This article was downloaded by: [152.180.134.

66]
On: 21 November 2014, At: 13:50
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Journal of Occupational and Environmental Hygiene

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/uoeh20

Exposure of Unsuspecting Workers to Deadly

Atmospheres in Below-ground Confined Spaces and
Investigation of Related Whole-Air Sample Composition
Using Adsorption Gas Chromatography
a b a c
Philip A. Smith , Bonnie Lockhart , Brett W. Besser & Michael A.R. Michalski
a
Health Response Team, U. S. Department of Labor, Occupational Safety and Health
Administration (OSHA), Sandy, Utah
b
St. Louis Area Office, U.S. Department of Labor, Occupational Safety and Health
Administration, St. Louis, Missouri (currently with Ameren Corporation, St. Louis, Missouri)
c
Industrial Hygiene Chemistry Laboratory, U.S. Department of Labor, Occupational Safety
and Health Administration (OSHA), Sandy, Utah
Accepted author version posted online: 23 May 2014.Published online: 31 Oct 2014.

To cite this article: Philip A. Smith, Bonnie Lockhart, Brett W. Besser & Michael A.R. Michalski (2014) Exposure of
Unsuspecting Workers to Deadly Atmospheres in Below-ground Confined Spaces and Investigation of Related Whole-Air Sample
Composition Using Adsorption Gas Chromatography, Journal of Occupational and Environmental Hygiene, 11:12, 800-808, DOI:
10.1080/15459624.2014.922687

To link to this article: http://dx.doi.org/10.1080/15459624.2014.922687

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Journal of Occupational and Environmental Hygiene, 11: 800–808
ISSN: 1545-9624 print / 1545-9632 online
DOI: 10.1080/15459624.2014.922687

Exposure of Unsuspecting Workers to Deadly Atmospheres

in Below-ground Confined Spaces and Investigation of
Related Whole-Air Sample Composition Using Adsorption
Gas Chromatography
Philip A. Smith,1 Bonnie Lockhart,2 Brett W. Besser,1
and Michael A.R. Michalski3
1
Health Response Team, U. S. Department of Labor, Occupational Safety and Health Administration
(OSHA), Sandy, Utah
2
St. Louis Area Office, U.S. Department of Labor, Occupational Safety and Health Administration, St.
Downloaded by [152.180.134.66] at 13:51 21 November 2014

Louis, Missouri (currently with Ameren Corporation, St. Louis, Missouri)

3
Industrial Hygiene Chemistry Laboratory, U.S. Department of Labor, Occupational Safety and Health
Administration (OSHA), Sandy, Utah

Hazardous atmospheres in confined spaces may be obvious
when a source of air contamination or oxygen (O2 ) deficiency is
recognized. Such is often the case in general industry settings,
especially with work processes which create hazardous atmo-
spheres that may be anticipated. Hazards present in active
sewers are also well recognized; but the possibility that O2
deficiency or high airborne contaminant concentrations may
exist in new construction sewers or storm drains has been re-
peatedly ignored with deadly results. Low O2 and high carbon
dioxide (CO2 ) concentrations may exist in new construction
manholes that have not yet been connected to an active sewer
or drain system, and these concentrations have been shown to
vary over time. A recent incident is described where workers
repeatedly entered such a confined space without incident,
but subsequent entry resulted in a fatality and a near-miss
for a co-worker rescuer. Additional cases are discussed, with
an emphasis placed on elevated CO2 concentrations as a
causative factor. A description is provided for the adsorptive
gas chromatography whole-air analysis methodology used to
quantitatively determine atmospheric conditions present at
this type of fatality site or others after an incident, and for
the gas chromatography-mass spectrometry method used to
provide confirmation of analyte identity with high certainty.
Many types of confined spaces may be encountered in addition
to the underground varieties discussed, and many possible
atmospheric hazards are possible. The definitive whole-air
analysis approach described here may be of use and should
be considered to investigate many confined space fatality and
near-miss cases, and to better understand the causes of danger-
ous atmosphere conditions that may arise in confined spaces.
Keywords Confined space, carbon dioxide, new-construction
manhole, whole-air composition analysis
Address correspondence to: Philip A. Smith, OSHA Health Re-
sponse Team, Salt Lake Technical Center, 8660 S. Sandy Parkway,
Sandy, UT 84070; e-mail: Smith.Philip.A@dol.gov
INTRODUCTION
Well-Understood Confined Space Hazards
The hazards of confined spaces in general industry, includ-
ing oxygen (O2 ) deficiency have been recognized for years,
and are the basis for the U.S. Occupational Safety and Health
Administration (OSHA) regulation titled “Permit Required
Confined Spaces” which was promulgated in 1993.(1) A spe-
cific and detailed OSHA rule to regulate confined space entry
in construction was proposed in 2007, but has not yet been
finalized. However, underground construction regulations may
pertain,(2) and a training requirement for confined spaces is
currently found within the OSHA construction standards. Em-
ployers in construction are to “instruct all employees who
enter into confined spaces about the nature of the hazard, the
necessary precautions to be taken, and in the use of protective
and emergency equipment required,” where a confined space is
defined as “any space having a limited means of egress, which
is subject to the accumulation of toxic or flammable contami-
nants or has an oxygen deficient atmosphere.”(3) Such confined
or enclosed spaces “include but are not limited to, storage
tanks, process vessels, bins, boilers, ventilation or exhaust
ducts, sewers, underground utility vaults, tunnels, pipelines,

