TESTIMONY OF KEVIN YESKEY, MD

Good morning. I am Dr. Kevin Yeskey, Acting Director of the Emergency and Environmental Health Services Division of the National Center for Environmental Health (NCEH), one of the Centers within the Centers for Disease Control and Prevention (CDC). I would like to thank the Committee for inviting me here today to discuss CDC's Congressionally mandated public health oversight role to the Department of Defense chemical demilitarization program and specifically our investigation of the agent release at Tooele, Utah. CDC provides independent oversight of Department of Defense's lethal chemical agent disposal facilities as a critical element of our role as a public health service agency. Furthermore, I would like to especially thank you, Chairman Hunter, for your interest in ensuring the safe disposal of our nation's chemical weapons stockpile.

As an extension of the U.S. Department of Health and Human Services, CDC reviews all Department of Defense (DOD) plans for destroying the nation's stockpile of chemical weapons. CDC evaluates any selected technology to ensure protection of the health and safety of the workers and the surrounding community. Once the technology is selected and implemented by DOD, CDC reviews facility design provisions and operating procedures aimed at protecting the workforce and surrounding communities. During operations, CDC continues to monitor and inspect the demilitarization facilities. Since air monitoring is a critical element in detecting any possible agent release incidents, CDC regularly examines air monitoring methods and strategies, as well as the quality of the data from these systems. Finally, to ensure local medical and emergency personnel are prepared to respond to an incident involving chemical agent, CDC provides consultation services to medical and emergency personnel near storage sites and reviews their capabilities to respond to lethal chemical agent-related emergencies. In its Congressionally mandated public health oversight capacity, CDC responded to the agent release at the Tooele Chemical Demilitarization Facility to ensure that public health was protected in this particular case and to help minimize risk of further releases. This response is in addition to CDC's regular and unannounced inspection visits to the TCDF prior to May 8 that also are part of CDC's role to protect the health of the public in communities near chemical weapons demilitarization facilities.

The Army notified CDC on May 9, 2000 of a chemical agent release from the common stack of the incineration plant at TOCDF. CDC immediately responded by dispatching a team to conduct an independent investigation of the incident. As part of the investigation, an initial assessment of the impact of the agent release on public health was conducted to ensure there was no danger to the workers or the surrounding community. CDC found the release was at a non-harmful exposure level, and consequently, no adverse health effects are anticipated. CDC then immediately followed our initial assessment with an in-depth analysis of the incident which is captured in our technical investigation report. Our report concludes that the TOCDF employees and the surrounding community were not exposed to a level of agent that would cause any short- or long-term adverse health effects from the release. However, CDC found several issues that simultaneously contributed to the release of sarin (GB) which should be addressed prior to resuming plant operations. In order to protect workers and the surrounding communities, CDC offers fifteen recommendations to help reduce the probability of similar events and enhance the utility of existing systems.

The overall objective of CDC's investigation was to provide an understanding of the operational events surrounding the release including the period after the release, evaluate the accuracy and validity of the air monitoring data, evaluate the potential impact on public health and provide recommendations as to how to prevent similar incidences in the future.

The CDC first reviewed the engineering data to develop a chronology of the operational events before, during, and after the agent release. In addition to reviewing the data, CDC also interviewed key plant personnel and Army representatives to discuss the operational events.

On May 8, 2000, the staff at TOCDF terminated M56 rocket burns (incinerating the chemical agent in the rockets containing GB) at approximately 4:00 pm because the deactivation furnace system lower tipping gate (this is the gate that drops the waste into the furnace) failed to close properly. At 8:10 pm the maintenance staff began the procedure to repair the gate and to clean the feed chute. The staff also cleaned the liquid agent strainer and placed the used strainer sock containing about one pound of waste contaminated with agent GB on the upper sliding gate of the feed chute. This waste is believed to be the major source of agent involved in the release. The maintenance staff completed their activities at 9:30 pm. High airflow rates through the pollution abatement system resulted in the scrubber fluid being drawn through the air flow meter into the demister. This caused the exhaust gas flow meter in the deactivation furnace pollution abatement system (the final air cleaning provision) to fail, causing an automatic shut-down of the burners. The plant operators attempted to relight the burners at 11:26 pm at which time the air monitors in the common stack went into alarm. The control room supervisors acknowledged the alarm, however, they allowed the control room operator to continue recovery efforts because they believed that the furnace was free of agent. However, after the pollution abatement system went into alarm at 11:41 pm, the supervisor directed the operator to isolate the furnace and the pollution abatement system.

Although the confirmational air monitoring samples had not yet been analyzed, the control room supervisor, still believing no agent was present, directed the operator to purge the deactivation furnace and relight the burners to maintain the after-burner temperature above 1000 degrees (the operating temperature required by the Environmental Protection Agency). During the second attempt to relight the burners at 12:28 am the air monitors in the common exhaust stack again went into alarm, although with lower agent level readings than in the first episode. The supervisor again directed the operator to isolate the furnace and pollution abatement system at 12:32 am.

In summary, the low deactivation furnace temperature from the loss of the burners and high air flow rates allowed a trace amount of GB agent to be released through the common stack. This release ultimately occurred as a result of a combination of unusual maintenance procedures, failure to recognize live agent in the system, and malfunctioning equipment.

The air quality is constantly measured both inside and outside the facility. These monitors detect any variation in the air quality and are specifically set to alarm if any agent is present.

CDC reviewed the operational status of both in-plant and outside air monitoring systems before, during, and after the release. The monitoring data, quality control data, and appropriateness of responses and activities of monitoring personnel were evaluated. The overall accuracy and validity of the monitoring data were carefully determined. The documentation, interpretation, and utilization of the monitoring results were examined. CDC also compared the air monitoring data and operational events to evaluate the chronological, spatial, and operational compatibility and consistency of these data. After careful review of the air monitoring data, CDC determined that all monitors functioned as designed in detecting agent and alerting the workers. Furthermore, the air monitoring data is consistent with the operational events leading to the release.

