Early in my career I worked for a university that had a large flight school, and an executive transportation department that consisted of a couple of Cessna 310s, a 402, a 411, and three DC-3s. I was a captain on all the aircraft, which kept things interesting.
One sunny spring day I had a flight from Carbondale, Illinois, to Tulsa, Oklahoma. The 26-seat DC-3 was to be filled with a collegiate baseball team in the back, and co-pilot Doug Fitz and me in the front. This DC-3 was originally a C-41, complete with Hap Arnold’s name plate still attached on a panel in the cockpit. There were some strange issues on that airplane, such as a non-standard hydraulic system, questionable fuel tanks, and a radar that only worked when it was in the mood.
I watched as Doug completed his usual thorough pre-flight of the aircraft. The baseball team had not yet arrived, so all gust locks were removed except for the rudder lock, which was to be removed just prior to closing the passenger door, when everyone was loaded. Doug drained the fuel tanks, using the laundry soap container which would hold almost a gallon of liquid. It was made of clear plastic, which made it easy to check for contaminants such as water, sediment, etc. Doug drained a little over a quart of fuel from each tank, noting that we had a little water in each tank. He sampled another quart out of each tank, to the point that the only liquid being drained was pure 100 octane.
The baseball team arrived, bags and ball bats were loaded, the rudder lock removed, two good engine starts, a normal taxi out, a good run-up, and all systems were go for takeoff. We departed and climbed to our cruise altitude of 6000 ft., headed for Tulsa. The weather was good VFR in Tulsa, and other than a few clouds at 4000 ft., beautiful along our route.
We did not have an autopilot, so we would normally take turns flying between VORs. Doug had just taken control of the aircraft and I settled in to enjoy the trees and lakes of Missouri going by my window.
After about 45 minutes, it was time to manage our fuel. There are four tanks on this aircraft, two in each wing. The front tanks were referred to as the “front mains,” and the auxiliary tanks were referred to as the “rear tanks.” We had departed with full front mains (400 gallons), and 75 gallons in each aux tank.
The fuel tank selectors on this aircraft are located on the pedestal, up near the instrument panel. Although either pilot can reach both fuel selectors with a long reach, our policy was each pilot was responsible for his/her fuel selector.
“We probably ought to get some fuel out of the back, don’t you think?” I asked.
A man of very few words, Doug said, “yep.”
Doug reached up, and I watched him rotate the fuel selector to the right rear position. I reached up about two seconds later and switched my fuel selector to the left rear tank position.
After another two seconds, it got quiet. Real quiet. As in no engine noise at all.
“Turn them back!” I yelled, trying not to sound scared to death. Doug and I had flown together many times, and he was already reaching for his fuel selector. I was right behind him on mine. Our philosophy had always been, “if something goes wrong, undo what you last did.” Both engines caught immediately; we had only lost maybe 75 feet of altitude, and we were back on our way. I turned around to look at the baseball players, expecting to see worrisome glances and wide-eyed athletes, but they were all asleep. They had no idea…
We concluded that we had bad fuel in the rear tanks. Our mains seemed OK, but our endurance was now reduced. We diverted to Springfield, Missouri, for an unscheduled stop to check the fuel, and to re-fuel as necessary.
I didn’t even have to ask, as Doug grabbed the laundry soap container and stormed out. “I’m draining the tanks to see what we got!” he growled.
Doug drained what seemed like a couple of gallons of water from the rear tanks. There was no water in the front main tanks, only the rear tanks. We topped off our front mains, which gave us enough fuel to get to Tulsa comfortably. We drained the tanks again after fueling, got a little bit more water from the rear tanks, and none from the mains. Off we went to Tulsa.
After the baseball team left on the bus, we peered down into the rear tanks, but of course, we could not really see anything. We drained the fuel tanks again, and only got a little bit of water in our fuel sample.
The preflight for our flight the next day included another serious draining of all the fuel tanks.
