初创公司尽调
尽调报告 Climate / Energy — Advanced nuclear SMR Late-stage private commercialization 2026-07-19

Kairos Power

先进核能开发商,客户验证扎实,但执行风险高

Kairos Power 是先进核能赛道最强的战略平台之一,但仅凭公开证据,目前更适合观察而不是买入:估值和机组经济性披露不足,首堆执行风险仍高。

封面要素

Google 订单储备 01
500 MWe by 2035 [CO020, CU004]
Hermes 支持包 02
629 USD M total [CI003, CI004]
团队规模 04
500+ employees [CO009]

公司概况

Kairos Power 是一家美国私营先进核能公司,正在开发 KP-FHR:一种氟盐冷却高温反应堆,使用 TRISO 环形球燃料,并按 150 MWe 商业电站设计。公司围绕田纳西州 Hermes 示范路线推进“测试中学习”模式,同时在 Alameda、Albuquerque 和 Oak Ridge 搭建制造与燃料能力。放在这个品类里,它的公开商业验证异常强:Google 签署主协议,目标到 2035 年最高 500 MWe;TVA 成为 Hermes 2 首个已披露的公用事业承购方。即便如此,公开证据显示 Kairos 仍未产生收入,也没有披露投资人做严谨判断所需的估值、股权结构表或合同经济性。

官网
kairospower.com
成立时间
2016-01-01
创始人
Mike Laufer, Ed Blandford
总部
Alameda, California, USA
产品
KP-FHR 先进反应堆采用氟盐冷却、高温运行和 TRISO 环形球燃料;Kairos 先开发示范资产,计划建设总计约 150 MWe 的商业双机组电站。
客户
超大规模数据中心、公用事业公司,以及最终其他需要长时段无碳电力、并可能看重工业热选择权的大型清洁稳定电力买方。
商业模式
开发、建设、持有并运营反应堆电站,再通过电力销售、辅助服务、环境属性,以及经公用事业公司或企业背书的购电结构变现。
阶段
Late-stage private commercialization / pre-revenue deployment
融资情况
公开证据显示,Hermes 在 $629 million 项目结构内获得最高 $303 million 的 DOE 支持,另有未披露的私人资本。留存来源没有公开当前股权结构表、现金或估值。
[CO001, CO002, CO004, CO005, CO006, CO007, CO009, CO011]

执行摘要

主要优势

  • 首个商业化验证已经出现:Google 最高 500 MWe 总协议,加上 Hermes 2 面向 TVA 的单项目落地路径。
  • 公私部门降风险组合罕见地厚,覆盖 NRC 施工推进、DOE 里程碑支持,以及可见的燃料 / 制造生态。
  • KP-FHR 架构差异化,迭代式试验学习开发模型也清晰;如果里程碑守住,公司有望打出成本和安全优势。

主要风险

  • Hermes 工期或预算若继续滑坡,估值会很快受压,因为投资逻辑高度依赖里程碑。
  • 商业化 HALEU 和 TRISO 供应正在改善,但仍未完全成熟;燃料链时间表仍是真瓶颈。
  • 公开证据没有披露当前估值、股权结构表、合同定价或现金跑道,入场价格难以评估。

未决问题

  • 当前普通股估值标记、优先股堆叠、稀释条款,以及是否有结构化下行保护。
  • Google/TVA 合同的完整经济性、期限、风险分配、抵押安排和终止权。
  • 披露项目延期后的 Hermes 最新预算、备用金使用和修订后时间表。
  • 从 DOE HALEU 配额到后续机组可规模化商业 TRISO 燃料供应的衔接路径。

目录

Chapter 01

01公司概况

1.1 身份定位、使命与全国布局

Kairos Power 把自己定位成一家使命驱动的核技术、工程和制造公司,重点是商业化一个反应堆家族,而不是运营多元化发电资产组合。留存的官方材料一致显示,公司总部在 California 的 Alameda,并形成三点运营布局:Alameda 负责设计、原型和盐实验室工作;Albuquerque 负责制造、盐生产和燃料开发;Tennessee 的 Oak Ridge 承载 Hermes 反应堆示范园区。因此,公司当前阶段更适合理解为后期商业化,而不是成熟公用事业运营:它在造硬件、验证建设和许可模式,并在拥有可产生收入规模的运营机组前签下未来产出合同。 Kairos 称,公司的使命是通过商业化 KP-FHR 设计,交付清洁、可负担、可靠的稳定电力。公司反复把市场窗口放在老化燃气机组、可调度无碳发电需求,以及数字基础设施带来的电力需求增长上。这个叙事很关键,因为它解释了为什么 Kairos 早早把反应堆开发和面向客户的商业化结构绑在一起,而不是把第一份客户合同当成许可之后的事件。布局、选址和使命表述指向同一个运营判断:把学习集中在少数可控地点,掌握关键制造环节,并在监管和交易对手允许的速度内,尽快从示范走向可复制的舰队式部署。[CO001, CO002, CO003, CO004, CO009, CO028]

Kairos Power 快照 KPI 表
指标数值 / 状态日期 / 资料时点置信度缺口 / 备注
成立20162016由官方关于页面材料和 2024 年公司标准简介支持
总部加州 Alameda2026-07官网还显示新墨西哥州制造园区和田纳西州反应堆园区
当前阶段后期商业化 / 商业部署准备2026-07正在建设示范项目并签署未来产出合同,但尚未运营商业机队
全国布局Alameda 总部;Albuquerque 制造园区;Oak Ridge 反应堆园区2026-07官网显示三地布局
已披露员工数全国 500+ 名团队成员2026-07官方田纳西地点页面给出唯一保留的公开员工数信号
核心反应堆产品KP-FHR;双机组电厂最高 150 MWe2026-07最小商业配置为 2 x 75 MWe
Hermes 135 MWth 非发电示范反应堆2025-03监管与 DOE 来源均确认其不发电
Hermes 2首个 TVA 电网部署,目标最高 50 MW 输出2026-04当前 50 MW 表述取代早期 20 MWe 示范配置
Google 订单簿2035 年前 500 MW;首次部署目标为 2030 年2024-10与 Google 的主电厂开发协议
DOE 支持最高 $303M、按里程碑拨付的技术投资协议2024-02支持 Hermes 设计、建设和调试
公开估值未公开披露2026-07保留的公开市场数据页面披露轮次元数据,但没有可核验的实时投后估值
收入 / ARR / 客户数未公开披露2026-07私营公司,没有保留的公开运营指标披露

快照整合官方、监管和独立报道。公开指标在估值和运营表现上仍不完整。

[CO001, CO002, CO003, CO009, CO012, CO014]
FO002: Kairos Power 公司快照逻辑

流程图展示 Kairos Power 如何把一个反应堆平台、受控制造、公共支持和客户 PPA 串成商业化系统。

1.2 创始人、领导层与治理依赖

公开证据支持一个扎根 UC Berkeley 先进反应堆研究生态的创始故事。Berkeley 来源把 Kairos Power 的起源与 Mike Laufer、Ed Blandford 和 Per Peterson 关联起来;当前 Kairos 材料则把 Laufer 列为 CEO 兼联合创始人,把 Blandford 列为 CTO。这个组合对尽调很重要,因为公司的公开身份仍高度依赖创始人技术背书:领导层可信度来自深厚的反应堆物理、许可和技术开发能力,而不是既有核电开发商常见的长期运营历史或受监管公用事业资产负债表。Jeff Olson 在商业公告中也很突出,说明随着客户签约变得更核心,Kairos 已补上一个可见的商务拓展声音。 治理图景只公开了一部分。Kairos 提供当前运营领导层名单,但留存来源没有给出公开发行人或项目融资型公用事业关联方通常会披露的董事会、所有权或委员会透明度。这种不透明不等于治理薄弱,但确实带来关键人物集中风险:Laufer 和 Blandford 仍然支撑技术可信度、监管信任、合作伙伴信心和资本形成。好处是叙事一致——一个反应堆家族、一个运营模式、一套信息。代价是公开尽调无法充分测试接班梯队、独立董事会约束力,或创始人与投资人、政府相关资金伙伴之间的战略控制权分配。[CO005, CO006, CO007, CO008, CO035]

领导层与创始人表
人物当前或公开角色类型有证据支持的背景关键人物 / 治理备注
Mike Laufer首席执行官;联合创始人创始人 / 高管Google、NRC 和商业化公告中的公开代表;创始人根源与 Berkeley 有渊源技术、监管和客户叙事中的核心外部可信度锚点
Ed Blandford首席技术官;联合创始人创始人 / 高管与 Berkeley 反应堆研究有渊源的联合创始人;在 ORNL 合作中被引用设计成熟度和公共部门合作的关键技术负责人
Jeff Olson业务发展与财务副总裁高管商业需求信号和客户开发逻辑的主要发言人技术进展与可融资承购叙事之间的重要转译者
Per PetersonBerkeley 研究负责人;被引用的技术源头创始人关联 / 顾问UC Berkeley 教授,聚焦高温反应堆、熔盐和许可公开的创始来源证据更能支撑技术根源,而不是当前治理权威
Linda Schenk财务运营高管Kairos 关于页面将其列为运营领导层成员财务负责人公开可见,但披露深度不及上市公司
Micah Hackett / Alan Kruizenga / Ravi Singaraju 等技术负责人燃料与材料 / 盐 / 工程设计与集成职能负责人具名负责人覆盖核心反应堆子系统和制造学科职能覆盖看起来较广,但公开来源没有提供继任或董事会监督细节

领导层列表仅反映公开运营角色。保留来源没有提供完整公开董事会治理资料包。

[CO005, CO006, CO007, CO008]

1.3 技术平台与商业化模式

Kairos Power 的商业故事离不开它的技术架构。KP-FHR 把熔融氟盐冷却剂和 TRISO 球燃料结合在一个高温、低压设计里;公司认为,这套设计既能带来固有安全性,也能比大型水冷反应堆降低建设成本。当前技术页面显示,最小商业电站由两个 75 MWe 机组组成,总计 150 MWe;Hermes 和 Hermes 2 则逐级示范许可、建设、燃料处理和供应链执行。换句话说,Kairos 不只是在证明反应堆物理可行,也在证明开发流程、制造栈和项目交付模式能足够重复,进而让未来客户放心。 这也是为什么迭代开发和垂直整合几乎贯穿所有留存的公司来源。Kairos 使用 ETUs、盐生产、燃料开发实验室和内部制造,不是把它们当副项目,而是把它们当作降本策略的运营骨架。这是对先进核能经典难题的成熟回答:首座电站很贵,第二座只有在开发商掌握足够多的学习回路、能吸收进度、制造和监管经验时才会变便宜。因此,商业模式混合了公共成本分摊、内部控制的制造能力,以及后续面向大型电力买家的 PPA。这套模式说得通,但资本开支也很重;学习链条上任何一环卡住,进度都会受拖累。[CO010, CO011, CO012, CO013, CO027, CO029]

利益相关方或投资者地图
利益相关方角色重要性公开证据信号尽调问题
Google锚定客户 / 开发伙伴创造需求信号和到 2035 年 500 MW 订单簿MPDA 与 Google 博客确认未来 PPA 和 2030 年首次部署目标审查里程碑触发条件、价格调整机制和终止条款
TVA公用事业电网交易对手将 Hermes 2 变成明确的首次部署,具备电网交付和清洁属性Kairos 称 TVA 签署了美国首份公用事业先进反应堆 PPA确认电价、承购和平衡责任
美国能源部成本分担出资方为 Hermes 提供按里程碑拨付的开发资本公开披露最高 $303M TIA索取完整付款计划和剩余里程碑
Oak Ridge National Laboratory技术伙伴提供商业化所需的燃料、材料、制造和乏燃料工作公开披露 2026 年五年期 $27M 合作澄清交付物、IP 处理和闸门依赖
KP-OMADA 联盟公用事业 / 发电商商业化联盟在许可、制造和机队部署上引入行业参与公司官网方法页称其为首创联盟确认具名成员、义务和治理
Breakthrough / Khosla / Prelude 及其他已披露投资者私人资本基础表明长期气候科技资本支持Caplight 披露轮次和投资者元数据,但未披露持股比例索取股权结构表、轮次规模和优先权堆叠
Oak Ridge 社区和田纳西州政治利益相关方本地选址与许可群体对劳动力、场地支持和区域部署动能很重要Kairos、ANS 和官方声明强调本地合作评估土地使用、劳动力和社区利益承诺
未来工业和数据中心买方扩张需求池决定 Kairos 能否从首座电厂走向系列化机队部署Google 先例指向更广泛的大负荷客户逻辑测试 Google 之外客户管线的广度

投资者行刻意保守,因为保留来源列出了投资者,但没有披露所有权、董事会权利或逐轮定价。

[CO023, CO025, CO026, CO029, CO030, CO035]
FO003: Kairos Power 快照 KPI

截至本次报告日期,概括 Kairos Power 商业化状态的关键成熟度和披露指标。

1.4 里程碑、资本信号与执行背景

Kairos 已远远不只是概念阶段信号。对一家私营先进反应堆公司来说,它的里程碑链条异常具体:2023 年 12 月 Hermes 获得许可批准;2024 年 2 月 DOE 签下基于里程碑的 Technology Investment Agreement;2024 年 7 月 Hermes 开工;2024 年 10 月达成 Google 500 MW 协议;2024 年 11 月 Hermes 2 获得许可;2025 年启动核安全相关建设;2026 年初与 ORNL 展开燃料和材料合作;2026 年 4 月 Hermes 2 破土动工。随后 Google-Kairos-TVA 的安排尤其关键,因为它把泛泛的未来订单储备转成一个明确的首个部署:最高 50 MW 接入 TVA 电网,并直接服务 Google 在 Tennessee 和 Alabama 的数据中心需求。 这些里程碑显著降低了公司风险,尤其是相较于只有纸面设计或政策支持的先进反应堆开发商。即便如此,公开记录也保留了关键限定:首堆先进核能项目仍面对不确定的前期建设成本、漫长周期和赞助方犹豫。TechCrunch 明确指出,在核能语境下,2030 年首座反应堆和 2035 年舰队时间表都很激进;Utility Dive 对 NIA 的报道也强调,早期项目赞助方在完全承诺前仍需要更多兜底。上市公司式资本透明度也还没有出现。留存的市场数据页面披露了投资人名称和融资轮次时间线,但没有可验证的当前投后估值、收入运行率或完整股权结构表。公司有势能,但还没有完整的公开财务可见度。[CO014, CO015, CO016, CO017, CO018, CO019]

里程碑表
日期里程碑类型金额 / 输出 / 状态参与方含义
2016公司成立创立已成立Laufer、Blandford、Berkeley 关联创始人Kairos 商业化论证的起点
2021-09-29Hermes 建设许可申请提交监管已提交申请Kairos、NRC启动首座示范反应堆的正式许可路径
2023-12-12DOE 发布 Hermes 许可批准说明监管许可已批准DOE、NRC、Kairos确认超过 50 年来首个非水冷反应堆许可
2024-02-21技术投资协议签署融资最高 $303MDOE、Kairos为 Hermes 路径提供按里程碑拨付的公共资金
2024-07-30Hermes 1 开工建设规模化场地建设进行中Kairos、Oak Ridge 利益相关方将项目从纸面推进到实体执行
2024-10-14Google MPDA 公布合作2035 年前 500 MW;首座电厂目标 2030 年前Kairos、Google为先进核能创造标志性企业订单簿
2024-11-20Hermes 2 许可获批监管两台 35 MWth 机组获许可NRC、Kairos打开后续发电示范路径
2025-05Hermes 核安全相关建设启动规模化核建设阶段启动Kairos显示场地工作后进度继续执行
2025-08-18Google/TVA/Kairos 部署合作公布合作Hermes 2 在 TVA 电网上最高 50 MWGoogle、TVA、Kairos将首座电厂转为并网客户部署
2026-02-23ORNL 燃料与材料合作公布技术五年期 $27M 合作ORNL、Kairos增强燃料、材料和乏燃料准备度
2026-04-17Hermes 2 奠基规模化Kairos 首个商业规模反应堆场地工作Kairos、DOE、田纳西州利益相关方释放向首次发电部署迈进的信号

公开证据只保留月份或年份时,表格保留该粒度,不臆造具体日期。

[CO001, CO015, CO016, CO018, CO019, CO020]
FO001: Kairos Power 里程碑时间线

这条公开里程碑时间线展示 Kairos Power 如何从 2016 年创立故事,走到拥有 DOE 支持、NRC 许可和 Google 牵头商业化订单储备的反应堆开发商。

1.5 展示

Chapter 02

02市场分析

2.1 市场边界与现状替代方案

分析 Kairos Power,不能停在模糊的“全球核能市场”大标题上。公司卖进的是一个更窄的清洁稳定电力问题:特定公用事业公司、超大规模云厂商,以及最终的工业场址,需要可靠的无碳电力,有时还要搭配辅助服务或工艺热,而且这些地区的电网供给正在收紧。EIA 的 2026 年展望在这里很重要,因为它显示美国电力需求在长期平台期后重新增长,数据中心被列为主要驱动因素。负荷重新加速,先进核能才从政策概念变成采购议题。 这个边界也说明哪些东西应该排除在外。Kairos 还不是在竞争所有传统核电替代、所有商用批发电力新建项目,或每一条泛气候科技预算线。它的实际竞争对手,是同一服务区域内能完成相同可靠性任务的技术组合:天然气、可再生能源加储能、需求响应、现有核电延寿或增容,以及其他清洁稳定电力选项。尽调的正确问题不是“核能市场大不大”,而是“在哪些场景里,150 MWe 级 Kairos 部署最像一个紧迫电力问题的可信解法”。[CM001, CM002, CM008, CM027, CM032, CM037]

市场定义表
细分 / 类别包含支出排除支出买方 / 付款方关联度
超大规模云厂商清洁稳定电力绑定特定数据中心负荷的 24/7 无碳电力长期 PPA、辅助服务和项目开发通用 REC 购买、短时备份、非稳定批发电力超大规模云厂商能源与基础设施团队当下最可见的验证点来自 Google;也是对签约新先进核能供应的迫切意愿最强的证据
公用事业或公共电力服务区域负荷增长地区的并网反应堆项目、资源规划、互联、选址和可靠容量没有锚定需求的纯商用电站、传统大型反应堆替代项目公用事业 IRP、发电规划和公共电力采购团队关键在于 Kairos 部署必须落在可操作的服务区域里,而不是抽象的全国市场
工业供热与供电园区先进反应堆可替代化石燃料热并提供稳定电力的供电 + 工艺热场景纯住宅零售供电和无关制造资本开支工业能源、运营和脱碳预算DOE 项目材料支持的相邻市场,但还不是 Kairos 主要公开滩头
核电和煤电场址再利用机会利用既有场址、基础设施和过往许可熟悉度的项目没有本地电网或社区逻辑的绿地囤地开发商,加上东道社区和公用事业重要的供给侧市场筛选器,因为选址可行性决定哪些需求真正能被服务
现状替代方案燃气电厂、可再生能源 + 储能、需求响应,以及解决同一可靠性任务的既有核电延寿电力可靠性之外的不相关气候科技类别同上终端买方比宽泛核能 TAM 营销更准确地界定 Kairos 的真实竞争

围绕清洁稳定电力采购界定可投资市场,而不是全球核能总支出。

[CM001, CM013, CM017, CM032, CM039]

2.2 规模测算视角:从广义 TAM 到 Kairos SAM

最宽但仍有用的市场视角,是 DOE 估计美国到 2050 年可能需要新增 550-770 GW 清洁稳定电力;如果先进核能在 2030 年前开始部署并快速扩张,可能贡献约 200 GW。这就是行业级 TAM 逻辑:可靠性缺口很大,先进核能可以填上一部分。IEA 和 EIA 数据强化了需求侧,因为它们显示数据中心负荷增长已经不再只是猜想。IEA 预计 2030 年全球数据中心需求将达到 945 TWh,并称美国数据中心到 2030 年可能贡献近一半电力需求增长;这有力说明,关键地区新增稳定发电不是可选项。 但 Kairos 自己的 SAM 更窄。公开证据显示,公司瞄准的是超大规模云厂商或公用事业需求能够长期锚定多机组电站的服务区域;起点是 2030 年前后的首个商业部署,以及到 2035 年最高 500 MW 的 Google 订单储备。场址复用会扩大漏斗——DOE 认为现有核电场址有 60-95 GW,煤电场址有 128-174 GW——但这些只是技术包络,不是已签需求。因此,Kairos 真正的 SOM 仍应按少数可融资的早期项目衡量,而不是按完整 200 GW 先进核能雄心衡量。[CM003, CM005, CM006, CM009, CM010, CM011]

TAM/SAM/SOM 或规模测算视角表
发布方年份地理范围数值CAGR / 时点方法置信度局限
DOE Pathways / Five Charts2024-2025美国到 2050 年新增 550-770 GW 清洁稳定电力到 2050 年的需求自上而下测算净零兼容稳定容量的电力系统需求行业整体需求,不是 Kairos 特定可触达收入
DOE Pathways / Five Charts2024-2025美国到 2050 年约 200 GW 先进核能容量假设 2030 年前开始部署,并在 2040 年前爬坡到 13 GW/年基于情景的先进核能部署路径需要快速降本和执行,而这些尚未在规模上得到证明
IEA Energy and AI 报告2026全球到 2030 年约 945 TWh 数据中心电力需求到 2030 年翻倍以上数据中心全球用电预测需求指标,不是 Kairos 的直接合同支出
IEA Energy and AI 报告2026美国到 2030 年,数据中心推动近一半电力需求增长到 2030 年美国新增电力需求份额份额估计没有说明多少由核能服务
DOE 选址分析2024美国现有或已退役核电站点 60-95 GW技术潜力站点占地与适配性筛选技术选址边界,不是已融资项目管线
DOE 选址分析2024美国煤电站点 128-174 GW技术潜力煤电转核电复用筛选技术选址边界,不是有购电协议支撑的需求
Google / Kairos2024美国到 2035 年最高已锁定 0.5 GW首个部署目标为 2030 年前具名开发协议 / PPA 支撑的订单簿公开锚定客户只有一个;价格未披露

需要多种规模测算口径,因为没有一个公开市场估计能干净隔离出 Kairos 专属的 SAM 或 SOM。

[CM003, CM005, CM006, CM009, CM010, CM011]
FM001: 市场规模测算视角

三层市场视角从美国清洁稳定电力需求,推进到先进核能容量潜力,再到 Kairos 当前公开点名的锚定订单储备。

为便于比较,数值均以 GW 表示。TAM 使用 DOE 区间中点。SOM 代理不是 Kairos 总市场份额;它只是当前公开点名的锚定订单储备。

[CM006, CM009, CM010, CM028, CM029]
FM002: 市场估算区间

以区间方式展示美国可复用场址容量包络,这些场址可能承接先进反应堆,说明选址可行性会实质性放大或收窄部署漏斗。

合并行只是已报告上下限的简单加总。这些是技术选址包络,不应视为已锁定需求,也不应视为 Kairos 独占机会。

[CM011, CM012, CM025, CM038]

