The Chemical Weapons Buried Under American Yards, and t...

In January 1993, a backhoe crew digging a utility trench in Spring Valley, a leafy neighborhood in the northwest corner of Washington DC, hit something it did not expect. Buried military ordnance, left over from an era when the ground next to the American University campus served as an Army chemical weapons proving station in World War I. The shells and buried glassware that surfaced there were the residue of experiments the government had covered over and mostly forgotten. Cleaning them up became one of the longest and most expensive single jobs handled by a standing federal capability most people have never heard of. It is called the Recovered Chemical Warfare Material Program, it is run by the Army, and the National Academies projects it will cost somewhere between $2.5 billion and $17 billion to work through over thirty years.

What lies under the yards

Between World War I and 1969, the United States made, tested, and disposed of chemical weapons at dozens of installations, and it did not always keep good records of where the leftovers went. Some were dumped at sea. A great many were buried, sometimes in dedicated trenches on a military reservation, sometimes on land that later became a suburb, a campus, or a golf course. Along with intact munitions, the ground holds chemical agent identification sets, small kits of glass ampoules of live agent once used to train troops to recognize a chemical attack by smell. Decades later, a shovel or a drill rig finds them.

The Recovered Chemical Warfare Material Program, or RCWM, is the government's standing answer to that problem. It is the national capability for responding when old chemical weapons or agent identification sets turn up inside the United States. The Army investigates a suspected site, safely recovers the munitions and any container holding an unknown liquid or a chemical agent fill, and destroys the items, often on the spot, using mobile treatment systems built for exactly this work. The program operates in support of the Defense Environmental Restoration Program, the broader effort to clean up the physical legacy of the country's military past. Unlike the destruction of the declared chemical stockpile, which had a known inventory and a finish line, RCWM deals with an unknown quantity of buried material scattered across the map, which is why it is built to be permanent rather than to end.

Who is in charge

Within the Department of Defense, the Army carries this mission as the DoD Executive Agent for Recovered Chemical Warfare Material. The official DoD Executive Agent registry lists the Army in that role, and DoD Directive 5101.17E, dated May 11, 2016, names the Secretary of the Army as the Executive Agent, a responsibility the Secretary delegated to the Assistant Secretary of the Army for Installations, Energy and Environment.

That 2016 directive is the current governing paper, but it formalized a role that already existed. When the National Academies examined the program in 2012, the designation did not trace to a numbered directive at all. It traced to a memorandum from the Under Secretary of Defense for Acquisition, Technology and Logistics, dated March 1, 2010, which designated the Secretary of the Army as Executive Agent and built on the Army's own Recovered Chemical Warfare Material Program Implementation Plan of July 2007, approved by the Secretary of the Army. The committee that reviewed all of this was pointed about the gap it saw, noting that long-range RCWM policy was, in its words, not clearly defined. In other words, the line of authority was real, but the durable, top-level policy document came years after the work was already underway. The chain today reads cleanly, from the 2007 Army plan, through the 2010 designation, to the 2016 directive, but it took the better part of a decade to write it down.

The money, projected and spent

The single number worth holding onto is a projection, and the band around it is enormous. The Army's July 2007 Implementation Plan, reproduced and cited by the National Academies in its 2012 report as (DOD, 2007), put the total 30-year life-cycle cost of the RCWM program at a low of $2.5 billion and a high of $17 billion. The report states the figure in its introduction and breaks it out later in the study. The low case is roughly $765 million for investigation, cleanup, and site close-out plus about $1.5 billion for assessment, destruction, and emergency response. The high case is about $1 billion plus $16 billion. This is money projected across three decades, not money spent, and a spread from $2.5 billion to $17 billion is not precision. It is a wide guess about an unknown quantity, which the report itself calls difficult to predict.

The scale of the problem behind that guess is large. The National Academies counted about 249 known or suspected buried chemical warfare material sites drawn from a 2007 Army inventory, plus roughly 699 more locations supported only by anecdotal evidence. The report's summary rounds the footprint to approximately 250 sites in 40 states, the District of Columbia, and 3 territories, though the underlying 249-site inventory maps more precisely to 35 states plus the District of Columbia and 2 territories, Guam and the U.S. Virgin Islands, with the broader 40-state figure coming from a 2011 program estimate. The largest single site is likely Redstone Arsenal in Alabama, where the disposal trenches run for more than five miles.

Against that projection sits a concrete figure for money actually spent, and it comes from Spring Valley. By the time the last property was closed out, the Army had spent an amount reported at roughly $300 million to clean one residential neighborhood of about 661 acres covering some 1,600 homes. That total is worth flagging honestly. The confirmed public waypoint is about $221 million spent by 2013, before the work paused, which is consistent with growth toward the roughly $300 million figure by the time the digging finished around 2021, but the exact final total, the acreage, and the close-out date could not be independently confirmed against primary sources for this piece. Program destruction activities are funded through the Defense-wide appropriation named Chemical Agents and Munitions Destruction, Defense, carried out through the Non-Stockpile Chemical Materiel Project. The lesson of the numbers is simple. One neighborhood alone consumed a sum in the same order of magnitude as a tenth of the entire program's low-end 30-year projection.

The efficiency critique

The government was warned early that this work would blow past its estimates, and it did. In 1997 the Government Accountability Office told Congress that non-stockpile chemical materiel disposal costs and schedules would likely exceed the projections, for reasons that had nothing to do with mismanagement and everything to do with the nature of the job. So much about the buried weapons was unknown, the destruction technologies were unproven, and environmental permitting was slow. GAO returned to the theme in 2002, flagging the many uncertainties around the effort. Spring Valley then proved every one of those points in a single ZIP code, with a dig that was supposed to wrap up slipping year after year for nearly three decades.