800 Journal of Occupational and Environmental Hygiene December 2014

 and open top spaces more than 4 feet in depth such as pits,
tubs, vaults, and vessels.”
Another relevant OSHA standard exists for manholes and
unventilated vaults that contain telecommunication-related
equipment and lines. For employee entry into these spaces “the
internal atmosphere shall be tested for combustible gas and, ex-
cept when continuous forced ventilation is provided, the atmo-
sphere shall also be tested for oxygen deficiency. When unsafe
conditions are detected by testing or other means, the work area
shall be ventilated and otherwise be made safe before entry.”(4)
An operational sewer system is recognized as a permit
required confined space (PRCS), as the basic requirements
for both the general industry PRCS standard(1) and the con-
struction training standard(3) are met: a sewer system may
be bodily entered for an individual to perform work, has
restricted egress, is not intended to have continuous employee
occupancy, and may suddenly present hazards that are widely
recognized. The potential for inadequate O2 , or the presence of
high levels of hydrogen sulfide (H2 S), carbon monoxide (CO),
Downloaded by [152.180.134.66] at 13:51 21 November 2014
and flammable gases or vapors, and recommended precautions
for sewer system entry are outlined in Appendix E of the
general industry PRCS standard.(1)
A Persistent and Deadly Misconception
It is commonly assumed that entry into underground spaces
such as a new construction sewer manhole that has not been
connected to an active sewer, or a vault that is not connected
to an obvious potential source of dangerous gases does not
require PRCS precautions. This is incorrect, as the experi-
ence of OSHA and the U.S. National Institute for Occupa-
tional Safety and Health (NIOSH) confirms. The accumulation
of high levels of CO in such a new-construction confined
space was confirmed by NIOSH, after the use of explosives
to open trenches for piping,(5) and on three occasions the
OSHA Health Response Team has investigated fatalities which
occurred in new construction sewer manholes that had not
yet been connected to an active sewer system. As will be
demonstrated, even without an event such as underground
blasting, the presence of low O2 concentrations, often with
high CO2 concentrations, has repeatedly led to worker deaths
from entry into below-ground confined spaces.
Reference is made in the literature to fatalities occurring
under conditions where low O2 and high CO2 concentrations
were unexpectedly encountered going back hundreds of years.
Carter cites an epitaph found in a small English country church
graveyard written in 1812:
[Died while] venturing into a well at Marton when it
filled with carbonic acid gas, or fixed air. From this unhappy
accident, let others take warning not to venture into wells
without first trying whether a candle will burn. If the candle
burns to the bottom, they may be entered with safety. If it
goes out, human life cannot be sustained.(6)
Journal of Occupational and Environmental Hygiene
Tankard et al. described a triple-fatality incident in 1930(7)
that resulted when workers entered a deep manhole that was
receiving sub-soil drainage from wet land. Flame testing and
calcium carbonate precipitation were used after the unfortunate
incident to detect low O2 and high CO2 concentrations in that
manhole atmosphere.
Similar cases are described in the more recent literature.
On October 10, 1950 two men died in a new sewer manhole
in Winona, Minnesota, that had not yet been connected to an
existing sewer system.(8) Raschka and Uber described analyses
of air samples collected on six different days from that fatality
scene during a period spanning nearly two months. These
showed that O2 concentrations ranged from 3.0 to 12.0%, while
CO2 concentrations were between 2.0 and 3.5%. Michaelson
and Park described a fatality case that occurred in June 1952.(9)
Workers had repeatedly entered a 4’ diameter manhole that was
9’ deep over a period of approximately two weeks to open and
close a gate valve for a culinary water line in Minneapolis,
Minnesota. The manhole had been completed on May 26 to
“provide a water connection for a new building under construc-
tion.” On the morning of June 10 an employee of a plumbing
contractor had entered the manhole to turn on the water.
This employee “. . .returned to the surface with no evidence
of distress.” At shift end another employee “. . .entered the
manhole and turned off the water. When he did not come
up in the expected time, his partner on the surface looked
down and called to him. There was no reply and he saw his
co-worker slumped over in a sitting position.”(9) Attempts
to revive the unconscious worker were not successful and
two measurements in this manhole showed that O2 and CO2
concentrations at the bottom varied over time, ranging from
16.1 and 4.3% respectively on June 16, to 0 and 14.3% 14 days
later.
Manning et al. described a triple-fatality case in a confined
space drainage pit at a baseball field where original measure-
ments taken by the local power company (likely in error) found
a methane level of 15%. The first worker to die had lowered
himself into the pit with a chain. Another worker assumed the
first had suffered a medical condition and entered the pit to
attempt rescue. When the second worker collapsed in the pit
a third worker then entered the pit in a “desperate attempt to
rescue his friends.”(10) The medical examiner could find only
a faint trace of methane in the victim’s tissues, and with help
from local OSHA personnel it was found that O2 and CO2
concentrations were 3 and 40%, respectively, at the bottom of
the drainage pit.
The examples listed above document a recurring problem,
where unsuspecting workers enter a confined space below
ground with low O2 and high CO2 concentrations. Details for
somewhat similar confined space fatalities and near-misses
have been extracted from more recent information maintained
by OSHA for fatality cases, and are presented in Table I.
Details regarding two specific cases found in Table I (Kansas,
2001 and Missouri, 2011) are discussed below.
December 2014 801
 TABLE I. Partial Listing of Confined Space Fatality Cases and Near-Misses Investigated by OSHA Between
1993 and 2012 (Excludes Cases Clearly Involving Deaths in Active Sanitary Sewer Systems)
Fatalities/Near-
State Year/Month Work Process Misses Atmosphere
MO 2011/10 Working in new-construction 1/1 low O2 /high CO2
manholeA,B
NC 2011/06 Working in manholeB 2/0 low O2 C
TX 2009/04 Fell in manhole/rescue attemptB 1/1 low O2 C
IL 2007/06 Working in new underground water 2/1 low O2 C-
vaultB
LA 2007/08 Grouting in manhole 2/0 low O2 C
GA 2006/09 Retrieving laser equipmentB 1/0 low O2 C
TX 2005/07 Testing newly installed storm sewer 1/0 low O2 C
lineB
ND 2005/07 Completing installation of new 1/0 D
−
manholeB
WI 2005/06 Testing prior to hookup to live sewer 2/0 D,E
−
Downloaded by [152.180.134.66] at 13:51 21 November 2014

