Bob

Shearing of a stainless steel cylinder in normal use is near impossible due to the tensile strength. Testing I did over 10 years ago with steel cylinders filled with over 50 lbs of water dropped at a 60 degree angle directly onto a steel plate from 6 ft damaged the valve but there was no leakage. I finally got it to leak after I increased the weight to over 100 lbs and a height of 7 ft. Even this did not shear.

 

The only incidents of valve shear are from brass. If you look at the Mythbuster video where they try to shear a brass valve, the first guillotine only cracked the valve. They really had to beef up the guillotine to shear it. This is why all the demos on the device was done using brass valves

 

The Bielli valve was an interesting concept but created to other problems. It “resolved” an event that was never going to happen

Eugene

 

 

Sent: Friday, October 02, 2009 1:06 PM

To: SEMI EHS Grapevine

Subject: Re: China Silane Incident

SEMI Environmental Health & Safety Grapevine

 

If is was a dual cylinder cabinet incident through a combined manifold like the 70-80's models and not the dual cabinet as two separately walled but combined cabinets as started in the 90's then this is a really likely.

 

We had an incident in Fab 4 on electronics lane in Carrollton as I stated before which was a release in a dual cabinet where the full one bled into the open pigtail but the second bottle was not yet in place on that one. So it was a gas release and explosion that left the tech deaf on one ear by 50%.

 

So if the second cylinder had been in place on the Carrollton incident and we had direct flame impingement then yes, those cylinders all had pressure relief valve. So was this incident in China where cylinders don't have pressure relief? I seem to recall at one point in the late 70's about discussions where silane would not have a relief in a manner like highly poisonous materials.

 

Or, did the techs in China panic and pull the full cylinder out to get it away from the flame. We saw techs bravely do that kind of thing in Fab 1 in Carrollton to brave the fire to put a relief valve back in as they cooled the flame opeing with CO2 extinguishers and used all the extinguishers in the fab. But as the techs in China pulled the cylinder out and away did they drop it and shear the valve? The internal safety shut off valve like Bielli was advocating would have saved the day in that case.

Bob

 

 

To: SEMI EHS Grapevine <sehs@ehs.semi.org>

Cc: SEMI EHS Grapevine <sehs@ehs.semi.org>

Sent: Thu, Oct 1, 2009 7:17 am

Subject: Re: China Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Hi Eugene,

It sounds like there was no ACV on the cylinder. Do you now if UV/IR detection was part of the detection scheme? Also, is the a web link which i can go to see any photos or incident report you know?

Thanks, Ed

 

"Eugene Ngai" <eugene_ngai@comcast.net>

09/28/2009 03:28 PM

Please respond to

"SEMI EHS Grapevine" <sehs@ehs.semi.org>

To "SEMI EHS Grapevine" <sehs@ehs.semi.org>

Subject China Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Based on additional reports, the silane incident I reported on a few weeks ago, did not occur as I described. I had assumed that the siilane was released unignited and suddenly exploding. It appears that a full cylinder (10 kg) was being placed into a 2 cylinder gas cabinet, something happened to cause the pigtail to release silane from the cylinder that was online. The flames impinged directly onto the cylinder from the pigtail. This heated the cylinder for some period of time and it ruptured

Eugene Ngai

Chemically Speaking LLC

 

China Silane Incident 1 2009-10-01

 

This was a manual cylinder valve. There is no public incident report other than the news media reports which were not accurate. I am relaying facts that I have been given/uncovered to insure that there is no misunderstanding as to what occurred.

 

This incident is significant, in my 35+ years of compressed gas industry experience, there has never been a silane cylinder rupture.

Eugene Ngai

Chemically Speaking LLC

 

 

Sent: Thursday, October 01, 2009 8:18 AM

To: SEMI EHS Grapevine

Cc: SEMI EHS Grapevine

Subject: Re: China Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Hi Eugene,

It sounds like there was no ACV on the cylinder. Do you know if UV/IR detection was part of the detection scheme? Also, is the a web link which i can go to see any photos or incident report you know?

Thanks, Ed Pryor

 

"Eugene Ngai" <eugene_ngai@comcast.net>

09/28/2009 03:28 PM

Please respond to

"SEMI EHS Grapevine" <sehs@ehs.semi.org>

To "SEMI EHS Grapevine" <sehs@ehs.semi.org>

Subject China Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Based on additional reports, the silane incident I reported on a few weeks ago, did not occur as I described. I had assumed that the siilane was released unignited and suddenly exploding. It appears that a full cylinder (10 kg)was being placed into a 2 cylinder gas cabinet, something happened to cause the pigtail to release silane from the cylinder that was online. The flames impinged directly onto the cylinder from the pigtail. This heated the cylinder for some period of time and it ruptured

Eugene Ngai

Chemically Speaking LLC

 

Correction on Silane Incident Information 2006-02-15

 

To All

 

I had cited in one of my Silane postings on an gas cabinet incident that I thought had occurred at IBM, Zurich in 1978. Upon further investigation of this, it appears that over these years, my memory merged

the following events together.

 

There was a tragic Silane cabinet explosion with one fatality in 1976 in Germany. (Company unknown) This caused the German government to conduct release studies (Conrad report) which determined that Silane could be released without ignition.

 

IBM at that time took the lead on funding many additional studies (Hazards Research, Battelle) which formed the design basis for RFO and ventilation.

 

I apologize for the error. Please let me know if this error is in any other document.

Eugene Ngai

 

Organic Silicon pyrophoric liquids 2009-01-26

 

Prof Fthenakis

I don't have any physical or chemical data for either silane. The total heat released should be able to be estimated assuming full oxidation to H2O and SiO2.

 

My recommendation for control of pyrophoric liquids would be the same as that for the pyrophoric water reactive organometallic liquids like TMAl or DEZ. Dry sand or dry vermiculite. These will absorb and crust over with an oxide layer. The challenge is to carefully stir the pile to react the absorbed material

Eugene Ngai

 

-----Original Message-----

Sent: Monday, January 26, 2009 3:07 PM

To: SEMI EHS Grapevine

Subject: Organic Silicon pyrophoric liquids

SEMI Environmental Health & Safety Grapevine

 

Dear grapevine members,

Does anyone have info on safety issues related to cyclopentasilane/hexasilane, liquid compounds that  are spontaneously combustible in air and the presence of water? We want to calculate how much energy would be generated in the worst case release from a lab flask and what are appropriate quenching agents if there is a spill.

Thank you,

Vasilis

 

Root Cause Finding - Silane Leak 2010-02-03

Liew

 

The leak integrity of a DISS connection relies on the finish of the polished toriod bead which deforms the nickel gasket as it is torque. Over time the bead can become damaged by fine SiO2 particles. Gas suppliers typically have a rigorous visual inspection program of the bead as part of the QA program. Have you used a 20X magnification to inspect the bead? The gasket?

 

You are using a helium mixture for leak checking, is it for pressure decay or are you also using a electronic leak detection device to sniff the connection?

How loud was the pop? Was there any damage?

Eugene Ngai

Chemically Speaking LLC

 

 

Sent: Wednesday, February 03, 2010 5:16 AM

To: SEMI EHS Grapevine

Subject: RE: Root Cause Finding - Silane Leak

SEMI Environmental Health & Safety Grapevine

 

Hi Eric,

I remember that we had an issue with identifying correct torque settings for pigtail connections on ion implanters at one time - could this be a factor here?

Best regards

John

 

Sent: 03 February 2010 08:54

To: SEMI EHS Grapevine

Subject: RE: Root Cause Finding - Silane Leak

SEMI Environmental Health & Safety Grapevine

 

Hi Eric,

Thanks for the input. For Silane gas cabinet we use 5%He/N2 to do pressure holding test for 3 minutes at 110 bar. Silane gas cylinder pressure is 75 bar.

 

 

Sent: Wednesday, February 03, 2010 3:13 PM

To: SEMI EHS Grapevine

Subject: Re: Root Cause Finding - Silane Leak

SEMI Environmental Health & Safety Grapevine

 

Liew Teck Liong ,

I don't have a definitive way to include either of your candidate causes, but it may be possible to eliminate "probable root cause 1" by:

 

a) cleaning the pigtail end carefully and inspecting it.

b) if there's no apparent defect, try making a connection with it and seeing if it seals properly. If it does, I'd eliminate "probable root cause 1", as the sealing surface is not likely to heal itself.

c) if the sealing surface is defective, it may be that the defective surface caused the leak, or that the surface was damaged by whatever caused the leak (such as a small, hard particle) or the leak itself.

 

It is possible that the gasket was defective. However, I would be inclined to consider the human error of improperly making the connection at least as likely as the gasket being defective.

 

On a related note, how (if at all) do you know that the gasket was replaced when the cylinder was

last changed?

 

I suggest you consult the part manufacturer's literature or the manufacturer directly for recommendations on the service life of the connector. It seems very unlikely, though, that it would be intended to survive only 41 make/break cycles. (I don't have a "place" with silane cylinders, so I can't answer your question  directly.)

 

Leak tests of piping to be used at pressures above atmospheric should be conducted at, as a minimum, the maximum foreseeable service pressure. (Some leaks do not exist at low pressures and you will not find them unless you subject the system to the use conditions.) Various codes and practices specify testing at higher pressures, such as 150% of service pressure.

In this case, the maximum foreseeable service pressure is the pressure of the silane in a full cylinder. If you are testing with He or an He mixture, it should not take more than a few minutes. If you are testing by pressure decay, however, you need to calculate the time based on the precision of your pressure gauge, the pressure, the volume, and the acceptable leakage rate.

 

If I can be of further assistance, please let me know.

TIA,

Eric

 

 

At 22:12 2/2/2010, you wrote:

SEMI Environmental Health & Safety Grapevine

 

Hi,

A Silane leakage detected at the pigtail connecting to a gas cylinder (standby) which installed to a gas  cabinet 3 days ago. A “pop” sound was heard and later the investigation found burnt mark (white powder) on the nickel gasket surface facing the pigtail side (gas panel), same gasket but surface facing the cylinder connection is clean.

 

Probable root cause 1 ~ pigtail toroid wear and tear off (41 cylinders changed)

Probable root cause 2 ~ gasket defect (EP grade)

 

How to confirm and prove which one is the actual root cause?

What is the maintenance criteria and frequency of change for the pigtail at your place?

What is the pressure and duration set to test the connection after new Silane gas cylinder installed?

 

Thanks & Regards

X-FAB Sarawak Sdn. Bhd.

 

Silane Accidents 2008-12-13

 

No, but do you have the Mostek gas cabinet explosion which left the technician deaf in one ear? The fab manager at the time was Dr. Reynold Kelm I believe as this was in the early 1980's. The facilities Mgr was John Smith who later started a silicon valley company I believe. This was the company started by the late Vin Prothro and also L J Sevin. It was a dual cylinder connection and one cylinder was online and the other was removed. The single valve isolating the second cylinder most likely was leaking and so

the silane came out the unattached pigtail and detonated when the cabinet door was opened.

 

 

-----Original Message-----

From: Ngai,Eugene Y. <NGAIEY@airproducts.com>

To: SEMI EHS Grapevine <sehs@ehs.semi.org>

Sent: Fri, 12 Dec 2008 11:34 pm

Subject: Silane Accidents

SEMI Environmental Health & Safety Grapevine

 

I am continuing my study of the root causes of silane gas cabinet explosions in the last 40 years, I am trying to locate the details of the Hitachi, Masashi, Japan, Dec. 13, 1990 incident which caused 3 injuries and 1 fatality. Does anyone have the details?

Eugene Ngai

 

Silane Cylinder Storage Temperature 2008-10-22

 

There is not a safety concern due to the temperature swing in 12 hours. Internationally, gas cylinder fill limits are based on a temperature of 65 C. We have facilities in Arizona where the outside temperature can get up to 50 C and cool to 35 at night.

 

Our safety practice is not to have a pyrophoric gas (Silane) in the same cabinet as a toxic (B2H6) or toxic – flammable gas (DCS)

 

The increased temperatures will cause a greater degradation of the B2H6 over time. This will not cause a safety problem since the full decomposition pressure is taken into account when the mixture is made.

Regards

Eugene Ngai

 

 

Sent: Wednesday, October 22, 2008 3:24 AM

To: SEMI EHS Grapevine

Subject: Silane Cylinder Storage Temperature

SEMI Environmental Health & Safety Grapevine

 

Dear all,

Is there any safety concern or requirement for Silane, 3% B2H6, DCS gas cylinder which sit in the cabinet (with ventilation) at ambient temperature (25oC @ 2am – 36oC @ 2pm)?

 

Gas cabinet is located in a small room with metal collapsible wall (explosion venting purpose), fresh air supply not chilled, no direct sunlight but facing west.

Regards,

Liew Teck

 

 

Silane Double Containment 2007-01-12

 

We continue to be concerned about Silane safety for the reasons Bob mentions, especially since there are many new types of users with the PV industry. As such we have championed the training of the users in Asia with Silane Seminars in Taiwan and one coming up in Korea both of which are standing room only (480 and 220). We will host one in China late 2007 and hope to have one in Southeast Asia and Europe if we can workout the details. I have spent the last 2 years testing and reviewing many of the earlier studies on the unpredictability of releases. I hope to present on some of this at Korea and SESHA in April.

 

The value of coaxial tubing has been a hotly debated issue. The original intent was to provide a barrier should the inner tube corrode from the product going through it or from the environment surrounding it. It does not however prevent the types of accidents in which someone cuts the wrong line, this happened with an Arsine and a Silane line recently. If you should choose to use coaxial tubing the monitoring is important as Bob indicates.

Eugene Ngai

 

 

Sent: Friday, January 12, 2007 10:29 AM

To: SEMI EHS Grapevine

Subject: Re: Silane Double Containment

SEMI Environmental Health & Safety Grapevine

 

Hello and thanks - I may be the safety renegade but I advocate double containment of silane and fought for it at more than one fab. I have been through the the design of 10 fabs and the building or complete rennovation of 9 (one fab was cancelled).

 

But there is a specific design that I advocate and it takes some money to do it. The annular space is  welded closed at each end. The annular space is pressurized with nitrogen below the operating pressure of the silane line when flowing gas. A pressure transducer is installed to sense the pressure in the annular space and connected to the gas control system for alarm. The alarm sounds if the pressure rises which would signal a leak of silane into the annual space or if the gas pressure falls which indicates a leak or low supply of the nitrogen.

 

I went through that period of time from 1978 to 1983 where the silane would explode for all sorts of reasons. One of the ones at fab 4 in our south building was when an R&M tech opened a door to a gas cabinet and the silane cloud inside exploded which left him 50% deaf. Others were silane roof top explosions in auto-ignition burn boxes. One of the pieces of metal flew down from the roof and landed next to Dr. Kelm. Elsewhere in the industry there were silane explosions in ducts. It also exploded in our vacuum pumps and routinely, several times a day, exploded inside of our vacuum collector exhaust systems but which were very heavy wall steel pipe and so no damage. You could stand under the pipe in the subfab and hear it go brrriiinngg. brrrinnng, brrrinng as it exploded inside the pipe.

 

Then there was the video tape of the silane gas releases in a controlled setting and it would explode. This would match our gas cabinet experience above.

 

So the stuff when it leaks is very unpredictable. It can simply dissipate or it can form pockets even in dead spots in gas cabinets. When it accumulates in pockets of gas it can explode.

 

What I recognized and could talk management into by the actual events stated above is that IF you do get a silane line leak then the gas may and can accumulate and explode.

 

So - Do You Feel Lucky with your single line - Do You? as Clint would say.

thanks - Bob

 

 

-----Original Message-----

To: sehs@ehs.semi.org

Sent: Thu, 11 Jan 2007 10:07 AM

Subject: Silane Double Containment

SEMI Environmental Health & Safety Grapevine

 

All,

It is my understanding that a lot of facilities are no longer installing double contained lines on Silane and I am wondering what the thought process has been on making that change? I know it's not required by code and is usually directed by the Authority Having Jurisdiction, but I also know some facilities out there have double contained lines and have no chosen for new installations to not double contain. Any background information from anyone would be appreciated.

 

Also what type of monitoring are some of you doing? There has been discussion whether you just do MDA monitoring or just IR/UV monitoring or both.

