bit-tech.net

Intel chemical leak injures 43 workers

Intel chemical leak injures 43 workers

A nitrogen trifluoride leak at Intel's Fab 32 in Arizona has injured 43 workers, 12 of whom needed hospitalisation.

Intel has confirmed reports of a chemical leak at its Arizona facility, the result of which has injured 43 people and hospitalised 12 through nitrogen inhalation.

Intel's two campuses in Chandler, Arizona combine to form the company's second-biggest facility in North America. The company has expanded its fabrication operations in Arizona over the years to the point where it employs an impressive 11,000 people and has a third manufacturing plant under construction at the site. The leak is reported to have occurred at Fab 32, which was Intel's first plant to mass-produce chips on a 45nm process node.

Thankfully, the chemical leak - identified by the Arizona fire department as nitrogen trifluoride, a chemical used during the plasma etching process - was spotted before it could build to serious levels. According to local news outlet Arizona Republic the leak occurred at six in the morning on Saturday, local time, and was discovered when an employee complained of breathing difficulties. The emergency services were contacted, and initial reports from Intel indicated that 24 people were treated locally for eye irritation and breathing difficulties while six were taken to hospital for further observation.

Those numbers rose steadily as they day progressed, with the final figure sitting at 43 treatments and 12 hospitalisations - although none of the injuries are reported as being life-threatening.

Intel spokesperson Jason Bagley has stated that the leak appears to have been the result of an O-ring failing in the gas exhaust system that should pipe the used nitrogen trifluoride out of the facility, with the result that the gas was vented into the plant itself. According to Bagley, production is being transferred to another area of the plant while the fault is repaired and output is not thought to be affected. Sadly, Bagley has been less forthcoming on the status of those hospitalised by the incident.

During use in plasma etching, nitrogen trifluoride breaks down into fluorine and nitrogen - the former of which acts as an active etching agent. Fluorine gas, which can be released as a by-product of the etching process, is extremely hazardous: concentrations above 25 parts per million (PPM) are high enough to cause significant irritation to the eyes, respiratory tract, lungs, liver and kidneys, while concentrations above 100PPM cause potentially fatal damage. Thus far, it is not clear whether the leak was of fluorine gas mixed with nitrogen trifluoride, or just nitrogen trifluoride. The latter by itself is much less irritating, but can lead to a potentially serious condition known as methemoglobinemia in high concentrations.

This isn't the first time Intel's Fab 32 has hit the news for employee safety concerns: back in 2011 seven workers were injured in a solvent fire, while later that same year a further eight suffered smoke inhalation when construction work caused a plastics fire.

14 Comments

Discuss in the forums Reply
theshadow2001 1st July 2013, 10:01 Quote
There should be safety systems in place to monitor air quality all around the plant in the event of such an occurrence.

From what I'm aware of Intel usually take worker health and safety to an almost patronising level. Which makes it surprising then that there was no safety system or it didn't trigger.

Then again my view of America is a more relaxed approach to health and safety.
Cerberus90 1st July 2013, 11:24 Quote
I'm surprised they don't have monitoring stuff too, if they're dealing with potentially fatal gasses (stop sniggering) then you'd think there'd be some system to detect it before employees start complaining of symptoms.

I'm surprised that sort of thing isn't required by health and safety law.
schmidtbag 1st July 2013, 15:18 Quote
I'm not surprised Bagely didn't give more attention to the conditions of the other employees. As long as they weren't starved of oxygen, there's no way any of them would have died. Nitrogen triflouride is just simply irritating but pretty hard to die from. Flourine is a ridiculously dangerous element (the most reactive of them all), but once it's bonded in a way like nitrogen triflouride or even sulfur hexaflouride, it's nearly harmless to encounter. In other words, unless 100% of the chemical dumped inside the facility, I'm sure many of the employees exaggerated their symptoms just to get out of work. Obviously though, there's such thing as too much of anything. You can die of too much oxygen.
Gareth Halfacree 1st July 2013, 15:31 Quote
Quote:
Originally Posted by schmidtbag
I'm not surprised Bagely didn't give more attention to the conditions of the other employees. As long as they weren't starved of oxygen, there's no way any of them would have died.
That's simply not true. Nitrogen trifluoride is significantly more dangerous than you give it credit for: just check out this MSDS safety sheet (PDF warning) for details.
Quote:
Originally Posted by Nitrogen Trifluoride MSDS