lineB
TN 2004/09 Working in new manhole 2/0 low O2 C
MO 2004/08 Grouting in manholeB 1/0 low O2 /high CO2
PA 2004/07 Entry into new manholeB 2/0 low O2 C
GA 2004/07 Opening valve in water vaultB 1/1 low O2 /high CO2
WI 2003/08 Opening valve in water vaultB 1/0 low O2 /high CO2
FL 2003/08 Leak repair in new manholeB 1/1 −D
NC 2003/03 Entry into water vault manholeB 1/0 low O2 C
TX 2002/07 Working inside manhole 1/0 −D
TX 2001/12 Checking grade in 24” water pipeB 1/0 −D
TX 2001/09 Entry into manhole 1/0 −D
KS 2001/08 Vacuum testing new sewer lineA,B 3/0 low O2 /high CO2
OH 2000/09 Working in new manholeB 2/0 low O2 /high CO2
CA 2000/05 Working in sump manholeB 2/0 low O2 C
NC 1999/10 Working in new manholeB 1/3 −D
CO 1999/9 Entry into 1-month old water line utility 1/F low O2 C
vaultB
IL 1996/08 Opening valve in water vaultB 1/0 low O2 /high CO2
MI 1993/09 Checking sight level in new sewerB 1/0 low O2 C
IL 1985/05 Entering new vertical shaft (20’ depth) B 1/5 low O2 /high CO2
A Details discussed in this manuscript.
B Incident information available precludes exposure to active sanitary sewer.
C No description of atmospheric testing other than for O
2
D No information available regarding atmospheric testing
EAutopsy indicated low blood [O ] through post-mortem measurement.
2
F Information on possible near-miss from attempted rescue not available.

O2 Deficiency in a New-Construction Sewer System,
Topeka, Kansas, 2001
In this case a new construction sewer line was under con-
struction for a subdivision in the Topeka area and it had not yet
been connected to an active sewer system. A worker entered
a manhole in this system at about 10 a.m. on August 28 to
place plugs into three 8” lines that entered this manhole, and
he exited with no apparent ill effects. During vacuum testing
the construction crew did not realize that a small crack existed
in one of the prefabricated concrete rings which had been used
to build the manhole vault, but a leak became apparent when
vacuum was applied. The crew had been having trouble with
the vacuum pump and hose used, and the workers incorrectly
assumed this was the cause of the vacuum testing problems.
From their experience with new construction manholes they
did not anticipate the possibility of O2 depletion. After un-
successfully attempting to draw a vacuum in the manhole for
about 45 min, a supervisor entered the manhole to verify that