2.3 买方、用户与付款方分层

最重要的客户分层洞察是,使用 Kairos 电力的一方和为 Kairos 项目付费的一方并不总是同一方。在 Google 协议里,超大规模云厂商是经济需求赞助方和关键承购方,但项目仍需要反应堆开发商、承载公用事业公司或服务区域伙伴,以及可行的并网地理位置。随后的 Google-Kairos-TVA 合作进一步说明了这一点:采用路径不是简单“卖一座反应堆”,而是“组装一个项目栈”,其中每个利益相关方都控制着不同的门槛资源。 这种结构形成了一张分层买方地图。超大规模云厂商带来迫切的负荷增长,并愿意签署长期清洁能源协议;公用事业公司和公共电力主体带来选址、电网整合和地方许可关系;工业客户是合乎逻辑的邻近市场,因为 DOE 自身的先进反应堆材料强调工艺热和负荷匹配。对 Kairos 来说,早期市场按项目推进、重关系。赢单靠的是对齐预算所有者、场址承载方、监管者、燃料来源和终端负荷,而不是往一个抽象电力市场里报一个平准化成本数字。[CM006, CM007, CM016, CM017, CM018, CM019]

细分市场 / 买方地图
细分市场买方使用方付款方工作流预算负责人采用触发因素
超大规模云厂商数据中心超大规模云厂商能源 / 基础设施团队数据中心园区和 AI 计算负载长期能源协议下的企业买方确定服务区域 -> 签署开发 / 购电协议 -> 协调选址和并网 -> 接收电网电力与属性企业能源、可持续发展和基础设施预算24/7 无碳电力需求叠加负载快速增长
公用事业或公共电力合作方公用事业发电规划团队或公共电力机构服务区域内电网和本地客户公用事业资产负债表、费率基数工具或结构化项目协议宿主站点评估 -> IRP / 规划对齐 -> 并网和电网集成 -> 项目支持发电规划和资本配置委员会增长服务区域需要可靠容量
开发商主导的示范路径Kairos 对接 DOE / NRCHermes / Hermes 2 学习站点和后续电站混合:DOE 支持、私人资本、客户背书合同示范反应堆和燃料 -> 验证许可与建设 -> 转化为可复制的商业电站Kairos 管理层和融资组合证明首批电站能拿到许可、建成并获得燃料
工业热能和电力用户大型工业能源买方需要热能和电力的工厂运营工业资本开支 / 能源采购评估脱碳需求 -> 判断热电匹配度 -> 谈判站点专属项目结构工厂运营和企业脱碳预算替代方案不足时,需要可靠热能 / 电力
区域经济发展或园区集群买方和本地合作伙伴组成的财团混合商业和电网负载混合项目融资和锚定合同聚合负载 -> 对齐本地站点和公用事业合作方 -> 签约分阶段部署锚定买方加区域合作伙伴单一买方无法独自支撑首批机组部署时

Kairos 早期商业化看来需要多方项目组合,而不是简单卖设备。

[CM016, CM017, CM018, CM019, CM033, CM036]
FM003: 买方 / 客群地图

关系图展示 Kairos 部署如何取决于反应堆开发商、锚定购电方、公用事业 / 供电服务区承载方和终端负荷协同。

[CM016, CM017, CM018, CM033, CM036, CM039]

2.4 增长驱动、瓶颈与采用节奏

市场顺风是真实的。数据中心需求在上升,公用事业公司在修正负荷假设,DOE 正在发布明确的先进核能商业化路径,Google 这类买方也证明先进核能可以进入 24/7 清洁能源采购。这些因素构成了该行业多年来最强的需求环境。它们也利好 Kairos 的迭代商业化模式,因为公司可以指向具体承购逻辑,而不是泛泛的脱碳口号。 但卡点仍然很硬。HALEU 供应有限,NRC 人手依然是有文件记录的许可瓶颈,同行反向证据——尤其是 NuScale 失败的 UAMPS 项目——说明即便政策支持很强,第一波经济性和承购结构也可能断裂。公开来源也没有披露 Kairos 具体 PPA 定价,或经过验证的首批舰队成本曲线。因此,核心不确定性已经不再是需求是否存在;而是 Kairos 和整个行业能否在客户公告写入的 2030-2035 时间表内,把需求转成可复制、可融资的项目。[CM019, CM020, CM021, CM022, CM023, CM024]

增长驱动因素和约束表
驱动因素 / 约束方向时间影响尽调问题
美国电力需求重新增长正向当前至 2050 年展望提高采购新增可靠发电的意愿,而不是只依赖能效提升或负载持平假设Kairos 管线中,哪些具体服务区域的备用容量裕度最紧、数据中心增长最快?
AI 和数据中心负载扩张正向当前至 2030 年催生迫切的大负载买方;他们看重 24/7 清洁电力,可能签长期协议除 Google 外,还有多少超大规模云厂商潜在客户正在实质讨论?负载规模分别多大?
DOE 商业化和选址支持正向当前提供政策可信度、需求聚合逻辑和可落地选址路径哪些 DOE 项目能直接给 Kairos 前五个项目降风险,而不只是改善行业观感?
负载匹配和工艺热灵活性正向中期将可服务场景从通用基荷电力拓宽到更多用途哪些非数据中心细分市场真正近在眼前,哪些只是战略相邻市场?
HALEU 供应受限负向当前至早期机队建设即使客户需求存在,也可能拖慢部署Hermes 1 之后的项目由哪些燃料配额或供应合同覆盖?
NRC 人员配置和许可吞吐量负向当前无论客户需求多强,都可能拖慢行业扩张Kairos 的 2030 年目标建立在哪个排队位置和哪些审查资源假设上?
FOAK 资本成本和融资风险负向当前至首批商业电站可能击穿公用事业购电结构,NuScale 已经示范过Kairos 首个商业站点由哪些固定价 EPC、预备费和业主成本假设支撑?
订单簿集中在一个旗舰买方负向近期让公开需求证明强但狭窄除 Google / TVA 挂钩项目外,客户管线有多分散?

关键问题是能否按计划把需求转化为可融资、可供燃料、可获许可的项目。

[CM019, CM020, CM022, CM023, CM024, CM026]
FM004: 采用漏斗 / 价值链地图

价值链图显示,先进核能要落地,客户需求、许可、燃料、场址就绪和项目执行必须同步推进。

[CM020, CM022, CM024, CM026, CM034, CM035]

2.5 展示

Chapter 03

03竞争格局

3.1 格局分层:同行、既有厂商与替代方案

Kairos 并不是在一个没有差异的“SMR 市场”里竞争。真实格局按买方任务拆开。直接先进反应堆同行包括 X-energy、TerraPower、NuScale、Holtec、GE Hitachi、Westinghouse、Terrestrial Energy,以及 Last Energy 这类更小的封装型进入者。邻近叙事包括微反应堆和聚变。现状替代方案同样重要:当买方更看重监管熟悉度、更大的电力块,或可见的建设案例,而不是更新颖的架构时,更大或更传统的核能路径也能赢。 这套分层很重要,因为采购团队买的不是抽象的反应堆类别,而是一组具体组合:兆瓦级电力块、场址匹配、许可路径、燃料容忍度、商业交易对手和进度信心。在部分维度上,Kairos 确实差异化明显,尤其是在分阶段部署和具名超大规模云厂商需求重要的场景里。在另一些维度上,它只是多个可信选项之一。因此,正确的竞争问题不是 Kairos 是否“最好”,而是相对于替代方案,它的规模、热力学和商业化路径组合最适合哪些客户任务。[CP001, CP002, CP014, CP017, CP033]

竞争对手画像表
竞争对手类别公开规模 / 资本信号目标买方差异化当前公开限制
Kairos Power研究对象 / 先进高温同业150 MWe [2 x 75 MWe] 商业电站;Google 到 2035 年最高 500 MW;Hermes 许可超大规模云厂商、服务区域公用事业、分阶段早期采用者氟化物盐高温设计,加上迭代式商业化和具名超大规模云厂商验证商业机队经济性和可复制性仍未公开验证
X-energy Xe-100直接同业 / HTGR80 MWe 模块、320 MW 四模块组合、Amazon 和 Energy Northwest 验证工业热能、大型园区、公用事业565°C 蒸汽、TRISO-X、高温工业定位第一波商业交付尚未到来
TerraPower Natrium直接同业 / 钠冷快堆345 MWe 设计、DOE 最高 $2B 50/50 成本分担、PacifiCorp 后续研究受监管公用事业、煤电站点替代、大型清洁负载增长更大的公用事业级电力模块和清晰的公私合作部署路径HALEU 和 FOAK 复杂度仍是实质风险
NuScale NPM已获许可的 LWR-SMR 同业77 MWe 模块、924 MWe 12 模块配置、公开 10-K 披露公用事业、数据中心、工艺热、氢最强的正式 NRC 设计批准信号和标准燃料CFPP 终止,且未披露有约束力的客户交付合同
GE Hitachi BWRX-300现有核电厂商 LWR-SMR 同业300 MW 级,具备 Darlington 和 TVA 参考路径公用事业和受监管项目西方公用事业参考路径和沸水堆技术谱系在热能和小规模分阶段部署上差异化较弱
Holtec SMR-300现有核电厂商 LWR-SMR 同业双机组 ~600 MW 电站边界公用事业、再供电站点、大型承购方PWR 熟悉度、被动安全特性、24/7 清洁电力定位更大电力模块不太贴合 Kairos 规模的区域增量
Rolls-Royce SMR公用事业级相邻同业每座电站 470 MW;英国竞赛获胜宣传国家级公用事业和大型电网规划方大型单站输出和强监管进展营销规模和部署方式与 Kairos 当前切入点差异明显
Last Energy / eVinci 项目微型 / 现场封装进入者PWR-20 现场模式或 5 MWe 微堆偏远、工业和表后买方工厂化交付和更小的现场封装与 Kairos 并网 150-MWe 级电站服务的买方任务不同

各行混合了产品规模、资本信号和客户验证,因为同业之间可比的实际定价没有公开披露。

[CP001, CP003, CP005, CP007, CP009, CP012]
FP001: 竞争定位图

定性等级图,把买方任务差异化与公开就绪度信号放在一起比较。

x 轴为买方任务差异化(1-5),y 轴为公开就绪度 / 证明(1-5)。评分概括 TP001-TP004 的证据,而不是量化市场份额。

[CP003, CP005, CP007, CP009, CP012, CP014]

3.2 能力与封装比较

Kairos 的公开方案很有辨识度,但并不在所有用例上占优。150 MWe 双机组商业配置、氟盐冷却、TRISO 球燃料和迭代开发模式,让它比部分更大型同行更容易讲清楚分阶段商业化故事。这支撑了那些想先下注服务区域级增量、或想在多 GW 舰队建设前拿到具名 24/7 清洁电力合同的买方。它也天然匹配 Google 多机组订单储备,以及随后与 TVA 关联的叙事。 但竞争对手各有强项。X-energy 在工业蒸汽和高温热上表述更明确。TerraPower、Holtec、GE Hitachi、Westinghouse AP300 和 Rolls-Royce 都呈现更大的公用事业级方案,或更熟悉的水堆血统。NuScale 仍保有最强的正式许可资质,但也背着更可见的商业采用伤痕。所有公司公开定价都很稀疏,所以能力比较只能靠规模级别、热 / 电输出特征、客户验证和监管姿态,而不能假装有可用的每兆瓦时成本表。这种不透明本身就是尽调结论。[CP003, CP004, CP005, CP007, CP009, CP010]

功能 / 能力矩阵
购买标准KairosX-energyTerraPowerNuScaleGEH / Holtec / Westinghouse AP300Rolls-Royce / 微型封装
分阶段 100-150 MWe 级部署低-中低 / 仅微型封装为高
工业热能 / 高温蒸汽低-中
超大规模云厂商验证低-中
公用事业级电力模块
标准燃料简单性混合
可见的监管或建设验证
现场 / 表后封装低-中

序数标签只概括公开证据,并刻意避免编造成本排名。

[CP003, CP005, CP007, CP009, CP014, CP021]
定价 / 封装对比
竞争对手公开定价可见度商业封装信号交易对手验证买方风险转移线索尽调含义
Kairos没有公开 PPA 价格或 LCOE开发商主导的多机组 PPA 和服务区域合作Google + TVA 挂钩公开信号公开来源未说明建设延期如何分配需要条款清单、资本开支模型,以及谁承担进度风险
X-energy没有公开交付价格披露供应商加合作伙伴项目,以及工业 / 公用事业开发路径Amazon、Energy Northwest、Dominion 探索价格升级和 EPC 风险转移未知需要 MW 和模块数量之外的项目经济性
TerraPower没有公开电价披露大型 DOE 支持示范,加公用事业规划通道PacifiCorp 和公私合作融资信号可见部分政策降风险,但合同经济性不可见需要公用事业级经济性和燃料风险假设
NuScale上市公司风险披露,但没有客户价目表模块化 LWR 产品,有公开文件10-K 中未披露有约束力的客户交付合同CFPP 后商业可行性仍承压将监管强度与客户转化分开看
GEH / OPG已审阅材料中没有公开价目表公用事业主导的建设参考路径Darlington 和 TVA 路径参考项目可信度强于价格透明度需要商业合同细节,判断客户侧经济性
Holtec / Westinghouse AP300已审阅页面中没有公开项目级价格面向大型承购方的公用事业封装品牌和技术可信度,但此处具名客户细节有限业主成本和 EPC 条款未知需要实际交易对手和融资结构
Last Energy / eVinci 项目没有直接可比的能源价格现场或微型封装电力模式商业模式本身就是差异化的一部分买方负担可能更低,但经济性未披露需要表后定价和所有权条款

缺少可比的公开定价本身就是一个核心竞争发现。

[CP010, CP011, CP030, CP031, CP034, CP037]
FP002: 功能广度 / 能力图

没有竞争对手能覆盖所有买方任务;在分阶段部署和超大规模云厂商验证重要的场景,Kairos 得分最高。

定性等级概括 TP002 的证据,并有意将缺乏支撑的成本单元格排除在范围外。

[CP020, CP021, CP022, CP023, CP026, CP027]

3.3 切换成本、供应姿态与护城河耐久性

Kairos 的护城河不是单一反应堆特征,而是一个组合:迭代测试学习模式、Hermes 的先发许可进展、具名 Google 订单储备,以及与 TVA 相连的电网伙伴故事。对想分阶段部署先进核能、而不是直接跳到超大型电站的买方来说,这个组合可能很有价值。但护城河并不排他。X-energy 现在有 Amazon 和 Energy Northwest;TerraPower 有 PacifiCorp 和 DOE 背书的大型参考路径;GE Hitachi 有 Darlington 和 TVA 势能;LWR 同行保留了标准燃料的熟悉度。 在买方承诺具体场址和许可路径前,多栖下注仍是真实威胁。一旦交易对手选定场址、公用事业伙伴、监管路线和燃料假设,切换成本会急剧上升。但在那之前,大型电力买方可以并行比较多条核能路径。这也是为什么整个品类获得客户验证,反而会削弱泛泛“核能服务数据中心”定位的价值。Kairos 的竞争耐久性来自把当前组合转成可复制部署,而不是假定竞争对手无法触达同一批客户。[CP016, CP018, CP024, CP025, CP028, CP029]

护城河耐久性 / 竞争风险登记表
护城河主张威胁严重性缓解措施 / 尽调问题
Google 主导的需求证明买方在多个核能供应商之间多栖检验 Kairos 是否掌握独家客户工作流,还是只有品类验证
迭代开发和 Hermes 学习路径同业先建出更强的公用事业级参考项目将 Hermes 到机队的学习曲线,与 Darlington、Natrium 和 X-energy 部署里程碑对比
高温架构X-energy 拥有最清晰的工业蒸汽叙事判断哪些客户任务需要 Kairos 热力学,而不是 X-energy 蒸汽规格
分阶段电站规模买方需要 300-470 MW 模块时,更大的同业胜出按站点负载和备用容量裕度需求绘制目标管线
新型燃料 / 冷却剂差异化在风险委员会眼里,标准燃料 LWR 同业看起来更简单评估每个目标客户对先进燃料和新型冷却剂的接受度
服务区公用事业合作既有公用事业公司在棕地或 IRP 上可能更有筹码核查电网合作伙伴深度、并网队列状态和地方政治支持
核能和 AI 负载带来的品类热度CFPP 这个反向样本显示,经济性仍可能失守索取首批商业站点的交付成本、价格升级和预备费假设
聚变与微反应堆相邻赛道资本和叙事关注可能分散买方 / 投资人焦点把近期客户替代方案与长周期叙事型竞争对手分开评估

护城河能否守住,取决于把当前证明点转化为可复制的项目交付,而不只是靠品类叙事。

[CP016, CP017, CP028, CP029, CP032, CP033]
FP003: 护城河 / 就绪度 KPI

最有决策价值的公开竞争 KPI,是产品单元规模、锚定客户证明和参考路径可见度。

[CP003, CP004, CP005, CP007, CP010, CP015]

3.4 竞争定位的尽调结论

不应把 Kairos 按先进核能赛道的通吃赢家来承销。更强的投资逻辑要窄一些:在具名大负荷客户、监管上可见的示范路径和可管理电站规模,比单一场址最大输出更重要的服务区域里,Kairos 是最可信的分阶段清洁稳定电力供应商之一。这个楔子可能非常有吸引力,尤其适合数据中心邻近或区域公用事业场景。 反面同样重要。对工业蒸汽买方、大型公用事业块状电力采购,或主要看重标准燃料和熟悉水堆谱系的客户来说,Kairos 并不显然是最佳答案。聚变也会争夺想象力和资本,哪怕它不争夺同样的近期部署。实际尽调任务是把 Kairos 放到买方具体替代方案旁边测试——热需求看 X-energy,规模看 TerraPower 或 Rolls-Royce,熟悉度看 GE Hitachi 或 Holtec,现场封装看 Last Energy 或 eVinci——而不是拿它对比一个泛化核能 TAM。这个框架既保留 Kairos 的强项,也保留它的竞争边界。[CP017, CP019, CP021, CP027, CP038, CP039]

3.5 展示

Chapter 04

04财务情况

4.1 收入模式与变现

Kairos 的财务模型现在更适合看作开发阶段桥梁,而不是运营中的核电公用事业 P&L。留存公开来源没有显示公司今天已有商业反应堆收入。相反,近期经济故事由 DOE 里程碑付款、私人资本、合作伙伴背书的开发工作,以及把原型和许可逐步转成未来项目收入组成。Google 公告尤其重要,因为它说明 Kairos 最终想变现的不只是交付反应堆,还包括 PPA 下的能源、辅助服务和环境属性。TVA 合作又给出第二条线索:最终商业模式更像服务区域和项目结构驱动,而不是简单硬件销售。 不过,今天的公开变现质量仍然不高。公司和合作伙伴披露容量目标和战略逻辑,但没有披露电价水平、项目 ASP、EPC 利润率、燃料定价或收入确认机制。因此,投资人应把 Google 订单储备视为未来需求验证,而不是当前收入质量。关键问题不是变现逻辑是否存在——它存在——而是这套逻辑何时变成可开票、可重复、能贡献利润率的收入。[CI001, CI002, CI009, CI010, CI011, CI012]

收入流表
收入流机制计量单位当前价值 / 状态收入质量尽调请求
DOE 里程碑付款按绩效支付、固定价格,绑定 Hermes 交付物的里程碑合同里程碑 / 可报销项目付款$629M、七年期奖励中,DOE 支持最高 $303M非稀释性资金,但有条件且依赖里程碑获取付款时间表、里程碑门槛和营运资金假设
反应堆项目交付为交易对手开发、建设并运营反应堆电站电站 / 站点 / 项目仅面向未来;未披露在运商业机组未来项目收入;不是当前经常性收入核查 EPC 范围、完工保证和里程碑开票
PPA 下的能源销售根据长期协议出售电力MWh / 容量 / 电站出力Google 合同披露容量目标,未披露已实现售电一旦投运,可能形成高质量收入;当前更像订单储备索取电价结构、容量付款、价格升级和结算机制
辅助服务和环境属性出售非能源电网价值和清洁能源属性服务 / 凭证 / 属性Kairos 官方材料提及,但公开资料未定价可能成为后续货币化尾部;当前未定价索取合同定义、市场假设和收入确认处理
生命周期工程 / 替换件 / 服务支持、维护、部件更换和机队服务服务合同 / 停堆检修 / 部件平台逻辑和 Berkeley 生命周期讨论有所暗示,但公开资料未定价可能形成经常性尾部收入;公开证据仍高度推测获取服务附加率假设和替换件毛利模型

公开资料支持其货币化逻辑,但不支持已实现收入、ASP 或利润率。

[CI002, CI010, CI011, CI012, CI032]
定价 / 货币化表
价格 / 合同项目公开证据标价与实际成交价未知项来源支撑的推论
反应堆 / 项目 ASP未找到公开反应堆售价未披露FOAK 折扣、业主成本、EPC 利润率、取消权无法凭公开数据为单站收入建模
PPA 能源价格Google 协议披露容量和时间表,未披露电价未披露执行价、价格升级、结算、平衡义务需求信号很强;已实现收入经济性不公开
辅助服务 / 环境属性官方公告称将出售这些项目未披露属性定价、市场基准、数量、会计处理可能带来额外收入层,但没有公开模型
DOE 里程碑支持公开披露为绩效型固定价格里程碑付款项目支持,不是商业价格合格成本、付款时点、配套出资负担改善资本充足性,同时把流动性绑定到里程碑完成情况
燃料 / 服务经济性燃料、换料、维护或替换件没有公开价格未披露燃料 ASP、换料节奏、服务附加率、替换件利润率长期经常性收入逻辑成立,但没有定价

公开货币化条款缺失,是财务尽调的核心堵点。

[CI009, CI011, CI016, CI022, CI037]
FI001: 收入模型桥接图

Kairos 如何从公共支持和原型堆走向未来经常性收入流。

[CI002, CI010, CI011, CI012, CI013]