INHALATION:
Nitrogen Trifluoride is irritating to the respiratory system. Nitrogen Trifluoride is also toxic by inhalation. Similar to the effects observed for Carbon Monoxide, the toxic effects of Nitrogen Trifluoride are related to formation of modified haemoglobin that prevents the absorption of oxygen in the blood and causes the destruction of red blood cells (hemolysis). Subsequently, cyanosis (development of a bluish skin colour) and chemical asphyxiation may potentially occur after over-exposures to this gas. At low level exposure this effect can reverse, however at high levels of exposure, the toxic effect of Nitrogen Trifluoride may require medical intervention. In the event of hemolysis, impaired function of the kidneys and damage to the liver and spleen can occur (based on animal tests). Inhalation of high concentrations may be fatal. Effects of exposure by inhalation can be delayed.

NOTES TO PHYSICIANS:
Treatment for chemical asphyxiation may be necessary after over-exposure occurs. At high levels of over-exposure and the onset of symptoms of hemolysis, therapeutic intervention may be necessary, including oxygen, methylene blue and exchange transfusion. The occurrence of hemolysis requires careful monitoring for the degree of anemia and the potential for impaired kidney function. Additionally, treatment similar to that for Hydrogen Fluoride over-exposure should be considered. The following information is to assist physicians in the treatment of exposure to fluoride compounds.

For Inhalation Exposure:
Administer 100% oxygen at half-hour intervals for three to four hours for victims of minor inhalation exposure. For serious inhalation exposure, 100% oxygen administration should begin immediately, under positive pressure (<4 cm) for half-hour periods for at least six hours until breathing is easy and the colour of the skin and mucous membranes is normal.

EFFECT OF MATERIAL ON PLANTS or ANIMALS:
Due to the toxic nature of this gas, animals exposed to this product may be killed. Plants contaminated with this product may be adversely affected or destroyed.

CANADIAN WHMIS CLASSIFICATION AND SYMBOLS:
This gas is categorized as a Controlled Product, Hazard Classes A, C and D1A as per the Controlled Product Regulations. Class A: Compressed Gas, Class C: Oxidizer, Class D1A: Poisonous and Infectious Material-Material Causing Immediate and Serious Toxic Effects

You don't get 100% oxygen at half-hour intervals for four hours (and that's for minor inhalation) when the stuff you've been exposed to is of no risk. The Occupational Safety and Health Administration (OSHA) sets a nitrogen trifluoride exposure limit of just 10 parts per million for a reason, y'know...
schmidtbag 1st July 2013, 15:46 Quote
Quote:
Originally Posted by Gareth Halfacree
That's simply not true. Nitrogen trifluoride is significantly more dangerous than you give it credit for: just check out this MSDS safety sheet (PDF warning) for details.



You don't get 100% oxygen at half-hour intervals for four hours (and that's for minor inhalation) when the stuff you've been exposed to is of no risk. The Occupational Safety and Health Administration (OSHA) sets a nitrogen trifluoride exposure limit of just 10 parts per million for a reason, y'know...