802 Journal of Occupational and Environmental Hygiene December 2014

 TABLE II. Results for On-Site O2 and CO2 Measure- The vacuum-testing which appeared to have played a role in
ments at the 2001 Topeka, Kansas Incident Site creating the dangerous atmosphere that claimed three lives was
recreated a few weeks later. Prior to testing this the manhole
Elapsed was thoroughly ventilated and the O2 concentration was found
Time O2 concentration CO2 Concentration to be normal, but as negative pressure was applied to the
(min) (%) (%) manhole CO2 gas was drawn in from the surrounding soil.
0:00 21 0.045 The CO2 and O2 concentrations measured at the exhaust of the
14:13 13 2.0 vacuum pump beginning at the start of vacuum testing even-
14:20 4.0 7.0 tually demonstrated that a deadly atmosphere existed within
14:34 1.2 9.8 the manhole (Table II). The 8” sewer lines and the top of the
manhole had been plugged as is the normal practice for this
type of vacuum testing, thus the only source of CO2 incursion
could have been through cracks in the concrete manhole vault.
The crew involved in this incident had been provided with a
the plugs had been properly sealed. He lost consciousness as four-gas meter and had been warned of the dangers associated
his head passed into the manhole and he fell to the bottom. A with entry into an active sewer. The crew’s four-gas meter was
worker was called to the manhole by the supervisor’s assistant found to be fully functional, but was in the crew truck at the
and he entered the manhole and also lost consciousness. The time of the triple fatality incident.
Downloaded by [152.180.134.66] at 13:51 21 November 2014

supervisor’s assistant ran across the job site to find another

worker, and when the third worker saw the two individuals at
the bottom of the manhole he entered and made it to the bottom,
but collapsed as he was attempting to exit. A fourth worker
was called to the manhole, and this individual did not enter,
but instead called the local emergency 911 center. A short time
later a fire department responder entered the manhole using a
self-contained breathing apparatus, and the third individual to
enter the manhole was removed after he was found to still
be alive. The first two individuals to enter were found to be
deceased at the scene, while the third individual eventually
died in a local hospital.
O2 Deficiency in a New-Construction Sewer System, St.
Louis Missouri, 2011
On the morning of Monday October 10, 2011, a foreman
was found unresponsive in a new construction sewer system
manhole (Figure 1). The sewer lines penetrating into this
manhole had not yet been connected to the local municipal
sewer system, but the manhole was connected to other new
construction manholes nearby through 8” diameter pipe. The
dimensions of the manhole were determined to be 6.3 m
(depth) × 1.2 m (diameter). This manhole had been entered

FIGURE 1. October 2011 St. Louis fatality site. A laser sight level used by the construction crew is still in place at the bottom of this
new-construction sewer manhole that had not yet been connected to an active sewer system.

