This technical report describes the details of the modeling and risk characterization of possible chemical warfare agent exposure in the Gulf War. This report describes the efforts that the Directorate for Deployment Health Support (DHS) of the Special Assistant to the Under Secretary of Defense (Personnel and Readiness) for Gulf War Illnesses, Medical Readiness, and Military Deployments (formerly the Office of the Special Assistant to the Deputy Secretary of Defense for Gulf War Illnesses) began in 1996. The report also details efforts to reconstruct the demolition of chemical-warfare-agent-filled munitions at Khamisiyah, Iraq, on March 10, 1991; to identify the servicemembers who possibly may have been exposed; and to evaluate the possible effects of that exposure.
To present the technical details of the modeling and risk characterization of the possible chemical warfare agent exposure due to the demolitions at the Khamisiyah Pit, this report first presents an Overview having Background and Summary sections, then follows with the format of the National Research Council’s (NRC) guidelines for a qualitative risk assessment (NRC, 1994) written in a layman’s language. Appendix A and Appendix B present the technical details of the modeling procedures and hazard and risk identification, respectively. The qualitative risk assessment steps and our approach for accomplishing them are as follows:
This process determines whether exposure to a substance can increase the incidence of adverse health conditions and characterizes the nature and strength of the evidence of causation.
For this report, DoD accomplished the initial hazard identification step by carrying out an extensive review of the toxicology of sarin and cyclosarin. This included reviewing reports describing both animal and human response to exposure to sarin and cyclosarin. Section II is a summary of that toxicology review, and Appendix B contains the complete toxicology review. The available information shows the high toxicity of these two nerve agents and describes the effects of exposure to a range of doses. The hazard is well defined, especially at levels that cause immediate symptoms in animals and humans.
Dose-Response Assessment: At what doses do adverse health effects become observable?
This assessment characterizes the relationship between the dose of an agent administered or received and the incidence of an adverse health effect in exposed populations; it also estimates the incidence of the effect as a function of human exposure to the agent. This process considers the intensity of the exposure, the time history of the exposure, variables that may affect response (e.g., gender, lifestyle), and other modifying factors. A dose-response assessment generally requires extrapolating from high to low doses and from animals to humans. Thus, such an assessment describes and justifies the methods of extrapolation used to predict the range of responses from the incident and should characterize the statistical and biological uncertainties inherent in these methods.
In this report, the dose-response relationship, described in Section III and detailed in Appendix B, relates the effects of varying doses. Dose-response relationship explains the terms used, how the relationship is determined, and factors affecting toxicity. It uses the concept of dose-response thresholds to explain how regulators determined exposure thresholds and occupational health limits, explains the General Population Limit (GPL) derivation and the rationale for its use, describes the relation between exposure concentration (C) and time of exposure (t), and why the short time over which the product (Ct) (i.e., dosage) is linear, explains why 8 hour or 24 hour exposure dosages were not calculated in this report.
Hazard area development and exposure assessment: What was the potential hazard area and what were the concentrations of toxic materials in that area?
This is the process of measuring or estimating the intensity, frequency, and duration of human exposures to a chemical warfare agent currently present in the environment or estimating hypothetical exposures that might arise from releasing new chemical warfare agents into the environment. In its most complete form, this method describes the magnitude, duration, schedule, and route of exposure; the size, nature, and classes of the human populations exposed; and the uncertainties inherent in all exposure estimates.
The approach to hazard area development and exposure assessment for the demolition in the Khamisiyah Pit consisted of reconstructing the source term, the prevailing meteorological conditions, and the dispersion patterns of the chemical warfare agents. After determining the resulting potential hazard area, DoD identified and notified possibly exposed individuals. In this report, Section III describes the exposure limits DoD used, Section IV describes the methods by which DoD and CIA used models to estimate the potential hazard areas covered by sarin and cyclosarin released from the Khamisiyah Pit and identifies how DoD identified and notified possibly exposed servicemembers. Appendix A provides the technical details of why the models were used, how they were selected, how they were used, and the results.
