EH-9309 September 1993 Occupational Safety Observer
                            SEPTEMBER 1993
                     Occupational Safety Observer

Part 1:

WATER HAMMER ACCIDENT KILLS HANFORD WORKER

      On June 7, 1993, a journeyman power operator at the Hanford Site was
fatally injured after a water hammer ruptured a cast-iron valve, filling
the operator's work space with steam.  The worker died a week later, on
June 14, 1993.  Before the accident, workers had taken precautions against
the potential dangers of working in a confined space, but they were
unaware of the potential for a catastrophic water hammer.

The accident
      At the time of the accident, Westinghouse Hanford Company (WHC) was
attempting to isolate the steam supply to Building 324 at Hanford.
Because Buildings 324 and 331 were both served by the same steam line, it
was necessary to arrange an alternate steam supply for Building 331 during
the outage.  Doing so required that an 8-inch Main Steam Supply valve
(MSS-25), located below ground in Pit U-3, be opened.
      Pit U-3 is a subterranean, reinforced-concrete vault with an
interior height of 16 feet and floor dimensions of 8 X 11 feet.  The pit
is entered through a 2-1/2-foot-square opening.  Pit U-3 contained four
valves, including MSS-25, and was classified as a permit-required confined
space by WHC because of (1) its limited space and high temperature
(approximately 140 degrees F), (2) its potential for oxygen deficiency and
for falls, and (3) the possible presence of asbestos.
      Because of these potential hazards, the preferred means of operating
valve MSS-25 was to do so remotely, from the surface.  Consequently, a
27-foot-long, T-handled reach rod was fabricated in an effort to open the
valve from the surface.  On June 7, 1993, workers attempted to open the
valve using the T-handled reach rod.  When the attempt failed, the
decision was made to send a power operator into the pit to open the valve.
      Workers took what they believed to be appropriate precautions before
entering the pit:  a confined-space entry permit was obtained, and a
respiratory technician reported to the worksite to take readings that
would identify oxygen deficiency or an explosive atmosphere.
The journeyman power operator who would actually enter the pit donned
personal protective equipment, including a disposable outer garment over
his coveralls, shoe covers, surgical gloves, leather gloves, and a powered
air-purifying respirator (PAPR).  This equipment was primarily intended as
protection against asbestos.  The operator was also wearing a rescue
harness, which he connected to the fall-arrester device -- but not to the
retrieval winch line provided for the pit.  A second power operator was
stationed outside the vault to support the entry.
      At 7:10 p.m., the operator entered Pit U-3 to remove the lock and
chain securing valve MSS-25.  He then planned to return to the surface to
get a valve wrench to facilitate valve operation.  However, he apparently
tested the valve hand wheel, found it to be free, and unseated the valve.
Workers on the surface noticed water and steam hissing from the pit's
steam trap discharge line, which vented outside the vault.  When the
venting stopped, the operator stationed on the surface saw the operator in
the pit move his hands and shoulders in a repetitive manner, which was
interpreted to mean that he was opening the valve further.
      Shortly thereafter, the respiratory technician heard the operator in
the pit exclaim, "Oh, no!" -- followed by a bang.  Steam, dirt, and rock
blew out of the pit.

Occupational Safety OBSERVER                                 Page 2

      Despite receiving severe injuries as a result of the "steam blast,"
the operator working in Pit U-3 was able to climb out without assistance.
An ambulance arrived at the scene within minutes of the accident.  The
injured worker was taken to Kadlec Medical Center in Richland, Washington,
where he was diagnosed with second- and third-degree burns to 65
percent of his body, as well as burns to his eyes and lungs.  Because of
the severity of injuries, he was airlifted to Harborview Medical Center in
Seattle, Washington, where he died 7 days later from lung injuries.

The investigation
      When investigators entered Pit U-3 several days after the accident,
they discovered that a 6-inch cast-iron valve located near the ceiling of
the vault had catastrophically failed, releasing steam into the pit.  The
6-inch valve terminated in a blind flange, was unnumbered, and had no
known use.
      Using an analytic approach, the investigation concluded that the
valve failed from thermal and mechanical shock and from overstress caused
by a thermal hydraulic transient water hammer.  Water hammer of this type
results from pressure pulses caused by sudden changes in momentum that can
occur when subcooled condensate and saturated steam are forced to mix
rapidly inside a closed piping system.
      Valve MSS-25 had been closed for 8 months before the accident
occurred, and the isolation had created a dead end in the piping system.
Valve MSS-25 was lower than the rest of the system, which allowed about
1,500 gallons of condensate to collect upstream of the valve.  There was
no steam trap or drain to remove the accumulated condensate from the
upstream side of valve MSS-25 while the valve was closed.
      Before this 8-month shutoff, valve MSS-25 had never been closed for
more than 72 hours, allowing a much smaller amount of condensate to
collect.  When the accident occurred, the temperature of the condensate
was estimated at 55 degrees F -- much cooler than the steam on the other
side of the valve, which was about 350 degrees F.  In addition, the
condensate upstream of MSS-25 was under 120 psig, whereas the steam
downstream was pressurized at 115 psig.  The operation of valve MSS-25
would not have been restricted by such a small difference in pressure.
      The power operator who entered the pit to open valve MSS-25 had
received no specific training for operating this particular valve, and
there was no formal procedure specific to this operation.  Workers at the
site stated that, on other occasions, they had routinely taken anywhere
from a few hours to several shifts to open MSS-25 gradually, thus
allowing the condensate to be removed by the steam trap on the downstream
side of the valve.  On this occasion, the operator opened the valve to the
halfway position within 1 - 2 minutes.  This action allowed the steam and
accumulated condensate to mix too quickly, thereby causing the water
hammer.
      It was determined that the cast-iron valve that failed met the
service conditions, but did not meet material/mechanical specifications,
for the steam system.  Preliminary analysis indicated that a pressure
pulse of 2,300 psig occurred, and the 6-inch valve failed.

Followup
      A meeting of senior safety representatives from throughout the DOE
complex was held August 25 - 26, 1993, to review this accident and develop
lessons learned and DOE-wide corrective action recommendations.  An
  
Occupational Safety OBSERVER                                 Page 3

upcoming issue of the Observer will discuss the root causes of the
accident and the results of that meeting.

References
ORPS #RL--WHC-WHC300EM-1993-0022
Type A Accident Investigation Board Report of the Hanford Site Steam Pit
U-3 Steam Valve Failure Resulting in a Fatality on June 7, 1993 (Draft,
July 1993)


Work Area Coordination:

INTERRUPTION OF BREATHING AIR TO CONSTRUCTION WORKER

      A combination of otherwise normal events recently created a
life-threatening occurrence -- a loss of breathing air -- that could have
been avoided if standard operating procedures had been followed. Further,
when the emergency occurred, workers did not fully understand or
appropriately implement immediate action procedures.  Luckily, these
events didn't result in injury or contamination, although they clearly
illustrate why workers must follow procedures and be prepared to react
effectively to off-normal events.

The incident
      On June 3, 1993, a construction worker at Savannah River Site was
assigned to perform tasks requiring that he wear a plastic protective suit
equipped with externally supplied breathing air.  The assignment involved
opening duct work and taking electronic measurements inside a "hut" (i.e.,
a temporary work space within a controlled environment).  A health
protection (HP) inspector accompanied the construction worker into the hut
to take readings for radiological contamination.
      While he was working, the construction worker inside the hut lost
breathing-air pressure without warning.  The worker immediately notified
the HP inspector, who instructed the worker to report to the airlock to be
"cut out" of the plastic suit.  Meanwhile, the standby operator, who was
responsible for helping workers exit the airlock and get out of their
suits, had left her post to see why the air supply had stopped.  She
climbed two flights of stairs, which placed her out of sight and earshot
of workers in the hut.  After discovering that the interruption in air
pressure had been caused by a maintenance worker replacing the filter for
the breathing-air system, the standby operator returned to the hut to help
the construction worker remove his plastic suit.  HP inspectors detected
no contamination when they surveyed the suit worn by the worker.

The cause
      The breathing-air stoppage occurred because a separations
maintenance mechanic had valved-out the breathing-air manifold to change
the air filter as part of a preventive maintenance program.  However, the
mechanic had not taken adequate precautions to ensure that no one was
using the breathing-air system.  The mechanic had contacted the Control
Room as dictated by procedure.  Control Room personnel in turn had checked
the log, which did not indicate that anyone was using breathing air.  A
visual search from the Control Room also indicated that no personnel were
on breathing air at that time.  Permission was therefore given to
  
Occupational Safety OBSERVER                                 Page 4

valve-out the breathing-air manifold.  Immediately before valving-out the
air, the mechanic performed a cursory walkdown of the system and then
proceeded with the job.
      Key factors contributing to this occurrence included the following:

   -- Failure to follow written procedures.  Procedures in place at the
site clearly state that the standby operator must call the Control Room to
register use of breathing air; the standby operator failed to comply with
this requirement.  The procedure further requires that the standby
operator remain at his or her post at all times:  the operator's first
duty in the event of an emergency is to help workers remove their plastic
suits.  However,the standby operator on duty during this incident had
temporarily left her post.