Journal of Occupational and Environmental Hygiene December 2014 803

 without problem on numerous days prior to this incident to set
up and retrieve a laser level device.
The worker who discovered the foreman (worker 1) initially
thought that he had fallen into the manhole, and his response
was to call to another worker (worker 2) who was some
distance away. With the arrival of worker 2, worker 1 walked
away to find a street sign while calling the local emergency
911 center. He observed worker 2 entering the manhole and
he then returned to the manhole opening where he heard
worker 2 calling for help. He quickly started a nearby electrical
generator to operate a fan and used this to push fresh air into
the manhole, and worker 2 was then able to leave the manhole
without assistance. The elapsed time between entry of worker
2 (coinciding with the emergency call) and the arrival of fire
department rescue personnel as worker 2 was emerging from
the manhole was about 4 min. The initial readings obtained by
the fire department using a 15’ hose to sample the air in the
space showed O2 content of 19.1%, and additional ventilation
of the confined space in the manhole commenced immediately.
Downloaded by [152.180.134.66] at 13:51 21 November 2014
The foreman who had collapsed at the bottom of the manhole
was removed, and was found to be deceased.
A consultant hired by the employer of the workers ar-
rived the afternoon of October 10 with a four-gas meter,
and found O2 levels in the manhole > 20% throughout. A
local OSHA compliance officer (author Lockhart) monitored
the atmosphere in the manhole for more than 2 months, and
was supported by the agency’s Health Response Team and
Industrial Hygiene Chemistry Laboratory located at the OSHA
Salt Lake Technical Center.
Objective
A knowledge deficit appears to exist within the industrial
hygiene and safety communities regarding the need for at-
mosphere testing for entry into the types of spaces where
the fatality cases listed in Table I occurred. The purposes of
this article are to (1) raise awareness of the potential haz-
ard among safety, health, and management professionals to
save lives, and (2) to describe methods used in the 2011 St.
Louis case that were helpful to definitively characterize the
composition of whole-air samples from that confined space.
In addition to the specific cases where below-ground confined
space atmospheres involved dangerously low O2 and high CO2
concentrations, other types of confined spaces with different
types of atmospheric hazards are also possible. The laboratory
methods employed in this case may be of use in any investiga-
tion where atmospheric gas percent composition information
is needed for any type of confined space.
MATERIALS AND METHODS
Direct-Reading On-site Measurements
Plastic tubing was lowered into a manhole of the new
construction sewer system where the 2011 St. Louis fatality oc-
curred to obtain O2 readings from different depths with a four-
gas meter (TMX412, Industrial Scientific, Oakdale Pa.). A cal-
ibration check with a certified National Institute of Standards
804 Journal of Occupational and Environmental Hygiene
and Technology-traceable gas mixture containing 102 ppm
CO, 20 ppm H2 S, 0.74% pentane, and 19.4% O2 (34L-402Ex,
Industrial Environmental Instruments, Worthington Ohio) was
performed each day before use of this meter. A pump device
(SP402, Industrial Scientific) designed to interface with the
meter was operated with a nominal flow rate of 33 mL/minute.
Readings were manually recorded after being allowed to sta-
bilize (about 2 min).
Due to the need to obtain instrumentation and detector tubes
specifically for the measurement of CO2 , the first measure-
ments for this gas in the new construction sewer system were
completed on October 14. Both handheld field instrumentation
(CEA, GD444XR, Westwood, N.J.) and colorimetric detec-
tor tubes (Gastec, 810-2H, Kanagawa, Japan) were used for
CO2 detection and field quantitation. The portable instrument
operated on the principle of non-dispersive infrared (NDIR)
absorbance, and had been calibrated within the past year when
the measurements were taken. Before each use in the field,
an atmospheric CO2 reading was obtained to verify correct
operation. This NDIR detector contained an on-board pump
to draw air into the meter, and sampling occurred through
a hose kept at the desired depth within a manhole and the
meter was operated until CO2 readings stabilized. For detector
tube field analyses air was drawn through a tube using a
manually operated gas sampling pump and an extension hose
of the same manufacturer as the tubes and in accordance with
the manufacturer’s instructions. This ensured that a pump
stroke would pull air from the bottom of the fatality scene
manhole directly into a detector tube at the end of the hose.
Whole-Air Sample Collection and Laboratory
Analysis
Whole-air samples were collected from the 2011 St. Louis
fatality manhole in five-layer aluminized gas sample bags (10 L
volume, Cali-5-Bond, Calibrated Instruments Inc., Hawthorne,
N.Y.) for gas chromatography (GC) analysis in the laboratory
to definitively characterize the composition of the air samples.
This type of sample collection and analysis approach has been
validated for quantitative determination of CO2 gas concen-
tration.(11) Analysis occurred at the OSHA Salt Lake Tech-
nical Center using two separate model 6890 GC instruments
(Agilent Technologies, Wilmington, Del.). For quantitation
of CO2 and O2 a GC instrument equipped with a pulsed dis-
charge detector (PDD, model D-3, Valco Instrument Company,
Houston, Texas) was used with the PDD operated in helium
photoionization mode. A six-port rotary valve was connected
to the injector inlet of the GC instrument, and with helium
carrier gas (ultra-high purity, Praxair, Salt Lake City Utah)
aligned to flow to the injector, gas from a sample bag was
drawn through a 1.0 mL offline sample loop by a vacuum
pump.
After an analysis event was started (30 sec delay), the
rotary valve was realigned to allow carrier gas to sweep the
sample loop contents into the injector of the GC instrument.
A porous layer open tubular (PLOT) column (RT-QPLOT,
December 2014
 Restek, Bellefonte, Pa.) with dimensions of 30 m × 0.32 mm
i.d. was used for the analysis of CO2 . Carrier gas linear velocity
was set to 50 cm/s with constant pressure of 14.1 psi, the
injector was operated with a 50:1 split ratio at a temperature
of 200◦ C, and the column oven temperature was maintained
at 35◦ C throughout an entire run.
The CO2 calibration ranges used encompassed the mea-
sured concentrations from field samples. The calibration
curves for the laboratory-analyzed samples typically used a
low concentration standard known to contain 1017 ppm CO2
due to the high concentration of CO2 in many of the samples
analyzed, even though the reporting limit for CO2 using the
methods described here has been determined to be 339 ppm
when the lowest concentration standard is at or below this level.
Calibration with CO2 standards produced a quadratic instru-
ment response and calibration curves with R2 values >0.99. As
a quality control metric the OSHA Industrial Hygiene Chem-
istry Laboratory verifies that the calculated concentrations for
injected standards based on GC peak area counts are within
Downloaded by [152.180.134.66] at 13:51 21 November 2014
of 57 cm/s. The 6890 GC instrument used for these analyses
was fitted with a GSC-GasPro PLOT column (Advanced Sep-
aration Technologies Inc., Whippany, N.J.) with dimensions of
30 m x 0.32 mm i.d. The initial column temperature was kept at
35◦ C for 2.0 min, followed by linear temperature programming
at 10◦ C/min to 250◦ C, with a 2-min hold time at the final
temperature. The GC column was plumbed directly to an
Agilent Technologies 5973 mass spectrometric detector with
the transfer line heater kept at 250◦ C. The mass spectrometer
quadrupole assembly and ion source were kept at 150 and
230◦ C, respectively. Electron ionization (70 eV) was used,
and the range m/z 10 to 350 was scanned.
RESULTS AND DISCUSSION
F igure 2 graphically demonstrates the O2 and CO2 mea-
surements obtained over time using direct-reading field-
quantitation methods in the manhole where the 2011 St. Louis