Risk Characterization: How does the received dose compare to the doses that produce adverse health effects?
This step integrates the information obtained in the previous three steps to develop a qualitative (or quantitative) estimate of the likelihood that the hazard would produce any adverse health effects in exposed individuals. These are the results of the health risk assessment.
Using the results of modeling efforts, this report evaluates the effects of any possible exposure of US forces in the Khamisiyah vicinity to chemical warfare agents. Section V presents a summary discussion and Appendix B contains the detailed discussion.
Immediately after Operation Desert Storm ended (February 28, 1991), US Army units occupied an area in Iraq that encompassed the Khamisiyah Ammunition Supply Point (ASP), a large munitions storage depot in southern Iraq (Figure 1). The XVIII Corps (Airborne) sent combat engineer and explosive ordnance disposal units to Khamisiyah to destroy its munitions and facilities. In carrying out this mission, US forces set off two large explosions; the first on March 4, 1991, in bunkers in the ASP, and a second on March 10, 1991 which included an area outside the ASP called the Pit. Figure 2 shows destruction in the Pit some days after the demolition.
Figure 2. Destruction of 122mm rockets in the Pit
In April 1991, United Nations Security Council Resolution 687 created the United Nations Special Commission on Iraq (UNSCOM) to identify and destroy Iraq’s surviving chemical and biological weapons and ballistic missiles. UNSCOM was to have the weapons moved to an Iraqi destruction facility or destroy the weapons in place. In 1991, 1992, 1996, and 1998, UNSCOM inspected Khamisiyah. In October 1991, Iraq’s officials led UNSCOM inspectors to three sites in the Khamisiyah area that contained chemical weapons (Figure 3), two of which they claimed were destroyed by US forces:
Figure 3. Site locations shown to UNSCOM
The US government did not immediately make the connection between the chemical munitions found by UNSCOM in three locations in and around the Khamisiyah ASP in October 1991 and any US demolitions operations there. In 1995, as a result of presidential concerns and the interest of two Central Intelligence Agency (CIA) employees, the CIA began a reexamination of relevant intelligence and the DoD formed the Persian Gulf Illnesses Investigation Team (later called Persian Gulf Investigation Team.) By October, the team had identified the 37th Engineer Battalion and other US forces that had occupied the area around Khamisiyah during the Gulf War and immediately following the cease fire. Then, in November 1996, the Secretary of Defense established the Office of the Special Assistant to the Deputy Secretary of Defense for Gulf War Illnesses (now called the Directorate for Deployment Health Support of the Special Assistant to the Under Secretary of Defense (Personnel and Readiness) for Gulf War Illnesses, Medical Readiness, and Military Deployments (hereafter called DHS)) to focus on DoD investigations and expand the investigation into Gulf War veterans’ complaints of undiagnosed illnesses.
Also in 1996, at the direction of the Presidential Advisory Committee for Gulf War Illnesses, a panel established to provide oversight to Gulf War illnesses investigations, the CIA modeled the possible chemical warfare agent released from the March 1991 demolition operations at Khamisiyah. The CIA’s report, published in August 1996 (CIA, 1996), centered on the Khamisiyah Bunker 73 explosion of March 4, 1991, and concluded that the chemical warfare agent released in this detonation likely moved to the east and northeast, away from US forces. For the demolition in the Pit on March 10, 1991, the quantification of the modeling parameters contained significant uncertainties, including how the chemical warfare agent was released from an open air demolition and the number of rockets destroyed. These uncertainties and the lack of contemporaneous weather measurements prevented definite conclusions about the possible impacts from the explosions in the Khamisiyah Pit. However, the CIA’s initial model runs indicated that any chemical warfare agent released by the March 10 detonation in the Pit generally would have moved in a southerly direction toward US forces.
At DoD and CIA’s request, the Institute for Defense Analyses (IDA) convened a panel of independent experts to review CIA’s modeling for the Pit at Khamisiyah. The IDA panel, also concerned about the uncertainties in the source characteristics used in the models, recommended:
"Collect all meteorological observations, including classified sources.
