Well, this was interesting:
We recently took a trip on a yacht that sailed through Raja Ampat and the Spice Islands. Diving and snorkeling were involved, although it wasn't the main focus of the trip. We got a dive or two most days.
I took my new Forensics CO detector with me and asked to sniff the first tank. The DM and cruise director graciously agreed. We got a reading of 3.6ppm*, which fell within my limits (as previously discussed here), but certainly pointed to an issue with the compressor setup.
The DM and director jumped on the problem, changing the filters and lifting the hatch** leading to the compressor. The next batch of fills tested at 0.0, which continued over the next 15 hours of fills.
The DM was very enthusiastic and interested in what was going on. It seems that most people/operations are not monitoring CO levels. As a goodwill gesture, I donated the CO detector to the DM and have already replaced it with a new copy. I'll be taking that along on our next trip to Raja Ampat this February.
* I dove the 3.6ppm tank with no noticeable side effects. In my mind, I had set a limit of 5ppm.
** As I understand it, the compressor was all-electric and was not generating CO. I could be wrong... language differences. But, the crew had been leaving the hatch mostly closed during filling so as to cut down on noise discomfort for the passengers. That hatch was below the smokestack for the 198-foot vessel, which at times churned out a good deal of exhaust and carbon soot, and was on the aft dive deck in the burble of the boat. Our sense was that the exhaust was coming through the mostly-closed hatch and collecting near the compressor intake. Opening the hatch wide -- and dealing with the noise -- allowed the area to breathe. Along with the new filters, the problem was solved.