fatality occurred. Dangerous conditions were not always present,

±10% of the respective known standard concentrations. A consistent with the history of the work site prior to the fatality
standard that falls within the calibration range is prepared as workers had entered the manhole on a number of previous
from a separate source, and analysis of this standard must also occasions without difficulty.
produce a calculated concentration that falls within ±10% of
its known concentration. Field blanks were not collected for Robust Air Analysis Investigation Methods Based
this type of bag sample. on Gas Chromatography
The OSHA method cited above(11) has been validated The typical configuration for a handheld four-gas meter
specifically for measurement of CO2 concentrations, but the includes detection for flammability (e.g., a catalytic bead pel-
concentrations of other atmospheric gases present in a sample listor type sensor), combined with electrochemical sensors
may also be determined following the same general approach. for O2 and usually H2 S and CO. In situations such as those
The full details for calibration of the GC instrumentation used described in this article where high CO2 concentrations were
for quantitation of O2 and nitrogen (N2 ) concentrations in the seen to exist along with very low O2 concentrations, a correctly
whole-air samples received from the field are not provided, but
were similar to the methods used for quantitative measurement
of CO2 discussed above. A different PLOT column (HP-PLOT
Molesiv, J&W Scientific, Folsom, Calif.) with dimensions of
30 m × 0.53 mm i.d. which was used for quantitation of O2 and
N2 would be suitable for most permanent gas analytes. It could
easily separate O2 and N2 from other gaseous components
present in a whole-air sample, however, it was unsuitable for
quantitative CO2 analysis due to excessive adsorption of this
analyte.
In many instances the OSHA Industrial Hygiene Chemistry
Laboratory does not use gas chromatography-mass spectrom-
etry (GC-MS) to confirm analyte identity, but this method is
useful in cases where overexposures have been determined,
or as needed for various types of investigations such as this
fatality case. Thus, GC-MS was used to confirm the identity of
CO2 in fatality scene samples concurrently with quantitation
by GC analysis with the PDD-equipped instrument. To accom-
plish this a second GC instrument with a mass spectrometric
detector was used. For GC-MS sample introduction, a 250 µL
aliquot of gas from a whole-air sample was removed from
a five-layer bag using a gas-tight syringe, and was directly FIGURE 2. Direct-reading field measurements of O2 and CO2
transferred into the heated injector of the GC instrument which concentrations over time in the 2011 St. Louis fatality manhole.
Field measurements for CO2 were not completed for days 0, 1,
was kept at 150◦ C. A split ratio of 10:1 was used, and pressure and 11.
programming provided an average carrier gas linear velocity

Journal of Occupational and Environmental Hygiene December 2014 805

 used four-gas meter would definitely warn workers of life-
threatening conditions and likely prevent fatalities. However,
it should be recognized that in some situations CO2 concen-
trations that would not immediately incapacitate a worker,
but that would be high enough to require the use of self-
contained breathing apparatus-type respiratory protection or
the immediate removal of workers would not be detected
using only the typical four-gas meter. While displacement or
dilution of atmospheric gases may lead to a low O2 reading, if
one assumes equal dilution of typical atmospheric gases with
a final 4% concentration of CO2 (which is the immediately
dangerous to life or health or IDLH concentration for this
gas), an O2 sensor alone would not detect that condition as
the O2 concentration would be about 20%. Also, the use of
electrochemical sensors for quantitative detection of O2 in
combination with additional simple detectors in a four-gas
meter will not definitively identify unsafe or unusual gases
actually present unless the specific sensors used happen to
correspond to these. Whether for purposes of worker protection
Downloaded by [152.180.134.66] at 13:51 21 November 2014
or investigation of an unfortunate confined space incident, the
correct selection of devices to measure atmosphere composi-
tion is important.
The ability to analyze a whole-air sample using a method
that separates and quantitatively measures a wide range of
gases that may be present is important to properly investigate
cases of serious injury or death where dangerous atmospheric
conditions are suspected. A whole-air sample may be analyzed
in the laboratory using adsorption chromatography with a near-
universal detector such as the PDD used here, and the same
analytical approach may also be useful for field analysis if a
small portable gas chromatograph with a thermal conductivity
detector is used. The confirmation of analyte identities made
possible through GC-MS analysis adds an additional level
of certainty to a high-concern investigation. The CO2 peak
seen in the Figure 3 chromatogram not only has the same
retention time as that produced from analysis of a CO2 standard
sample (mixed in helium), but a 70 eV electron ionization mass
spectrum consistent with CO2 further confirms the identity of
this analyte. The various chromatography data obtained in the
course of the OSHA investigation into the 2011 St. Louis
fatality showed unequivocally that elevated CO2 and low O2
concentrations existed in the new-construction manholes that
were not connected to an active sewer system. While the target
analyte in this case was CO2 , the same general approach could
be used to verify the identity of the major (percent level)
components in any gas sample.
Safe Confined Space Entry
The hazards of dangerous confined space atmospheres have
been widely recognized for years, and the requirements of the
OSHA general industry PRCS standard are widely understood
and followed for entry into many confined spaces. However,
the fact that new construction manholes which are not con-
nected to an active sewer system can be affected by CO2
intrusion seems not to be well understood, as evidenced by
repeated incidents. Considerable lengths of time may occur
806 Journal of Occupational and Environmental Hygiene
FIGURE 3. Chromatogram resulting from GC-MS analysis of
a fatality scene sample collected on October 23, 2011. The
quantitative results obtained by GC-PDD analysis of the same
whole-air sample showed CO2 and O2 concentrations of 5.3
and 9.3%, respectively. On-site direct-reading instrumentation
provided corresponding CO2 and O2 concentration values of 6.1
and 9.9%. The GC column used was selected to separate CO2
from other permanent gases present in a whole-air sample and
peak 1 consists of unresolved air components such as nitrogen
and O2 , while peak 2 resulted from the presence of CO2 in
the sample. Chromatograms resulting from the use of the HP-
PLOT Molesiv column with the GC-PDD instrument (not shown)
produced complete resolution of O2 from nitrogen and CO2
allowing quantitation of O2 , while chromatograms resulting from
the use of the RT-QPLOT column with the GC-PDD instrument
(also not shown) produced complete resolution of CO2 from the
other gas components for quantitation of CO2 .
between the placement of a sewer manhole or storm drain
vault and completion work due to weather or other construction
issues. This delay can allow soil gases to accumulate, and it is
also possible to actively draw soil gases into a manhole during
vacuum integrity testing.
The geological or soil-related processes that may lead to
high CO2 concentrations in below-ground structures such as
new-construction manholes are not clear, and thus there is no
way to know beforehand that dangerous conditions related
to this gas may exist in these locations. Similarly, it is not
possible to identify other types of confined spaces that might
have deficient atmospheres from visual inspection. For this
reason, all crews working in confined spaces should be pro-
vided with properly selected and calibrated monitors to detect
and measure the correct atmospheric hazards, and should be
trained in the correct use of this detection equipment. The
safety of a confined space atmosphere must be monitored
prior to entry to ensure adequate O2 , and the absence of
hazardous gases such as (but not limited to) CO, H2 S, and
flammable gases and vapors should be verified. The possibility
that other gases present in soil besides CO2 may contribute
to dangerous conditions should not be ignored, but specific
conditions and processes that may lead to dangerous con-
fined space atmospheric conditions from a gas other than CO2
December 2014
 (e.g., sub-surface blasting leading to accumulation of CO in
a new-construction manhole(5)) are beyond the scope of this
article. In a case where a gas such as CO may kill, even
when O2 concentrations are adequate to sustain life, a properly
selected detection device that will respond to the specific gas is
imperative. Also, in a small-volume confined space with poor
ventilation, the effect on the atmospheric conditions of O2
consumption and CO2 production by a worker should not be
overlooked. The use of continuous ventilation and continuous
monitoring with a correctly selected meter will further protect
workers.
Wilson, Madison, and Healy recently studied confined
space emergency response and noted that a substantial portion
of employers may be relying on public fire departments for
PRCS response.(12) To respond as quickly as possible with
