By Joel A. Turpin, ATP CFII  FAA Master Pilot

The Way We Were

Please understand that when this incident happened in 1976, few of the smaller airports with a control tower and an approach control were equipped with radar. Also, enroute center radar coverage in the Midwest was nonexistent below 5,000 feet. This meant we operated mostly in a non-radar environment which added to the difficulty of flying in those days. In addition, localizer back course approaches were present anywhere a front course ILS was available which, thankfully, is not true today.

The back course used the same localizer antenna and the same front course bearing, but there was no glide slope which made it a non-precision approach. And because you were flying toward the runway in the opposite direction of the front course ILS, your course corrections were the exact opposite of what would be used when flying the front course.

For example, if the ILS needle was off to the right, on a back course, you turned left to bring it back to the center of the indicator. Flying a localizer back course was very challenging, if not a bit confusing.

Another difference from those days that is mostly not true today was that all of our airplanes were equipped with analog instruments as there was no such thing as GPS. Our compasses were called directional gyros, or DGs for short. They were powered by suction from the vacuum pump that pulled air through the instrument that spun a gyro.

Since there was no external input, a directional gyro compass had to be manually set using a knob to set it to what was seen on the magnetic compass. On cross country flights, the pilot had to occasionally reset the DG to the magnetic compass. That was due to the fact that DGs would sometimes drift off of the heading that was previously set, especially when flying in turbulence, a phenomenon known as precession.

Skyway Airlines

I was a pilot for Skyway Airlines, a small but vibrant commuter airline that was based at Fort Leonard Wood in the Ozarks of southern Missouri. I flew for them from 1973 to 1978. At Skyway, we flew the 9 passenger, radial engine, tailwheel Beech 18, the 15-passenger turboprop Beech 99, and a small fleet of single engine planes that included a Piper Arrow and two Cherokee Sixes. We would add a DC-3 to our fleet in 1974 which I would later fly as a captain.

We flew these airplanes in scheduled passenger service from Fort Leonard Wood to Lake of the Ozarks, Kansas City, Rolla, Missouri, and St. Louis. We also flew on mail routes at night as well as offline passenger charter flights. Since my hire date in 1973, by 1976, I had climbed the ladder at Skyway Airlines from copilot to chief pilot, training officer, and company check airman on all of our airplanes as a 26-year-old.

Skyway had a satellite operation in Columbia, Missouri where we had two Cherokee Sixes and two pilots based. From Columbia, those two pilots flew several round-trip flights each day to both St. Louis and Kansas City.

Beech Training

By late 1976, our passenger business at Columbia had grown so much that the 5-passenger Cherokee Sixes could not handle the demand. A decision was made by my boss to base a Beech 18 at Columbia to provide up to 10 seats to meet that demand. We could fly the Beech 18 single pilot on VFR flights which added the copilot’s seat to the 9 cabin seats,

If it was VFR, ten seats would double what a Cherokee Six could handle. But on IFR flights, we were required by the FARs to have a copilot which took one of the ten passenger seats out of the equation.

However, before dedicating one of our Beech 18s to Columbia, we first had to get both of their pilots trained and qualified to fly the formidable tailwheel Beech 18 for both single pilot and two pilot operations. And that would be my responsibility since I was the company training officer.

I had enjoyed several rare days off over the Christmas holidays from my tough and demanding job where I was routinely flying 100 hours each month. Prior to the Christmas of 1976, I had spent a great deal of time in Columbia training our two pilots on how to fly the Beech 18 with an emphasis on making crosswind takeoffs and landings. The Beech 18 simply demanded that its pilot have unerring crosswind skills or a ground loop would be the penalty he would have to pay. 

I would routinely board one of our Cherokee Sixes, fly by myself to Columbia, then conduct training flights with the two pilots based there, Ron Arnett and Andy Baughman. Both pilots had made good progress during the arduous task of getting them checked out, even though neither had any previous tailwheel flying experience.

Once the training flights for the day were completed, I would then fly the Cherokee Six back to Fort Leonard Wood solo. And it was on one of these flights back to Fort Leonard Wood that I experienced a dramatic chain of events that seriously challenged my flying skills as well as my decision-making skills! And it all began when I said to myself “All I’ve gotta to do is…”

“All I’ve Gotta Do Is …”

December 28, 1976, was clear and cold in Columbia where I had two training flights scheduled, one for each of the Columbia based pilots. After the training flights, Ron Arnett would fly flight 65, a scheduled passenger flight, in the left seat to St. Louis, and Andy Baughman would then fly flight 66 in the left seat back to Columbia. I would occupy the right seat as a safety pilot on each leg thus giving both of them some supervised real world operating experience.

