Decontamination and decommissioning
Financing future liabilities25 March 2010
Two generations will pass before a new civil nuclear reactor will need to be decommissioned. There are many provisions now for paying for those obligations. Jonathan Mirrlees-Black examines the ways that countries have chosen to finance them.
Generation of electricity from nuclear power leads to two main types of deferred costs: the cost of dismantling the power plant and returning the site to non-nuclear use, and the cost of managing spent nuclear fuel. These costs will be incurred years or even decades after the production of electricity. There are numerous issues of public concern about these costs, including the scale of costs, ensuring that operators make appropriate provision for future costs and that there will not ultimately be a call on public finances to meet them.
An appropriate financing structure for these long-term liabilities can make an important contribution to the public acceptability of nuclear power. This is particularly important at a time when there is a real prospect of a very large increase in nuclear power because of its role in reducing carbon dioxide emissions and increasing energy security of supply.
Three major changes in the electricity industry have occurred since the last rapid increase in nuclear power. First, many countries have liberalised their electricity markets, which means that revenues for electricity generation are based on market prices. In such markets, there is no guarantee of future revenues, which will depend on market characteristics and other fuel prices. Second, many governments have sold off stakes in their electricity companies, or at the very least expect the private sector to shoulder much of the burden of future investment. Third, concern about climate change has led to a range of policy interventions to support non-fossil fuel electricity generation (sometimes, but often not, including nuclear). All of these developments can affect the way in which policy for financing nuclear liabilities should develop.
Liabilities vs. pensions
Nuclear liabilities have a number of characteristics that are analogous to the liabilities of defined benefit pension schemes.
• The liability will often crystallise many years after it was created, with commitments of several decades.
• Nuclear and pension obligations can be large when compared to underlying assets or market capitalisation. For example, the German utility RWE, with an installed capacity of 10GW of nuclear generation, reports nuclear liabilities of EUR9.5 billion representing 28% of its December 2009 market capitalisation.
• There is uncertainty about the size of the obligation. With pension funds, uncertainty relates to mortality assumptions and the way that promised payments are indexed. For nuclear, there is uncertainty about regulation (there may be changes in the way in which companies are required to do the decommissioning or manage waste), costs (inflation of different cost components), and the date when the liability will crystallise (power station lives are uncertain).
• In both cases, assets are held against the future obligations, and an investment policy needs to be devised so that those obligations can be met at least cost.
• Stakeholders will have differing views about the appropriate way of managing investment and managing other risks.
• There are occasions when both a pension fund sponsor (such as a bank) and a nuclear operator could conceivably face a situation of ‘moral hazard’, such as in a financial crisis, when a company might be tempted to pay shareholders instead of the decommissioning fund.
These characteristics mean that pension fund beneficiaries are typically protected by detailed regulation. Most countries have also adopted regulation of nuclear liabilities. However, concerns about the nature of the underlying liabilities (particularly their large size, long duration and cost uncertainty) lead some commentators to conclude that they can only be managed in a public-sector organisation.
The focus of this article is on provisions for future costs associated with electricity generation, rather than the larger-scale arrangements for total nuclear liabilities. It is important to ensure that funding arrangements will be adequate. However, it is not straightforward to benchmark decommissioning liabilities for different electricity generation companies. Cost estimates are based on different assumptions about the method used for decommissioning, different waste storage regulations, different nuclear generation technology, different assumptions about the escalation of future costs, and different discount rates. Generally speaking, the present, or discounted value (PV) is the amount that needs to be set aside today to meet future cost (C) in n years, assuming a discount (interest) rate of r. It is is calculated using: PV=C/(1+r)n.
The Nuclear Energy Agency (NEA) analysed decommissioning costs for reactors of different sizes, and found cost estimates (undiscounted, in 2001$) of between $93/kW and $909/kW, with most estimates in the range
$2-500/kW with an average of $320/kW . Costs for larger, more modern reactors are lower. Reported discounted decommissioning estimates are about this range, suggesting cost estimates are reasonable.
