Are you ensuring the integrity of tank roofs and gauging platforms?

When I was a shift in charge in a naphtha based ammonia plant in the 80's, we used to gauge the level in the tank by climbing up the staircase of the floating roof tank. The tank was provided with a gauging pipe, which we had to open and we used to drop the measuring dip tape with a brass bob attached to the end. We used to apply a paste on the tape, at the approximate level. After we performed the dip, we could observe the exact place where the paste colour had changed and that told us the level. All this while we used to stand on the gauging platform which was mounted on the tank roof. In my 40 years experience since, I have read and heard about quite a few incidents where the person performing the gauging fell into the tank as the roof and gauging platform structure were badly corroded and gave away. In one of the cases in an oil refinery in India, the body got stuck in the heavy oil and they had a tough time removing it.

Another incident reported in OSHA.gov mentions this:

"An employee was taking measurements of Bunker C fuel oil in a tank. He was going to access the tank through a hatch located on its roof. When the employee stepped on the roof, a section of it collapsed because of corrosion. The employee fell inside the tank and died of asphyxiation".

LESSON: Maintain the integrity of your tank gauging platforms and roof, along with the rest of the tank.

Improper isolation incident

Part of a benzene plant was shutdown, as part of the annual shutdown programme. As part of the preparations for maintenance the main process sections were drained, purged and steamed in accordance with the set procedures. Work then began on the stripper column reboiler circuit, including two heat exchangers. The actions required for the preparation of one of the exchangers had been highlighted, and so it was assumed these actions had been completed. Under a Permit to Work the foreman and 4 of his team commenced on unbolting the exchanger end plate and the main channel end flange.
The work was not completed and was carried forward to the next shift. During the work it was noticed that the exchanger surface was still hot. This was assumed to be due to steaming operations in the shell side of the exchanger. The following day under a re-signed Permit to Work, the team continued with unbolting and the exchanger end plate seal was released. Hot condensate spilled out of the bottom section of the exchanger end channel. When the flow ceased the final bolts
were removed from the end plate flange and the end plate cover was rigged ready for lifting down to ground level. Approximately 10 minutes after the end plate was removed, a fitter working adjacent to the area was hit by a large flow of hot condensate, which flowed from the exchanger, impinged on a tube baffle plate and then sprayed over the fitter. He crawled away and colleagues put him under a safety shower until the ambulance arrived. The fitter received scalds to his back and neck. Investigations showed that there had been ineffective isolation of the exchanger system from the live LP plant steam supply. There was also passing valves on the condensate system which contributed to the presence of hot condensate. The highlighted had not in fact been completed and
there had been inadequate physical checking of the isolation work prior to handover for maintenance. The Permit to Work system had not highlighted potential hazards, and due to work overload was not being operated effectively.
Lessons
The following recommendations were made:
1. Key isolation valves should be checked for passing.
2. All work packs were re-checked for proper system isolation before shutdown work recommenced.
3. The organisation and supervision for the shutdown were reviewed and clear requirements for detailed recording and handover of progress between shift
teams were set.
4. A schedule was to be set up for a management review of the progress of the new coordination routine and for general safety auditing of the shutdown
activities on the plant.
5. The lessons learnt from the incident were to be circulated to other plants undergoing shutdown, to identify Best Practice for the future.
6. Generic recommendations from other condensate related incidents were to be reinforced.
Source:IChemE

Pneumatic testing fatality

A worker was killed and another seriously injured during leak testing on a heat exchanger.
The workers were using inert gas when a tube bundle ejected with great force striking them both.
An investigation into the cause of the incident found the following immediate causes:
1. Use of an unsafe work procedure for leak testing of the heat exchanger, no test ring was used and the use of high risk pneumatic test method.
2. Failure to stop test when instructed.
3. Inadequate protection from the potential of tube bundle propelling outwards.
Source:IChemE

Low temperature failure incident

Three nozzles on top of a reactor suffered cracks in the welds during decommissioning of a high-pressure lube oil hydrogenation unit when it inadvertently discharged liquid nitrogen into three reactors. Excessive shrinking occurred, caused by thermal shock.
Damage that occurred to equipment is estimated to be approximately US$55,000 (1999).
Source:IChemE

Coal dust explosion

Six workers were injured and one killed in a coal dust explosion and fire at a power generating plant.
Three injured mechanics and electricians were in critical condition at hospital with third degree burns over more than half of their body.
The accident occurred in a unit of the plant's coal burning plant minutes after workers restarted a coal pulverizer.
The pulveriser had been taken off-line for some maintenance work. The mechanics had finished the maintenance and were testing it. The cause of the explosion is not known.
Source : CNN.COM, U.S. NEWS, AUGUST 16, 1999, (http://www.cnn.com).

Empty drums are dangerous!

A worker welding a pipe onto a 55-gallon drum was seriously injured when oil vapours from the drum ignited causing an explosion. The drum had been used to store waste oil.
Source: csb.gov

Inadequate breathing air capacity kills two

Two workers carrying out sandblasting work inside a 22-by-27 foot boiler were found dead by a third worker at a chemical plant. It is not known what hazardous chemicals were involved as the boiler used water to create steam and did not handle chemicals.
The workers were provided with breathing air due to the nature of the space they were in and because of the sandblasting operation.
An investigation into the cause found a low level of oxygen in the cylinders used.
Source : CNI NEWS, 21 MARCH 2000, (http://www.cnionline.com)

Fire water systems should not be connected to other sources

A fire occurred at a refinery when fire fighting water became contaminated with fuel. An investigation into the incident found a small leak in a closed valve that is meant to separate the fire fighting water used to wash out fuel processing vessels.
Four other valves where meant to serve as backup devices to prevent contaminated water from flowing backward into the fire fighting water. But three were stuck in the open position and the forth one had a broken spring.
The incident occurred when the fire fighting water was sprayed underneath a welding job to quickly extinguish sparks that might ignite any stray vapours from refining units. But the water released a cloud of gas that burst into flames. The worker holding the hose and the welder suffered burns in the fire.
Source:csb.gov