9 April 1997
On February 27, in response to President Clinton's tasking to his Advisory Committee (PAC) on Gulf War Veterans' Illnesses, I appointed Robert Walpole to be my Special Assistant for this issue. I asked him to have a Persian Gulf War Illnesses Task Force running by 3 March. One of its first tasks was to determine what the Intelligence Community knew about the Khamisiyah storage facility, when we knew it, and what we did with that information. Former task forces had focused on identifying areas of potential exposure to chemical agents and on assessing what had happened in March 1991 at Khamisiyah.
This paper and the accompanying documents do not contradict previous intelligence warnings before Desert Shield/Desert Storm: that Iraq was likely to have chemical warfare (CW) munitions in the theater of operations and that Iraqi CW munitions might not be marked. It also does not change our judgment that Iraq did not use chemical weapons during Desert Storm.
The paper does, however, illustrate that intelligence support associated with Operations Desert Shield and Desert Storm--particularly in the areas of information distribution and analysis--should have been better. Key issues include problems with multiple databases; limited sharing of "sensitive" but vital information; and incomplete searches of files while preparing lists of known or suspect CW facilities. This Task Force is preparing recommendations to address these problems and will continue to assess how we can improve. We will move aggressively to implement those recommendations.
Finally, I would like to thank the United Nations Special Commission for its part in this public release of information. I also want to reiterate my commitment to the men and women who served this country in the Persian Gulf. We owe them a full and accurate accounting of what happened. This paper is a part of that commitment. But this commitment also extends to enhancing intelligence support to men and women who will serve in the future.
The US Intelligence Community (IC) (1) has assessed that Iraq did not use chemical weapons during the Gulf war. However, based on a comprehensive review of intelligence information and relevant information made available by the United Nations Special Commission (UNSCOM), we conclude that chemical warfare (CW) agent was released as a result of US postwar demolition of rockets with chemical warheads in a bunker (called Bunker 73 by Iraq) and a pit in an area known as Khamisiyah.
Before the Persian Gulf war, the IC assessed that Iraq had a significant chemical weapons capability, including chemically armed Scuds. The IC also assessed that Iraq had used chemical weapons on numerous occasions against Iran and its own citizens. At the time of the US deployments to the Persian Gulf, the IC had reached consensus that Iraq had chemical weapons in its arsenal, had likely forward-deployed these weapons, and was prepared to use them against Coalition forces.
When Desert Shield began, our concerns about the Iraqi use of weapons of mass destruction became the focus of our chemical weapons analytic and collection efforts. IC analysts sought to identify possible Iraqi CW facilities for targeting purposes. Sites throughout Iraq were identified, albeit on incomplete information.
Several CIA chemical and biological warfare analysts maintained internal 24-hour coverage during the start of the air war and later through the ground campaign to provide support to senior CIA officials and key policymakers. Although there were many reports of chemical weapons use, analysis of all-source information indicated that these were false alarms and that chemical weapons were not used. CIA later published an assessment concluding that Iraq had never deployed chemical weapons to its frontline units, subsequently decided to move them out of the theater prior to war, and never used them against Coalition forces.
Map of Iraq
In the months immediately following the Gulf war, the IC turned its assets to identifying and characterizing Iraq's surviving CW and other weapons-of- mass-destruction capabilities. As the following intelligence chronology demonstrates, the IC did not focus on the possible release of chemical agent until after veterans' health concerns surfaced.
Predemolition photo of Khamisiyah ammunition storage area showing Bunker 73 and pit area.
When viewed with the clarity of hindsight, the history of events at the Khamisiyah facility appears relatively simple. The following intelligence chronology, however, underscores the complexity of the issue and the ambiguity intelligence analysts face in piecing together sometimes conflicting information.
