The Big Short(fall): Colorado’s Upcoming Power Plant Closures and Planned Replacements

By Ethan Cornell*

Introduction

The data summarizing planned generation closures in Colorado signals a rapid and profound infrastructural transformation. The schedule details the planned retirement of 10 major coal-fired units between 2025 and 2031, collectively representing a loss of nearly 4,200 megawatts (MW) of nameplate capacity.

The primary policy question this raises is whether the state’s proposed replacement sources can be deployed with sufficient speed and scale to ensure grid reliability. This report argues they cannot. An analysis of formal resource adequacy forecasts reveals that this transition, as planned, will create a severe and widening gap between reliable energy supply and the capacity required to power the grid.

This projected shortfall stems from a critical flaw in the replacement strategy: swapping high-accredited-capacity coal plants for intermittent, low-accredited-capacity renewables. While the “nameplate” capacity of new solar and wind projects may seem to replace what is lost, their “accredited” capacity—the reliable power they can be counted on to produce—is far lower. This report will quantify this impending reliability crisis, showing how it leads to a forecasted failure to meet the state’s own reserve margin targets, with the gap growing to critical levels by 2030.

This report’s analysis is based on data from the U.S. Energy Information Administration (EIA) Form 860M, with a data cutoff of September 2025. It must be noted that the data included in this report does not include smaller, distributed generation projects. The scope of this paper is focused on large, centralized generation assets being retired by major utilities, as their replacement is most critical to system-wide reliability.

Energy Subtraction

Table 1 outlines the current retirement schedule of Colorado’s last remaining coal units in accordance with state decarbonization mandates.

Table 1: Total Planned Generation Closures[i]

Table 2 presents a list of the planned generating units that will soon be placed in service on the Xcel system, a representation of the utility’s strategy to help replace coal capacity retiring over the next five years. This list serves as a counterpart to the closure table, detailing the key components of the resource portfolio intended to fill the 2,269 MW capacity gap Xcel is creating.

The plan hinges on large-scale solar and wind projects, specifically Rocky Mountain Solar (325 MW), the Singing Grass Wind Project (603 MW), and the Arroyo Solar facility (335 MW), which are all scheduled to begin operation in 2026.

However, a careful review shows an immediate capacity deficit. While the closures total 2,269 MW, the combined new renewable capacity is 1,463 MW.

The list’s inclusion of a 505 MW Natural Gas unit at the Pawnee site highlights the reality that a utility must maintain reliable, on-demand capacity to balance the intermittency of wind and solar during this transition.

The reliability challenges posed by this timeline are not hypothetical; they are already forcing changes. The planned 2025 closure of Comanche 2 (396 MW), for example, is reportedly being delayed, with the unit now expected to remain operational through 2026. This delay underscores the central argument of this report: that existing, reliable capacity is being retired faster than it can be replaced, and utilities are already finding that the planned retirement dates are not feasible from a reliability standpoint.

Table 2: Planned Opening Generation Plants under Xcel (To replace 2,269 MW) (2025-2031)

Table 3: Planned Opening Generation Plants to Replace Craig, Rawhide, and Ray Nixon (To replace 1,392.8 MW) (2025-2031)

Xcel’s Replacement Plan

The capacity gap identified in Xcel’s replacement plan is not just shown in a loss of nameplate capacity over the years, it is confirmed by formal resource adequacy forecasts. Data from Xcel’s latest Resource Adequacy Report, shown in the table below, provides this stark projection in grid reliability. While the target reserve margin is 6.4% (the safety buffer of extra capacity needed), the forecast shows this margin becoming negative as early as 2026. This deficit is projected to worsen significantly, plummeting to a -10.8% margin in 2029 and -17.6% margin by 2030. This data directly quantifies the critical challenge and primary question around whether grid reliability can be maintained during this transition.

Table 4: Projected Resource Adequacy vs Target[ii]

Accredited capacity is reliable capacity a power plant can be counted on to produce, not its maximum “nameplate capacity”. As Xcel’s replacement plan shows, this value is calculated using an “ELCC” (effective load carrying capacity) percentage. This percentage heavily discounts intermittent sources like wind and solar, reflecting the reality that they are not available 24/7.

The swap of high-accredited-capacity sources for low-accredited-capacity sources is the direct cause of the reliability shortfall shown in the table below. Table 5 compares the total “Accredited Capacity” (the projected reliable supply) against the total capacity required to power the grid and maintain its safety buffer (Load + Reserve Margin).

Table 5: Projected Capacity Supply vs Total Required Capacity (2026-2030)[iv]

As shown, the reliable “accredited capacity” is not projected to keep pace with the system’s total requirements. This shortfall grows to a staggering 2,187 MW by 2030, which is the physical megawatt gap that produced the -17.6% margin.

It is important to note that Xcel is in discussions to expedite additional resource acquisitions (primarily renewables) to help further offset its looming capacity shortfall stemming from the closure of its coal plants. Any new projects, however, even if approved, must also grapple with transmission constraints and be physically connected to the grid in time to offset the MW shortfall.[v]

Conclusion

The data examined in this report, from the ambitious 4,200 MW coal retirement schedule to the state’s own required resource adequacy models, points to a single conclusion: Colorado’s current energy transition strategy, as planned, does not ensure grid reliability.

The policy’s central flaw is its failure to account for the fundamental difference between “nameplate capacity” (the maximum output of a plant) and “accredited capacity” (the amount of power a plant can be counted on to produce). The plan to swap thousands of megawatts of high-accredited-capacity, on-demand coal power for intermittent, low-accredited renewables is not a like-for-like replacement.

As Xcel’s Replacement Plan forecast data demonstrates, this swap is the direct cause of a projected, quantifiable reliability gap.

This is not a theoretical risk. A forecasted reserve margin plummeting from a required +6.4% safety buffer to a critical -17.6% deficit by 2030 is a clear mathematical projection of a system-wide failure. This gap, which equates to a physical shortfall of 2,100 MW, moves the state to a state of energy scarcity, where rolling blackouts are no longer a distant possibility.

As Colorado proceeds with this transformation, policymakers must pivot from a focus on renewable MW goals to a non-negotiable standard of reliable accredited capacity. The primary question of whether the grid can be maintained is, by the data’s own admission, being answered in the negative. A reevaluation of the plant closure timelines, replacement technologies, and reliability standards is required to avoid a self-inflicted energy crisis.

[i] https://www.eia.gov/electricity/data/eia860m/.

[ii]https://xcelnew.my.salesforce.com/sfc/p/#1U0000011ttV/a/R3000006QOWH/8BVhp_eRQJ3fOfVvEpzeu2spaM7t1GDhZnRlUKLy7N8

[iii] This column represents the total gap between the required safety target and forecasted reality using the following formula: [Target Planning Reserve Margin %] – [Forecasted Planning Reserve Margin %] = [Projected Shortfall Percentage]

[iv]https://xcelnew.my.salesforce.com/sfc/p/#1U0000011ttV/a/R3000006QOWH/8BVhp_eRQJ3fOfVvEpzeu2spaM7t1GDhZnRlUKLy7N8

[v] https://i2i.org/wp-content/uploads/Comanche-Unit-2-Variance-Petition.pdf

*Ethan Cornell is pursuing a master’s degree in natural resources and energy policy at Colorado School of Mines and is a recent graduate of the Future Leaders Program at Independence Institute.