Interesting, what you showed has a higher health risk than what this says:
http://www.mathesongas.com/pdfs/msds/MAT16650.pdf

Regardless, I'm aware it's not actually safe but it isn't fatal unless it's mostly all you're breathing in, but like I said, the same can be said about almost any gas. I'm assuming it's heavier than molecular oxygen, in which case you can "drown" in it if you're not careful.
Gareth Halfacree 1st July 2013, 15:55 Quote
Quote:
Originally Posted by schmidtbag
Interesting, what you showed has a higher health risk than what this says:
http://www.mathesongas.com/pdfs/msds/MAT16650.pdf
I just picked the first likely-looking file that came up on a Google search for "nitrogen trifluoride coshh." The file you link isn't exactly kind to nitrogen trifluoride, though, as this extract from page one shows:
Quote:
Originally Posted by Matheson Gas Nitrogen Trifluoride MSDS
SHORT TERM EXPOSURE: harmful if inhaled, irritation, nausea, vomiting, difficulty breathing, irregular heartbeat, headache, drowsiness, dizziness, bluish skin color, lung congestion, blood disorders, convulsions, coma
Doesn't sound like my idea of fun, and nor does it sound "nearly harmless!"
Quote:
Originally Posted by schmidtbag
Regardless, I'm aware it's not actually safe but it isn't fatal unless it's mostly all you're breathing in, but like I said, the same can be said about almost any gas. I'm assuming it's heavier than molecular oxygen, in which case you can "drown" in it if you're not careful.
Re-read what I posted: the problem is not direct asphyxiation, but chemical asphyxiation. Inhale nitrogen trifluoride, and it can prevent your blood from carrying oxygen efficiently enough. The result: you asphyxiate, even though you've been moved to fresh air. The symptoms can develop anything up to several hours after initial exposure - hence the employees most directly affected by the leak being admitted to hospital for observation, to make sure they don't keel over stone dead. To date, as far as I can tell, nobody has died from nitrogen trifluoride exposure - but it's certainly possible, if you breathe in a decent amount for a short period of time and don't get medical attention.

Sure, it can also kill you in the traditional "displaces oxygen in the air" sense, but so can any heavier-than-air (or lighter-than-air-in-a-sealed-enough-room) gas - but nitrogen trifluoride is toxic. You can get seriously ill - and potentially die, although so far the worst that has ever happened to anyone is a coma - from short-term exposure, even after having been moved out of the leak zone.

EDIT: The LC50 (median lethal concentration) of nitrogen trifluoride is given in the MSDS you linked to as exposure to 6,700PPM for one hour. That's a pretty low concentration, and a pretty short time - and the concentration isn't high enough for traditional asphyxiation to have taken place, meaning that the cause of death would be chemical asphyxiation. Granted, that's based on it killing a rat, mostly because science councils tend to frown on using human test subjects for toxicity studies these days, but still... In contrast, the LC50 for butane is 10,325PPM over four hours' exposure, again for a rat.

This also ignores the point, raised in the article, that the pipe was reportedly extracting used nitrogen trifluoride. The gas is used in plasma etching, where it breaks down into nitrogen and fluorine - and if there was any fluorine in the gas that the workers' inhaled, that's Serious Business. The 6,700PPM/1hr LC50 for nitrogen trifluoride? Yeah, it's 185PPM/1hr for fluorine. Like you said, it's ridiculously dangerous.
Diellur 2nd July 2013, 13:06 Quote
Quote:
Originally Posted by theshadow2001
There should be safety systems in place to monitor air quality all around the plant in the event of such an occurrence.

From what I'm aware of Intel usually take worker health and safety to an almost patronising level. Which makes it surprising then that there was no safety system or it didn't trigger.

Safety systems aren't 100% so even if they're in place, it is possible they don't trigger. This should be managed through the company's safety management system, where the safety systems are tested and maintained regularly - if they're not, the likelihood of them not working increases. Also, we don't know what mods have gone on. If the safety systems were in place for an original design, and this pipework has been re-routed due to a mod, sometimes the safety systems aren't updated in line with this. So in effect, it is possible for there not to be a safeguard in place for the leak. There are plenty of industry examples where safety systems which should have protected against an event have not operated as expected and an accident has occurred. Sadly, this is often due to human error or failures in management systems and generally not directly attributable to a failure of an engineered safeguard.
Quote:
Originally Posted by theshadow2001
Then again my view of America is a more relaxed approach to health and safety.