The IC has access to a large volume and multiple sources of information, but individual analysts rarely have access to all information on a given topic. Furthermore, not all information we receive is clear or correct. Analysts normally must sort through large volumes of reporting, much of which is contradictory, inaccurate, incomplete, or ambiguous, to reach a single analytic judgment. Finally, resource constraints and conflicting priorities limit the number of intelligence issues that can be addressed in depth. (2)
Intelligence on Khamisiyah was buried in a large volume of reporting that needed to be sorted and analyzed. Only after a massive interagency effort was this evidence identified, isolated, analyzed, and prepared for release. The sheer volume of reporting on Iraq greatly complicated our ability to single out this one facility--which was only a small part of the Iraqi CW effort--and properly exploit information once received. We will continue to search for relevant documents and to release useful information.
Before its demolition by US forces in 1991, the Khamisiyah facility was a large ammunition storage depot in southeastern Iraq, approximately 100 kilometers (km) from the Kuwaiti border. The facility we now call Khamisiyah was first identified in intelligence information from September 1976, while it was under construction. The IC identified the facility as a conventional ammunition depot. In June 1977, it was assigned the name Tall al Lahm--after a nearby town--in our imagery database.  This remained the most common name the United States used for the facility until mid-1996, when the name used by the Iraqis--Khamisiyah--was adopted to avoid confusion. Information available to the IC identified the facility's location as 304700N/0462615E. 
The first known reference to the depot using the Iraqi name Khamisiyah occurred in intelligence reporting in April 1982, when the "Al Khamisiyah ammunition depot" was mentioned in connection with the transfer of munitions in support of Iraqi military operations during the Iran-Iraq war.  This report did not specify the facility's location, but subsequent reporting associated it with the geographic coordinates of the nearby town of Khamisiyah (3046N/04629E).  Neither this reporting nor the intelligence from 1976 hinted at any connection with chemical weapons. This facility was maintained in a National Security Agency database as Khamisiyah, and in the imagery database as Tall al Lahm. No apparent effort at the time was made to reconcile the facility names.
While not discovered until 20 March 1997, intelligence acquired in July 1984 currently provides the earliest potential indication that chemical weapons or chemical warfare activities might have been associated with the Khamisiyah depot at the time. As part of an ongoing review of historical files on Khamisiyah, we discovered information indicating that a decontamination vehicle normally associated with tactical chemical defense was at the depot. This activity was not associated with any specific bunker or other storage structure and, by itself, does not provide confirmation of chemical weapons storage.
The first recognized connection between Khamisiyah and chemical weapons--and the only such evidence prior to Iraq's August 1990 invasion of Kuwait--appeared in a CIA human-source report obtained in May 1986. (3) This report was a translated copy of an Iraqi CW production plan and discussed the transfer of chemical weapons to Khamisiyah:
3,975 155-mm mustard-loaded artillery grenades [sic] have been issued (from June 1984 to March 1985) to al-Khamisiyah warehouses. We do not have official data about using this quantity by the third army corps. The warehouses currently have 6,293 150-mm mustard bombs [sic], enough to meet front demands for four days on a 15-minute mission. (4) 
Khamisiyah Ammunition Storage Area
This report was made available to select individuals in the policy and intelligence communities--including DoD officials--but did not receive broad distribution because of its sensitivity. (5) Of note, the munitions mentioned above were artillery shells containing mustard agent. Thus, they were different from those blown up by US troops at Khamisiyah in 1991; those were 122-mm rockets containing the nerve agents sarin and GF, which--according to Iraqi declarations--were moved to Khamisiyah in January 1991.
A CIA assessment in November 1986 used the above information to conclude that chemical weapons were stored during the Iran-Iraq war "at the southern forward ammunition depot located at Tall al Lahm." (6) This assessment shows that a connection had been made at that time between Khamisiyah and what we knew as Tall al Lahm. It also stated that "a new generation of 16 bunkers will expand Iraq's capability to store CW munitions at six airfields and at three ammunition storage depots that are strategically located throughout the country."  Subsequent analytic efforts focused on this new generation of bunkers--dubbed "S-shaped" bunkers by the IC because of their unusual shape--as the most likely storage sites for forward-deployed Iraqi chemical weapons.  None of these bunkers was located at Khamisiyah: the nearest were located at Tallil Airfield and the An Nasiriyah Southwest depot. Over time, the IC developed a bias toward the S-shaped bunkers as intended for CW storage. By 1991, this bias led analysts to conclude, erroneously, that reporting about Khamisiyah referred to the An Nasiriyah SW depot.