Thanks,

Laurie

 

Silane fire video 2008-11-07

 

The Silane videos you may have seen were probably from one of my presentations. I had many videos of poppers, RFO releases, fires, explosions, etc The videos that Air Products has made during our release

testing programs (2006-now) are not publicly available for liability reasons. There are also videos that were produced by CGA as part of the tube trailer release testing in 1996. This was paid for by 10 companies ($1 million) and was the basis for CGA G-13, Silane. This is available along with the report from CGA at a fee. The most memorable videos are the ones from Hazards Research done in the early 1980's for IBM. Numerous gas cabinet and duct explosions. Many are of poor quality since they are VHS copies of 8 mm film. IBM distributed these to a few Semiconductor and Gas companies in the mid 1980's as part of the silane safety effort which they championed. You would have to ask IBM for copies since they own the rights to them. Samir Shaban, Intel also conducted numerous release studies in the 1990's on confinement, PPE and Scrubbing. Some of his videos may still be around.

Eugene Ngai

 

 

-----Original Message-----

Sent: Friday, November 07, 2008 3:23 PM

To: SEMI EHS Grapevine

Subject: Silane fire video

SEMI Environmental Health & Safety Grapevine

 

Greetings:

We're trying to locate a video on silane fires. I saw one several years ago, possibly at a SESHA  conference.

 

We've heard of one perhaps sponsored by Bell Labs, 1989 time frame...

 

Any help appreciated

Jim

 

Silane Gas Rooms 2010-01-27

 

Rowena

In the US grade means ground level. Below grade means a open pit or basement where a heavier than air gas can accumulate. Silane is slightly heavier than air so it can accumulate in areas below grade.

 

The high pressure gas regulations in Japan have a secondary toxicity definition which includes anything with a TLV less than 200 ppm. Some other countries have adopted this. I have worked to change this with some success in Korea and Taiwan.

Eugene Ngai

Chemically Speaking LLC

 

Sent: Wednesday, January 27, 2010 8:30 PM

To: SEMI EHS Grapevine

Subject: RE: Silane Gas Rooms

SEMI Environmental Health & Safety Grapevine

 

I think when looking at silane, there is a need to look at it from a health point of view (toxic) and a safety risk point of view (pyrophoric).

 

1. Safety – result in physical damage e.g. fire, property, people injury

2. Health – result in body damage through exposure to a substance e.g. usually something toxic

 

(not quite the best definition but just for simplification and ease of understanding)

 

Countries that are classifying the gas as toxic because it falls under the definition of toxic materials set by that particular country. Also, most countries would be “piggybacking” either US, European or Japan standards when it comes to this. Why? I can imagine it saves on resources and time, as well as coming from a credible source. What the countries are different are the control methods used to reduce exposure to the toxic material.

 

The term “indoor storage” can have different interpretations towards the control measures.

 

If it’s from a safety point of view, I assume then it would mean a separate room outside of the main fab building or outside of the subfab but still in the fab building. It becomes a question whether it’s going to have a huge impact if there was a leak in one of those rooms.

 

If it’s from a health point of view (toxic), I assume that it may mean that it needs to be segregated from people either through a cabinet with some interlocks and ventilation features. The question of physical location doesn’t come into play.

 

As for “making the grade”, it probably just needs to be defined better. The question on whether how many makes the grade or not, that’s the concern of the factory and not a standard development team (or however it’s called). Better to have a standard than none. If a factory refuses to become compliant to changes to a standard, they have to bear the risk of such a decision. So why worry if there are factories compliant or not? That shouldn’t stop the development of a standard. Else no one would stop worrying

and nothing gets done when the cows do come home (if ever).

Rowena Michelle

 

 

From: Eugene Ngai [mailto:eugene_ngai@comcast.net]

Sent: Wednesday, January 27, 2010 10:28 PM

To: SEMI EHS Grapevine

Subject: Silane Gas Rooms

SEMI Environmental Health & Safety Grapevine

 

As you might be aware, we have been working hard to update ANSI/CGA G13 Silane and Silane Mixtures. A lot of industry statistics on failure rates is being compiled, release calculations and dispersion analysis being done.

 

One question that has come up from another gas association is

 

In G-13, 7.2.2 we state “Silane shall not be stored in locations below grade.” In Asia this type of storage is allowed in some countries. While this is not the best practice, some countries still classify silane as toxic, therefore requiring indoor storage. I know that many gas rooms are in the subfab. How many of these truly meet the definition of below grade and not just inside at grade?

Eugene Ngai

Chemically Speaking LLC

 

Silane Incident 2009-09-17

 

It was a chinese vendor. I don’t know the name

Eugene

 

Sent: Thursday, September 17, 2009 9:44 PM

To: SEMI EHS Grapevine

Subject: Re: Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Thanks Eugene.

Do you happen to know who the gas vendors was?

Mike

 

In a message dated 9/17/2009 8:35:09 A.M. Pacific Daylight Time, eugene_ngai@comcast.net

writes:

SEMI Environmental Health & Safety Grapevine

Mike

Thanks, It’s amazing that after an explosion like that, that the fire dept was not called until 4 hrs into the incident.

 

 This leak sounds like a small cylinder valve thread leaker. It was damaged by impact, most likely the cylinder cap was not on. A 50 cc/min leak will produce a good size flame. A 10 kg cylinder has 7,500,000 cc. A leak that will last 104 days! If it doesn’t plug with SiO2 which it often does. The best ER is to vent the cylinder or cascade it.