The weather forecast indicated a major snow storm was expected to affect Columbia at our arrival time on the night flight from St. Louis. This was due to a cold front that was rapidly moving down from the north.

Our IFR flight plan from St. Louis to Columbia had us flying a Victor Airway westbound that passed about 20 nm north of Columbia. We would remain on the Victor Airway until intercepting the localizer back course for Runway 20, then fly it to the airport.

As we cruised westbound at 4,000 feet in perfect VFR weather, the leading edge of the cold front overtook us from the north, and consequently we flew into snow. At that point, we had unlimited visibility to the south, but not to the north due to the snow. Soon, the snow intensified and we found ourselves flying on instruments in heavy snow and the good visibility to the south had disappeared.

“All I’ve gotta do is, take off on Runway 20, then fly as fast as possible to the south, and I will quickly out run the storm and be in the clear in about 5 minutes.”   

Since we were approaching a point almost due north of the Columbia airport, we had our #1 VOR set to the localizer back course and used the #2 VOR to fly the airway. Soon, the localizer indicator began to move and we intercepted the back course and flew it towards Runway 20. After flying the back course on a heading of 222 degrees for about one minute, we flew out of the snow and into perfectly clear skies. This meant the snow line was just 20 nm to the north of the Columbia airport and rapidly moving south.

Since I still had to fly the Cherokee Six back to Fort Leonard Wood that night, it would be a race against time. I said to myself “All I’ve gotta do is take off from Columbia on Runway 20 and make a slight turn to the south before the snow begins.”

We landed the Beech 18 and since time was of the essence, I instructed Andy and Ron to secure the airplane for the night leaving me free to get the Cherokee ready to fly back home. However, as I was doing my preflight, I discovered that I had forgotten to have it refueled when I had parked it many hours earlier. I quickly called for the fuel truck and as the fueler was topping off my tanks, it began to snow. This meant I had lost the race with the rapidly moving cold front.

With this new development, I said to myself, “All I’ve gotta do is, take off on Runway 20, then fly as fast as possible to the south, and I will quickly out run the storm and be in the clear in about 5 minutes.”   

Once the fueling was complete, I started the engine and checked the ATIS. The wind had shifted to the north with the passage of the cold front and now Runway 2 was the active. That meant taking off into the storm flying north, the exact opposite of what I had planned.

But I said to myself, “All I’ve gotta do is take off to the north and request an immediate right turn to the south.” That way I would only be flying in the snow for a short time before I outran it and into clear skies. Once I got to the end of Runway 2, the weather had worsened to a ceiling of 100 feet obscured, visibility of 1/8^th of a mile in heavy snow.

I was not concerned with icing as we had not picked up any when flying in the snow in the Beech 18. Plus, you almost never get wing icing when flying in heavy snow because the supercooled water droplets in the clouds that turn to ice upon impact with your wings had condensed into snow. With this thought, I was cleared for takeoff on Runway 2 with instructions to contact the non-radar approach controller once airborne.

While I would not normally have departed into such vile weather, I did so thinking I would quickly outrun the storm and be flying in clear skies in just a few minutes after takeoff.

Prior to takeoff on Runway 2, I reset my DG to 022 degrees, the heading of the runway. Once cleared, I made the night take off into the teeth of the storm and was met with zero forward visibility along with moderate turbulence. I contacted the approach controller and, as per my plan, I requested an immediate right turn to a heading of 180. His reply was a shock!

He said he could not clear me for a turn to the south because he had one airplane on the ILS approach to Runway 2, and another plane in the holding pattern over the outer marker. After all, he was working without radar and a heading of 180 could not guarantee traffic separation. What to do now?

Plan B

I quickly came up with a new plan when I said to myself, “All I’ve gotta do is to fly due east until intercepting the 360 radial of the Jefferson City VOR (located 15 miles southeast of Columbia), then fly it southbound and out run the storm that way.” I made this request to the Columbia Approach controller and he cleared me as requested.

While flying due east at 3,000 feet in heavy snow, I selected the 360-degree radial of the Jefferson City VOR on my #1 Nav and waited for the indicator to move. When it did, I would simply turn right to the south and track it to the Jefferson City VOR and outrun the storm.

While flying a heading of 090, the turbulence increased, making instrument flying difficult. I checked my magnetic compass and it showed 105 degrees while the DG compass indicated 090 degrees. I reset the DG to 105 degrees and turned left back to a heading of 090. I reasoned that it had precessed due to the turbulence.

A few minutes later, my DG once again did not agree with the magnetic compass, as the magnetic compass was showing 110 degrees. So I reset the DG to 110 degrees and once again turned left to 090. This happened several times and I was always turning left to make the DG agree with the magnetic compass. And when will that VOR needle ever move indicating that I had intercepted the 360 radial off the Jeff City VOR? I was starting to sense that something was amiss with my gyro compass when I was startled by a loud CRACK!

