Chapter One:

Purpose of the Report

The purpose of this work is to conduct a comprehensive scientific literature review on the possible health effects due to exposures to the Kuwait oil well fires and to produce a stand-alone, short scientific report on the findings that would pass peer-review scrutiny.

Concerns about the possible health effects experienced by the troops deployed in the Gulf due to exposure to emissions from the oil fires triggered an extensive effort to determine the extent of those environmental exposures. That effort yielded quantitative data on what pollutants were present in the Gulf and at what concentrations.

The target audience for the present report is the scientific community familiar with the issues, methodologies, and processes that are the basic tools of environmental health science. The report contains a summary geared for a policymaker.

This work involved comprehensive library searches that yielded about 2500 titles. Of these, about 500 abstracts were examined and then reduced to approximately 250 peer-reviewed papers that were analyzed. About 180 of these are cited in this report. Also, about 35 pertinent reports from U.S. and international agencies and institutions were reviewed.

Organization of the Report

A brief introductory chapter provides a concise description of the geographical characteristics of the Gulf region, the U.S. troop presence there during Operations Desert Shield and Desert Storm, and the characteristics of the Kuwait oil fires.

The introduction is followed by a chapter reviewing the U.S. air quality standards and recommended limits for chemical hazards, and the environmental measurements available for the Gulf region before, during and after the Kuwait oil well fires. This chapter establishes the list of pollutants present and their concentration levels, shows the consistency of these results across studies, and compares the prevalence of the pollutants to U.S. ambient measurements and occupational standards. The first two chapters are a necessary background for the stand-alone report described above and in no way presume to be either a critical or comprehensive analysis and evaluation of all environmental data.

The third chapter describes the possible health effects that could be associated with pollutants present in the Gulf region. Because of the volume of references on many pollutants, the presentation is limited to possible health effects for exposures closest to those measured in the Gulf. The last section summarizes the results of several health studies related to Gulf War veterans. It also includes an overview of the veterans' symptoms information to complete the stand-alone report and in no way presumes to be either a critical or comprehensive review.

The final chapter presents the findings and conclusions of the present literature review based on the exposures measured in the Gulf region. It discusses what health effects are plausible at the exposure levels measured in the Gulf, how those compare with the list of veterans' symptoms, and what areas need further research to elucidate the relationship between exposure and effect.


Iraq invaded Kuwait on August 2, 1990. In support of United Nations Resolution 660, the United States responded by sending troops to the Persian Gulf in Operation Desert Shield. On January 16, 1991, Operation Desert Storm began with an air war against Iraq that was followed, 39 days later, by a four-day ground war. By the time hostilities concluded, 697,000 American troops had served in the Persian Gulf. Of that number, 17 percent were from the Reserve Components, and 7 percent were women. Although nearly 700,000 troops passed through the theater, the peak number there at any one time was about 560,000.

As the Iraqis withdrew from Kuwait, they set fire to over half of Kuwait's 1000 oil wells and damaged most of the rest. Industry experts estimated at the time that these fires were burning 5-6 million barrels of crude oil per day and 70-100 million m3 per day of natural gas. The greater Al Burqan oil field, with approximately 700 wells (of which 365 were ignited), was the most important field in terms of oil production, number of wells on fire, and amount of smoke generated. Oil in the field continuously rose to the surface due to high subsurface pressure. This field probably had the greatest effect on exposed humans since it was the closest to Kuwait City and other coastal communities. The first fires were extinguished in early April 1991 and the last fire was extinguished on November 6, 1991. Figure 1.1 shows the relationship between U.S.-troop presence in the Gulf and the oil fires (USAEHA, 1994). It also shows when most of the air-quality monitoring occurred.

Figure 1.1--Kuwait Oil Fires and U.S. Troop Presence

Figure 1.2--Gulf War Region

The Arabian Peninsula

The region's geography and climate are important to any consideration of the effects of the oil well fires. Saudi Arabia occupies most of the Arabian peninsula. It consists of about 2,150,000 km2 of plains with scant vegetation (except near oases), and has a population of about 16 million. Kuwait lies between Saudi Arabia, Iraq, and the Persian Gulf. It has about 18,000 km2 of sandy desert and coastal beaches, and had an estimated population of about 800,000 natives and 1,300,000 immigrants as of 1990 (Husain, 1995).

The weather in Kuwait and eastern Saudi Arabia is typical of the Sahara region. It is characterized by long, hot, and dry summers; it also has short, warm, and sometimes rainy winters, with 4 inches of rain in Kuwait and 1 inch in coastal Saudi Arabia. The region is subject to violent storms. It also is prey to fierce winds (called Shammal winds) that blow sand and dust that can reduce daytime visibility to a few meters. Average winds in Kuwait are from the northwest at approximately 14 mph. The highest recently recorded temperature was 51oC (124oF) in July 1978; the lowest was -6oC (21oF) in January 1964. Temperatures also vary widely by season, ranging from an average of 45oC (113oF) in summer to an average of 8oC (46oF) in winter (Nasralla, 1983).

