Powering Colorado's Next Economy

The Short Answer

Colorado is losing its competitive edge at the exact moment the national economy is being reorganized around one scarce resource: reliable, clean electricity. The states that can deliver abundant power will attract the next generation of employers — in tech, manufacturing, and advanced industry. The states that can't will watch those employers, and their tax base and jobs, grow somewhere else.

We can lead this. We have coal communities with the workforce, the land, and the grid connections already in place. We have a statewide clean-energy mandate that nuclear can help us actually meet. And we have a national power shortage that turns our biggest liability — an aging fossil fleet that is closing anyway — into our biggest opportunity.

The proposal is straightforward: clear the path for next-generation nuclear power, and pair it with data-center demand — done intentionally, with ratepayer protections built in from day one. This gives businesses a reason to stay and grow here, gives displaced coal workers a future, and positions Colorado as a national leader in energy, data, and AI.

Part I: America's Power Crisis

After a decade in which U.S. electricity demand grew just 0.2% per year, we have entered what every major energy analyst now describes as the largest demand surge since post-WWII industrialization. Bank of America projects 2.5% annual compound growth through 2035 — a 12x acceleration driven by four simultaneous forces: building electrification (heat pumps replacing gas furnaces), data centers and AI, industrial reshoring, and electric vehicles.

The numbers are staggering. Goldman Sachs projects U.S. data center power demand alone will double from 31 GW to 66 GW by 2027 and could reach 122 GW by 2030. Electric vehicles will add 651-721 TWh of demand by 2035 — an 18% increase over current total generation. Industrial reshoring is driving 3% annual growth in industrial power consumption. And building electrification is the single largest driver at roughly 1 percentage point of annual demand growth.

The Supply Gap

The grid is not ready. 104 GW of firm, dispatchable generation capacity (coal and gas plants that run on demand, regardless of weather) is set for retirement by 2030. Only 22 GW of new firm capacity is planned to replace it — an 82 GW net deficit of reliable generation. Schneider Electric projects the total shortfall will reach 175 GW by 2033.

New capacity is being added — 53 GW in 2025, 86 GW planned for 2026 — but most of it is intermittent solar (23% average capacity factor) and wind (34% capacity factor). One hundred megawatts of solar delivers the equivalent of only 23 megawatts of firm power. The math doesn't work.

Meanwhile, the infrastructure to deliver power is failing. Power transformer lead times have reached 128 weeks (2.5 years), with some units taking 4 years. Domestic manufacturers meet only 20% of demand. Roughly 40 million distribution transformers are past their expected service life. The interconnection queue — the line to connect new generation to the grid — stretches to 2,061 GW, with an average wait of 4-5 years. For every 1 GW that actually reached commercial operation in 2025, approximately 14 GW withdrew from the queue.

How Far Ahead Is China?

In 2025, China added approximately 540 GW of new power capacity. The United States added 53 GW. China's buildout was more than ten times America's — and China's was not a record year. China's total installed capacity now stands at 3,890 GW, compared to America's 1,281 GW — a 3:1 ratio.

The nuclear gap is especially acute. China has 33 reactors under construction and the capacity to build 50 simultaneously. Its nuclear target is 200 GW by 2035 and 400-500 GW by 2050. The United States has zero reactors under construction. The only U.S. reactors completed in 30 years are Vogtle Units 3 and 4 (2023-2024).

China will have a commercial, operating small modular reactor (the Linglong One) before the United States has broken ground on one at commercial scale. China's grid investment plan for 2026-2030 is $580 billion. Al Jazeera calls China's abundant cheap energy its "secret weapon" in the AI race. Every year of American delay widens the gap.

Part II: Colorado Communities at Risk

Colorado's clean energy mandates — HB19-1314, SB19-236, SB23-016, and the state's greenhouse gas reduction targets — are driving the retirement of every coal plant in the state. At least 11 communities across 9 counties are directly impacted, designated as Tier One or Tier Two Coal Transition Communities by the Office of Just Transition. The human cost is real and concentrated.

