[labnetwork] TGMS Emergency Response Cordination

[Hathaway, Malcolm]

Hi Corey,

So, at Harvard, our response is pretty much the same as what Noah described at UPenn.  Any ambient "sense event" is treated like a major leak.  All trigger levels are very conservative (low = 1/2 x the 8-hour TLV, high = 1 x 8-hour TLV).  For ambient leaks, all available ERT members should respond on-campus, even nights/weekends.

One of the points of discussion here has been exactly what the Emergency Response Team will do in various situations.  Our current goal to provide fast, accurate information to the Fire Department in the event of a "gas sense event".  We call it that to be clear that we do the same thing whether there is an actual leak, or just a "apparent leak".  Our challenge has been to maintain a "fighting posture" until we have definitively demonstrated there is no leak.  This can be tricky as there is a tendency to silence the alarms and get everything "settled down" as quickly as possible.

Most of our (relatively few) events have been sensors needing to be changed (1 or 2 times), cross-sensitivity (a couple of times), and maintenance chamber openings.  The cross-sensitivity issues involved IPA triggering a TEOS sensor (we decommissioned the TEOS sensors), and some acetylene triggering an LEL sensor.  The chamber-maintenance issues (and smells) have gone away since we go a hepa/carbon-filtered portable fume abatement "snorkel".  We have our bump-test guys come every four months, and that has worked pretty well keeping the sensors happy.  We also have Cambridge Fire come to visit annually (the whole lot of them, in three big batches).

One nice/bad thing about this being Harvard is that if anything bad happens (or might happen), it is automatically a Big Deal.  So, the Cambridge Fire Dept. and certainly the campus police/EHS take everything really seriously.  In the case of your support organizations not seeing the seriousness of the situation, the best thing is to get them in, maybe even for the captains/chiefs to come in for a 40 min powerpoint show featuring the Best of Eugene Ngai (Silane Guy), blown up buildings, destroyed fabs, and the like, followed by a tour, so they can see what could go wrong.  Just having face time with these folks (and university administrators) helps build mutual respect and cooperative attitudes.  Unless you have industry fabs nearby, I can imagine locals may not have much familiarity with the hazards.

And, as Vince at NIST pointed out, fix those false alarms.  Real leaks last longer than a few seconds, so trigger after two polling cycles (for instance).  Make sure vendors know what triggers gas alarms.

Hope this helps.  Good luck!


Mac Hathaway
Toxic Gas Monitoring Guy
Harvard CNS



________________________________
From: labnetwork-bounces at mtl.mit.edu [labnetwork-bounces at mtl.mit.edu] on behalf of Noah Clay [nclay at seas.upenn.edu]
Sent: Wednesday, July 22, 2015 3:55 PM
To: Corey David Wolin
Cc: labnetwork at mtl.mit.edu
Subject: Re: [labnetwork] TGMS Emergency Response Cordination

Corey,

Here’s our general protocol for the University of Pennsylvania in Philadelphia.  Everyone has agreed to and responds according to this plan.  Below is a from a memo sent last year to our safety group, EHRS (R=radiation), campus police and fire group.

Best,
Noah Clay

Quattrone Nanofabrication Facility
University of Pennsylvania
Philadelphia, PA

*******

-All of our gas cabinets and VMBs can be shut-down remotely by the TGMS.
-All equipment that handles toxic gases can be shut-down or addressed remotely by the TGMS as well.
-In some cases, we don’t shut-down the equipment completely; we shut-down gas pods/boxes attached to the equipment.

*******

We are monitoring hazardous gases and alarming at two levels of detection:

- Level 1: 1/2 TLV or 10% LEL
- Level 2: TLV or 20% LEL

The system also warns us for non-zero concentration in any sensor.  To avoid noise blips and false alarms, any Level 1 or Level 2 detection is considered real (sounds alarms) if it is measured consecutively for three seconds.

Local evacuation is with horns and blue strobes, which are zoned.  For example, if there is a detection of toxic gas in the cleanroom ambient, then horns and blue strobes will be active there, but not in the gas bunkers.

The TGMS is under service contract with two visits annually (gas challenging) and has an active autodialer (analog line, not VOIP).

The autodialer has four outputs (two for exhausted enclosure detection and two for ambient detection):

1. Alarm 1: Level 1 gas detection in an exhaust
Response:
- Shut-down detected gas supply
- Shut-down equipment connected to detected gases
- Contact Nanofab facility staff

2. Alarm 2: Level 2 gas detection in an exhaust
Response:
- Alarm #1 response
- Local evacuation with horns & blue strobes
- Contact PennComm (campus police), EHRS on-call pager, EHRS Director, Nanofab facility staff

3. Alarm 3: Level 1 gas detection in the ambient
Response:
- Alarm #1 response
- Ring fire alarm
- Local evacuation with horns & blue strobes

4. Alarm 4: Level 2 gas detection in the ambient
Response:
- Alarm #3 response
- Contact PennComm (campus police), EHRS on-call pager, EHRS Director, Nanofab facility staff

Gas cabinets and VMBs shut-down themselves for loss of exhaust and the cleanroom is evacuated.  We do not turn on the TGMS in this case; it’s not a detection event.

On Jul 17, 2015, at 11:57 AM, Corey David Wolin <cdwolin at ucdavis.edu<mailto:cdwolin at ucdavis.edu>> wrote:

I'm curious as to how many of you coordinate with campus first responders and EH&S on the serious nature of the the gases to which the sensors are monitoring?  What do many of you feel the proper emergency response to a high level silane alarm should be? Low Level? In addition, at which point should the TGMS evacuate the building if automatic shutoff valves exist for both high and low level alarms?

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20150724/96645a4e/attachment.html>