4.2 单位经济性与项目融资

对 Kairos 来说,分析单位不是软件订阅或零部件销售,而是一座融资后的电站或一组电站项目。公开商业单元是 150 MWe,而正在产出学习的发电资产更小。这个错配很关键,因为示范栈可以验证反应堆物理和建设学习,但商业单元经济性在公开信息里仍大多没有定价。EIA 对先进核能资本成本的基准,以及 Lazard 对新核能数据有限的提醒,都强化同一个判断:成本、融资或进度假设的小错误,都可能吞掉整个利润率逻辑。 一个简单资本开支代理能把赌注具体化。如果 Kairos 的 150 MWe 电站贴近 EIA 公布的先进核能资本强度基准,那么在业主成本和建设期利息之前,隔夜成本区间已经会很大。公开来源并没有说 Kairos 会落在那个基准附近——它的整套逻辑正是模块化学习和较低压力架构应改善经济性——但这个基准仍有用,因为它定义了公司必须把成本和进度改善推到什么程度,股权回报才跑得通。[CI014, CI015, CI016, CI017, CI018, CI019]

单位经济性表
驱动因素公开锚点粗略数值 / 状态利润率含义置信度局限
商业电站规模Kairos 技术页面150 MWe [2 x 75 MWe] 电站定义商业部署中收入和资本开支的计量单元未披露公开电站预算
学习型资产规模TechCrunch / ANS 对 Hermes 2 的报道2 x 35 MWth 反应堆加 20 MW 涡轮示范项目经济性不等于商业单元经济性测试资产产出不能直接等同于货币化产出
先进核能资本强度基准EIA AEO 2025 资本成本报告先进核能棕地情景为 $7,861/kW若进度或融资出问题,高资本开支可能吞掉利润率基准并非 Kairos 专属,且使用 AP1000 情景
示意性隔夜资本开支代理值作者用 150 MWe x EIA 基准测算业主 / 融资成本前约 $1.18B显示回报对降本假设有多敏感仅作示意;不是 Kairos 指引
生命周期服务尾部收入Berkeley 关于可更换高温部件的专题部分部件可能有 15-25 年更换周期可能支撑经常性服务 / 部件经济性没有公开定价或范围
燃料可得性WNN / DOE HALEU 来源Hermes 首炉燃料已锁定;更广供应仍受限燃料约束可能拖慢未来机队扩张,或推高成本未量化项目层面燃料成本

由于 Kairos 公开披露的单位经济性信息稀疏,本表刻意同时使用公司专属锚点和外部基准。

[CI014, CI015, CI017, CI019, CI025, CI032]
FI002: 单位经济模型桥接图

简化的单位经济模型桥接图,展示 Kairos 如何从电厂规模走向 capex 敏感性和最终变现。

[CI014, CI017, CI018, CI019, CI020]
FI003: 财务估算区间

区间式公开财务锚点,展示已披露支持和基准强度的规模,但不声称未披露的公司经济性。

该图有意在区间图内使用固定点,因为公开证据只支撑少数财务锚点,不支撑完整的低 / 基准 / 高公司指引。capex 行是作者计算,使用 Kairos 公开电厂规模和 EIA 基准,不应视为公司指引。

[CI003, CI004, CI019]

4.3 资本充足性与融资依赖

在私营核能初创公司里,Kairos 获得的支持异常充足,但业务仍依赖融资。最强公开证据是 DOE 合同架构:在 $629 million、七年期 Hermes 奖励内,通过基于里程碑的协议最高获得 $303 million。这份支持很实质,固定价格结构也能提升纪律性和可信度。但它没有消除资本依赖。公司仍在经常性收入出现前,把钱投向制造基础设施、原型机组、许可和燃料路径。公开来源也没有披露手头现金、债务、烧钱速度或现金跑道,所以外部投资人无法判断,如果里程碑后移或舰队经济性成熟更慢,Kairos 有多少回旋余地。 同行证据在这里很重要。X-energy 的 $700 million Series D 和 NuScale 的上市公司风险披露都显示,先进核能在稳定客户现金流到来前仍会吃掉巨额资本。在这个背景下,DOE 支持降低了却没有解决承销问题。下一个有意义的财务去风险事件,不是另一条头条公告,而是项目级预算、资本结构、风险转移和现金节奏证据。[CI003, CI004, CI005, CI006, CI007, CI022]

资本充足性表
资本来源 / 依赖项公开证据当前状态为何重要尽调请求
DOE ARDP / 里程碑合同$629M、七年期 Hermes 奖励中,DOE 最高 $303M已确认且重要降低融资负担,但覆盖不了项目总需求核查付款时点、配套出资义务以及任何成本超支责任
私募股权和投资人支持Caplight 显示可见轮次 / 投资人;抓取输出中的细节不完整存在性已确认,当前精确规模不清楚决定里程碑之间的现金跑道和灵活度索取股权结构表、上一轮融资文件和当前现金余额
制造和原型支出熔盐设施、ETU 3.0 和 Hermes 推进都需要收入前资本开支持续需求已确认经常性商业收入出现前,现金会先被消耗索取逐项目资本开支预算和预备费
燃料链依赖仅为 Hermes 首炉锁定 HALEU首个示范项目已部分去风险燃料是后续电站的门槛性支出和时间风险索取机队燃料采购计划和成本假设
未来项目融资无公开披露的债务、贷款人或项目融资承诺未知 / 未披露决定订单储备能否转化为可建设电站索取首批商业电站融资策略

资本充足性方向上可信,但公开证据在量化层面披露不足。

[CI003, CI004, CI007, CI022, CI023, CI024]
FI004: 资本强度 / 现金流地图

矩阵展示 Kairos 公开财务故事哪里最强,哪里披露仍最弱。

定性等级单元格描述证据密度,不描述表现质量。“低”通常意味着披露不足,而不是实力弱。

[CI022, CI023, CI024, CI025, CI033, CI039]

4.4 公开缺口与反向信号

财务尽调最大的阻塞点是披露。公开来源没有给出现金余额、烧钱速度、现金跑道、债务契约、客户预付款、项目级 EPC 或有项、已实现模块成本,或任何可被银行接受的电价数据。Caplight 的公开输出确认 Kairos 有可见融资轮和投资人,但即便在那里,抓取到的信息也太不完整,无法用于精确估值或融资轮承销。这迫使判断回到保守立场:官方 DOE 承诺、燃料里程碑和具名客户合同,比第三方估值页面权重高得多。 反向证据也真实存在。TechCrunch 明确把 2030 年时间表视为乐观,整个行业也有 NuScale 10-K 和同行资本需求这些清晰的警示可比项。这些都不能否定 Kairos 最终的财务上行空间,只说明公开数据结论仍是继续研究。在承销收入质量、利润率或估值前,投资人需要私人预算、合同经济性、现金数据,以及更清楚的视角,看到里程碑节奏如何转化成流动性和项目融资。[CI025, CI026, CI027, CI031, CI032, CI033]

公开财务缺口表
缺失指标为何重要公开状态当前最佳代理尽调请求
手头现金 / 现金跑道决定公司能否熬过里程碑滑坡或项目延期未披露DOE 支持加上可见基础设施建设索取最新资产负债表、现金跑道计划和董事会现金预测
烧钱速度 / 现金使用显示里程碑之间的资本强度是否可控未披露原型、许可和设施公告索取月度现金消耗以及固定支出与可变支出
债务 / 受限现金影响融资灵活度和契约风险未披露无留存公开披露索取债务明细表、权利负担和抵押条款
PPA 价格 / 项目经济性把需求证明转化为收入质量未披露仅有 Google 容量目标和服务描述索取条款清单、执行价、价格升级和信用支持
电站层面资本开支和预备费对测算回报至关重要未披露仅有外部 EIA 基准索取 P50/P90 预算、预备费和业主成本模型
燃料 / 服务利润率长期经常性收入逻辑需要这一指标未披露仅有 HALEU 和替换件逻辑索取换料假设、服务附加率和替换件定价

最大缺口恰恰是测算收入质量、利润率路径和稀释风险所需的指标。

[CI016, CI022, CI033, CI034, CI035, CI039]

4.5 展示

Chapter 05

05产品与技术

5.1 产品定义与资产地图

Kairos 的产品最容易理解为一套交付栈,而不是单个反应堆模块。对客户来说,公司卖的不只是反应堆设计;它卖的是一条通往清洁稳定电力的路径,这条路径围绕 KP-FHR 商业电站、逐级推进的 Hermes 示范资产,以及让路径可复制所需的制造、盐和燃料能力搭建。就初创公司而言,公开记录在这一点上异常扎实:地点页面和项目更新清楚说明 Alameda、Albuquerque 和 Oak Ridge 在产品系统里各自承担不同角色。 这很重要,因为先进反应堆承销常常卡在一个点:核心设计很有吸引力,但制造、场址就绪或燃料链执行证据很弱。Kairos 至少对三者都给出了部分公开答案。Alameda 是工程和原型层,Albuquerque 是制造和盐 / 燃料开发层,Oak Ridge 是建设学习和运营示范层。因此,这个产品不只是一座未来也许能卖出去的反应堆,而是一套已经在公开视野里被组装起来的多场址运营模式。[CE001, CE002, CE004, CE005, CE006, CE015]

产品模块 / 资产矩阵
资产 / 模块用户或买方任务状态 / 成熟度关键证据差异化尽调缺口
KP-FHR 商业电站数据中心附近的清洁稳定电力和服务区部署已披露商业设计,暂无在运电站150 MWe [2 x 75 MWe] 氟盐冷却高温反应堆使用 TRISO 球形燃料和低压冷却剂的中等规模电站没有公开运营经济性或可用率
Hermes 1 低功率测试反应堆技术和许可证明在建 / 示范NRC 和 DOE 将其描述为低功率、非发电测试反应堆50+ 年来首个非 LWR 建设施可尚无发电或商业运行时长
Hermes 2 示范电站通向完整商业机队的发电桥梁已获许可且在建两座 35 MWth 反应堆,共用 Rankine 系统首条商业规模、可发电 Gen IV 许可路径仍是学习型资产,不是完整商业机队
Albuquerque 制造开发园区模块制造、盐生产、燃料开发、大型非核测试运营站点 / 能力扩张中选址页面加 WNN 模块制造细节将制造和燃料工作集中到一个站点吞吐量和良率指标未披露
Alameda 工程中心工程设计、快速原型、迭代测试运营站点California 选址页面棕地工程和原型基地没有公开测试产出指标
TRISO + HALEU 燃料路径供应商业和示范反应堆燃料已部分去风险,仍在成熟中BWXT 合作,加上 DOE HALEU 流程和首批燃料分配将内部球形燃料能力与外部制造支持结合机队规模供应、成本和时间仍披露不足
熔盐生产能力支撑冷却剂和材料工作流设施建设可见DOE 熔盐设施文章让盐生产成为可见的内部能力没有公开产能或成本数据

Kairos 的公开产品面覆盖反应堆资产和支撑性基础设施,因为公司的商业化模式两者都需要。

[CE001, CE002, CE003, CE005, CE007, CE009]
工作流 / 用例表
用户任务当前工作流痛点Kairos 产品 / 资产已披露收益当前限制
数据中心密集型负载增长需要 24/7 清洁电力清洁稳定电力稀缺,通用电网电力可能达不到无碳目标KP-FHR 商业电站,加上 Google / TVA 部署路径与服务区匹配的清洁电力、辅助服务和环境属性尚无公开电价或完整运营证明
机队部署前需要为新型反应堆去风险公用事业公司和买方抗拒直接跳到 FOAK 全尺寸项目Hermes 1 到 Hermes 2 的示范阶梯逐步验证许可、建设、燃料和运营学习型资产不等于完整机队经济性
需要比传统定制核电成本更低的建设路径传统核电项目慢,且高度按站点定制借助 Albuquerque + Oak Ridge 做模块化建设和场外制造预制、模块化和偏工厂化的方法,目标是降低成本、压缩工期尚无公开已实现降本曲线
先进反应堆部署需要燃料和材料就绪HALEU 和 TRISO 供应是行业瓶颈DOE HALEU 流程加 BWXT TRISO 路径面向示范项目和机队目标的首批燃料及制造路径可见机队规模的量和成本仍不确定
核运营前需要工程迭代新型系统到项目后期再改,风险很高Alameda 设计 / 原型和 ETU 非核测试单元核部署前开展迭代测试公开运行时长测试指标稀疏

这条工作流解释了 Kairos 为什么搭出一串资产阶梯,而不是直接跳到首个商业电站。

[CE001, CE006, CE014, CE016, CE020, CE023]
FE001: 产品架构图

从客户用电结果到场址、示范堆和燃料链,Kairos 公开证据能支撑的产品栈层级。

该产品栈只包括公开可见的产品层级,不声称覆盖完整内部控制或软件架构。

[CE001, CE002, CE003, CE012, CE018, CE032]

5.2 架构与运营模式

Kairos 公开架构的核心是 KP-FHR:一种氟盐冷却高温反应堆,使用 TRISO 环形球燃料;披露的商业电站为 150 MWe,由两个 75 MWe 机组组成。公司公开材料和 NRC 记录显示,这套前端规格搭配的是一个后端运营模式,围绕迭代式设计-建造-测试循环、垂直整合和更小的学习资产展开。Hermes 1 对核心技术做非发电示范,Hermes 2 则加入发电桥接资产:两个 35 MWth 反应堆,共用一套能量转换系统。 因此,技术故事比许多先进反应堆叙事更容易读懂。买方和投资人能看见反应堆设计、模块制造、场址建设和能量转换如何连接。即便如此,最强证据仍是架构和项目层面的,而不是运营层面的。公开材料描述 Kairos 计划建什么、打算怎么建;它们还没有提供舰队运行数据、已实现可用率或长期运行验证。[CE003, CE007, CE008, CE009, CE010, CE014]

技术 / 运营架构表
层级公开披露设计运营角色优势开放技术问题
核心反应堆设计采用 TRISO 环形球形燃料的氟盐冷却高温反应堆产生稳定热量,用于转换为清洁电力低压、高温架构,与传统水冷系统拉开差异没有公开长周期运行数据
燃料系统使用 HALEU 的 TRISO 球形燃料提供稳健燃料形态,并贡献功能性包容能力NRC 材料明确描述了燃料及其包容作用机队规模燃料成本和认证节奏不公开
功率转换Hermes 2 采用共享蒸汽 Rankine 循环;商业电站为 150 MWe将反应堆热量转换为上网电力让产品定位成发电电站,而不只是测试反应堆商业涡轮和常规岛性能仍未验证
建设模式模块化建造、预制混凝土、隔震基础、场外制造压缩工期和成本,提高可复制性反应堆初创公司公开披露了具体建造方法尚未披露已兑现的工期或成本节省
制造 / 熔盐 / 燃料骨干Albuquerque 园区、熔盐设施以及与 BWXT 合作支撑反应堆可复制生产和燃料就绪看得见的支撑基础设施,而不是抽象供应主张缺少实际产能和良率数据
支持 / 控制 / 网络安全公开来源主要覆盖物理和监管层保障电站运行安全、可靠监管流程提供了一定信任面OT 网络安全架构和具名控制供应商仍披露不足

本表把可见架构与披露最少的数字和运行层分开。

[CE003, CE010, CE013, CE017, CE019, CE028]
FE002: 客户流程 / 运营流

代表性路径,展示 Kairos 打算如何从客户需求走向可复制的商业部署。

[CE006, CE009, CE010, CE014, CE018, CE023]

5.3 部署、依赖与差异化

Kairos 的差异化不只是一个反应堆物理主张,还包括一套制造和部署逻辑:Albuquerque 的模块制造,Alameda 和 Oak Ridge 的棕地复用,Hermes 2 的模块化施工方法,以及一条把内部能力、外部伙伴和 DOE 支持混合起来的燃料路径。这个组合支撑了公司的核心论点:相较传统核电建设模式,它可以压缩周期、降低成本。面向客户的版本已经出现在 Google 和 TVA 路径里,公司呈现的是服务区域、数据中心邻近的商业化模式,而不是泛泛的基荷电站叙事。 反面是依赖集中。燃料是最清楚的例子。DOE 的 HALEU 材料、BWXT 的 TRISO 合作,以及 Kairos 的首批燃料分配都显示出真实进展,但也确认产品无法脱离燃料链就绪度。盐生产、施工方法和被监管接受的材料工作也适用同样逻辑。Kairos 之所以差异化,正因为它试图同时协调这些层;这也意味着任何一层延迟,都可能限制部署速度。[CE011, CE012, CE013, CE017, CE018, CE019]

路线图 / 发布 / 发展阶段表
阶段 / 资产目的当前公开状态进展证据剩余阻碍
Alameda 工程与原型现场部署前设计并迭代核心系统运行中场地地点页面和棕地建设内部原型性能指标未公开
Albuquerque 制造与熔盐 / 燃料园区打造可制造模块、熔盐能力、燃料开发和非核测试运行中 / 扩建中地点页面、DOE 熔盐设施、BWXT 合作产能和成本曲线未公开
Hermes 1验证许可路径和低功率 KP-FHR 行为已获许可并在建NRC 和 DOE 许可报道尚无发电输出验证
ETU 3.0 及相关测试单元试点制造与建造流程原型资产DOE ETU 3.0 文章从成熟度到机组群复制的转化率未知
Hermes 2衔接发电型、商业规模演示已获许可并开工NRC Hermes 2 页面、官方破土动工、WNN/ANS 报道仍需证明能量转换和可复制执行
商业机组群 / Google-TVA 路径扩展为服务辖区内可复制部署未来 / 运营前Google 和 TVA 路径公告燃料、进度、经济性和运营验证需要同时收敛

Kairos 的路线图可见且逐步推进,但商业规模仍处于运营前。

[CE006, CE009, CE010, CE016, CE018, CE023]
FE003: 关键依赖图

这些依赖最直接决定 Kairos 公开产品承诺能否转化为可复制的商业机队。

该 DAG 只展示公开依赖;它不声称合作伙伴具有排他性,也不披露隐藏的数字控制依赖。

[CE012, CE018, CE019, CE020, CE021, CE022]

5.4 信任、质量与开放技术缺口

Kairos 的信任逻辑,在物理安全和监管互动上最强。NRC 关于 Hermes 和 Hermes 2 的页面,对燃料、功率水平和安全壳方法给出了异常具体的描述,其中包括 TRISO 燃料在功能性包容中的作用。公开 DOE 和公司材料也让原型资产、施工技术和选址变得可见。这些质量证据强过许多早期能源公司。 但公开记录仍没有达到完整运营尽调室的程度。留存来源没有给出已实现可用率、舰队维护指标、控制系统供应商细节或 OT 网络安全架构。对核能产品来说,这些遗漏并不推翻技术逻辑,但会塑造尽调姿态。结果是,本章对可见物理和监管栈有较强信心,同时也清楚承认:商业运营验证和数字控制保障仍是未来工作,而不是公开事实。[CE027, CE028, CE030, CE031, CE033, CE034]

信任 / 质量 / 合规表
信任面公开证据当前置信度重要性剩余缺口
NRC 建造许可证Hermes 和 Hermes 2 页面列出详细特征与里程碑显示监管互动和审查都很深入,而不只是营销说法许可证不等于运营验证
燃料 / 功能包容逻辑NRC 说明了与 TRISO 燃料绑定的功能包容相比泛泛的被动安全表述,安全论证更具体尚无现场性能记录
DOE 和国家实验室支持DOE 文章、HALEU 流程、ORNL 合作中-高支撑技术生态以及材料 / 燃料成熟度支持不保证进度或商业就绪
选址与棕地策略Alameda、Oak Ridge 和 Heritage Center 细节中-高指向一个借助既有工业场地落地的实用部署模型尚无重复场地改造速度的证据
燃料配给与首批装料规划DOE HALEU 流程和 WNN 首批燃料报道说明演示燃料已有路径机组群规模燃料安全仍未解决
数字控制与网络安全保留下来的公开细节很少对运行期信任和电站运营至关重要需要 OT 架构、供应商和网络安全治理

监管方和具体设施可见的地方,公众信任最强;通常应出现运行软件和机组群指标的地方,信任最弱。

[CE020, CE021, CE027, CE028, CE030, CE033]
FE004: 产品成熟度 / 能力图

Kairos 已披露产品层的相对成熟度,并区分监管证明与运行证明。

序数值概括的是证据密度,不保证项目成功。

[CE020, CE021, CE027, CE029, CE033, CE038]

5.5 展示

Chapter 06

06客户情况

6.1 锚定客户、买方逻辑与地理位置

对一家商业化前反应堆公司来说,Kairos 的客户故事异常具体,但它不是常规的一客户一合同公用事业销售。首个可见商业结构有三层。Google 是需求锚:它需要面向特定数据中心运营的稳定无碳电力和清洁能源属性。Kairos 是开发运营方,负责建设并运营电站。TVA 是公用事业承购方和电网交付层,将购买 Hermes 2 出力,并通过自己的系统转移这些属性。这种分工很重要,因为它让 Kairos 看起来不像一个商用反应堆卖方,更像是一个为超大规模负荷服务、由公用事业赋能的采购平台。 公开地理也比大多数先进核能客户故事更具体。2025 年合作并不只是抽象提到数据中心,而是指向 Google 在 Tennessee 的 Montgomery County 和 Alabama 的 Jackson County 的运营。这种具体性强化了买方-用户-付款方映射,也降低了客户叙事纯属愿景的风险。最弱点在广度,而不是清晰度。公开证据显示一个极强的终端客户和一个极重要的公用事业中介,但还没有显示跨区域、跨用例的独立付费买方组合。[CU001, CU002, CU006, CU007, CU009, CU010]

客户分群表
分群买方 / 用户 / 付款方用例规模证据战略价值缺口
Google 数据中心负载Google / Google 数据中心运营 / 公用事业联动采购下的 Google为超大规模计算负载获取 24/7 无碳电力和清洁能源属性500 MWe 总协议;已点名 TN 和 AL 具体设施锚定需求来源,具备重复部署潜力无公开定价、期限或照付不议细节
TVA 公用事业承购TVA / TVA 电网 + Google 终端负载 / TVA 结算层购买 Hermes 2 电力,并通过 TVA 系统转移属性最高 50 MW Hermes 2 PPA;美国首个公用事业 Gen IV PPA提供可融资的公用事业交易对手和并网路径目前只披露一份特定电站合同
Carolinas 大负载电价参与方Google、Amazon、Microsoft、Nucor / 大型工商业负载 / Duke 电价框架下各参与方借助电价创新支持清洁稳定电力增量2024 年与 Duke Energy 签署 MOU显示 TVA 服务辖区外相邻买方也有需求不是 Kairos 专属采购协议
未来超大规模云厂商买方可能的大型科技买方 / 数据中心 / 企业或公用事业居中付款如果 Hermes 2 成功,清洁稳定电力采购可复制保留的公开来源中没有具名的非 Google 买方若模型证明可复制,总可用市场(TAM)很大尚无公开转化证据
未来工业或公共电力买方公用事业或工业买方 / 高耗电运营 / TBD潜在清洁稳定电力或工业能源用例公开材料只作暗示,未列为 Kairos 具名客户可分散单一超大规模云厂商风险除 Google-TVA 外,无公开具名合同或试点

