Centrus and TerraPower Lead US Push to Cut Russian Nuclear Fuel Monopoly

Bloomberg Originals

The projected tripling of nuclear power hinges on scaling domestic HALEU enrichment before supply bottlenecks derail new small modular reactors.

AI-assisted summary of Making New Nuclear Fuel for an Atomic Renaissance | Bloomberg Primer — original source. Check the original for full context.

Source published: 2026-05-21

Detailed notes

## Background Published on May 13, 2026, by Bloomberg Originals on YouTube, this video is a Bloomberg Primer documentary focusing on the global race to manufacture new nuclear fuel amid an atomic resurgence. The page description highlights how the nuclear fuel supply chain is being radically reshaped by rising energy demand, next-generation reactor technology, intense lobbying, and critical geopolitical friction.

## TL;DR The documentary explores the complex challenges of rebuilding the Western nuclear supply chain to support a new generation of small modular reactors (SMRs). Driven by massive Big Tech investments to power AI data centers and a cultural shift led by 'Nukefluencers,' the industry is racing to produce High-Assay Low-Enriched Uranium (HALEU) to break Russia's near-monopoly on enrichment capacity. However, this atomic renaissance faces steep hurdles, including a $1.5 billion, 20-year timeline for greenfield uranium mining in Canada, a severe 'chicken-and-egg' dilemma between fuel enrichers and reactor developers, and unresolved environmental concerns regarding increased radioactive waste per unit of electricity produced by SMRs.

## Atmosphere & style The video feels like a high-stakes, planet-spanning corporate thriller that blends cutting-edge technological optimism with stark geopolitical reality. It carries an aura of tight industrial secrecy and urgency, moving from the heavily guarded, high-security 'King Kong doors' of an Ohio enrichment facility to the piercing, high-pitched alarms of radioactive discovery in the remote Canadian wilderness. The narrative balances the sleek, modern enthusiasm of tech-backed advocates with the sobering financial warnings of veteran nuclear engineers, leaving the viewer with a sense of precariously balanced momentum where massive fortunes and the future of global energy hang in the balance.

## Key takeaways - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) The resurging interest in nuclear energy is driving a massive overhaul of global uranium supply chains, from mining to enrichment, to power next-generation reactors. - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) High-Assay Low-Enriched Uranium (HALEU) provides extreme energy density, with just three tablespoons satisfying an average person's lifetime electricity needs. - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) Sanctions against Russia, which controls roughly 50% of world enrichment capacity, have forced Western nations to rapidly build out independent domestic supply chains. - [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) Social media 'Nukefluencers' and massive financial backing from tech industry leaders are driving a dramatic cultural and economic shift in favor of nuclear power. - [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) Explosive energy demands from AI data centers have reshaped the economics of nuclear energy, leading to lifespan extensions and restarts for older plants. - [07:44](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=464s) The multi-step nuclear fuel cycle transforms raw ore containing only 0.7% fissile U-235 into enriched ceramic fuel pellets. - [09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) Developing new greenfield uranium mines like NextGen's Arrow deposit requires upwards of $1.5 billion and more than two decades from discovery to production. - [11:55](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=715s) Centrus Energy operates the only licensed U.S. facility producing HALEU, but scaling up faces a six-to-seven-year timeline and DOE funding support. - [11:55](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=715s) A 'chicken-and-egg' dilemma exists between enrichers needing guaranteed reactor demand and SMR developers needing guaranteed fuel supplies to build. - [15:31](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=931s) TerraPower is developing Generation 4 SMRs using sodium cooling and molten-salt thermal storage to provide flexible baseload power. - [15:31](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=931s) Small modular reactors face economic challenges because smaller capacities offer lower economies of scale relative to required concrete, steel, and oversight. - [19:23](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1163s) HALEU supply bottlenecks are directly causing project delays, pushing back launch schedules for flagship advanced reactors like TerraPower's Natrium. - [20:09](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1209s) Unlike past projects funded via ratepayer surcharges, modern nuclear developments rely heavily on private tech capital and government grants to manage upfront risk. - [21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) Spent nuclear fuel management remains unresolved, with critics noting that many SMRs could actually generate more waste per unit of electricity than traditional plants. - [22:11](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1331s) The ultimate success of the nuclear renaissance depends on rebuilding lost manufacturing skillsets and competing against cheap natural gas and renewable battery storage.

## Detailed walkthrough Here is a detailed, faithful walkthrough of the provided transcript segment.

---

[00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) Ohio's Centrifuges and the Promise of HALEU

In a small Ohio town, hidden behind layers of barbed-wire fencing and high-voltage warning signs, lies a vast industrial footprint containing roughly 11,000 holes in the ground. At this site, massive centrifuges are actively churning out High-Assay Low-Enriched Uranium (HALEU), a potent new class of nuclear fuel. The sheer energy density of HALEU is remarkable: just three tablespoons can supply a typical person’s entire electricity needs for a lifetime. The private company operating the site currently has 16 centrifuges built, but its ultimate goal is to fill all 11,000 holes to scale up production.

This ambitious expansion reflects an industry surging back after years on life support, propelled by what many are calling a "nuclear renaissance." Big Tech companies are aggressively investing in nuclear energy to meet soaring electricity demands, while political leaders from President Donald Trump to international figures advocate for quadrupling national nuclear capacity and constructing reactors rapidly. As global powers race for energy dominance—with China currently outpacing the United States in reactor deployment by at least a decade—HALEU has emerged as the essential fuel for next-generation small modular reactors (SMRs). However, securing the nuclear supply chain remains a complex challenge. Uranium must undergo intricate enrichment, and with global uranium demand projected to more than double by 2040, the market faces significant geopolitical friction. Russia’s state corporation Rosatom currently controls nearly 50% of world enrichment capacity, forcing Western nations—bound by sanctions following the war in Ukraine—to completely overhaul their supply chains from mining to final fuel fabrication.

- **Energy Density:** 3 tablespoons of HALEU provides a lifetime of electricity for an average individual. - **Facility Capacity:** 16 active centrifuges out of 11,000 total planned installations in Piketon, Ohio. - **Market Demand:** Global uranium demand is projected to more than double by the year 2040. - **Geopolitical Concentration:** Russia's Rosatom controls roughly 50% of global uranium enrichment capacity.

---

[02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) "Nukefluencers," Tech Money, and a Mindset Shift

Inside a power plant simulator for Unit 1 at California’s Diablo Canyon, a trainee sits at the control joystick attempting to balance control rod positions. The operator-in-training is Isabelle Boemeke (known online as "Isodope"), a Brazilian fashion model who became a prominent nuclear influencer on TikTok. Boemeke started making abstract, one-minute videos to make nuclear energy engaging and accessible to younger audiences. She represents a broader movement of "Nukefluencers" using short-form video to promote atomic power, though critics from environmental groups like Greenpeace argue that snappy social media campaigns gloss over the persistent dangers and waste hazards of nuclear energy.

Boemeke’s influence extends far beyond social media advocacy. Married to Airbnb co-founder and former Chief Design Officer Joe Gebbia, the couple has donated millions to nuclear causes, and Boemeke has actively provided input on draft executive orders aimed at streamlining nuclear regulation. This wave of tech-backed advocacy represents a dramatic reversal from previous decades, when the dominant narrative in American nuclear power was defined by plant closures such as San Onofre and Crystal River. Operating nuclear plants has historically carried staggering financial risks; for example, Units 3 and 4 at Southern Company's Vogtle plant in Georgia came online seven years behind schedule and ran over $16 billion over budget. However, the explosive growth of windowless AI data centers requiring massive, continuous baseload power has fundamentally reshaped the economics, leading to extended lifespans for existing facilities like Diablo Canyon (now approved to operate until at least 2030) and the planned restart of Three Mile Island Unit 1 under the name Crane Clean Energy Center.

- **Global Investment:** Estimated global investment in nuclear energy since 2020 has reached approximately $300 billion. - **Vogtle Overruns:** Georgia's Vogtle Plant expansion finished 7 years late and ran more than $16 billion over budget. - **Diablo Canyon Lifespan:** Originally slated for decommissioning, the California plant's operational license was extended to at least 2030. - **Three Mile Island Restart:** Constellation Energy is working to reopen Unit 1 as the Crane Clean Energy Center.

---

[07:44](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=464s) The Nuclear Fuel Cycle: From Raw Ore to Fission

Transforming a natural lump of uranium ore into clean electricity is an intricate, multi-step process governed by physics and chemistry. Raw uranium ore contains three isotopes, though U-234 exists in trace amounts under 0.01% and is negligible. Non-fissile U-238 accounts for 99.3% of raw ore, while the vital, fissile U-235 isotope makes up only about 0.7%. To turn this gamma-emitting rock into reactor fuel, the ore is first milled and refined into "yellowcake"—a bright yellow, powdery substance resembling cake mix.

Next, the yellowcake is transported to a conversion facility where it is converted into a gas called uranium hexafluoride ($\text{UF}_6$). To enrich the fuel, this gas is fed into rapidly spinning centrifuges, where centrifugal forces cause the heavier U-238 atoms to push outward while the lighter U-235 atoms concentrate near the center. Standard Low-Enriched Uranium (LEU) used in traditional commercial reactors requires a U-235 concentration of 3% to 5%. If enrichment continues past commercial thresholds up to 90% or higher, the material becomes weapons-grade, which is why centrifuge technology is subject to strict international regulations. Once enriched, the uranium hexafluoride gas is chemically converted into metal, pressed into ceramic pellets roughly the size of a gummy bear, stacked into metal fuel rods, and loaded into reactor cores. Inside the reactor, U-235 atoms undergo nuclear fission, splitting apart to create extreme heat that boils water into steam, spinning a turbine to generate electricity.

- **Isotopic Breakdown:** Natural uranium consists of 99.3% U-238 and only 0.7% fissile U-235. - **Enrichment Levels:** Standard LEU requires 3%–5% U-235 enrichment; weapons-grade material requires 90% or higher. - **Fuel Form Factor:** Enriched uranium is manufactured into solid pellets roughly the size of a gummy bear before being loaded into fuel rods.

---

[09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) Upstream Mining: Mining the Arrow Deposit at Rook One

For mining companies attempting to capitalize on the nuclear resurgence, the fuel cycle begins deep underground. At the Rook One project in Saskatchewan’s Athabasca Basin, NextGen Energy is targeting the massive Arrow deposit. Handheld radiation detectors at the site trigger high-pitched alarms as counts register above 65,000 counts per second, signaling exceptionally high-grade uranium mineralization. NextGen founder and CEO Leigh Curyer highlights that the Arrow deposit represents the world’s largest and highest-grade undeveloped uranium project, projected to produce approximately 20% to 25% of the world’s mined uranium supply once operational.

Located 700 kilometers north-northwest of Saskatoon and 155 kilometers north of La Loche and the Clearwater River Dene Nation, developing a greenfield mine like Rook One demands vast capital and patience. The project carries an estimated capital cost of $1.5 billion. Beyond the financial burden, navigating regulatory permits, environmental reviews, and formal agreements with local Indigenous communities (who represent 13% of Athabasca County) means the timeline from initial discovery to first commercial output can easily span over 20 years. Despite these lengthy lead times, NextGen is already executing off-take agreements with utility companies across North America, Europe, Asia, and the Middle East.

