Workplace Chemical Monitoring - Occupational Health and Workplace Monitoring at Chemical Agent Disposal Facilities - page 11

general population limit values for VX and HD may be
revised downward as a result of a review of agent stan-
dards now being done by the Army (Ruetter et al.,
2000).

Another weakness of the airborne monitoring sys-
tem is the lack of real-time (< 10 seconds) agent detec-
tion. The committee has recommended that the Army
develop a real-time system that uses a measurement
technology independent of the gas chromatography
with flame photometric detector methods used by the
ACAMS and DAAMS systems (NRC, 1994b). To date,
the Army's attempts to develop and demonstrate such
a system have not been successful (NRC, 1999a). New
interest in chemical agent detection as a key compo-
nent of antiterrorism activities has spurred government
and commercial activities focused on developing better
airborne agent sensors (IOM, 1999). The committee
has previously urged the Army to continue to monitor
technological advances and to consider implementing
any that are appropriate for chemical agent disposal
facilities (NRC, 1999a).
The recurrent problem with the airborne monitoring
system is false positives—which occur when an
ACAMS alarm goes off but the presence of agent can-
not be confirmed by later DAAMS tube analysis. The
resulting tendency to discount alarms and to proceed as
if agent were not present was graphically illustrated by
an incident involving a minor release of GB at TOCDF
in May 2000 (CDC, 2000).

Assessment

ACAMS monitors, the principal agent quantifica-
tion instruments throughout any disposal facility, alarm
when a preset level of agent (usually 20 percent of the
relevant control level for that location) has been ex-
ceeded. Signals much larger than the preset response
level may saturate the signal processing algorithm, and
because the duty cycle of an ACAMS monitor is less
than 100 percent (i.e., samples are collected only dur-
ing part of the duty cycle), confirmation that agent ac-
tually caused the signal depends on the analysis of
DAAMS tubes at the same location. This analysis can
give only the average agent concentration over the
DAAMS tube's total exposure period, although it is
reasonable to assume that most agent accumulation
occurred during the period when the associated
ACAMS monitor was in an alarm mode. A DAAMS
tube without associated ACAMS monitoring can only
indicate the average agent level between the time of
deployment and the time of collection for analysis.
One potential weakness of the current airborne agent
monitoring program is that ACAMS monitors are typi-
cally set to detect only the single agent currently being
processed. Because individual ACAMS monitors can
detect only one agent at a time, multiagent monitoring
requires different ACAMS monitors for each agent.
Moreover, only the agent currently being processed is
usually addressed during routine DAAMS tube analy-
sis. Thus, an accidental release of a chemical agent not
being currently processed might go undetected. For
instance, leaks from a mislabeled munition or a projec-
tile filled with an unexpected or mislabeled agent in
nominally agent-free areas would be missed, and con-
tamination of the downstream processing area by the
unexpected agent could go undetected. This issue was
raised in the committee's 1994 monitoring report
(NRC, 1994c), but the Army has judged the probability
of “mislabeling” to be low enough that routine deploy-
ment of ACAMS monitors for multiagent detection is
currently restricted to the plant-air carbon filtration sys-
tem. Recent briefings on JACADS closure planning
have indicated that multiagent monitoring will be
implemented during closure operations (U.S. Army,
1999c).

MONITORING AGENT IN LIQUIDS AND SOLIDS

The primary processing requirement for monitoring
agent in liquid media is to analyze the hydrolysates
produced at Aberdeen and Newport to ensure that the
mustard or VX has been thoroughly destroyed before
proceeding to the secondary treatment step—biodegra-
dation at Aberdeen and SCWO at Newport. It will also
be necessary to ensure that no significant amount of
agent is present in any process stream that is ready for
discharge.
Because of the relatively low solubility of VX and
several of its hydrolysis products in water and the salt-
ing-out effect of the high ion concentrations involved
in caustic hydrolysis, there will be two liquid phases
present during and after VX hydrolysis. Because the
hydrolysate must be certified to be free of agent (de-
fined as a destruction and removal efficiency [DRE] of
99.9999 percent) before it goes to the high-tempera-
ture, high-pressure SCWO reactor, both phases will
have to be included in the analysis. VX, being lipo-
philic, is likely to partition selectively into the less
dense oily phase, which constitutes about 5 percent of
the hydrolysis mixture.