Page 1 of 1

N-butane leak detection

Posted: Mon Apr 06, 2009 9:06 am
by kevham
We produce a temperature sensing device which contains N-butane. I need to check the components for leaks on a high volume production basis. Which sensing technology would be best? we currently use PID but find the machine unreliable.

regards
Kevham

Posted: Mon Apr 06, 2009 10:41 am
by gcguy
Are you looking for a quant or qualitative test?

GCguy

Posted: Mon Apr 06, 2009 11:58 am
by kevham
quantative. I'll dig out the spec for the max allowable leak rate and post

Posted: Mon Apr 06, 2009 12:05 pm
by kevham
max leak rate allowable is 1.76 * 10 to the -7 cc/sec

Posted: Fri Apr 10, 2009 10:14 pm
by freemab
Simple thermal conductivity measurement (NOT including GC separation) might work just fine. It would depend on whether the expected concentration is sufficient for the relatively insensitive TCD. Check GOW-MAC's website for a TCD detector used for checking for helium leaks. One that uses hot-wire filaments would be inappropriate for a air stream, but possibly a thermistor-based TCD could be used to measure butane in air.

Which reminds me that flammable-gas detectors are another option. Google that.

Those may or may not be FIDs. Free-standing FIDs (NOT including GC separation) are available, and would certainly be a sensitive way to measure butane.

If distinguishing butane from other hydrocarbons were necessary, THEN you could consider a dedicated GC.

Posted: Sat Apr 11, 2009 1:45 am
by Bruce Hamilton
I wasn't going to comment, because my experience is decades old, but I would caution about using any sort of non-specific detector for nC4.

One problem is that low molecular HCs are almost ubiquitous at ppb concentrations around industrial areas. C1 is natural and from IC engine exhausts and evaporative emissions, C2-C5 can also be from liquified gases used for heating and fuels. As background concentrations vary, so sensitivity can change rapidly.

Note that HC flammability detectors are quite insensitive because of the narrow flammability limits ( 6 -14% ) in air.

Another problem is that, at such low leak rates, the sampling is critical. Having worked in a production environment with thermionic and electron capture detectors seeking similar leak rates of CFCs from refrigeration hermetic systems, I've seen the issues of directional leaks being missed by some handheld devices.

It is essential to have a contamination-free area, with some form of whole module sampling system, and yet be able to trap and remove the leaked emissions from devices that leaked badly, as well as those that leaked slightly.

My understanding is that in some situations, small MS systems have used, however I would suggest contacting some of the suppliers of industrial systems, eg
http://www.lacotech.com/Leak_Detection.html
A Google search should find many others...

Bruce Hamilton

Posted: Wed Apr 15, 2009 10:10 am
by kevham
Thanks Bruce and Freemab, I'll start doing some digging based on your advice.