Lessons in process safety from Fukushima

The Fukushima nuclear accident continues to unfold. There are lessons to be learnt in process safety from the accident:

1. Were your assumptions for the the worst case scenario correct? (The Fukushima plant suffered from both the earthquake and tsunami). The question I have asked is difficult to answer correctly because during the determination of worst case scenario, nothing bad has happened as yet and this will be weighing heavily on the minds of the decision makers. In other words, they may be wondering, why spend money on something that has never happened? The key word is "Never". How do you decide on probability of an event happening? Is your basis right? Are they supported by data?
2. Are the set points of your automatic shutdown systems correct? (In the Fukushima accident, though some of the operational reactors automatically shutdown once seismic activity was detected,  residual heat continued to be generated) 
3. Are your back up systems truly "back up"? There was no passive cooling water system available at Fukushima that would work even though cooling water pumps failed. It is reported that now in Indian nuclear reactors, authorities are planning to provide automatic shutdown for all reactors for seismic activity at a much lower setpoint than what the reactors are designed for. They are also thinking of providing batteries as power back up and connections for hook up of portable cooling water systems.
4. I am sure that there will be also lessons to be learnt in disaster management once the full details of the accident are out.

But we should not forget one thing - if it could happen in Japan, which is most technologically advanced and disciplined, we in India should be much more careful and conduct a thorough review of disaster preparedness.

Be aware of hazards of chemical cleaning

In 2009 an incident occurred during a chemical cleaning of a heat exchanger. 4 people reportedly died. Apparently, residual polymer was being removed from a heat exchanger by a chemical cleaning operation using 70% nitric acid when a chemical reaction caused pressure to build up  and blew the contents of the heat exchanger out. Chemical cleaning of heat exchangers is often done during turnarounds. Ensure that a proper hazard analysis is carried out before you embark on such a cleaning. Often, the information about hazards will be available within the company but may not be known to the personnel who are carrying out the job.
Read about the incident in this link.

Another steam turbine bites the dust - the dangers of overspeed

In February 2011 a steam turbine in a power plant in South Africa failed catastrophically when the overspeed trip system was being reportedly being tested. Luckily there were no casualties.  
Steam that you cannot control can kill. See pictures of the accident in this link.
See my earlier post on dangers of turbine overspeed in this link.

Process safety - Fires in insulation

Many fires, some of them devastating, have occurred due to fires caused by lube oil/ thermic fluid soaked insulation.
An article by Don Drewry and Dominique Dieken mentions the following on lube oil fires in steam turbines:
"Most steam turbines use mineral oil with a flash point ranging between 375 F and 500 F (190 to 260 Deg C). When sprayed onto a hot surface the oil will self-ignite at about 675 F (357 Deg C). Pressurized supply oil lines, if damaged, can compound the problem as atomized oil would be sprayed onto hot surfaces. Either self-ignition or a nearby ignition source can result in a three-dimensional fire at the bearing or lube oil piping with the burning oil flowing downward and collecting at ground level in the form of a pool fire.
The flame temperature of a lube oil fire, up to 2,100 F (1148 Deg C), can cause heat to be transferred to various turbine components through conduction, convection and radation. If there is flame impingement, the surface temperature of any exposed area can be expected to reach 2,100 F (1148 Deg C) within five minutes. As the temperature rises the turbine generator components expand at different rates. If this expansion is prevented because of geometric limitations, thermal stresses occur, which can exceed the yield strength of the materials and cause the components to fail.When a fire occurs the heat rising from the fire can also collect at the ceiling. If this happens the temperature can rise above 900 F and failure of the roof can occur. Likewise, a pool fire at ground level will quickly involve control and power cables beneath the turbine deck. Any extended exposure to heat will also damage the turbine's concrete pedestal".
Read the article in this link.

Fires have also occurred in reactors with coils that use thermic fluids to heat/cool the reactor. A presentation by John Griffiths mentions the following:
"Gas phase or liquid phase reaction?
Exothermic reaction of the liquid occurs as a result of oxidation by atmospheric oxygen. The liquid is dispersed over an enormous surface area within the structure.
How close to classical “thermal ignition”?
It all depends: if the fluid is very involatile at a typical temperature for exothermic reaction then the problem reduces to “thermal ignition”.The principles of lagging fires are the same – a breakdown of the balance between heat release and heat loss leads to thermal runaway.
So what are the distinctions from “thermal ignition”?
1.Vaporisation of the liquid can occur. It may be sufficiently rapid that most is dispersed, preventing self-heating taking place.
2.There can be an depletion of oxygen within the porous structure as a result of fuel vapour movement, but not necessarily enough to preclude oxidation.
3.The endothermic effect of vaporisation, contributes to the “heat loss” component.
4.Condensation is possible elsewhere in the structure (“giving back” the enthalpy of vaporisation).
Studies prior to ours erred towards involatile liquids so the most important distinctions (and features) of lagging fires were masked.
Is autoignition temperature of the fluid (AIT) relevant?
No (other than giving some indication of how reactive a substance might be)".
See the complete presentation in this link.

Explosions in power transformers and process safety

Numerous process incidents have been reported due to the failure of electrical power supply. There have been cases where back up power supply has also not come on line. Maintain your electrical power supply systems. A presentation by Dan Perco highlights the dangers of transformer oil tank explosions and the protection systems. Do not miss the last slide with a baby's photo in the presentation.
See the presentation in this link.

Lessons from emergency response

That's me standing in yellow bunker gear (First left, next to Instructor) during the intensive emergency response training course at Dubai in 1996! In my 32 year career, I have seen many plant emergencies and there are always some lessons to be learnt. Murphy's law always applies during emergencies! But it is always better to be prepared both mentally and physically. An article by FEMA on one of the worst chlorine releases in the USA (70 MT was released) summarises the lessons learnt from that incident:
"Lessons Learned
1. The problems associated with a high-risk occupancy in one jurisdiction, creating a prob­lem in a different jurisdiction, present obvious challenges for emergency planning response agencies.
In this case the responding agencies worked well together, but the deficiencies of the regulatory and planning processes were a major focus of attention after the incident.
2. The Incident Command System (ICS) proved to be extremely effective in this incident, par­ticularly in coordinating the efforts of several different agencies at the scene.
The ability to assign major responsibilities to command officers from different fire depart­ments, without any problems, is evidence that the personnel are trained and prepared to operate effectively.
3. The lack of effective radio communications between agencies was a problem at this incident.
Cellular telephones were used very effectively to supplement public safety radio capabilities and proved to be reliable under these circumstances. In other situations cellular telephone service has been compromised by the number of persons trying to use the systems under emergency conditions and particularly the heavy use of the systems by news media personnel. The loca­tion and time of day may have been key factors in making the cellular network responsive in this incident. Note: Centel Cellular will block off communications allowing only emergency personnel phones to work if the system starts to overload.
4. The decision between evacuating residents and warning them to remain indoors, with win­dows and outside air inlets closed, is often critical.
In this case, it was considered more practical to keep patients inside the hospital than to expose them to the outside atmosphere. This took into consideration the susceptibility of the patients to chlorine exposure, the ability to exclude outside air from the ventilation system and the avail­ability of medical personnel and equipment inside the hospital. Some area residents reported that they were notified to evacuate and were exposed to the chlorine cloud while waiting for busses to pick them up. The risk of exposure during evacuation may have been greater than the risk if they had remained indoors.
5. The use of buses operated by fire department personnel is a practical means to evacuate residents.
It is more feasible to have firefighters drive buses than to train bus drivers to use SCBA. It is equally difficult for firefighters using SCBA to convince residents to expose themselves to the outside atmosphere in order to evacuate, unless the residents are already in distress. The contin­gency plan, sending a crew of SCBA-equipped firefighters on a bus to enter and evacuate an area in immediate danger, is a practical innovation.
6. Police officers, who are not provided with or trained to use SCBA, were effective in evacuat­ing areas ahead of the contamination, but could not function in the contaminated areas.
Several police officers who were assigned to traffic control or to assist with evacuation were exposed to the chlorine cloud and transported themselves to medical facilities for evaluation in the later stages of the incident.
7. Due to the relatively low concentration of chlorine in the gas cloud, in this case, the pre­dominant medical condition was limited to short duration respiratory irritation.
Chlorine is detectable by odor at very low concentrations and is a respiratory irritant between 3 and 30 ppm. Individuals with chronic respiratory problems, such as asthma, were quickly affected and accounted for most of the hospital admissions.
8. It proved to be extremely difficult to determine the size, shape, and movement of the chlorine cloud.
Helicopter observation was a valuable asset, particularly with increasing daylight. Ground sam­pling over large areas is difficult to coordinate and requires careful mapping to be effective. An attempt was made to predict dispersion of the chlorine using CAMEO (a computer model pro­gram), but complicated factors of terrain, slope, temperature, wind velocity, relative humidity, and an unknown rate of release made predictions extremely difficult.
9. The application of water to the vapor cloud was considered in this situation to accelerate the evaporation of the pooled liquid.
Conventional wisdom suggests that massive applications of water spray could absorb chlorine from the air, resulting in a dilute liquid solution. Chlorine has a low rate of solubility in water, and, with a large leak, there is a concern that applying less-than-sufficient volumes of water would create a corrosive fog. Applying water to a container of liquefied chlorine could heat the contents to their boiling temperature and cause the container to rupture. It is often difficult or impossible to estimate the flow and rate of vaporization from a leak to make such determinations.
10. The delay in notification of the fire department and other agencies indicates a problem with plant personnel and the established standard operating procedures at the facility.
A review of communications tapes reveals that no call was received by the police or fire depart­ments for this incident from the facility. A private-sector ambulance provider had been requested to respond to transport plant employees who had been exposed to the chlorine gas.
11. While emergency procedures had been planned for the chlorine facility itself, there was no specific plan for notification or evacuation in the event of a chlorine leak or other emer­gency extending beyond the property line.
Due to the risk created by the chemical industries in the area, the need for emergency warning systems should be evaluated. This level of planning requires both private- and public-sector participation.
12. The fact that the incident occurred at the facility where the CHLOREP Team equipment was stored caused unusual problems. Most of the equipment that could have most valuable in trying to secure the leak could not be reached because it was in the Hazmat area. This included the SCBA units normally used by the team members, which are of a different type from those used by area fire departments. Another chemical plant in the area was able to provide the needed SCBAs for the plant members on the entry team".

Read the excellent article in this link.

Hot work incident and hydrogen peroxide fire

An accident during a hot work has claimed the life of one person in the USA. It is reported that hot work was going on a pipeline containing coal tar solvent when the explosion occurred. Make sure that your equipment are free of flammables before you allow any hot work.Read the news article and see the video in this link.
Thanks to Abhay Gujar for sending information on a hydrogen peroxide fire in Australia.About 1000 L of H2O2 apparently caught fire in this incident. See this link.
A good write up on the hazards of Hydrogen peroxide by Solvay Chemicals is given in this link The write up mentions the following
"Hydrogen peroxide is a powerful oxidant,and improper handling or use of the product can create potential hazards. For example:
• If hydrogen peroxide solutions come in contact with eyes, severe injury or even blindness can result.
• Hydrogen peroxide will irritate and possibly cause chemical and/or thermal burns on the skin.
• Ingestion may be fatal.
• Decomposition of hydrogen peroxide generates heat and gas which can result in rapid pressure buildup leading to pressure bursts of inadequately vented containers.
• Decomposition of hydrogen peroxide can generate sufficient heat and oxygen to initiate combustion of ignitable materials.
• Oxygen enrichment of hydrocarbon vapors resulting from the decomposition of hydrogen peroxide can result in vapor phase explosions.
• Hydrogen peroxide can form explosive mixtures with some organic substances.Hydrogen peroxide aqueous solution is noncombustible. If involved in a fire, it may decompose, yielding oxygen that supports combustion. Decomposition in confined spaces may result in pressure burst. If involved in a fire, keep containers cool by spraying with water. For fire fighting, use only water; do not use other extinguishing agents. Keep upwind and operate from a safe distance. Firefighters should wear full personal protective equipment (bunker gear) and SCBA (self contained breathing apparatus)".

Facility siting - fiberglass tanks and radiated heat

An incident in the US highlights the need for a careful evaluation of siting (location) of fiberglass tanks. In the incident, a fire from a hydrocarbon relaease due to an equipment failure led to the melting of a fiberglass tank containing hydrochloric acid. The HCl spilled into a dyke area. If you are using fiberglass tanks for storing hazardous chemicals, take into consideration the heat radiated from nearby potential sources of fires.
Read about the incident in this link.

