A 42-year-old laborer leak testing joints inside a 54-inch round pipe suffered fatal blunt force injuries in October 2015, when an inflatable ¿bladder¿ ruptured at a waste-water treatment plant. Inspectors from the U.S. Department of Labor¿s Occupational Safety and Health Administration found his employer failed to train him properly on the testing procedure.

Workers can be killed when employers fail to protect construction workers from the many dangers in confined spaces, said Assistant Secretary of Labor for Occupational Safety and Health Dr. David Michaels. These are among the first citations under OSH¿s new Confined Space Standard. Employers can prevent more tragedies like this one if they ensure proper training of workers and communication among multiple employers whose workers are on the same site.

In August 2015, OSHA implemented its confined space in construction standard after research showed proper safety procedures would protect hundreds of workers each year from life-threatening hazards. Hazards include the risks of toxic exposure, electrocution, explosion and asphyxiation present for workers in confined spaces such as pipes, manholes, crawl spaces and tanks. In an emergency, it can be difficult to exit these spaces quickly or for rescuers to enter safely.

The agencys investigation also found the four companies failed to continuously monitor confined spaces for atmospheric and other hazards and train workers in hazards. 

Source:Osha.gov

WASTE WATER TANK INCIDENT

On September 9, 2003, Employees #1 and #2 were extending the guardrail system to the hatch and platform area on top of a tank located on the Brewery property. The tank was used for the storage of solid waste at the waste water treatment facility. The tank was 40 ft in height and 48 ft in diameter. The employees were arc welding the guardrail system, 40 ft above the ground, when an explosion occurred. Employee #1 fell to the ground receiving fatal injuries. Employee #2 fell to the inside of the tank, an opening of approximately 10 ft left due to the explosion. The tank had to be drained in order to recover the body. Employee #2 apparently died from injuries due to the explosion

Credit:Osha.gov

ARE YOU CONSIDERING HUMAN FACTORS DURING YOUR HAZARD IDENTIFICATION?

 An ammonia leak occurred in the machinery room of an unoccupied arena. An employee was attempting to add oil to an ammonia compressor when he observed a leak (fill hose was not attached). Approximately 200 lbs ammonia was released.
Qualified person was trained, but with minimal experience in this procedure. No written procedure was available, and an error occurred while executing the procedure. The shut-off valve type (wrench-operated,
mufti-turn, no position indicator) added complexities to the process.

Source: British Columbia Safety Authority 

WHY HAZOP STUDY IS IMPORTANT

 Approximately 200 lbs ammonia was released to atmosphere. The condenser safety valve activated due to ‘no cooling’ in the condenser while the ammonia compressor was operating. While restarting the plant after a power failure, the operator forgot to start the condenser circulating pump (which should be started before starting the compressor). The compressor was started without condenser cooling, and as a result, ammonia gas temperature began to rise, thus raising the gas pressure in the system. Eventually the gas pressure rose more than the safety valve setting, activating the safety valve which released the
ammonia to atmosphere.
The compressor’s high pressure safety cut off did not activate. The high pressure cut off is supposed to activate and shut off the compressor unit when the system senses a high pressure condition.

Source: British Columbia Safety Authority 

ARE YOU SPECIFYING PRESSURE GAUGES CORRECTLY?

 Approximately 100lbs ammonia was released into an unoccupied processing room of an industrial facility when a pressure gauge failed on the liquid line to an ammonia evaporator. Inspection revealed that a second pressure gauge (on the hot gas line for the same installation) was pinned at maximum pressure. Both pressure gauges had a range of 0 to 150psi and were installed in a system with an operating pressure of 150 to 160psi. The pressure gauge failed from over-pressure operation.

Source: British Columbia Safety Authority

ARE YOU INPECTING SMALL BORE TUBING?

 An ammonia leak occurred at a commercial-industrial facility. Approximately 10 lbs ammonia entrained in approximately 200 litres of compressor oil was released when a suction side 3/8-inch pressure sensing line failed. The suction side oil pressure pushed approximately 200 litres of oil from the reservoir onto the floor where the entrained ammonia then escaped to atmosphere. The 3/8-inch stainless steel tubing within the compression fitting failed when a circumferential crack completely fractured. The crack within the
3/8-inch stainless steel compression fitting did not show up on external inspection. Metal fatigue appeared to be a factor, along with unit vibration and initial metal stress within this type of compression joint.
 

