Carbon monoxide (?) mystery

[Michael]

On two consecutive mornings this week, the carbon monoxide gas alarm was sounding when I got to Hali.  On the first morning, the propane gas detector was also sounding.  In each case, the boat had not been used, and the engine had not been run, since the night before.  The boat was relatively buttoned up, with only the vent hatches in the head and the aft cabin having been left open.  The CO monitor read-out was in the 80-90 ppm range but subsided to zero within about five minutes after the hatches and port lights were opened.  There was no noticeable smell (such as of propane or sewer gas) in the boat.

On the third and fourth mornings, in similar conditions except apparently one, no alarm was sounding.  The exception is that the night before the third day, a co-owner checked the propane tank valve and found she could tighten it somewhat more firmly.

Recent work on the boat had included replacing the two 4D house batteries (the new batteries have removable cell caps whereas the old ones apparently didn't) and putting a marine "no odour" product into the shower sump drain and toilet-waste line-holding tank.

The carbon monoxide gas detector is surface mounted on the starboard side of the engine box in the aft cabin about a foot from the floor.  It is a Kidde (3 AA battery operated) Carbon Monoxide Alarm with digital display and peak level memory Model KN-COPP-BCA Assembly 900-0146 manufactured on October 23, 2006.  It has a five year warranty.  Its User's Guide says it was "not designed to detect smoke, fire or any other gases."  It is an electrochemical device.  A note in the User's Guide says, "The following substances an affect the sensor and may cause false readings: Methane, propane, iso-butane, ethylene, benzene, toluene, ethyl acetate, hydrogen sulphide, sulfur dioxides, alcohol based products, paint thinners, solvents, adhesives, hair sprays, after shaves, and some cleaning agents."  The unit had been powered up in March or April, 2007, and not re-powered or "peak display" tested since then.  When the "Peak Level Display" button was pressed and held (which results in the display showing the highest CO reading recorded since the last peak level test or power-up) the reading given was 159.  This was after the second morning.

I read at the Centres for Disease Control website that the "Occupational Safety and Health Administration (OSHA) has two permissible exposure limits (PELs) for CO exposure. Exposures may not be over 35 ppm averaged over 8 hours and may never be over 200 ppm."

The propane gas monitor on Hali is surface mounted within 6" of the cabin sole at the base of the galley isthmus and is a hardwired 12 volt model GS/3 manufactured by Electro Systems Inc. in June, 2002.  I haven't checked yet whether it is out of warranty.  It was sounding an alarm (occasional beeping) that does not appear to be consistent with the presence of propane (but whether it would be consistent with the presence of some other gas hasn't been ascertained yet).

Nonetheless, a check on the propane gas line is in order...but hasn't been done yet.

Hali is docked in a marina.  A powerboat across the finger from Hali was gone on the first morning that the alarms were sounding...and possibly had left shortly beforehand but that is not know yet.  No nearby boat seems to have departed or been run on the second morning.

I understand that while carbon monoxide is generally only associated with incomplete combustion, it is at least theoretically possible for carbon monoxide to be produced when hydrogen sulphide (sewer gas) comes into contact with hypochlorite, an ingredient of bleach.  http://web.bsu.edu/IEN/archives/111607.htm

My current theories are:

- there was a propane leak
- there was carbon monoxide, probably produced from a nearby boat but possibly built up the evening before when Hali was run but which caused the CO alarm to sound only after its time sampling was sufficient (after we left) and which continued until the next morning
- the most unlikely theory is that the CO was produced in Hali's drain or waste line
- the new batteries are throwing off a gas that is tripping the alarms
- more than one of the above (and therefore a gumption trap)

Comments and suggestions would be appreciated.

See a fascinating posting by sailingdolphin about CO poisoning at http://c34.org/bbs/index.php/topic,2874.msg15204.html#msg15204.

[sail4dale]

I had one that went off without any traceable reason.  I also have a RV with a detector that has worked for years.

I traded detectors and both are working well.   WHY?

I read some fine print in the writeup on the offending detector and it casually mentions that sometimes a battery charger that is near by can cause interference that sets off the unit.   I still don't know why but both are still working well.  You might check out the battery charger possibility -  shut it off for a while and see it the unit works right.

[Michael]

Interesting.  The battery charger is very close to the propane detector but not so close to the carbon monoxide detector.  However, my working hypothesis (possibly wrong but safer than doubting the instruments, which I have learned to be wary of doing since finding a broken piston ring on an airplane engine when the magneto check showed slightly more than the expected drop in RPM) is that something bad is being detected, particularly because opening the hatches and letting the boat air has stopped the alarm from sounding.  But I will certainly add the possibility you have mentioned to the list.  Thank you.