Apply an ensemble of credible atmospheric models, including civilian models, to enhance confidence in the weather simulation.
Use a combination of transport and diffusion models to determine how the agent is dispersed by the simulated weather."
Because US forces may have been exposed to low levels of chemical warfare agents as a result of the Pit demolition, the early work of DHS placed an emphasis on researching US military operations at Khamisiyah. DHS began a program to acquire additional data about the Pit detonation by interviewing participants in the demolition, and on February 21, 1997, published the first Khamisiyah case narrative. The narrative provided important insights into what actually took place and which US military units were involved.
D. Summary of DoD Coordinated Modeling and Risk Characterization
The Directorate for Deployment Health Support began a comprehensive review of the events on and after the March 10 Khamisiyah demolition and, with CIA, began interviews to reconstruct source information and acquire additional data about the Pit detonation. These interviews obtained information on the numbers of rockets in the Pit and the quantities and placement of the explosive charges (C4) used to destroy the rockets. DHS and CIA, subsequently began a new modeling effort as recommended by the IDA panel.
CIA used the additional information regarding number of rockets and the placement of charges, along with estimates of the heights of the stacks of 122mm rocket crates in the Pit, information from Iraq’s declarations to UNSCOM, and results from UNSCOM inspections to estimate the number of damaged rockets in the Khamisiyah Pit as 500. Subsequently, DHS and CIA used this number, together with the weight of chemical warfare agent in each rocket’s warhead (6.3 kilograms), agent purity (approximately 50%), and the ratio of sarin to cyclosarin (3:1) to estimate the source characteristics of the agent released for the modeling.
In May 1997, DHS and CIA conducted a field test having several trials using simulant-filled 122mm rockets in exact replicas of the shipping crates in the Pit to figure out how the rockets released their contents when the soldiers detonated C4 explosives on the outside of the warheads (Nudell et al., 1997). The test was designed to determine the effects of the explosives on the rockets as described by those who originally placed them for the Pit demolition.
The results of the test indicated that it was highly probable that many of the rockets’ central burster tubes exploded, ruptured the surrounding chemical warhead, and allowed the liquid chemical warfare agent to spill onto the ground and onto the wooden packing crates. According to the CIA data analysis, the bulk of the chemical warfare agent formed liquid pools, soaking into the sandy soil and into the wooden crates before slowly evaporating. Only a small amount was instantaneously aerosolized at the detonation.
These estimates of the source characteristics (i.e., amount of and manner in which the chemical warfare agent was released) further defined the modeling parameters for DHS to use in predicting the potential hazard areas (i.e., the areas where forces may have been exposed to nerve agents)
In conjunction with developing the source characteristics, DoD carried out a general review of sarin’s and cyclosarin’s toxicologies to identify the toxicity of the hazard and to examine and describe the possible adverse health hazards from exposure to these nerve agents. DoD found that human studies specifically designed to address chronic (i.e., lifetime) effects from short-term (less than 24 hours) exposure to sarin were not available. This conclusion agrees with the results of several recent reviews performed by the RAND Corporation (Augerson, 2000), the Institute of Medicine, National Academy of Sciences (IOM, 2000), and Romano et al. (2001) of the effects of acute exposure to low levels of chemical warfare agent—"There is inadequate/insufficient evidence to determine whether an association does or does not exist between exposures to sarin at low doses insufficient to cause acute cholinergic signs and symptoms and subsequent long-term health effects." (Augerson, 2000)
DoD next reviewed the current US Army and US Department of Health and Human Services occupational and general population guidelines for exposure to sarin vapor and the critical toxicological studies used to develop these guidelines. The Department of Health and Human Service’s Centers for Disease Control and Prevention (CDC) derived the existing occupational and general population guidelines for exposure to sarin vapor from amounts that produce no adverse effects to amounts known to produce the mildest detectable effects. CDC extrapolated the exposure guidelines (with safety margins incorporated) to chronic exposures. First Noticeable Effects (FNE) represents the time integrated concentration, below which no effect manifests, including the most sensitive, miosis. The FNE dosage for sarin is approximately 1 milligram-minutes per cubic meter of air (1 mg-min/m�), an amount about 70 to 100 times less than sarin’s lethal exposure dosage. The literature suggests associating averaging periods ranging from 10 minutes to 4 hours with this dosage. The General Population Limit (GPL) exposure level represents the concentration level of a substance anyone in the general population could inhale continually for 24 hours a day, every day, for a lifetime (70 years) without experiencing any adverse health effects. Because there were no studies of continuous, daily, lifetime exposures, the CDC derived the GPL for sarin for a lifetime exposure from short-term exposure studies to be 0.000003 mg/m3. Though CDC developed these occupational and general population limits to address chronic exposures, the exposure guideline levels at least define a certain threshold (i.e., upper bound) for a potential hazard area. The GPL for cyclosarin was assumed to be the same as the GPL for sarin in 1997, due to lack of toxicity data.