分群按买方—用户—付款方结构划分,因为 Kairos 首条商业路径把需求、公用事业承购和反应堆运营拆给不同实体。

[CU001, CU002, CU007, CU013, CU018, CU019]
客户商业化路径表
阶段公开记录证据商业含义主要未解缺口
战略需求表达Google 将先进核能与 AI 时代负载增长和 24/7 CFE 目标绑定说明真实买方问题存在需要按区域和时段拆分的需求预测
机组群框架到 2035 年的 500 MWe 订单簿显示有意愿扩展到单机组之外需要逐电站转化机制
特定电站公用事业合同TVA Hermes 2 PPA 最高 50 MW显示与电网整合的客户交付路径需要合同经济性和期限
区域场址 / 负载映射Tennessee 和 Alabama 的 Google 数据中心已具名提升客户证据具体度需要设施层负载和交付核算
可复制采购架构Duke/ACE 电价工作和 TVA 模型暗示存在超出单一服务辖区的扩张渠道需要 Kairos 专属后续承诺

加入这张额外路径表,是因为 Kairos 的商业化流程与具名客户数量本身同样重要。

[CU004, CU006, CU018, CU026, CU027, CU034]
FU001: 客户旅程图

Kairos 的第一条客户旅程从超大规模负荷增长出发,落到由公用事业公司居中交付,而不是简单的双边售电。

[CU001, CU002, CU004, CU006, CU013, CU022]

6.2 采用轨迹与具名客户验证

采用轨迹可信,是因为它从宽泛框架走向了具体电站合同。2024 年 10 月,Google 与 Kairos 签署到 2035 年的 500 MWe 主开发协议。2025 年 8 月,这个宽框架收窄为一个已披露、有约束力的 Hermes 2 路径:通过 TVA 实现最高 50 MW,并以 2030 年运行为目标。这个顺序很重要:它说明 Kairos 已从概念性的企业兴趣,推进到一个具名公用事业承购安排,并且绑定具体电站、具体区域和具体终端负荷。 与此同时,具名验证集仍然很窄。Google 和 TVA 都是真实且质量异常高的交易对手,但后续舰队机组仍受里程碑约束;Duke 2024 年合作更适合解读为未来采购架构证据,而不是 Kairos 客户转化。因此,本章支持一个强结论:客户质量高;也支持一个弱结论:客户广度仍不足。Kairos 拥有先进核能领域最好的公开具名客户验证包之一,但还没有多元化的公开客户名单。[CU003, CU004, CU005, CU008, CU014, CU015]

客户增长 / 采用轨迹表
指标数值日期来源置信度含义缺失分母
Google-Kairos 机组群协议到 2035 年最高 500 MWe2024-10-14Kairos + Google 公告显示客户目标是多机组,而非单一试点没有逐机组合同时间表
首座电站目标2030 年 Hermes 22025-08-18Kairos + Google + TVA 公告为框架转交付设定明确的时间里程碑未披露详细 COD 关键路径
首份特定电站公用事业 PPA最高 50 MW2025-08-18TVA/Kairos/行业报道把客户验证从概念推进到已披露承购安排无公开合同期限
具名终端使用地点位于 Montgomery County, TN 和 Jackson County, AL 的 Google 数据中心2025-08-18Kairos + DCD具体地点提升证据质量和地理映射无设施层负载拆分
Google 数据中心用电需求增长同比 27%2025-06-27Google 环境报告支撑清洁稳定电力采购的紧迫性无按市场划分的区域拆分
Google 2025 年清洁能源采购签约 12 GW 净新增清洁能源2026-06-26Google 可持续发展说明 Kairos 嵌在一个很大的采购引擎里未披露核能占比
Duke 大客户框架参与方4 家具名公司2024-05-29Duke Energy释放出更广泛买方对公用事业居中清洁稳定电力结构的需求信号无特定电站的 Kairos 配额

这条轨迹在具名交易对手和里程碑日期上最强,在逐电站经济性以及后续机组转为已签 PPA 上最弱。

[CU003, CU004, CU006, CU010, CU011, CU015]
具名客户证据表
客户 / 证据面分群部署 / 用例生产与试点结果限制
Google超大规模数据中心买方最高 500 MWe 先进核能容量的 Master Plant Development Agreement;为数据中心提供清洁能源属性已签约 / 交付前首份企业多反应堆先进核能协议,并有具名区域用例无公开定价、条款清单或已交付电力记录
TVA公用事业承购方有约束力的 Hermes 2 PPA 接入 TVA 电网,支撑 Google 区域负载已签约 / 交付前美国首份由公用事业签署、采购先进 Gen IV 反应堆电力的 PPA只披露单座电站;后续机组尚未纳入公开 TVA PPA
Duke Energy / ACE 联盟公用事业赋能的采购框架与 Google、Amazon、Microsoft、Nucor 签署 MOU,建设新的清洁能源电价结构仅框架证明相邻买方愿意支持有利于核能的公用事业电价不是 Kairos 专属承购或部署承诺

公开具名证据的质量少见地强,但数量仍少。

[CU003, CU005, CU006, CU014, CU018, CU019]
FU002: 采用 / 部署漏斗

公开客户证据从广泛的大负荷兴趣,收窄到 1 份具名公用事业 PPA;截至目前,实际供电客户队列为 0。

该漏斗刻意区分广泛买方意愿和已签约的 Kairos 专属交付证明。

[CU015, CU018, CU019, CU024, CU031]
FU003: 客户验证矩阵

Google 的终端需求证据最强,TVA 的公用事业购电证据最强,Duke 的侧面采购证据最强,但没有 Kairos 专属采购承诺。

[CU008, CU023, CU030, CU036]

6.3 留存、耐久性与变现可见度

公开留存证据几乎全部来自结构,而不是经验数据。核电合同如果进入运行,粘性应当很强:资产寿命长、场址特定,并深度嵌入公用事业和客户规划。Google 的订单储备概念,以及 TVA 作为公共电力公用事业公司的角色,也支撑一个合理的复购判断。但公开来源仍没有显示合同期限、续约机制、流失、NRR、最低购买义务或终止权。因此,留存逻辑仍是从市场结构和交易对手推断出来的,而不是经过衡量的队列历史。 变现透明度同样不完整。POWER 报道了 TVA 交易的分时固定市场定价,但公开定价栈和风险分担瀑布都不可见。因此,投资人必须把两个常被混在一起的概念拆开:Kairos 已经赢得很强的战略客户验证,但公开披露还不足以把这份验证当成充分承销过的收入质量。正确解读是,客户耐久性可能非常好,但公开记录还不能证明。[CU016, CU021, CU022, CU024, CU025, CU035]

留存 / 重复使用 / 满意度表
指标数值 / 公开状态分群置信度尽调追问
续约率或流失率未披露Google / TVA索取合同续约机制、里程碑退出权以及任何介入条款
平均合同期限未披露Google / TVA索取 Hermes 2 PPA 期限和后续电站选择权
复购证据订单簿结构显示,只要达成里程碑,就可能重复部署Google索取后续反应堆订单的选择权时间表或触发条件
客户满意度或 NPS暂时意义不大;尚无运营客户交付全部询问客户指导委员会节奏和里程碑评审结果
切换成本耐久性若电站投运,结构上很高;交付绑定特定场址并与公用事业整合Google / TVA验证并网、属性核算和替代电力兜底权利

本表把结构性耐久性和实测留存分开。公开来源对前者的支撑远强于后者。

[CU016, CU021, CU022, CU024, CU025, CU035]

6.4 扩张循环与集中风险

Kairos 最好的扩张论点是,它的第一位客户正是那种能够支撑重复部署的买方。Google 持续报告 AI 驱动电力需求上升和创纪录清洁能源采购,并且已经同时支持 Kairos 的舰队式协议,以及在另一区域与 Duke 的电价创新。TVA 的参与加入了第二个重要条件:通过公用事业居中,把清洁稳定电力送入电网,而不是只依赖复杂的表后结构。如果 Hermes 2 跑通,商业化循环可能自我强化——经济性更好,公用事业意愿更强,超大规模云厂商兴趣更高。 主要公开风险是集中度。Google 仍是 Kairos 舰队唯一具名终端客户,Duke 只是框架伙伴。DOE 关于核能数据中心的材料也提醒,新反应堆需要时间、成本高,并面临计量和燃料链障碍。所以,上行空间真实存在,但路径还不够多元,也没有足够去风险,不能把它视为今天已经形成广泛客户采用。Kairos 已跨过“严肃购买兴趣验证”这条线;还没有跨过“广泛市场转化验证”这条线。[CU011, CU012, CU026, CU027, CU028, CU029]

扩张与集中度风险表
扩张驱动因素集中度风险影响尽调路径
Google 由 AI 驱动的负载增长和 24/7 CFE 目标Google 仍是唯一具名终端客户若重复机组转化,上行很高;若 Google 放慢或调整优先级,下行很高索取选择权时间表、区域路线图,以及按电站划分的客户集中度假设
TVA 公用事业居中交付模型迄今只有一个公用事业买方公开绑定 Kairos若能复制到其他服务辖区,上行很高索取更多公用事业谈判管线和选址标准
Duke/Carolinas 电价试验监管批准可能停滞或削弱电价结构获批则中等利好;延迟则中等利空跟踪 NCUC/SCPSC 批准,以及任何 Kairos 专属后续协议
FOAK 核能作为数据中心供给建设周期长、燃料供应和计量问题可能拖慢采用若里程碑延误,下行很高将客户爬坡假设与 DOE 和许可时间线对齐
强具名交易对手证据除 Google 和 TVA 外,公开名单仍窄验证层面中等利好;分散化层面高利空索取按垂直行业和地域拆分的完整商务拓展漏斗

目前最高上行驱动因素和最高集中度风险来自同一实体:Google。

[CU020, CU026, CU028, CU029, CU031, CU032]

6.5 展示

Chapter 07

07风险

7.1 监管和法律风险已经体现在进度里,不只是原则上的假设

公开记录里,置信度最高的风险是监管排期继续滑坡。Kairos 不是泛泛提醒首台套反应堆建设可能很慢;Hermes 1 已经需要拿到 NRC 修正案,把完工期限从 2026 年底延到 2029 年 4 月。NRC 自己的安全评估把原因说得很清楚:设施首台套属性带来的开发问题,加上比预期更复杂的场地准备工作。这个点关键,因为它把泛泛的创业风险变成了带时间戳的先例。公司可以拿强监管互动来说明自己在推进,但已经不能再讲一条没有滑坡的首次运行路径。 缓释因素也真实存在。同一份 NRC 文件称,修正案没有授权新活动,认定不存在重大危害影响,并给出无重大环境影响结论。Hermes 2 也已经拿到建设许可,文件链条透明。所以法律与监管图景不是被拒或遭敌视,而是合作但严苛的监管。承销时这个差别很重要。Kairos 比陷在许可不透明里的公司更好,但不如一条干净准时的开发故事。后续日期如果再滑坡,会是重大负面信号,因为“第一次延误”已经发生。[CR001, CR002, CR003, CR004, CR005, CR006]

监管 / 法律风险登记表
规则 / 许可 / 案件司法辖区状态可能性严重性缓释措施剩余敞口尽调路径
Hermes 1 建造完工期限和许可证延期先例NRC / 联邦FOAK 延误后,最新完工日期已延至 2029 年NRC 已批准延期,且未发现重大危害或环境损害继续跟踪后续修正请求、里程碑滑坡,以及与修订后 2028 年完工预期的偏差。
若再次错过期限,施工许可可能到期NRC / 联邦许可条件中仍有活跃法律风险中高委员会可基于正当理由延期,但前提是有证据且留足时间中高要求提供内部关键路径、许可条件跟踪表,以及触发再次申请修正的阈值。
Hermes 2 持续监管与后续许可步骤NRC / 联邦许可已经拿到,但监管和文件审查仍在推进文件链条透明,且此前许可已跑通审查待审议主题、检查节奏,以及哪些后续批准仍在关键路径上。
向 Part 53 下的新先进反应堆框架过渡联邦 / NRC规则已经生效,但 Kairos 当前路径实操上能用到多少仍好坏参半中等长期看,这一可选空间应会改善后续路径设计询问管理层,哪些未来申请会使用 Part 53,以及实际预期能获得哪些许可效率。

各行按剩余严重程度排序;已经发生的进度延误排在更抽象的监管流程问题之前。

[CR001, CR002, CR004, CR006, CR007, CR008]
FR001: 风险热力图

Kairos 最高的剩余敞口集中在进度、燃料供应、对手方与 FOAK 商业化时点的交汇处。

[CR001, CR011, CR020, CR027, CR038, CR041]

7.2 燃料、建设与运营风险都卡在同一个瓶颈:太多首台套系统必须一起成熟

核心运营风险不是某个孤立工程问题,而是反应堆建设、燃料供应、燃料制造和并网节奏要同时成熟。Hermes 使用含 HALEU 的 TRISO 球形燃料,DOE 明确称,国内 HALEU 供应商还无法提供先进反应堆开发商最终所需的规模。Kairos 为 Hermes 1 锁定 DOE 来源的 HALEU,合理压低了近期风险,但这只是过桥安排,不是广泛商业市场已经成型的证明。ANS 和 DOE 材料说得很清楚,配给项目存在,正是因为供应链还不完整。 执行复杂度会在这里叠加。Centrus 目标是在 2029 年转向更大规模 HALEU 生产;BWXT 现在才开始与 Kairos 探索商业 TRISO 制造;DOE 仍把配给描述为持续流程。这些事实不意味着 Kairos 会失败,但意味着时间错配仍然可能。再加上 DOE 自己提醒,首台套数据中心反应堆造价高、建设慢,运营风险栈就很清楚:即便反应堆物理可行,商业路径仍可能在燃料、制造或建设学习曲线上放慢。[CR010, CR011, CR012, CR013, CR014, CR015]

运营 / 质量 / 安全风险登记表
失效模式发生概率严重程度缓释成熟度剩余敞口未解决缺口
国内 HALEU 供应仍不足以支撑先进反应堆大规模部署中 — Hermes 1 已拿到 DOE 配额,但批量部署级供应仍依赖尚未成形的市场建设需要按电站拆清 DOE 配额到商业燃料供应之间的数量、时间和成本桥接。
FOAK 施工学习拉长工期并推高烧钱中 — 非核示范和迭代建设降低技术意外,但拉长时间线需要修订后进度、应急预算,以及场地准备和基础工程的实际生产率。
TRISO 燃料制造和商业化产能爬坡落后于电站需求中高中低 — 与 BWXT 的合作有前景,但面向后续反应堆仍处探索阶段需要明确 Hermes 1 之后燃料球的责任分工、产能假设和合格鉴定里程碑。
并网和数据中心商业化结构面临计量 / 成本分摊摩擦中高中低 — 公用事业合作有帮助,但政策问题仍在外部中高需要为未来辖区的表计边界、成本分配和电费用户承担安排给出明确解决路径。

这张登记表聚焦最可能把好技术故事拖成延误商业故事的运营瓶颈。

[CR010, CR011, CR012, CR013, CR014, CR015]
FR003: 依赖图

Kairos 位于一张紧密依赖网的中心,网络横跨监管机构、DOE 燃料支持、商业浓缩、燃料制造、公用事业交付和锚定客户。

[CR011, CR015, CR020, CR021, CR022, CR026]

7.3 合作伙伴、客户与市场结构风险可能把强验证点变成集中的商业化漏斗

Kairos 的具名交易对手质量很高,但名单仍短,这种质量本身也带来集中度风险。Google 是机组组合唯一具名终端客户,TVA 是唯一具名公用事业承购方。这不削弱客户验证的重要性——很少有先进反应堆创业公司拿得出类似证据——但意味着公司的商业叙事异常暴露于少数交易对手的行为。如果 Google 需求情景或 TVA 交付路径任一走弱,信号损伤都会被放大。 第二个市场结构风险是,商业化不仅取决于硬件准备度,还取决于公用事业和电价架构。DOE 与 Latitude 都强调,面向数据中心的核电模式会遇到成本分摊、计量和费率设计问题。因此,Duke 的大负荷电价争议即便不是 Kairos 合同,也有参考意义。它展示了未来清洁稳定电力路径可能如何围绕谁承担电网升级成本、大负荷需要缴纳多少抵押品、以及数据中心客户是否得到补贴或被恰当隔离而引发反弹。Kairos 需要的不只是一台能跑的反应堆;还需要采购结构在政治和经济上都扛得住。[CR020, CR021, CR022, CR023, CR024, CR025]

合作伙伴 / 依赖风险登记表
依赖项交易对手角色集中度失败情景严重程度缓释因素剩余敞口
锚定客户需求Google订单簿和首个区域负荷案例中具名终端客户很高需求放缓、里程碑耐心下降,或后续机组未能转化为重复采购Google 长期 24/7 CFE 战略和庞大清洁能源计划
公用事业承购与交付路径TVA首个披露的公用事业 PPA 和电网交付层Hermes 2 延误、费率经济性低于预期,或公用事业优先级转移公共电力使命和明确的先进核能战略中高
燃料来源桥接DOE HALEU 配额计划近期 HALEU 提供方和政策抓手配额时间或数量无法匹配反应堆进度Kairos 已敲定 Hermes 1 合同;DOE 继续发放配额
商业 HALEU 产能爬坡Centrus后续部署潜在商业浓缩骨干中高商业扩张延误或成本上升,导致需要燃料时批量部署无料可用DOE 支持的大合同和商业化过渡计划中高
TRISO 制造产能爬坡BWXT后续反应堆燃料潜在商业制造伙伴中高制造就绪晚于反应堆部署需求现有合作以及双方对商业燃料路径的共同利益中高

公司风险不是依赖单一机构,而是依赖一条很短的机构链,且每个环节都必须同时到位。

[CR020, CR021, CR025, CR026, CR034, CR036]
FR002: 风险传导图

Kairos 最重要的风险沿着一条短链传导:从进度和燃料,传到客户信心、资本需求和估值支撑。

[CR005, CR017, CR025, CR034, CR036, CR038]

7.4 最重要的下行先例是 NuScale,最重要的缓释手段是严守里程碑闸门

检验 Kairos 风险画像的最佳外部压力测试,不是泛泛的核能警示故事,而是 NuScale 的 Carbon Free Power Project 已记录的失败路径。DOE 监察长称,关键风险没有得到有效评估或监控,前置的公共成本分担让纳税人资金暴露在风险中,项目最终仍未达成核心目标。UAMPS 与 CATF 补上商业层面的教训:订购疲弱、首个客户匹配不佳,即使项目已经取得多年监管进展,也足以拖垮反应堆项目。这对 Kairos 高度切题,因为 Kairos 同样在试图把示范学习接到商业承购上,同时依赖强交易对手和公私支持。 Kairos 的缓释逻辑是先学习再放大。Hermes 1、Hermes 2、DOE 燃料支持、BWXT 合作,以及 Google/TVA 推出路径,都符合迭代式、里程碑闸门管理的策略。方向上,这比直接跳进大型首台套商业电厂更好。但这条路同样耗资本、耗时间,也更依赖人才、供应和顺序纪律。因此,尽调应紧盯一份简短的止损条件清单:进度进一步滑坡、燃料节奏错配、客户承诺恶化,或成本与复杂度更像 NuScale 剧本而不是 Kairos 的学习剧本。[CR027, CR028, CR029, CR030, CR032, CR034]

人员 / 执行风险登记表
角色 / 职能依赖或缺口发生概率严重程度缓释因素尽调路径
Hermes 1、Hermes 2 和商业化的项目管理过多相互咬合的关键路径可能超过管理层带宽中高迭代开发理念和按里程碑推进的示范审查综合总进度表、关键路径责任归属和跨项目升级治理。
燃料与材料工程HALEU、TRISO 和制造项目依赖专门人才及外部实验室 / 合作伙伴Kairos 拥有 DOE、Los Alamos 和 BWXT 联系要求提供燃料项目中的具名人员配置、伙伴 SLA 和单点故障。
Oak Ridge / Tennessee 核能劳动力区域建设可能收紧熟练劳动力和供应商供给中高本地生态强,TVA 支持的核能活动也在增加审查劳动力计划、关键分包商承诺和工资 / 成本升级假设。
监管与跨机构协调场地工作已经需要大量 DOE 和第三方协调中等已经证明能赢得修正批准并继续推进施工梳理所有跨机构依赖,找出仍然卡住的瓶颈机构。

这些执行风险不是泛泛的初创公司问题;它们来自 Kairos 开发模式的并行推进和物理复杂度。

[CR003, CR033, CR034, CR035]
缓释措施与终止标准表
风险可监测触发项阈值 / 事件行动含义
进度滑坡Hermes 1 和 Hermes 2 里程碑日期任何进一步 NRC 期限延期请求,或较修订后内部进度出现跨季度延误下调时间线信心,并重新测试客户 / 资本假设。
燃料时点错配HALEU 和 TRISO 就绪度DOE 配额中断、Centrus 商业化延迟超过 2029 年,或燃料制造里程碑缺失假设批量部署推迟,融资需求更高。
客户集中Google / TVA 承诺强度后续机组未转化、合同明显收缩,或区域交付经济性恶化将客户证明视为单项目验证,而非机组群验证。
FOAK 成本失控公开或尽调中的成本证据资本开支上升,或监管失灵开始接近 NuScale/UAMPS 先例把公司重估为高风险示范项目,而不是商业部署故事。
费率和市场结构反弹电费用户 / 监管方回应有证据表明,未来大负荷核能采购模型无法隔离成本或获得批准下调 TVA 辖区之外的复制预期。

这张表强调商业全面运营前就能观察到的触发项;对一家后期但仍未收入的反应堆公司来说,这一点很关键。

[CR016, CR022, CR027, CR029, CR041, CR042]