[tonywright]

Hi Michael

Logic seems to suggest that you have a propane leak:

1) One detector can malfunction, and give a false reading. Two going off at the same time says that there is a real problem. The chances of two giving a false reading at once are very small.

2) The CO detector says that it will register in the presence of propane.

3) Opening hatches and providing fresh air cleared both alarms.

4) Tightening the propane tank valve stopped alarms from sounding.

I would suggest a thorough check of your propane system and stove.

Tony

[Michael]

Hi and thank you, Tony.  Yes, I completely agree that the propane system needs a thorough check.  Yet the high "peak" CO reading still concerns me.  So while I will "hope" to find a propane leak, if there is one I still won't rule out the possibility of a CO issue.  Regards. Michael

[jfssail]

Michael, I had a similar problem last season with the same CO detecter. I had the problem when equalizing my batteries, due to the battery acid vapors produced during the equilalization procedure without adequate ventilation. You could be getting the same vapors during normal charging of your batteries in a closed spaced?
I don't think propane vapors would affect this CO detecter.

Jack F Stewart
1993 C36 #1233 "Windancer"
Port Clinton, OH

[Michael]

Thank you, Jack.  Its good to have input from those who have been in a similar situation.

Another co-owner of Hali has also weighed in.  Lionel remembered that we spilled oil on the engine during an oil change on May 13. He speculated that the burning-off of the oil produced CO.  If we count his vote, yours, Tony Wright's, and sail4dale's, the votes and competing theories now are:

Battery/battery charger related: 2
Propane: 1
Carbon monoxide: 1

We probably need to ride all the theories.

Does anyone else have any thoughts/questions?

[Stu Jackson]

1.  You've identified the suspects

2.  Until you can conclusively narrow it down or eliminate one  or more you HAVE TO continue to consider all three

3.  You need to develop and implement a plan to begin to conclusively eliminate them one by one

4.  Once you determine the cause, you'll obviously need a fix to avoid recurrence

[joe]

it would seem to me that if you enclosed your monitor in the battery compartment the alarm would go off pretty quickly if that were the problem

[Craig Illman]

One other consideration is that it could be a combination of two as well.

Craig

[tonywright]

It might be worth trying to find a tech with the appropriate detection/analsysis equipment. (Might be worth giving the local fire dept a call, they might be interested in helping you track down the problem on a slow day..?). There are detectors out there that can tell you exactly what you are looking at, and will help sniff out the precise location of a leak. Bacharach is a name of manufacturer that comes up if you google. While their basic detectors only cost $300-500 each, it is probably more than the average sailor wants to spend. On the other hand, proving the presence or absence of a propane leak - priceless?

I wonder if there is somewhere that rents them by the day?

Of note is that CO detectors apparently need calibration every 1 -2 years.

Tony

[Ron Hill]

Michael : I don't believe that propane will confuse a CO detector.
I'd send that detector back to the company for a check and/or have the local Fire Dept (as suggested) take a look at it!  CO detection is something that you don't want to miss!!  A thought

[Michael]

Ron, Stu, and everyone else who has weighed in, thank you for your input.

Heeding the advice so far, I will continue to assume there is danger of CO poisoning, battery fume poisoning or explosion, and propane explosion, ...but that something else also might be at work. 

Tony's idea, seconded by Ron, of asking the fire department to investigate seems a good one.  Or, with the CG station almost next door, their dock may be a first port of call.

I have now emailed the alarm companies for feedback.

There has been no peep of an alarm since those first two mornings...but they rattled the cage well.

Thanks again.  I'll post the conclusion (all going well) but meanwhile if anyone has further thoughts they are welcome.

[Randy and Mary Davison]

On Gorbash, my CO detector goes off anytime the boat is closed up and the batteries are charging heavily.  The same detector trips at home when I leave in my office/ham shack closed up with the 4d I use for radio equipment charging heavily.  I've read that some CO detectors are sensitive to battery charging fumes.  This one certainly is.  I have another one to try out.

[Michael]

The propane gas alarm manufacturer, Electro Systems Inc., kindly provided the following response:

"From your description of the situation, our best guess is that the
batteries are discharging Hydrogen, which could trigger the GS/3
propane sensor, which is designed specifically for propane, but is
sensitive to most hydrocarbons. A description of the way it works is on
our website at: http://www.es-web.com/sensor.html

We don't know if hydrogen could trigger the CO detector.

....the fact that your co-owner could tighten the propane valve...
suggests a small propane leak may have occurred. Propane has a compound
called mercaptan added as a safety precaution to make it smell so you
notice a leak. It is usually described as smelling like rotten cabbage...

I hope this is helpful. Kidde should be able to provide some insights
on the CO alarm."

With that input and Randy Davison's, the betting continues to favour the batteries as the cause of the alarm(s) sounding.  But we will keep propane and carbon monoxide, like Hilary Clinton, in the race.