Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Findings and Recommendations Finding 1. To perform the statistical analyses presented in this report, the Stockpile Committee, with assistance from the Army, expanded the parameters measured or inferred from existing databases on the chemical stock- pile condition. These additional data fields may be in- structive and useful to the Army. The Minitab software that was used is an example of the statistics programs used by industry for analysis. Statistical software of this type can also support an overall program to im- prove performance, including safety performance. Recommendation 1. The Army should collect data on degraded agent and other munition anomalies in a man- ner that provides for maintaining the parameters in the data matrix format developed by the committee or in a similar format. The Army should establish additional data fields, such as documenting inspection activities in which no leaks are found. The Army should also consider the use of a statistical program such as Minitab or an equivalent program to assist in the overall quality of stockpile analyses and safety improvement. Finding 2. The data on leaking munitions and contain- ers in the chemical agent stockpile from 1992 onward were obtained using a consistent set of protocols. How- ever, the information required for doing certain trend analyses, such as the age of the munition when leaking was detected and agent subtype, is not readily discern- ible and often not available. 47 Recommendation 2. The Army should continue using its current monitoring protocols. It should set up a pro- gram for acquiring data, maintaining it in a database, and entering it into a statistical software package. Staff will have to be given adequate training in statistical analysis. Finding 3. The sampling program as currently con- ducted is biased toward munitions and containers hold- ing GB agent. What data there are indicate that VX leaks are relatively rare. However, one concern is that the stabilizer in VX munitions may be dropping to criti- cal levels. Recommendation 3. The Army should verify that VX leaks are much less frequent than GB leaks by monitor- ing VX munitions and containers more closely. The sta- tus of stabilizers in VX munitions should be ascertained. Finding 4. The database on leaking chemical muni- tions and containers developed by the Army and ex- panded to include additional fields by the committee is useful for predicting leaker trends, but not for predict- ing trends for other anomalies such as gels and foam- ing or frothing rounds. The search for causes and ef- fects of leaks in munitions and for remedial actions to address them could similarly be applied to other stock- pile anomalies. End-of-campaign reports are one source of information on anomalies other than leaks.
48 EFFECTS OF DEGRADED AGENT AND MUNITIONS ANOMALIES ON CHEMICAL STOCKPILE DISPOSAL OPERATIONS Recommendation 4. The Army should carefully ana- lyze the data in end-of-campaign reports from chemi- cal stockpile disposal operations and any other avail- able information to determine if a predictive technique can be developed for anomalies other than leaking munitions for example, gelled agents. Finding 5. The agent decomposition mechanisms de- scribed in Chapter 2 include a number of reactions with positive temperature coefficients (EaC~ > 0~. Corrosion reaction rates are also expected to increase with tem- perature, implying that the occurrence of leaks may increase at higher ambient temperatures. The available data did not validate the hypothesis of a temperature dependence. Recommendation 5. The Army should consider instru- menting some existing storage facilities for the con- tinuous monitoring of temperature, particularly those facilities where materials most susceptible to degrada- tion are stored namely, munitions with problematic GB subtype lot numbers. The temperature data could then be correlated with leaker rates to identify any tem- perature dependence. If such a dependence is shown to exist, the Army could consider mitigating measures, including temperature control in the igloos and the re- evaluation of processing sequences for the various GB munitions. Finding 6. Pressurized hydrogen gas has formed from mustard agent degradation in some munitions and con- tainers. This can present a significant risk to workers during disposal operations. Recommendation 6. Special safety measures are re- quired during disposal of mustard agent munitions and containers to minimize the risk of hydrogen gas ex- ploding. Finding 7. The degradation mechanisms for GB agent outlined in Chapter 2 imply that degradation can pro- ceed by autocatalysis once inhibitor is exhausted. Au- tocatalytic reactions are characterized by a slow induc- tion period followed by a rapid rise in reaction rate. While the available data do not support an acceleration in leaker frequency, the current state of reaction within GB rockets could still be on a relatively flat portion of the reaction rate curve, indicating that most prob- ably only the usual reactions with stabilizer present are occurring. However, there is a possibility that the rapidly rising portion of the curve that would be char- acteristic of an autocatalytic reaction is being ap- proached. Recommendation 7. The Army should maintain an aggressive monitoring and data analysis program, par- ticularly in the case of GB munitions and containers, in order to identify any significant upward trend in the rate of leakers as early as possible. Expeditious dis- posal of GB munitions can help to avoid an encounter with the rapidly rising portion of the curve that would characterize an autocatalytic reaction resulting from GB degradation. Finding Ha. Corrosion has been observed on valves and plugs of GB ton containers. This corrosion could cause agent leaks or spills. Finding Xb. The corrosion that has been observed in burster tubes of 105- and 155-mm projectiles was caused by agent leaking into the burster as a result of incorrectly brazed joints. In such cases, corrosion is the result of a leak, not the source of the leak. In time, this corrosion may result in the release of agent. Munitions in this condition may be in the stockpile. Recommendation X. GB munitions and GB in con- tainers should be destroyed as soon as possible. Finding 9. The frequency of monitoring M55 rockets is based, in part, on observed leakage frequencies. Given that the number of GB M55 rocket leakers is greater than the number of VX M55 rocket leakers, monitoring by the Army focuses on the former. Recommendation 9. In view of the toxicity of VX, the Army's monitoring protocols should be reviewed to increase confidence that the frequency of leakers and the mechanisms associated with VX degradation are well understood. Finding 10. Continuing degradation of M55 rocket pro- pellant, even when accelerated by agent contamination, is at present only a minor contributor to storage risk. However, the expeditious destruction of GB M55 rock- ets is an effective way to reduce risk to the public since accidental ignition due to lightning or earthquakes re- mains the largest contributor to storage risk. Significant delays in the destruction of GB M55 rockets could ne- cessitate monitoring the condition and stability of the
FINDINGS AND RECOMMENDATIONS propellant more intensely as agent continues to degrade and perhaps contribute to propellant instability. Recommendation 10. If destruction of the GB M55 rockets is significantly delayed, the condition and sta- bility of the M28 propellant in these rockets should be closely monitored. Finding 11. It may be possible to prioritize the destruc- tion of M55 rockets by GB subtype and thereby reduce the likelihood of future leaks. Recommendation 11. If the committee's analyses of the role of GB subtypes in leaker development can be substantiated by the Army, consideration should be given to prioritizing the order of destruction of M55 GB rockets by GB subtype, as long as this process does 49 not adversely impact the overall M55 GB rocket de- struction schedule. Finding 12. Currently available records, such as end- of-campaign reports, do not allow for directly linking system and equipment downtimes to the processing of anomalous munitions. The early communication of anomaly detection by direct contacts between all sites and by means of the program's lessons-learned data- base is important for all concerned. Recommendation 12. The Army should assemble corrective actions from lessons learned to address anomalous conditions of all types, including the modi- fications of process rates, sequences, schedules, and procedures, and it should ensure the prompt, direct, and adequate distribution of this information to all disposal sites.
