• Webinars

The State and Outlook for Nuclear Generation

Watch Webinar
View presentation
Explore the future of U.S. nuclear energy and how shifting policies, next-generation technologies like SMRs and microreactors, and global competitiveness are shaping the nuclear landscape. Discover cost outlooks, integration challenges in high-renewable grids, and why data centers could be nuclear’s next big customer.

UPDATE: 12-1-2025

The Paris Agreement targets limit global warming below 2°C, with an additional goal to keep global temperatures below 1.5°C, from pre-industrial.[1] Signatories are required to review their targets periodically and to commit to more stringent ones where it is cost effective to do so.

The United States had adopted a trajectory to reduce emissions by 50-52% of 2005 baseline levels by 2030 under its NDC to the Paris Agreement but is expected to alter or scrap the targets per the withdrawal from the Paris Agreement.

As shown in the Figure 1 below, a number of states have committed to the same or more stringent targets and are expected to maintain them. The key question for these states is whether nuclear generation could reduce the cost of their power system transition to zero carbon?

Figure 1 – Historic and Projected Emissions for the US and Key States, Source: Source: Energeia Research, US EIA (2023)

Nuclear Policy

Given the significant cost of developing and maintaining a nuclear capability, with potential military linkages, pursuing nuclear generation is very much a national policy question. The US has traditionally supported it, but its role in the energy transition is unclear, with solar PV, wind and battery storage being the focus to date.

Table 1 summarizes current nuclear energy targets of the US and other key jurisdictions globally. It shows that most countries with existing nuclear generation are expecting to increase their overall share of nuclear energy over time. A notable exception to this is Japan, which is expected to reduce its overall dependence on nuclear power going into the future. Canada notably has no explicit targets despite being a world leader. However, they are currently investing in nuclear power on an ongoing basis. France has reversed their 2014 policy to lower nuclear penetration from 70% to 50% and, since 2022, plans to install 6 new reactors.

Table 1 – Summary of Worldwide Nuclear Targets, Source: Energeia Research, Note: BRI = Belt and Road Initiative

Table 2 summarizes the incentives being used in each jurisdiction to encourage investment in nuclear energy. The research shows that jurisdictions are funding their nuclear energy programs through a mix of federal subsidies, tax incentives, and loans. Both the UK’s and France’s policies offer developers a contract-for-difference (CfD) style arrangement, thereby locking in a bankable sale price for all energy generated by the power station once it is constructed.

Table 2 – Summary of Worldwide Nuclear Incentives, Source: Energeia Research Note: CfD = Contract-for-Difference, RAB = Regulatory Asset Base

The impacts of these varied targets and incentive approaches to nuclear will heavily depend on the progression of nuclear technologies in future. The key question is whether nuclear technology is able to compete with and/or complement solar PV and wind based zero carbon energy systems.

Nuclear Technologies

The following section summarizes key nuclear energy technologies, current historic rates of development, and future technologies.

Key Definitions for Nuclear Technologies

The most common and/or prospective nuclear energy technologies are summarized in Table 3 below. Generation IV nuclear reactors are highlighted in light green, and the Small Module Reactor row is highlighted in light blue as is the focus of most next generation technology solutions and appears to hold an edge over conventional reactor designs for reasons spelled out below. Additionally, the microreactor is an emerging area of technological development, due to its even more compact design.

Table 3 – Nuclear Reactor Technologies, Source: Energeia Research Note: * - Indicates Generation IV Technology

Of the types listed in the table above, the most common nuclear reactor technology that is used today is the Light Water Reactor (LWR). The two most common types of LWRs are the Pressurized Water Reactor (PWR) and the Boiling Water Reactor (BWR), which have a global market share of 71% and 14% respectively. Their operation mechanism is similar, except that a PWR transfers heat through two fluid circuits, compared to a BWR, which only uses a single fluid circuit.

Figure 2 shows a diagram of the operation of a PWR.