Eugene Ngai

Chemically Speaking LLC

 

 

Sent: Thursday, September 17, 2009 11:20 AM

To: SEMI EHS Grapevine

Cc: MikeWEHS@aol.com

Subject: Re: Silane Incident

SEMI Environmental Health & Safety Grapevine

 

Ed/SEMI Grapevine,

Here's additional information that was in the newspaper.

 

Silane Explosion in China, Changsha

 

On 17th August 2009, a Silane leak was reported from Hunan Solar Company, located in Changsha, Hunan. As it has a possibility of chain-reaction explosion, three fire engines were despatched.

 

The Hunan Solar Company was a 3 storey factory, opposite sides of the factory walls are window and had an area of 1000m2. There were 35 SiH4 cylinders, each 10 kg.

 

The leak happened at 2pm and a technician was working on that cylinder (no details of how the leak occurred was provided in the news report). After the leak, the technician evacuated and huge explosions followed. The factory was badly damaged by the explosions and other facilities within were thrown off their place. All the windows at the sides were shattered. A large part of the properties were lost but luckily nobody was injured.

 

After the explosion, the factory team went back into the site and found a 10m high wall at the west side was blown apart exposing the inside of the factory. All the SiH4 cylinders were lying on the floor and 2 SiH4 cylinders were found leaking. The factory team tried to contain the leak but minor explosion still happened. After 4 hours of failed attempt, the fire brigade was called.

 

The huge damage was caused by only 1 unit of 10kg SiH4 cylinder. And yet after 4 hours, the situation is still not under control, other cylinders were still leaking and explosions were still happening. It will be unimaginable if all the 35 cylinders were to explode. The fire brigade went into the factory. The first action that could be taken was to cool the cylinder to prevent the cylinders from heating and exploding. After about 2 hours, they managed to move the leaking cylinders and dump them into a nearby pond.

 

According to the supplier who was on site, the leaking cylinder will take about 3 months for it to leaking naturally and completely. They suggested to trans-fill to good cylinder on site (!?).

Michael

 

 

In a message dated 9/15/2009 3:18:18 P.M. Pacific Daylight Time, eugene_ngai@comcast.net writes:

SEMI Environmental Health & Safety Grapevine

 

On Aug 17th a PV facility in Changsha, China experienced a significant incident with silane. They were using 10 kg cylinders and for reasons unknown they experienced a leak. An immediate evacuation occurred which probably prevented serious injury or fatalities. The silane then ignited and detonated. The explosion blew out windows on all sides of the 1,000 m2 (10,000 ft2) 3 story metal frame building. The metal walls on the process side was blown out. There was no fire afterward.

 

35 cylinders of silane in the area were knockedover by the blast. Two started to leak. One leak was sealed by facility personnel. The second leak could not be stopped and was placed in a field 1000 m away by the fire dept.

 

This information came from 2 chinese news articles. If anyone has anything further to add I would appreciate the information

Eugene Ngai

Chemically Speaking LLC

 

Silane Incident 2005-11-29

 

I don't have the answer to either question since it was not an Air Products cabinet or cylinder, so we don't have the details. We were asked to assist in some of the cleanup activities and preliminary investigations. The theory was based on summaries of the damage and the pictures of the cylinders, room and cabinets. Eugene

 

 

Sent: Tuesday, November 29, 2005 8:13 AM

To: SEHS Grapevine

Subject: RE: Silane Incident

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Eugene-

Thanks for distributing the details of this tragic incident.

 

I'm curious, do you know if the cabinet was outfitted with a gas detector and was that in an alarm condition?

 

-Thanks!

Beth

 

 

Sent: Monday, November 28, 2005 10:47 PM

To: SEHS Grapevine

Subject: Re: Silane Incident

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

Hi Eugene,

Do you know where the leak originated?

Thanks,

David

 

 

----- Original Message -----

From: Ngai,Eugene Y.

To: SEHS Grapevine

Sent: Monday, November 28, 2005 9:12 PM

Subject: Silane Incident

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

On the morning of Nov. 23, 2005, a tragic accident occurred at a Semiconductor Fab located in the Tainan Science Industrial Park, Taiwan causing severe damage to the facility and a fatality. While the investigation is continuing, it has been theorized that the following may have occurred.

 

An operator was responding to a local gas cabinet alarm in a 2 cylinder Silane gas cabinet. We beleive that silane was leaking and not burning in the cabinet. When he opened the door, the Silane/Air mixture exploded and blew out the gas cabinet door. The explosion and door threw him against a 2 cylinder silane gas cabinet immediately behind him. It also threw the 2 Silane cylinders out of the cabinet onto the room floor severing the pigtails for both silane cylinders from the gas panel. Both cylinders emptied their contents into the room severely burning him. The cylinders were found with their valves pointed at the operator only a few feet away The fire caused other Silane and Ammonia cylinders in the gas room and cabinets to also empty their contents, adding to the fire which burned for an hour before firefighters could bring it under control.

 

In the last 25 years there have been similar incidents where Silane was accidentally released into a gas cabinet initially without ignition. A flow disturbance or other event then caused the mixture to detonate. These have caused a number of fatalities and severe injuries. Testing by Hazards Research,

Factory Mutual, Intel and others have demonstrated that Silane under certain conditions can be released without ignition and then suddenly detonate. This can even occur with cabinets being exhausted at fairly high exhaust rates (500 cfm, 14,158 lpm). In one experiment the cabinet detonated 5 seconds after the

flow of Silane was stopped and the cabinet had an estimated 2 complete air exchanges.

