This is the second article in a three-part series on existing nuclear in the United States. Part 1 identified and discussed major economic and policy challenges. Part 3 will look at other state, regional, and federal policy solutions.
- Although the economic struggles of the existing US nuclear fleet are clear, potential solutions to these challenges remain somewhat undefined and unanalyzed.
- Any solution must provide sufficient incentives to generators, reduce uncertainty, and be politically feasible.
- This is a clear situation where a one-size fits all approach will not work – instead, planners will likely have to consider many different policy and regulatory avenues.
- Short term policy fixes may be sufficient to protect the most vulnerable nuclear reactors but more systemic policy reforms are needed for the long-term sustainability of the entire nuclear power fleet.
Nuclear Retirements Under Increased Scrutiny
In the last several months, the announcement of planned retirements at 5.1 GW of nuclear capacity in Illinois and California underscored the challenges facing the U.S.’s nuclear fleet.
While there are many factors at play there are four dominant issues:
- Restructured electricity markets structurally favor short run marginal costs and largely ignore long term considerations
- Low natural gas prices reduce electricity prices in these markets and also challenge the cost effectiveness of nuclear plants in regulated markets
- Continuing growth in renewable energy, especially solar, will only exacerbate these two issues in the long term
- The nuclear fleet is aging, leading to multiple issues including the need for large additional capital expenditures
Unless market or policy design changes, a large portion of the existing nuclear fleet faces retirement in the short term, resulting in increased national carbon emissions. As renewable energy continues to grow rapidly and plants continue to age, the rest of the fleet will also face retirement risks if current market design persists.
Although policymakers are increasingly recognizing these issues, questions remain about how to address them.
Assuming that we want to maximize generation from existing nuclear units, what are the policy or regulatory mechanisms to do so?
This article profiles the design, implementation, and risks associated with three prominent and noteworthy solutions at the state government level:
- Maintaining or reinstating cost-of-service regulation
- Including nuclear in state renewable portfolio standards
- New York’s recent proposal for a carbon value market adjustment
Key Considerations to Regulatory and Policy Solutions
Any solution to keep existing nuclear power plants online will have to manage several considerations:
- Increase nuclear revenues sufficiently to maintain profitability
- Reduce regulatory and market uncertainty for existing nuclear plants
- Be politically feasible
Of these considerations, numbers 1 and 2 deal with largely technical issues: how does the policy function and what design issues does it pose?
The third consideration, politically feasibility, may be the most challenging. The economic struggles of the U.S. nuclear fleet are occurring against the backdrop of sweeping regulatory changes in the power sector.
In large portions of the country, states have enacted regulatory restructuring during the last twenty years. This is a major cause of the economic challenges facing nuclear units but also has ramifications for regulatory solutions. In many regions, regulatory power is shifting from state legislatures and PUCs to the quasi-independent ISOs overseen by FERC.
Any solutions to nuclear’s challenges are thus constrained by the distribution of regulatory power: policymakers, industry, and nuclear advocates need to know which levers are best suited to addressing challenges at specific facilities and at what times they should employ each lever.
This part of our series specifically addresses the options available at the state level. Critically, it highlights how the recent shift of power from the states to FERC may legally constrain how state’s exercise those options.
There are several important temporal aspects to all potential solutions:
- Some plants face short term challenges while many face long term competitiveness issues.
- All of the existing nuclear fleet will inevitably retire when the plant’s reach their lifetimes and their licenses expires. The question is when.
- Considering that reactors will face different retirement risks at different times, policy solutions at the state, regional, and federal levels should be sequenced.
As usual with the U.S. electricity system, many different approaches are likely to be pursued by different states and regions. One solution may not fit all circumstances – understanding the pros, cons, and feasibility of each solution is essential to develop the best solution for different parts of the electric system.
Maintain or reinstate Cost of Service Regulation
Of the U.S.’s existing 99 reactors, 50 operate in states with cost-of-service regulation, reducing their retirement vulnerability in the short term.
Of the five reactors that retired in the last five years and the nine reactors plan to retire in the next ten years, only two were in regulated states. Even these two plants are outliers compared to the entire nuclear fleet:
- The 860 MW Crystal River unit closed due to a crack in its containment dome
- The 478 MW Fort Calhoun unit was the smallest in the fleet with generating costs above $70/MWh (much higher than the ~$44/MWh average for single-unit plants)
The most pressing economic headwinds for the U.S. nuclear fleet result primarily from marginal price-based compensation in competitive wholesale electricity markets. The 50 reactors operating in regulated states still compete indirectly with natural gas but are much less likely to retire than reactors in deregulated states.