Figure 2 – Pressurized Water Reactor, Source: United States Nuclear Regulatory Commission
Table 4 – Conventional vs Small Modular Reactors (PWR Case Study), Source: Energeia Research. Note: * – Based on current NuScale SMR estimates. For the cost assumptions, this includes no subsidy

The above nuclear technologies can generally be grouped into the following three key categories:

  • Light water reactors – PWRs and BWRs. Both are the most widely implemented technologies to date
  • Small modular reactors – This approach can be applied to most of the other reactor technologies
  • Generation IV nuclear reactors – These aim to be the next step in reactor designs, targeting improvements to sustainability, economics, safety and reliability, and proliferation resistance
  • Microreactors – Latest technology innovation of SMRs under 20 MWe

Generation IV reactors are still novel, with very limited grid deployment to date. However, with nuclear becoming increasingly relevant for major flat loads including data centers and industrial electrification applications, combined with the planned construction of Gen IV nuclear reactors, they’re expected to capture a greater worldwide generation market share moving forward.

Conventional vs Small Modular Reactors

SMR is an umbrella term for reactors with the following features, as determined by the International Atomic Energy Agency (IAEA)[2] and World Nuclear Association (WNA):

  • Reactors and other major components are designed to be standardized
  • Smaller physical footprint
  • Designed with inherent passive cooling, in the case of a power failure
  • Minimum capacity 30-300 MWe, which can be daisy-chained.

Table 4 compares conventional nuclear reactors to a small PWR modular reactor.

Estimates for the SMR are based on NuScale, which is not yet fully commercialized, but is the only SMR to be certified by the US Nuclear Regulatory Commission to date.

Small modular reactors have recently begun operation in China and Russia, and more are under construction. Table 5 summarizes the SMRs that have been constructed to date. The table shows that to date, only two SMRs have seen grid-connection deployment worldwide, with a total of 280 MWe capacity across the two reactors.

Table 7 – Small Modular Reactors Proposed in the US, Source: Energeia Research Note: CPA = Construction Permit Application

The six identified SMRs under construction are summarized in Table 6 and are projected to add 1,127 MWe of capacity. Current estimates show that SMR pilots are expected to deliver the following cost and build time ranges:

  • ~3-5-year deployment times, with median install time on the higher end of this range
  • ~$18.5m ($/MW) unit rates, with the Canadian reactor an outlier to this range

The actual delivery times and costs for the planned projects won’t be known until the projects are completed.

Table 4 – Conventional vs Small Modular Reactors (PWR Case Study), Source: Energeia Research. Note: * – Based on current NuScale SMR estimates. For the cost assumptions, this includes no subsidy

US Government policy is to catch-up to nuclear leaders China and Russia. The list of proposed reactors shown in Table 7 below, if continued into construction, will see the US share of operating SMRs increase significantly by the start of the next decade.

Table 7 – Small Modular Reactors Proposed in the US, Source: Energeia Research Note: CPA = Construction Permit Application
WATCH WEBINAR
DOWNLOAD PRESENTATION

Listen or click through at your own pace

The summary of installed and currently under construction SMRs show that SMR deployments are not limited to conventional PWR designs. There are reactors currently operating, as well as those under construction, that use advanced reactor technologies such as:

  • High Temperature Reactors,
  • Fast Breeder Reactors, and
  • Molten Salt Reactors.

The use of these ranging technologies results in different operations and build characteristics, as summarized in the following section.

Summary and Benchmarking of Technologies

Key issues with nuclear technology in the past have included cost overruns, minimum sizing, flexibility, leaks, and waste disposal, all of which need to be considered when comparing different technology options.

The following section summarizes the results of our research into cost, lead time, sizing, and technical capabilities of different nuclear energy technologies. These metrics impact their competitiveness in future energy systems, especially those with variable loads and/or variable resources.

