Why It Matters
The federal government is spending nearly $1 billion a year to maintain aging nuclear facilities and it still can't say with confidence what it actually needs, or what it would cost to fix it. A new federal watchdog report finds that the Department of Energy's nuclear waste cleanup program is managing crumbling infrastructure with flawed data, inconsistent tracking methods, and a planning document that most of its own sites say doesn't reflect reality. Meanwhile, Congress is being kept in the dark about at least $120 million in potential savings sitting on the table.
The Government Accountability Office report, released May 5, finds systemic gaps in how the DOE's Office of Environmental Management tracks, prioritizes, and communicates its infrastructure needs, raising urgent questions about whether one of the federal government's most complex and hazardous cleanup missions is being managed responsibly.
A Mission Measured in Decades
The Office of Environmental Management exists to address one of the most consequential environmental liabilities in American history: the radioactive waste remediation left behind by decades of nuclear weapons production and research. The work includes treating radioactive waste, cleaning up groundwater contamination, and decommissioning facilities that handled some of the most hazardous materials ever produced.
To carry out that mission, EM operates approximately 4,300 facilities spread across 13 sites nationwide. The problem is that many of those facilities are between 50 and 70 years old, well beyond their designed operational life. They were built during the Cold War, and they are still being used to contain, treat, and manage nuclear waste cleanup activities that have no near-term end date.
The cost of keeping those facilities operational has surged. EM's fiscal year 2026 budget request included more than $950 million in nuclear facility maintenance costs, an 80 percent increase since fiscal year 2020. Site officials told GAO those costs are expected to climb further over the next five years. And as of June 2025, EM reported more than $1.5 billion in outstanding repair needs across its operating facilities.
The Data Problem
To manage aging nuclear infrastructure at this scale, EM relies on a system called the Facilities Information Management System, or FIMS, which is supposed to capture facility condition data from all 13 sites. That data is then validated through a scorecard process and used to inform maintenance decisions and budget planning.
But GAO found that the scorecards themselves contained inaccurate or unsupported data, and that some sites had not completed corrective action plans to address known problems. More troubling, different sites were using different methods to generate the same data elements - data that DOE's own internal order, DOE Order 430.1C, requires to be captured consistently. The result is that the numbers feeding into headquarters-level planning may not be comparable across sites, and may not be reliable at all.
That matters because EM headquarters uses a document called the Master Asset Plan to document its maintenance needs and make the case to Congress for funding. If the underlying data is inconsistent or inaccurate, the plan built on top of it is compromised from the start.
Sites Say the Plan Doesn't Reflect Their Needs
The Master Asset Plan has another problem: the people closest to the work say it doesn't capture what they actually need. Eight of EM's 13 sites told GAO the plan fails to reflect their maintenance needs, in part because sites have more granular, site-specific data than what headquarters incorporates into the plan.
That disconnect means the document EM uses to communicate infrastructure needs to Congress may be systematically understating the scale of the problem. It also means that project prioritization decisions made at the site level (where staff have the most detailed knowledge of facility conditions) are not making their way into the plan in a meaningful way.
GAO is recommending that EM better incorporate site-level information, including site project prioritization decisions, into the Master Asset Plan to more accurately reflect what the program actually needs.
$120 Million in Hidden Savings
Perhaps the most striking finding in the report involves money that EM's own analysis identified but never told Congress about.
The Master Asset Plan contains 19 projects identified through a cost savings model. According to the model, completing those projects using surplus funds could generate approximately $120 million in savings for the program. The projects are already identified. The analysis has already been done. But EM has not communicated the potential cost and risk reductions associated with those specific projects to Congress.
GAO is recommending that EM do exactly that: communicate to Congress what completing those 19 unfunded maintenance projects could achieve. EM partially concurred with that recommendation, as it did with the recommendation to better incorporate site-level data into the Master Asset Plan. EM fully concurred with the two recommendations focused on fixing data accuracy and comparability problems in FIMS.
The Legislative Backdrop
The report was mandated by a provision in Senate Report 118-188, which accompanied a bill for the fiscal year 2025 National Defense Authorization Act. The NDAA is the annual defense policy vehicle that Congress passes each year, and the provision directed GAO to evaluate the status of EM's infrastructure and how the agency completes and prioritizes maintenance.
The mandate reflects congressional unease about the trajectory of EM's maintenance costs and whether the program has the management infrastructure to handle a mission that will stretch well into the future. With maintenance spending up 80 percent in five years and repair needs exceeding $1.5 billion, the concern is not abstract. All four of GAO's recommendations remain open.
Access the Legis1 platform for comprehensive political news, data, and insights.