Analysis of total nuclear liabilities for a selection of European utilities is depicted in Figure 1, below. This shows that – as would be expected – provisions are higher for companies whose power stations have relatively shorter operating lives remaining.
It is, nevertheless, crucial that there is a process for estimating liability costs, and challenging them, that will facilitate the public acceptance of the costs.
There are two main ways in which nuclear liabilities are financed. Some countries levy charges on nuclear utilities, and in exchange for this the government takes responsibility for nuclear liabilities. Alternatively utilities remain responsible for nuclear liabilities, and make provision for the costs on their balance sheet. In the latter case, governments may require the establishment of segregated funds to restrict the way that accumulated funds are invested, and to ensure they are available when required.
The former approach is taken by Spain. A state-owned company, Enresa, is responsible for both waste management and decommissioning of all nuclear power stations in Spain. Its operations are funded by a levy on electricity consumers. Revenues from this contribute to a fund to finance future operations, as well as Enresa’s current operating costs. As a result of these arrangements, the companies that own nuclear power stations do not report nuclear liability provisions on their balance sheets. If future costs turn out higher than currently expected, the contributions to Enresa would rise, but the levy is made on consumers.
A similar approach is taken in Sweden, with payments from a levy on operators made to a fund to cover all future nuclear liabilities. However, formal guarantees are made by utilities to cover additional costs in the event that contributions are not sufficient to cover all liabilities.
Japan also uses this type of system, with a fee charged to operators for waste management based on electricity production, and the government also requires contributions from utilities into a fund to cover decommissioning .
In contrast, utilities operating in other European countries typically bear direct responsibility for the liabilities. They make a provision in their balance sheet to reflect the future costs that they will need to incur. In Europe, companies are required to report such provisions under International Financial Reporting Standard (IFRS) rules (IAS37). These require the use of the best estimate of future costs and a discount rate that reflects the time value of money as well as the risks appropriate to the liability. In many of these countries there is a requirement to contribute to a fund, the accumulated value of which can only be used for authorised payments associated with nuclear liabilities. EDF, for example, was required to establish ‘dedicated assets’ under 2006 legislation. In Switzerland, utilities are required to pay into a government-run fund, the value of which appears as a receivable in the accounts offsetting the measured liability. In Germany, though, there is no requirement for operators to establish segregated funds.
The USA’s approach is a mixture of these. Utilities make payments for decommissioning liabilities into funds regulated by the Nuclear Regulatory Commission, which reviews contributions to ensure that funds will be sufficient. For waste management, 1982 legislation required utilities to enter into contracts with the appropriate government department under which they would pay a fixed fee, in exchange for which the government would transfer waste into storage. However, because the storage site has not been agreed, utilities are storing waste on site, and the government is reimbursing them for the additional costs incurred.
In the UK, liabilities associated with British Energy’s AGR stations and the Sizewell B PWR will be met by the Nuclear Liabilities Fund (NLF), to which British Energy still has obligations to contribute. (Liabilities for the Magnox stations are the responsibility of the Nuclear Decommissioning Authority.) Following the restructuring of British Energy, prior to its relisting, the UK government gave an indemnity against any shortfall in funding. However, the financial position of the fund is now healthy compared to the liabilities (liabilities were GBP5.6 billion in September 2009, and the NLF fund should now exceed GBP8 billion following the GBP4.4 billion payment into the fund following the acquisition by EDF).
For the new build nuclear in the UK, new arrangements are being put in place . Utilities will need to put in place a funding plan that will satisfy the Secretary of State that sufficient funds will be available to meet future obligations. For decommissioning, operators will be liable to build up funds to meet expected costs. For waste management, the funds will need to be sufficient to cover expected fees, which will be fixed in advance by the government. A new body, the Nuclear Liabilities Financing Assurance Board (NLFAB), advises ministers on whether to accept funding plans and to monitor the arrangements. It is intended that a future government would not be required to provide additional funds.