Table 8 captures this information, where the data is known in the public domain. The table shows that there are still many data gaps in the public domain regarding key characteristics of reactors, including how they perform when grid-connected. These important characteristics include:

  • Ramp Rate – The rate at which a nuclear power station can change its output, measured in MW/minute or % of rated capacity/minute. This indicates how well it can follow load and/or other co-installed generation technologies, such as variable renewables
  • Maintenance Time – The downtime per year required for refueling and other maintenance activities, including forced outages, which reduce its availability to generate
  • Minimum Load – The minimum level at which a reactor can be safely and stably operated, measured in MW or % of rated capacity, which can lead to negative bidding behavior to remain online rather than turning off, and incurring restart costs
  • Restart Costs – The cost of turning a reactor off and on again, comparable to a cold, warm, or hot start for a coal power station, when components need to heat up, i.e., boilers. Higher costs will lead to larger negative prices to avoid shutting down

All of the above technical constraints are key to consider when integrating into a grid that has an existing generation mix, and known demand characteristics, including ramping requirements due to rapid changes in solar PV or wind generation.

Table 8 – Comparison of Key Nuclear Reactor Technologies, Source: Energeia Research

A new generation of technology is promising to address these concerns. The question is, will they be cost-competitive?

Cost Competitiveness

There are two main use cases for nuclear technology at the moment. The first is playing a role in the energy transition, enabling a zero-carbon power system at lower cost than without it. The second is playing a role in powering new major loads such as power centers, which traditionally have had relatively flat loads, with some weather sensitivity. Data centers could be grid connected or in a microgrid.

Figure 3 and Figure 4 below show the nuclear cost capital expenditures (CAPEX), fixed operational and maintenance cost (FOM), and variable operational and maintenance cost (VOM) outlook comparing conventional nuclear with SMRs. As shown, the CAPEX in the long run for both SMR and Conventional Nuclear is comparable (in $/kW) however in the short term it is higher for SMRs. In contrast, both FOM and VOM is higher for conventional nuclear than SMRs.

Figure 3 – Nuclear Capex Outlook, Source: Energeia Research
Figure 4 – Nuclear FOM and VOM Outlook, Source: Energeia Research, NREL ATB (2024)

A comparison of the levelized cost of energy (LCOE) for nuclear with other generation sources including gas turbines among renewables, shows that Nuclear could be more cost competitive compared to dispatchable resources like gas turbines (open and combined cycle) and utility scale 8-hr battery as shown in Figure 5.

In reality, a system solution will require more than a single generation resource in most cases (excl. gas turbines, which can ramp with load). Solar or wind will require battery storage due to their intermittency. Nuclear may also require battery storage where load changes faster than the nuclear station can.

Table 7 – Small Modular Reactors Proposed in the US, Source: Energeia Research Note: CPA = Construction Permit Application

There are two main use cases for nuclear technology at the moment. The first is playing a role in the energy transition, enabling a zero-carbon power system at lower cost than without it. The second is playing a role in powering major new loads such as power centers, which traditionally have had relatively flat loads, with some weather sensitivity. Data centers could be grid connected or in a microgrid.

The following section shows Energeia’s analysis of the cost competitiveness and system impact of nuclear energy in the current (2024) and future (2040) CAISO grid. Ideally, this analysis would have been nationwide, as there is significant trade between the states. The analysis is therefore limited in that it does not consider inter-state trade.

UPDATE: 12-1-2025 —Energeia reviewed its key inputs and assumptions, and linear solver-based results post webinar to confirm the 2040 CAISO results. As a result of additional linear solver runs, the optimized system mix and average pricing have changed, but the overall result has not; that under the simplifying assumptions made, and the IEPR 2024 load forecasts and NREL ATB costs assumed, our modelling shows that nuclear may result in a lower average system cost.

Figure 6 shows CAISO’s average daily load profile in 2024, which exhibits a distinctive “duck curve” shape from rooftop solar generation during the day and an evening peak immediately following the solar period. Seasonality also influences the average day profile, with lower minimum demand occurring in winter.