Dangers of pneumatic testing

Pneumatic testing is carried out in certain cases where water should not be permitted in the system and for other design considerations. However pneumatic testing can be deadly if certain precautions are not followed. The Alberta Boiler Safety Association has provided the following tips:

"Due to the large amount of energy stored in compressed gas and the potential hazard of a sudden release of this energy, pneumatic testing should be avoided if at all possible. The data in Table 1 illustrate the comparative risk of a pneumatic test versus a hydrostatic test:

A pneumatic pressure test should only be considered if a hydrostatic test has been carefully reviewed and determined not to be feasible. When pneumatic testing is necessary, there are many critical safety precautions that must be considered. Some important considerations are:
a) Code of construction requirements for pneumatic test (e.g., UW-50 for ASME Section VIII Div. 1).
b) A determination of the energy stored in the test fluid. Calculations may be made based upon the isentropic expansion of a confined gas. 
c) Test site preparations and related precautions including removal of unauthorized personnel, isolation of test site and a determination of the restricted distance for the pneumatic pressure test. The restricted distance is the distance from the item(s) under test at which barriers are placed to prohibit access, and the distance at which the test is monitored.
For the example cited above, with approximately 2,000,000 ft-lbs of stored energy (equivalent to one pound of nitroglycerine) the restricted distance using the NASA Glenn Research Center methodology would be 80 feet.
d) Test medium, pressure source and pressure and temperature ranges during testing.
e) Provision of pressure relief valves, which must be sized to handle the maximum output of the pressure source, to avoid excessive testing pressure.
f) Material specifications of the vessel or system involved in the test. For materials whose resistance to brittle fracture at low temperature has not been enhanced, a test temperature above 60 ºF (16 ºC) should be used to reduce the risk of brittle fracture during the pneumatic test.
g) Precautions taken to prevent gas expansion temperature drop and thermal stresses due to temperature gradients".
Read the article in this link.

In February 2009, a LNG pipeline undergoing pneumatic testing killed a worker when it exploded.Here is the link for that accident along with pictures.

Confined space deaths

A fatality has been reported on Monday at a refinery in Singapore where a 34 year old Indian Worker reportedly died while carrying out maintenance work in a confined space.filled with nitrogen. Be very careful when working in confined spaces and with nitrogen inside. You cannot afford to relax your vigil even for a minute. Ensure your companies procedures are clearly understood by all contractors.
Read the article in this link.
Today's Hindu newspaper has reported that three workers in a Ranipet tannery were asphyxiated when they inhaled toxic gas in a chamber in a tannery. They had entered the chamber to lay new lines.
For all of you who work in the industry, remember that safety has no holiday. 

Process Safety - best built in design

I read an article about North Americas largest sulphuric acid plant (4500 STPD). The design of the plant took into account the requirements of PSM and they have used 3D modelling,simulation and CFD modelling etc to build in process safety and control of emissions at the design stage itself. One of the innovations is the use of novel expansion joints for the large diameter piping.The article mentions the following:
"The large plant size and expansion movements required special attention to thermal growth. MECS developed a Swivel Expansion Joint to use in place of a multi-bellows expansion joints when the thermal movement of the duct connecting the associated equipment was in more than one plane. An internal cable allows for multi-directional movement and also accounts for the internal pressure thrust being exerted from the single convulsion joint. This design eliminates the need for the external tie-backs that would be required when using a typical multi-bellows expansion joint".
It is always cheaper to build in process safety at the design stage itself instead of retrofitting later.
Download the article (large file 10mb) from this link.

Process Safety - enforcement is the key

A refinery in the USA has been proposed to be fined US$ 207,500/= for 45 alleged serious and 13 other-than-serious violations for exposing workers to possible fires, explosions and other hazards.
The news article mentions the following:
"OSHA inspected the refinery under its Petroleum Refinery Process Safety Management National Emphasis Program. The PSM standard emphasizes the management of hazards associated with highly hazardous chemicals and establishes a comprehensive management program that integrates technologies, procedures and management practices.
OSHA’s Baton Rouge Area Office began its inspection Sept. 15, 2010, at the company’s facility on Old Highway 7. Serious violations include failing to conduct adequate inspections and testing of piping and pressure vessels; ensure that employees in process and administrative buildings were provided adequate protection in case of an explosion; implement written operating procedures; resolve recommended actions resulting from compliance audits; provide an adequate confined space program; and provide an adequate lockout/tagout program for the control of hazardous energy.
Other-than-serious violations include failing to provide the required machine guarding and correct electrical hazards".
I was reading an article about two deaths due to an accident in a chemical factory in India. The article quotes the Director of Factories as follows 
"We filed charges against the company with negligence, issues of safety and standards of equipment, among others," .He said if the court agrees with the charges, there can be a fine ranging from Rs 25,000 to Rs 2,00,000 under the Factories Act, 1992, for causing deaths". Rs 2,00,000/= is about US $ 4500. You can draw your own conclusions!!
Read the article on the US refinery in this link
Read the article about the accident in the Indian Company in this link.