Overnight Ice Factory Explosion Leads to Chemical Hazard, at Least 60 People Injured

[pizzachang]

5 hours ago, Georgealbert said:

employee forgot to close a valve,

The most common reason for larger leaks.  We used a check system and even had actual key locks for valves that were normally closed and a tag when you removed the lock to use the valve.  Ammonia does have to be added to the systems periodically, because ammonia does leak, even through the tiniest of openings - the good thing is you can smell a leak at 5ppm and that is when action needs to be taken.

[Georgealbert]

11 minutes ago, pizzachang said:

Well, as I said maintenance is the key in systems like this. Interesting enough "oil" is essential in maintaining the integrity of piping and valves.  I witnessed an explosion of the ammonia tank at a fish processor in Alaska - across the street from our ice plant - 4000 lbs of liquid ammonia.  The resulted cloud dispersed relatively quickly because of the location in a bay, but the fire took some time to extinguish.  This accident started with an rusty valve in a supposedly unused section of piping. The engineer (a very competent tech that I knew personally) had his mask and an assistant, and were aware of all the possibilities; the threads at the valve broke (corroded from the inside - ammonia is corrosive - much less so if there's enough oil in the system.) and they ran outside and then attempted to turn off another valve that "could" supply that piping, but too late. Neither were injured but did get some serious inhalation - ammonia turn chemically into sulfuric acid when it contacts the moisture in your lungs. We used sulfur "sticks" to locate small leaks and those are terrible to inhale too, but the masks work very well. So, in a guess, a leak unattended, a rusty pipe corroded inside and looking perfectly normal outside, or just not double-checking a refill procedure could have caused this one. The storage tanks are usually regulated and inspected but do require safety valve replacement by time in service, at least in the US.

Thanks, so it that situation it was not possible to close the valve, or did the responders try to use any sort of hazmat leak kit or leak sealer?

 

You said the fire took longer to contain, was that because a decision was made to allow the fire to burn,  containing the fire spread, to burn off the leaking gas or because it was a large out of control fire?

 

So would you expect any automatic warning device to detect the leak, or should the area have some form of detection system, or just rely on staff smelling the leak?

 

Sorry to ask, but I have some friends, who are first responders and they like to quiz me on what I would do. Personally I would want as much information from someone at the factory, when doing the initial size up, SDS/hazmat guidance, then a plan of rescues, containment, isolation of leak, vapour dispersal and finally clean up.

[pizzachang]

12 hours ago, Georgealbert said:

it was not possible to close the valve

I'll try and respond in order.  It is quite possible that the valve couldn't be closed (various reasons, including corroded parts inside or simply to much escaping gas) Remember, it's under pressure to keep it liquid. To my knowledge, limited to my years of service ending in 2008, there is no "leak stopper" or sealer. Maybe there is now, I haven't kept up. #2 Ammonia likely was already fully expended from the system in the Thailand leak - up to the next closed valve. In the case of say, a 100 lb. refill tank (looks like a big lpg tank but beefier) a faulty valve or unscrewing past a certain point will be a big surprise to the tech. Once I had an assistant who did exactly that on a routine replenish to our holding tank, and the 100 lbs of instantly converting liquid to gas, emptied in about 3 or 4 seconds. I had time to stop him from trying to rethread the valve - even with heavy rubber gloves, he would have had frostbite or worse. This much ammonia dispersed quickly, but the surrounding businesses complained a lot. By the time the complaints reached the city's manager, even the smell was gone. Remember, 5ppm can be detected but it's not really a danger - think of household ammonia sold in the US ( I never see it in Thailand) The fire was out of control in our situation before the f.d. arrived and the decision was made by them & plant managers to let it finish but contain it to the building. The bad part was the various materials other than ammonia that burned, insulations, paint, oils, and lots of paper/plastic packaging supplies. But the ammonia was long dispersed. #3 Location of sensors, their maintenance would be the factor here. The sensors we used had a industry setting 25 ppm, set in potential leak areas at around 6 feet. They require periodic maintenance to do their job. I'd guess that the leak (Thailand) was small enough to be contained  and don't know if the particular location of the leak could have been tied to any automatic shut-off.  Usually these sensors are tied to a sound warning for the particular area; then any responder had to have the proper gear to enter the area; first to assess the location, turn off adjacent valves or see the situation is serious enough to allow the expulsion to finish and then fix the problems. Remember, NH3 dissipates very quickly and containing an active leak depends on different factors - of responders, equipment quickly available and the actual amount of NH3 that is involved.  A train tank car is much different than a small ice plant's storage or refill vessels. Here's some photos of the one I witnessed. As you can see, the location was about as good as they get for quick dispersion of the ammonia. 

