The buried shells the Pentagon still cannot price

Somewhere under a subdivision outside a mid-century Army post, or a cattle pasture that used to be a bombing range, or a stretch of desert now zoned for solar, there is a piece of metal in the ground. A metal detector cannot tell you what it is. It might be a rusted fence staple. It might be a 75mm shell that failed to go off in 1943 and is still capable of killing whoever digs it up. For most of the last century the only way to find out was to excavate it, and because harmless scrap outnumbers live ordnance by a wide margin on a typical old range, the country spent decades paying to dig up fence staples. The office charged with making that problem smaller is the Unexploded Ordnance Center of Excellence, or UXOCOE, a small Department of Defense coordination function run by the Army. It does not dig up anything. It exists to make the digging cheaper, and the size of the bill it is trying to shrink runs, by the government's own 2003 estimate, from $16 billion to $165 billion.

The scrap that might be a shell

Unexploded ordnance is the residue of a country that has been firing weapons on its own soil since before it had an air force. Every artillery range, bombing range, grenade court, and rocket impact area leaves behind a fraction of rounds that did not detonate on impact. Those rounds do not disappear. They sink, they corrode, they migrate with frost heave, and some stay live for a very long time. When a range closes and the land is handed back for farming, housing, or recreation, the ordnance stays behind unless someone pays to remove it. That someone is almost always the federal government, because it did the firing.

The technical heart of the problem is discrimination, the plain question of whether a buried metal signature is a threat or trash. On a former range, the overwhelming majority of what a survey detects is harmless clutter, shrapnel, spent brass, horseshoes, wire. If you cannot tell clutter from a live shell without digging, you dig everything, and the cost of clearing an acre balloons. That single unsolved problem, telling a buried shell from junk without excavating it, drove enormous waste for decades and is exactly the problem UXOCOE was created to help solve. It is a research and coordination shop, not a cleanup crew. Its stated mission is to improve the effectiveness and economy of unexploded ordnance clearance research across the whole department, so that the services stop each solving the same problem in isolation and start fielding tools that reduce the number of holes anyone has to dig.

The office that coordinates and the offices that dig

The structure here is easy to misread, so it is worth being precise. UXOCOE is a DoD Executive Agent function, and the designated Executive Agent is the Secretary of the Army, with the day-to-day representatives sitting in the office of the Assistant Secretary of the Army for Acquisition, Logistics, and Technology. The designation is set by DoD Directive 5101.13E, dated March 2, 2006, and certified current as of November 1, 2011. The directive names the Army as Executive Agent for the center and defines its job as centralizing the management of unexploded ordnance research, development, test, and evaluation information across the department. The DoD Executive Agent registry carries the same designation and, as of its March 2024 update, still lists the office with active points of contact. The registry, not a press release, is the authoritative record that this Executive Agent role exists and belongs to the Army.

The hands-on work of the office runs through the Joint UXO Coordination Office, JUXOCO, at Fort Belvoir, Virginia. JUXOCO is the coordination hub: it tracks who is researching what, works to cut duplication across the Army, Navy, Air Force, and Marine Corps, and pushes promising detection and clearance technology toward the field. What it does not do is own the cleanups. The actual dirt gets moved under separate programs, chiefly the Formerly Used Defense Sites program run by the Army Corps of Engineers for land the government no longer holds, and the services' own range-sustainment budgets for land still in use. Keeping that line straight matters, because the office is tiny and the liability it serves is vast, and the two get conflated constantly.

A price tag between $16 billion and $165 billion

Here is the number that defines the stakes, and it deserves to be stated carefully. As of April 2003, DOD identified 10,444 operational ranges covering 24.6 million acres, and estimated it would cost between $16 billion and $165 billion to clean up the unexploded ordnance, discarded military munitions, and munitions constituents on them. That figure comes from the Government Accountability Office's 2004 report on operational ranges, GAO-04-601. It is a projected liability, not money spent, not a contract ceiling, and not obligated funds. It is the government's own guess at what the eventual bill might be, and the fact that the low and high ends differ by more than a factor of ten is itself the finding. GAO judged the estimate questionable precisely because a range that wide cannot guide a budget.

Set that against a second, better-bounded number for land the government has already given up. For Formerly Used Defense Sites, DOD estimated in 2021 that it would cost about $11.9 billion to clean up roughly 1,700 munitions and hazardous sites. Unlike the operational-range figure, the FUDS program has real spending to point to: DOD obligated over $1 billion for FUDS from fiscal year 2016 through fiscal year 2020, money actually put to work. These are different buckets and should never be added together or swapped for one another. The $16 billion to $165 billion is a projected estimate for active ranges. The $11.9 billion is a projected cost-to-complete for former sites. The over $1 billion is spent. Three numbers, three meanings, one underlying problem.

Where GAO keeps finding holes

The Government Accountability Office has returned to this enterprise for thirty years and has rarely come away satisfied. The 2004 operational-range report did not just call the top-line estimate too wide; it explained why. The services built their numbers on unvalidated assumptions and inconsistent inventory methods, and the per-acre cleanup costs they fed in varied roughly tenfold between them, about $755 an acre for the Air Force against about $7,577 an acre for the Army. Congress was handed a range that spanned an order of magnitude with no transparent basis for either end. When the same acre can be priced at either figure depending on which service is holding the pen, the estimate is not a plan, it is a placeholder.