7.5 图表

Chapter 08

08估值

8.1 投资逻辑与反向逻辑

Kairos 的正向案例强过普通先进核能故事,因为公司已经拥有这个行业里极其重要的三件事:具名超大规模买方、具名公用事业承购方,以及可见的首座电厂路径。Google 的多机组订单、TVA 的 Hermes 2 PPA、DOE 燃料支持,以及 Kairos 的迭代开发模型,合在一起构成了商业化叙事;对一家私营反应堆公司来说,这个叙事异常清晰。客户需求背景也真实存在。Google 持续披露因 AI 和数据中心规模带来的大量且增长的清洁电力需求,这意味着 Kairos 并不是在向一个想象中的未来市场销售。 反向逻辑不是市场虚假,也不是技术缺乏吸引力,而是公开证据仍难以给公司精确定价。Hermes 1 进度已经滑坡,通往机队规模的燃料桥仍不完整,合同经济性未公开,NuScale/UAMPS 的下行先例又太切题,不能忽略。换句话说,Kairos 也许是最强的私营先进核能平台之一,但公开记录支持的仍是依赖里程碑兑现的可选性故事,而不是可以干净承销的价值故事。[CV001, CV002, CV003, CV004, CV005, CV006]

投资逻辑 / 反向逻辑表
维度投资逻辑反向逻辑 / 何种情况会改变判断
客户证明Google + TVA 为一家私人反应堆公司提供了异常强的具名交易对手交易对手质量不能替代合同经济性披露,也不能替代客户多元化
市场需求AI 和清洁稳定电力需求持续扩张,Kairos 的方向因此具备吸引力并非所有需求都能转化为成本分摊可接受、可融资的核能项目
执行模式迭代式 Hermes 路径比直接跳向巨型 FOAK 商业电站更聪明同一条迭代路径更慢、更耗资本,而且已经延误
燃料路径DOE 燃料支持叠加 BWXT 和 Centrus 进展,拼出了一座可见的桥这座桥在机组群规模上仍未完成,且时间敏感
可比公司支撑TerraPower、X-energy、Oklo 和 NuScale 证明资本市场仍关注先进核能同一批同行也证明,波动性、资本强度和失败结构能迅速毁掉价值
估值可见度如果里程碑持续兑现,战略溢价可能合理没有披露当前标记、股权结构表和预算,就无法有信心地为溢价定价

这张表把公司质量论点和价格 / 结构反对意见拆开。

[CV001, CV003, CV006, CV008, CV012, CV020]
FV001: 投资建议逻辑

一旦把定价不透明、进度延期和燃料时点纳入门槛,Kairos 就会从战略吸引力很强降为仅值得跟踪。

该图是决策框架,不是机械财务模型。

[CV001, CV003, CV006, CV008, CV024, CV027]

8.2 建议、信心、风险评级与估值立场

当前合适姿态是观察 / 条件性,而不是买入。这并不是否定 Kairos 的战略位置。就公开证据看,Kairos 是一家严肃公司,客户验证优于许多同业,去风险步骤也比概念阶段气候科技创业公司更可见。需要谨慎的原因在价格,而不在质量。投资者不知道当前普通股估值、当前客户经济性、股权结构表的优先权安排,或已披露进度延期后的修订项目预算。缺少这些,公开来源不足以支持一个强判断:当前私营价格——无论是多少——显然有吸引力。 因此,信心为低到中,风险仍高。公司已有足够证据值得持续主动跟踪,也可能足以支持进一步尽调,但公开透明度还不足以支撑激进承销。正确的估值立场是“披露不足,但战略价值高”:如果执行守得住,这个类别显然有值得持有的资产;但缺失变量太重要,不能靠行业热情一笔带过。[CV004, CV005, CV023, CV024, CV025, CV026]

建议摘要表
维度评估依据
总体建议观察 / 有条件公开证据显示,公司质量强于价格支撑;只有在更深入尽调或严守入场条件时才继续推进
信心水平低至中客户和监管证明很强,但股权结构表、估值标记和经济性仍未公开
风险评级进度、燃料和交易对手集中风险仍在相互作用,而非彼此分散
估值立场披露不足,但具战略价值公开证据足以支撑兴趣,但不足以给出精确点位或激进定价
决策含义密切观察;不要追逐不透明定价上调判断需要更清晰的预算、客户经济性和普通股价值可见度

评估有意对价格敏感,不是泛泛的质量评分。

[CV024, CV025, CV026, CV027]
FV004: 投资 KPI

仅基于公开证据的 Kairos 投委会式评分。

[CV024, CV025, CV026, CV027, CV036, CV041]

8.3 乐观 / 基准 / 悲观情景分析与可比公司框架

这里唯一站得住脚的估值方法是情景分析。乐观情景假设 Hermes 里程碑守住、燃料桥按时推进,Google/TVA 开始像可复制订单簿的前端,而不是一次性项目。在那个世界里,Kairos 相比普通商业化前能源创业公司,可能值得一个有意义的战略溢价,因为它同时拥有客户验证、监管进展和重复部署的可见路径。基准情景让 Kairos 维持在私营独角兽区间,但在公司披露更多预算、股权结构表和合同经济性证据之前,不给重大重估。悲观情景假设进一步延期或燃料错配,把故事拖成一个漫长、昂贵且近期可复制性有限的示范项目。 可比集合支持这种宽区间框架。TerraPower 证明大型私营资本仍支持先进核能;X-energy 显示,当里程碑可见时,公开市场可以奖励“反应堆 + 燃料”平台;Oklo 显示,即便尚未商业化的标的也能吸引公开市场热情;NuScale 则提醒投资者,一旦结构和客户匹配破裂,首台套商业化会多快压缩价值。没有单一可比公司能干净给 Kairos 定价,但同业集合有助于框定战略上行与下行的边界。[CV008, CV009, CV013, CV014, CV015, CV016]

乐观 / 基准 / 悲观情景表
情景概率信号关键假设估值区间含义
乐观低概率上行情景Hermes 里程碑守住,燃料时间同步,客户转化扩大,披露折价明显收窄$1.6B-$2.4BKairos 开始更像可复制商业平台,而不只是一个背书很强的示范项目
基准核心情景战略位置仍强,但进度 / 燃料风险和不透明折价仍在,里程碑也只是渐进推进$0.9B-$1.4BKairos 仍值得观察,但没有私下尽调,不显得明显便宜
悲观实质下行尾部进一步延误、燃料推迟或客户转化疲弱,把故事推向漫长示范经济性$0.25B-$0.75B即使技术故事仍未破产,股权价值也可能大幅压缩

区间是分析师情景带,不是可观察市场标记,也不是 DCF 输出。

[CV028, CV029, CV030, CV031, CV032, CV033]
可比估值表
可比公司状态 / 指标重要性对 Kairos 的参照价值主要限制
TerraPower私人公司;2025 年宣布融资 $650M;首座 Natrium 电站获得 DOE 大额成本分摊支持显示私人资本仍愿意大规模支持长周期先进核能对资本强度和战略溢价语境参照价值高反应堆技术不同,公私资本堆叠规模也大得多
X-energy2026 年上市;IPO 净筹资约 $1.1B;2026 年 Q1 收入和补助收入 $43M最接近的同行,显示市场如何奖励反应堆 + 燃料可见度和里程碑对商业化和燃料链比较参照价值高反应堆类型不同,公开财务披露也多于 Kairos
Oklo上市先进核能公司;尽管商业证明有限,仍披露 2025 年业绩和大量风险显示在广泛运营收入出现前,期权价值也可能很高对公开市场热情和风险感知语境参照价值中等规模、燃料回收角度和接近微反应堆的叙事不同
NuScale上市 SMR 公司;2025 年有收入和 $1.3B 现金,但也有 CFPP 下行先例同时界定商业化潜力和 FOAK 下行压缩作为压力测试可比公司,参照价值高轻水 SMR 路径,客户和资本结构历史也非常不同

Kairos 没有完美的一一对应上市可比公司,因此这组可比混合了溢价、证明和失败参照。

[CV013, CV015, CV017, CV019, CV020, CV021]
FV002: 估值敏感性

不同里程碑证明与披露质量组合下的示例估值输出。

数值为分析师估算、以十亿美元计,仅用于说明估值对执行和披露的敏感性。

[CV028, CV029, CV030, CV034, CV042]
FV003: 估值 / 回报区间

Kairos 是私营公司,当前公开估值标记未披露,因此采用宽情景区间更合适。

这些区间是情景输出,依据里程碑转化、燃料时点和披露质量,而非观察到的市场定价。

[CV028, CV029, CV030, CV032, CV033]

8.4 退出准备度、投资逻辑破裂触发点和最终尽调问题

下一条最可能的资本路径更像私募,而不是公开上市。相比近期 IPO,Kairos 更适合再做一轮私募融资、项目层面资本筹集,或引入与客户和燃料链建设绑定的战略资本。公开市场确实有先进核能投资者,但能进入这些市场的公司仍要承受剧烈波动,并且披露远多于今天的 Kairos。在 Kairos 给出更清晰的预算桥、合同经济性、股权结构表透明度和燃料时点证据之前,IPO 式估值讨论还太早。 尽调清单也相应具体。投资者需要当前普通股估值和优先权堆叠、Google/TVA 协议的经济性摘要、延期后修订的 Hermes 预算和应急资金使用情况,以及一条从 DOE HALEU 配给通向商业规模燃料供应、带日期的桥。如果这些事项干净落地,建议可能上调。如果它们恶化——或者进度和燃料时点再次滑坡——公开证据将支持从有条件兴趣转向更强硬的等待姿态。[CV036, CV037, CV038, CV039, CV040, CV041]

投资逻辑破裂与终止触发因素表
触发因素阈值对投资逻辑的传导行动含义
Hermes 进度进一步滑坡任何新的期限延期请求,或较更新计划出现可见的跨季度延误把执行折价变成核心商业化损伤从观察 / 有条件转向明确等待
燃料桥恶化商业 HALEU 或 TRISO 时间较反应堆进度走弱削弱从示范走向可复制机组群部署的路径扩大估值折价,并向下重测乐观 / 基准区间
客户转化停滞首座电站之外没有实质进展,或 Google/TVA 姿态走弱把订单簿从平台证明压缩成单项目证明削减附着在战略验证上的溢价
融资条款不透明或不利新资本附带严苛优先权、棘轮条款,或沉重项目负担普通股价值可能远弱于头部战略叙事推进前要求完整股权瀑布分析
FOAK 成本失控信号预算或应急储备恶化开始接近 NuScale 式教训叙事从纪律化迭代转向失控的资本强度除非价格大幅重置,否则从观察转为回避

这些是当前估值逻辑最快可能破裂的公开或尽调可见路径。

[CV003, CV008, CV020, CV040]
最终尽调问题表
主题缺失证据重要性负责人或尽调路径
当前估值和股权结构表当前普通股估值标记、清算优先权、认股权证和特殊权利没有这些,投资者无法把战略强度转化为真实普通股价值支撑索取最新融资文件和股权结构瀑布表
客户合同经济性PPA 定价机制、期限、下行保护、里程碑和选择权转化逻辑只有理解经济性和义务,具名交易对手才有价值索取 Google/TVA 协议的商业摘要
项目预算和应急储备延期后的 Hermes 1 和 Hermes 2 最新预算、完工尚需成本和应急储备消耗判断延误能否消化,还是会毁掉价值审阅最新预算包和挣值报告
燃料衔接到机组群规模从 DOE HALEU 配额到商业供应和 TRISO 产能的数量 / 时间衔接如果燃料节奏赶不上电厂节奏,平台故事就站不住索取包含交易对手和缓冲的整合燃料供应计划
复制项目管线按概率加权列出 Google / TVA 之外的新增公用事业、地区或客户只有它能解释为何按平台价值、而不是单项目价值给溢价审阅业务拓展漏斗和 NDA 覆盖的交易对手

这些尽调要求是把当前公开投资逻辑转成真实定价决策所需的最低集合。

[CV038, CV039]