Effects of Regulated Status on Nuclear Power Plant Retirement Risk
Source: SparkLibrary based on data from EIA
Why the difference between deregulated and regulated nuclear reactors?
In deregulated markets, nuclear revenues are dependent solely (or primarily) on energy markets – they are highly vulnerable to even short periods of low wholesale electricity prices. Long term contracts are uncommon due to structural, market, and political reasons.
Comparably, traditional rate-regulation is based on average cost compensation. State PUCs determine what the costs are to keep individual plants open and allow plant owners to earn a specified rate of return. There is some indirect competition from natural gas and other energy sources, but any decision to close a plant would have to be approved by the commission and (most likely) proposed by the utility. As long as the owners of regulated nuclear generators receive their required rate of return, they have limited incentive to close existing plants.
Importantly, many of these rate-regulated nuclear reactors are smaller and/or are single-units – if exposed to competitive markets, these units would face much higher retirement risks.
If the status quo persists, COS regulation will protect almost half of the nuclear fleet, even relatively uneconomic smaller units. To the degree that these smaller units are vulnerable in a regulated system, economic arguments at the PUC level may be most effective at keeping plants online.
However, if ISOs continue to grow and more states deregulate, many of these currently ‘protected’ reactors could become vulnerable.
Source: EIA (as of 2010)
Importantly, a number of the reactors in regulated states are in SPP and MISO, where competitive wholesale electricity markets are operated. Despite these markets, competition remains relatively limited in these areas and cost-of-service regulation still dominates. These areas face elevated retirement risks compared to other regulated reactors; the presence of transparent, functioning wholesale markets could make it easier for a utility to retire a nuclear unit and replace its output.
Importantly, COS regulation is not a fleet-wide policy solution due to the many nuclear units in deregulated states. It is highly unlikely that regulators will reinstate COS regulation in restructured states, if only because such a shift would move counter to the prevailing regulatory trends of the last twenty years.
Political and regulatory power in these states has already shifted from state public utility commissions to FERC and the ISOs. To re-regulate only nuclear plants in these states would invite significant opposition from many interest groups, most other generators, and likely many regulators and politicians.
Making State RPS Policies into Low Carbon Standards
One of the more recent and most promising proposals to protect nuclear calls for an expanded role for state RPS policies.
RPS programs already provide a broad (if diffuse) policy platform: twenty-nine states and the District of Columbia already require that load serving entities obtain a certain level of retail sales from renewable resources. These policies generally exclude nuclear and primarily support wind and solar.
A Low-Carbon Portfolio Standard expands these existing policies to include nuclear power. Under such a proposal, existing nuclear power plants would sell nuclear energy credits (NECs) to utilities obligated by the state to purchase them. This could provide sufficient revenues to make up for insufficient revenue from the marketplace.
In most states, the process of transforming the RPS into a LCPS is relatively straightforward: state legislatures pass legislation increasing the stringency of the RPS to match existing nuclear generation in that state and adding nuclear as an eligible resource. The Breakthrough Institute estimates that such an approach would increase the amount of generation in 2030 covered by RPSs from 420 TWh to 940TWh.
There is significant merit to the proposal. Most states in PJM, ISO-NE, and NYISO have existing RPS policies, meaning that most states with deregulated nuclear plants have a RPS they could readily modify. Similarly, most states in MISO have RPS policies, potentially protecting the regulated nuclear units in that hybrid market.
RPS Policies by State
RPS policies have proven robust once passed. If well designed, a LCPS could provide critically needed certainty for struggling nuclear reactors.
However, there is a major political challenge to this proposal: getting it passed in the first place. Each state legislature would need to pass legislation to include nuclear in the state’s respective RPS policy. Unlike renewable energy, nuclear energy does not enjoy universal public support and such legislation would likely face significant public opposition.
Technical Details of a Nuclear-Inclusive RPS
Aside from political challenges, nuclear-inclusive RPS policies entail significant technical considerations. Nuclear reactors qualitatively differ from renewable energy in political, economic, and technological terms.
RPS policies function through the use of market-based renewable energy credits – eligible renewable facilities receive a price for their renewable attributes based on competition between multiple generators, minimizing the cost of the policy. Comparably, the large size of nuclear units and relatively limited number of units in any single state make determining the value of nuclear energy credits (NECs) challenging.
If NECs and RECs are treated the same, several problems arise:
- Due to their large capacity and output, nuclear plants could exercise market power, distorting the NEC/REC price
- If designed poorly, such a structure can create competition between nuclear and renewables in the secondary NEC/REC market, leading to conflict between policy goals.