Figure 6 – CAISO Average All, Summer and Winter Days (2024, Source: IEPR 2024, CAISO (Baseline Scenario)

Figure 7 shows the price duration curve in CAISO in 2024. This shows that at 2024 prices, an SMR plant with an adjusted (removing est. RA payments) levelized cost (LCOE) of $78/MWh would have received around $41/MWh on average for each MWh sent out, assuming the reactor operates for ~100% of the hours in a year at a fixed output. This includes operating at negative prices, which were observed to occur for ~3% of the time.

Figure 7 – CAISO NP15 Price Duration Curve (2024), Source: CAISO (NP15) (2024)

The plant could avoid negative prices, but that would incur additional restart costs and increase the investment as well as fixed operating and maintenance costs per MWh delivered. These negative prices typically occur periodically, correlating with daily solar generation, meaning that a run strategy including avoiding negative prices could require up to daily or intra-daily plant restarts.

CAISO 2040 Case Study

In 2040, CAISO load is expected to evolve due to increased electrification and adoption of behind-the-meter resources. Figure 8 shows CAISO’s forecasted load for 2040, showing an exacerbated “duck curve” effect with a more significant load throughout the entire day from building and transport electrification. Overall load significantly increases from 2024, but the impacts of seasonality are the same.

Figure 8 – CAISO Forecasted Load, Summer and Winter Days (2040), Source: IEPR 2024, CAISO (Baseline Scenario)

Energeia developed a simplified CAISO wholesale cost-of-service model to investigate least-cost solutions (from an annual capex and opex perspective) to meet 2040 operational demand with varying constraints on fuel type inclusions. This analysis was developed using the following simplifications and assumptions:

  • Analysis uses 2040 resource capex and opex costs from the 2024 NREL Annual Technology Baseline (ATB) dataset, identifying the least-cost combination that serves 100% of load
  • Assumes ramping and excess nuclear generation able to be smoothed out by reported ramping rates and 5-hour battery,
  • Importantly, the analysis does not account for existing capacity (consistent with a long-run-marginal-cost analysis), forced or planned outages, or inter-state trade
  • The analysis relies on 2040 CAISO renewable energy traces from the 2024 CPUC IPER (Baseline Scenario)

Figure 9 and Figure 10 show the key outcomes of this analysis, being the modeled least cost capacity mixes for each of the two 100% (24/7) renewable energy scenarios, one that allows for SMR nuclear technology, and one that does not. The nuclear scenario requires no other renewable energy solution, relying on its internal ramping and 5-hour battery storage instead.

Figure 9 – CAISO Capacity in 2040, With and Without Nuclear, Source: Energeia Analysis using NREL ATB and CPUC IPER ISP Data (2024), Note: OS = Off-Shore

Under the highly simplified modeling and CPUC/NREL assumptions used, the nuclear scenario comes in lower cost than the non-nuclear scenario. The key driver of this outcome is wind droughts; This forces additional renewables and battery firming to be built to overcome these wind droughts and ensure all load is met, pushing up the average cost.

Figure 10 – CAISO Cost ($/MWh) in 2040, With and Without Nuclear, Source: Energeia Analysis using NREL ATB and CPUC IPER ISP Data (2024), Note: OS = Off-Shore

Given the significant assumptions and simplifications made for this modeling, the considerable complexity of the real-world markets, and the immaturity of SMR nuclear technologies and their associated costs, it is uncertain whether these outcomes would hold true.

Additional analysis of regional power dynamics, nuclear ramping capabilities and costs, load flexibility, renewable energy diversity, and allowed unserved energy are needed to increase certainty of findings. The role of existing generation in the transition also needs to be factored in. However, if nuclear and storage is truly the long-term lowest cost, existing sunk costs should not impact the result.

Data Centre 2024 Case Study

Data centers typically run a flat load because much of their power consumption comes from non-time-dependent processing, which is overlaid on synchronous business use.