Craig / Moffat County: The Epicenter

Craig Station's three units (1,285 MW total, owned by Tri-State Generation) are retiring between 2025 and 2028. Combined with the closure of the Colowyo Mine (133 layoffs, January 2026) and the impending closure of Trapper Mine (185 employees), Moffat County faces approximately 437 direct job losses in a place where coal underpins 43% of county property taxes and contributes $321 million to regional GDP. One analysis estimated this is the equivalent of Denver losing 141,000 jobs.

The settlement with Tri-State provides $22 million in direct payments and up to $48 million in minimum tax revenue through 2038. A $8.5 million federal commerce grant targets tourism and 200+ new jobs. But these are holding measures, not replacements for an industry that defined the community for generations.

Pueblo County: $845 Million in Lost Revenue

Comanche Generating Station (1,410 MW total, Xcel Energy) is Colorado's largest coal plant. Unit 1 retired in 2022. Unit 3's closure was accelerated from 2070 to 2031 — a loss of nearly four decades of expected operation. The total economic impact is 411 jobs (77 direct, 161 indirect, 173 induced) and $196 million in annual economic activity. Comanche generates $31 million per year in property taxes — 10.5% of everything Pueblo County collects. Closing Unit 3 in 2031 instead of 2070 results in $845 million in cumulative lost tax revenue.

The Pueblo Innovative Energy Solutions Advisory Committee (PIESAC) studied replacement options and declared advanced nuclear "the clear winner" for replacing Comanche.

Hayden / Routt County

Hayden Station (233 MW, Xcel) is closing 2027-2028 — accelerated 8 years from its original 2036 retirement. Combined with Twentymile Mine (233 employees, Peabody Energy), the closure eliminates approximately 294 direct jobs. The Hayden School District loses 38% of its tax revenue. West Routt Fire Protection District loses 44%. Across Routt, Moffat, and Rio Blanco counties, the coal industry supports nearly 2,900 jobs paying more than $228 million in labor income — more than one-fifth of the three-county GDP.

Other Impacted Communities

The Human Cost Is Not Abstract. It Is Measured in Lives.

When a community loses its economic anchor without an equivalent replacement, the consequences follow a predictable and devastating pattern. This is not speculation — it is documented across every coal community that has been through it, and the data is unequivocal.

Deaths of despair. A 2026 Bangor University study found former coal mining communities have 53% higher rates of deaths from drug overdoses, suicide, and alcoholic liver disease compared to non-mining areas. The association was strongest for alcohol-specific and drug-poisoning mortality. Counties that became more economically distressed experienced a five-fold increase in age-adjusted overdose deaths. Rural Appalachian opioid death rates reached 62 per 100,000 in 2021 — nearly double the urban rate. If life expectancy increases had continued at prior rates, 600,000 more Americans would be alive today.

Domestic violence. Oxford's Review of Economic Studies found domestic violence increases 32% after men lose their jobs and 56% after women lose theirs, sustained for at least four years after layoff. A county's foreclosure rate is the strongest predictor of all violence subtypes; unemployment rate is the second strongest. Financial stress doesn't stay at the office — it comes home.

Mental health collapse. Rates of diagnosed depression reach 50% among those unemployed 12 months or more. Rural suicide rates are 49% higher than urban rates. During economic downturns, psychological distress among job losers is up to 8.4 times greater than those who continue working. The coal fields of Appalachia have life expectancy 15 years shorter than wealthier U.S. communities.

Substance abuse. 33% of unemployed people binge drink; nearly 15% have an alcohol use disorder — almost double the rate of employed persons. Opioid use and coal decline feed on each other in Appalachian communities. Rural areas consistently have the highest per-capita drug overdose death rates in the nation.

The death spiral. The Philadelphia Federal Reserve found a job loss multiplier of 2.0 for coal closures — every coal job lost eliminates an additional job in the local economy. West Virginia University found the multiplier is as high as 3.7. In affected Appalachian counties, closures produced unemployment increases of 0.9 percentage points, payroll job reductions of 4%, wage decreases of 8%, and GDP declines of 15%. McDowell County, West Virginia — once the world's largest coal producer with 100,000 residents — has lost 60% of its population. Gary, Indiana lost 61% of its population after steel collapsed and has 6,500 abandoned buildings.