- **Mine Location:** Rook One is located in the Athabasca Basin, 700 km NNW of Saskatoon, Saskatchewan. - **Production Projection:** Arrow deposit is expected to supply 20% to 25% of the global mined uranium supply. - **Capital Cost & Timeline:** Rook One carries a $1.5 billion construction cost and a 20+ year development timeline from discovery to production. - **Indigenous Demographics:** Local Indigenous communities make up approximately 13% of Athabasca County.

---

[11:55](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=715s) Centrus, HALEU, and the Chicken-and-Egg Dilemma

Once uranium is mined and converted, enrichment presents the next major industrial hurdle. At Centrus Energy's facility in Piketon, Ohio, operators drive transport vehicles through massive security barriers known internally as "King Kong doors." Matt Snider, a former Nuclear Navy mechanic with 15 years at Centrus, oversees operations at the facility, which is currently the only licensed U.S. producer of HALEU. Centrus operates 16 demonstration centrifuges alongside thousands of empty floor mounts, planning to expand to 11,000 centrifuges to meet future market demand.

HALEU is enriched up to 20% U-235, making it far more potent than standard 3% to 5% LEU. Because Russia and China currently hold a virtual monopoly on commercial HALEU supply, Western governments are stepping in; the U.S. Department of Energy awarded $900 million to Centrus and two other fuel developers to establish domestic production. However, scaling up will take six to seven years even with full funding. Furthermore, strict national security classifications prevent plant workers from discussing specific enrichment operational timelines, repeatedly responding with "I can't comment" or "That's classified." This creates a severe "chicken-and-egg" economic dilemma: enrichers cannot justify spending billions to scale up HALEU production without guaranteed demand from commercial SMRs, while SMR developers cannot build reactors without a guaranteed, steady supply of HALEU fuel.

- **Enrichment Potency:** HALEU is enriched to roughly 20% U-235, compared to 3%–5% for traditional LEU. - **Federal Backing:** The U.S. Department of Energy allocated $900 million toward domestic HALEU fuel production. - **Scaling Timeline:** Fully building out the 11,000-centrifuge plant at Centrus is projected to take 6 to 7 years. - **International Competition:** Commercial-scale HALEU production is currently limited to state-backed entities in Russia and China.

---

[15:31](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=931s) TerraPower, Reactor Generations, and the Economics of SMRs

Working to break the fuel-and-reactor deadlock is TerraPower, a nuclear technology company founded by Bill Gates in 2008. Headquartered outside Seattle, TerraPower is building its flagship SMR, known as the Natrium reactor, in Kemmerer, Wyoming. TerraPower CEO Chris Levesque—a 30-year nuclear industry veteran who began his career in submarine engine rooms—notes that while the existing fleet of ~90 U.S. reactors and ~400 global reactors were designed in a "slide-rule and paper drawing" era, modern SMRs are engineered entirely within advanced computing environments. The Natrium design produces roughly one-third the electricity of a conventional large-scale plant, but it features an integrated molten-salt energy storage system ("Energy Island") that acts as a thermal battery to boost output during peak demand.

The evolution of commercial nuclear technology spans several distinct generations: * **Generation 1 (1950s–1960s):** Early prototype reactors that were often small, unreliable, and lacked adequate containment structures or automated controls. * **Generation 2 (1970s–1990s):** Commercial-scale plants featuring automated emergency shutdowns; however, high-profile accidents like Three Mile Island and Chernobyl highlighted ongoing risks from equipment failure and human error. * **Generation 3 / 3+ (1990s–Present):** Simplified, highly efficient designs featuring passive safety systems; Gen 3+ SMRs are essentially scaled-down versions of these light-water reactors. * **Generation 4:** Advanced designs utilizing novel coolants (such as liquid sodium or molten salt) and advanced fuels like HALEU.

While proponents argue that modular, factory-built SMRs will eventually reduce costs through standardized assembly, critics and economists warn that early SMRs face significant economic hurdles. Because large reactors benefit from economies of scale, smaller reactors produce less energy per unit of concrete, steel, and operational oversight. Distributed across more individual sites, SMRs require more regulatory oversight, specialized security, and operating personnel, threatening to drive up power costs on a per-kilowatt-hour basis.

- **TerraPower Founders & Location:** Founded by Bill Gates in 2008; flagship Natrium SMR is under construction in Kemmerer, Wyoming. - **Fleet Scale:** Worldwide operating fleet consists of approximately 400 reactors, with over 90 located in the United States. - **Natrium Output:** Generates roughly one-third the electrical output of a traditional large light-water reactor, supplemented by molten-salt thermal storage.

[19:23](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1163s) Small Modular Reactors and the HALEU Supply Bottleneck

While the potential for small modular reactors (SMRs) is immense, the field is undergoing a aggressive survival-of-the-fittest competition. The number of active SMR designs globally has ballooned to around 127 today, up from roughly 80 just a few years ago, but industry experts caution that "they won't all survive nuclear selection." A primary operational bottleneck for these designs is securing High-Assay Low-Enriched Uranium (HALEU). For example, TerraPower's flagship Natrium reactor was originally scheduled to go online in 2028, but faced a two-year delay explicitly due to a lack of available HALEU.

To overcome this roadblock and secure fuel for its first core load, TerraPower identified an alternative partner: ASP Isotopes, a NASDAQ-listed U.S. firm that will carry out enrichment operations in South Africa. Funding for these capital-intensive projects remains massive; TerraPower reports receiving $2 billion from the U.S. Department of Energy alongside raising $1.4 billion in private capital.

* **SMR Market Expansion:** Global SMR design proposals have expanded from ~80 to 127 designs. * **TerraPower Timeline:** Natrium reactor launch delayed by two years (from 2028) due to global HALEU shortages. * **Alternative Fuel Solution:** Initial HALEU supply for Natrium sourced via South Africa-based operations of U.S.-listed ASP Isotopes. * **Capital Investment:** TerraPower has secured $2 billion from the U.S. DOE and $1.4 billion in private funding.

[20:09](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1209s) Funding the Renaissance: Ratepayers vs. Private Tech Capital

The financial model powering nuclear development is undergoing a fundamental shift compared to previous decades. Recalling childhood memories in New Hampshire during the construction of the Seabrook Nuclear Station, a commentator recalls how working-class families living paycheck to paycheck were burdened with a "Seabrook Surcharge" directly on their electricity bills to fund capital expenditures. That legacy model—where ratepayers directly subsidize utility reactor construction—is no longer viable in today's U.S. market.

Instead, the new paradigm relies heavily on private capital, particularly massive tech industry investments, to shoulder the high upfront risks and costs. This shift relieves taxpayers of direct capital burdens, though it simultaneously shifts controlling influence over the nation's future energy landscape into the hands of private enterprises.

* **Geographic Expansion:** TerraPower has established preliminary agreements with utilities in Utah, Kansas, and the United Kingdom, alongside its flagship project in Wyoming. * **Broader Momentum:** Multiple SMR concepts in both the U.S. and Europe are advancing through key commercial development stages.

[21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) The Environmental Dilemma of Nuclear Waste

Despite the hype surrounding advanced reactors, a critical challenge remains unaddressed: long-term radioactive waste management. Deploying hundreds of small reactors around the globe inherently leads to a geographic proliferation of spent nuclear fuel stored in temporary facilities, with no unified long-term repository operational. Outside typical nuclear plants, spent fuel sits in massive steel and concrete casks—materials that remain dangerously radioactive for thousands of years.

Adding to the complexity, recent technical evaluations suggest that due to their compact core designs, many SMRs may actually generate *more* nuclear waste per unit of electricity produced than traditional, large-scale commercial reactors. While some SMR vendors counter that future designs will eventually recycle up to 96% of spent fuel, any accidental exposure poses severe risks to ecosystems, water supplies, the food chain, and public health.

> "If you go to any nuclear power plant, if you go out back, you're gonna see these giant steel and concrete casks, and inside them is the spent fuel rods and those things are deadly. They'll be deadly for a thousand years."

[22:11](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1331s) The Competitive Horizon and Uncertain Future

For a full nuclear renaissance to succeed, industrial nations in Europe and the U.S. must rebuild specialized engineering and manufacturing skillsets that have "been atrophying over the past decades." However, whether nuclear's share of U.S. power generation can increase from its current ~20% baseline to 30% or 40% isn't determined solely by nuclear tech. It depends heavily on rival energy markets—specifically whether natural gas frackers yield market share, and how rapidly cheap battery storage scales to complement wind and solar power.

While industry insiders remain confident that a broader variety of advanced nuclear technologies will be operational by the early 2030s, the overall nuclear revival remains precariously balanced. Real momentum, money, and social media attention are currently flowing into the space, but nuclear projects remain notoriously slow to build and extraordinarily expensive. If the newly forming, fragile uranium and enrichment supply chains fail, the entire nuclear renaissance risks collapsing under its own weight.

## Notable quotes > [11:15](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=675s) I can't comment

> [11:15](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=675s) That's classified.

> [19:23](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1163s) they won't all survive nuclear selection.

> [21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) If you go to any nuclear power plant, if you go out back, you're gonna see these giant steel and concrete casks, and inside them is the spent fuel rods and those things are deadly. They'll be deadly for a thousand years.

## Mentioned Centrus Energy, High-Assay Low-Enriched Uranium (HALEU), Donald Trump, Rosatom, Isabelle Boemeke, Isodope, TikTok, Greenpeace, Joe Gebbia, Airbnb, San Onofre, Crystal River, Southern Company, Vogtle plant, Diablo Canyon, Three Mile Island, Crane Clean Energy Center, Constellation Energy, Rook One project, Athabasca Basin, NextGen Energy, Arrow deposit, Leigh Curyer, Clearwater River Dene Nation, Matt Snider, U.S. Department of Energy, TerraPower, Bill Gates, Natrium reactor, Chris Levesque, Chernobyl, ASP Isotopes, Seabrook Nuclear Station

[Source](https://www.youtube.com/watch?v=rkCg7wR3_hI)

中文简报

Centrus与TerraPower领衔美国打破俄罗斯核燃料垄断核电规模预计增长两倍的前景，取决于能否在供应瓶颈阻碍新型小型模块化反应堆前扩大本土HALEU浓缩规模。该纪录片探讨了重建西方核供应链以支持新一代小型模块化反应堆（SMR）所面临的复杂挑战。在科技巨头为AI数据中心供电提供巨额投资，以及由“核能网红”（Nukefluencers）引领的文化转变推动下，整个行业正竞相生产高丰度低浓缩铀（HALEU），以打破俄罗斯对浓缩产能的近乎垄断地位。然而，这场核能复兴面临重重障碍，包括加拿大新建铀矿开采耗资15亿美元且长达20年的时间周期、燃料浓缩商与反应堆开发商之间严重的“鸡生蛋还是蛋生鸡”困境，以及关于SMR单位发电量产生更多放射性废物的未决环境隐忧。高丰度低浓缩铀（HALEU）具备极高的能量密度，仅三汤匙即可满足一人一生的用电需求，但由于俄罗斯 Rosatom 控制了全球近 50% 的浓缩产能，面对 2040 年预计翻倍的铀需求，西方国家正加速重建本土供应链。 (0:03) AI 数据中心爆发式的基荷电力需求与科技行业领袖的巨额投资正在重塑核电经济格局，不仅推动了加州 Diablo Canyon 电厂延寿至至少 2030 年，还促成了三里岛核电站 1 号机组的重启计划。 (2:54) 核燃料浓缩面临严重的“先有鸡还是先有蛋”困局：Centrus Energy 虽然是全美唯一获得 HALEU 生产许可的企业，但全面扩建 11,000 台离心机需耗时 6 至 7 年，在下游小型反应堆需求确立前难以孤立推进规模化量产。 (11:55) 全球小型模块化反应堆（SMR）设计已激增至约 127 种，但比尔·盖茨创立的 TerraPower 旗下 Natrium 反应堆就因 HALEU 燃料短缺而推迟两年上线，且 SMR 单位发电量面临更高核废料产出的潜在环保争议。 (19:23)