8.5 图表

免责声明

本报告基于截至 2026-07-19 的公开信息。Kairos Power 是一家私营公司,未公开披露精确估值或承销所需的完整运营、合同和资本结构细节。

证据索引

结论
编号陈述可信度来源
CO001 Kairos Power was founded in 2016. SO002, SO007
CO002 Kairos Power is headquartered in Alameda, California. SO001, SO005
CO003 Kairos Power operates a national footprint that includes headquarters in Alameda, a manufacturing campus in Albuquerque, and a reactor demonstration campus in Oak Ridge. SO001, SO002, SO006
CO004 Kairos Power's mission is to commercialize the KP-FHR as a clean, affordable, and reliable source of firm electricity. SO004, SO007
CO005 Kairos Power's technical roots are publicly tied to UC Berkeley molten-salt and high-temperature reactor research associated with Mike Laufer, Ed Blandford, and Per Peterson. SO030, SO031
CO006 Mike Laufer is Kairos Power's CEO and co-founder. SO002, SO007
CO007 Ed Blandford is Kairos Power's chief technology officer and a co-founder. SO002, SO024
CO008 Jeff Olson is the executive publicly associated with Kairos Power's business development and finance messaging around customer demand. SO002, SO007
CO009 Kairos Power says it now employs more than 500 team members nationwide. SO006
CO010 Kairos Power's commercialization strategy emphasizes iterative development and vertical integration to improve cost and schedule certainty. SO004, SO005
CO011 The KP-FHR commercial reactor uses fluoride-salt cooling, high-temperature operation, and TRISO annular pebble fuel. SO003, SO020
CO012 Kairos Power's minimum commercial KP-FHR configuration is a dual-unit plant with two 75 MWe reactors producing up to 150 MWe. SO003, SO034
CO013 Kairos argues that low-pressure Flibe coolant and TRISO fuel allow a simpler, inherently safer plant design than conventional high-pressure water-cooled reactors. SO003, SO026
CO014 Hermes 1 is a 35 MWth low-power test reactor in Oak Ridge that will not produce electricity. SO017, SO019
CO015 The NRC issued Kairos Power a construction permit for Hermes in December 2023, the first U.S. approval to build a non-water-cooled reactor in more than 50 years. SO017, SO019
CO016 Kairos Power began physical construction on Hermes in July 2024. SO012, SO024
CO017 Kairos Power began nuclear safety-related construction on Hermes in May 2025. SO006, SO027
CO018 The NRC issued Hermes 2 construction permits in November 2024 for two 35 MWth test reactors using the same KP-FHR technology base. SO018, SO011
CO019 Kairos Power broke ground on Hermes 2 in April 2026 as its first commercial-scale reactor and the first power-producing Gen IV reactor with an NRC construction permit. SO009, SO025
CO020 Kairos Power and Google signed a Master Plant Development Agreement on October 14, 2024 for a fleet totaling 500 MW by 2035. SO007, SO026
CO021 Under the Google agreement, Kairos will develop, construct, and operate plants and sell energy, ancillary services, and environmental attributes under PPAs, with first deployment targeted for 2030. SO007, SO020
CO022 Google and Kairos described their October 2024 deal as the first corporate agreement for multiple deployments of a single advanced reactor design in the United States. SO007, SO026
CO023 The later Google-Kairos-TVA collaboration made Hermes 2 the first TVA-grid deployment and raised the plant's planned output to up to 50 MW for Google-linked data-center demand in Tennessee and Alabama. SO008, SO013
CO024 Kairos Power and TVA characterize the Hermes 2 arrangement as the first U.S. utility PPA for electricity from an advanced Gen IV reactor. SO008
CO025 DOE and Kairos executed a Technology Investment Agreement in February 2024 that provides up to $303 million of milestone-based support for the Hermes demonstration program. SO010, SO029
CO026 ORNL and Kairos announced a five-year, $27 million partnership in 2026 covering TRISO fuel, materials, spent-fuel pebbles, and remote-maintenance work needed for commercialization. SO027, SO028
CO027 Kairos Power uses Engineering Test Units and in-house manufacturing to learn construction and cost drivers before full commercial deployment. SO014, SO015
CO028 Kairos Power selected brownfield or legacy nuclear-industrial sites in Alameda and Oak Ridge as part of its deployment strategy. SO005, SO006
CO029 Kairos Power's current business model combines cost-shared public development funding, vertically integrated manufacturing, and future corporate or utility PPAs rather than existing electricity sales. SO004, SO007, SO008
CO030 KP-OMADA is Kairos Power's alliance with leading North American utilities and generating companies to support licensing, manufacturing, construction, and commercialization. SO004
CO031 Public sources consistently frame Kairos Power's first Google-linked commercial deployment as a 2030 target and the broader fleet as a 2035 target. SO007, SO020, SO023
CO032 The Google agreement bakes milestone-based accountability into development, making Google a de-risking partner rather than only a future offtaker. SO007
CO033 Neither Kairos Power's retained official materials nor the retained market-data pages disclose public revenue, ARR, or customer-count figures. SO002, SO032
CO034 Retained public sources do not disclose an exact current post-money valuation for Kairos Power, even though private-market trackers expose round and investor metadata. SO032, SO033
CO035 Caplight identifies Kairos Power investors that include Breakthrough Energy Ventures, Khosla Ventures, Prelude Ventures, Lowercarbon Capital, Schooner Capital, The Yucaipa Companies, and the Department of Energy. SO032
CO036 Brand Hopper summarizes Kairos Power as having assembled more than $600 million of combined public and private funding, but the article does not provide a full primary-source ledger. SO033
CO037 Utility Dive's Nuclear Innovation Alliance coverage says first-of-a-kind advanced-nuclear projects still face unclear business cases, uncertain development costs, and long preconstruction timelines that slow sponsor commitments. SO022
CO038 TechCrunch characterizes Kairos Power's 2030 first-reactor and 2035 fleet schedule as aggressive because a decade is a short window in nuclear power. SO034
CO039 Kairos Power is best classified as a late-stage commercialization startup because it is building demonstrations and contracting future output rather than operating a revenue-scale power fleet today. SO008, SO017
CM001 Kairos Power’s relevant market is the sale of clean firm electricity and related grid services into specific U.S. load pockets—especially hyperscaler data centers, utility service territories, and selected industrial campuses—rather than the entire global nuclear sector or all power-generation capex. SM003, SM012, SM018, SM019
CM002 EIA says that after roughly 15 years of nearly flat U.S. electricity consumption, demand increased by 2.1% per year on average over the prior five years and is projected to keep growing through 2050 at 0.9% to 1.6% annually, with data-center server energy use a major factor. SM003, SM004
CM003 The IEA projects global electricity demand from data centers will more than double by 2030 to around 945 TWh, creating a materially larger pool of power demand than existed when many first-wave SMR market decks were written. SM002
CM004 The IEA further says electricity demand from AI-optimized data centers is projected to more than quadruple by 2030, making AI a specific demand accelerator rather than generic digital-load growth. SM002
CM005 In the United States, the IEA expects data centers to account for almost half of electricity-demand growth through 2030, which strengthens the commercial case for new clean firm generation located near large-load campuses. SM002
CM006 Google and Kairos signed the first corporate agreement to buy power from multiple SMRs, targeting up to 500 MW by 2035 with the first deployment intended by 2030. SM005, SM008, SM013, SM019
CM007 Google explicitly frames the Kairos deal as part of its 24/7 carbon-free energy strategy and as infrastructure needed to support AI-driven power demand, showing that the buyer is purchasing reliability and clean-energy attributes together. SM012, SM013
CM008 Data-center power demand has become a concrete nuclear demand signal rather than an abstract decarbonization theme, because hyperscalers are now signing named contracts and public agencies are updating load-growth forecasts around those facilities. SM002, SM003, SM005, SM013
CM009 DOE’s Pathways to Commercial Liftoff work estimates advanced nuclear could add about 200 GW of U.S. capacity by 2050 if new deployment begins by 2030 and ramps to 13 GW per year by 2040. SM016, SM017
CM010 DOE also characterizes the broader U.S. clean-firm-power need at 550-770 GW by 2050, implying that even a successful advanced-nuclear buildout would address only a portion of the reliability gap. SM016, SM017
CM011 DOE’s preliminary siting work indicates operating or recently retired nuclear sites could host roughly 60-95 GW of new nuclear capacity, depending on reactor size. SM015
CM012 The same DOE siting analysis says former coal-plant sites could support an additional 128-174 GW of nuclear capacity, expanding the practical site funnel for advanced reactors. SM005, SM015
CM013 DOE’s advanced-reactor program materials describe SMRs as suitable for power generation, process heat, desalination, and other applications, with modularity and load-matching as core advantages. SM006, SM018
CM014 Kairos’ usable SAM is narrower than the full advanced-nuclear TAM because its current commercial model is oriented toward multi-unit deployments in selected service territories and load centers, not every geography or reactor application. SM008, SM012, SM019, SM020
CM015 Third-party coverage reports Kairos’ minimum commercial plant as two 75-MWe units, or 150 MWe total, with a smaller 50 MWe KP-X option also described—sizes that fit campus-scale or regional procurement better than one-gigawatt replacement projects. SM008
CM016 Kairos’ early customer pattern pairs a hyperscaler offtaker with a utility or service-territory partner, first via Google and later via the Google-Kairos-TVA collaboration. SM019, SM020, SM021
CM017 The buyer map for Kairos divides into hyperscalers as economic demand sponsors, utilities and public-power actors as siting/grid integrators, and industrial users as an adjacent future segment where heat and power can be co-sold. SM018, SM019, SM020
CM018 Budget ownership for first-of-a-kind Kairos deployments is more likely to sit with large-load buyers under long-term PPAs and structured development agreements than with merchant wholesale power markets. SM012, SM013, SM019
CM019 DOE and trade reporting both emphasize demand aggregation from large power consumers as an important mechanism for enabling serial advanced-nuclear deployments and lowering development costs. SM005, SM016
CM020 HALEU remains a real adoption bottleneck because many advanced reactor designs require it and commercial supply is currently limited. SM010, SM014
CM021 Kairos has secured the HALEU needed for Hermes’ first fuel load, reducing fuel risk for the first demonstration while leaving longer-term fleet fuel availability dependent on broader supply-chain development. SM010, SM014
CM022 GAO found that NRC had difficulty hiring and retaining the staff needed to license advanced reactors, so licensing throughput can bottleneck sector deployment even when customer demand is present. SM009, SM011
CM023 Independent reporting says skeptics still view advanced reactors as largely unproven in the Western world and warn that early deployments may cost more than advertised. SM007, SM023
CM024 NuScale’s cancelled 462-MW Carbon Free Power Project is a concrete adverse signal that first-wave utility-offtake structures can fail on cost, financing, and buyer-concentration grounds. SM007, SM023
CM025 Existing or retired nuclear and coal sites can reduce siting friction by reusing infrastructure, community familiarity, and in some cases prior licensing work. SM005, SM015
CM026 For advanced nuclear, adoption timing is governed as much by siting, fuel availability, and licensing capacity as by raw electricity-demand growth. SM002, SM010, SM011, SM015
CM027 The best near-term markets are regions where rising large-load demand collides with decarbonization goals and a shortage of clean firm alternatives, making project-specific 24/7 power more valuable than generic grid energy. SM002, SM003, SM013, SM020
CM028 Google’s 500 MW agreement is strategically important as an anchor orderbook, but it remains tiny relative to DOE’s roughly 200 GW advanced-nuclear capacity outlook and should not be mistaken for broad market capture. SM013, SM016, SM017
CM029 Kairos’ near-term SOM is best measured in early fleet deployments and service-territory-specific projects, not in generic global reactor counts or undifferentiated nuclear market values. SM019, SM020, SM022
CM030 NEI says 90 nuclear projects are in development across North America, eight are already breaking ground, and many aim to start by 2030, signaling rising competition for sites, regulators, fuel, and construction talent. SM009
CM031 Kairos’ market timing aligns with a 2030-2035 window in which utilities, hyperscalers, and DOE want clean firm capacity before advanced nuclear becomes a mature commodity industry. SM005, SM012, SM016, SM019
CM032 Status-quo substitutes for the same reliability job include combined-cycle gas, peakers, renewables paired with storage, uprates or life-extension of existing nuclear, and demand-response portfolios. SM003, SM013, SM018
CM033 Because Google wants Kairos plants sited in relevant service territories to power specific data centers, interconnection geography matters almost as much as total national demand growth. SM013, SM019, SM020
CM034 Public sources do not disclose Kairos-specific PPA pricing, achieved levelized cost of energy, or a bankable first-fleet construction-cost curve, leaving the most important SAM-to-SOM bridge unverified. SM007, SM019, SM020
CM035 The central contradictory estimate in this market is not whether demand is rising—the demand case is strong—but how quickly advanced nuclear can convert that demand into financeable projects at acceptable cost. SM002, SM007, SM011, SM016
CM036 SMRs’ ability to be matched with loads and scaled to demand makes them structurally more relevant to campus and regional procurement than traditional gigawatt-class baseload projects. SM006, SM018
CM037 Kairos itself now presents rising electricity demand and data-center load growth as part of the macro thesis for clean nuclear, indicating its GTM story is aligned with sector demand pull rather than pure technology push. SM012, SM019
CM038 The combined DOE site-reuse envelope of 188-269 GW across existing nuclear and coal sites is large enough to matter for sector growth, but DOE still notes that utilities, communities, and capital costs will determine whether technical siting potential becomes real projects. SM005, SM015
CM039 Advanced nuclear demand extends beyond simple baseload energy into ancillary services, environmental attributes, and industrial heat, broadening the revenue logic relative to a plain merchant power sale. SM012, SM018, SM019
CP001 Kairos competes inside a crowded clean-firm-power landscape that includes direct advanced-reactor peers (X-energy, TerraPower, NuScale, Holtec, GE Hitachi, Westinghouse, Terrestrial Energy, Last Energy), adjacent microreactor and fusion narratives, and status-quo substitutes such as larger reactor projects and other dispatchable power options. SP003, SP006, SP008, SP011, SP013, SP019, SP021, SP026
CP002 For buyers, the sharper competitive cut is by job-to-be-done—utility-scale replacement, staged service-territory deployment, industrial heat, on-site modular power, or long-dated fusion optionality—rather than by the generic label “SMR.” SP001, SP003, SP008, SP018, SP026
CP003 Kairos’ public technical differentiation is a fluoride-salt-cooled high-temperature reactor using TRISO annular pebble fuel, packaged as a 150 MWe commercial plant in a two-over-one, two-by-75-MWe configuration and developed through an iterative test-to-learn model. SP001, SP002, SP028
CP004 Kairos’ commercial differentiation is not only its reactor architecture but its staged commercialization path: Hermes licensing, a named Google orderbook, and a later TVA-linked service-territory collaboration. SP028, SP029, SP030, SP031
CP005 X-energy positions the Xe-100 as an 80 MWe, 200 MWt high-temperature gas-cooled reactor that can deliver approximately 565°C steam, use TRISO-X fuel, and scale to four-to-twelve units per site. SP003, SP004
CP006 Amazon and Energy Northwest give X-energy unusually strong customer proof: the Washington project is described as four SMRs for about 320 MW initially with an option to scale to 960 MW, and Amazon says its investment supports more than five gigawatts of X-energy equipment manufacturing capacity. SP005, SP022
CP007 TerraPower’s Natrium sits in a larger near-term scale class than Kairos, with NRC materials describing a 345 MWe sodium fast reactor and TerraPower highlighting DOE’s 50/50 ARDP cost share with up to $2 billion authorized for the project. SP008, SP027
CP008 TerraPower’s utility-grade signal is strengthened by PacifiCorp planning work: TerraPower says the 2023 IRP includes two additional Natrium systems, for 1,500 MW of advanced nuclear across three total Natrium reactors. SP009, SP023
CP009 NuScale retains the strongest formal regulatory credential in this peer set because its official product page says the NuScale Power Module is the first and only SMR to receive NRC design approval. SP006, SP007
CP010 NuScale markets a 77 MWe light-water module using standard LWR fuel, with a 12-module plant configuration capable of up to 924 MWe and use cases spanning data centers, process heat, hydrogen, and microgrids. SP006, SP007
CP011 NuScale’s 2025 10-K discloses that it has not yet entered into a binding customer contract to deliver modules and warns that cost competitiveness, commercialization, and future funding remain material risks; adverse reporting on the CFPP termination reinforces those concerns. SP007, SP020
CP012 GE Hitachi and OPG provide one of the clearest western utility-led LWR-SMR reference paths: GE Hitachi says TVA has submitted the first U.S. construction permit application for a BWRX-300 and OPG’s Darlington materials describe a 300 MW SMR at Darlington by the end of the decade, pending approvals. SP011, SP012
CP013 Holtec and Westinghouse both compete with more familiar water-reactor lineages: Holtec emphasizes an advanced PWR with approximately 600 MW in a dual-unit envelope, while Westinghouse presents AP300 as a proven and readily deployable SMR solution. SP013, SP024
CP014 Westinghouse eVinci and Last Energy represent a different packaging challenge from Kairos because they target decentralized or on-site power with factory-built, smaller-scale offerings rather than a 150-MWe-class grid-connected clean-firm-power wedge. SP025, SP026
CP015 Rolls-Royce SMR is aimed at a much larger grid block than Kairos: Rolls says each plant will generate 470 MW of low-carbon energy and power about one million homes for at least 60 years. SP001, SP021
CP016 Terrestrial Energy remains visible as a molten-salt-adjacent entrant, but the current public evidence pack is materially thinner than the disclosures available for Kairos, TerraPower, X-energy, or NuScale. SP014, SP015
CP017 Fusion firms such as Commonwealth Fusion Systems and TAE are narrative competitors for capital and future clean-energy mindshare, but they are not equivalent near-term commercial power substitutes for Kairos’ 2030-oriented deployment plan. SP016, SP017, SP032
CP018 CFS explicitly positions SPARC to achieve Q>1 in 2027 and describes ARC as the successor grid-scale fusion power plant, keeping fusion relevant to long-duration investor narratives even if it does not solve Kairos’ immediate buyer job. SP016, SP032
CP019 TAE frames commercial fusion as a compact, cost-effective, practical clean-energy solution and says its stable-plasma design now offers twice the performance with half the hardware, but those claims remain much earlier in commercial delivery than Kairos’ reactor-specific deployment path. SP017
CP020 Kairos appears strongest where buyers value staged deployment increments, a named hyperscaler anchor, and a build-test-learn commercialization path rather than immediate 300-470 MW block power. SP001, SP002, SP029, SP030, SP031
CP021 TerraPower, Holtec, Westinghouse AP300, GE Hitachi BWRX-300, and Rolls-Royce look stronger than Kairos where the buyer wants a larger utility-scale block, familiar water-reactor lineage, or a high-visibility reference project. SP011, SP012, SP013, SP021, SP024
CP022 X-energy looks stronger than Kairos for buyers who explicitly need industrial steam, because X-energy’s public materials make 565°C process-heat capability central to the product story. SP003, SP005, SP018
CP023 NuScale’s standard LWR fuel and conventional water-cooled architecture make its fuel-supply and licensing story simpler in one dimension than many advanced-fuel peers, even if its commercial packaging has been weaker. SP006, SP019, SP024
CP024 Kairos and X-energy share an important competitive weakness as well as a strength: both are differentiated high-temperature concepts, but both remain exposed to first-wave advanced-fuel and first-of-a-kind execution risk. SP001, SP003, SP004, SP019
CP025 TerraPower’s competitive strength is offset by shared advanced-reactor bottlenecks, including HALEU dependence and sodium-system complexity, which keep its moat from being purely a function of size or DOE support. SP008, SP019, SP027
CP026 LWR-based peers benefit from familiar BWR or PWR technology and standard fuel, reducing one category of buyer anxiety relative to advanced-fuel and novel-coolant entrants. SP006, SP011, SP013, SP024
CP027 That same LWR familiarity can also limit differentiation for high-temperature industrial heat or smaller staged campus deployments where Kairos and X-energy can present a more tailored product story. SP001, SP003, SP018, SP024
CP028 Before a buyer commits to a specific site, fuel path, and regulatory strategy, switching costs remain limited and multi-homing across nuclear pathways is feasible. SP005, SP022, SP030
CP029 After a buyer commits to a specific deployment path, switching costs rise sharply because licensing, site preparation, utility relationships, and engineering assumptions become technology-specific. SP012, SP027, SP028
CP030 Public evidence does not support an apples-to-apples price ranking across Kairos and peers; most vendors disclose architecture and milestones but not delivered cost, PPA price, EPC terms, escalation clauses, or who bears delay risk. SP007, SP014, SP015, SP020, SP029
CP031 Customer proof is now visibly multi-homed across the category: Kairos has Google and TVA-linked positioning, X-energy has Amazon and Energy Northwest, TerraPower has PacifiCorp, and GE Hitachi has OPG and TVA reference paths. SP005, SP012, SP022, SP023, SP029, SP030
CP032 Kairos’ moat therefore depends less on being the only company selling nuclear to large loads and more on whether Hermes, Hermes 2, Google, and TVA become a repeatable commercialization system. SP002, SP028, SP029, SP030
CP033 Status-quo substitutes still matter because larger or more established nuclear pathways can offer lower perceived FOAK risk even when their power blocks are less tailored to Kairos’ preferred use cases. SP012, SP020, SP021, SP024
CP034 Last Energy’s build-own-operate, behind-the-meter PPA model is a real packaging challenge in on-premises industrial and AI-load markets because it reduces customer burden around plant development and operation. SP026
CP035 Westinghouse eVinci is better understood as a microreactor competitor for remote or decentralized use cases than as a direct substitute for Kairos’ current commercial plant configuration. SP025
CP036 The Darlington project materially improves GE Hitachi’s competitive credibility because it provides a concrete 300-MW SMR deployment path in a western utility setting rather than only a conceptual product page. SP011, SP012
CP037 Data Center Frontier’s treatment of the CFPP cancellation is an important adverse base-rate signal: rising costs and stronger alternative power options can still undermine the first commercial SMR site even after years of sector enthusiasm. SP007, SP020
CP038 Kairos should therefore be underwritten as a specific buyer-job solution—especially for staged clean-firm-power deployments in the Google/TVA mold—rather than as a blanket category winner across all advanced nuclear segments. SP001, SP002, SP029, SP030, SP031
CP039 For data-center-centric buyers, plausible alternatives to Kairos still include X-energy, utility-grade LWR-SMRs, microreactor models, and non-Kairos nuclear procurement structures; which vendor wins depends on site constraints, timeline, power-block size, and risk appetite more than on category labels. SP003, SP006, SP025, SP026, SP029, SP031
CI001 There is no retained public evidence that Kairos Power is generating commercial reactor revenue as of the run date; the visible business remains development-stage, demonstration-stage, and pre-fleet. SI007, SI010, SI011
CI002 Kairos’ near-term cash inflows are better understood as a mix of private capital, DOE milestone payments, and partner-backed development work than as recurring electricity sales. SI001, SI003, SI008, SI011
CI003 Kairos’ official funding contract states DOE will provide up to $303 million using a performance-based, fixed-price milestone approach to support Hermes design, construction, and commissioning. SI001, SI002, SI003, SI025
CI004 DOE’s ARDP materials and independent coverage state the total Hermes project award value over seven years is $629 million, with DOE contributing $303 million. SI002, SI003
CI005 Using the public numbers, DOE’s share of the $629 million Hermes award is approximately 48%, implying the balance must be matched or otherwise financed by Kairos and its partners. SI002, SI003
CI006 The milestone structure means Kairos does not simply bill cost-plus expenses; cash receipts depend on completing agreed performance milestones, which can strengthen discipline but also create liquidity sensitivity if schedules slip. SI001, SI002
CI007 The molten-salt production facility and ETU 3.0 test-unit work show Kairos is spending capital on manufacturing, process, and construction-learning infrastructure before it has a recurring reactor-revenue base. SI004, SI005
CI008 DOE says ETU 3.0 is being used to pilot new manufacturing techniques and construction processes intended to lower the cost of building Kairos’ commercial reactor, making present spending part of a future cost-down thesis. SI005, SI011
CI009 Google’s agreement is financially important, but public sources describe future capacity targets and project timing rather than current recognized revenue, tariff levels, or gross margin. SI008, SI015, SI016
CI010 Kairos’ likely long-run monetization model spans reactor project delivery, electricity sales under PPAs, ancillary services, environmental attributes, and potentially services or lifecycle value around the installed fleet. SI008, SI009, SI011
CI011 Kairos’ own Google announcement says the company plans to develop, construct, and operate reactor plants and sell energy, ancillary services, and environmental attributes to Google under PPAs. SI008, SI009
CI012 The TVA collaboration implies Kairos’ monetization path includes utility and service-territory structuring, not merely one-off reactor equipment sales. SI009, SI011
CI013 Google’s 500 MW orderbook is best read as backlog-style demand validation rather than current revenue quality, because no public source discloses operating plants, invoiced energy sales, or realized project margins. SI008, SI015, SI016
CI014 Kairos’ commercial plant size is 150 MWe, which means the core revenue unit is a mid-scale plant rather than a 300-1000 MW block typical of larger utility nuclear projects. SI010, SI016
CI015 TechCrunch reports Hermes 2 will use two 35 MWth reactors connected to a 20 MW turbine, underscoring that Kairos’ learning assets are smaller than its planned 150 MWe commercial unit. SI014, SI016
CI016 Public monetization terms remain opaque: retained sources do not disclose reactor ASPs, PPA strike prices, fuel prices, EPC margins, O&M cost, or warranty/liquidated-damages economics. SI001, SI008, SI019
CI017 EIA’s capital-cost benchmark for an advanced nuclear brownfield case shows $7,861/kW for a 2 x AP1000, illustrating how capital-intensive new nuclear remains even before financing costs are included. SI020