Reporting from early 1988 with the same high reliability, sensitivity, and handling as the May 1986 report, stated with regard to Iraqi chemical weapons storage locations:
As of early 1988, Iraqi artillery shells, bombs, and rockets loaded with chemical warfare (CW) materials were stored either at Samarra or in a large ammunition dump near the town of Muhammadiyat. This facility was located about 12 [sic] kilometers outside of Baghdad. Additionally, 122-mm rockets temporarily were stored at the airbase in Kirkuk for further transport to Sulaymaniyah. 
This report, especially with the "either-or" construction, suggested that chemical weapons were not stored at Khamisiyah or any other location in southern Iraq at that time. In addition--because we had previously identified an S-shaped bunker at Kirkuk airfield--mention of CW storage at "the airbase in Kirkuk" in the 1988 report further strengthened the IC's focus on S-shaped bunkers and the assessment that they would be used for forward deployment of chemical munitions, but were not intended for long-term storage.
This information, the strengthened analytic bias toward S-shaped bunkers, and several other factors may have played a role in Khamisiyah's omission from CW facility lists generated by the IC between 1986 and 1991. For example, following the May 1986 report and the November 1986 assessment, some analysts believed the reported activity at Khamisiyah represented temporary, forward-deployed storage. (7) We have located no additional reporting suggesting chemical weapons were stored at Khamisiyah from May 1986 to the end of the Iran-Iraq war in 1988--a period in which Iraq used thousands of tons of CW agents against Iran.
Additional information concerning possible chemical weapons storage at Khamisiyah was obtained shortly after Iraq invaded Kuwait, but was not recognized until early 1996 during a review of the Khamisiyah facility as a possible CW agent release site. Intelligence acquired on 18 August 1990 showed what was reported only as munitions transloading activity. Because CW analysts did not carry Khamisiyah on their lists of CW-related facilities in 1990, the information was not reviewed by chemical weapons specialists at the time. We now judge that this activity might have been a chemical weapons transfer under way outside a bunker at Khamisiyah; we have determined that this was not Bunker 73.
Khamisiyah was not mentioned as a chemical weapons storage location in any finished intelligence document or list of facilities produced during the months leading up to Desert Storm. At the time, the IC unanimously identified S-shaped bunkers as the most likely locations for forward deployment of chemical weapons when tasked to identify Iraqi CW facilities. As a result, Khamisiyah was not added to IC lists of suspect Iraqi CW facilities. Analysts emphasized at the time, however, that chemical weapons could be stored anywhere--even in the open.  Nevertheless, the Tall al Lahm facility was mentioned in 28 February 1991 military intelligence information requests as suspected to have possibly contained chemical munitions prior to the ground war. 
A report pertaining to chemical weapons at a location we now know to be Khamisiyah was obtained during Desert Storm. On 23 February 1991, a CIA reporting cable indicating potential storage of chemical weapons was sent to CIA Headquarters and Desert Storm support elements in Saudi Arabia. This cable reported the location to be 3047N/04622E. The cable did not provide the name of the facility or any details about the chemical weapons, but mentioned the information corresponded to a storage area "east of Juwarin." The chain of acquisition of this report was quite tenuous. The source was reportedly in the Iranian Air Force or Air Force--related industry; he apparently passed the information through foreign intermediaries.  In Saudi Arabia, this report was immediately made available to Central Command (CENTCOM) and some subordinate US military elements in Riyadh.  Review of the cable shows the coordinates to be at or near the town of Tall al Lahm on various maps, and the storage area (unnamed) on the Joint Operations Graphic (JOG) series map to be near "Al Khamisiyah." This storage area is the Khamisiyah storage facility.