The most critical element of our investigation was to assess the potential public health impact to determine if the estimated amount of GB released was at a high enough level to harm the plant workers or the citizens in the surrounding community.

GB is a volatile chemical warfare agent, which makes it primarily an inhalation hazard. It is toxicologically related to organophosphate insecticides, which produce adverse effects on the nervous system by inhibiting cholinesterase (ChE) enzyme activity. The route of exposure of GB can include the eyes, respiratory tract, and skin. In high concentrations of GB, exposure can lead to central nervous system failure and ultimately death.

CDC collected the air monitoring data, real-time meteorologic data, and the operational data then applied this information using an accepted Environmental Protection Agency (EPA)-approved dispersion model to characterize the agent plume to evaluate the potential exposure to workers and the general public. CDC used this dispersion model based on the highest concentration taken from the air monitoring data over a 30-minute period, which is approximately equal to the duration of the two-phase event. Using this calculation, the total agent release based on CDC's conservative model and exposure analysis would be equivalent to 46 mg over the entire release. CDC believes this to be a substantial overestimation of the actual release amount, and it is therefore, a very conservative figure to use for the examination of human health implications.

Based upon the results of this modeling and a 30-minute duration of the release, the model predicted the maximum possible exposure to GB at ground level was less than 1/10 of 1% (or 1/1000th) of the exposure one would receive if exposed to the general population limit (GPL), a limit considered to have no adverse health impacts. The GPL set by the Department of Health and Human Services for GB is .000003 mg/m³.

To evaluate the maximum possible impact on the health of workers at TOCDF during the GB release, the model was again run using worst-case parameters. Although local meteorologic data indicated that downwash conditions (wind speed and direction conditions that result in rapid movement of stack gases to ground level near the plant) probably would not have occurred during the release, the downwash option was used to ensure a worst case scenario. Even with this most conservative approach, estimated maximum agent concentrations were well below the established concentration of GB for the 8-hour occupational exposure limits (OEL), a limit considered to have no adverse health impacts for workers. The OEL set by the Department of Human Health Services for GB is .0001 mg/m³. Considering potential exposure for the entire release period, the maximum possible exposure was less than 1% of the allowable concentration. In addition, since the TOCDF workers masked early during the release their exposure was far less than the amount used for this analysis. In conclusion CDC determined there was no adverse health impact on the workers or the surrounding community based on the workers responding to the release by wearing their masks, no reports of GB exposure symptoms, and results of the conservative model.

Based on our investigation of the incident, CDC has offered several recommendations to further improve the public health and safety surrounding the plant operations, to improve the performance and utility of the monitoring system, and improve overall community safety. The details of our recommendations may be found in CDC's Technical Report.

The incidents leading to the release were so called "non-normal" procedures conducted during abnormal plant conditions. The process for implementing an "non-normal" procedure should be reviewed and additional training provided as needed to ensure the personnel have the capability to conduct the "non-normal" procedure. Additionally there should be an evaluation of all "non-normal" procedures to ensure they can be performed safely.

Once the air monitoring system goes into alarm, the most qualified control room operators should be controlling operations. Furthermore, even when plant personnel do not believe agent is in the system, when an agent-related alarm activates, the alarm should be considered valid until it can be proven false. Once the alarm is activated, the control room supervisor should make all decisions consistent with guidelines from the Contingency Procedure for Agent Detected in the Stack manual. The contingency procedures should be reviewed and updated to reflect this incident.

Prior to the restart of plant agent operations, CDC recommends a careful review and evaluation of the entire system. Intensified investigation is needed to identify and eliminate the source of false stack alarms. All air monitors should be challenged and reviewed for quality assurance. An engineering evaluation of the location and operation of the flow meter should be conducted. The deactivation furnace feed chute and related gates should be evaluated to optimize proper functioning. Finally, all components of the deactivation furnace system including the pollution abatement systems need to be systematically examined to ensure proper functioning.

Since there is some question regarding the notification policy during an agent release incident and the current policy is unclear, the procedure to ensure that Tooele County is informed in a timely manner of potential and confirmed agent released should be reviewed and clarified. In addition, the notification process for other organizations with potential involvement following a release should be reviewed.

CDC will continue to monitor DOD's response to our recommendations and will assist in any way possible to ensure these recommendations are addressed to the intent and detail necessary to minimize the potential for any future chemical agent release to the plant workers and the surrounding community. CDC has participated in public meetings and will continue to attend public meetings in the State of Utah to address questions and comments from the community members regarding our report. CDC believes our independent investigation was diligent and when combined with the State of Utah's report provides a comprehensive analysis of the incident, addressing events leading to the release, the response afterwards, and the necessary corrective actions to prevent occurrence of future events. Additionally during the week of August 7, 2000, CDC conducted an on site review of the Army's implementation of our recommendations. Based on our final assessments of TOCDF and the Emergency Operations Center, CDC has determined that our recommendations are being addressed and implemented. CDC will continue to closely monitor the DOD as they implement their necessary actions to ensure that TOCDF workers and the surrounding community of TOCDF are protected from exposure to chemical agent. As a public health service agency, we believe that protecting the public's health is the number one priority in the destruction of these chemical weapons. We will continue to work with DoD, the State of Utah, and the local communities to demilitarize our nation's stockpiled lethal chemical weapons in the safest manner possible while reducing the inherent risk of exposure from continued storage.

Thank you for the opportunity to testify today. I am happy to answer any questions that you may have.