 

The incidents and the studies created considerable concern within the Semiconductor and Gas industries as well as the Fire Service. In the late 1980’s, Sematech Inc funded a series of studies to further understand the release behavior of Silane and to review the results of the earlier studies. A series of

Technical Symposiums were conducted throughout the US in the mid 1990’s to educate the users, suppliers and the fire service. The Compressed Gas Association (CGA) and the member companies were active participants in these efforts As part of this effort, CGA spearheaded the risk assessment and

release testing of bulk Silane ISO Modules. The safety guidelines developed by these studies have been incorporated into the CGA Silane pamphlet P-32 which is currently undergoing ANSI balloting and will become CGA G-13.

 

This is a reminder to all Silane users and suppliers to remain aware of the hazards associated with Silane

Eugene Ngai

 

Silane Incident 2005-11-29

 

Jay

Good questions.

 

We don't have any more details of the alarm or detection since it was not our cylinder or cabinet. We are helping with the investigation and cleanup.

 

The response to a Silane leak in a cabinet is more difficult since a lot will depend on the

 

design of the cabinet

visual indications

alarms, levels

was the cylinder just changed

Is there a pneumatic cylinder valve or is it on the pigtail?

 

Sematech concluded that one of the most important design criteria for a cabinet was to insure that any potential leak point is rapidly diluted to <1% conc. This requires a specially designed cabinet with baffles and inlets.

 

This would be an excellent topic for SERF to take on. R. Barnes will be posting this on the site for discussion

Eugene

 

 

Sent: Tuesday, November 29, 2005 7:37 AM

To: SEHS Grapevine

Subject: RE: Silane Incident

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

Eugene,

Thanks for sharing this incident.

 

Do you know any more about:

 

1. the cause of the “local gas cabinet alarm”?

2. the type of detection that was being used?

 

Also, what would your advice be about responding to a silane gas cabinet leak?

 

Specifically, if a leak has been detected should knowledgeable responders open a cabinet to  troubleshoot?

 

Thanks again for getting this info out quickly.

Jay

 

 

Sent: Monday, November 28, 2005 9:13 PM

To: SEHS Grapevine

Subject: Silane Incident

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

 

On the morning of Nov. 23, 2005, a tragic accident occurred at a Semiconductor Fab located in the Tainan Science Industrial Park, Taiwan causing severe damage to the facility and a fatality. While the investigation is continuing, it has been theorized that the following may have occurred.

 

An operator was responding to a local gas cabinet alarm in a 2 cylinder Silane gas cabinet. We beleive that silane was leaking and not burning in the cabinet. When he opened the door, the Silane/Air mixture exploded and blew out the gas cabinet door. The explosion and door threw him against a 2 cylinder silane gas cabinet immediately behind him. It also threw the 2 Silane cylinders out of the cabinet onto the room floor severing the pigtails for both silane cylinders from the gas panel. Both cylinders emptied their contents into the room severely burning him. The cylinders were found with their valves pointed at the operator only a few feet away The fire caused other Silane and Ammonia cylinders in the gas room and cabinets to also empty their contents, adding to the fire which burned for an hour before firefighters could bring it under control.

 

In the last 25 years there have been similar incidents where Silane was accidentally released into a gas cabinet initially without ignition. A flow disturbance or other event then caused the mixture to detonate. These have caused a number of fatalities and severe injuries. Testing by Hazards Research,

Factory Mutual, Intel and others have demonstrated that Silane under certain conditions can be released without ignition and then suddenly detonate. This can even occur with cabinets being exhausted at fairly high exhaust rates (500 cfm, 14,158 lpm). In one experiment the cabinet detonated 5 seconds after the

flow of Silane was stopped and the cabinet had an estimated 2 complete air exchanges.

 

The incidents and the studies created considerable concern within the Semiconductor and Gas industries as well as the Fire Service. In the late 1980’s, Sematech Inc funded a series of studies to further understand the release behavior of Silane and to review the results of the earlier studies. A series of

Technical Symposiums were conducted throughout the US in the mid 1990’s to educate the users, suppliers and the fire service. The Compressed Gas Association (CGA) and the member companies were active participants in these efforts As part of this effort, CGA spearheaded the risk assessment and

release testing of bulk Silane ISO Modules. The safety guidelines developed by these studies have been incorporated into the CGA Silane pamphlet P-32 which is currently undergoing ANSI balloting and will become CGA G-13.

 

This is a reminder to all Silane users and suppliers to remain aware of the hazards associated with Silane

Eugene Ngai

 

 

Silane Leak Check 2006-02-21

 

Most of the gas cabinets we design, install and/or operate test for leaks based on pressure decay over time The sensitivity is a function of time and dead volume tested. More importantly the leak check must be done at the full Silane cylinder pressure. Incidents have occurred when a user leak tested at a much lower pressure and a leak occurred when the Silane cylinder was opened.

 

Some locations also use a outboard Helium leak check using a handheld thermal conductivity detector. An inboard Helium leak check for a typical gas cabinet is too expensive to install and can sometimes give

the user the wrong result. Sometimes the vacuum for an inboard leak test can create enough of a seal to pass the leakiest. When it is pressurized the leak occurs.

Eugene Ngai

 

-----Original Message-----

Sent: Tuesday, February 21, 2006 10:23 AM

To: SEHS Grapevine

Subject: Silane

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Hi,

by reading all the information which have been exchanged since end of November, almost everybody could build a safe silane gas cabinet. We can also protect our workers with Nomex and aluminized PPE.