The refrigeration contractor identified the main cause and factors leading to the failure as a severe vibration condition of the compressor. This severe vibration condition only occurs when only happens when the control slide valve is at, or at near its minimum position. The vibration was so intense that the contractor immediately shut the compressor down. Also, the contractor’s investigation discovered the ‘PHD’ vibration monitoring system was inactive. When the monitoring system was activated, the compressor in fact shut down on ‘high vibration.’ 

Source: British Columbia Safety Authority

ARE YOUR SAFETY DEVICES WORKING?

Ammonia was released at a recreational facility. A high pressure cut out switch failed to shut down a compressor when the compressor experienced a high pressure condition. Pressure continued to build until a safety relief device operated releasing ammonia gas to atmosphere via the relief stack,
which triggered the ammonia alarm.
The water supply line to the condenser had no protection and was subject to freezing during cold weather. The high limit switch was old (1986) and is mounted on the compressor base subjecting it to vibration. The safety relief operated as designed, venting gas to atmosphere, preventing a possible
catastrophic failure.
Source: British Columbia Safety Authority

Lessons Learned: Pressure Vessels

NON METALLIC GASKETS - AVOIDING LEAKS AND BLOWOUTS

 https://www.plantengineering.com/articles/nonmetallic-gaskets-avoiding-leaks-and-blow-outs/

"The compressive stress on a gasket plays a larger role in its ability to maintain pressure than its tensile strength.
By Dave Burgess October 1, 1999

The compressive stress on a gasket plays a larger role in its ability to maintain pressure than its tensile strength. Why is this significant? It is the reason for many gasketed joint problems!

In a classic scenario, a joint is assembled without controlled bolt loads; that is, without known and controlled bolt torques. The joint withstands a hydro test at 1.5 or 2 times rated pressure, yet leaks or blows out after a period of service at pressures well below the test pressure.

Was this failure due to a loss of gasket tensile strength? Probably not. Gasket tensile strength alone cannot be counted on to hold system pressures. What very likely occurred was loss of compressive stress on the gasket."

READ THE COMPLETE ARTICLE IN THE LINK

ETHYLENE OXIDE RELEASE INCIDENT

 https://dnrec.alpha.delaware.gov/croda-questions-answers/

 "What happened at the Croda Atlas Point plant?

At 4:15 p.m., Sunday, November 25, 2018 ethylene oxide (EO) was accidentally released into the air from Croda’s ethylene oxide plant. Since EO is very soluble in water, Croda deployed its water deluge system to contain as much of the release as possible. Croda personnel also began transferring the EO from the leaking vessel in the processing equipment to a secure containment vessel. Local responders and DNREC Emergency Response arrived. Water suppression continued. With input from Croda personnel, responders determined that by opening two nitrogen valves in the processing equipment, the remaining EO from the leaking vessel could be transferred to the non-leaking vessel, stopping the release."

 READ THE LINK FOR COMPLETE DETAILS...

 

 

 

 

 

 

 

SODIUM CYANIDE LEAK

 Incident
A plant operator responded to a bund alarm and observed sodium cyanide solution overflowing from a storage tank. A pool of approximately 84 000 litres of solution had formed and was contained within the tank bund. The site emergency plan was enacted and an exclusion zone was established around the area whilst the product in the bund was transferred to another tank. The concrete bund contained the spill, and the recovered solution and all of the washings were kept for use in future blending operations.

Cause
The combination of a faulty valve and failure by an operator to follow standard operating procedure, led to the tank overfilling. The overflow spilt into the bunded area and activated the bund alarm.

Consequences
The defective valve was repaired and the tank high-level critical alarm system has been reviewed and improved. The tank overflow piping is also under review to identify options to prevent recurrence. The responsible operator has been counselled on appropriate self-check work systems. There were no injuries or damage to any property as a result of the incident and the emergency response plan functioned as intended.