Reading summary

A tech-backed push to deploy next-generation nuclear reactors faces severe fuel and supply chain bottlenecks, as Western nations attempt to break Russia's near-monopoly on high-assay low-enriched uranium (HALEU). Tech Demand and the Push to Revive Nuclear (2:54) Spurred by the immense 24/7 baseload power needs of AI data centers, nuclear energy has seen roughly $300 billion in global investment since 2020. This financial influx marks a sharp break from decades of plant closures and financial disasters, such as Georgia's Vogtle Units 3 and 4 running seven years late and more than $16 billion over budget. Instead of saddling utility ratepayers with construction surcharges—as occurred during legacy projects like New Hampshire's Seabrook Station—private tech capital and affluent advocates are now financing new atomic ventures. This momentum has pushed regulators to extend the life of existing plants, such as California's Diablo Canyon through at least 2030, and motivated Constellation Energy to plan the restart of Three Mile Island Unit 1 as the Crane Clean Energy Center. The Fuel Cycle and the HALEU Chokepoint (7:44) Natural uranium ore contains only about 0.7% fissile U-235, which must be converted into uranium hexafluoride gas and spun through centrifuges to produce usable fuel. While conventional commercial reactors run on low-enriched uranium (LEU) at 3% to 5% U-235, advanced small modular reactors (SMRs) rely on HALEU, which is enriched up to roughly 20%. HALEU packs immense energy density—three tablespoons can meet an average person's lifetime electricity demand—making it vital for Generation 4 reactor designs. However, Russia's state-owned Rosatom controls nearly 50% of global enrichment capacity, and commercial HALEU supplies remain effectively monopolized by Russia and China. This geopolitical vulnerability has forced Western governments to attempt a rapid rebuild of their domestic enrichment infrastructure. Upstream Mines and Centrifuge Scaling Delays (9:50) Rebuilding the fuel pipeline requires decades-long lead times across every stage of production. On the mining front, NextGen Energy is advancing the Arrow deposit at the Rook One project in Saskatchewan's Athabasca Basin, aiming to supply 20% to 25% of global mined uranium. However, the greenfield project faces a $1.5 billion capital cost and a 20-plus-year timeline from discovery to output due to regulatory hurdles and consultations with local Indigenous communities. Downstream, Centrus Energy operates the only licensed U.S. HALEU facility in Piketon, Ohio, but it currently has just 16 active centrifuges out of a planned 11,000. Even with $900 million in federal backing from the U.S. Department of Energy, scaling the facility is projected to take six to seven years. The Reactor Bottleneck and SMR Economic Dilemmas (15:31) Advanced reactor developers are trapped in a chicken-and-egg dilemma: enrichers hesitate to spend billions scaling fuel without guaranteed demand, while developers cannot deploy reactors without reliable fuel supplies. Bill Gates–backed TerraPower, which is constructing its Natrium SMR in Kemmerer, Wyoming with $2 billion in U.S. Department of Energy funding and $1.4 billion in private capital, was forced to delay its planned 2028 launch by two years due to HALEU shortages, eventually partnering with ASP Isotopes in South Africa for initial fuel. At the same time, the field of roughly 127 competing SMR designs faces harsh economics. Unlike large plants that benefit from economies of scale, smaller reactors produce less electricity per unit of concrete and steel, requiring more specialized security, site licensing, and operational oversight per kilowatt-hour. Unresolved Waste and External Market Competition (21:03) The rapid deployment of distributed SMRs threatens to worsen long-term radioactive waste management. Because no unified permanent repository exists, spent fuel remains stored on site in steel and concrete casks that stay hazardous for thousands of years. Furthermore, technical assessments indicate that compact SMR designs may produce more radioactive waste per unit of electricity than traditional reactors, despite industry hopes for future fuel-recycling systems. Beyond operational hurdles, nuclear power's ability to expand beyond its roughly 20% baseline share of U.S. generation remains constrained by rival technologies, particularly cheap natural gas from domestic fracking and the falling costs of battery storage paired with renewables. While tech capital and social media advocacy have generated substantial momentum, the atomic renaissance rests on a precarious foundation. If the West cannot successfully rebuild its atrophied industrial manufacturing and resolve its severe enrichment bottlenecks, high capital costs and fuel shortages could stall the next generation of nuclear power before it achieves commercial scale.

精炼阅读摘要

彭博节目指出，在科技巨头电力需求激增与政界支持的推动下，全球正掀起核能复兴浪潮，但第四代小型模块化反应堆（SMR）的商业化落地，正面临高丰度低浓缩铀（HALEU）浓缩瓶颈、高昂建造成本与长期核废料处理等多重制约。算力需求与民间资本逆转核能叙事 (2:54) 在经历数十年因成本失控与安全顾虑导致的关停潮后，美国核电产业迎来了态度逆转。例如佐治亚州Vogtle核电站3号和4号机组的扩建工程延期7年交付，超支逾160亿美元；而加州Diablo Canyon核电站原本计划退役，如今运营许可已延至至少2030年，Constellation Energy亦计划重启三里岛1号机组。自2020年以来，全球核能领域吸引了约3000亿美元投资，这股动能主要由短视频领域的“核网红”（如Isabelle Boemeke）和科技行业庞大资本驱动，背后的关键驱动力是无窗AI数据中心对全天候基荷电力的强劲需求。与过去依赖向普通用电户加征附加费（如新罕布什尔州Seabrook核电站建设时期）的公用事业集资模式不同，当前核电复兴严重依重科技资本与私人投资承担前期高风险。包括比尔·盖茨创立的TerraPower等项目，虽然减轻了纳税人的直接资本负担，但也使私营企业对国家未来能源格局拥有更大的话语权。从矿石到HALEU的产业链与地缘困境 (7:44) 天然铀矿石中仅含有约0.7%的可裂变铀-235同位素，其余99.3%为不可裂变的铀-238。制造燃料需要先将矿石加工成“黄饼”，再转化为六氟化铀气体（UF6），并通过高速离心机分离浓缩。传统商用轻水反应堆仅需3%至5%浓度的低浓缩铀（LEU），而下一代SMR则需要浓度达20%的高丰度低浓缩铀（HALEU），其能量密度极高，仅三汤匙即可满足一人一生的用电需求。全球铀需求预计到2040年将增长一倍以上，但上游供应链建设周期漫长，如加拿大NextGen Energy在萨斯喀彻温省的Rook One项目虽预计可供应全球20%至25%的开采铀，但面临15亿美元资本开支及长达20年以上的开发与原住民审查周期。更严峻的瓶颈在于下游浓缩环节。目前俄罗斯国家原子能公司（Rosatom）控制着全球约50%的浓缩能力，且商业级HALEU几乎由中俄两国的国有实体垄断。乌克兰危机引发西方制裁后，重塑供应链变得极为紧迫。美国能源部已向Centrus Energy等企业拨款9亿美元以建立本土生产能力，但Centrus位于俄亥俄州Piketon的设施目前仅有16台示范离心机在运行，要扩建至规划的11000台规模预计需要6至7年时间。第四代反应堆技术竞赛与先鸡先蛋困境 (15:31) 全球SMR设计方案已从早期的约80种激增至目前的127种，行业内部正面临残酷的淘汰赛。以比尔·盖茨旗下的TerraPower为例，其在怀俄明州Kemmerer开发的旗舰Natrium反应堆属于第四代钠冷快堆设计，发电量约为传统大型轻水反应堆的三分之一，并集成了熔盐储能系统以应对用电高峰。然而，由于全球HALEU供应断供，Natrium反应堆的投运时间直接推迟了两年，不得不转而寻求与在南非运营的纳斯达克上市公司ASP Isotopes合作以获取初始燃料。这构成了典型的“先有鸡还是先有蛋”经济死结：商用SMR没有稳定可靠的燃料供给便无法破土建设，而离心浓缩企业在缺乏确定的大规模反应堆订单前，也难以承担数十亿美元的扩产投入。此外，经济学家和批评人士指出，大型传统反应堆享有规模经济优势，而SMR由于单位混凝土与钢材产生的电能较低，加之分散化布局需要更多监管、安保和运营团队，可能在商业化初期推高每度电的度电成本。核废料扩散疑虑与外部能源竞争 (21:03) 尽管支持者强调下一代核反应堆在安全系统上的进步，但长寿命放射性废料的妥善处理仍是未解难题。在全球尚未建成统一长期深地质处置库的情况下，高放射性废燃料只能储存在核电站户外的钢制与混凝土干式贮存罐中，危害可持续数千年。部分技术评估进一步警告，由于堆芯设计更为紧凑，许多小型模块化反应堆在每发单位电量时产生的核废料总量，可能反而高于传统大型商业反应堆。核能能否将其在美国总发电量中的占比从目前的约20%提升至30%或40%，不仅取决于其自身的工程制造能力能否在数十年萎缩后重新建立，还受到外部替代能源价格的直接挤压。页岩气水力压裂开采能否持续让出市场份额，以及廉价电池储能技术能否迅速成熟以补足风能与太阳能的间歇性，都将决定核电在经济上的生存空间。先进核能技术虽然获得了前所未有的资本注入与社会关注，但新一代反应堆高度依赖脆弱且受地缘政治制约的HALEU浓缩供应链。如果数年内无法解决燃料短缺与建造成本问题，在面对水力压裂天然气和可再生能源储能系统的竞争时，这场核能复兴或将面临难以兑现预期的系统性风险。

