[aprssig] APRS to the planet rescue?

[KF4LVZ]

On 2017-04-03 07:56, Robert Bruninga wrote:
> Here's a project.  Add a $4 Methane detector to your APRS mobile and help us 
> find the bad methane leaks.  Google is adding it to their maping cars.  Maybe 
> Amateur Radio and APRS could help?
> 
> https://mail.google.com/mail/u/0/#inbox/15b2521854ccaa63
> 
> 
> Apparently the 2:1 advantage of replacing the high carbon emissions of coal with 
> natural gas, overlooks the fact that the methane released with natural gas 
> extraction methods is 84 times worse than carbon as a green house gas and may 
> more than offset any advantages.
> 
> Just one methane leak a few years ago did more damage than about a million cars 
> a day before it was plugged months later
> 
> .http://www.businessinsider.com/methane-gas-leak-porter-ranch-los-angeles-disaster-largest-in-history-2016-1
> 
> Anyway, a $4 sensor and a tiny processor could add a flag to our mobile packets 
> somehow  when levels exceeded some background level
> 
> Though I am not sure if the sensitivity of these $4 sensors is good enough to 
> find the outdoor area low level leaks, or only detects high concentrations...
> 
> I wish I had the time to make one and experiment with the concept.  It would 
> make a great fun project. I used to thing a cheap radiation detector would let 
> APRS contribute to plotting radiation hot spots, but I fear the natural gas 
> leaks and fracking damage is going to kill us first.
> 
> My thoughts on adding this to APRS is to simply have a processor send out a 
> modified STATUS text when a detection is made.  Since APRS was designed around 
> the status being merged-on-receipt with the most recent posit, this is a way of 
> nailing down the location of high methane levels or leaks.
> 
> Status could simply be >CH4=XXX where XXX is the reading.  Format of XXX will be 
> determined to  give reasonable scale over expected detections.  Then the APRS 
> map could selectively plot any detections.
> 
> Anyone have more experience with these sensors and the levels of detections 
> expected?
> 
> Bob, wb4apr

No, the sensor you want to use is a passive unit designed for indoor air
quality monitoring (specifically looking for a hazardous leak that could
result in an explosion).  The unit in the StreetView car is a gas
sampling analyzer (sniffer) and probably costs closer to $1000  as it's
looking for concentrations on the level of single and double digit
parts-per-billion.

More specifically, according to the paper the unit in the StreetView car
is from Picarro.  It's an infrared spectrometer based detector.
Because it's a spectrometer it can measure into the low
parts-per-billion levels which is critical for detecting a leak outdoors
where the atmosphere will dilute the gas quickly.  The cheap sensor you
describe is a chemical cell sensor whcih is only good at concentrations
above 200 parts-per-million, so already five orders of magnitude poorer
sensitivity than the StreetView car.  Again, the $4 sensor would work
for detecting a sudden leak in your kitchen in order to sound an alarm
or turn off a valve in some automated fashion (mainly if the home is
empty).

However, even this cheap sensor is worse than your Mark 1 olfactory
sensor (human nose) because the main additive to natural gas for
detecting a leak is t-butyl mercaptan and the nose can pick up the smell
to a level of 0.33 parts-per-billion, equal to the much more expensive
Picarro sensor.