Lecture 6: Nuclear Reactor Cost Around the World and the Potential for Cost Reduction

Overview

This lecture analyzes global nuclear reactor costs, highlighting how the high capital costs of nuclear power, while historically justified for baseload by low fuel costs, are increasingly challenged by cheaper renewables like solar and wind. The analysis reveals that historical cost reductions in countries like France and Japan were often due to standardization and specific regulatory environments, but recent projects like Flamanville and Fukushima demonstrate escalating costs and reliability issues. The lecture emphasizes that cost reduction in nuclear power is not guaranteed through simple learning curves and requires addressing factors like reactor size, safety standards, and construction methodologies.

Key takeaways

Nuclear reactor costs are heavily influenced by scale; smaller reactors (SMRs) are predicted to be ~2.35 times more expensive per kW than larger ones due to loss of economies of scale.
Safety regulations significantly impact cost; reactors built after Three Mile Island cost approximately 40% more than those built before.
Historical cost reductions in nuclear power were often driven by standardization and increasing reactor size, not necessarily by learning curves.
The learning rate for nuclear reactors is very slow (~1% per doubling of experience), meaning simply building more reactors will not drastically reduce costs.
France's low reactor availability (~74.5%) compared to the US (~99%) highlights a trade-off between upfront cost and long-term reliability and maintenance expenses.
The Combined Operating License (COL) process in the US, while intended to streamline operations, can lead to significant construction delays and cost increases due to required modifications.

Chapters

0:00 Introduction to Screening Curves and Technology Optimization

Screening curves illustrate optimal technology choice based on usage frequency.
Peaker plants favor low capital, high variable cost technologies.
Baseload operations (8000+ hours/year) favor high capital, low fuel cost technologies like nuclear.
The y-axis represents the cost to have capacity available, independent of generation.

0:15 The Role of Nuclear Power in the Energy Mix

High capital cost technologies like nuclear have a role when geothermal is unavailable.
Solar and wind cost reductions in the 2010s significantly impacted nuclear's competitiveness.
Nuclear remains relevant in regions with poor solar/wind conditions (high latitude, cloud cover, limited land).

10:07 Nuclear Reactor Cost Estimates: Optimistic vs. Empirical

Vendor predictions (e.g., Oklo, TerraPower) are often optimistic and unproven.
Empirical costs, averaging Western countries over the last decade, are significantly higher.
The Future of Nuclear Power report median estimate is $7,520/kW (overnight cost).
Overnight costs exclude financing and construction delays, measuring inherent technology complexity.

14:16 Factors Influencing US Reactor Construction Duration and Cost

US reactor construction duration increased significantly from the 1970s onwards.
Delays were often due to safety-related issues (e.g., concrete strength, rebar installation).
Construction duration is a more comparable metric than dollar cost due to accounting variations.

18:48 French Nuclear Reactor Costs and Standardization

Early French nuclear fleet costs were low due to state-owned utilities and standardization.
Flamanville (Gen III+ AP1000) is five times more expensive than previous French reactors.
Standardization in France led to lower engineering costs (15% vs. US 25%).
French electricity prices are around $70/MWh, competitive with some renewables but more expensive than onshore wind/solar.

25:56 French Reactor Reliability and Maintenance Trade-offs

French reactors have a low availability of ~74.5%, compared to US ~99%.
Unplanned outages and issues like stress corrosion cracking drive up maintenance costs.
Building cheap upfront can lead to higher long-term maintenance and potential safety risks.

31:44 Japanese Nuclear Reactor Construction: Speed vs. Safety

Japan historically controlled construction duration (average 4 years) and costs.
Many Japanese reactors were US designs built by Japanese companies.
Weak regulatory oversight (pre-Fukushima) allowed for rapid construction but led to safety issues.
TEPCO lacked internal diagrams during the Fukushima crisis, indicating poor record-keeping.

33:23 Fukushima Accident and its Long-Term Impact on Japanese Nuclear Power

Fukushima led to widespread reactor shutdowns, with many still offline 14 years later.
Operational Japanese reactors have low availability (~60%) due to ongoing safety fixes.
The effective cost per kilowatt for Japanese nuclear power has reached $44,000/kW.
Building fast and cheap proved unsustainable, leading to higher long-term costs and risks.

39:00 South Korean Nuclear Reactor Development and Scandals

Korea initially saw decreasing costs and construction durations.
Post-Fukushima, safety standards increased, driving costs and durations up.
Falsified safety certificates for reactor parts led to shutdowns and retrofits costing ~7% of construction cost.
Low reliability (~23% downtime) and rising costs have led to the industry mirroring US trends.

47:20 Indian Nuclear Power: Cost Reduction Through Scale

India has systematically reduced reactor costs, but construction durations have increased.
Financing primarily through government budget, not loans, reduces pressure for rapid construction.
Slower construction is linked to higher availability (~95.6%) and safety.
Cost reductions are primarily due to dramatically increasing reactor size, not technological learning.

51:55 Global Nuclear Cost Trends and Labor Productivity

Major nuclear countries (US, France, Japan) show rising costs; India shows decreasing costs.
Asia's higher labor productivity (fewer safety standards) contributes to lower labor costs.
A correlation exists between lower costs, more outages, and higher safety risks.
New Gen III+ reactors (Vogtle, EPR2) have outrageously high costs compared to historical data.

55:07 The Safety vs. Cost Dilemma in Nuclear Reactor Design

Cutting costs can lead to reliability issues and potential safety risks.
Designing for high safety (e.g., Gen III+ reactors) significantly increases costs.
Determining 'acceptably safe' requires comparative analysis with other energy technologies.
US safety culture is driven by liability, INPO standards, and insurance costs.

58:57 Challenges in Data Collection for Non-Market Nuclear Economies

Data from China and Russia is unreliable as plants are outside free market systems.
Cost accounting is difficult when components are produced by state-funded institutes.
China's published prices are untrustworthy; they may not know their true costs.
Loss of industrial capacity in the West may contribute to rising costs.

1:00:55 Indian Nuclear Reactor Designs and Regulatory Structure

India uses indigenous designs, evolving from modified CANDU designs.
Regulatory structure features 'gods' of specific safety areas, leading to concentrated expertise.
This unique system differs significantly from global norms.
The 'gods' are deep experts, not political appointees.

1:05:03 Comparative Regulatory Structures: US vs. France

US NRC operates independently, reviewing designs after submission.
French regulator works intimately with designers during the design phase.
The French system aims for efficient trade-offs between safety and cost.
Japanese and Korean regulators are modeled on the US NRC but historically less empowered.

1:09:04 Critique of Learning Curves and Cost Models

NREL's cost model assumes cost reduction with time, regardless of actual construction (flawed).
Wright's law (capacity-based cost models) is statistically superior to time-based models.
Early learning curve studies confused cost reduction with increased reactor size.
Cobb-Douglas production function is used to model cost inputs and their elasticities.

Summary, takeaways, and chapters were generated by AI from the video's transcript and may contain errors. The video belongs to its creator, MIT OpenCourseWare.