   -- Failure to communicate.  The standby operator and supervisor had
discussed this job earlier that day.  The standby operator asked whether
the job was still "on."  The supervisor indicated that everything was
ready and that the work would be conducted as scheduled.  The
standby operator mistakenly thought that the supervisor had arranged for
breathing air, and the supervisor assumed that the standby operator would
make these arrangements.  Clearly defined responsibilities for both the
standby operator and the supervisor would have enhanced communications and
might have prevented this incident.

   -- Recent modifications.  The construction worker in the hut was
working in a section of the building that was being remodeled, and the
breathing-air supply for that section had recently been removed.  A new
air supply was not scheduled to be installed for approximately
2 months.  Thus, a "work-around" was being used to supply the workers in
the hut with breathing air through an air-supply cart connected to a
manifold located two floors away from the job site.  As a result, the
worker was not only out of sight of the breathing-air manifold,
the work was performed in an area not normally served by the manifold
being used.

   -- Inadequate procedure.  The maintenance procedure for this facility
did not require a system walkdown before replacing air filters.  The
separations maintenance mechanic checked the manifold but did not follow
every airline to its end.

Lessons learned
      Even though this incident occurred in a radiological control area,
the lessons learned are applicable to all situations involving the
shutdown of an operating system.  The events surrounding this incident
reaffirm the necessity to perform the following actions:

   -- Conduct critical system walkdowns before beginning tasks that will
change the normally expected state of these systems.  When changing air
filters, for example, good operating procedures call for a walkdown of the
entire system -- regardless of what information is contained on the
Control Room log sheet.

   -- Be knowledgeable about standard operating procedures for assigned
tasks and about immediate action procedures.
  
Occupational Safety OBSERVER                                 Page 5

   -- Be aware of how changing conditions can affect a work area.
Personnel must be ready, as the old saying goes, "to expect the
unexpected."

   -- Ensure that the shutdown of a system will not disrupt operations in
progress.  Several good practices can provide such assurance:  (1) use a
formal lockout/tagout program; (2) attach an "in use" tag whenever a key
system is activated; or (3) require users of key systems to sign
operations clearance forms or logs before beginning and after completing
work.

   -- Establish clear lines of responsibility for ensuring the
availability of system operations.  In addition, require that work
packages include shutdown authorizations.

The regulations
      In this particular instance, 29 CFR 1910.134 offers guidelines for
cases in which supplied breathing air is used.  OSHA regulations specified
in 29 CFR 1910.134(e)(3)(i) require (1) that workers using supplied-air
respirators in areas where respirator failure could result in the workers
being overcome by toxic or oxygen-deficient atmospheres must be
supported by standby personnel, and (2) that communications (visual,
voice, or signal line) should be maintained between all parties involved.
Job planning for such tasks must include at least one individual who will
be unaffected by any likely incident and who will have access to
appropriate rescue equipment in case of an emergency.
      The presence of the HP inspector during the above-described incident
suggests that these guidelines were generally followed.  This precaution
proved to be a key factor in the safe resolution of the above-described
incident.

Reference
ORPS #SR--WSRC-HCAN-1993-0046



Cold Burns:

LIQUIFIED PETROLEUM GAS CYLINDER ACCIDENT

      Because compressed-gas cylinders are so common, workers sometimes
overlook their potential hazards.  In a recent incident at Los Alamos
National Laboratory (LANL), two workers received cold burns (i.e., tissue
damage resulting from exposure to extreme cold) from liquified petroleum
gas when one of the cylinders they were inspecting suddenly vented.

The incident
      On May 7, 1993, workers inside a Roads and Grounds Office at LANL
smelled the odor of liquified petroleum gas coming from the facility's
open storage lot.  Two workers went outside to investigate and identified
the source of the odor as one of eight 100-pound propane cylinders stored
in the area.  Because they detected an intermittent hissing noise
coming from the cylinder, they decided to check the shut-off valve to
ensure that it was completely closed.  While the valve's safety cap was
  
Occupational Safety OBSERVER                                 Page 6

being removed, the cylinder discharged a stream of liquified petroleum
onto the workers' arms.
      The workers were immediately sent to the local medical facility for
evaluation, and site safety representatives were summoned.  After a review
of the situation, the safety representatives decided to don protective
equipment and move the cylinder to an isolated section of the yard.  The
cylinder vented again during the attempted relocation.  At this point,
site safety representatives ordered the evacuation of the building and
storage yard and called 911.
      Personnel representing the Los Alamos Fire Department, the
Laboratory Emergency Management and Response organization, and the
Hazardous Materials Team all responded to the call for assistance.  The
Hazardous Materials Team removed the cylinder to a safe area to
be vented.  On completion of these actions, the area was declared safe and
the building was reoccupied.  The two injured workers were treated for
first-degree cold burns and returned to work later the same day.