Detailed notes — zh

## Background 该视频由彭博社旗下栏目 Bloomberg Originals 于 2026 年 5 月 21 日发布在 YouTube 平台上。视频属于《Bloomberg Primer》系列，聚焦探讨核电试图卷土重来之际所面临的核燃料短缺危机，深入剖析了地缘政治、游说团体、激增的能源需求以及新型反应堆技术如何正在重塑从铀矿开采到浓缩加工的全球核供应链。 ## TL;DR 在人工智能数据中心巨额电力缺口与零碳转型的双重驱动下，核电产业正迎来一场由硅谷巨头、民营资本和政策力量合力推进的“核能复兴”。然而，支撑下一代小型模块化反应堆（SMR）的核心命脉——高纯度低浓缩铀（HALEU）——却深陷严峻的供应链瓶颈与地缘断供风险，目前全球产能仍被中俄高度垄断。西方世界正试图通过开采极地超级铀矿床、在俄亥俄州扩建离心机基地以及比尔·盖茨创立的 TerraPower 等前沿堆型来重构本土核燃料循环，但整个行业依然困在燃料与堆型“先有鸡还是先有蛋”的产业胶着、超长建设周期、高昂改造成本以及长期核废料处理的深刻危机之中。 ## Atmosphere & style 全片充斥着一种宏大野心与残酷工业现实交织的冷峻张力。镜头在俄亥俄州密布上万个空基孔的幽暗离心机厂房、加拿大极北苦寒荒原上读数爆表的辐射探测器，以及加州核电模拟室内时尚网红的谈笑风生之间剧烈切换。叙事节奏紧凑洗练且兼具技术严谨性与商业警醒感，既有目睹人类尖端能源野心时的震撼与狂热，又在冷酷的数据与千年不灭的核废料阴影中，向观众传递出通往能源独立之路极度脆弱而迷茫的警示之感。 ## Bulletin - [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) 核电产业正迎来新一轮复兴，AI 算力中心对零碳基荷电力的渴求扭转了核反应堆退役颓势，自 2020 年以来全球核电领域的投资规模已飙升至约 3,000 亿美元。 - [11:37](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=697s) 新一代反应堆所需的 HALEU（高纯度低浓缩铀）目前仅由俄罗斯和中国实现商业规模化生产，西方国家在对俄制裁后面临极为脆弱的燃料断供风险与漫长的重构周期。 - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 美国唯一持牌生产 HALEU 的 Centrus 基地仅运转 16 台离心机，距离规划的 11,000 台目标仍需 6 至 7 年建设期，折射出整个核供应链前端严重的供给瓶颈。 - [19:36](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1176s) 比尔·盖茨创立的 TerraPower 因燃料短缺将首座 Natrium 反应堆并网推迟至 2030 年，其资金转向由能源部与科技/风投资本共同承担，彻底打破了传统纳税人兜底的融资范式。 ## Key takeaways - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 高纯度低浓缩铀（HALEU）拥有极高能量密度，仅需 3 汤匙即可满足一人一生的用电需求，是推动小型模块化反应堆（SMR）的核心命脉。 - [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 俄罗斯国家原子能公司（Rosatom）掌控着全球近 50% 的浓缩铀市场份额，地缘冲突迫使西方必须从采矿到加工全面重塑核供应链。 - [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) AI 算力中心的爆发性电力需求彻底逆转了核电退役潮，恶魔谷核电站成功延寿至 2030 年，三里岛部分机组也被重启并更名为 Crane 清洁能源中心。 - [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) 新型核电的公众认知正通过社交媒体被重塑，但反核团体仍对废料与事故风险提出尖锐批评，科技资本深度参与了相关监管政策的推动。 - [07:18](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=438s) 天然铀矿石中可裂变的铀-235 仅占 0.7%，必须经历黄饼加工、化学气化转化与超高速离心分离，才能浓缩为反应堆可用的燃料颗粒。 - [09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) 位于加拿大的 Rook I 项目蕴藏世界级高品位铀矿床，投产后预计供应全球矿山铀开采量的 20% 至 25%，但从勘探到商业交付周期长达 20 年以上。 - [11:37](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=697s) Centrus Energy 规划部署 11,000 台离心机以打破中俄在 HALEU 商业供应上的垄断，但彻底建成仍需数年时间，且工艺细节受防扩散严格涉密保护。 - [14:12](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=852s) HALEU 供给端与 SMR 反应堆建设端陷入典型的“鸡与蛋”循环困境：燃料商因缺乏商用堆不敢扩产，开发商因缺乏本土燃料不敢贸然开工。 - [15:44](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=944s) TerraPower 旗下由盖茨投资的 Natrium 反应堆采用液态钠冷却并配备熔盐储能系统，可在用电高峰期作为热力电池动态调峰并网。 - [17:01](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1021s) 核反应堆经历四代演进，现存超百种 SMR 方案正面临激烈市场洗牌，业界对其能否真正超越传统大型堆的度电规模经济效益存在分歧。 - [19:36](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1176s) 核电项目出资模式发生深刻变革，告别了由终端用电户附加费兜底的公用事业机制，转向由科技巨头与私募风投资本分担前期巨额风险。 - [21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) 小型模块化反应堆的分散化布局将产生更多散落各地的核废料，人类目前仍缺乏长期深地处置方案，部分小型堆单位发电产生的废料体积甚至更高。 - [23:10](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1390s) 核电未来在电网中的份额不仅受自身供应链脆弱性制约，还将直面廉价页岩气与大规模储能电池等外部替代技术的动态竞争挑战。 ## Detailed walkthrough ### [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 俄亥俄州地下的万孔之谜与 HALEU 带来的“核能复兴” 在俄亥俄州一个偏远小镇的铁丝网深处，警示危险的标语随处可见。这里的地面上密布着大约 11,000 个预留基孔，伴随着机器的轰鸣，巨型离心机正在加紧生产一种备受瞩目的新型核燃料——高纯度低浓缩铀（HALEU）。这种新型燃料所蕴含的能量密度令人咋舌：“三汤匙的 HALEU 释放的能量，就足以满足一个普通人一生的全部用电需求。”然而，运营这座设施的私营企业（Centrus Energy）目前仅建成了 16 台离心机。要将地下那 11,000 个基孔全部填满离心机，看似天方夜谭，但如果原本濒临绝境的核电产业正迎来一场真正的“核能复兴”（Nuclear Renaissance），这个野心勃勃的目标就成了必然。这场核电回归的浪潮由多方力量共同推动。特朗普政府提出了将美国核电容量提升至四倍的宏大构想，直言“这是涉及数千亿美元的产业，我们将重新做到每年新建一座反应堆，在全美遍地开花”；而在全球竞技场上，中国在核电建设速度上已领先美国至少十年，印度也将其视作国家未来的关键支柱。与此同时，亚马逊、谷歌等科技巨头也纷纷重金押注小型模块化反应堆（SMR）。随着人工智能技术爆发，未来六七年内有望涌现大批小型核反应堆，为数据中心源源不断地供电。而这些前沿反应堆的核心命脉，正是亟待增产的 HALEU。然而，燃料供应链的现实却格外严峻。核反应堆不能直接烧地里挖出的天然铀矿石，必须经过极其精密复杂的前端浓缩加工。据行业预测，到 2040 年全球铀需求将翻倍以上。但在这条核燃料生态链中，俄罗斯国家原子能公司（Rosatom）占据着无可匹敌的统治地位，掌控着全球近 50% 的浓缩铀产能。随着俄乌冲突爆发引发的一系列对俄制裁，西方国家不得不痛定思痛，被迫从采矿、转化、浓缩、燃料制造到反应堆建设，彻底重构整条核供应链。这不仅关乎能源安全，更是一场涉及数千亿美元利润的新兴产业争夺战；一旦供应链构建失败，不仅那些地下基孔将继续空置，西方公众数十年好不容易建立起来的核能信心也将再度崩塌。 - **核心数据**：仅需 3 汤匙 HALEU 即可提供一个人终身所需的全部电力。 - **建设现状**：俄亥俄州基地规划了约 11,000 个离心机孔位，目前仅运转 16 台。 - **全球格局**：俄罗斯 Rosatom 独占全球近 50% 的浓缩铀市场份额；预计到 2040 年全球铀需求将增长一倍以上。 --- ### [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) 从时尚名模到“核能网红”：公众认知重塑与硅谷的电力饥渴在加州恶魔谷（Diablo Canyon）核电站一号反应堆的全真模拟控制室里，一场略带戏剧性的场景正在上演。正在教练指导下推动控制棒手柄的学员，并不是传统的核电工程师，而是一位留着精致美甲的巴西超模伊莎贝尔·博梅克（Isabelle Boemeke）。在推拉操作杆时，她开玩笑问是不是因为自己指甲太长卡住了，教练则幽默地建议“试着从侧面推”，并在她成功对齐读数后宣布：“好了，现在你是一名核电操纵员了。”博梅克在社交网络上更为人熟知的身份是 TikTok 上的头部核能科普博主“Isodope”。她用一分钟左右超现实、前卫又幽默的短视频向年轻人科普小型模块化反应堆等硬核技术，打破了核电枯燥晦涩的传统刻板印象。博梅克的走红折射出近年来“核能网红”（Nukefluencer）群体的崛起。尽管绿色和平组织等反核团体坚持批评核电“依然危险且污染环境”，指责这种快节奏的短视频宣传掩盖了核废料与事故风险的复杂争议，但博梅克的影响力早已突破了社交媒体平台。她的丈夫是 Airbnb 联合创始人、被特朗普任命为白宫首席设计官的乔·吉比亚（Joe Gebbia），两人共同向核电公益事业捐助了数百万美元。博梅克不仅亲自投资了核电初创企业，甚至自述曾深度参与起草了旨在简化核能监管的美国总统行政令草案，成为推动硅谷资本与华盛顿政策结合的重要民间纽带。这种公众与资本的积极态度，与过去数十年美国核电的颓势形成了鲜明对比。自 2019 年以来，关于核电的新闻几乎全是一座接一座老反应堆“永久关闭”的讣告（如加州的 San Onofre、佛罗里达的 Crystal River）。传统核电站的牌照审批、人工与基建投资耗资巨大且进度莫测——例如佐治亚州南方公司建造的沃格特尔（Vogtle）核电站两座新机组，完工足足拖延了 7 年，预算超支逾 160 亿美元。与近年来成本骤降、建造灵活的太阳能相比，传统核电毫无优势。直到近年来，全美各地拔地而起一座座没有窗户的神秘巨型建筑——科技巨头的人工智能算力中心。AI 算力对全天候、零碳基荷电力的贪婪需求，彻底改变了核电的游戏规则。原本即将退役的加州最后一座核电站恶魔谷，在民意抗议与政界协调下被紧急延寿至至少 2030 年；就连发生过严重核事故的三里岛核电站，其停运的机组也被星座能源（Constellation Energy）计划重新启用，更名为“克兰清洁能源中心”（Crane Clean Energy Center）。自 2020 年以来，全球在核电领域的投资规模已飙升至约 3,000 亿美元。 - **沃格特尔案例**：佐治亚州 Vogtle 核电站新机组超期 7 年交付，超支金额高达 160 多亿美元。 - **退役逆转**：恶魔谷核电站成功延寿至 2030 年；三里岛部分机组正被翻新并更名为 Crane 清洁能源中心以重启供电。 - **资本热潮**：2020 年至今，北美、欧洲与亚洲在重启与新建核电项目上的全球总投资额已达到约 3,000 亿美元。 --- ### [07:18](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=438s) 拆解核燃料循环：从开采天然矿石到微型燃料颗粒的蜕变核燃料供应链之所以受到全球极度严苛的监管与控制，原因在于民用核燃料的浓缩技术与制造核武器的核心技术完全处于同一同位素物理轨道，一旦失控将带来灭顶之灾。天然开采出的铀矿石在物理性质上其实是一块放射出微弱伽马射线的石头，其中包含三种同位素：铀-234 含量不足 0.01%，几乎可以忽略不计；铀-238 占据了绝大多数，比例高达 99.3%；而真正具备裂变能力的天然钥匙，是稀缺的“幸运儿”铀-235，其天然丰度仅有微不足道的 0.7%。