Source: Department of Minerals and Energy, Australia

CHLORINE LEAK DUE TO STRESS CORROSION CRACKING OF SS BOLTS

 Incident
Approximately 150 kg of chlorine gas was released over a period of 140 minutes, due to failure of bolts on a vacuum gas regulator attached to an on-line liquefied chlorine gas drum, used for water treatment. The incident occurred outside work hours, however plant operators were not called out, as the plant alarm system was not correctly programmed. Impact from the release to the surrounding community was limited to chlorine odour being detected by a passing motorist who reported the odour to the operator of the
site. The leak was isolated by a plant operator who arrived at the scene after the release was reported.
Cause
Inspection of the vacuum gas regulator revealed that the release was due to the failure of four bolts on the regulator. Material test results show that the failure of the stainless steel bolts was due to chloride-induced-stress cracking from chlorine attack, which is assumed to have resulted from an extremely small leak over a short duration.
Consequences
As part of the investigation, a number of changes to prevent recurrence of the event have been identified and implemented. These include introduction of a weekly leak check, the company committing to inspection of all vacuum regulators, the replacement of any inappropriate bolts and reprogramming of the chlorine alarm. In addition, the company will investigate the use of different alloy bolts.

Source: Department of Minerals and Energy, Australia 

NITROGEN HOSE BURST AND RELEASED AMMONIA

Incident
A release of ammonia occurred from a chemical plant when a hose burst following maintenance to an ammonia filter. The release of ammonia was detected by operators due to ammonia alarms and a high flow of ammonia to the plant. The site emergency siren was activated to alert people of the incident and operators isolated the supply of ammonia to the plant. Operators donned personal protective equipment
and doused the leak with water in order to gain access to the area to isolate the leak.
Cause
Ammonia filters were used to remove contaminants from the liquid ammonia, prior to it being processed in the plant. An essential step in the maintenance of filters is a nitrogen purge of the system. After purging occurs, the filter is changed, resealed and the nitrogen hose disconnected. The relevant valves are then opened to recommission the system with ammonia. In this instance, the nitrogen hose remained connected to the filter and drain valves were left open allowing the hose to become
pressurised with liquid ammonia. While the hose was suitable for the pressures normally experienced under service with nitrogen, the hose was not suitable for the much higher pressures of liquid ammonia and as a result the hose burst in two places.
Consequences
The incident occurred as a result of the failure to follow the appropriate procedure and the company has taken the action of reviewing the competence of each operator to carry out the task. The company has also reviewed the operating procedures associated with the filter cleaning process to ensure they are appropriate. 

Source: Department of Minerals and Energy, Australia

RELEASE OF TOXIC LIQUID THROUGH STEAM COIL LINE GASKET LEAK

 Incident
A release of toxic liquid (containing arsenic trioxide) from a storage tank occurred following the shutdown of a chemical plant. The release occurred from a steam coil passing through a storage tank used to store the solution during the shutdown. The release flowed into a surface-drainage system,through a series of secondary catchment sumps and ponds, into Cockburn Sound.
Cause
An investigation into the incident has identified that the toxic liquid was released through a hole in a gasket of a flange. The flange was located on the steam coil pipework within the tank. It was found that the steam-line terminated outside the bunded area in a partially-covered surface-drainage system.
Consequences
To prevent a recurrence, all steam-lines within the plant were cut to ensure that they terminated within the bunded area. The company has also decommissioned the plant and will commission a new plant that does not utilise arsenic compounds.

Source: Department of Minerals and Energy, Australia 

FAILED GASKET DURING START UP

 Incident
A mixture of process gases was released to the atmosphere through a failed gasket during the start-up of a chemical plant. Operators had just completed the start-up when they heard a large steam release and received alarms from ammonia detectors. The plant was shutdown, however the plume released travelled off-site necessitating the evacuation into refuges of a small number of workers on a neighbouring site. None was injured as a result of the release.
 