Because the evaporative release from the liquid pools that soaked into the ground and packing crates probably continued for several days after the initial detonation, modeling toxic agent dispersal in the atmosphere required knowledge of the prevailing meteorological conditions over those days. However, local weather data for the Khamisiyah vicinity from March 10 to 13, 1991, are severely limited because Iraq stopped reporting meteorological data to the World Meteorological Organization in 1981 during the Iran-Iraq War. Therefore, the detailed meteorological conditions required by transport and dispersion models had to be developed from computer modeling.
DoD and CIA expanded the quest for initial parameters for the atmospheric modeling effort by using global weather conditions contained in the Global Data Assimilation System (GDAS), Naval Operational Global Atmospheric Prediction System (NOGAPS), and European Centre for Medium-Range Weather Forecasting (ECMWF) databases; photography of soot patterns resulting from other simultaneous demolitions at the Khamisiyah ASP on March 10; and satellite photos of Kuwait’s oil well fires. DoD and CIA then used state-of-the-art prognostic meteorological models to reconstruct the meteorological conditions following the demolition in the Khamisiyah Pit. These models are highly intensive and are based on fundamental conservation laws of mass, momentum, and energy, allowing them to project meteorological conditions with only minimal data.
The DoD and CIA modeling team determined the meteorological conditions near Khamisiyah during March 10-13, 1991, to the highest possible fidelity by applying these mesoscale weather models whose version numbers and dates evolved over time (specific versions and dates are identified in the details of analyses in Appendix A):
The outputs from the above meteorological models were then used as inputs for the two primary DoD transport and dispersion models, which also evolved over time—Hazard Prediction and Assessment Capability/Second Order Closure, Integrated Puff (HPAC/SCIPUFF); and Chemical/Biological Agent Vapor, Liquid, and Solid Tracking (VLSTRACK) computer model. Using the CIA-defined source term and the simulated meteorological conditions, these transport and dispersion models (often simply called dispersion models) DoD estimated the concentration of the released chemical warfare agent as a function of space and time. The model results, together with the exposure limits (e.g., the GPL), then defined the potential hazard areas. DoD used the composite (or union) of the hazard areas predicted by each model combination as the basis for identifying the potential hazard area. DoD used the union (referred to as an ensemble) to compensate for the bias that is inherent in each model; that is, to produce a more robust result by maximizing the strengths of each model and minimizing the weaknesses.
Once the modeling teams delineated the hazard areas, the US Army Center for Health Promotion and Preventive Medicine (USACHPPM) determined the possibly affected units. USACHPPM obtained unit locations from the US Armed Services Center for Unit Records Research database, which contains daily unit geographic locations within the Operation Desert Storm theater of operations. USACHPPM then used geographic information system techniques to place units on an operational base map on a daily basis and overlaid it with the modeled daily potential hazard areas. The Defense Manpower Data Center and the DHS then took the unit identification codes for units within the potential hazard area boundaries and identified the individuals in those units. In 1997, no US forces were identified within the FNE contours nor had any observed acute symptoms, but DoD sent letters to approximately 99,000 veterans, advising them that they were located in the potential hazard area where they may have been exposed to low levels of nerve agent.