Results of the investigation
      An informal investigation was conducted for this incident.  The
cylinder was DOT-certified for propane and had a valid inspection tag.
Response team personnel suggested that the cylinder was probably
overfilled at the recharge facility.  Despite LANL's system of quality
controls (e.g., random cylinder weight checks, valve inspections, and gas
sniffers), such an explanation is plausible.
      Another explanation faults the storage practices used at the Roads
and Grounds Office.  Cylinders were stored in the open and exposed to
direct sunlight for several hours a day.  Safety personnel speculate that
the heat of the direct sunlight expanded the contents of the
cylinder and caused the overpressurization.  Moving the cylinder agitated
its contents, creating additional pressure and worsening the situation.
The safety valve activated as designed to release excess cylinder pressure
to the atmosphere.
      The final ORPS report concludes that the incident resulted from an
overfilled cylinder being improperly stored in direct sunlight.

Lessons learned
      Gases expand when heated, and compressed gases contained in
pressurized storage containers are no exception.  CFR 1910.110(f)(2)(i)
requires that pressurized containers be stored in a manner that minimizes
their exposure to excessive temperature changes.  This generally means
that compressed-gas cylinders should be stored in the shade.
      Because compressed-gas cylinders are so common, the potential
hazards they pose are often dismissed by workers.  In this case,
management probably should have summoned the Hazardous Materials Team when
the propane leak was first detected.
      It is recommended that management periodically schedule activities
to increase worker awareness of the potential hazards associated with
compressed-gas cylinders.

Reference
ORPS #ALO-LA-LANL-SERVICESS-1993-0014
  
Occupational Safety OBSERVER                                 Page 7

120-Volt Shock:

WORKER SUFFERS ELECTRICAL SHOCK WHILE USING HAND DRILL

      When power tools must be used in the vicinity of live electrical
wires, proper precautions must be taken.  On March 3, 1993, a construction
worker at Savannah River Site failed to exercise such care and received an
electrical shock while attempting to drill through a wall.

The incident
      The worker involved in this incident used an ungrounded drill in an
attempt to create a hole through a penetration seal.  The worker knew that
there was a live wire that he could not see behind the penetration seal,
but he believed that the wire could be avoided during the drilling
operation.  In fact, the wire formed a loop behind the wall and was in the
path of the drill bit.  When the bit struck the wire, an electrical
current passed through the worker's left hand and went to ground through
his left biceps.
      In this instance, the worker was fortunate.  The shock could have
been much more serious.  The penetration seal was located on a wall behind
an instrument rack.  Because the space between the wall and the rack was
very narrow, the worker's arm touched the rack as he drilled.  In this
position, the current from the wire ran through the worker's arm to his
biceps, where it transferred to ground through the rack.  If a different
path to ground had been formed through the worker's body, the resulting
electrical shock could have caused serious injury or death.
      After sustaining the electrical shock, the worker underwent medical
observation and testing.  No medical treatment was required, and he was
sent back to work without any restrictions.

OSHA regulations
      The worker involved in this incident violated OSHA regulations as
well as basic, common-sense work practices.  Because the position of the
wire was not confirmed, proper safeguards should have been in place for
the work area.  Instead, the operator acted on his belief that the area
was safe for drilling.
      29 CFR 1910.333, "Selection and Use of Work Practices," states that
"live parts to which an employee may be exposed shall be deenergized
before the employee works on or near them unless the employer can
demonstrate that a greater hazard or extreme operational difficulty would
result."  If the live parts are not deenergized, safety practices
described in 29 CFR 1910.333(c) should be used.
      29 CFR 1910.333(c) applies to all employees who may be exposed to
hazardous energized parts.  Only qualified workers are allowed to work on
electrical circuits or equipment that has not been deenergized.  These
workers must be familiar with the proper use of special precautionary
techniques, personal protective equipment, insulating and shielding
materials, and insulated tools.

Lessons learned
      Before beginning a drilling operation, workers must be fully aware
of the position and nature of all electrical components in the work area.
  
Occupational Safety OBSERVER                                 Page 8

   -- Don't drill into a wall if you don't know what's inside.  Basic
precautions include walking down both sides of the wall, consulting wiring
and piping diagrams, and using an electrical tick tracer to check for live
wires.  (Be aware, however, that a tick tracer will not detect wires that
are not carrying current or other hazards such as water lines.)

