Fuel cap mishap

Taken from: FLIGHT SAFETY AUSTRALIA JAN–FEB 2008

In October 1998, I was working as a
casual pilot for a Perth/Jandakotbased,
aerial survey and photography
company. I had about 700 hours’ flying
time, so I was relatively inexperienced in
comparison to the company’s flight crews;
however, I had more tail wheel hours than
nose wheel, and felt extremely privileged
when I was invited to captain the company’s
first aircraft: a Cessna 185, VH-KPA.
She was bought new in 1963 and virtually
kept due to sentimentality by the
company owner. She was the early model
with a 260 HP IO-470, rather than the 300
HP IO-520 that most of 185s have, and she
had been ground looped on several occasions,
so she flew a little sideways. But
after each event she was lovingly rebuilt
and returned to service.
When I got to know her, she was a
beautiful 35-year-old, resplendent in the
company’s red, gold and white colours,
and I cared for her as if she was mine.
I had been engaged to fly a dog-baiting contract commencing in Madura on the
Nullarbor. We had done a couple of days;
however, we were going to be operating in
increasingly remote areas, so following
the day’s flying, I was to ferry her alone to
Kalgoorlie for the fitting of an HF unit for
our SAR ops. I landed at Madura in the
late afternoon.
I organised the refuelling from drum
stock. I filled the left tank, and the hose
was passed to one of the crew to fill the
right. She was not NVFR, so the rush was
on to get airborne and to Kalgoorlie before
dark. As they completed the refuelling I
went to my tent to gather my bag.
As I approached her from behind, I
noted both tank caps were in place; I had
fitted the left, the crew member had fitted
the right, but in the need for haste I did
not get up and check the right.

First and second factors
I ran through my pre-flight checks and got
airborne, streaking away from the sun. The right-hand cap had been fitted incorrectly,
and at some point early in the flight
was removed by the low pressure, laying
back silently on its chain in the airflow.
As it was the tank furthest from me, I
did not see the escaping fuel;

and as she
was an early model, she did not have a
fuel selector, just an on/off, pull-out knob,
with both tanks feeding simultaneously
into the fuel bowl. She was fitted with two
150-litre bladder tanks connected by a
balance tube.
The low pressure siphoned the contents
of the right tank overboard, then pulled
the bladder up, holding the fuel indicator
float up, which gave me a half-tank indication.
The low pressure set about siphoning
the left-hand contents through the
balance tube.

Third factor
She had a tendency to burn part of one
tank, then would balance as the flight
progressed, so I noted nothing out of the ordinary. Nearly at Kalgoorlie, I could
see the aerodrome 6 nm away, and had
commenced my descent, passing through
about 3000 ft.
The engine stopped.
I went through the checks: the righttank
gauge suggested it was half full; the
left less so, but still indicating. I gave a
radio call expressing the occurrence, and
then commenced an approach on a wide,
straight road. At about 500 ft I saw a truck
whose day was about to get a lot more
exciting, and the engine restarted.
In retrospect, it was the extreme deck
angle bringing the last of the fuel to the
line; however, the fuel gauge indicated
half tank, and I had no reason to disbelieve
it. I set climb, rushed out a radio call,
and prepared for a straight-in approach
on Kalgoorlie’s main runway.
At 1000 ft the engine stopped again.
I rushed out another radio call, and
prepared an approach on a mine site road.
I recall the deck angle; even though the
plane was fitted with STOL droop wing
tips, the glide was steep. At 100 ft I noted,
to my dismay, a low chain-link fence and
gate about 10 metres beyond my proposed
touch-down point. I selected half flap and aimed for a point just beyond the gate.
She staggered over the gate in the
three-point attitude, and with her last
remaining energy, arrived on the gravel.

I alighted from the cockpit, got on the step
and gazed at the wing. The right-hand
cap was lying on the wing on its chain. I
went to the other side, replaced the cap,
and rang the company. I stepped out the
distance from touch-down to stop: a mere
100 paces.
On further inspection I noted a fleck
of paint about the size of a twenty-cent
piece missing from the left lower inboard
STOL tip that hadn’t been missing at the
start of the day. I looked back up my glide
path and noted that, had I not evaded the
fence, I would had put the left wing into a
tailings pile not evident from above, with
an ensuing messy semi-airborne ground
loop, probably destroying the aircraft.
I climbed the tailings pile and noted
a shallow groove running across its top,
in line with my glide path. I pondered
the circumstances, and ran my hand
down her engine cowling, thanking her;
I’d taken care of her and she had looked
after me. The first mine site vehicle with
surprised-looking occupants arrived.The company sent an engineer, and
overnight, we cleaned the fuel strainer,
put some fuel in her, and flew her out the
following morning on an adjacent longer
road.
The HF was fitted, and the operation
continued the following day. I have
considered since that a landing on the road
with some power may have been a better
decision; however, I had no reason not to
believe my fuel gauge indication, and the
semi-trailer I was head-to-head with was
a significant obstacle, the driver’s reaction
difficult to predict.
The factors of pressure with time and
darkness, failure to physically check the
fuel cap, reliance on visual inspection, the
nature of the fuel system, the erroneous
fuel indication, and my own inexperience
all conbined to cause this incident.
All things considered, the outcome was
as good as I could have hoped for despite
some questionable command decisions. I
can only hope this account serves to help
others to avoid my evident mistakes.
There’s AL WAY S time to check fuel cap
security, and doubly so when non-flight
crew are involved.
Safe and happy landings! ■

VH-POO's Profile

Home

FLIGHT SAFETY AUSTRALIA JAN–FEB 2008