This may sound dramatic and it really was fairly routine and done as a precaution. I wanted to set the stage before I started using phrases like “smoke in the cockpit” and “declared an emergency” so that you have some context that this really wasn’t a big deal.
I was doing an IFR flight from Arlington to Orcas Island. The evening before I had flown IFR to Arlington through heavy rain. The 206 I fly sat on the ramp at Arlington overnight in the rain. The next day I had some trouble starting her, which is super common with this plane and an issue we’ve been chasing down all summer. I took the cowling off to put the battery on charge, and then got her started a while later using an airport GPU to jump start the plane to avoid stressing the battery.
I departed IFR from Arlington and headed up into the clouds. I entered one particularly dark cumulus cloud and encountered light turbulence. In pilot speak light turbulence makes passengers uncomfortable, moderate turbulence will shake your dentures out, and severe turbulence will damage the plane. So light is a noticeable shaking.
I had a brief puff of smoke in the cockpit with a slight odor of oil. Not an acrid odor. Just subtle. I was in IMC at the time (in the clouds) so I thought it may have been an optical illusion from all the vapor I was flying through. I looked to the right in the cockpit and against the black dash I could see a slight haze. I checked my carbon monoxide monitor which always reads zero and it said “8”. That got my attention.
I closed all heating vents and opened a window. The late Richard McSpadden’s voice was in my head and I remember the story about the pilot who got CO poisoning, fell asleep and his plane luckily landed itself in a snow field. Thank you sir.
I was IFR and on with Whidbey Approach so I called them up and let them know what happened. They asked what I needed. It took about 2 seconds, but felt like a comfortable 2 minutes where I assimilated the data, and my logic went something like this:
I work in cybersecurity so I think about mitigating risk using layers of security. I think about risk in the same way: When you have layers of issues stacking on top of each other, your risk increases. Any single one of them may be no big deal. But stack two or three and you have the makings of a problem. In this situation I had smoke, an odor, a carbon monoxide detection, and I had jump started the plane on the ground which adds some stress to the battery and starter. Those issues stacked were enough for me to make the decision to get back into VMC (out of the clouds) quickly, and on the ground. So I let Whidbey know I was “going to go ahead and declare” and that I’d like to get down, and direct to Orcas Island.
A new voice came on the radio – probably a supervisor – just incredibly calm and helpful just like the previous voice. I was immediately put on a heading to Orcas which cuts the corner and descended down from 4000 to 2700 which is their minimum vectoring altitude in the region. This is an altitude that you can find if you look for it, but it’s not published on our IFR charts. It’s the absolute lowest a controller can descend you without cancelling IFR.
Once I was at 2700 I was in and out of VMC so I told him I’m cancelling IFR and immediately descended into one of the holes in the clouds into clear air. I headed straight to Orcas and landed a few minutes later. I called Whidbey on their phone as per their request and they just wanted to know that I made it safely onto the ground. Nothing to discuss about declaring.
On the ground I closed the heating vents, closed the window, and did a high power runup trying to reproduce the CO detection or smoke or smell, and absolutely nothing. And the entire time in the air the engine monitor was absolutely fine. Normal CHT’s, normal EGTs, TIT normal, oil normal etc.
A buddy of mine who flew F16s for 20 years and is a rock solid CFII on the island helped me inspect the engine with the cowling off and do a sniff test of the entire thing, and we didn’t find anything that looked out of place or smelled burnt or weird in any way.
This happened yesterday. Today I took the plane to my A&P and he started asking me a series of questions. Great guy. From an IT/ops background, so just super methodical when it comes to isolating root causes. “Did you fly through rain at all on the way there?”. Why yes, as it happens a lot of it, and then it rained on the ground, and then I took the cowling off and water probably ran from that into the engine compartment.
What we figured out probably happened is that as I entered that dark cloud, pooled water sitting in the front of the engine splashed up onto the hot muffler and created a brief flash of steam. Oil naturally gets into parts of the engine compartment during the normal functioning of the engine and it accumulates over the years. So the water was mixed with oil and caused some combustion products which turned it from clean steam into steam with some CO and a light oily odor. That explains why it wasn’t acrid smoke, and why I doubted that I had even seen anything. And why I got an indication on the CO monitor.
The plane is in for a service and getting a full inspection and an electronic magento upgrade (SureFLY) which I’m excited about because it’ll probably solve the starting issues which I think are related to a bad impulse coupler in one or both mags.
Shout out to Whidbey Approach, Victoria Terminal and Seattle Approach for having our backs in any weather day or night.
Leave a Reply