要让这 0.7% 的可裂变物质驱动现代电网，需要经历一连串精细且极度危险的工业蜕变过程： ```mermaid flowchart TD accTitle: 核燃料全生命周期制造流程 accDescr: 简明展示从天然铀矿石开采、化学转化、离心机浓缩到制成燃料颗粒并最终在反应堆中裂变发电的完整工序流。 subgraph FrontEnd["前端原料处理与气体转化"] A["天然铀矿石开采"] --> B["粉碎提纯制成黄饼"] B --> C["化学转化为气态六氟化铀 (UF6)"] end subgraph Enrichment["离心浓缩"] C --> D["高速离心机旋风分离"] D -->|分离较轻 U-235 与较重 U-238| E["商业核反应堆燃料: 约 5% 丰度"] D -.->|过度浓缩: 极其严管| F["武器级高浓缩铀: 90% 以上丰度"] end subgraph FuelFabrication["元件制造与能量释放"] E --> G["转化为金属形态并压制为小软糖大小颗粒"] G --> H["组装为核燃料棒并装入堆芯"] H --> I["核裂变释放巨大热能烧水产生蒸汽"] I --> J["驱动汽轮机发电 (伴随放射性核废料处理)"] end ``` 第一步是将开采出的铀矿石加工粉碎提炼成蓬松鲜艳的粉末——因其颜色和质地类似蛋糕而得名的“黄饼”（Yellowcake）。随后，黄饼被送往专门的转化工厂，经过氟化反应转变为气态的六氟化铀（$\text{UF}_6$）。进入气态后，最核心的离心机浓缩程序拉开帷幕：在离心机超高速旋转形成的内部飓风中，质量稍轻的铀-235 原子与稍重的铀-238 原子发生极其细微的物理分离。通过数千台离心机的级联循环，铀-235 的浓度被富集提升至民用常规动力堆所需的约 5%（如果继续浓缩超过 90%，就会跨过极其危险的门槛成为武器级浓缩铀）。浓缩达标后的气体随后被运往燃料制造厂，重新转化回固体金属化合物形态，并被机械高压压制成一块块仅有“小软糖大小”（gummy bear）的高密度核燃料颗粒。这些小颗粒被整齐紧密地封装进耐高温腐蚀的锆合金燃料棒中，集成为燃料组件装填入反应堆堆芯。在此处，铀-235 原子核遭遇中子轰击分裂并释放出难以想象的巨大热能，将水煮沸转化为超高压蒸汽喷向汽轮机发电。“每一次你打开家里的空调，都仿佛在暗中激活了一座精密无比的鲁布·戈德堡机械。”至于裂变反应遗留的高放射性乏燃料核废料，正如现今整个工业界的惯常做法一样，被留待日后集中封存解决。 - **天然同位素构成**：天然铀中 U-238 占 99.3%，具有裂变活性的 U-235 仅占 0.7%，U-234 占不足 0.01%。 - **浓缩比例对比**：民用常规轻水反应堆需将 U-235 浓缩至约 5%；而武器级浓缩铀丰度需达到 90% 以上。 - **物理尺寸与形态**：最终入堆的燃料颗粒被压制成仅如橡皮软糖大小的圆柱体，排列封装在金属燃料棒内部。 --- ### [09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) 极北荒原的超级铀矿：NexGen 的 Rook I 勘探与漫长开发周期随着新一轮核电繁荣与对俄断供预期的持续发酵，全球资本在核供应链最底层的矿石开采端展开了激烈争夺。在加拿大萨斯喀彻温省著名的阿萨巴斯卡盆地（Athabasca Basin），NexGen Energy 公司的放射性探测仪正发出极其刺耳的密集蜂鸣声，读数瞬间爆表冲破了每秒 65,000 次计数（counts per second）。这种惊人的超高计数频率，宣告了地底深处蕴藏着全球罕见的顶级高品位铀矿化层。这里正是 NexGen 旗下备受瞩目的 Rook I 地下开采项目，其核心依托的是 Arrow 铀矿床。NexGen 创始人兼首席执行官利·库里尔（Leigh Curyer）在介绍该矿床时毫不掩饰其战略分量：“Arrow 矿床是全球规模最大、品位最高的新建铀矿项目，一旦正式投产，其单个矿区的产量预计将占到全球矿山铀总供应量的 20% 到 25%。”虽然前景诱人，但要在地图上找到它绝非易事。它深嵌于加拿大极北的苦寒荒原之中，距离萨斯卡通（Saskatoon）西北偏北 700 公里，距离最近的市政定居点拉洛什（La Loche）及克利尔沃特河德内原住民部族领地也有 155 公里之遥。即便手握这一地下巨额宝藏，将其转化为实际供应依然隔着巨大的现实沟壑。首先是天量的前期资本开支，仅 Rook I 项目的建设启动成本就高达约 15 亿美元。更耗时熬人的是极其冗长的合规审批、环境评估、开采牌照申请，以及与周边占当地人口约 13% 的原住民部落达成具有法律约束力的利益补偿协议。库里尔坦言：“从最初发现 Arrow 矿床，到真正挖出第一磅能够商业交付的铀，整整需要历时 20 年以上。”尽管距离破土动工与全面投产仍有时日，但在全球燃料短缺迫在眉睫的焦虑驱动下，NexGen 已经开始在建造阶段之前，与全球各大核电运营商密锣紧鼓地展开长期供货协议的商业谈判。 - **产能预期**：Rook I 项目全面投产后，预计单一矿区即可供应全球天然铀矿总开采量的 20% 至 25%。 - **地理坐标**：位于萨斯喀彻温省阿萨巴斯卡盆地，距萨斯卡通市西北偏北方向 700 公里。 - **周期与成本**：项目启动资本需约 15 亿美元；从地质发现到首批商业铀矿石开采出厂，周期长达 20 年以上。 ### [11:37](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=697s) 铀浓缩与离心机森林：探访 Centrus 生产基地从地底开采铀矿不仅耗费巨资，而且难度极大，但真正艰巨的挑战始于分离棘手的铀-235同位素。铀浓缩是地球上受到最严密管控、资本密集度最高的工艺之一，它是化学、物理学与地缘政治在同一“旋转脱水周期”中的结合体。在工厂内部，工作人员在后方装配离心机，随后通过一条被称为“转运走廊”的通道将它们推向车间，穿过一扇被戏称为“金刚之门”（King Kong doors）的巨型厚重防护门。驾驶转运车飞驰的是 Matt Snider，他曾是美国核潜艇部队的机械师，如今在 Centrus 拥有 15 年工龄。Centrus 是目前唯一获得美国核管会许可生产 HALEU（高纯度低浓缩铀）的本土公司。眼下的工厂景象展现出强烈的对比：车间里仅有 16 台离心机正在平稳运转，而地面上布满了密密麻麻的预留孔洞。公司的宏伟目标是将整个厂区填满 11,000 台高速离心机，全面扩大 HALEU 产能，为下一代新型核反应堆提供动力。这种布局让人真切感受到，一场产业变革正在拉开帷幕，但远未彻底落成。常规的商用核燃料被称为低浓缩铀（LEU），其铀-235的丰度通常仅在 3% 到 5% 之间；而 HALEU 则强效得多，富集度达到了约 20%。仅需 3 汤匙的 HALEU，就足以供给一个人整整一生的全部用电需求。目前，全球只有俄罗斯和中国能够实现商业规模化生产 HALEU。为了打破垄断，Centrus 若能部署完这 11,000 台机器，其产能将能够填补对俄制裁留下的巨大市场真空。美国能源部（DOE）此前已向 Centrus 及另外两家燃料制造商划拨了 9 亿美元专项资金，但在适当资金支持下，这片“离心机森林”仍需要 6 到 7 年的时间才能完全建成。当被记者追问“富集实际铀材料究竟需要多长时间”时，受访人员直截了当地拒绝：“关于这一点，我完全不能谈。”由于离心机技术同样可用于提炼武器级浓缩铀，所有核心工艺细节均属于高度涉密范畴。 - **LEU（低浓缩铀）**：铀-235 浓度约为 3% 至 5%，为现有传统反应堆标准燃料。 - **HALEU（高纯度低浓缩铀）**：铀-235 浓度约为 20%，能量密度极高（3 汤匙满足单人一生用电）。 - **产业现状与周期**：商业级 HALEU 生产目前由中俄掌控；Centrus 规划的 11,000 台离心机满载工厂需要 6 至 7 年建设周期。 --- ### [14:12](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=852s) 下一代反应堆与“先有鸡还是先有蛋”的产业僵局押注 HALEU，本质上也是在押注计划使用该燃料的下一代核反应堆——尤其是小型模块化反应堆（SMR）。SMR 的核心构想是在工厂中标准化预制构件，直接运输并在现场完成组装。比如一些雄心勃勃的落地构想，包括在加拿大北部的偏远荒野城镇部署小型反应堆，为远离电网的社区供电。在这些极端场景下，使用 HALEU 可以大幅减少更换燃料棒的出勤频次，从而显著拉低长期的运维成本。然而，整个行业正深陷于典型的“先有鸡还是先有蛋”的悖论中：燃料生产商不敢贸然扩建产能，因为目前尚无实际反应堆带来的商业需求；而反应堆开发商又不敢贸然开工建造，因为市场上还没有稳定充足的本土燃料供应链。这种胶着目前仅在极小范围内通过点状投资被打破，如欧洲铀浓缩财团（Urenco）和法国法马通/欧安诺（Orano）已承诺投资新的 HALEU 产线。全行业接下来面临的关键考验，是能否有足够多的初创企业突破“首堆示范”（first-of-a-kind）阶段，真正迈向标准化批量流水线生产。 ```mermaid flowchart TD accTitle: HALEU与SMR的先有鸡还是先有蛋困境 accDescr: 展示燃料供应商因缺乏堆型需求不敢扩产，与反应堆厂商因缺乏燃料不敢建厂的循环僵局 subgraph FuelSupply ["燃料供给端 (HALEU)"] F1["缺乏商业规模订单"] --> F2["不敢贸然投资扩大离心机产能"] end subgraph ReactorDemand ["反应堆建设端 (SMR)"] R1["担忧本土燃料短缺与断供"] --> R2["不敢贸然建设和部署商业反应堆"] end F2 -.->|"没有燃料保障"| R1 R2 -.->|"没有反应堆运行"| F1 ``` --- ### [15:44](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=944s) 微软创始人入局：TerraPower 与 Natrium 钠冷快堆位于西雅图郊外的一家科技公司正试图成为率先破局的那只“鸡”或“蛋”——这就是比尔·盖茨于 2008 年创立的先进核能公司 TerraPower。该公司的首座旗舰级 SMR 项目名为 Natrium（钠冷反应堆），正在怀俄明州的凯默勒（Kemmerer）破土动工。如果一切进展顺利，它有望成为全美首座并网发电的商用小型模块化反应堆。 TerraPower 首席执行官 Chris Levesque 拥有 30 年核工业资历，早年曾在核潜艇担任军官。在吵闹的测试车间背景噪音中，他打趣说自己有点耳背是因为年轻时“在潜艇轮机舱呆得太久了”。Levesque 强调，目前全美运行的 90 多座反应堆以及全球的 400 多座老反应堆，都是诞生于计算尺和纸质图纸时代；而 Natrium 是彻头彻尾在现代先进计算模拟环境中设计出来的。Natrium 的额定基础发电输出约为传统常规核电站的三分之一，但它配备了一个名为“能源岛”（Energy Island）的熔盐储热罐系统，这套系统扮演着热电池的角色，能在用电高峰期释放高额调峰电力。 - **公司背景**：比尔·盖茨于 2008 年创立 TerraPower，探索革命性裂变构型。 - **旗舰堆型 Natrium**：采用液态钠冷却，配有熔盐储热系统（能源岛），可作为热力电池动态调峰。 - **建设进展**：位于怀俄明州凯默勒，力争成为全美首座投入运行的商用 SMR。 --- ### [17:01](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1021s) 核反应堆的代际进化与 SMR 的经济学争论核反应堆的演进是一部吸取历史惨痛教训的前行史。20 世纪 50 至 60 年代投运的“第一代”（Gen 1）反应堆多数规模小、可靠性差，存在诸多严重设计缺陷，例如缺少合格的安全壳、燃料管理粗放且严重依赖手动控制。70 年代至 90 年代建造的“第二代”（Gen 2）反应堆迈入全球大规模商业化普及阶段，提高了热效率并配备了自动紧急停堆机构，但设备老化与人为失误依然导致了宾夕法尼亚州的三里岛事故以及苏联的切尔诺贝利灾难。90 年代以后的“第三代”（Gen 3）反应堆则简化了内部管线、改善了热工效率，大幅增强了被动安全冗余度。在当前的产业图景中，下一代反应堆涵盖了“三代+”（Gen 3+）和“第四代”（Gen 4）反应堆，它们普遍被归入 SMR 的范畴。Gen 3+ 本质上是现有技术的微型化和模块化版本；而 Gen 4 则采用了截然不同的物理设计，引入了诸如气冷、铅冷或液态金属钠冷等多元冷却剂，并多以 HALEU 作为高能核心。对于 SMR 的经济前景，学界与业界分歧尖锐：部分乐观派认为模块化工厂预制会显著降低每千瓦建设成本；反方专家则指出，大型传统反应堆是经过数十年极限优化才把平准化度电成本压下来的，SMR 堆芯本身容积较小，其单位千瓦时的固有能量输出受限，失去了单机规模效益。更重要的是，SMR 的分散化部署意味着监管审批点、持证操纵员和实物保卫人员成倍增加，这些沉重的经常性开支极易抹平预制组装省下来的资金。