It was caused by a 6-inch-long spark that jumped from the inside of my windscreen to the top to the instrument panel!

I had never seen such a thing! Then I caught sight of a living web of electrical sparks crawling on the outside of the windscreen. It was Saint Elmo’s Fire, a phenomenon caused by an incredible buildup of static electricity on my airplane caused by flying through heavy snow.

I then discovered that the magnetic compass, which was mounted on the post that separated the two windscreens, was spinning wildly due to the magnetism from the sparks on the windscreen. That solved the mystery of why my DG had supposedly precessed so badly.

The DG heading had been correct all along! And I had been resetting it to a magnetic compass that had been lying to me due to the influence of the electrical sparks on the windshield. So, I had not been flying due east, which was why I never intercepted the 360 radial off of the Jefferson City VOR. Then, a scary question arose in my mind. If I had not been flying due east then, where the hell am I?

Now What?

Since I was operating in a non-radar environment, a simple call to the approach controller to give me a position report was out of the question.

And at that moment, my engine began to run rough. I checked that the carburetor heat was on, checked both mags, and saw that all engine gauges were in their green arcs, but could not solve the problem of a rough running engine. But it was producing enough power to maintain altitude and a normal cruise airspeed, so I put this new problem out of my mind.

Being lost while flying in a snowstorm, at night, without a reliable compass was more than concerning, it was terrifying! Without a reliable compass, you cannot track a VOR radial, so simply centering the needle on the Jefferson City VOR and flying direct to it was not an option.

But then, a plan for salvation emerged in my brain. I said to myself, “All I’ve gotta do is take a cross-bearing check between the Columbia VOR and the Jefferson City VOR. Where the two VOR radials meet is where I am.”

This I did in quick order and found that I was well northeast of Columbia instead of due east. I needed to fly south to outrun the storm which meant I had to find a heading to fly using a DG that was set in error, and a useless magnetic compass.

I did the math and found that since I was northeast of where I should have been, I first needed to turn right about 60 degrees to have me flying due east. And then I had to turn right another 90 degrees to fly south, a total heading change of 150 degrees. I looked at the erroneously reading directional gyro, counted off 150 degrees to the right from what it was currently showing, then flew whatever heading that was.

Once I had made the 150-degree heading change to the right, my plan was to hold it until I outran the storm, along with the Saint Elmo’s Fire that had rendered my magnetic compass useless. Now, all I had to do was wait to see if my math was correct, and if I truly was flying due south.

Freedom

Suddenly, like an apparition, the skies below me began to brighten. My heart was racing when I realized what had caused the skies to brighten through the snow below me. It was the city lights of Jefferson City! Soon thereafter, I flew out of the snow and into clear, but very dark skies. The plan, and my math, had worked to perfection!

And the rough running engine smoothed out. At the same time, the Saint Elmo’s Fire disappeared, giving me back my magnetic compass. I quickly reset my DG to the magnetic compass and everything was totally back to normal. I reasoned that the heavy snow might have caused one or more spark plug ignition wires to arc resulting in the rough running engine. It could also have been due to the incredible amount of static electricity that had built up on my airplane which could have affected the magnetos.

Not Over Yet

But then a new problem arose when I made contact on the company frequency with another Skyway plane, a Beech 99, that was flying into Fort Leonard Wood from Kansas City. The captain told me that the winds at Fort Leonard Wood were 180 degrees at 20 knots gusting to 30 knots.

 The single runway at Fort Leonard Wood was Runway 14 and 32. That meant there was way too much crosswind to land a Cherokee Six at Fort Leonard Wood, so I diverted to our maintenance base at Vichy, that served the city of Rolla. Vichy was about 40 nm from Jefferson City.

After landing, and as I was putting my flight bag into a company car that I would drive to Fort Leonard Wood, it began to snow.

Fate Is Still the Hunter!

The mystery of why the Cherokee Six I was flying that night had accumulated so much static electricity was solved when I returned to the Vichy airport a few days after the snow storm had abated. I was supposed to ferry the Cherokee Six back to Fort Leonard Wood solo as a positioning flight.

While doing my preflight inspection in bright sunshine, I discovered that both of the bonding straps that connected wings to the the ailerons were broken. These ¾ inch wide braided steel straps were designed to bleed static electricity from the wing to the ailerons and then to their sharp trailing edges where it would be disapated into the slipstream. This setup was used in lieu of static wicks, which did the same thing, only more efficiently.

With the straps no longer connected to the ailerons, static electricity on the wings could not be bled off. This is what  caused the giant spark that leapt from the windscreen to the top of the instrument panel, causing the Saint Elmo’s Fire to occur.

Fate is still the hunter!