A good method of financing decommissioning should have the following characteristics:
• Ensure funds are available to meet future costs;
• Provide assurance of availability of funds in the event of an early closure or increase in costs;
• Have sufficiently transparent process and management to be acceptable to the public;
• Be robust to withstand changes in corporate control;
• Ensure that all is done at least cost.
Most countries that have not already done so are moving towards the use of segregated funds (exceptions are Germany and the Netherlands, ). It is possible to devise good financing arrangements without segregated funds. However, now that future revenues are often not guaranteed and that a change in corporate control is likely over long timescales, segregating funds is the most straightforward way to safeguard them.
Segregated funds do not necessarily need to be fully-funded. A deficit in the funding arrangements will still be a liability of the sponsor firm, which will need to meet payments when they fall due. What matters is that there is assurance that future obligations can be met: this can be through the segregated fund, but it could also be met by a type of insurance arrangement. If the sponsor were to default, this could trigger a payment into the fund to make up the deficit. The insurance arrangement, therefore, could be a simple form of credit default swap (CDS), the price of which will depend on market perceptions of risk.
Discount rates for calculating the net present value of liabilities are a very real issue, which has implications for funding profiles as well as corporate reporting. A change of 0.25% in the discount rate changes the calculated liabilities by around 10%. Under IFRS, the discount rate used for reporting of long-term liabilities such as pensions is an appropriate corporate bond rate, and there is often considerable discretion about which rate to use.
What rate is appropriate? Rates ranging from 3-5.5% nominal are used for nuclear liabilities, which is a large range. Given that costs are probably uncorrelated with movements in the stock market and economy as a whole, a case can be made for using the risk-free rate. The risk free rate is the return on government securities. The rate varies with time and is typically 2-3%, although currently in the UK it is around 1%.
What investment policy is appropriate? Guidance from regulatory authorities (such as the UK) normally points to lower-risk securities such as government bonds. However, is this necessarily the lowest-cost way to accumulate sufficient funds?
A strategy involving equities would obviously be riskier but gives the prospect of lower costs. With careful design of appropriate insurance contracts, government need not assume any additional risk from a riskier investment portfolio designed to meet the obligations over the longer term.
Does the state need to fix the level of certain types of payment? The proposed arrangements for the UK involve fixing the fees for waste management, as has been done in the US. Is this necessary? When there is significant risk of cost inflation that is out of the control of the utility, this does seem appropriate. Without fixed fees, governments could impose additional costs after the fact, which would jeopardise the investment. However, if there is a track record of a government working closely with industry, their mutual trust might remove the need for such a contract.
Does the state need to be the ultimate guarantor of payments? An EU guidance note recommends that the government is explicitly made the guarantor . Arguably, government is a guarantor by default (much as the government guaranteed the banking system by default in the current financial crisis). Nevertheless, insurance arrangements discussed above could be required in place of a government guarantee, or indeed the government could take a fee for providing the guarantee instead.
The privatisation of the industry since the last big build up of nuclear power means that new ways of financing nuclear liabilities are emerging, such as insurance type arrangements used to guarantee funds as well as new ways of overseeing segregated funds. The evidence suggests that these arrangements need not be a barrier to investment by the private sector. However, it is important to put the costs into context – decommissioning costs represent only 10-15% of the construction cost of new generation. Achieving growth in the industry depends on many other factors, including effective management of large complex construction programmes, and clarity about the future remuneration of all types of low-carbon generation.
Jonathan Mirrlees-Black is a management consultant and investment analyst with over 20 years experience of the energy industry. Email: [email protected]Related ArticlesNew unclear liabilities Financing new build RVA to assess liabilities at BE's plants Nuclear in the UK – where did it go wrong?
 NEA (2003). Decommissioning Nuclear Power Plants: Policies, Strategies, and Costs. OECD.