此外，目前全行业的 SMR 概念设计从数年前的 80 种激增到如今的 127 种，但在残酷的“核能适者生存”洗牌中，绝大多数都将被淘汰。 | 反应堆代际 | 发展时间窗口 | 技术特征与演进路径 | 代表性短板或典型事故 | |:--- |:--- |:--- |:--- | | **第一代 (Gen 1)** | 1950 - 1960 年代 | 早期原型堆与试验性示范设施，规模较小 | 缺乏有效安全壳结构，燃料利用率低，重度依赖人工手动干预 | | **第二代 (Gen 2)** | 1970 - 1990 年代 | 走向商业规模化并网，引入自动化紧急停堆 | 设备故障叠加人为误操作，发生三里岛与切尔诺贝利事故 | | **第三代 (Gen 3)** | 1990 年代至今 | 简化系统构型，显著强化非能动安全，提升热效率 | 建造成本昂贵，现场施工周期动辄大幅拖延超支 | | **三代+ / 四代 (SMR)** | 当下在建及未来 | 工厂模块化预制，采用 HALEU 燃料，包含液态金属等多元冷却 | 缺乏度电规模效应，127 种堆型竞争激烈，审批与运维监管门槛高 | --- ### [19:36](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1176s) 融资范式转变：私有资本与科技巨头的入局 Natrium 原定于 2028 年并网发电，但由于缺乏 HALEU 燃料供应，投产节点被迫推迟两年至 2030 年。为了填补首炉堆芯料的供应缺口，TerraPower 最终敲定了一家替代供应商——在纳斯达克上市但生产基地位于南非的 ASP 公司。在资金端，TerraPower 迄今已获得美国能源部 20 亿美元的资助，并在民营领域成功筹集了 14 亿美元风投资本。这种资金结构标志着核电行业投融资底层逻辑的巨变。Chris Levesque 回忆起自己在西布鲁克核电站（Seabrook Nuclear Station）开建时于新罕布什尔州度过的童年。当时，他那些靠薪水勉强度日的父母每个月都会在电费账单上看到一笔“西布鲁克附加费”（Seabrook Surcharge），相当于普通纳税人和终端用电居民直接为核电站高昂的固定资产资本支出买单。在当下的美国社会，这种让纳税人兜底的公用事业模式已彻底行不通。取而代之的是由私有风投资本和对电力如饥似渴的科技巨头承担早期财务风险，这虽在客观上减轻了居民电费负担，却也悄然将国家核心基载核电资产的控制权拱手交到了私营企业手中。目前，TerraPower 除了怀俄明项目外，已与犹他州、堪萨斯州以及英国的公用事业公司签署了前期合作意向书。 - **投产延期与化解**：因受困于 HALEU 断供推迟两年，首炉燃料最终由南非提纯设施（美股上市公司 ASP）承接。 - **资金构成**：美国能源部资助 20 亿美元，民间私募资本融入 14 亿美元。 - **出资模式裂变**：从 20 世纪由“终端纳税人/用电户附加费”兜底，全面转向由“私营风投资本与科技巨头”直投承担高额风险。 --- ### [21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) 棘手的后方现实：千年的核废料阴影与未来博弈在宏大的复兴叙事背后，潜藏着极具争议的隐形成本：全球遍地开花的小型反应堆同样会产生高放射性废料。成百上千座 SMR 散落各地，意味着更多极具危险性的核垃圾将堆积在全球各处的临时存储点内，而目前人类根本没有终极长效的深地处置解决方案。任何参观过传统核电站后方堆场的人都会看到那些巨大的钢筋混凝土干式贮存罐，罐体内封存的乏燃料棒具有致命的放射性，并且“这种致命危险将持续整整上千年”。部分 SMR 研发商声称其闭门反应堆未来能循环利用高达 96% 的乏燃料，但核泄漏一旦沾染生态圈、地下水源或食物链，其带来的生物破坏和长久健康灾难是不可逆的。更讽刺的是，许多学术分析表明，受制于紧凑设计的物理约束，部分 SMR 在每产生一单位电力的过程中，其产出的核废料体积甚至超过了体量庞大的传统反应堆。欧美国家正在艰难重拾过去几十年间因去工业化而逐渐萎缩的核工程建造技能，但这还远非核能复兴面临的全部挑战。核电在美国整体电网中的发电占比未来究竟能从 20% 爬升至 30% 还是 40%，很大程度上不取决于核能本身，而是取决于外部竞争技术的动态博弈：页岩气水力压裂开采商是否愿意让渡其天然气市场份额？新能源领域能否发明出极度廉价的大规模储能电池来平抑风光波动？尽管面临重重阻力，业界乐观预计到 2030 年代初将有更多新型核电机组运转。资本、公众注意力、政治野心与社交媒体的追捧让当前的核能赛道展现出空前动能，但这个行业历来以“步伐极度缓慢”而著称。 ### [23:10](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1390s) 供应链脆弱性与核能复兴的不确定未来高昂且快速攀升的账单给整个产业带来了沉重的财务负担。正在艰难重构的新铀燃料供应链面临着巨大的不稳定性，一旦这一核心供应链发生动摇甚至崩塌，围绕核能展开的宏大“核能复兴”愿景恐将全盘崩溃。尽管核能承诺为未来提供零碳、稳定的基荷电力与技术突破，但现实却极为严峻。正如片中所强调的，面对错综复杂的供应链重组与巨额成本，“通往真正能源独立的道路依然充满了变数”（the path to true energy independence remains anything but certain）。 - **财务与成本压力**：核能产业链投资规模庞大，开销与账单积累极其迅速。 - **供应链系统性风险**：新兴的铀燃料供应链极度脆弱，若无法顺利成型，整个产业复兴将被连根动摇。 - **能源独立的不确定性**：摆脱对外依赖、实现真正自主的清洁能源之路依旧道阻且长。 ## Notable quotes > [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 三汤匙的 HALEU 释放的能量，就足以满足一个普通人一生的全部用电需求。 > [00:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=3s) 这是涉及数千亿美元的产业，我们将重新做到每年新建一座反应堆，在全美遍地开花 — 特朗普 > [02:54](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=174s) 好了，现在你是一名核电操纵员了。 — 教练 > [07:18](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=438s) 每一次你打开家里的空调，都仿佛在暗中激活了一座精密无比的鲁布·戈德堡机械。 > [09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) Arrow 矿床是全球规模最大、品位最高的新建铀矿项目，一旦正式投产，其单个矿区的产量预计将占到全球矿山铀总供应量的 20% 到 25%。 — 利·库里尔（Leigh Curyer） > [09:50](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=590s) 从最初发现 Arrow 矿床，到真正挖出第一磅能够商业交付的铀，整整需要历时 20 年以上。 — 利·库里尔（Leigh Curyer） > [11:37](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=697s) 关于这一点，我完全不能谈。 — 受访人员 > [21:03](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1263s) 这种致命危险将持续整整上千年 > [23:10](https://www.youtube.com/watch?v=rkCg7wR3_hI&t=1390s) the path to true energy independence remains anything but certain ## Mentioned HALEU (高纯度低浓缩铀), LEU (低浓缩铀), Centrus Energy, Rosatom (俄罗斯国家原子能公司), 特朗普, 亚马逊, 谷歌, 小型模块化反应堆 (SMR), 恶魔谷核电站 (Diablo Canyon), 伊莎贝尔·博梅克 (Isabelle Boemeke / Isodope), 乔·吉比亚 (Joe Gebbia), Airbnb, 绿色和平组织, 沃格特尔核电站 (Vogtle), 南方公司, 三里岛核电站, 星座能源 (Constellation Energy), 克兰清洁能源中心 (Crane Clean Energy Center), 黄饼 (Yellowcake), 六氟化铀 (UF6), NexGen Energy, Rook I 项目, Arrow 矿床, 利·库里尔 (Leigh Curyer), Matt Snider, 美国核管会 (NRC), 美国能源部 (DOE), 欧洲铀浓缩财团 (Urenco), 欧安诺 (Orano), 法马通 (Framatome), 比尔·盖茨, TerraPower, Natrium 钠冷反应堆, Chris Levesque, ASP 公司, 西布鲁克核电站 (Seabrook Nuclear Station) ## Video description Nuclear power is trying to make a comeback but the fuel required is in short supply. From uranium mining to enrichment, the supply chain is being reshaped by geopolitics, lobbyists, rising energy demand and new technologies. As countries and companies race to build next-generation reactors, the question is whether the world can produce the fuel needed to power them. Watch more: https://www.youtube.com/show/VLPLqq4LnWs3olWuFFpsiq0cFkGs7WMb6MtA?sbp=QAE%3D Bloomberg Primer cuts through the complex jargon to reveal the business behind technologies poised to transform global markets. This six-part, planet-spanning series offers a comprehensive look at the "impossible" industries vying to shape our future. #Fuel #Nuclear #documentary -------- Like this video? Subscribe: http://www.youtube.com/Bloomberg?sub_confirmation=1 Get unlimited access to Bloomberg.com for just $1.99 your first month: https://www.bloomberg.com/subscriptions?in_source=YoutubeOriginals Bloomberg Originals offers bold takes for curious minds on today’s biggest topics. Hosted by experts covering stories you haven’t seen and viewpoints you haven’t heard, you’ll discover cinematic, data-led shows that investigate the intersection of business and culture. Exploring every angle of climate change, technology, finance, sports and beyond, Bloomberg Originals is business as you’ve never seen it. Subscribe for business news, but not as you've known it: exclusive interviews, fascinating profiles, data-driven analysis, and the latest in tech innovation from around the world. Visit our partner channel Bloomberg News for global news and insight in an instant. [Source](https://www.youtube.com/watch?v=rkCg7wR3_hI)