Cause
An investigation showed that the gasket failed as a result of a hole in a boiler tube which had allowed water to pass from the boiler side into the process side. The temperature generated during start-up caused the pooled water to rapidly boil leading to a surge in pressure which resulted in the failure of the gasket. Non-destructive testing of the boiler tubes showed gouge-type corrosion believed to have been caused by flow distribution problems in the boiler. This resulted in excessive metal temperature, which led to corrosion of the tube. 

Consequences
A complete boiler inspection has now been conducted and boiler tubes showing signs of corrosion have been plugged off. Actions to prevent a recurrence include the addition of insulation at the top of all boiler tubes and further investigation into the boiler water chemistry to minimise potential for corrosion. The company has also modified the boiler to incorporate a drain valve in order to provide early warning in the event of water leakage.

 Source: Department of Minerals and Energy, Australia

START UP INCIDENT IN AMMONIA PLANT

 Incident
A mixture of synthesis gas and ammonia was released from an ammonia plant during the start-up process following a shutdown for maintenance. The release occurred from a high vent after ammonia had been allowed to collect in the vent system through a valve left open as a result of failure to tag the valve as out of service.
During the start-up, operators failed to control levels in the process, resulting in a high level alarm in an ammonia catchpot which led to a release of synthesis gas (as designed) into the vent system. However, as the vent system already contained ammonia, the gas mixture was forced out of the high vent. The wind direction took the gas cloud over an adjacent construction area on the same site resulting in fourteen
contract employees experiencing irritation and discomfort and later seeking medical attention. Six of these contractors were physically affected as a result of the exposure but nobody suffered serious or long-term injuries. The ammonia release dispersed and was not detected at neighbouring premises.
Cause
The incident was caused by a combination of factors including a failure of the tagging system and the failure to adequately control the catchpot level during start-up.
Consequences
The tagging system for equipment which is out of service, has been reviewed to ensure during maintenance there is no unauthorised operation of equipment. Resource allocation has been reviewed to ensure that process control is secured and level indicators in the catchpot have been assessed to ensure accuracy for process control. Also, workers at the adjacent construction site will now carry gas respirators at all times and will be notified of any operating conditions, such as start-ups and shutdowns, with the potential to lead to releases of gas.

Source: Department of Minerals and Energy, Australia 

AMMONIA LEAK FROM REFRIGERATION SYSTEM

Incident
Approximately 56 kg of anhydrous ammonia was released to the atmosphere from
a 25-year-old refrigeration system. The emergency services stopped the leak by
closing an isolating valve. There was no injury. The site was not licensed for the
storage of anhydrous ammonia.
Causes
One valve failed to close and stop filling up the accumulator of the refrigerating
system. This resulted in flooding of the refrigeration compressor by liquefied
anhydrous ammonia, causing one of the compressor O-rings to fail, thereby causing
the release of anhydrous ammonia to the atmosphere. This points to an inadequate
maintenance program.
Consequences
The operator has been instructed to take appropriate measures so that the
refrigerating system complies with the relevant safety standard and the Dangerous
Goods Regulations. The operator was successfully prosecuted for failing to placard
the site and have a licence to store anhydrous ammonia.

Source: Department of Minerals and Energy, Australia

Explosion due to hydrocarbon being drawn into boiler firebox

A leak in a hydrocarbon refrigerant system formed a vapor cloud that was drawn into the inlet of a steam boiler. The increased fuel to the boiler caused rapidly rising pressure within a steam drum. The rapidly rising pressure exceeded the capacity of the boiler’s safety valve and the steam drum ruptured. The boiler rupture was close enough to the gas leak to ignite the vapor cloud and produce an explosion due the confined nature of the gas lea  and an ensuing fireball. The fire took eight hours to extinguish. The explosions and fire destroyed a portion to the LNG plant and caused 27 deaths, and injury to 72 more.

LNG leak causes sewer explosion

 LNG leak from open run-down line during a pipe modification project. LNG entered an underground
concrete storm sewer system and underwent a rapid vapor expansion that overpressured and ruptured
the sewer pipes. Storm sewer system substantially damaged.

The Case of the FATAL TANK ENTRY

 https://www.assp.org/docs/default-source/psj-articles/mtricketts_0220.pdf?sfvrsn=ba188947_2