- There is no guarantee that a combined NEC/REC price will provide sufficient revenues to the most vulnerable nuclear generators
This last point is especially important as decreasing renewable energy costs could make financial support for renewables from RPS policies less important. REC prices may decline over time while nuclear revenue needs from NECs may need to rise due to merit order issues.
The likely solution is to treat NECs as a separate generation class under RPS policies. REC prices would still be determined by renewable competition; NEC prices would need to be administratively determined. Proper market competition for NECs is impossible when there only a handful of reactors in a state.
Administrative determinations are technically challenging and would need to be transparent. The staff of the PUC would need:
- To determine how much nuclear generation the policy requires
- Identify the NEC price needed to keep the marginal nuclear plant operational
- Adjust the NEC price based on variations in market revenue
Legal Issues Present Policy Design Ramifications
Each of these raises issues. However, #1 may be the most challenging due to legal limitations in state’s ability to regulate interstate commerce.
Functionally, RPS policies work by requiring covered entities to purchase RECs equivalent to a specified share of their retail load. While there are some barriers due to REC tracking systems, this design allows utilities to use out-of-state RECs to meet RPS mandates in almost all cases. RPSs are ‘supply blind.’
Interstate REC Trade is a Key Component of RPS Compliance
A nuclear-inclusive RPS would likely be different: the in-state nuclear generaiton would determine the amounnt of NECs required. If the state allowed out-of-state NECs to be applicable, nuclear units in neighboring states without nuclear-RPS policies would be able to sell NECs, undermining the state’s efforts to keep its own nuclear facilities open. For renewables, this is not an issue as they are growing (no incumbent generator with free RECs to interfere).
By specifically favoring in-state nuclear units, a nuclear-RPS may invite a legal challenge under the dormant commerce clause, which prevents states from favoring in-state business interests. Multiple parties are actively challenging state RPS policies on these grounds. The supply blind nature of RPS policies has (so far) kept them from being overturned.
The 10th Circuit Court of Appeals recently highlighted the issue when examining a challenge to Colorado’s RPS:
“But as far as we know, all fossil fuel producers in the area served by the grid will be hurt equally and all renewable energy producers in the area will be helped equally. If there’s any disproportionate adverse effect felt by out-of-state producers or any disproportionate advantage enjoyed by in-state producers, it hasn’t been explained to this court.”
A nuclear-RPS policy favoring in-state nuclear units clearly provides an advantage vis-à-vis all out-of-state electricity generators.
This does not mean that a nuclear-inclusive RPS policy is certain to be illegal. Rather policymakers will need to consider the legal ramifications as they design such a policy. Coordinating many states to include nuclear in their RPS policies without in-state restrictions could get around this issue (but would be more difficult).
- New York’s Carbon-Value Based Price Floor
In early July, the staff of the New York Public Service Commission (NY PSC) released a proposal to provide revenues to New York’s upstate nuclear power plants under the developing New York Clean Energy Standard. In August, the Commission approved the proposal. Critically, the policy is not actually the nuclear equivalent of a RPS. Although this policy is unique to New York, it may provide a model for future programs.
New York’s proposal (available here) specifically targets four reactors in upstate New York with a combined capacity of 3.2 GW: Nine Mile Point 1+2, Ginna, and James A. Fitzpatrick. Fitzpatrick is currently scheduled to retire but Exelon plans to purchase the plant and keep it operational if the proposal is approved.
The proposal excludes the remaining two nuclear reactors in New York at the 2.1 GW Indian Point power plant on the basis that it receives much higher revenues from downstate electricity markets. In the staff’s view, the plant is not at risk of closure due to economic reasons. This is a key difference between New York’s proposal and a nuclear-inclusive RPS: a RPS would not discriminate between in-state plants like this.
Based on historic generation, the four eligible reactors contract to sell zero carbon electricity credits (ZECs) to NYSERDA for two year periods during the next twelve years. Load-serving entities would then be required to purchase ZECs from NYSERDA to reflect their load-share in the state.
How ZEC Price is Calculated
Due to market power issues, the ZEC price is administratively determined based on the following formula (from the proposal):
Effectively, this formula creates a price floor for the covered nuclear facilities.
If forecasted energy and capacity prices are at or below $39/MWh plus the social cost of carbon (SCC) converted to a $/MWh value (modified by the ‘priced-in’ value of RGGI credits), the ZEC price makes up the difference. If market revenues stay at the $39/MWh level or below, the ZEC price will be between $17.48-$29.15/MWh, depending on the year and RGGI.
Based on the output from the reactors, these prices would translate into total costs of up to $482 to $805 million annually.