The US use a prescriptive approach, where they demonstrate safety by complying with a set of requirements; if all those requirements are met or exceeded, then in the eyes of the law they have discharged their duty of care to keep people safe. The UK use a goal-based approach with a principle called ALARP at the heart of it (As Low As Reasonably Practicable). This requires anyone imposing a risk to reduce that risk ALARP, so even if a certain level of risk reduction is achieved by putting in a safety system if further risk reduction is possible without being excessively costly then the duty holder is legally obligated to do so (or he must have a robust rational for why he hasn't). Again, even with the best safeguards in place, a poor management system can lead to the safeguards being eroded over time.
theshadow2001 2nd July 2013, 14:22 Quote
The bottom line is Intel failed their employees. How that happened is anyone's guess. The exact details will never come out.
Cthippo 2nd July 2013, 15:40 Quote
Quote:
Originally Posted by theshadow2001
The bottom line is Intel failed their employees. How that happened is anyone's guess. The exact details will never come out.

While you're probably correct, I've seen enough totally off-the-wall, unpredictable accidents to give the benefit of the doubt until shown otherwise. Things like two completely separate systems failing on an aircraft in a way no one has ever seen before while landing on a short, icy runway, or a multi-ton air conditioner unit sliding down a 3/4" steel cable for 100 yards until it decides to fall off and squish somebody. Most accidents are preventable, but sometimes life just decided you're screwed.
Diellur 2nd July 2013, 16:20 Quote
Quote:
Originally Posted by theshadow2001
The bottom line is Intel failed their employees. How that happened is anyone's guess. The exact details will never come out.

It depends on the regulatory regime in the US. I would expect the opposite, as any industrial accident of this magnitude should trigger an investigation and subsequent report. Whether we see it is another matter; it won't arrive for many months and by then we'll have probably forgotten about this incident. Legally, it is extremely unlikely that this will be just swept under the rug...Intel may try, if they are inclined that way (which i doubt, not in the US anyway) but the authorities get the final say.
theshadow2001 2nd July 2013, 16:21 Quote
Quote:
Originally Posted by Cthippo
While you're probably correct, I've seen enough totally off-the-wall, unpredictable accidents to give the benefit of the doubt until shown otherwise. Things like two completely separate systems failing on an aircraft in a way no one has ever seen before while landing on a short, icy runway, or a multi-ton air conditioner unit sliding down a 3/4" steel cable for 100 yards until it decides to fall off and squish somebody. Most accidents are preventable, but sometimes life just decided you're screwed.

Quote:
This isn't the first time Intel's Fab 32 has hit the news for employee safety concerns: back in 2011 seven workers were injured in a solvent fire, while later that same year a further eight suffered smoke inhalation when construction work caused a plastics fire.

Thats only what has made it to the news.

Their record is bad enough for me to not give the benefit of the doubt.
theshadow2001 2nd July 2013, 21:46 Quote
Quote:
Originally Posted by Diellur
It depends on the regulatory regime in the US. I would expect the opposite, as any industrial accident of this magnitude should trigger an investigation and subsequent report. Whether we see it is another matter; it won't arrive for many months and by then we'll have probably forgotten about this incident. Legally, it is extremely unlikely that this will be just swept under the rug...Intel may try, if they are inclined that way (which i doubt, not in the US anyway) but the authorities get the final say.

Your right that there will be an investigation and a report and hopefully some sort of mitigation strategy. Its unlikely that any of that will make the news though. Which was the intention of what I said. Not some sort of conspiracy theory style cover up.
Diellur 3rd July 2013, 10:56 Quote
Quote:
Originally Posted by theshadow2001
Your right that there will be an investigation and a report and hopefully some sort of mitigation strategy. Its unlikely that any of that will make the news though. Which was the intention of what I said. Not some sort of conspiracy theory style cover up.

Well, you said "the exact details will never come out", which has somewhat sinister and conspiracy-like overtones. :)
Xir 3rd July 2013, 11:46 Quote
Actually, considering the enormous amount of dangerous chemicals and processes used in the Semiconductor industry, the percentage of accidents is rather low.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums