Fire and explosion at Bayer Cropscience facility-W.Virgina
On 29th Aug 2008, there was a large fire and explosion at a Bayer Cropscience facility near Charleston, W. Virginia. A team from the US Chemical Safety baord reportedly visited the site after the accident, but they have not released any investigation report as of now. The exact cause of the accident remains unknown, but here is some raw video footage, sourced from the citizen news site liveleak.com
There were unconfirmed reports of a tank that stored effluents and chemical waste from different streams exploding and causing the fire, but much remains unknown.
This is another gray area. Many times, the companies who process effluents may have no idea what kind of mixtures and reactions could occur between the effluents themselves, since the proportions may vary from time to time. Also who classifies under which category such unknown waste material goes? For example if you are storing Acetone than, it is classified under Group D (Hazardous Area Classification), similarly how does one decide under which category such unknown ( and possibly explosive and/or toxic waste mixtures) should go? Any thoughts on this would be welcome.
Also the emergency response wasn’t much to feel comfortable about, as per reports in some local newspapers and media.
New wireless instruments can improve safety monitoring
There is now yet another generation of instruments on the market-wireless! The developments in instruments are faster than what a traditional chemicals manufacturing plant was used to. First we had pneumatic/hydraulic & mechanical instruments, then analog instruments, then digital “smart” instruments, fieldbuses and now wireless. However, in this case these may prove to be very useful for safety monitoring.
How? Recall that in most plants, safety valves and rupture disks are devices that are not monitored remotely (except perhaps if there are CCTV cameras installed around them). Thus other than a bang (when the rupture disk goes off), there is no apparent indication. If you cannot hear it, nothing can be done.
There is now however a way, with the advent of wireless safety relief valves and wireless rupture disks. These devices have a small wireless transmitter mounted inside them, so that you can remotely monitor their status in far-away control rooms;suddenly they are now visible on a plant’s DCS or a SCADA system.
Isn’t that great? Coupled with some newer instruments that can be mounted on safety showers and eyewash fountains (that transmit a digital signal wirelessly when operated), it will lead to a greater degree of safety, even in older plants (as presumably one can retrofit the older relief devices and showers with these wireless transmitters).
Have any of you any experience in such a wireless safety instrument implementation? We will be glad to hear from you about your actual experiences. Please respond through the comments form (look for the link near the title of this post).
Can using Gas detectors simplify hazardous area classification?
Is it possible that if we use gas detectors to continuosly monitor the explosive limits in a classified (hazardous) location, we could manage it better?
Of course! If you recall, the area classification concept itself is based on the amount of time in a year, an explosive gas mixture (or a combustible dust-air mixture) is likely to be present in a given hazardous area. Thus a Divsion 1 classified location will have more incidences of explosive mixtures being present in a year than Division 2. Alternatively in the IEC system we have three Zones, which are Zone 0, Zone 1 and Zone 2. If you recall the definitions, Zone 0 is a place where hazardous gas/vapor and air mixtures are likely to be present for more than 100 hours in a year.
Why I am I again drumming in definitions that all of you already know? Because if you read the above carefully, it talks of a gas or vapor-air mixture as “likely” to be present.
Well, if you install an explosive gas detector in the area, you could actually measure if any such mixtures are actually present. Not likelihoods, but real occurences, if any. This means that from the uncertainty of likelihoods we are now talking about actual hours of explosive gas or vapor mixtures being present.
This is a big leap from earlier years and could lead to a revolutionary change in how hazardous areas are classified in the future.
Read more about it here.
Gas Detector calibration & testing-ideal frequency
In a typical large chemical manufacturing plant or facility, typically there may be hundreds of gas detectors and gas monitors that warn operators about lurking hazards related to leaking chemicals and vapors. These detectors are installed during the plant startup or possibly are added after a few incidents or as the result of HAZOP or safety studies. So far so good. How do you, as the plant engineer or plant manager or the safety manager, ensure that they continue to work as intended?
Well, you calibrate them or just “bump test” them with a known gas mixture in a gas bottle. Fine, but after how many months or years?