   -- Use grounded or double-insulated power tools.  A grounded drill
would have protected the worker involved in this incident from receiving a
shock.  Additional safety measures might have included using a
double-insulated electric drill or wearing insulated gloves and
protectors.

   -- Deenergize circuits when in doubt.  If there is any doubt about the
location of a wire, deenergize the circuit before drilling.

   -- Don't assume unnecessary risks.  The worker assumed an awkward work
position that could have resulted in back injury, as well as electrical
shock.

   -- Provide clear guidance for dealing with unknowns.  Effective
guidance would have outlined the precautions listed in this section.

      Always know what's inside a wall before you drill.  In this
instance, the worker was lucky not to receive a more serious shock, but he
was also lucky in other ways -- he could have disabled a control circuit.

Reference
ORPS #SR--WSRC-WVIT-1993-0007



Unanticipated Vessel Failure:

LAB TECHNICIAN CUT WHEN GLASS VESSEL BURSTS

      Even when a program provides appropriate training, supervision, and
procedures, accidents can happen if hazards in the workplace have not been
identified.  The following incident shows that large glass vessels are
prone to failure while being pressurized and that such failures are
difficult to predict.  When a lab technician was cut after a large glass
vessel burst, her management learned the importance of anticipating
unexpected accidents and of ensuring that protective devices, such as
shields, are always in place to ensure worker safety.

The incident
      This incident occurred at the Idaho National Engineering Laboratory
Research Center on June 4, 1993.  While working as a lab technician in a
summer intern program, an undergraduate student was pressurizing a
12-liter glass carboy to lift biological media out of the carboy for
transfer to another vessel.  The biological media being transferred was a
nontoxic culture suspension containing water and mineral salt.  The
culture in the suspension consisted of nonpathogenic environmental
microbes that were being studied for phosphate ore solubilizing.
  
Occupational Safety OBSERVER                                 Page 9

      The intern was working under the direction of a principal
investigator, had been trained in the proper use of glass vessels, and was
following an established procedure that had been successfully performed
several times.  While the intern was pressurizing the carboy, the
glass vessel unexpectedly burst and the intern received cuts to her cheek,
chin, and ear.  The intern was taken to the Occupational Medical Office
for medical treatment and sutures.  She returned to her duties that same
day, with no lost time or work restrictions.  The damaged carboy and
spilled media were cleaned up and removed without further complications.

Cause of the incident
      Facility personnel investigated the incident to determine why the
carboy had burst.  The initial conclusion was that the carboy had burst
because it had been subjected to excessive pressure during the procedure.
Facility personnel conducted stress analyses on similar carboys to
determine a pressure below which the glass would not fail.  Data from
these analyses indicated that, because glass vessels are inherently
brittle and have slight imperfections, a pressure below which a large
glass vessel will never fail could not be predicted.
      An evaluation of the training protocols used at the laboratory was
also conducted to determine whether some aspect of training, such as
inadequate procedures or failure to follow established procedures, might
have contributed to the incident.  The intern was working with
a principal investigator, who personally monitored the pressurization of
the carboy.  Thus, the intern had adequate supervision at the time of the
incident.  In addition, the intern and the principal investigator had both
been trained in the proper use of glass vessels; both also wore
appropriate apparel, including safety goggles.  Finally, the intern and
principal investigator were following an established procedure that had a
solid record of success.
      Based on the investigation of this incident, facility personnel
concluded that the accident occurred because no one at the laboratory had
anticipated that the carboy might fail under the conditions described.
Thus, the laboratory itself was not adequately equipped with
shields to protect workers from the consequences of unanticipated
equipment failures.  Shields were subsequently installed in the
laboratory, and procedures were modified to require workers using glass
under pressurized conditions to conduct their work behind protective
shields.

Lessons learned
      The lessons learned from this incident are two-fold:  First, this
accident serves as a general reminder to all workers to review procedures
and processes for possible sources of failure.  Second, the incident
reminds laboratory personnel in particular to be aware of the
potential defects inherent in glass equipment.  Because a glass vessel can
fail whenever it is subjected to pressure, laboratory workers are required
to wear protective eye and face equipment as specified by OSHA in 29 CFR
1910.133 and in ANSI Z87.1.  The safety of the laboratory environment may
be further enhanced by installing protective shields for use by
personnel who work with pressurized glass equipment.

Reference
ORPS #ID--EGG-ERATOWNFAC-1993-0006