There was no water detected in our front mains, and barely a quarter cup of water from both rear auxiliary fuel tanks.
After the ballgame, we headed to Wichita. We did not use any fuel from the rear tanks on this leg. We checked for water again in the rear tanks and found hardly enough to get excited about. The next day we headed back to Carbondale. At 7000 feet, while cruising along nicely, we tried utilizing fuel from the rear tanks again. However, this time, we switched to the right aux tank, and waited a couple of minutes to see if the right engine would continue to run. Then we did the same on the left side. Both engines purred like kittens, but for a few minutes, we were on edge to say the least. We were both ready to undo what we did.
It was determined that both aux tank caps had leaky seals that would allow rainwater to drip into the fuel tanks. When the tanks were examined, it was noted that water was adhering to the fuel tank wall. Vibration during flight would cause the moisture to drop into the fuel, and cause contamination to the point where there was more water than fuel going to the engines. The tanks were cleaned, seals were replaced, and the problem was never seen again, although we still drained all the fuel tanks significantly before each flight.
Ever since that day nearly forty years ago, “undo what you just did” is always in my pocket when things don’t go as expected.
- There was no checklist for this one… - July 5, 2021
- DC-3 vs. blizzard—and that’s just the beginning - May 6, 2021
- “We was on fire; I could see the flames!” - March 4, 2021
A well-written and good reminder of something I too often see skipped, as many seem to think it’s merely an academic risk. Early into my flying career I flew traffic watch, where one of the ride-along observers asked, “What are you doing?” as I painstakingly sampled each tank, every time, as all the others in the fleet just got in and flew off. Hearing my explanation, the observer (who rode with all of us) said, “the others don’t do that” — which was as evident among these “pros” as It was with the weekend warriors who regularly took off without sampling after stopping for fuel. Others who rode along in my other endeavors, each also having experienced flights with other professional pilots, said the same. One day a flight instructor I knew told me of a close call he had during an instructional flight where he pulled the engine to idle for a simulated emergency landing drill. The “landing” assured, he added power to climb away from the clearing — and watched in horror as that only emergency landing site slipped behind the wings as the engine then failed to produce power. The culprit? Like your story, Steve, it was found to be defective seals which let in rain water. In his case, the flight instructor ended up very lucky in that the engine ingested the water for awhile before choking back to life just before they nearly hit the rocks and trees. They missed by a foot. How lucky was he? After a rainstorm, I had removed water from the tanks of that same aircraft a week before for 45 minutes before getting a clear sample, squawked it, then the aircraft was brought back on line (evidently without repair) for his flight. So, if I had pulled water out for 45 minutes, then the next rain replaced a similar amount of water, you can imagine how much water had to flow through to reach fuel. I’ve been flying for dozens of years and seeing others skip fuel sampling still continues to be a regular part of what I see on the ramps. Of all the foolish reasons to crash, come on, this has to be about the top of the list. Oh, and remember that it takes time for the 8#/gal water to fully sink through the 6#/gal fuel (one published figure was 20 minutes), so do wait a bit after fueling to perform that check. Thank you, Steve, for writing an important article on two important points – consider undoing last action first during trouble, and please sample fuel before each flight.
Great story. A 727 lost all three engines overflying Chicago at night. Happened during fuel balancing – – combinations of boost pump and fuel shutoff switches….. Remember the old saying, “turn something on before turning something off?” The flame-outs were as attributed to a different accumulator configuration for the new fuel efficient engines.
They finally recovered partial power, low on final into ORD.
As I recall, the crew did what you practiced – – “undo was you last did.”
Not to mention, when changing configuration ie, fuel tanks or going from low to high blower we always did one, wait then did the other.
DC 3, 4,6,7,C 46, many multi tank aircraft, because things happen when you do change fuel selections.