Figure 11 compares the average price observed by selected US Balancing Authorities for a flat load in 2024. This analysis varies from the foregoing analysis in that it is focused on today’s costs, to inform discussion as to whether a data center is better off connecting to the grid where it can or connecting to a microgrid with or without it where it cannot.

Figure 11 below shows that a data center in most states in 2024 would have seen an average wholesale price of $32.90/MWh, which is above . This suggests that a data center would pay more at the average grid price than the best-priced nuclear station in 2024. Note that this assumes that the additional data center load does not itself increase wholesale market prices, which is increasingly a dubious assumption to make.

Figure 11 – Data Center ($/MWh) by Balancing Authority in 2024, Source: Energeia Analysis using CAISO, PJM, and ERCOT Data (2024)

Where a data center is not able to connect to the grid, or where it may be lower cost not to, the alternative cost of a microgrid system is relevant. Figure 12 shows the costs of different standalone generation solutions to meet data center demand in 2024, again using the 2024 NREL ATB inputs. Results show gas solutions to be least-cost at current costs, and renewables with battery as the least-cost zero emissions option, followed by nuclear as the most expensive.

Figure 12 – Data Center Cost of Service by Solution (2024, $/MWh), Source: Energeia Analysis using NREL ATB and CPUC IPER ISP Data (2024)

It is important to note that even though an OCGT delivers a lower cost LCOE, California’s zero emissions target means that it will face rising CO2 emissions costs over time, which are not factored into the above analysis. Once CO2 emissions costs are factored into the analysis, which are expected to rise over time, it is likely that the nuclear option may become lowest cost, with the assumptions made.

Key Takeaways and Recommendations

Energeia’s key takeaways and recommendations for tackling the workforce skills gap and implementing decarbonization are summarized below.

Key Takeaways:

  • The transition to zero emissions power systems, combined with the rapid growth in major flat loads like data centers, is driving renewed interest in nuclear technologies
  • Most countries with nuclear capabilities are aiming to invest in next-generation technology, or at least keep their options open
  • Small modular reactors promise smaller form factors, faster construction times, without too much of a cost impact from the smaller scale
  • There are very few small modular reactor designs in the world that have been certified
  • Key limitations of the technology still include nuclear waste management, water consumption, ramping rates, and restart costs
  • A few pilot projects are currently underway, and significantly more are planned, with significant results expected in the next 3-5 years or so
  • The high utilization/baseload nature of the technology may find it hard to integrate into a high renewables and/or highly variable net load system without adding in battery storage
  • Data centers, with their flat loads, could be a key future customer, especially where there are connection limitations, but access to water will be key
  • Our analysis shows that even assuming a flat load profile, a modular reactor would not be competitive with a solar + wind + BESS solution in 2024, economically speaking
  • It also shows that for CAISO’s 2040 profile, under simplified assumptions, nuclear may be able to deliver a lower cost, zero emissions outcome than without it

Key Recommendations:

  • Undertake more sophisticated modeling of nuclear and alternative resource mixes, including battery storage and regional dynamics
  • Monitor the outcomes of small modular reactor pilots in the US and around the world to see if they deliver on timelines and cost targets
  • For data centers and other major industrial loads, prioritize consideration of load flexibility to maximize grid connection capacity and minimize standalone costs
  • Focus analysis on the limits and costs of nuclear outage rates and flexibility, including how excess generation may be best utilized
  • It will also be important to monitor the cost of alternative technologies, especially long-duration storage

[1] The United States of America Nationally Determined Contribution
https://unfccc.int/sites/default/files/NDC/2022-06/United%20States%20NDC%20April%2021%202021%20Final.pdf

[2] https://www.iaea.org/newscenter/news/what-are-small-modular-reactors-smrs

WATCH WEBINAR
DOWNLOAD PRESENTATION

You may also like

Scroll to Top