After releasing the results in July 1997, DHS asked a peer-review panel of world-renowned atmospheric scientists to review the overall technical approach to the modeling of the Khamisiyah Pit demolition. In response to the recommendations from the peer-review panel, and with new intelligence and scientific information becoming available after July 1997, DHS renewed its efforts to refine the Khamisiyah modeling results. DHS made refinements that affected the 2000 modeling results mainly in six areas: (1) meteorological modeling, (2) estimate of the total number of rockets destroyed, (3) consideration of agent removal mechanisms, (4) unit location and personnel data, (5) sarin and cyclosarin exposure thresholds, and (6) the combined toxicity of sarin and cyclosarin.
Meteorological Modeling: Improvements in the meteorological modeling were made by using more recent versions in the models and adding available additional input data.
Number of Rockets Destroyed: After releasing the 1997 results, the intelligence community continued to refine the estimate of the Khamisiyah source term[2] . According to an unclassified October 14, 1999, memorandum, CIA revised its estimate of the number of 122mm rockets damaged during the Khamisiyah Pit demolition from 500 to 225.
Agent Removal Mechanisms: The dispersion models account for the turbulent mixing processes in the atmospheric boundary layer and estimate the concentration of the released chemical warfare agent based on time and distance. In the 1997 modeling, HPAC/SCIPUFF assumed a program default for dry gaseous deposition velocity without specific properties of sarin, while VLSTRACK assumed no dry gaseous deposition because the model did not account for that physical process. Neither dispersion model considered agent decay (sometimes called degradation). The 2000 modeling considers the dry gaseous deposition by properly accounting for agents’ physical properties, and the Naval Surface Warfare Center revised the VLSTRACK model to account for dry gaseous deposition.
Unit Location and Personnel Identification: After 1997, continuing DoD efforts improved both the quantity and quality of the unit location and personnel data. For example, the initial unit location and personnel database was built primarily through paper records. Starting in 1997, DHS held a series of meetings with Army brigade and division units’ operations officers to further refine the data. In addition, the 1997 data did not contain information about Air Force personnel, a shortcoming the 2000 unit identification efforts corrected.
Exposure Thresholds: DHS used the FNE, GPL, and M8A1 alarm thresholds to define potential hazard areas. The GPL indicates low-level exposure, whereas the FNE and M8A1 indicate or warn of acute exposure. The FNE and GPL are expressed as dosages (concentrations accumulated over time), and the M8A1 as a concentration. After a careful review in 1999, the USACHPPM recommended that the short-term GPL of sarin for the Khamisiyah modeling should be 0.00003 mg/m3 (USACHPPM, 1999). As a result, the 2000 modeling used the sarin GPL concentration of 0.00003 mg/m3 times 1440 minutes (24 hours)—a dosage of 0.0432 mg-min/m3. Because cyclosarin was determined to be three times more toxic than sarin, we used a GPL dosage of 0.0144 mg-min/m3.
Agents’ Combined Toxicity: With the toxicity data for cyclosarin now determined, we used both sarin and cyclosarin toxicity values in the 2000 modeling.
Combining the improved location and personnel data with the results of the 2000 modeling, DHS estimates that roughly 102,000 servicemembers were within the GPL contours, again with none within the FNE contours. Of the 102,000 servicemembers, about 66,000 servicemembers were within the GPL contours in the 1997 simulation. The 2000 count is slightly higher than that of 1997, despite the fact the 2000 hazard areas are smaller, because of the more accurate force location and personnel data for 2000.
Based on the available data about the apparent health status of the forces at Khamisiyah, modeling and exposure data, and toxicological data, the DoD concludes the exposures the forces possibly received would have been below those expected to cause acute health effects, such as miosis, or long-term effects, such as organophosphate-induced delayed neuropathy (OPIDN.) Although exposure to sarin itself at the estimated concentrations may not result in any adverse health effects, this finding does not preclude the possibility of adverse health effects resulting from any number of combinations of noxious chemicals and other stressors.