What I am missing is the procedures or BKMs for handling all that safe equipment.

 

Only the procedures make safe equipment really safe.

 

Here' s the question:

 

How do you leak check your gas cylinder gaskets?

Helium leak check, pressure drop over time, or just trust and believe?

 

Looking forward hearing from you!

regards

Roland

 

Silane Mixtures 2008-08-19

 

Thanks -

 

We had two explosions on the roof of our south building in a burn box made of sheet metal into which a 1/4 line vented from a gas cabinet. The sheet metal disentigrated each and landed on the lawn by the building. It was 100% silane.

 

We had an explosion in a gas cabinet in fab 4 when a tech accidentally opened a valve on a line not connected to a bottle in a dual gas cabinet but the first bottle was still connected. So the release was through a 1/4 inch line. The tech was made deaf in one ear from the explosion. It was 100% silane.

 

Low flow releases of 100% silane through threaded fittings would burn nicely and evenly on the lines on standoffs in fab 1.

 

At Samsung we had the 100% silane vent above the roof of the gas building and the high pressure release through the line which gave a puff as it burned on slowing of the velocity when shutoff.

 

I have no experiences with 20/80 silane/N2 mixtures.

 

thanks - Bob

 

 

-----Original Message-----

From: Ngai,Eugene Y. <NGAIEY@airproducts.com>

To: SEMI EHS Grapevine <sehs@ehs.semi.org>

Sent: Tue, 19 Aug 2008 8:18 am

Subject: Silane Mixtures

SEMI Environmental Health & Safety Grapevine

 

In the last 2 days, Prof Chen and I continued our Silane release testing experiments in Taiwan. So far we have conducted release testing through a variety of RFOs, 1/8” tubes and ¼” tubes at pressures varying from 150 psig to 1400 psig. Of the 40 plus releases we were not able to obtain immediate ignition only puffs when the pressure decayed.

 

An interesting result was a 20% Silane/Nitrogen mixture would autoignite at low release velocities. As the velocity increased it went out. It would not reignite as the pressure was reduced. If we allow it to just flow at low flow rates it burns very nicely. I understand others in the US and Japan have observed this in the 70’s and 80’s

 

Has anyone experienced this?

Eugene Ngai

 

 

 

 

Silane PPE 2008-11-10

 

 

Eugene Ngai

 

 

Sent: Friday, November 07, 2008 8:35 PM

To: SEMI EHS Grapevine

Subject: Two Items

 

Hello Everyone:

I would be appreciative if I could get some feedback on two items:

 

1.     what procedure do you use to test showers and eyewashes that are located within a fab or cleanroom? I am concerned in particular about making sure that water doesn't run all over the fab floor and more importantly, not contaminating the area with the resultant aerosols. We have tried using the "funnel" type containment device but it is open at the top so it will not contain the generated water mist. Our pull down eyewash stations do not have a catch basin that will contain the discharged water - has anyone with the same type of unit found a successful method for testing?

2.     what type of PPE do you require for silane cylinder changes?

 

Thanks for your assistance.

Regards,

Faith

 

Silane PPE 2005-12-06

 

Jay

Aluminized suits hold up better to radiant heat and direct flames. The issue we have for routine use is that they are not as flexible. You should note that Nomex is designed only for flashover, any direct flame impingement for a second or more will char it. We had an employee injured when he stuck his arm in front of a Silane flame to turn off a valve. The flame burnt through in a few seconds.

 

The user should also note that Nomex will conduct the radiant heat very quickly.If the user wears a synthetic fiber underneath it will melt to their body, only natural fiber (wool, cotton, etc) should be worn.

 

We also use leather welders jackets in certain applications

Eugene Ngai

 

 

Sent: Wednesday, November 30, 2005 9:17 AM

To: SEHS Grapevine

Subject: RE: Silane PPE

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

We use Nomex suits for our silane cylinder changes

Jamie

 

 

Sent: Wednesday, November 30, 2005 7:03 AM

To: SEHS Grapevine

Subject: Silane PPE

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

Thanks to Eugene’s report, we have been reviewing our silane procedures.

 

In regard to PPE we have used aluminized fire resistant suits. It appears aluminized suits are designed for environments with high radiant heat.

 

 

In reading various documents on the web including a 94 SEMATECH report and Air Products Safetygram, there is reference to flame resistant clothing (e.g. Nomex), but nothing about aluminized suits.

 

What are others using for routine cylinder changing?

Thanks, Jay

 

Silane Safety 2005-12-09

 

The lower limit of Silane flammability has been debated and tested by a number of people. While the current safety data shows Silane LFL to be 1.37%, you should be aware that this is strongly influenced by geometry, temperature, humidity, etc.

 

Under test conditions by Horiuchi he has reported a lower flammability concentration of

 

·         0.64 % in Nitrogen

·         0.17% in Argon

·         0.16% in Helium

 

Addition of Hydrogen increases flammability range to almost all concentrations! A mixture of 0.88% Silane, 1.14% Air Bal. Hydrogen exploded.

 

Horiuchi, S., et al, "Flammability Limits of Monosilane", High Pressure Gas (Japan), Vol. 24, No. 3, 1987

 

Eugene

 

-----Original Message-----

Sent: Friday, December 09, 2005 8:35 AM

To: SEHS Grapevine

Subject: RE: Silane Safety

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

All,

In 1989, Larry Britton of Union Carbide published a paper of, "Combustion Hazards of Silane and its Chlorides." The report is 44 pages and extremely comprehensive. The paper (12b) was presented at

the Loss Prevention Symposium of the AIChE Spring National Meeting in April 1989. Anyone wanting an comprehensive tutorial on silane should try to get the paper either through Union Carbide or AIChE. The paper is copyrighted so I can't post it.