Colorado already has a case study. When Nucla Station closed in 2019 (just 170 jobs), Naturita's main street hit 40% vacancy. School and fire districts faced 56% revenue cuts. Two-thirds of Nucla's tax base was eliminated. Residents say: "We've already transitioned, but we haven't recovered." Now apply that to Craig, where 437 jobs underpin 43% of county property taxes and workers earn $100,000/year in a county where the average wage is $40,000. Those aren't just jobs. They are the mortgages that keep families in their homes, the spending that keeps Main Street open, and the tax base that funds the schools and fire department. When they go, everything goes.

Part III: Nuclear Siting — Five Communities That Could Host a Plant

Colorado's coal communities already have what nuclear developers need most: transmission infrastructure, land, water access, and a skilled energy workforce. The 2025 passage of HB25-1040 — which added nuclear to Colorado's statutory definition of "clean energy" with overwhelming bipartisan support (43-18 in the House, 29-5 in the Senate) — was step one. HB26-1337 requires Xcel Energy to solicit community interest and identify potential sites by August 2027.

1. Pueblo (Comanche Station) — Ranked #1

The strongest candidate by every measure. The 1,410 MW Comanche site is owned outright by Xcel Energy, has existing high-voltage transmission, secured Arkansas River water rights, and a city of 114,000 providing workforce depth. PIESAC's recommendation of nuclear gives Pueblo the most advanced community engagement process of any site. The committee estimated a nuclear plant would generate $95 million annually in state and local taxes. State bill sponsors have explicitly named Pueblo as a leading candidate.

2. Craig (Craig Station) — Ranked #2

Craig can tell a nuclear developer what very few communities in America can: we have the site, the grid connection, and the trained people — ready now. The 1,285 MW transmission infrastructure is already built. The Yampa River provides ample cooling water, and Moffat County secured water rights in its Tri-State settlement. A 2024 NCEI survey found 61% of NW Colorado residents support nuclear. The Northwest Colorado Energy Initiative has been actively advocating for nuclear as the long-term anchor. The primary limitation is Tri-State's cooperative ownership structure, which adds complexity compared to Xcel's investor-owned model.

The critical insight for Craig: rather than closing the coal plant and leaving the community in limbo for a decade, delay the retirement of Craig Station until the nuclear replacement is operational. This is already happening reactively — DOE forced Unit 1 to stay open into 2026, and Xcel has proposed running all remaining coal through 2030 due to a 608 MW supply shortfall. Making this intentional policy rather than crisis management changes the entire equation: workers transfer directly from one plant to the next, the tax base stays intact through construction, and the grid keeps 1,285 MW of firm capacity online until clean replacement power is ready.

3. Colorado Springs (Ray Nixon) — Ranked #3

Colorado Springs Utilities is actively exploring nuclear through a DOE-funded year-long feasibility study with Idaho National Laboratory and Oak Ridge National Laboratory. The utility board unanimously endorsed including nuclear in its updated Sustainable Energy Plan. With a population of 490,000 (metro 750,000), the city provides the largest workforce and ratepayer base of any candidate site. Military installations (Fort Carson, Peterson Space Force Base, NORAD) provide additional demand and security synergies.

4. Hayden (Hayden Station) — Ranked #4

Adequate physical infrastructure and shared workforce with Craig. However, Routt County commissioners have expressed skepticism about nuclear, and the community's identity is shifting toward recreation and tourism. The site works best as a potential secondary location if demand warrants multiple plants.

5. Brush / Fort Morgan (Pawnee Station) — Ranked #5

The weakest candidate for near-term nuclear. Pawnee is being actively converted to natural gas, eliminating the brownfield opportunity. The small agricultural communities lack demonstrated interest in nuclear, and water constraints in the South Platte basin present additional challenges. Growing transmission infrastructure makes the eastern Colorado corridor interesting for longer-term consideration.