Source quotes

[0:03] 三汤匙的 HALEU 释放的能量，就足以满足一个普通人一生的全部用电需求。 [0:03] 特朗普: 这是涉及数千亿美元的产业，我们将重新做到每年新建一座反应堆，在全美遍地开花 [2:54] 教练: 好了，现在你是一名核电操纵员了。 [7:18] 每一次你打开家里的空调，都仿佛在暗中激活了一座精密无比的鲁布·戈德堡机械。 [9:50] 利·库里尔（Leigh Curyer）: Arrow 矿床是全球规模最大、品位最高的新建铀矿项目，一旦正式投产，其单个矿区的产量预计将占到全球矿山铀总供应量的 20% 到 25%。 [9:50] 利·库里尔（Leigh Curyer）: 从最初发现 Arrow 矿床，到真正挖出第一磅能够商业交付的铀，整整需要历时 20 年以上。 [11:37] 受访人员: 关于这一点，我完全不能谈。 [21:03] 这种致命危险将持续整整上千年 [23:10] the path to true energy independence remains anything but certain

Source quotes

[11:15] I can't comment [11:15] That's classified. [19:23] they won't all survive nuclear selection. [21:03] If you go to any nuclear power plant, if you go out back, you're gonna see these giant steel and concrete casks, and inside them is the spent fuel rods and those things are deadly. They'll be deadly for a thousand years.