Dave

 

 

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Terry,

We use aluminized flame resistant suits to uncap cylinders, and to make and break connections. Otherwise same procedure as other HPM gases.

 

For the group,

Along the lines of Terry's question about PH3/SiH4 mixture......Is there a point at which the "detonatability" is not an issue. For example, we have 2% and 5% silane in inert gases. I understand that they are pyrophoric up to approx 1.3%, however are they also detonable in those mixtures? Any data or examples of mixtures being detonable?

Thanks, Jay

 

 

Sent: Friday, December 09, 2005 2:51 AM

To: SEHS Grapevine

Subject: RE: Silane Safety

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Hi,

The summary (linked below) is excellent (and an eye-opener), and it seems to single-out silane as the "problem" gas. We use a 65/35 silane/phosphine mix here. Is it likely to have the same unpredictability as "pure" silane? Also I would be interested to know of other facilities' procedures for changing pyrohoric (esp. silane) gas cylinders.

 

Look forward to your reply.

Kind regards,

Terry

 

 

From: Ngai,Eugene Y. [mailto:NGAIEY@airproducts.com]

Sent: 08 December 2005 22:26

To: SEHS Grapevine

Subject: RE: Silane Safety

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

To All

Just a correction on the IBM incident, it was not in 1983. The best I can determine it was 1976 in Germany

Eugene

 

Sent: Wednesday, December 07, 2005 3:07 PM

To: SEHS Grapevine

Subject: RE: Silane Safety

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

I believe there's good summary of this incident and silane info in general in a presentation that Eugene Ngai did that's on Air Products' website at

 

http://www.airproducts.com/NR/rdonlyres/SafetyOverviewPart2431brulyqnrtk

 

kooiqsg.pdf

________________________________

Sent: Wednesday, December 07, 2005 8:26 AM

To: SEHS Grapevine

Subject: Silane Safety

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

To the best of my recollection, that there was another tragic incident in Japan some years ago, I think in a university in Osawa, when a student, responding to an alarm, opened a silane gas cabinet and created a (delayed) ignition and explosion.

 

I am trying to put some statistics and some lessons learned together for the benefit of our university labs where researchers may not have all the sophisticated prevention options that our fabs have, and I will appreciate any related information or links to information sources. The information will be disseminated only to the stakeholders (i.e., researchers using or planning to use SiH4)

 

Thank you

Vasilis

 

-----Original Message-----

From: Ngai,Eugene Y. [mailto:NGAIEY@airproducts.com]

Sent: Thursday, February 17, 2005 11:58 AM

To: SEHS Grapevine

Subject: RE: Yellow smoke from exhaust, is it from silane?

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

Silane under ideal conditions will oxidize to SiO2 in the burnbox and is a white amorphous solid. With the  right filter (HEPA) system this should be removed. If the burn box is not working properly or there are contaminants the Silane is not completely oxidized forming SiOH which is yellow to brown

Eugene Ngai

 

________________________________

Sent: Thursday, February 17, 2005 11:34 AM

To: SEHS Grapevine

Subject: RE: Yellow smoke from exhaust, is it from silane?

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS) GRAPEVINE

 

I haven't seen a bulk silane operation, however some of our process tools run silane, which goes to a burn box, then water scrubbing (for the other chems), then diluted 40000cfm of house exhaust. Our air

permit dictates that our exhaust stacks do not form a plume. We haven't had any problems. I would suggest talking to the abatement device vendor, and they should be able to provide efficiencies and opaqueness guarantees.

 

-----Original Message-----

Sent: Thursday, February 17, 2005 8:59 AM

To: SEHS Grapevine

Subject: Yellow smoke from exhaust, is it from silane?

SEMICONDUCTOR INDUSTRY ENVIRONMENTAL HEALTH & SAFETY (SEHS)GRAPEVINE

 

Hi

A gas supplier told me that if we purge bulk silane into a burnbox, there will be yellow smoke generated from the exhaust. However, an environmental aspect-impact study by the company indicates

that the smoke should be colourless.

 

Anyone has experience on this? Does anyone know of the chemistry behind any yellow smoke in your plant?

 

Regards

Danny

 

Silane Safety Seminar 2010-01-18

 

Continuing the campaign to educate silane users and suppliers, I am happy to announce that plans have been finalized for a silane safety seminar at the Intersolar Munich Exhibition on June 10, 2010. This will be sponsored by REC Silicon and open to anyone preregistered for Intersolar Munich. I have a agenda which I can forward since it is too big for the system This follows the successful the major silane seminars that we held in

 

·         Taiwan 2006

·         Korea 2007

·         China 2007

·         Singapore 2008

·         US 2008

·         Belgium 2008

·         US 2009

 

There were also numerous silane safety presentations provided in other conferences.

 

At this seminar Prof Chen, Kaohsiung First University will present his latest findings in his 4 year testing program on silane autoignition behavior He believes that he is close to determining the conditions for silane to always autoignite. I will also be presenting on major silane incidents that have occurred in the PV industry, as recently as Sept 2009. I also plan on showing a short video clip of a 2009 gas cabinet emergency response which will dramatize the problems that the responder faces even in a small release and fire.

 

This will be another worthwhile seminar

 

Eugene Ngai

Chemically Speaking LLC