Part IV: The Data Center Opportunity

In the first four months of 2026 alone, 75+ data center projects worth approximately $130 billion were successfully blocked across the United States. That figure matches the total value of projects blocked in all of 2025. The industry faces simultaneous crises: power infrastructure bottlenecks (a data center deploys in 18 months, but power infrastructure takes 5 years), community opposition (70% of Americans now oppose a nearby data center, up from 50% in late 2025), and legislative backlash (533+ local moratoriums across 42 states, 300+ statehouse bills in the first 6 weeks of 2026 alone).

The Demand Is Real and Growing

Despite the opposition, underlying demand continues to accelerate. Global data center spending hit $582 billion in 2026 (Gartner). Big Tech's combined capital expenditure reached $725 billion. Northern Virginia's vacancy rate is 0.3% — essentially sold out. Four out of five facilities under construction are already pre-leased. Goldman Sachs projects US data center power demand will reach 122 GW by 2030. S&P Global estimates 183 GW.

Colorado Is Falling Behind

Five Colorado jurisdictions — Denver, Jefferson County, Boulder County, Larimer County, and Longmont — enacted data center moratoriums in 2026. Colorado offers zero data center tax incentives while 37 states do, despite hosting 8 data center developer headquarters. Xcel Energy has 5.8 GW of pending data center power applications, and data centers are projected to drive 62% of Colorado's energy growth.

Meanwhile, Wyoming approved a 2.7 GW data center campus. Meta chose Cheyenne for its mega data center "Project Cosmo." Utah has 948 MW operating and 30,652 MW planned. Every month of delay pushes investment to competing states. Infrastructure decisions made in 2026-2027 lock in economic trajectories for 20+ years.

The Nuclear + Data Center Model Is Proven

Every major hyperscaler has now signed nuclear power purchase agreements. Total committed nuclear capacity for data centers exceeds 9.8 GW across 13 announced projects. The Three Mile Island restart (835 MW, Constellation + Microsoft) is projected to contribute $16 billion to Pennsylvania's GDP and generate $3 billion+ in taxes. Iowa's Duane Arnold restart (615 MW, NextEra + Google) carries a $9 billion projected economic impact.

The logic is straightforward: nuclear provides 24/7 carbon-free baseload power at a 92-93% capacity factor — exactly what data centers need. When a data center's demand finances dedicated new nuclear generation, it adds to the grid rather than competing with households for existing supply. The data center gets reliable power; the community gets jobs, taxes, and clean energy. This is not theory. It is happening right now in Pennsylvania, Iowa, Illinois, and Washington state.

Part V: What This Means for Colorado — The Numbers

Using real-world data from Pennsylvania, Iowa, the Nuclear Energy Institute, and multiple data center economic impact studies, we can project what a nuclear + data center buildout would mean for Colorado:

Colorado currently has approximately 30,000 fossil fuel jobs (coal, oil, natural gas combined), declining at 4.5% per year. A 5 GW buildout would create 24,650-45,800 permanent jobs — more than replacing every fossil fuel job at risk, in a growth industry paying 30-50% higher wages, with plant lifespans of 60-80 years.

Part VI: The Honest Tradeoffs

Nuclear is expensive and slow to build. Nuclear megaprojects have historically run about 117% over budget. NuScale's SMR project was cancelled in 2023 after costs ballooned toward $9+ billion. The path from planning to power is typically 10-15 years. How to manage it: Prioritize proven reactor designs over first-of-a-kind experiments, use federal cost-share aggressively, and structure state incentives so ratepayers are protected if a project slips.

The timeline gap between coal closure and nuclear startup. Craig's units are scheduled to close by 2028; even a fast-tracked nuclear plant won't produce power until the mid-2030s. How to manage it: The strongest solution is to delay coal plant closure at sites selected for nuclear replacement until the new plant is operational. This is already happening reactively — DOE forced Craig Unit 1 to stay open, Xcel proposed running all coal through 2030 due to supply shortfalls, and the legislature extended Nixon to 2032. Make it intentional policy: tie coal retirement dates to nuclear commissioning milestones. Workers transfer directly, the tax base stays intact, and the grid keeps its capacity. Where timing doesn't align, pair expanded Just Transition support with nuclear as the permanent anchor.