On 24 February, CIA was informed that CENTCOM/Collections tasked its assets to investigate this facility. On 25 February 1991, CIA/DO telephoned a CIA analyst and relayed some of the information in the cable. The analyst noted that the coordinates were close to the An Nasiriyah depot and Tallil airfield, both of which were carried as suspect CW storage facilities because of the presence of S-shaped bunkers. The analyst consulted with the National Photographic Interpretation Center (NPIC) and learned that CW-related activity had been reported at An Nasiriyah in mid-January 1991. On the basis of this activity, the analyst suspected that the report referred to the An Nasiriyah depot. (8)  Nevertheless, this misidentification was never relayed to DoD. Instead, CIA indicated that "WE ARE UNABLE TO IDENTIFY SPECIFIC CHEMICAL STORAGE FACILITY AT [referenced] LOCATION."  The second paragraph of the 23 February 1991 cable was subsequently sent to select CIA analysts.
During 23-25 February 1991, Army Central Command (ARCENT) issued a collection emphasis for the coordinates mentioned in the 23 February CIA cable; this emphasis, however, requested confirmation that Iraqi troops were present and did not mention chemical weapons.  In addition, it is unclear if there is any direct relationship between this information and a 26 February 1991 XVIII Airborne Corps log entry stating that there were "possible chemicals on Objective Gold," a location at or near Tall al Lahm. (9) 
Also in February 1991, DIA completed a review of nonrefrigerated "12-frame" bunkers. (Just as the previously mentioned S-shaped bunkers were associated with the storage of chemical weapons, 12-frame bunkers were believed to be potential storage sites for biological and possibly chemical weapons.) In late February, DIA notified CENTCOM that such bunkers were at Tall al Lahm and at five other facilities. 
On 28 February 1991, CENTCOM's National Military Intelligence Support Team (NMIST) requested that ARCENT determine by 4 March whether chemical or biological weapons were present at 17 suspected CBW storage locations occupied by ground forces. The request stated that "THESE SITES WERE SUSPECTED TO HAVE POSSIBLY CONTAINED SPECIAL MUNITIONS PRIOR TO THE GROUND WAR." The Tall al Lahm depot and the adjacent revetted storage area were included in this list.  A response from VII Corps on 1 April states that no chemical weapons were found at either part of Tall al Lahm or at 11 other sites on the list occupied by US troops. Four of the facilities were not occupied by US troops and could not be surveyed. (10) 
Postwar reports received by the IC indicated that no chemical weapons were found in the Kuwaiti Theater of Operations (KTO).  These reports were generally accepted by the IC. While most national-level sources said that Iraq's chemical munitions were probably not marked, lower-level tactical units were disseminating information on markings that was gathered from enemy prisoner of war (EPW) interrogations and other local sources.  As a result, either the standard US CW marking system or incorrect markings data gleaned from EPWs were mistakenly used by some CENTCOM troops as the basis for determining if captured Iraqi munitions contained chemical agents. On 6 March 1991, in an attempt to gain clearance to enter the KTO, CIA analysts relayed concerns about the markings issue to CENTCOM J-2 and J-3 officers in Saudi Arabia through the Joint Intelligence Liaison Element in Saudi Arabia (JILE/Saudi):
ALTHOUGH THERE HAVE BEEN EPW REPORTS THAT IRAQ'S CHEMICAL MUNITIONS HAVE COLORED BANDS [or] OTHER MEANS OF IDENTIFICATION, OUR EXPERIENCE WITH THE MUNITIONS IRAQ USED IN ITS WAR WITH IRAN INDICATES THAT THE IRAQIS DID NOT/NOT MARK THEIR CHEMICALLY FILLED MUNITIONS. WE BELIEVE THE EPW REPORTS ON MARKINGS MAY REFLECT TRAINING CLASSES ON CHEMICAL MUNITIONS USING SOVIET EXAMPLES...IF PERSONNEL IN THE KTO ARE NOT AWARE OF THIS POSSIBILITY, OPPORTUNITIES TO SUCCESSFULLY IDENTIFY CHEMICALLY FILLED MUNITIONS MAY BE MISSED. WHEN CACHES OF UNMARKED MUNITIONS ARE DESTROYED, THERE IS ALSO THE POSSIBILITY THAT INDIVIDUALS COULD