Kuwait Oil Well Fire Characteristics

Typically, burning crude oil produces smoke containing a large number of particulate and gaseous species such as carbon dioxide (CO2), carbon monoxide (CO), sulfur dioxide (SO2), nitrogen oxides (NOx), volatile organic compounds (VOCs), polycyclic aromatic hydrocarbons (PAHs), hydrogen sulfide (H2S), acidic aerosols, and soot.[1]

However, the composition of crude oil differs by strata and region. Calculations based on airborne measurements of the smoke from the Kuwait oil fires in May and June 1991 (Ferek, 1992) indicate that the combined oil and gas emissions were equivalent to the burning of approximately 4.6 million barrels of oil per day, somewhat less than estimated by industry experts. The combustion was relatively efficient; about 96 percent of the fuel carbon burned was emitted as CO2. Particulate smoke emissions averaged 2 percent of the fuel burned, of which approximately 20 percent was soot. About two-thirds of the mass of the smoke was accounted for by salt, soot, and sulfate. The salt most likely originated from oil-field brines that were ejected from the wells along with oil. These oil brines are formed from salt deposits in certain layers of the wells, which penetrate the Jurassic strata typically 1.5-3.0 km below the surface (Cahalan, 1992). Crude oil may contain nickel, vanadium, arsenic, and other metals in small or trace quantities, many of which are removed during refining. General characteristics of Kuwait crude oil appear in Table 1.1.

Table 1.1
Characteristics of Kuwait Crude Oil
Component Amount
Sulfur 2.44 %w
Nitrogen 0.14 %w
Nickel 7.7 ppm
Vanadium 28 ppm
Low-boiling point naphtha < 205oC 22.7 %w
High-boiling point naphtha > 205oC 77.3 %w
Aromatics 23.3 %w
Paraffins 20.9 %w
Insolubles 3.5 %w
SOURCE: Ferek, 1992.

NOTES: %w = percent by weight; ppm = parts per million.

Although the composite smoke plumes in satellite photographs appeared black, individual fires emitted plumes with a variety of colors and densities. About 25 percent of the plumes were white or light gray and others black or dark gray. The densest black plumes were from large ignited pools of oil on the ground. The white plumes contained higher concentrations of NaCl and CaCl crystals (many in the fine fraction (<2.5 µm) range) and of SO2 than the black plumes (Stevens, 1993).

Initial calculations of worst-case effects of the burning oil wells assumed about a 10 percent soot content (Bakan, 1991). This level of soot could alter not only the region's climate but also perhaps that of the entire globe (Bakan, 1991). Actual measurements showed a much lower concentration. The soot concentration was only 0.45 percent of the fuel burned, or 0.53 percent of the total carbon emitted (Ferek, 1992). This measurement is over an order of magnitude lower than the 10-percent estimate (Bakan, 1991).

Fear of large emissions of SO2 caused a local health-hazard concern, as well as concern that long-range transport of the smoke might produce acid rain globally. Partly because of the early and successful extinction of the oil fires, these concerns did not materialize. Measurements of the removal rates of SO2 and NOx after emission show that, on average, they were 6 percent per hour for SO2 and 22 percent per hour for NOx. A likely explanation for the rapid removal of SO2 and NOx is through reactions with coarse-mode sand dust. The large concentrations of soil dust probably scavenged SO2 and NOx from the plumes (Hobbs and Radke, 1992).

Natural winds tended to move smoke plumes toward the southeast, but occasionally a plume moved toward the northeast, merging with a southeast-bound plume and creating a composite or "superplume" that continued southeast.

Through the summer, there were local inversion episodes, and the plumes could not rise. Such inversions may have reinforced themselves by keeping the smoke optically thicker than it would otherwise be, thus leading to more solar absorption in the atmosphere and cooling below. This effect helped to maintain the confinement (Cahalan, 1992). The smoke plumes were never observed to rise above 6 km, even after traveling 1600 km in 48 hours. In general, the plumes were depleted and dispersed for several thousand miles downwind over a period of several weeks, but always well below the base of the stratosphere (approximately 13 km). Such altitudes are low enough to prevent global distributions of the smoke-plume contaminants. Therefore, the effects of the contaminants were limited to the region, although the diluted plumes were detected worldwide.

Sometimes local atmospheric inversion patterns returned high concentrations of pollutants to ground level (Hobbs and Radke, 1992). Smoke particles in the plume coagulated as they were transported downwind and became hydrophilic as a result of being coated with sulfate. This substantially increased their removal efficiency by clouds and rain. Within a few kilometers of the fires, the circulation generated by the heat of combustion carried the pollutants aloft, with relatively clean air flowing into the fire region from all sides. Heavy particles and unburned oil droplets, large enough to be visible to the naked eye, fell through this clean air and caused a visible "plume" that was largely not associated with the combustion products.

The smoke plumes were 15 to 150 km wide from the source up to distances of 1000 km from the fires. The plumes absorbed sunlight, making the ground surface colder and darker. Daytime temperatures below the plumes were reportedly 10o C or more below normal under the optically thickest part of the plume, within 100 km of the source (Cahalan, 1992). Despite the darkness, the air at ground level retained relatively low pollution levels, except during the inversion episodes. Satellite images indicated that frequent intense dust storms mixed with the smoke in the same region either at the same level or in stratified layers (Limaye, 1992).

[1]Soot: A conglomeration of particles impregnated with tar, formed by the incomplete combustion of carbonaceous material.
Next Section