Data centers can raise electricity bills. PJM's capacity auction jumped nearly ninefold in one cycle. National rates are up ~30% since 2020. How to manage it: Require data centers to bring, build, or buy their own power through a separate rate class. This is how you attract data centers without the backlash that led 5 Colorado jurisdictions to enact moratoriums.

Water and waste are real, not zero. Closed-loop cooling reduces but doesn't eliminate water use. Spent fuel requires long-term storage. How to manage it: Set clear siting and water standards upfront. Treat waste as a solvable engineering question — the U.S. Navy has managed reactor waste safely for 70 years.

Demand could not fully materialize. If AI growth slows, generation built on speculation can leave ratepayers holding stranded costs. How to manage it: Tie approvals to firm, long-term customer contracts before construction begins. The existing model — 20-25 year PPAs — already provides this protection.

Part VII: The Ask

1. Streamline permitting for clean baseload generation. Air-quality permits in Colorado averaged over two years to issue last fiscal year. Set 12-month hard deadlines for state permitting decisions on qualifying projects.

2. Create targeted nuclear incentives with ratepayer protection. Pair state incentives with federal cost-share (DOE loan guarantees, IRA tax credits, ADVANCE Act streamlining). Structure every incentive so taxpayers and ratepayers are protected if a project slips.

3. Adopt data center ratepayer protections now. Create a separate rate class with "bring/build/buy your own power" requirements so growth doesn't land on household bills. Oregon, Georgia, and Maryland have versions. Codify ratepayer-protection pledges as policy.

4. Tie coal retirement to nuclear replacement — close the gap. For communities like Craig, the best transition plan is no gap at all. Delay coal plant closure at sites selected for nuclear until the replacement facility is operational, with an intentional overlap so workers transfer directly. This is already happening ad-hoc through DOE emergency orders and Xcel extension proposals — make it deliberate policy. Tie retirement milestones to nuclear commissioning dates, keep the tax base intact during construction, and give workers a direct bridge. Where coal-to-nuclear timing doesn't align, expand Just Transition support as the interim bridge, with community input driving site selection through Xcel's 2027 solicitation.

Sources

Goldman Sachs, "US Data Center Power Demand Projected to Double by 2027" (2026) • Bank of America Institute, "US electricity demand to grow 2.5% annually through 2035" (2025) • NERC, "2025 Long-Term Reliability Assessment" • NERC, "2026 Summer Reliability Assessment" • DOE, Grid Reliability Report (2026) • Schneider Electric, "175 GW Capacity Gap" (2026) • CBRE, "Global Data Center Trends 2026" • Tom's Hardware / Data Center Watch, "$130B in blocked projects" (2026) • Construction Dive, "Data center cancellations quadrupled" (2026) • Constellation Energy, "Crane Clean Energy Center" (2024-2026) • SMR Intel, "Nuclear Data Center Deals Tracker" (2026) • Oxford Economics, "Economic Contribution of US Nuclear Power Industry" • Nuclear Energy Institute, "Nuclear in the Neighborhood" (2021) • Colorado Sun, Comanche Station coverage (2022-2026) • Colorado Sun, Craig Station coverage (2025-2026) • CPR News, nuclear legislation and CS Utilities coverage (2025-2026) • PIESAC Committee Report, Xcel Energy (2024) • Colorado General Assembly, HB25-1040 and HB26-1337 • World Nuclear Association, country profiles (China, USA) • EIA, capacity additions and retirements data (2025-2026) • Ember, Global Electricity Review 2026 • IEA, World Energy Investment 2025 (China) • TerraPower, Kemmerer construction announcement (April 2026) • ADVANCE Act of 2024, NRC implementation guidance • Colorado Office of Just Transition, Action Plan and Community Grants • AGC, "Economic Impact of Data Center Development" • Brookings Institution, "Data Center Employment Effects" (2026)