 

[Georgealbert]

18 minutes ago, pizzachang said:

I'll try and respond in order.  It is quite possible that the valve couldn't be closed (various reasons, including corroded parts inside or simply to much escaping gas) Remember, it's under pressure to keep it liquid. To my knowledge, limited to my years of service ending in 2008, there is no "leak stopper" or sealer. Maybe there is now, I haven't kept up. #2 Ammonia likely was already fully expended from the system in the Thailand leak - up to the next closed valve. In the case of say, a 100 lb. refill tank (looks like a big lpg tank but beefier) a faulty valve or unscrewing past a certain point will be a big surprise to the tech. Once I had an assistant who did exactly that on a routine replenish to our holding tank, and the 100 lbs of instantly converting liquid to gas, emptied in about 3 or 4 seconds. I had time to stop him from trying to rethread the valve - even with heavy rubber gloves, he would have had frostbite or worse. This much ammonia dispersed quickly, but the surrounding businesses complained a lot. By the time the complaints reached the city's manager, even the smell was gone. Remember, 5ppm can be detected but it's not really a danger - think of household ammonia sold in the US ( I never see it in Thailand) The fire was out of control in our situation before the f.d. arrived and the decision was made by them & plant managers to let it finish but contain it to the building. The bad part was the various materials other than ammonia that burned, insulations, paint, oils, and lots of paper/plastic packaging supplies. But the ammonia was long dispersed. #3 Location of sensors, their maintenance would be the factor here. The sensors we used had a industry setting 25 ppm, set in potential leak areas at around 6 feet. They require periodic maintenance to do their job. I'd guess that the leak (Thailand) was small enough to be contained  and don't know if the particular location of the leak could have been tied to any automatic shut-off.  Usually these sensors are tied to a sound warning for the particular area; then any responder had to have the proper gear to enter the area; first to assess the location, turn off adjacent valves or see the situation is serious enough to allow the expulsion to finish and then fix the problems. Remember, NH3 dissipates very quickly and containing an active leak depends on different factors - of responders, equipment quickly available and the actual amount of NH3 that is involved.  A train tank car is much different than a small ice plant's storage or refill vessels. Here's some photos of the one I witnessed. As you can see, the location was about as good as they get for quick dispersion of the ammonia. 

 

Many thanks for giving such a detailed and interesting response. My comments are more generic and general hazmat procedures, hence why I said before, at any incident, the on site specialist is the person to advise emergency responders.

 

The hazmat leak sealers kits I have used before, were normally made of plugs, stoppers, patches, bandages, clamps and sealing putty and were designed for sealing up punctures, gashes, cracks and fractured pipes/valves, not relevant in this situation, as this now seems to have been a valve left open, and emptying the tank.

[pizzachang]

2 hours ago, Georgealbert said:

hazmat leak sealers kits

Interesting. We never used these types of aids.  Just as a reference, an ammonia leak would have to be extremely small to be able to use such an item, since the ammonia is under a lot of pressure, and  if it's out of control by valves, you'd need protection to get close.  I can certainly see the use, in a remote, field condition situation, as in a tiny leak in a large tank.  Thanks for the conversation

[bristolgeoff]

A ice factory if it makes dry ice may have  have chemicals  so possible danger for sure

[Georgealbert]

4 hours ago, bristolgeoff said:

A ice factory if it makes dry ice may have  have chemicals  so possible danger for sure

Sorry but this was a ice factory using ammonia, as a refrigerate and probably water purification , standard process everywhere.

 

Dry ice is solid carbon dioxide, totally different process

[Mr Meeseeks]

On 4/18/2024 at 6:16 AM, hotchilli said:

Yet    again commercial businesses in the residential areas.

 

A well known NDT company in Thailand keeps its radioactive sources in a house in a gated village in the suburbs of Bangkok. I have personally seen it.

 

There is absolutely no proper risk assessment done that factors in the danger to the public for most commercial activities in Thailand. 

[Mr Meeseeks]

On 4/18/2024 at 1:22 PM, Georgealbert said:

100 km from any chemical processing in Thailand, would be a long way from civilisation.
 

Maybe the poster should consider contacting Elon Musk, no chemical works on Mars, but wait you have that risk of the rocket and the fuel. 5555555

 

The Thai Government have already mandated that all those types of commercial activities will be restricted to the EEC on the Eastern Seaboard, and not permitted elsewhere in the country. 

 

A long, long way to go before that is the reality, but that is the vision.