CI018 Lazard explicitly notes limited public and observable data for new-build nuclear projects, which makes generic LCOE comparisons less robust than for mature generation technologies. SI021
CI019 Applying EIA’s $7,861/kW benchmark to Kairos’ 150 MWe commercial plant yields a rough overnight-capex proxy of about $1.18 billion before owner and financing costs, highlighting the scale of capital that a first fleet could require if cost curves do not improve materially. SI010, SI020
CI020 Because nuclear projects have long build cycles, schedule slips, and high financing sensitivity, small changes in capex, interest rates, or milestone timing can dominate future margin outcomes. SI001, SI020, SI021
CI021 Kairos still appears pre-revenue and development-stage even though it has credible contracts and regulatory progress, because the visible metrics are permits, prototype units, and future capacity targets rather than sales or utilization. SI006, SI007, SI008, SI010
CI022 Capital adequacy cannot be fully underwritten from public evidence because cash on hand, burn rate, runway, debt, restricted cash, and committed project-level capex are not disclosed in retained sources. SI001, SI008, SI019
CI023 Public funding reduces but does not eliminate financing dependency: DOE’s ARDP support is meaningful, yet official materials still imply a substantial private-match burden and continued execution dependence. SI001, SI003, SI004
CI024 Hermes, the molten-salt facility, and ETU 3.0 show Kairos is making pre-revenue investments across licensing, manufacturing, and construction-learning layers before any recurring fleet cash flow exists. SI004, SI005, SI006
CI025 Securing HALEU for Hermes’ first fuel load narrows one critical demo-stage input risk, but DOE’s own HALEU materials make clear that commercial supply remains limited for the wider advanced-reactor sector. SI012, SI026
CI026 The ORNL / DOE collaboration is financially relevant because it can lower technical-development burden and improve commercialization support even if it is not itself recognized revenue. SI013, SI011
CI027 TechCrunch’s October 2024 coverage explicitly called the 2030 commercial timeline optimistic, which means schedule slippage remains a direct risk to revenue onset and capital efficiency. SI015, SI016
CI028 The Big Tech-backed fission-startup coverage shows Kairos is operating in a capital-rich but crowded financing landscape, where peer fundraising and customer narratives are also attracting attention. SI017, SI023, SI024
CI029 X-energy’s approximately $700 million Series D shows peer nuclear startups still require very large private rounds even after strategic backing and public-private support, which is an important comparable for Kairos’ future capital needs. SI023, SI024
CI030 NuScale’s 2025 10-K is a useful adverse comparable because it combines strong regulatory credentials with disclosures that it has not yet entered a binding customer-delivery contract and continues to face significant commercialization and funding risk. SI022
CI031 For Kairos, future margin quality should be judged more against capital intensity, fuel availability, contracting structure, and schedule execution than against category enthusiasm alone. SI019, SI020, SI022, SI027
CI032 Berkeley’s nuclear-renaissance feature supports a long-run services angle by noting that Kairos’ high-temperature components would likely be replaced every 15 to 25 years and may be lighter than those of conventional reactors. SI018
CI033 Caplight confirms that Kairos has a visible funding-round and investor history, but the fetched public output is too incomplete to underwrite exact valuation, round size, or current secondary-market marks with confidence. SI019
CI034 Compared with tertiary valuation pages, the stronger financial-quality signals in public evidence are official DOE commitments, named customer contracts, and construction or fuel milestones. SI001, SI003, SI008, SI019
CI035 Sales-efficiency proxies such as CAC, payback, or funnel conversion are not publicly available for Kairos; the more relevant GTM proxy is the length and complexity of utility, regulator, and hyperscaler procurement cycles. SI008, SI009, SI016
CI036 Kairos’ eventual capital stack likely combines corporate equity, DOE milestone receipts, and later project finance, yet no retained public source discloses debt scale, lender commitments, or balance-sheet cash. SI001, SI003, SI022
CI037 The novel fixed-price milestone contract likely improves execution discipline and external credibility, but it can also increase working-capital pressure if milestone timing and cash disbursement diverge. SI001, SI002, SI025
CI038 Public traction metrics are mostly non-financial: cited sources emphasize permits, prototype installations, capacity targets, and contracts rather than revenue, backlog conversion, or utilization. SI005, SI006, SI008, SI016
CI039 The correct public-data verdict is research-more: Kairos has unusually credible demand and public support for a private advanced-nuclear company, but investors still need private budgets, cash data, contract economics, and project-finance terms before underwriting revenue or margin. SI001, SI008, SI019, SI020, SI022
CE001 Kairos’ product is best understood as a commercialization stack that sells clean firm power outcomes, not a standalone reactor SKU: the public surface spans a commercial KP-FHR plant, demonstration reactors, fuel and salt capabilities, and site-specific deployment infrastructure. SE001, SE002, SE024, SE025
CE002 The public asset stack already includes the commercial KP-FHR plant, Hermes 1, Hermes 2, Alameda engineering and prototyping, Albuquerque manufacturing and salt/fuel work, and the Oak Ridge demonstration campus. SE003, SE004, SE005, SE006, SE007
CE003 Kairos’ technology page defines the KP-FHR commercial reactor as a 150 MWe [2 x 75 MWe] fluoride salt-cooled high-temperature reactor using TRISO annular pebble fuel. SE001, SE009
CE004 The Alameda site is positioned as the engineering-design, rapid-prototyping, and iterative-testing hub for Kairos, with brownfield reuse and about $25 million of disclosed investment supporting technology development. SE005, SE006
CE005 Kairos describes its Albuquerque campus as dedicated to manufacturing, salt production, fuel development, and large-scale non-nuclear testing, with more than 140 jobs and about $125 million invested to date. SE004, SE006
CE006 The Tennessee site at Heritage Center is where Kairos is building the Hermes reactor series and piloting advanced construction techniques plus operating experience to optimize cost and support commercial-fleet deployment. SE003, SE007
CE007 Hermes 1 is a low-power test reactor that supports development of KP-FHR technology and, unlike later plants, is not intended to produce electricity. SE009, SE011
CE008 Hermes 1 was the first non-light-water reactor to receive a U.S. construction permit in more than 50 years, giving Kairos unusually strong regulatory proof for a startup product line. SE009, SE011
CE009 Kairos and the NRC frame Hermes 2 as the first commercial-scale, power-producing Gen IV reactor to receive a construction permit, making it the bridge asset between non-power demonstration and revenue-grade deployment. SE007, SE010, SE020
CE010 NRC materials describe Hermes 2 as two low-power test reactors at 35 MWth each, fueled by HALEU TRISO pebbles and connected to a shared steam-powered Rankine conversion system. SE010, SE022
CE011 World Nuclear News says Hermes 2 will supply up to 50 MW of electricity to the TVA grid and is the immediate precursor to Kairos’ full-scale commercial plants. SE007, SE020
CE012 Kairos’ manufacturing model is explicitly geographic: equipment modules for Hermes 2 are fabricated in Albuquerque and shipped to Oak Ridge for assembly. SE004, SE007
CE013 Hermes 2 is intended to use modular construction methods, including precast concrete and a seismically isolated foundation, to shrink timelines, lower nuclear construction costs, and support a repeatable design. SE003, SE007
CE014 Kairos’ operating model is anchored in iterative design-build-test cycles and vertical integration rather than a single-step leap to a large commercial reactor. SE002, SE004, SE005
CE015 The architecture is therefore better described as reactor physics plus manufacturing, salt, fuel, and construction-learning loops than as a conventional utility plant sold off a fixed catalog. SE001, SE002, SE003, SE004
CE016 DOE says ETU 3.0 is being used to pilot new manufacturing techniques and construction processes to lower the cost of Kairos’ commercial reactor. SE012, SE002
CE017 The molten-salt production facility makes salt production a disclosed internal enabling asset for Kairos rather than an invisible outsourced input. SE013, SE004
CE018 Kairos’ public fuel path combines in-house annular graphite pebble capability with BWXT collaboration on commercial TRISO manufacturing and DOE-managed HALEU allocation. SE001, SE015, SE016, SE021
CE019 BWXT says its collaboration with Kairos will explore commercial TRISO production for Hermes 2 and subsequent reactor deployments, moving the company toward a more scalable fuel-manufacturing pathway. SE021, SE007
CE020 DOE’s HALEU materials state that most advanced reactor designs require HALEU and that supply gaps can delay deployment, making fuel availability a first-order product dependency rather than a back-office procurement detail. SE015, SE016, SE017
CE021 Kairos’ first-fuel allocation reduces risk for the Hermes demonstration line, but public DOE and WNN material still imply that fleet-scale fuel readiness remains a broader sector dependency. SE015, SE017, SE018
CE022 The ORNL / DOE collaboration is a product-strengthening dependency because it adds fuel, materials, and spent-fuel research support to Kairos’ commercialization system. SE019, SE002
CE023 The customer workflow implied by public sources runs from a clean-firm-power buyer need to site and service-territory alignment, through licensing and construction-learning assets, into power-producing demonstration and then repeatable fleet deployment. SE003, SE007, SE010, SE024, SE025
CE024 Kairos differentiates from conventional water-cooled nuclear designs through its low-pressure fluoride-salt coolant, TRISO pebble fuel, and explicit demo ladder. SE001, SE009, SE010
CE025 Brownfield reuse across Alameda and Oak Ridge suggests the company’s product strategy includes repurposed industrial infrastructure as part of deployment practicality. SE003, SE005, SE006
CE026 Kairos’ differentiation also includes a factory-leaning manufacturing narrative, where construction learning and module fabrication are meant to drive repeatability instead of relying only on custom site work. SE003, SE007, SE008
CE027 The public trust case is stronger on physical design and NRC process than on field performance because the key proof points are permits, test units, and facilities rather than operating commercial plants. SE009, SE010, SE011
CE028 NRC’s Hermes 2 page explicitly describes functional containment as being implemented principally by the high-temperature TRISO particle fuel. SE010, SE009
CE029 No retained public source provides achieved uptime, steady-state operating data, or commercial reliability metrics for a Kairos reactor because the fleet is still pre-operational. SE009, SE010, SE011
CE030 Public sources remain thin on digital controls, OT cybersecurity architecture, and named control-system vendors, leaving a material diligence gap in the technology stack. SE009, SE010
CE031 Google’s public comment in The Time is Now frames Kairos’ factory-based manufacturing approach as a path toward lower-cost, cleaner power. SE008, SE024
CE032 The location pages and About page collectively show vertical integration across design, manufacturing, salt, fuel, testing, and deployment geographies. SE003, SE004, SE005, SE006
CE033 Product maturity is strongest on prototypes, facilities, permits, and construction methods, and weakest on operating commercial proof and runtime performance. SE007, SE009, SE010, SE012
CE034 The current support model is physical and programmatic: test assets, facilities, and regulatory workstreams are visible, while post-sale operations support and field-maintenance metrics remain undisclosed publicly. SE003, SE004, SE009, SE010
CE035 DOE’s ARDP page reinforces Kairos’ use-case fit by noting that SMRs can provide power generation, process heat, desalination, and can be matched to loads and scaled to demand. SE014, SE024
CE036 The Google and TVA pathways show that Kairos’ product is being shaped for data-center-adjacent 24/7 clean power and service-territory deployment, not just generic baseload generation. SE024, SE025
CE037 World Nuclear News ties the Hermes series to Oak Ridge-origin technologies including TRISO fuel and Flibe molten fluoride salt coolant, connecting Kairos’ product story to a specific reactor heritage rather than purely startup branding. SE007, SE023
CE038 The overall technology verdict is that Kairos has unusually strong public visibility into assets, sites, and regulatory progress for an advanced-reactor startup, but still material diligence gaps around operating performance and cyber/control detail. SE009, SE010, SE017, SE021
CU001 Google is the clearest public economic buyer in Kairos Power’s commercial story: it initiated the 2024 master agreement to secure clean firm electricity for its data centers rather than simply endorsing the technology as a strategic investor. SU001, SU002, SU003
CU002 The customer stack is split between an end-demand buyer and a grid intermediary: Google anchors the load and clean-energy-attribute demand, while TVA is the utility counterparty that will purchase Hermes 2 output and move those attributes through its system. SU006, SU010, SU021
CU003 Google’s October 2024 agreement with Kairos was described by both parties as the world’s first corporate agreement for multiple advanced-reactor deployments of the same design. SU001, SU002, SU021
CU004 The master agreement targets up to 500 MWe of carbon-free electricity by 2035, starting with Hermes 2 in 2030. SU001, SU002, SU021
CU005 Kairos’ dedicated Google page states that the TVA power purchase agreement for Hermes 2 is the first-ever U.S. advanced-reactor PPA. SU021, SU006, SU010
CU006 The August 2025 collaboration specifies that Hermes 2 will deliver up to 50 MW to the TVA grid powering Google data centers in Tennessee and Alabama. SU006, SU011, SU012
CU007 The first disclosed regional use case is not generic grid decarbonization but specifically Google data center operations in Montgomery County, Tennessee, and Jackson County, Alabama. SU006, SU011
CU008 Public customer proof is therefore stronger on contract structure than on delivered-energy usage: no reactor is operating yet, but the buyer, offtaker, locations, and milestone dates are named. SU006, SU010, SU011
CU009 Google’s own explanation for the deal centers on rising electricity demand from AI and data centers plus its 24/7 carbon-free-energy objective, not on a speculative science-project narrative. SU002, SU018, SU025
CU010 Google reported a 27% increase in electricity demand for its data centers in 2025 while still pushing record clean-energy procurement, reinforcing why it is willing to contract for advanced nuclear capacity. SU018, SU025
CU011 Google’s 2026 sustainability reporting says it signed more than 12 GW of net-new clean energy in 2025, showing that Kairos sits inside a much broader and growing procurement program rather than as a one-off symbolic deal. SU025, SU018
CU012 Google also described 2024 as its largest-ever clean-energy procurement year and linked new utility rate structures with the need to scale firm clean resources such as nuclear. SU023, SU025
CU013 The buyer-user-payer logic of the first Kairos deployment is unusual but legible: Google is the ultimate user-beneficiary, Kairos develops and operates the plant, and TVA remains the utility settlement and grid-delivery layer. SU006, SU010, SU021
CU014 TVA is the first U.S. utility publicly disclosed as a direct purchaser of electricity from a Gen IV reactor, giving Kairos utility-grade offtake proof earlier than many advanced-reactor peers. SU006, SU008, SU010
CU015 The 2025 TVA agreement is the first plant-specific commercial contract publicly shown under the broader Google orderbook. SU006, SU008, SU011
CU016 POWER reported that Hermes 2 electricity would be sold to TVA at market-based prices fixed by time of day, but the contract length and full pricing structure were not publicly disclosed. SU010
CU017 Later units in the 500 MWe program remain milestone-gated: the public record shows a master framework and one binding Hermes 2 utility PPA, not a fully specified fleet-wide set of plant PPAs today. SU002, SU006, SU010
CU018 The Duke Energy collaboration announced in 2024 is not a Kairos-specific power purchase agreement; it is a memorandum-of-understanding framework to create large-customer tariffs that could support future carbon-free generation. SU014, SU020, SU023
CU019 That Duke framework matters because Google joined Amazon, Microsoft, and Nucor in signaling willingness to support novel rate structures for clean firm capacity in the Carolinas. SU014, SU015, SU016
CU020 Duke said the ACE and Clean Transition Tariff pathways would still require regulatory approvals in North Carolina and South Carolina, so this channel is real but not yet bankable customer proof for Kairos. SU014, SU020
CU021 Google explicitly said the orderbook approach is intended to accelerate repeated reactor deployments and improve cost certainty for Google and other customers over time. SU002, SU001
CU022 The commercialization path therefore includes a plausible repeat-purchase loop: one demonstration-linked plant, followed by additional deployments through 2035 if milestones and economics improve. SU002, SU006, SU021
CU023 Public proof quality is unusually strong for a startup because Google, TVA, Duke, and multiple trade outlets all describe the customer pathway in their own names rather than only via company marketing. SU002, SU014, SU026
CU024 Even with strong named counterparties, no public source discloses Kairos customer retention metrics, churn, renewal rates, NRR, or average contract duration. SU001, SU006, SU010
CU025 Because no Kairos reactor is yet operating commercially, any claim about customer durability is still a structural inference from long-lived infrastructure and counterparties rather than a measured cohort outcome. SU006, SU022
CU026 The next expansion channel may depend as much on utility-rate design and service-territory structures as on direct corporate PPAs, as shown by both the TVA model and Google’s Duke tariff work. SU006, SU014, SU023
CU027 Google’s willingness to help shape tariffs in the Carolinas suggests the buyer is not just purchasing electrons but trying to create repeatable procurement mechanisms for clean firm power. SU014, SU023
CU028 DOE warns that new reactors for data centers will take years to license and build, so customer adoption for advanced nuclear should be underwritten as gradual rather than immediate. SU022, SU024
CU029 DOE also highlights first-of-a-kind cost, metering disputes, fuel-chain buildout, and spent-fuel handling as hurdles for nuclear-powered data center models. SU022
CU030 The public named-customer roster remains partial: Google and TVA are clear, while Duke represents an enabling tariff framework rather than a signed Kairos offtake contract. SU006, SU014, SU020
CU031 Customer concentration is still material because Google is the only publicly named anchor end customer for Kairos’ advanced-reactor fleet. SU001, SU006, SU021
CU032 There is some counterparty diversification at the first-project level—TVA as offtaker, Google as end-demand anchor, and Oak Ridge/Tennessee Valley as host ecosystem—but not yet broad diversification across independent paying buyers. SU006, SU010, SU026
CU033 Expansion upside is credible because Google has publicly committed to repeated deployments through 2035 and continues to report large-scale clean-energy needs from AI-linked data-center growth. SU002, SU018, SU025
CU034 The TVA arrangement is important not only for one plant but because it offers a template for utility-mediated delivery of clean energy attributes to hyperscale buyers without relying on purely behind-the-meter structures. SU006, SU009, SU022
CU035 Public sources still do not reveal termination rights, collateral terms, minimum-take volumes, or the exact contract waterfall between Google, TVA, and Kairos. SU010, SU012
CU036 Customer quality is therefore high on strategic validation but only medium on monetization transparency: the counterparties are strong and named, while the economics remain mostly hidden. SU006, SU010, SU021
CU037 If Hermes 2 reaches 2030 operations and its economics improve as Google expects, the Kairos-TVA-Google structure could become a template for other utilities and hyperscale buyers. SU006, SU010, SU021
CU038 Until those milestones are met, investors should treat Kairos customer traction as milestone-backed commercial validation rather than as already-recurring delivered-power revenue. SU008, SU010, SU022
CR001 The NRC amended Hermes 1’s latest construction completion date from December 31, 2026, to April 30, 2029. SR001, SR003, SR007
CR002 The NRC accepted Kairos’ delay justification as good cause because of developmental problems attributable to the first-of-a-kind nature of the Hermes facility. SR001, SR007
CR003 The NRC safety evaluation says site preparation and removal of legacy structures required significantly more work and coordination than originally planned. SR007
CR004 The extension amendment did not authorize new activities and the NRC issued a finding of no significant environmental impact for the schedule change. SR002, SR007
CR005 A material schedule slip has already occurred before Hermes has operated, so timeline risk is no longer hypothetical. SR001, SR003, SR007
CR006 Hermes 2 has a construction permit, but the reactor remains subject to ongoing NRC oversight and a live document trail rather than being a finished regulatory story. SR005, SR006
CR007 Construction-permit projects carry real legal expiry risk because the permit can lapse if the latest completion date is missed without an approved amendment. SR001, SR007
CR008 The NRC’s part 53 rule became effective in April 2026 and creates an optional technology-inclusive path for future commercial reactors. SR025
CR009 Kairos still faces a hybrid regulatory reality because its current demonstration path is already moving under earlier permitting structures even as part 53 arrives for future applicants. SR004, SR025
CR010 Hermes uses HALEU-bearing TRISO fuel pebbles and molten fluoride salt coolant, making fuel supply and fabrication central operational dependencies. SR004, SR007, SR010
CR011 DOE says HALEU is not currently available from domestic suppliers and that supply gaps could delay advanced-reactor deployment. SR016, SR023
CR012 Kairos’ DOE-sourced HALEU contract reduces near-term startup risk for Hermes 1 but does not solve long-run fleet-scale fuel availability. SR010, SR011, SR023
CR013 ANS notes that the United States still lacks domestic HALEU enrichment capacity, making DOE allocation a bridge rather than a complete market solution. SR011, SR023
CR014 Centrus says its first new commercial-scale HALEU capacity is expected to come online by 2029, leaving timing risk if Kairos’ fleet ramps before the broader supply chain matures. SR012
CR015 BWXT’s collaboration with Kairos shows that commercial TRISO manufacturing for future reactors is still being built rather than already de-risked. SR022, SR011
CR016 DOE warns that new reactors for data centers will take years to build and that first-of-a-kind deployments are expensive. SR016, SR013
CR017 DOE also warns that metering arrangements and cost allocation can become regulatory hurdles for nuclear-powered data center models. SR016, SR015
CR018 Spent-fuel handling remains part of the residual risk stack for advanced nuclear even when it is not the immediate gating item for Hermes. SR016, SR008
CR019 Kairos itself says the Hermes series is intended to mitigate technology, licensing, supply-chain, and construction risk before the commercial fleet. SR003, SR010
CR020 Google is still the only publicly named end customer for the Kairos fleet, so customer concentration remains high. SR019, SR021, SR026
CR021 TVA is the only publicly named utility offtaker and delivery route for Kairos’ first commercial power path. SR019, SR020, SR028
CR022 Duke-style tariff experimentation shows a possible future procurement channel for clean firm power, but it also introduces ratepayer and regulatory backlash risk. SR015, SR017
CR023 Latitude reports that Duke’s proposed large-load tariff could spread grid-upgrade costs broadly and make it harder to isolate data-center cost causation. SR015
CR024 Latitude also reports that large-load customers under Duke’s proposal could face long terms, minimum-demand payments, early-exit penalties, and collateral requirements. SR015
CR025 The public Google-TVA-Kairos disclosures still do not reveal the full pricing stack or risk-allocation waterfall. SR019, SR020
CR026 Dependence on DOE funding and DOE fuel programs remains material to Hermes and to the pace at which Kairos can move into the commercial fleet. SR010, SR023, SR024
CR027 DOE OIG found that DOE did not effectively evaluate, structure, or monitor critical risks in the NuScale Carbon Free Power Project and that roughly $183 million was spent without achieving the key project objective. SR009
CR028 DOE OIG also said front-loading the NuScale cost share put nearly $143.5 million at risk when the project terminated. SR009
CR029 UAMPS and NuScale terminated the Carbon Free Power Project because subscription appeared unlikely to reach a level that could support deployment. SR014, SR013
CR030 CATF argues that the NuScale/UAMPS failure illustrates the danger of forcing a demonstration-oriented first-of-a-kind reactor into a full commercial model too early. SR013
CR031 CATF also argues that investor-owned utilities may not be ideal first movers for FOAK reactors and that data centers, heavy industry, and public utilities may be better suited to bear early deployment risk. SR013
CR032 The NuScale precedent shows that customer subscription and first-customer fit can sink a reactor program even after major licensing progress. SR009, SR013, SR014
CR033 The NRC safety evaluation shows that even basic site-preparation work required extensive coordination with DOE and other organizations, underscoring multi-party execution risk. SR007
CR034 Kairos is managing multiple parallel dependencies—Hermes 1, Hermes 2, Google/TVA commercialization, fuel fabrication with Los Alamos, and future TRISO scaling with BWXT—so execution bandwidth is a material risk variable. SR010, SR019, SR022
CR035 The Oak Ridge/Tennessee region offers ecosystem support, but growing advanced-nuclear build activity also creates competition for specialized labor and suppliers. SR018, SR030
CR036 Google’s 500 MWe orderbook and rising AI-driven electricity demand can amplify schedule pressure on Kairos because the customer-side need is scaling faster than reactor delivery can. SR021, SR026, SR029
CR037 FONSI and permit approvals reduce binary denial risk, but they do not remove schedule, cost, fuel, or commercialization risk. SR002, SR004, SR007
CR038 Large-load and behind-the-meter policy debates show that Kairos’ commercialization risk now extends beyond reactor physics into tariff design and grid cost allocation. SR015, SR016, SR017
CR039 DOE says the HALEU allocation process is ongoing and additional companies may receive allocations, implying competition for limited near-term material. SR023, SR024, SR011
CR040 Centrus’ transition toward commercial HALEU production improves the long-run fuel outlook but still leaves an interim dependency on government-supported bridging arrangements. SR012, SR023
CR041 The most monitorable thesis-break triggers are additional schedule slippage, failure to secure fuel on time, weakening of Google/TVA commitment, or FOAK cost escalation that begins to resemble the NuScale precedent. SR003, SR009, SR020, SR023
CR042 Kairos’ overall residual risk is high but not existential: regulatory progress and named customer proof are unusually strong, yet the commercialization path still depends on several interlocking first-of-a-kind systems working on time and on budget. SR004, SR010, SR019, SR021