Flying from Knoxville, TN to upstate New York in my aging Aztec one lovely afternoon, my three passengers and I were having a nice conversation. As I customarily did, I had switched tanks at a point on my clock to go to outboard aux tanks which were full. My brother in law decided on a nap and got comfortable, putting his feet up on the pedestal. Within moments, I became aware of the right engine shutting down! A quick glance around showed the right tank-selector being in the “Off” position! My errant passenger sat up abruptly and began to apologize profusely! “Did I do that?” he called out. “Umm, yeah maybe,” I said, realizing the old levers allowed the worn detents on the selector to slide forward easily to the center position, effectively shutting off the fuel supply to that engine.
It was a startling lesson in fuel management and passenger involvement!
Many years ago my Twin Comanche was parked outside on a small airport in South Western Oregon for a week. 60 to 70 inches a year is a common rainfall there and it seemed the rain never stopped during that week while we were working the fall cattle roundup chores.
When it was time to leave I took particular care to check the 4 tanks for water and found none.
Take off to the north put me over the river and town for about a mile over inhospitable landing sites. I had climbed to 1000 feet and just cleared the town when BOTH engines quit at the same time. I turned 90 degrees right switched both engines from mains to aux and set up for a base leg to a small patch of open ground that was too short to save the airplane but I felt would save me. (I have air bags and shoulder harness!). Just as I was tuning “final” BOTH engines caught ! I started an immediate circling climb and went to 8000 feet to let my pulse settle down. I then flew to a nearby FBO where we found water in ALL tanks. The problem was that the rubber seal in the caps, even though adjusted to be hard to cam over and seal, had hardened so they were not sealing the caps. Rest assured that I maintained soft rubber seals on all the tanks for the remaining 45 years I flew the airplane. A H Powers. N88AP
In the late 1970’s I had a Cessna 210- with struts – that I flew on a courier route. After landing at Red Bluff some fuel was added and the on to Chico, Sacramento and Orland where it was topped off. Then up to Red Bluff, on to Chico and to Sacramento….but on the way in to Chico I felt a bit of a miss and guessed at fouled plugs. Then only got almost to Marysville and gradually lost power until diverting to MYV. I called my Red Bluff office to bring the C-182 down to finish the route, thinking something in the induction plumbing had fallen apart. But they told me tow other planes had just reported problems and to check the tanks. I got 3-4 samplers of water from both and taxied to a gravel area and pulled the fuel drain for several seconds. Ran great ever after but got bits of water for several landings and fuelings.
I learned this lesson about ten years ago. I was doing practice night landings in my ’62 Skylane on a pretty, moonlit winter evening, enjoying flying over the snow-covered meadows below. I had just taken off on what was probably my fifth or sixth circuit, adjusting the prop and glancing at the tachometer as I twisted the knob. I remember thinking, “Boy, it seems to be taking a while for the prop to respond.” Suddenly the engine quit at about 400′ AGL – just quit, although the prop was still windmilling.
I’d always wondered how I would respond to a real emergency. Surprisingly, I didn’t panic; I lowered the nose immediately to best glide speed, looked at the fields below, and said our loud, “Sh*t. I’m going to get my plane muddy.” I thought later, what a strange reaction ….
This was before I learned the “Undo the last thing” mantra, so while I was gliding into a muddy, snowy pasture, I thought, “Well, it’s winter; maybe I picked up some carb ice.” (The O-470 is a famous icemaker.) I pulled carb heat. Amazingly, the engine started running again – not great, but running. It made no sense, but I didn’t want to look the proverbial gift horse in the mouth, so at about 150′ AGL I began a slow turn toward the airport, climbing about 100’/minute.
Safely over the airport, I started a little troubleshooting, only to discover the mixture knob nearly all the way out. In the dark, I had been so relaxed that I failed to confirm the knob’s shape and had been twisting the mix instead of the prop control; pulling carb heat had enrichened the mixture enough to get the engine somewhat running again.
I relayed my interesting flight to my instructor, who clewed me in on the “Undo the last thing.” Obviously, lesson well learned, and learned the best way: no penalty for doing it wrong.