Episode Summary
Executive Summary: The episode argues that nuclear power should be judged less as a standalone climate savior and more as one clean-firm tool among many in a transmission-constrained, demand-growing grid. Jigger Shaw says nuclear failed mainly because of lack of standardization, lost workforce capacity, and first-of-a-kind project risk—not just regulation. He sees near-term demand from data centers and industry, with standardized large reactors and emerging microreactors both potentially viable if costs and financing are stabilized.
Main Topics: Why clean firm power matters (Priority: 5/5): Shaw explains that a wind/solar-centered grid still needs clean resources that are available on demand to fill gaps, use constrained transmission more efficiently, and serve fast-rising demand from electrification and data centers. Why U.S. nuclear has been so costly (Priority: 5/5): He attributes nuclear’s poor performance to custom one-off reactor designs, loss of skilled labor, and repeated first-of-a-kind construction rather than simply to activism or regulation. Standardization as the core fix (Priority: 5/5): Shaw argues that repeating a small number of designs at existing nuclear sites is the path to lower cost and lower risk, similar to how other industries achieve learning curves. SMRs and microreactors (Priority: 4/5): He distinguishes between SMRs as smaller versions of conventional reactors and microreactors as very small, specialized units for industrial or remote uses, but says both still need standardization and real customers. Data centers and industrial demand (Priority: 4/5): The conversation frames hyperscale data centers, industrial growth, and 24/7 clean-energy commitments as key market signals that could justify new nuclear and other clean-firm projects. Waste, reprocessing, and thorium (Priority: 3/5): Shaw downplays the scale of nuclear waste, says storage is manageable at existing sites, notes reprocessing could make sense in the future, and dismisses thorium as not worth worrying about now. Near-term deployment outlook (Priority: 4/5): He predicts the first new commercial nuclear projects will likely be in Canada before the U.S., with U.S. construction following later in the decade and first service in the early 2030s.
Key Arguments: Clean firm power is needed because wind and solar are variable and transmission is becoming the binding constraint; firm clean resources help use the grid more efficiently. The main historical failure of U.S. nuclear was not just politics or safety rules, but a collapse of learning because each plant became a bespoke design. Nuclear plants should be built in repeated series, ideally at existing sites where infrastructure and workforce already exist, to move from first-of-a-kind to nth-of-a-kind costs. SMRs are not a magic fix; they are still reactors with fixed costs that remain high unless enough units are built to establish a learning curve. Microreactors are more plausibly modular and factory-built, but they are still early and will need real orders, demonstrations, and scale to become economical. Data centers and industrial customers are shaping the market by demanding 24/7 clean power and are willing to pay above-market prices for it. Regulation is not the main bottleneck today; the larger issue is execution capacity, design stability, financing, and customer commitment. The U.S. should also consider non-grid uses for nuclear, including industrial heat, hydrogen, and behind-the-meter applications. Waste is much smaller in volume than commonly assumed and is manageable with current dry-cask and site-security practices. Reprocessing and alternative fuel cycles may eventually matter, but they are not the near-term solution to nuclear’s deployment problem.
Data Points: U.S. operating reactors: 92 - Shaw says the U.S. has 92 operating reactors, and no two are the same. Existing reactor site count: 53 locations - He says the 92 reactors are spread across about 53 sites, many originally built for four reactors each. Projected clean firm need: 2x to 3x nuclear capacity by 2050 - He cites DOE modeling and the Biden administration’s goal for nuclear expansion to meet climate targets. Wind/solar build share since 2016: ~80% - He says roughly 80% of grid megawatts built since 2016 have been wind or solar. Solar industry growth: $1 billion/year to $1 billion/day - He uses this as an example of how quickly solar scaled. Current wind/solar build rate: 47 GW/year - He says the current rate needs to rise to 100 GW/year and then 200 GW/year. Transmission buildout outlook: 1.8x - He says the U.S. may only reach about 1.8x transmission growth, below the 3x level modeling suggests is needed. Geothermal near-term potential: 3-5 GW by 2030 - He describes immediate geothermal additions as modest in the near term. Geothermal broader potential: ~90 GW - He says geothermal could eventually reach close to 90 GW, with East Coast economics lagging. Hyperscale data-center load growth: 200,000 GWh / 200 TWh - He cites BCG/Gramllich estimates of new electricity demand by 2030. Microsoft nuclear demand estimate: 10,000 MW - He mentions Microsoft alone is said to need 10 GW for ChatGPT-related needs. Vogtle Unit 4 relative cost: 30% discount vs Unit 3 - He says Vogtle Unit 4 came in at about a 30% discount to Unit 3. Potential cost for repeated Vogtle-style build: ~$130/MWh - He suggests a repeated next unit at the same site could land around this level. Current nuclear operating cost: 3.5 cents/kWh - He says fully paid-off existing nuclear plants can run at about this cost. Potential SMR operating cost: 5.5 cents/kWh - He estimates that fully paid-off SMRs would still cost more to run than existing large reactors. Microreactor power range: 1-17 MW electric - He describes microreactors as very small units, often producing heat as well as electricity. Microreactor projected cost at scale: sub-10 cents/kWh - He says this is possible if factories build about 100 units per year over time. Project Pele size: 1 MW - He says the DOD’s Project Pele is expected to be a movable one-megawatt unit. Vogtle workforce: 13,000 trained people - He says the National Building Trades trained 13,000 people for the Vogtle project. U.S. nuclear waste volume: football field 10 yards high - He uses this to describe the cumulative waste from all U.S. nuclear power plants. Coal waste comparison: 10x more than nuclear - He says coal plants have produced about 10 times more radioactive waste than nuclear plants. Hydro relicensing risk: 37 GW - He says roughly 37 GW of hydro may not be relicensed over the next 15 years without upgrades. IRA nuclear tax credit: up to 50% - He says a nuclear project can get a 30% credit, plus 10% for energy community and 10% for domestic content. Loan applications: $128 billion / 1706 applications - He cites the DOE Loan Programs Office’s active queue for transmission and grid-related projects. Ontario nuclear project: 4 BWX-300s - He says Ontario Power Group has already decided to pursue four small reactors. Ontario timing: construction in next few years; service in early 2030s - He predicts Ontario could start construction in the next few years and begin service early next decade.
Pivotal Quotes: "we don't actually live in a world of the trade-offs that we're presenting" — Jigger Shaw: He says the usual framing of nuclear versus solar-plus-storage is misleading because nuclear retirements have often been replaced by gas. "nuclear is really a trust exercise" — Jigger Shaw: He uses this to argue that standardization, repetition, and predictable delivery matter more than hype around new reactor branding. "This is not a nuclear problem. This is a can America do big things problem." — Jigger Shaw: He reframes Vogtle and other megaproject failures as a broader U.S. industrial-capability issue.
Implications: Nuclear’s future depends less on slogans than on execution: standardized designs, real customers, financing backstops, and workforce rebuilds. If data-center demand holds, nuclear may return as one option in a broader clean-firm portfolio.