Original transcript

[0:03] - In a small town in Ohio [0:05] behind layers of barbed wire fence [0:07] and many warnings of danger, [0:09] there are about 11,000 holes in the ground [0:13] and these giant things which are churning out an exciting [0:16] new type of nuclear fuel called HALEU. [0:19] - Three tablespoons of HALEU is enough [0:22] to supply a typical person's entire use of electricity [0:25] for their entire lifetime. [0:27] - The private company behind this new fuel production [0:30] wants to fill all those 11,000 holes with centrifuges. [0:34] Right now they've built 16. [0:38] It's the kind of ambitious goal that only makes sense [0:41] if an industry that's been on life support suddenly has a [0:44] - Nuclear renaissance. [0:45] - Nuclear renaissance. - Nuclear renaissance. [0:47] Big tech is bringing nuclear power back in a big way. [0:51] - President Trump has laid out a target [0:53] of quadrupling the USS nuclear capacity. [0:56] - We're talking hundreds of billions of dollars. [0:58] We will build a reactor a year again, [1:00] we will build them across the country. [1:01] - This is no longer about just electricity, [1:03] it's about India's future. [1:05] - China is outpacing the United States [1:07] by at least a decade. [1:08] - And this new fuel may be crucial to power a new wave [1:11] of smaller reactors that can be deployed in more places [1:14] to help power the future of technology. [1:16] - Amazon and Google meanwhile are investing in small [1:19] modular reactors. [1:21] - I think in the next six, seven years, [1:23] I think you're gonna see a whole bunch [1:24] of small nuclear reactors. [1:26] - But supplying this fuel gets tricky. [1:29] - Nuclear reactors run on uranium, [1:31] and it's not just uranium you dig outta the ground. [1:34] They have this very complicated process [1:36] where they have to enrich the uranium into nuclear fuel [1:40] - And the need for nuclear fuel is growing. [1:43] By some estimates, the global uranium demand is expected [1:47] to more than double by 2040. [1:49] - Far and away. The biggest company involved [1:52] in the nuclear fuel ecosystem is Russia's Rosatom. [1:55] They control almost 50% of the capacity to enrich uranium. [2:00] - But due to sanctions against Russia over its war in [2:03] Ukraine, the world's uranium market is transforming. [2:06] And since the nuclear ecosystem is interconnected, [2:09] western countries are being forced [2:11] to rewrite their entire nuclear supply chains [2:14] from mining, conversion, enrichment, [2:16] and fabrication to building all those small reactors. [2:20] Success at each link could mean more than new fuel. [2:23] It could grow new industries [2:25] and the potential to rake in lots and lots of money. [2:29] And failure wouldn't just mean these [2:31] holes in the ground stay empty. [2:33] It could cause a crack in nuclear confidence [2:35] that could take generations to seal. [2:54] - All right, so yeah, push that. [2:55] Make it in. It might be hard. [2:58] - Is it my nails? - Maybe [3:00] Try going sideways. [3:02] - That's funny. - There you go. [3:04] You're gonna drive rods in [3:05] and you're gonna try to keep [3:07] that number close to that number. [3:09] So go ahead. [3:11] Now you're a nuclear operator. [3:14] - This is a simulator [3:16] of reactor one at California's Diablo Canyon power plant [3:20] and the trainee working The joystick is a Brazilian fashion [3:23] model turned nuclear influencer named Isabelle Boemeke [3:27] or for those deeply into the nuclear energy space on TikTok, [3:31] she's known as "Isodope." [3:33] - This is a plant. No, not this one. This one. [3:36] A smile modular reactor. [3:37] What is my purpose? [3:40] I started making short one minute long videos [3:42] that were very abstract, and I also try to make it very fun. [3:49] - She's not alone. In the last few years, [3:51] there has been a rise in so-called "Nukefluencers." [3:54] Some critics note that this new wave [3:56] of snappy short form social media advocacy can gloss over [4:00] nuanced positions from anti-nuclear groups like Greenpeace. [4:04] - We don't need nuclear energy. [4:06] It's still dangerous and polluting. [4:09] - But Boemeke is not your average Nukefluencer. [4:12] She's reportedly invested in a nuclear startup [4:15] and is married to Joe Gebbia co-founder of Airbnb [4:18] and Donald Trump's chief design officer. [4:20] The two have donated millions to nuclear causes. [4:24] This rise in social media influence comes at the [4:26] same time the nuclear power industry. [4:28] and some Silicon Valley startups are pushing the Trump [4:31] administration to help kickstart development. [4:34] And according to her own account, Boemeke was able [4:36] to provide input on draft executive orders targeting the [4:40] regulation of nuclear energy. [4:42] This renewed push is a major change in the trajectory [4:45] of nuclear power in the United States. [4:48] - San Onofre nuclear plant is shutting down for good. [4:51] - Crystal River nuclear power plant [4:53] now set to close for good. [4:55] - Closed for good. [4:56] - Shut down for good. - For good. [4:57] - For good. [4:59] - I've been writing about nuclear energy since 2019 [5:02] and when I started, pretty much the only story, [5:05] especially in the US was which one is gonna close down next [5:09] - From permits to labour to concrete building [5:12] and operating a nuclear plant can cost a fortune [5:15] long before it produces a single watt, [5:17] and the road to operation can be unpredictable. [5:21] The two newest reactors built in the US at Southern [5:24] Company's Vogtle Plant in Georgia came online seven years [5:28] behind schedule and more than $16 billion over budget. [5:32] At the same time, solar energy has gotten much cheaper [5:35] to produce and a lot more reliable to build and operate. [5:39] But then all these unmarked windowless buildings started [5:43] popping up everywhere. [5:44] - The big tech companies started saying, "we need [5:48] to build out these huge data centers [5:50] for artificial intelligence." [5:52] And the thing about AI is [5:54] those data centers need a lot of electricity, [5:57] and so they started looking at nuclear power. [5:59] - Meanwhile, the operating license [6:01] for California's last nuclear power plant was [6:03] about to expire. [6:05] - So at the same time, [6:06] Diablo Canyon in California was also on the chopping block. [6:11] - This time, it was a confluence of economic concerns [6:13] and activism pushing against the proposed closure. [6:17] We had a giant blimp rolling through the streets. [6:20] *CHANTING* We're on a mission to stop all emissions! [6:23] - To the surprise of literally everybody, [6:25] the plant is open at least until 2030. [6:28] - That's not all. Other decommissioned plants are getting a [6:31] new lifeline like Three Mile Island. [6:34] Now, Constellation Energy is actively working to reopen one [6:37] of its reactors under a new name. [6:39] The Crane Clean Energy Center. [6:41] - Changing the name really, you cannot erase history. [6:45] You cannot rebrand a nuclear disaster. [6:48] - But still, global investment in nuclear energy since 2020 [6:52] is estimated to be around $300 billion [6:56] with ambitious plans in Europe, North America, and Asia, [6:59] to revive or kickstart their nuclear energy sectors. [7:03] Every step of the supply chain from mining to enrichment [7:06] to building reactors is highly regulated and controlled [7:10] because a single misstep could be catastrophic [7:14] and erode the public trust [7:15] nuclear has worked decades to regain. [7:18] It's a real shift from when I came of age [7:21] and everyone was worried about nuclear war [7:23] because the process of enriching uranium for nuclear fuel, [7:27] same technology that they use for enriching uranium [7:31] to make nuclear weapons. [7:33] - It also means that turning a lump of rock like this [7:36] into enough electricity to power all [7:39] of this is complicated. [7:41] Let's break it down. [7:44] Uranium ore contains three unique isotopes: [7:47] U-234 accounts for less than 0.01%. [7:52] So let's just forget about that one. [7:54] Poof! [7:55] U-238 on the other hand, makes up most of it: 99.3%. [8:00] But the key to nuclear energy lies within this one: [8:04] lucky number U-235. [8:06] Meaning only about 0.7% is fissile type of uranium. [8:12] But now, this thing is basically a gamma emitting rock. [8:16] So let's turn it into fuel. [8:18] - You need to first convert the ore into yellow cake, [8:21] distinguished by its color [8:23] and its fluffy composition similar to cake. [8:26] - That stuff is then sent to a conversion facility [8:29] where it's converted into this gas, uranium hexafluoride. [8:33] - To make the fuel, you take the gas [8:36] for a spin in a centrifuge inside the whirlwind, [8:39] the lighter U-235 atoms slowly separate from the heavier [8:44] U-238 ones. [8:46] - So they go through the centrifuges [8:48] until the U-235 is at about a 5% concentration. [8:53] - If you keep things cooking for longer, [8:54] you'll reach weapons grade, which is 90% or more, [8:59] - And that's really - Tightly regulated. [9:01] - So let's take it out before then. [9:03] - That uranium hexafluoride gas is sent to another facility [9:07] where it can be turned into metallic form [9:10] - And then pressed into pellets each about the size [9:13] of a gummy bear, then stacked into rods [9:16] and loaded into nuclear reactor cores. [9:19] And this is where we finally get energy through the form [9:22] of nuclear fission U-235 isotopes split, [9:27] which makes heat, which boils water, which shoots [9:30] through a turbine, which makes energy. [9:33] It's like activating a secret Rube Goldberg machine every [9:36] time you turn on the AC. [9:39] Oh, there's also the matter of radioactive nuclear waste, [9:41] which we'll put off to deal with later. [9:44] Just like the real thing. [9:50] For companies chasing the next big uranium boom, [9:53] every link in this chain looks like an opportunity [9:55] starting from the ground up. [9:58] - Now this sound going off is telling us we're seeing lots [10:02] of counts per second here, [10:03] very high counts per second at 65,000+. [10:07] Now that is some high scale, high grade [10:10] uranium mineralization. [10:12] *Machine beeping* [10:14] - Rook One is an underground uranium mining project [10:17] from NextGen targeting the arrow deposit in [10:19] Saskatchewan's Athabasca Basin. [10:21] - The arrow deposit is the world's largest, highest grade [10:26] uranium mining project, [10:28] and we'll be producing approximately 20 to 25% [10:32] of the world's mine supply. [10:34] - It hasn't begun production yet, [10:36] but if you want to check out the site, Leigh Curyer [10:39] the founder and CEO of NextGen is happy [10:41] to give you directions. [10:42] - It's 700 kilometers NNW of Saskatoon. [10:46] 155 kilometers north of the nearest municipal [10:50] and reservation of La Loche [10:52] and the Clearwater River Dene Nation. [10:57] - But even if you are lucky enough [10:58] to find an underground treasure trove of uranium ore, [11:01] you can't just grab a shovel and start digging it up. [11:04] First. You need to raise a lot of capital. [11:08] Rook One for example, costs around $1.5 billion. [11:13] And then there's the paperwork like licenses and permits [11:16] and signoffs from surrounding indigenous communities, [11:19] which make up around 13% of Athabasca County. [11:23] - You're looking basically from discovery of arrow [11:26] to likely first pound of production, 20+ years. [11:32] - But even before entering the construction phase, [11:34] NextGen is already negotiating deals [11:37] with utilities in the us, Europe, Asia, and the Middle East. [11:42] Pulling uranium from the earth is no small feat [11:45] or small expense, [11:47] but the real challenge begins when you try [11:49] to separate those pesky U-235 isotopes. [11:55] Uranium enrichment is also one [11:56] of the most tightly controlled capital intensive [11:59] processes on earth. [12:01] It's chemistry, physics, [12:02] and geopolitics all in one spin cycle. [12:07] This should look familiar. [12:09] - We assemble the centrifuges behind us, [12:11] then we take them down this transfer corridor. [12:14] - And you got these massive doors? [12:16] - Yes, we call them King Kong doors. [12:19] - The speed demon behind the wheel is Matt Snider, [12:22] a former mechanic in the nuclear navy [12:24] and a 15-year veteran at Centrus, [12:27] the only US company licensed [12:28] to make HALEU, a concentrated type of nuclear fuel. [12:32] - They've got this enormous facility [12:35] with 16 centrifuges running [12:37] and then holes in the ground [12:39] where they're gonna have the rest. [12:42] - They plan to jam pack this place with 11,000 centrifuges [12:46] so they can scale up their HALEU production [12:48] to power a future market of new nuclear reactors. [12:51] - It really struck me how this is starting to happen, [12:55] but it's not here yet. [12:57] - Standard nuclear fuel is called "low enriched uranium," [13:01] or LEU, and has a U-235 enrichment concentration of 3 to 5%. [13:07] HALEU is more potent. It's about 20%. [13:11] - Three tablespoons of HALEU is enough [13:13] to supply a typical person's entire use of electricity [13:17] for their entire lifetime. [13:19] - Right now, only Russia [13:21] and China are making HALEU at commercial scale. [13:24] - So if we are to expand our operations [13:26] with those 11,000 machines, that's enough [13:29] to meet the demand left by the Russian sanctions. [13:32] - The US Department of Energy awarded Centrus along [13:35] with two other fuel makers, $900 million, [13:39] but the centrifuge forest won't grow overnight [13:42] - With appropriate funding we expect our full capacity plant [13:46] to be built within the next 6 to 7 years. [13:49] - How long does it take to to enrich the, the uranium action [13:52] - For I can't. I can't go into that [13:54] - Because centrifuges can also make weapons-grade fuel. [13:57] Many topics around the technology a classified, [14:00] - I can't really comment. - I can't answer that. [14:02] - I can't talk about that. [14:04] - Can't really comment. [14:06] - So let me try to get into this next part without revealing [14:09] any national security secrets. [14:12] Betting on HALEU also means betting on the next generation [14:16] reactors that plant to use it like many SMRs, [14:19] Small Modular Reactors, [14:21] that can be built in factories [14:22] shipped and assembled on site. [14:24] - If you are looking at some [14:26] of the more ambitious SMR ideas like placing small [14:31] reactors in the wilds of Northern Canada to supply power [14:36] for remote towns, you want to use HALEU as the thinking goes [14:41] so that the number of refueling trips out can be reduced [14:46] and that therefore would drive down the cost of operation. [14:51] - We'll get into the economics of SMRs a bit later, though. [14:54] Hang tight. [14:57] - It's a chicken and an egg problem. [14:59] They can't have the fuel unless there's demand, [15:02] but there's no demand until they have reactors. [15:05] But they don't want to build the reactors [15:07] unless they know that there's gonna be fuel. [15:09] You can see where I'm going here. [15:10] - The chicken and egg problem is being solved [15:12] at a small scale right now. [15:14] Urenco has committed to invest in new HALEU production. [15:18] Orano in France has as well. [15:20] The next stage is whether there are enough companies [15:24] that can go beyond a first of a kind SMR model [15:28] and actually begin serial production. [15:31] - One company outside of Seattle is looking to become [15:34] that first SMR chicken, or SMR egg, [15:38] or whichever one comes next. [15:44] TerraPower is a nuclear technology company founded [15:46] by Bill Gates in 2008. [15:49] Their first SMR called Natrium is currently under [15:52] construction in Kemmerer, Wyoming. [15:54] If all goes to plan, it could be the first SMR [15:57] to go online in the U.S. [15:59] - If you looked at the 90+ reactors [16:02] that are operating in the U.S. today, [16:04] or the 400 around the world, [16:06] they were born and designed in slide rule space [16:10] in the space of paper drawings. [16:13] But our reactors started their development in an advanced [16:17] computing environment. [16:19] - The CEO of Terra Power, Chris Levesque, [16:21] has spent 30 years in the nuclear industry [16:23] beginning his career as an officer in the nuclear navy. [16:26] - Sorry, with the background noise, [16:28] I'm having a little hard time hearing you. Yeah. [16:30] - Oh, sorry. Was just giving you a quick introduction. [16:32] - Part of the reason I have trouble hearing is I've spent [16:34] too much time in engine rooms in in submarines. [16:38] But you know, Bill knew he wanted to work on the next form [16:42] of vision, you know, which for us is our natrium reactor. [16:47] - Natrium outputs around a third of the electricity [16:49] of a standard nuclear reactor [16:51] or more when it taps into its molten salt storage tap, [16:55] otherwise known as energy island, which sort [16:57] of acts like a thermal battery. [17:01] Each generation of reactor is built on the [17:03] lessons of the past. [17:06] Reactors from the 1950s [17:08] and 1960s are often called Gen 1 reactors. [17:11] Many were small, unreliable, [17:13] and had a number of design flaws like inadequate [17:16] or nonexisting containment structures, [17:19] inefficient fuel management, [17:21] and a heavy dependence on manual control. [17:24] Next came Gen 2 reactors built between the 1970s [17:27] and the 1990s. [17:29] They were more efficient and built for commercial scale [17:31] and global deployment. [17:34] Gen 2 featured safety upgrades like automated emergency [17:37] shutdown mechanisms, but equipment failure and human error [17:41] led to accidents at Three Mile Island in Pennsylvania [17:45] and Chernobyl in the USSR. [17:47] Reactors from the 1990s and beyond are called Gen 3 [17:51] and feature simplified designs, better fuel efficiency [17:55] and improved safety systems. [17:57] - The next generation of nuclear power plants, [18:00] there's gonna be Gen 3+, [18:02] and what that is is pretty much a smaller version of [18:06] what we've got now. [18:08] After Gen 3+ is gonna come Gen 4. [18:11] Those are different. There's a lot [18:13] of different designs people are working with. [18:15] There's different cooling materials. [18:17] They use different fuels like HALEU, but Gen 3+ [18:20] and Gen 4, they're all generally what we call SMRs. [18:25] That's a small modular reactor. [18:28] - Some experts estimate [18:30] that SMRs will be less expensive per kilowatt [18:33] than their predecessors. [18:35] While others argue that [18:36] first-of-a-kind risks, high maintenance requirements, and [18:40] regulatory hurdles will ultimately make them more costly. [18:45] - Large nuclear reactors have been optimised over decades [18:49] to bring down the cost of generating the electricity. [18:53] And because SMRs are smaller, the amount [18:57] of energy that they will be able to generate inside [19:00] of their machines will be smaller [19:03] on a per kilowatt-hour basis, [19:05] meaning the economics will be challenging. [19:07] Obviously, SMRs imply there's gonna be more nuclear reactors [19:11] at more sites, which means more regulation, [19:15] more people to operate them. [19:16] Those are all things that raise the cost [19:18] of SMRs potentially. [19:20] - Another hurdle with SMRs is there are a lot [19:23] of different design ideas, [19:24] but they won't all survive nuclear selection. [19:27] - The amount of SMR designs has grown [19:29] to about 127 today from just about 80 a few years ago. [19:36] - Natrium was initially slated to go online in 2028, [19:40] however, the company was delayed two years due to a lack [19:43] of HALEU, but they found a solution. [19:45] - That new source for the first core load is going to be [19:49] a company called ASP in South Africa. [19:52] It's a U.S. NASDAQ listed company, [19:54] but the enrichment operations will occur in South Africa. [20:00] - So far, TerraPower says it has received $2 billion from [20:04] the U.S. DOE, and raised another $1.4 billion [20:07] in private capital. [20:09] - I grew up in New Hampshire [20:10] where Seabrook Nuclear Station [20:12] was being built when I was growing up. [20:14] And you know, my parents who lived paycheck to paycheck [20:17] on their electric bill, they had something called [20:19] the "Seabrook Surcharge." [20:21] So the rate payers were literally paying for the CapEx of [20:24] that first nuclear plant. [20:26] That's not gonna work in the U.S. today. [20:29] - So rather than rate payers subsidizing utilities reactors, [20:33] a new shift could see private capital footing [20:36] that massive bill, including the influx of tech investment, [20:39] which could take financial burden off [20:41] of taxpayers while putting control [20:43] of the country's nuclear power in the hands [20:45] of the private sector. [20:49] In addition to Wyoming, [20:50] TerraPower says it's also signed preliminary agreements [20:53] with utilities in Utah, Kansas, and the U.K. [20:56] and many other SMRs in both the U.S. [20:58] and Europe are in promising stages of development. [21:03] Still, there's a hidden cost that comes with hundreds [21:07] of small reactors being deployed around the world. [21:09] Just like larger reactors, SMRs create radioactive material [21:14] when the fuel becomes depleted. [21:15] More SMRs deployed globally means more nuclear waste piling [21:20] up and temporary storage around the world. [21:22] And currently there's no long-term solution for [21:25] where to put the stuff. [21:27] - If you go to any nuclear power plant, if you go out back, [21:30] you're gonna see these giant steel and concrete casks, [21:34] and inside them is the spent fuel rods [21:35] and those things are deadly. [21:37] They'll be deadly for a thousand years. [21:39] - Some SMR developers claim their reactors will eventually [21:42] recycle up to 96% of their spent fuel. [21:46] But even a little bit of nuclear waste exposure [21:48] could do irreparable harm to the ecosystem, water supply, [21:51] food chain, and cause long-term health effects. [21:55] And ironically, many SMRs, due to their compact designs, [21:58] could actually produce more nuclear waste per unit [22:02] of electricity than they're larger reactor counterparts. [22:11] - In Europe and the U.S. they have got to relearn some [22:14] of the skills that have been atrophying over the past [22:18] decades, but there's no reason [22:20] that they cannot overcome the challenge. [22:23] Whether nuclear power is going to increase from 20% [22:27] of generation in the U.S. to 30% [22:30] or 40% of generation in the future is largely a function [22:33] of other technologies. [22:35] It's a function of whether gas frackers are going [22:38] to be willing to forego market share of their product. [22:41] It's gonna be a function of whether we find new ways [22:44] of making cheap batteries to store wind or solar power. [22:48] - It's gonna take a few years, [22:50] but by the early 2030s, [22:51] I'm pretty confident we're gonna have more nuclear power [22:55] plants and more different nuclear technologies in operation. [22:59] - There's real momentum behind nuclear right now, [23:03] money, attention, ambition, social media influence. [23:06] But this is an industry that moves famously slow [23:10] with enormous bills that pile fast. [23:12] And if the newly-forming uranium supply chain begins [23:15] to topple, the whole nuclear renaissance could collapse. [23:19] So for all its promise, the path [23:21] to true energy independence remains anything but certain.