CV001 Kairos has stronger named commercial proof than many private advanced-nuclear peers because its story already includes Google as anchor buyer, TVA as utility offtaker, and a plant-specific first deployment path. SV001, SV003, SV027
CV002 The 500 MWe Google orderbook plus the 50 MW Hermes 2 TVA pathway move Kairos beyond concept-stage commercialization and into milestone-backed deployment. SV001, SV003, SV004, SV025
CV003 Hermes 1’s deadline extension means any valuation framework must apply a meaningful execution discount for schedule risk. SV023, SV024
CV004 Retained public sources do not disclose Kairos’ current common-equity valuation, current cap table, or preference stack. SV001, SV003, SV027
CV005 Retained public sources also do not disclose full pricing, term, or risk-allocation details for the Google-TVA-Kairos customer contracts. SV003, SV004
CV006 Google’s 2026 sustainability reporting shows a very large and growing clean-power demand base, supporting real customer-side optionality for Kairos if milestones are met. SV002, SV005
CV007 DOE-backed HALEU supply for Hermes materially reduces early existential risk for the first unit and supports the credibility of the commercialization path. SV010, SV028
CV008 Commercial HALEU timing remains a valuation haircut because DOE allocation is a bridge and Centrus only expects its first new commercial-scale capacity around 2029. SV011, SV028
CV009 BWXT’s collaboration with Kairos is a positive signal for fuel manufacturing, but it also confirms that a commercial TRISO supply chain is still being assembled. SV012, SV028
CV010 The DOE OIG NuScale audit shows that FOAK advanced-nuclear value can be destroyed by weak risk oversight and poorly structured commercialization even when strategic logic is compelling. SV006
CV011 The UAMPS/NuScale termination demonstrates that customer subscription and buyer fit are first-order valuation variables for advanced-reactor projects. SV007, SV008
CV012 CATF argues that data centers and public utilities may be better FOAK customers than conventional investor-owned utilities, which is favorable to Kairos because it already combines Google and TVA. SV008, SV003
CV013 TerraPower’s 2025 $650 million fundraise proves that private capital still supports long-horizon advanced-nuclear platforms with strong milestones and strategic narratives. SV014, SV015
CV014 TerraPower’s Natrium program also shows that first-plant advanced nuclear can require very large capital stacks and public-private cost sharing. SV014, SV015
CV015 X-energy’s 2026 IPO and roughly $1.1 billion in net proceeds show that public markets can reward advanced-nuclear platforms when reactor, fuel, and project milestones are visible. SV016, SV017
CV016 X-energy’s $43 million of first-quarter 2026 revenues and grant income indicate a level of public operating disclosure that Kairos has not yet matched. SV016
CV017 Oklo’s public-market status and full-year 2025 results show that investors can assign material value to a pre-commercial advanced-nuclear platform before broad operating revenue exists. SV019, SV020
CV018 Oklo’s own disclosures also highlight the same risks Kairos still faces—fuel access, regulatory uncertainty, financing needs, and PPA execution—so public-market enthusiasm does not remove fundamental risk. SV020
CV019 NuScale’s 2025 results show that public listing, cash, and some revenue do not by themselves eliminate commercialization risk in advanced nuclear. SV021
CV020 Taken together, NuScale plus UAMPS provide the clearest downside precedent for valuation compression if FOAK commercialization stumbles. SV006, SV007, SV008, SV021
CV021 The peer set argues for wide valuation bands instead of point precision because public advanced-nuclear outcomes range from richly financed private programs to public-market volatility and outright project termination. SV014, SV016, SV019, SV021
CV022 Kairos is strategically stronger than a pure concept company because it has named customer proof, utility integration, and a visible fuel-risk reduction path. SV003, SV010, SV027
CV023 Kairos is still weaker than the best public peers on disclosure depth because it does not publish audited statements, cash balances, or SEC-level risk factor detail. SV004, SV018, SV022
CV024 The appropriate current recommendation is TRACK / CONDITIONAL rather than BUY because public evidence supports company quality more than it supports entry-price certainty. SV003, SV006, SV021
CV025 Confidence should be low-to-medium because the core thesis is unusually strong while the key underwriting variables remain private. SV004, SV010, SV023
CV026 Risk rating should remain high because schedule, fuel, and customer concentration still interact rather than diversify one another. SV003, SV023, SV028
CV027 The valuation stance is best described as under-disclosed but strategically valuable, not obviously cheap or fully supported. SV004, SV005, SV006
CV028 A reasonable bull-case range is about $1.6B-$2.4B if Hermes and TVA milestones hold, fuel timing synchronizes, and later units begin converting from framework into repeat deployments. SV003, SV011, SV014, SV016
CV029 A reasonable base-case range is about $0.9B-$1.4B if Kairos retains strategic leadership but remains private, opaque, and only incrementally de-risked over the next milestone window. SV003, SV006, SV021
CV030 A reasonable bear-case range is about $0.25B-$0.75B if further slippage, fuel bottlenecks, or weak customer conversion push Kairos toward a prolonged demonstration story rather than a commercial fleet story. SV006, SV007, SV023
CV031 The bull case depends much more on milestone conversion and repeat orderbook progress than on near-term revenue multiples. SV002, SV003, SV008
CV032 The base case assumes Kairos remains one of the strongest private advanced-nuclear franchises while still trading with a meaningful opacity discount. SV005, SV006, SV023
CV033 The bear case is explicitly informed by NuScale/UAMPS lessons that commercialization structure can fail even when technology and regulation have progressed. SV006, SV007, SV008
CV034 Public-market enthusiasm for advanced nuclear should not be copied mechanically into Kairos because public names disclose more and still remain highly risky and narrative-sensitive. SV016, SV019, SV021
CV035 Comparable logic suggests that reactor developers with visible orderbooks, fuel paths, and public financing access can earn strategic premiums, but only when disclosure and milestone credibility are strong. SV014, SV016, SV020, SV021
CV036 Exit readiness looks stronger for follow-on private financing or project-level capital formation than for a near-term IPO because public disclosure remains too thin. SV004, SV006, SV022
CV037 A strategic sale looks less likely near term than continued private capital formation because Kairos is still building foundational assets rather than operating a mature fleet. SV010, SV014, SV015
CV038 Final diligence should focus on cap table and preference terms, customer contract economics, revised schedule and budget, and the fuel bridge from DOE allocation to commercial supply. SV004, SV023, SV028
CV039 If Kairos disclosed a clean common-equity mark, customer economics, and an updated project budget with limited slippage, the recommendation could improve materially. SV004, SV023, SV027
CV040 If Hermes slips again or the fuel bridge weakens, the current track call would likely deteriorate toward an avoid-or-wait posture. SV011, SV023, SV028
CV041 Market tailwinds from AI load growth, utility interest, and policy support are real enough that Kairos remains worth active monitoring despite the valuation opacity. SV005, SV013, SV029
CV042 The right valuation method is milestone-sensitive scenario analysis rather than a conventional DCF or current revenue multiple alone. SV006, SV008, SV021
来源
编号出版方标题引文
SO001 Kairos Power Kairos Power | Advanced Nuclear Reactor Technology
SO002 Kairos Power About | Kairos Power
SO003 Kairos Power Technology | Kairos Power
SO004 Kairos Power Our Approach | Kairos Power
SO005 Kairos Power California Location | Kairos Power
SO006 Kairos Power Tennessee Location | Kairos Power
SO007 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SO008 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SO009 Kairos Power Kairos Power Breaks Ground on Hermes 2 Demonstration Plant | Kairos Power
SO010 Kairos Power License to Build: Progress on Hermes and the ETU Series | Kairos Power
SO011 Kairos Power Nuclear Regulatory Commission Approves Construction Permits for Hermes 2 Demonstration Plant | Kairos Power
SO012 Kairos Power Kairos Power Begins Construction on Hermes Low-Power Demonstration Reactor | Kairos Power
SO013 Kairos Power Clean Electricity for the Tennessee Valley | Kairos Power
SO014 Kairos Power The Engineering Test Unit Program: Learning How to Build Kairos Power Reactors | Kairos Power
SO015 Kairos Power Building Small Modular Momentum | Kairos Power
SO016 Kairos Power An Opportune Moment for Clean Nuclear Energy | Kairos Power
SO017 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SO018 Nuclear Regulatory Commission Hermes 2 – Kairos Application
SO019 U.S. Department of Energy Office of Nuclear Energy NRC Approves Construction for Hermes Reactor
SO020 POWER Magazine Google Bets Big on Nuclear: Inks Deal with Kairos Power for 500-MW SMR Fleet to Power Data Centers
SO021 Utility Dive Google, Kairos Power ink 500-MW advanced nuclear reactor deal
SO022 Utility Dive Nuclear Innovation Alliance offers road map to catalyze advanced nuclear development
SO023 World Nuclear News Google and Kairos Power team up for SMR deployments
SO024 World Nuclear News Kairos, DOE enhance collaboration on advanced reactor design
SO025 American Nuclear Society Kairos Power breaks ground on first power-producing reactor in Oak Ridge
SO026 Google New nuclear clean energy agreement with Kairos Power
SO027 American Nuclear Society TRISO pebble life cycle studied in new ORNL, Kairos Power partnership
SO028 Oak Ridge National Laboratory ORNL, Kairos Power partner to advance deployment of next-gen nuclear energy | ORNL
SO029 Oak Ridge Economic Development Initiative U.S. Department of Energy and Kairos Power Execute Novel Performance-Based, Fixed-Price Milestone Contract - Oak Ridge Economic Development Initiative
SO030 Research UC Berkeley Nuclear Power Renaissance | Research UC Berkeley
SO031 Research UC Berkeley Per F. Peterson | Research UC Berkeley
SO032 Caplight Kairos Power | Valuation, Funding Rounds & Stock Price | Caplight
SO033 The Brand Hopper Kairos Power – Founders, Business Model, Funding & Competitors
SO034 TechCrunch Future Google supplier Kairos gets approval to build two small nuclear reactors | TechCrunch
SO035 TechCrunch Here are the nuclear fission startups backed by Big Tech | TechCrunch
SM001 International Energy Agency Electricity 2026 – Analysis - IEA
SM002 International Energy Agency AI is set to drive surging electricity demand from data centres while offering the potential to transform how the energy sector works - News - IEA
SM003 U.S. Energy Information Administration Annual Energy Outlook 2026 - U.S. Energy Information Administration (EIA)
SM004 U.S. Energy Information Administration EIA releases the Annual Energy Outlook 2026
SM005 Utility Dive Google, Kairos Power ink 500-MW advanced nuclear reactor deal
SM006 Utility Dive Data center boom fuels demand for nuclear projects
SM007 Utility Dive What’s next for advanced nuclear technology?
SM008 POWER Magazine Google Bets Big on Nuclear: Inks Deal with Kairos Power for 500-MW SMR Fleet to Power Data Centers
SM009 Nuclear Energy Institute State of the Nuclear Industry 2026
SM010 U.S. Department of Energy Office of Nuclear Energy HALEU Technologies
SM011 U.S. Government Accountability Office Nuclear Power: NRC Needs to Take Additional Actions to Prepare to License Advanced Reactors
SM012 Kairos Power An Opportune Moment for Clean Nuclear Energy | Kairos Power
SM013 Google New nuclear clean energy agreement with Kairos Power
SM014 World Nuclear News Kairos secures HALEU for Hermes' first fuel load
SM015 U.S. Department of Energy Office of Nuclear Energy Could the Nation’s Nuclear Power Plant Sites Support New Reactor Builds?
SM016 U.S. Department of Energy Office of Nuclear Energy Commercializing Advanced Nuclear Reactors Explained in Five Charts
SM017 U.S. Department of Energy Office of Energy Demonstrations Sector Spotlight: Advanced Nuclear
SM018 U.S. Department of Energy Office of Nuclear Energy Advanced Reactor Demonstration Projects
SM019 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SM020 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SM021 World Nuclear News Google and Kairos Power team up for SMR deployments
SM022 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SM023 Utility Dive Nuclear Innovation Alliance offers road map to catalyze advanced nuclear development
SM024 World Nuclear News Kairos, DOE enhance collaboration on advanced reactor design
SM025 American Nuclear Society Kairos Power breaks ground on first power-producing reactor in Oak Ridge
SP001 Kairos Power Technology | Kairos Power
SP002 Kairos Power Our Approach | Kairos Power
SP003 X-energy Xe-100: High-Temperature Gas-Cooled Nuclear Reactors (HTGR) — X-energy
SP004 X-energy TRISO-X: Advanced TRISO Particle Fuel for Gen 4 Nuclear Reactors — X-energy
SP005 Amazon Amazon signs agreements for innovative nuclear energy projects to address growing energy demands
SP006 NuScale Power The NuScale Power Module | NuScale Power
SP007 Securities and Exchange Commission NuScale Power Corp. Form 10-K for fiscal 2025
SP008 TerraPower TerraPower Natrium | Advanced Nuclear Energy
SP009 TerraPower Wyoming Nuclear Energy Milestones
SP010 Holtec International Small Modular Reactor
SP011 GE Vernova Hitachi Nuclear Energy BWRX-300 Small Modular Reactor | GE Vernova Hitachi Nuclear
SP012 Ontario Power Generation Darlington New Nuclear Project Insert
SP013 Westinghouse Electric Company AP300™ SMR | Westinghouse Nuclear
SP014 Terrestrial Energy Terrestrial Energy
SP015 Terrestrial Energy Terrestrial Energy
SP016 Commonwealth Fusion Systems SPARC: Proving commercial fusion energy is possible | Commonwealth Fusion Systems
SP017 TAE Technologies Clean energy solutions for a bright future
SP018 U.S. Department of Energy Office of Nuclear Energy Advanced Reactor Demonstration Projects
SP019 World Nuclear Association Small Modular Reactors - World Nuclear Association
SP020 Data Center Frontier Commercial SMR Prospects Dim On Cancellation of First Planned U.S. Site?
SP021 Rolls-Royce SMR To Deliver Clean, Affordable Energy For All
SP022 Energy Northwest Amazon & Energy Northwest Announce Plans To Develop Advanced Nuclear Technology in Washington
SP023 TerraPower PacifiCorp Forecasts Need for Two Additional Natrium Reactors in New Regulatory Filing
SP024 Holtec International SMR-300
SP025 Westinghouse Electric Company eVinci™ Microreactor | Westinghouse Nuclear
SP026 Last Energy Last Energy | 20 MWe SMR | Fully modular, factory made
SP027 Nuclear Regulatory Commission Natrium | Nuclear Regulatory Commission
SP028 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SP029 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SP030 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SP031 Google New nuclear clean energy agreement with Kairos Power
SP032 Commonwealth Fusion Systems ARC: Putting fusion energy on the grid | Commonwealth Fusion Systems
SI001 Kairos Power U.S. Department of Energy and Kairos Power Execute Novel Performance-Based, Fixed-Price Milestone Contract | Kairos Power
SI002 POWER Magazine DOE, Kairos Unveil Milestone-Based Funding Agreement for Advanced Nuclear Demonstration Project
SI003 U.S. Department of Energy Office of Nuclear Energy Energy Department’s Advanced Reactor Demonstration Program Awards $30 Million in Initial Funding for Risk Reduction Projects
SI004 U.S. Department of Energy Office of Nuclear Energy Kairos Power Breaks Ground on Molten Salt Production Facility
SI005 U.S. Department of Energy Office of Nuclear Energy Kairos Power Installs Reactor Vessel for Third Test Unit
SI006 U.S. Department of Energy Office of Nuclear Energy NRC Approves Construction for Hermes Reactor
SI007 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SI008 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SI009 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SI010 Kairos Power Technology | Kairos Power
SI011 Kairos Power Our Approach | Kairos Power
SI012 World Nuclear News Kairos secures HALEU for Hermes' first fuel load
SI013 World Nuclear News Kairos, DOE enhance collaboration on advanced reactor design
SI014 American Nuclear Society Kairos Power breaks ground on first power-producing reactor in Oak Ridge
SI015 TechCrunch Google signed a deal to power data centers with nuclear micro-reactors from Kairos — but the 2030 timeline is very optimistic
SI016 TechCrunch Future Google supplier Kairos gets approval to build two small nuclear reactors | TechCrunch
SI017 TechCrunch Here are the nuclear fission startups backed by Big Tech | TechCrunch
SI018 Research UC Berkeley Nuclear Power Renaissance | Research UC Berkeley
SI019 Caplight Kairos Power | Valuation, Funding Rounds & Stock Price | Caplight
SI020 U.S. Energy Information Administration Capital Cost and Performance Characteristics for Utility-Scale Electric Power Generating Technologies
SI021 Lazard Lazard LCOE+ (June 2024)
SI022 Securities and Exchange Commission NuScale Power Corp. Form 10-K for fiscal 2025
SI023 X-energy X-energy Closes Oversubscribed $700 Million Series D Financing Round to Continue Expansion to Meet Global Energy Demand
SI024 Business Wire X-energy Closes Oversubscribed $700 Million Series D Financing Round to Continue Expansion to Meet Global Energy Demand
SI025 Nuclear Engineering International Kairos Power receives DOE funding to support development of Hermes reactor
SI026 U.S. Department of Energy Office of Nuclear Energy HALEU Technologies
SI027 U.S. Department of Energy Office of Nuclear Energy Advanced Reactor Demonstration Projects
SE001 Kairos Power Technology | Kairos Power
SE002 Kairos Power Our Approach | Kairos Power
SE003 Kairos Power Tennessee Location | Kairos Power
SE004 Kairos Power New Mexico Location | Kairos Power
SE005 Kairos Power California Location | Kairos Power
SE006 Kairos Power About | Kairos Power
SE007 Kairos Power Kairos Power Breaks Ground on Hermes 2 Demonstration Plant | Kairos Power
SE008 Kairos Power The Time is Now | Kairos Power
SE009 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SE010 Nuclear Regulatory Commission Hermes 2 – Kairos Application
SE011 U.S. Department of Energy Office of Nuclear Energy NRC Approves Construction for Hermes Reactor
SE012 U.S. Department of Energy Office of Nuclear Energy Kairos Power Installs Reactor Vessel for Third Test Unit
SE013 U.S. Department of Energy Office of Nuclear Energy Kairos Power Breaks Ground on Molten Salt Production Facility
SE014 U.S. Department of Energy Office of Nuclear Energy Advanced Reactor Demonstration Projects
SE015 U.S. Department of Energy Office of Nuclear Energy U.S. Department of Energy HALEU Allocation Process
SE016 U.S. Department of Energy U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers
SE017 U.S. Department of Energy Office of Nuclear Energy HALEU Technologies
SE018 World Nuclear News Kairos secures HALEU for Hermes' first fuel load
SE019 World Nuclear News Kairos, DOE enhance collaboration on advanced reactor design
SE020 American Nuclear Society Kairos Power breaks ground on first power-producing reactor in Oak Ridge
SE021 BWX Technologies Kairos Power and BWXT to Collaborate on Commercial TRISO Manufacturing
SE022 TechCrunch Future Google supplier Kairos gets approval to build two small nuclear reactors | TechCrunch
SE023 Research UC Berkeley Nuclear Power Renaissance | Research UC Berkeley
SE024 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SE025 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SE026 Oak Ridge National Laboratory ORNL, Kairos Power partner to advance deployment of next-gen nuclear energy | ORNL
SE027 Oak Ridge National Laboratory 3D printing reshapes construction for nuclear energy | ORNL
SE028 Oak Ridge National Laboratory Analysis and Design of High-Power TRISO Fuel Compact Irradiation in HFIR
SE029 Idaho National Laboratory Molten Salt Reactors
SE030 Kairos Power Careers | Kairos Power
SU001 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SU002 Google New nuclear clean energy agreement with Kairos Power
SU003 Utility Dive Google, Kairos Power ink 500-MW advanced nuclear reactor deal
SU004 POWER Magazine Google Bets Big on Nuclear: Inks Deal with Kairos Power for 500-MW SMR Fleet to Power Data Centers
SU005 ESG Today Google Signs First Nuclear Energy Deal to Address Growing AI Carbon Footprint
SU006 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SU007 PR Newswire Google, Kairos Power, TVA Collaborate to Meet America's Growing Energy Needs
SU008 World Nuclear News Google, Kairos Power, TVA announce collaboration
SU009 American Public Power Association Kairos Power, TVA and Google Unveil Advanced Nuclear Power PPA Tied to Data Centers
SU010 POWER Magazine TVA Inks First U.S. Utility PPA for Gen IV Nuclear Power in Landmark Three-Way Deal with Google, Kairos
SU011 Data Center Dynamics TVA signs 50MW PPA with SMR developer Kairos for Google data centers in Tennessee and Alabama
SU012 RTO Insider Kairos Power, TVA Announce Nuclear PPA
SU013 Nuclear Engineering International Kairos, TVA to power Google data centres
SU014 Duke Energy Responding to growing demand, Duke Energy, Amazon, Google, Microsoft and Nucor execute agreements to accelerate clean energy options
SU015 Enlit Duke Energy| Amazon| Google| Microsoft and Nucor to co-develop new clean tech tariffs
SU016 T&D World Duke Energy and Parties Execute Agreements to Accelerate Clean Energy Options
SU017 Latitude Media Duke Energy’s 'light-touch' large load tariff could be a problem.
SU018 Google Read Google’s 10th annual Environmental Report
SU019 Power Technology Google and Kairos Power partner TVA for advanced nuclear energy
SU020 Duke Energy Investor Relations Responding to growing demand, Duke Energy, Amazon, Google, Microsoft and Nucor execute agreements to accelerate clean energy options
SU021 Kairos Power Kairos Power and Google | Kairos Power
SU022 U.S. Department of Energy Office of Nuclear Energy Advantages and Challenges of Nuclear-Powered Data Centers
SU023 Google 6 unique ways we advanced energy solutions in 2024
SU024 U.S. Department of Energy DOE Identifies 16 Federal Sites Across the Country for Data Center and AI Infrastructure Development
SU025 Google Sustainability Our 2026 Environmental Report - Google Sustainability
SU026 Tennessee Valley Authority A New Nuclear Heyday
SR001 U.S. Government Publishing Office In the Matter of Kairos Power LLC; Hermes Test Reactor; Extension of Latest Date for Completion of Construction
SR002 U.S. Government Publishing Office Kairos Power LLC; Hermes Test Reactor; Environmental Assessment and Finding of No Significant Impact
SR003 World Nuclear News Regulator extends Hermes 1 reactor construction deadline
SR004 Nuclear Regulatory Commission Hermes – Kairos Application | Nuclear Regulatory Commission
SR005 Nuclear Regulatory Commission Hermes 2 – Kairos Application
SR006 Nuclear Regulatory Commission Construction Permit Application Review Documents for Hermes 2 – Kairos Power
SR007 Nuclear Regulatory Commission Final Safety Evaluation to Amend Hermes 1 CP to Extend Construction Completion
SR008 U.S. Government Accountability Office GAO-26-107969, Department of Energy: Action Needed to Approve Advanced Test Reactor Spent Fuel Plan
SR009 U.S. Department of Energy Office of Inspector General Audit: DOE-OIG-26-25
SR010 Kairos Power U.S. Department of Energy to Provide HALEU for Hermes Demonstration Reactor | Kairos Power
SR011 American Nuclear Society Kairos Power finalizes contract on HALEU for Hermes
SR012 Centrus Energy Centrus Signs Contract with Department of Energy for $900 Million Award; Intends to Transition HALEU Production Cascade to Commercial Operation
SR013 Clean Air Task Force Lessons learned from the recently cancelled NuScale-UAMPS project
SR014 Utah Associated Municipal Power Systems Utah Associated Municipal Power Systems (UAMPS) and NuScale Power Agree to Terminate the Carbon Free Power Project (CFPP)
SR015 Latitude Media Duke Energy’s 'light-touch' large load tariff could be a problem.
SR016 U.S. Department of Energy Office of Nuclear Energy Advantages and Challenges of Nuclear-Powered Data Centers
SR017 Duke Energy Responding to growing demand, Duke Energy, Amazon, Google, Microsoft and Nucor execute agreements to accelerate clean energy options
SR018 Tennessee Valley Authority A New Nuclear Heyday
SR019 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SR020 POWER Magazine TVA Inks First U.S. Utility PPA for Gen IV Nuclear Power in Landmark Three-Way Deal with Google, Kairos
SR021 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SR022 BWX Technologies Kairos Power and BWXT to Collaborate on Commercial TRISO Manufacturing
SR023 U.S. Department of Energy Office of Nuclear Energy U.S. Department of Energy HALEU Allocation Process
SR024 U.S. Department of Energy U.S. Department of Energy to Distribute First Amounts of HALEU to U.S. Advanced Reactor Developers
SR025 U.S. Government Publishing Office Risk-Informed, Technology-Inclusive Regulatory Framework for Advanced Reactors
SR026 Google New nuclear clean energy agreement with Kairos Power
SR027 Utility Dive Google, Kairos Power ink 500-MW advanced nuclear reactor deal
SR028 Data Center Dynamics TVA signs 50MW PPA with SMR developer Kairos for Google data centers in Tennessee and Alabama
SR029 Google Sustainability Our 2026 Environmental Report - Google Sustainability
SR030 Tennessee Valley Authority Clinch River Nuclear (CRN) Site
SV001 Kairos Power Google and Kairos Power Partner to Deploy 500 MW of Clean Electricity Generation | Kairos Power
SV002 Google New nuclear clean energy agreement with Kairos Power
SV003 Kairos Power Google, Kairos Power, TVA Collaborate to Meet America’s Growing Energy Needs | Kairos Power
SV004 POWER Magazine TVA Inks First U.S. Utility PPA for Gen IV Nuclear Power in Landmark Three-Way Deal with Google, Kairos
SV005 Google Sustainability Our 2026 Environmental Report - Google Sustainability
SV006 U.S. Department of Energy Office of Inspector General Audit: DOE-OIG-26-25
SV007 Utah Associated Municipal Power Systems Utah Associated Municipal Power Systems (UAMPS) and NuScale Power Agree to Terminate the Carbon Free Power Project (CFPP)
SV008 Clean Air Task Force Lessons learned from the recently cancelled NuScale-UAMPS project
SV009 U.S. Department of Energy Office of Nuclear Energy Advantages and Challenges of Nuclear-Powered Data Centers
SV010 Kairos Power U.S. Department of Energy to Provide HALEU for Hermes Demonstration Reactor | Kairos Power
SV011 Centrus Energy Centrus Signs Contract with Department of Energy for $900 Million Award; Intends to Transition HALEU Production Cascade to Commercial Operation
SV012 BWX Technologies Kairos Power and BWXT to Collaborate on Commercial TRISO Manufacturing
SV013 Tennessee Valley Authority A New Nuclear Heyday
SV014 TerraPower TerraPower Announces $650 Million Fundraise
SV015 TerraPower TerraPower Natrium | Advanced Nuclear Energy
SV016 X-energy X-energy Reports First Quarter 2026 Results | X-Energy News
SV017 X-energy News Releases — X-Energy, Inc.
SV018 Oklo Oklo Inc. - Investors - Financials
SV019 Nasdaq / Business Wire Oklo Publishes Full-Year 2025 Financial Results and Business Update
SV020 Oklo Oklo Publishes Full-Year 2025 Financial Results and Business Update
SV021 NuScale Power NuScale Power Reports Fourth Quarter and Full Year 2025 Results
SV022 NuScale Power SEC Filings | NuScale Power
SV023 Nuclear Regulatory Commission Final Safety Evaluation to Amend Hermes 1 CP to Extend Construction Completion
SV024 World Nuclear News Regulator extends Hermes 1 reactor construction deadline
SV025 Data Center Dynamics TVA signs 50MW PPA with SMR developer Kairos for Google data centers in Tennessee and Alabama
SV026 Utility Dive Google, Kairos Power ink 500-MW advanced nuclear reactor deal
SV027 Kairos Power Kairos Power and Google | Kairos Power
SV028 U.S. Department of Energy Office of Nuclear Energy U.S. Department of Energy HALEU Allocation Process
SV029 Duke Energy Responding to growing demand, Duke Energy, Amazon, Google, Microsoft and Nucor execute agreements to accelerate clean energy options
SV030 Latitude Media Duke Energy’s 'light-touch' large load tariff could be a problem.