How often should you check, whether your installed gas monitors & gas detectors are working OK? In order words, what should be an ideal calibration frequency for these instruments? This is one of the questions that many engineers and technicians ask us, after they learn about gas monitors (from our excellent e-learning course on gas detectors and gas monitors).
Nobody seems to have a common answer.
Some experts suggest every year, some every half year and others, every quarter. So who is right? Some may feel the more the frequency, the better. However, the catch is, that in most electrochemical type gas detectors, every time a calibration or even a bump test is carried out, a small amount of the electrolyte is depleted. This means that the useful life of the gas detector gets reduced , the more you test it. It may so happen, that on a particular test (say the fifth one on the same sensor since its installation), almost all the electrolyte will get depleted. However because sufficient electrolyte was present during the test, the detector will pass out with flying colors. BUT, suppose the next day there is a gas leak AND the electrolyte is now depleted, the instrument that was just declared healthy yesterday will fail in the actual emergency !
Does this bother you? It should. Probably some manufacturer will start indicating the level of useful electrolyte left (there are some models that have this, but I am not sure) or there is some kind of other sensor diagnostic available, but in the vast majority of these detectors, it does not seem to be present.
What about the catalytic combustion type? Well, frequently exposing them may to %LEL gas mixtures may cause damage of the bead (repeated explosions taking place on the bead) and could render them ineffective in an actual gas leak.
The only types that can escape this “destruction by calibration” seem to be the semiconductor and the optical/infrared types, that may be unaffected by frequent exposures to gases to which they are sensitive. Does it mean that we should replace all of the electrochemical types and the catalytic combustion types with IR sensors and semiconductors? Not really practical as they are much more expensive than the electrochemical type and the catalytic combustion type.
So do we use the “lightbulb replacement method”, that used to be present in many factories in the 80s (for those who have forgotten about it, here is a refresher-some smart guy had calculated that replacing all light bulbs-even the working ones-every six months was cheaper than replacing only those bulbs that failed, due yto the inventory carrying costs, the payroll costs of the “light-bulb-changers” ,etc, etc).
Do we simply replace all the electrochemical sensors and catalytic combustion sensors every two years? Any thoughts on this issue?
Please use the comments form below.
(P.S. BTW if you wish to know more about gas detection or have a good training course on gas monitors, why don’t you download our excellent e-learning course ? )
Static Electricity Hazards-Mere bonding and grounding may not be enough!
Continuing with our series on the US Chemical Safety Board accident investigation videos, we discuss today the most insidous and hidden hazards in the chemical processing industry-Static Electricity. Why is static electricity so hazardous in chemical plants? Simply because many inflammable liquids (like most solvents), tend to accumulate static charges on their surface, especially when the liquid becomes turbulent. Static charges can then cause sparking and a single spark can then ignite a fire or cause an explosion of a very high magnitude. The consequences can be frequently disastrous.
Prevention of static buildup is done by bonding the metallic parts of all the liquid carrying equipment to each other (known as “bonding”) and then tying these to a single earth point (known as “grounding”). Together this technique is known as “bonding and grounding” and this helps the static charge that is built up on the liquid surface, to flow to the earth (ground), thereby preventing any sparking.
This is why workers are careful to carry out “bonding and grounding” while emptying the contents of a typical tank truck into a storage tank. The question is whether these measures alone can prevent static buildup entirely?
The answer is no. Surprised? Watch the video below. I”ll bet you never thought about “bonding and grounding” inside a storage tank!
Our conclusion? Safety is ensured by multiple and diverse layers of protection. Simply depending on a single protection mechanism does not ensure safety. Thus relying on just the “bonding and grounding” scheme did not help, as shown in the incident.
Hence the plant management should have thought of using Nitrogen blanketing or purging, as an additional layer of protection. Alternatively, a non contact level measurement technique like a Radar level gauge could have been used. Care should be taken to ensure however that the instrument is suitable for installation in a hazardous area, most likely Zone 0.