The estimates themselves have wandered widely enough to undercut confidence in any of them. A 2003 DoD Inspector General report worked from a separate buried-munitions figure of about $11.7 billion that rested on an un-updated 1996 estimate, a reminder that the program's price tag has been a moving target for as long as anyone has tried to fix it. And when the Inspector General looked directly at how the Army was running its live response sites in 2018, it found a curious split verdict. The Army had complied with its own rules at all three of its active response sites. The problem was the rules. The interim guidance the Army was following was eight years old, and a governing Corps of Engineers pamphlet was thirteen years old, both well past the five-year currency standard set by Army Regulation 25-30. Compliance with stale guidance is not the same as sound management. Underneath all of it sits the structural critique that no audit can resolve. Because the program is open-ended by design, there is no firm end date and no total price the public can hold it to.

The public-good defense

Then there is the other side, which is just as real. The hazard is not theoretical, and it often sits under someone's lawn. Spring Valley alone yielded 556 munition items, 23 of them filled with chemical agent, along with dozens of sealed glass containers of agent and thousands of tons of contaminated soil, all inside a residential neighborhood that includes a university and several foreign embassies. That is not a paperwork risk. When a contractor's backhoe hits a live World War I mustard shell in a place where children walk to school, someone has to be able to identify it, contain it, and destroy it without hurting anyone, and RCWM is the only national body that can do all three.

Viewed that way, the deliberate pace that reads as delay in a GAO report reads as caution in a neighborhood. The record at the active sites shows compliance and, importantly, an absence of accidents, which is the outcome that matters most when the material is nerve or blister agent a century past its shelf life. Destroying these non-stockpile items also keeps the United States in line with its obligations under the Chemical Weapons Convention, which requires the country to eliminate recovered chemical weapons, not just the ones it declared. A cleanup that is slow and expensive but safe and law-abiding is a defensible thing to buy.

Reading the ledger

Both readings are true at once, and the honest move is to let them stand together rather than collapse them into a headline. On one side of the ledger, this is a program with a projected cost band so wide, $2.5 billion to $17 billion, that it barely qualifies as an estimate, running on guidance that went years out of date, with no finish line and a single neighborhood that swallowed roughly $300 million and most of thirty years. On the other side, it is the only mechanism the country has for safely removing live chemical weapons from under family homes, it has done that work without killing anyone, and it keeps the United States compliant with a treaty it signed. The waste critique and the public-good defense are not in competition here. The uncertainty that makes the program impossible to price is the same uncertainty that makes it necessary to keep. What the buried shells of Spring Valley finally teach is that the bill for a war does not close when the war does. It comes due, in pieces, decades later, wherever someone happens to dig.

Fact-check notes and sources

The Army is the DoD Executive Agent for Recovered Chemical Warfare Material. The current designating paper is DoD Directive 5101.17E (May 11, 2016), which names the Secretary of the Army, delegated to the Assistant Secretary of the Army for Installations, Energy and Environment. At the time of the 2012 National Academies study the role instead traced to a March 1, 2010 memorandum from the Under Secretary of Defense for Acquisition, Technology and Logistics, building on the Army's July 2007 Implementation Plan, and the committee noted long-range policy was "not clearly defined." DoD Executive Agent registry, DoD Directive 5101.17E.
The program responds to and destroys recovered chemical warfare materiel inside the United States and funds destruction through the Defense-wide Chemical Agents and Munitions Destruction, Defense appropriation, carried out through the Non-Stockpile Chemical Materiel Project. DoD RCWM Program mission and funding (DENIX, OSD).
The projected 30-year life-cycle cost of $2.5 billion to $17 billion is a projection, not money spent, and the band is deliberately wide. It originates in the Army's July 2007 Implementation Plan, reproduced by the National Academies as "(DOD, 2007)," and appears in the report's Introduction and is broken out in Chapter 7, not Chapter 3. Site counts (about 249 known or suspected, plus roughly 699 anecdotal) and the geography are approximate. The report's summary rounds to "approximately 250 sites in 40 states, the District of Columbia, and 3 territories," while the underlying inventory maps to 35 states plus DC and 2 territories, with the 40-state figure from a 2011 program estimate. National Academies, Remediation of Buried Chemical Warfare Materiel (2012).
GAO warned in 1997 that non-stockpile disposal costs and schedules would likely exceed estimates because of uncertainty about the buried weapons, unproven destruction technology, and slow environmental permitting. The DoD Inspector General found in 2018 that the Army complied with policy at all three active response sites, but its interim guidance was eight years old and a Corps of Engineers pamphlet thirteen years old, both past the five-year currency standard in Army Regulation 25-30. GAO/NSIAD-97-18, DODIG-2018-042.
The Spring Valley cleanup specifics are partly unverified against primary sources. The roughly $300 million total spent, the approximately 661-acre footprint, and the late-2021 close-out date are reported and plausible but could not be independently confirmed for this piece; the confirmed public waypoint is about $221 million spent by 2013 before the digs paused. The recovered items (556 munitions, 23 with chemical agent) are drawn from the public record of the site. Program context: DoD RCWM Program (DENIX, OSD).

This post is informational and journalistic, drawn from public records, and is not legal, financial, or policy advice, with all dollar figures attributed to their stated fiscal year or source date and projected figures labeled as projections rather than money spent.

The Chemical Weapons Buried Under American Yards, and the Bill to Dig Them Up