The pattern held for former sites. GAO's 2022 review of environmental liabilities at Formerly Used Defense Sites, GAO-22-104744, found that DOD had set no completion goal for FUDS munitions cleanup at all, scored risk inconsistently between its hazardous-waste sites and its munitions sites, and had finished only about 45 percent of the munitions sites, with roughly 1,200 more still under investigation and their eventual cost unknown. A program with no finish line and no consistent risk yardstick cannot tell Congress when it will be done or what the last mile will cost.

And the fragmentation UXOCOE was built to cure was documented before the office existed. GAO's 1995 report, NSIAD-95-197, found unexploded ordnance detection and clearance research spread across the department with duplication and no central visibility, and recommended a multiagency clearinghouse and a DoD executive agent to coordinate it. That recommendation is the direct ancestor of the center. The honest reading is that the office was a reasonable answer to a real diagnosis, and that the diagnosis, fragmented research chasing an unsolved discrimination problem, persisted for years after the office opened its doors.

What better sensors actually bought

The other side of this ledger is genuine, and it is measurable. The reason to coordinate research rather than let each service reinvent it is that a breakthrough in discrimination pays off everywhere at once, and a breakthrough is roughly what happened. Advanced geophysical classification tools developed and demonstrated through DoD's environmental research and demonstration programs, SERDP and ESTCP, can now distinguish hazardous unexploded ordnance from harmless metal scrap at live demonstration sites 70 to 90 percent of the time. Because most of what a survey detects is clutter, being able to leave high-confidence clutter in the ground instead of excavating it cuts combined geophysical survey and excavation costs by roughly 30 to 40 percent.

That is not a marketing figure, it is the whole economic case for the office stated in one sentence. The old regime dug everything because it could not trust the sensor to sort threat from trash. The new tools let a project characterize a signal well enough to skip the harmless digs, and on a range where clutter outnumbers ordnance many times over, skipping the harmless digs is where the money is. The stakes are also human and do not reduce to dollars. Live ordnance on former ranges maims and kills civilians who stumble onto it, so a technology that returns contaminated land to safe use faster and at lower cost is buying safety as well as savings. Effectiveness and economy are the two words in the office's mission, and on the technology it helped shepherd, both moved in the right direction.

Reading the ledger

Two true things sit side by side here and neither cancels the other. The first is that the government still cannot price the problem it is solving. A cleanup liability quoted as $16 billion to $165 billion is an admission that after decades of work the department does not know its own bill within an order of magnitude, and GAO has said so plainly, in 2004 about the estimate, in 2022 about the missing finish line, and as far back as 1995 about the fragmented research. That is a real failure of accounting and management, and the tiny office at Fort Belvoir cannot fix a range-wide estimate that depends on how forty years of ranges were inventoried.

The second true thing is that the coordination worked where it could. The discrimination problem that made everything expensive is meaningfully less unsolved than it was, and the 30 to 40 percent savings on survey and excavation is the kind of return that justifies a clearinghouse many times its own cost. The office is small, largely invisible, and folded inside Army research accounts with no separate budget line to point at, which is why almost no one has heard of it. The fair verdict is mixed and should stay mixed. The enterprise around it wastes money it cannot count, and the specific thing this office was built to improve, the technology that decides whether you dig, got better and cheaper. Both entries belong in the ledger, and anyone who reports only one of them is telling half the story.

Fact-check notes and sources

The Army is the designated DoD Executive Agent for the Unexploded Ordnance Center of Excellence, a role set by DoD Directive 5101.13E (March 2, 2006, certified current as of November 1, 2011) and confirmed in the department's registry, which was last updated in March 2024 with active points of contact. DoD Executive Agent Registry, UXOCOE entry; DoD Directive 5101.13E.
The load-bearing figure is a projected liability, not spent money and not a contract ceiling. As of April 2003, DOD estimated $16 billion to $165 billion to clean up unexploded ordnance and munitions on about 10,444 operational ranges covering 24.6 million acres, and GAO judged the estimate questionable because it rested on unvalidated assumptions, with per-acre costs varying roughly tenfold between services (about $755 for the Air Force versus about $7,577 for the Army). GAO-04-601, DOD Operational Ranges.
The Formerly Used Defense Sites figures are a separate accounting from the operational-range estimate. In 2021 DOD estimated about $11.9 billion (projected cost-to-complete) to clean up roughly 1,700 FUDS sites, obligated over $1 billion (spent) from FY2016 through FY2020, had completed about 45 percent of FUDS munitions sites, and had set no completion goal for FUDS munitions cleanup. GAO-22-104744, Environmental Liabilities: Formerly Used Defense Sites.
The office grew out of a documented 1995 finding of fragmented, duplicative UXO research with no central visibility, which recommended a multiagency clearinghouse and a DoD executive agent. GAO/NSIAD-95-197, Unexploded Ordnance.
The efficiency gains are demonstrated, not projected: advanced geophysical classification developed through SERDP and ESTCP distinguishes hazardous ordnance from metal scrap at live demonstration sites 70 to 90 percent of the time, cutting combined survey and excavation costs by roughly 30 to 40 percent. SERDP-ESTCP UXO focus area.

This post is informational and journalistic, drawn from public records, and is not legal, financial, or policy advice; dollar figures are attributed to their fiscal year and labeled as projected, obligated, or approximate where the sources so indicate.