rescuers who are familiar with actual hazards present, large
employers should consider the recommendations these authors
make to equip employees working on-site to properly address
technical rescue requirements. However, even if local public
Downloaded by [152.180.134.66] at 13:51 21 November 2014
fire and rescue personnel must be relied on by employers
with only a few workers at a work site some distance from
headquarters, there are steps that an employer should take
to protect the lives of PRCS workers. All employers with
such workers must prepare for, and be capable of competently
responding to, an incident in a confined space. In the United
States, employees who can speak and understand English must
be on duty outside of a confined space at all times when an
entry is to take place. These employees must be trained to
never enter a space to attempt rescue, and must understand the
need to call rescue personnel if an incident occurs. To maintain
safety in confined spaces such as new-construction manholes,
proper rescue apparatus such as a tripod and harness should be
used. Employees should be provided with training on proper
procedures in using this gear and should be trained in retrieval
procedures.
CONCLUSION
B ased on the numerous incidents that have led to the deaths
of unsuspecting workers for many years, a need clearly
exists to reiterate the dangers inherent to below-ground con-
fined spaces that do not present traditionally obvious hazards.
Problems with H2 S and other gas contaminants are fairly
well recognized when dealing with active sewers, and most
workers who routinely enter active sewers are equipped with
O2 sensors or some type of four-gas meter that would warn
of commonly encountered lethal atmospheres. A more vexing
problem occurs in atmospheres that many would assume to be
safe, such as within new-construction or inactive sewers, or in
other below-ground structures that are not intended for human
occupancy. Even when not specifically identified in confined
space standards, all sewer, storm drains, and utility vaults that
are deep enough so that a worker’s breathing zone will be
within the confines of an enclosed area should be considered
confined spaces and should be treated as such. While not an
exhaustive listing of all potential circumstances, this definitely
Journal of Occupational and Environmental Hygiene
should apply to situations where a worker steps inside a vault
through an opening, or enters any manhole. This concern must
be taken seriously, as demonstrated by the number of fatalities
which have occurred where employees were working in new
construction manholes not connected to an active sewer, or
other obvious chemical sources. It must be recognized that the
soil itself may contain gases or can act as a conduit for gases
that can penetrate and contaminate such spaces. In addition
to the use of a properly functioning and calibrated detection
device prior to entry into a confined space, experience has
shown that compliance with other requirements of the OSHA
PRCS standard will save lives, and this should be the goal of
any employer, industrial hygienist, or safety professional with
responsibility to see that confined space entry operations are
carried out safely.
To improve the information available with regards to con-
fined space fatalities or for any other circumstance where the
atmospheric composition of a confined space is called into
question, highly specific whole-air analysis methods should
be considered. Reliable quantitative analysis methods based
on adsorption chromatography have been shown to fill this
need for CO2 measurements, and as long as a particular gas
mixture is stable within a sampling bag and a suitable ad-
sorptive GC column is available the same methodologies may
also be used for other gases. This type of analysis should
be considered whenever a serious confined space incident
occurs, especially if it results in a fatality. A determination
of whole-air composition is important to understand the cause
of an incident and to facilitate further study into the variables
that affect the concentrations of various gases in confined
spaces.
DISCLAIMER
T he views expressed here are those of the authors. This
document is not a standard or regulation. It creates no new
legal obligations and alters no existing obligations created by
OSHA standards or the Occupational Safety and Health Act.
It may contain recommendations that are advisory in nature,
informational in content, and intended to support the provision
of safe and healthful workplaces. This article is not subject to
U. S. copyright law.
REFERENCES
1. “Occupational Safety and Health Standards, Subpart J - General
Environmental Controls, Permit-Required Confined Spaces,” Code of
Federal Regulations, Title 29, Part 1910.146. 2013.
2. “Safety and Health Regulations for Construction, Subpart S - Under-
ground Construction, Caissons, Cofferdams, and Compressed Air,” Code
of Federal Regulations, Title 29, Part 1926.800. 2013.
3. “Safety and Health Regulations for Construction, Subpart C - General
Safety and Health Provisions, Safety Training and Education,” Code of
Federal Regulations, Title 29, Part 1926.21. 2013.
4. “Occupational Safety and Health Standards, Subpart R - Special
Industries, Telecommunications,” Code of Federal Regulations, Title 29,
Part 1910.268. 2013.
December 2014 807
 5. Decker, J.A., L. Santis, S. Deitchman, J.P. Flesch, and R.T. Hagedorn:
Carbon monoxide poisoning and death after the use of explosives in
a sewer construction project. Appl. Occup. Environ. Hyg. 17:152–153
(2002).
6. Carter, J.T.: Vitiated air: A Victorian villain? J. Roy. Soc. Med.
74:914–919 (1981).
7. Tankard, A.R., and D.J.T. Bagnall: Fatalities due to vitiated air
produced by the oxidation of vegetable refuse. Analyst 656:673–676
(1930).
8. Raschka, G.J., and W.J. Uber: Oxygen depletion in a sewer manhole.
Sewage Ind. Wastes 23:802–807 (1951).
Downloaded by [152.180.134.66] at 13:51 21 November 2014
808 Journal of Occupational and Environmental Hygiene
9. Michaleson, G.S., and W.E Park: Asphyxiation in street manholes. Pub.
Health Reports 69:29–36 (1954).
10. Manning, T.J., K. Ziminski, A. Hymann, G. Figueroa, and L. Lukash:
Methane deaths? Was it the cause? Am. J. Foren. Med. Pathol. 2:333–336
(1981).
11. “Carbon Dioxide in Workplace Atmospheres: Method ID-172.” Available
at http://www.osha.gov/dts/sltc/methods/inorganic/id172/id172.html (ac-
cessed February 14, 2013).
12. Wilson, M.P., H.N. Madison, and S.B. Healy: Confined space emer-
gency response: Assessing employer and fire department practices. J.
Occup. Environ. Hyg. 9:120–128 (2012).
December 2014