Importantly, the ZEC is not a true price floor:
- The ZEC maxes out at the SCC-RGGI value. If combined energy and capacity prices fall below $39/MWh, the ZEC remains capped at SCC-RGGI-$39/MWh.
- The actual maximum price of a ZEC is a bit higher than $17.48-$29.15/MWh used above; if RGGI prices approached $0/ton, the maximum ZEC would be ~10-20% higher.
While there is a maximum price to the ZEC, the ZEC can also decline as market revenues rise above $39/MWh.:
- If market revenues exceed $39/MWh plus the SCC-RGGI, the ZEC would be $0/MWh. To put it another way, the ZEC price would be $0/MWh when energy and capacity prices exceed ~$55-$70/MWh (depending on the year).
- Similarly, if RGGI prices rise, the ZEC would be reduced in value correspondingly; if RGGI was higher than the SCC, the ZEC price would be $0/MWh.
By guaranteeing that the covered nuclear facilities will receive a specific revenue amount, the proposal provides economic certainty in a way that market revenues alone cannot. Barring severe maintenance or safety issues these plants will remain open for the 12 years of the program (which ends in 2029 when Ginna and Nine Mile Point 1’s NRC licenses expire).
Technical Design May Have Unexpected Consequences
There are some drawbacks and potential questions about NY PSC’s methodology though.
Importantly, it is value-based not cost-based. Initial estimates from the staff of the NY PSC indicated that keeping the vulnerable upstate nuclear reactors would only cost $8 to 94 million annually. Even if this were a low estimate, it is <20% of the cost of the new program. If we are looking to keep nuclear reactors open at the least cost, this proposal may not be the best solution.
Relatedly, the $39/MWh combined energy and capacity price used to calculate the ZEC is arbitrary. It is determined based on forecasted average energy prices between April 2017 and March 2019 and based on a capacity price forecast for April 2017 to March 2018. Both power and capacity prices are very volatile – there is no solid justification to use limited 1 or 2 forecasts as the basis for a 12-year subsidy.
Similarly, the ZEC price for any specific two-year period is determined based on a two-year forecast for power prices and a one-year forecast for capacity prices made the year before:
As power prices are dependent on natural gas prices, they can be exceptionally volatile. In this case, the power prices forecasts occur before a gas-market resetting winter and span two natural gas storage cycles. A lot can happen to power prices over the course of two years. While less volatile, capacity prices can also vary year to year.
Predicating the ZEC based on forecasted prices instead of realized prices could lead to the financial effects of the program diverging from their intended purpose:
- If forecasted power prices are high and exceed SCC-RGGI-$39/MWh, the ZEC price would be $0/MWh and the nuclear reactor would be taking market-only prices. If power prices subsequently fall, the nuclear reactor would lose money again.
- Conversely, if forecasted power prices are $39/MWh or less, the ZEC price would be at its highest. If power prices subsequently rise, the nuclear reactor would receive the ZEC benefit even if power prices alone were sufficient to keep the plant operating.
Legal challenges could be more stark than a nuclear-inclusive RPS
Despite these considerations, New York’s proposal is one of the most concrete policy solutions to date that could prevent nuclear retirements.
There are notable legal risks. A similar dormant commerce clause challenge to the in-state nuclear-RPS issue discussed earlier may also occur here. As the ZECs are being sold to NYSERDA directly, this may be less of an issue for New York.
However, there is a potential legal challenge to tying the ZEC price to wholesale market prices due to the Supreme Court’s recent ruling in Hughes v. Talen. In that case, the Supreme Court ruled that Maryland cannot directly modify the capacity price received by new generators from PJM’s wholesale market.
Importantly there are major differences between Maryland’s case and New York’s proposal:
- The ZEC is a separate product than energy or capacity. Maryland’s case invovled specifically modifying the price for a product regulated by FERC (capacity).
- While modified based on the wholesale price, the ability to receive a ZEC is not directly tied to participation in the wholesale markets.
- Maryland’s intent was to specifically modify revenue from an existing capacity market. While New York’s proposal may functionally do so, the intent is to prevent carbon emissions.
These issues require further investigation. If litigation delays the implementation of the policy, Exelon may decide not to purchase the Fitzpatrick plant reactor from Entergy, leading to it retiring in 2017.
- The Breakthrough Institute’s report on including nuclear in state RPS policies.
- State policy to protect nuclear plants need to be legally sound – this guide describes how minimize constitutional risk.
- Solid numbers regarding the economics of the nuclear fleet.
- Good article covering the discussion on the future role of nuclear.