Some calculations, all figures rounded, and short analysis.
Empty weight twin otter: 7000 lbs
Fuel in main tanks: 1500L, equals 3300 lbs buoyancy from empty fuel tanks.
Assuming full rigid cabin ferry tanks: 6000 lbs (as reported by PilotDAR).
Specific weight aluminum: 3000 kg/m3 = 6500 lbs/m3
Specific weight iron: 8000 kg/m3 = 17500 lbs/m3
Specific weight jet fuel: 780-840 kg/m3, assume 800/m3
Assuming a 50:50 weight ratio in iron and aluminum, the "weight"
of the submerged airplane would be:
Aluminum: 3500 - (3500/6500 * 2200) = 2300 lbs
Iron: 3500 - (3500/17500) * 2200) = 3060 lbs
Total submerged weight: 5400 lbs
Subtract the buoyancy of the empty fuel main tanks, giving the remaining weight: 2100 lbs
The buoyancy of 6000 lbs Jetfuel:
6000/0.8 - 6000 = 1500 lbs
When the ferry tanks are full, the submerged weight becomes: 2100 - 1500 = 600 lbs
The positive buoyancy contribution of 2 human bodies is neglectable.
The crew probably would have some additional stuff with them, assuming 200 lbs.
The cabin ferry tank structure would add some weight, assuming 200 lbs.
Gives a submerged weight of 1000 lbs.
Or so to say, the twin otter with empty main fuel tanks and full rigid cabin ferry tanks will clearly sink.
And, the twin otter with empty main fuel tanks and full or empty flexible material cabin ferry tanks will clearly sink.
Given the twin otter does float, the cabin ferry fuel tanks are rigid and
aren't full, IE the tanks are at least partially empty.
When the ferry tanks aren't rigid, the buoyancy contribution situation will
be worse (when emptying out) and will give no extra buoyancy due to the
ferry tanks inflating on emptying out.
Given, the twin otter did run out of (usable) fuel, it's unlikely they did
manage to use part of the ferry tank's fuel (either all fuel is available
or nothing, some way or another). So, given the wreckage is found floating,
it looks like the cabin ferry tanks are (nearly) empty.
The fact, the twin otter seems to be floating upside down, is an indication of
a strong buoyancy due to empty main fuel tanks and empty rigid cabin ferry tanks.
Now the departure situation: Assuming experienced twin otter pilots, I can not imagine,
they would not notice the anomaly of "empty" cabin ferry tanks, since that would
represent a pretty low-weight twin otter and certainly noticeable due to its STOL
take-off capabilities (vs. an expected overweight departure).
Or, so to say, the twin otter quite probably lost the cabin ferry fuel when in the air, on its way to its destination.
How to accomplish that:
- Hose detaching, somewhere down the line to the main fuel tanks.
- Given the gravity fuel flow to the main fuel tanks, maybe a lost sealing of the main fuel tank caps, fuel tank venting being open(-ing up), or something like that.
Feel free to correct the figures, etc.