy Law should be regarded by also considering EU developments such as Regulation (EU) 2019/943 on the internal market for electricity (the "Regulation"), which entered into force on 1 January 2020. According to the Regulation, in order to allow market participants to be protected against price volatility risks on a market basis, and mitigate uncertainty on future returns on investment, long-term electricity supply agreements are negotiable over-the-counter, subject to compliance with Union competition law. The Regulation makes no distinction based on the commissioning date of the power generation capacity and, therefore, this amendment to the Romanian Energy Law can be seen as an important first step in the efforts to harmonise local energy rules with European Union law. The Hinkley Point C decision raises considerable concern that indirect mitigation of nuclear construction risk through mechanisms meant to mitigate electricity market volatility is vulnerable to being incorrectly cast as impermissible State aid.

Watch this space

Undoubtedly, C 3&4 are key to ensuring the country's energy security over the next decades. On the other hand, the combination of CfDs and EU competition laws will make financing C 3&4 difficult. It forces Romania to look to as yet unproven financing mechanisms for nuclear power projects. It also likely materially reduces the pool of viable funders.

Although we have said this before on numerous occasions, it bears repeating: nuclear projects are different from conventional infrastructure projects in crucially important ways and those who insist on approaching a nuclear new-build project as if it was just another infrastructure project very likely have a future appointment with disappointment.

Hopefully, the Romanian Government will take the current opportunity to reimagine the pathway to the successful financing and completion of C 3&4. It would go a long way to achieve multiple objectives of the country — energy independence, national security, and as a concrete example of successful foreign investment.


[1] See https://www.anre.ro/ro/energie-electrica/rapoarte/puterea-instalata-in-capacitatiile-de-productie-energie-electrica (available in Romanian only, accessed on 28 May 2020).

[2] SNN has been listed on the Bucharest Stock Exchange since November 2013. The Romanian state currently holds 82.49% of the company's shares.

[3] For more details, see https://cnpp.iaea.org/countryprofiles/Romania/Romania.htm. See also: Ioan Rotaru, Mircea Metes, Mihnea Serban Anghelescu, CERNAVODA NPP UNIT 1 – A PLANT OF SEVERAL GENERATIONS, https://inis.iaea.org/collection/NCLCollectionStore/_Public/33/011/33011192.pdf

[4] Considering the average 90% capacity factor (e.g. 92.37% in 2018) at which C 1 has been operating since its commissioning, the limit of its operational life is estimated to be reached towards the end of 2023. The refurbishment of C 1 is therefore very high on SNN's priorities list.

[5] Commission Decision (EU) 2015/658 of 8 October 2014 on the aid measure SA.34947 (2013/C) (ex 2013/N).

[6] OECD, Trade and Agriculture Directorate, Arrangement on Officially Supported Export Credits, TAD/PG(2020)1.

[7] C 1 would be effectively shut down for refurbishment works during the December 2026-December 2028 period, as per the NECP.