The ATSB has done ground smoke and digital detector tests. The problem faced was there wasn’t a definitive path from exhaust crack into accessory compartment. This was solved by locating a relatively large diameter hose in the Beaver’s exhaust outlet and pumping raw, undiluted exhaust into the accessory compartment. Hose entry into the compartment was made possible by lifting an accessory cowl corner. No surprise then that CO via the missing bolts did register in the cabin.
Hardly a valid test. To get CO into the cabin, you can figure how the ‘door ajar’ theory came about.
Even so, the highest level detected was 144 ppm at pilot’s feet near the mag panel. To put that into perspective, 100 ppm will result in a slight headache in 2 – 3 hours. The CO spot detector as fitted to NOO will begin to darken at 100 ppm in 15 – 45 minuets. No head height readings are shown in the report.