MIR FIRE DURING NASA 4: TOXICOLOGICAL ANALYSIS OF
SAMPLES RETURNED ON THE SOYUZ
1. Introduction
This report describes the analytical results from 4 grab sample containers (GSCs) and 20 formaldehyde badges returned from Mir on a Soyuz spacecraft in early March, 1997. The samples were returned early because most were taken after a fire occurred in the supplemental oxygen generation system in the Kvant module of Mir at about 2245 on February 23, 1997. The goal of analyzing these samples was to toxicologically assess the air quality after the fire to determine if the crew could have been exposed to an excess of potentially harmful pollutants. The crew was able to don respiratory protection within 90 seconds of the start of the fire, so the pollutants found in the air samples are not directly related to the crew’s exposure to pollutants. Furthermore, the samples were taken after several hours of atmospheric scrubbing had occurred; therefore, they do not provide the data necessary to estimate the pollutant levels shortly after the fire. The samples do provide data on the spatial distribution of pollutants and the rate of scrubbing from the atmosphere.
2. Analytical Methods
The GSCs were analyzed by gas chromatography (GC) for methane, carbon monoxide, hydrogen, and carbon dioxide. A targeted list of 78 volatile organic compounds was measured by GC/mass spectrometry (MS) along with many additional nontarget compounds (table 1). Two of the target compounds required special analytical techniques to quantify the concentrations. The methods are described in applicable standard operating procedures and have been reviewed and accepted by the Environments Subgroup of Medical Working Group 8.
Toxicological analysis (table 2) was performed by comparing the measured concentrations of each pollutant to the 7-day spacecraft maximum allowable concentration (SMAC) to give a T value. The method is described in detail in JSC 20584. The crew was probably not exposed to fire-generated pollutants for 7 days; however, the 7-day values were used because many compounds do not have a 1-hour or 24-hour SMAC assigned.
Twelve formaldehyde badges were exposed about 8 hours after the fire, one was exposed during the fire, and three more had been exposed 3-4 days before the fire (table 3). These badges, along with 2 positive and 2 negative controls, were analyzed by the chromotrophic-acid method. By agreement with the Russians, this method is considered for reference purposes only at this time.
DISTRIBUTION LIMITED TO PERSONNEL IN U.S. AND RUSSIAN HUMAN SPACEFLIGHT PROGRAMS
3. Results
The analytical results of the GSCs are summarized in Table 1. The GSC sample which was labeled as taken, but not actually taken was designated a “trip control” for analytical purposes. The concentrations of volatile organic compounds in this GSC were very low with no more than traces found for a few compounds. This sample will not be discussed further.
The 3 remaining samples provided interesting data on spatial differences (Kvant vs. Core module) and temporal differences (0130 vs. 0200 samples in Core module). We will assume that the timing of samples was accidentally reversed as indicated in a message from Dr. Linenger; therefore, sample AA01585 will be referred to as the 0200 sample and AA01587 as the 0130 sample.
Temporal and spatial differences can be calculated by focusing on compounds that would be expected from combustion and are normally present only in low concentrations. The tables below illustrate this point.
CompoundSpatial Difference Temporal Difference
Percent Difference, Kvant - CorePercent Difference, 0130 - 0200
at 0130in Core module
benzene 20 % 29 %
styrene 47 % 21 %
2-methyl-2-propenenitrile 29 % 20 %
2-methylpropanenitrile 25 % 33 %
carbon monoxide 18 % 28 %
This table suggests that even 3 hours after the fire there was still a slight (18 to 47 %) localization of likely combustion products in the Kvant. It also suggests that scrubbing of the atmosphere from 2 ¾ hours to 3 ¼ hours after the fire resulted in a decrease in fire-associated pollutants of 20 to 33 %.
The formaldehyde data (table 3) indicate that several hours after the fire, there was no evidence that formaldehyde was increased in the atmosphere as a result of the fire. All area samples were at or below the formaldehyde 7-day SMAC of 0.04 ppm. One of the badges worn by a cosmonaut showed a concentration of 0.08 ppm; however, other crewmembers’ badges showed a range of <0.01 to 0.05 ppm. The reason for the higher badge value on one cosmonaut is unknown.
Quality control and quality assurance data were as follows. Formaldehyde positive controls (including the lab control) were 75%, 96% and 80%. Recovery of surrogate standards during the GC/MS analysis was as follows:
surrogate standardsample designation number
AA01584AA01585AA01587
acetone-C1389 %92 %98 %
fluorobenzene-D590 %89 %93 %
chlorobenzene-D586 %82 %83 %
The GC/MS analysis passed daily criteria for ion abundance of bromofluorobenzene and daily calibration checks. The table below shows the variations of the daily standard during GC analysis of the samples:
daily standardconcentration (ppm)range of percent deviation
carbon monoxide 5.23-8 to +11
carbon dioxide2000-2 to +6
methane 501-1 to +4
hydrogen 202-9 to -3
4. Discussion and Conclusions
Inspection of the T values at the bottom of table 2 shows relatively large numbers that decreased with time (4.3 at 0130 to 3.4 at 0200 in the core module) after the fire or distance from the source of the fire (5.2 in the Kvant Module at 0130 to 4.3 in the Core Module at 0130). The primary components of the T values (>0.1 unit contribution) were benzene, furan, 2-methyl-2-propenenitrile, carbon monoxide, methane, and carbon dioxide. All compounds except methane can be linked to pyrolysis of materials. Likewise, except for methane, all would be scrubbed from Mir air rather rapidly, so it is reasonable to conclude that sometime later when the crew removed their breathing masks, they were not exposed to harmful concentrations of pollutants.
Even with only 3 canister samples returned, we showed that the pollutants had not reached spatial uniformity and that those pollutants associated with the fire seemed to be rapidly scrubbed from the air. There was no evidence that the fire generated excess formaldehyde.
SD2-97-513
TO:DistributionApr 10, 1997
FROM:SD2/John T. James
SUBJECT:Toxicological Assessment of Air Samples Taken after the Oxygen- generator Fire on Mir
The enclosed report describes the analytical results from canister and badge samples of air obtained several hours after the fire that occurred in the oxygen generation system in Kvant on February 23, 1997. The analysis of canister samples taken 3 hours after the fire showed several compounds that would be expected from a fire; however, their concentrations were below levels that could harm the crew. The concentrations may have been much higher immediately after the fire, but the crew donned respiratory protection very quickly after the fire began. Analysis of formaldehyde badges gave no evidence that this irritant was produced by the fire.
The concentrations of fire-associated compounds in the 3 canister samples indicated that 3 hours after the fire these pollutants were still slightly concentrated in the Kvant module. The data also suggest that the compounds were being rapidly scrubbed from the atmosphere. Additional data will be available when the remaining canisters are returned aboard STS-84.
John T. James, Ph.D.
NASA Life Sciences Data Archive (
Investigator Name: John T. James, Ph.D.
Mission (Payload): NASA 4
Experiment Title (ID) Toxicological Assessment of Airborne Volatile Organic Compounds (MSD007)
File Name/Inventory ID: JMMIRMSD007_177.doc
NASA Life Sciences Data Archive (
Investigator Name: John T. James, Ph.D.
Mission (Payload): NASA 2, 3, 4, 5, 6, 7
Experiment Title (ID): Toxicological Assessment of Airborne Volatile Organic Compounds (MSD007)
File Name/Inventory ID: