A look at RV accident information by Bob Collins

Before I begin this article, this disclaimer: I'm not an accident investigator, I'm sure as heck not a statistics expert, and I haven't spent as much time dissecting NTSB accident data as a lot of other RVators. I'm just a writer with a sense of curiosity.

Last week, like many of you, I was pouring over data and articles following the Cirrus accident in New York. Working in a newsroom and all, I tried -- and failed -- to intercept the "how safe is the Cirrus" story and direct it more to a look at pilots, not planes. It wasn't the best week I've ever spent in the business and, frankly, it was the first time in 30 years in the news business that a newsroom didn't turn to me for help in covering a story about aviation.

But all was not lost. I learned a lot about the Cirrus, about its pilots, and about RVs, oddly enough. See, I still have a theory that the marketing push to put the average rich guy into a brand new Cirrus, as described in the fine James Fallows book, "Free Flight," would show up in accident data. Maybe it has. Maybe it hasn't. As usual, nobody can seem to agree.

I ran across a very comprehensive blog by Cirrus owner Phillip Greenspun, after I found a terrific review of the SR20 and SR22 in which he addressed this suspicion. "I think the beginner can be safe in an SR2x provided he or she schedules an immediate 50-hour intensive instrument training cross-country program. He or she will have 15 hours of UND training at the Cirrus factory and then 50 hours of dual traveling, say, from Duluth to San Francisco to Florida and back to San Francisco," he said.

I also noted the comments of the head of the Cirrus owners grou, Mike Radomsky, who made the observation that all the safety features of the Cirrus, ironically, would entice more people into the air who "have no business flying."

And I recalled a 2001 article in which, as I interpreted it at the time, Richard L. Collins seemed to question the effect of novice pilots in a hotrod, but as I look back, he was pointing out that the accident rate for the new plane was higher than the general fleet. According to a reporter at my workplace, one who isn't real happy with me right now, Collins denied suggesting such a thing in a phone coversation last Thursday afternoon. Thanks, Dick!

Ironically, after getting home from work last Friday after a terrible day of trying to get a group of editors and reporters to explain how a newsroom doesn't turn to an aviation-knowledgeable editor on matters of aviation (I failed), there was this months' Flying Magazine, and Collins' article in which he takes on, you guessed it, accident records.

He noted that his son has been studying accident records and has noticed a correlation between time in an airplane, and the accidents that happen. The 2001 article, had an intriguing statistic. At the time, Cirrus was listed as the best of the new planes at the time, since the first fatal accident didn't occur until 145 of the company's planes were flying. High, indeed, but not for new planes.

That's what got me thinking about RV accident rates. Think about that figure -- 145. There are, according the Van's Web site, 360 RV7/7As flying now. According to the NTSB records, there's been only 1 fatal accident involving that aircraft. Unfortunately, I don't know how many were flying at the time of the December 2003 accident . Since I am builder #240 and I received my kit in June 2001, I'll bet it's close to, but probably a little short of that magical 145 figure .

The RV-9/9A is more impressive. There has never been a fatal crash involving a 9/9A and there are 240 of them flying as of last Sunday. Beat that, Cirrus. And with the completion of 52 more RV-10s, that model will also eclipse the Cirrus record.

(Update: Kevin Horton sends the following: You'll probably get this from a bunch of people, but there has been one fatal RV-9A accident. N96VA, the prototype RV-9A (called the
RV-9, as at that time they didn't plan to do a TW version) was lost in a fatal accident. Bill Benedict (Van's employee) and his son Jeremy were killed on the way to Sun n Fun in April 2000. For some reason the FAA registration has the aircraft as an RV-6T, but it was
in fact the prototype RV-9.)

That, to me, is an impressive record, especially when these are built by amateurs. Why the fine record? Caution: guesswork ahead. Although I don't know for sure, it wouldn't surprise me a bit if the answer is the background of the pilots. Few RV builders are novice fliers; low-time, yes, sometimes, but rarely is the RV the first plane you fly. That's not the case with Cirrus, which has often considered its competition to be BMW, or Jaguar. You know, cars. Fast cars. Fast luxury cars.

Let's look a little deeper into the individual models. According to the NTSB statistics, available online only back to early 2001, the RV 6/6A model has had the most accidents involving fatalities. This is hardly surprising; the RV 6/6A is the most popular model in the Van's fleet. There are almost twice as many 6/6As flying than the next-most-popular plane -- the RV 4.

The high total number of accidents in the 6/6A, however, improperly suggests a high accident rate. According to some quick calculations, .72 percent of the 6/6As in the fleet have been involved in a fatal accident. In fact, that is a record that is miniscule, although it is higher than the RV-7/7A (.28 percent), the RV-4 (.67 percent) and, of course, the RV-9/9A (0%).

I don't have the background -- or the experience -- to make any conclusions other than the one the statistics suggest, recognizing that there's probably a lot more that goes into these sorts of things: the fatal accident record of new Van's aircraft is likely better than the Cirrus or most production aircraft.

How could this be? What am I missing? We're amateurs building homebuilt airplanes?

A potential answer, as it does in the crashes involving Cirrus, could involve a look at the pilots more than the planes. As Richard L. Collins' son apparently has discovered, there's a relationship between a fatal accident and the time a pilot has spent in the individual airplane.

Unfortunately, this relationship is difficult to assess in homebuilt airplanes because, for some reason, NTSB investigators do not have a consistent policy, apparently, of determining -- or at least documenting -- the time spent in an accident aircraft, even though pilot logs are quite often available. A cursory look at the statisics seems to reveal the failure to do so depends on what NTSB/FAA field office is in charge of the reports.

However, there is enough anecdotal evidence to allow the "researcher" to engage in speculation. Given Cirrus' -- and many production aircraft manufactuer's -- intent to get non-pilots into their aircraft, it seems a given that no-time pilots are jumping into the accident aircraft. This is not to say that the marketing plan is responsible for the accident, but it's not to say it's not either.

We know, from people smarter than me, that there is a "killing zone" of pilot time. In "The Killing Zone: How and Why Pilots Die" (available in the RV Builder's Hotline book section), Paul Craig concludes the first 50-350 flying hours -- the killing zone-- is the time when most pilots die in accidents.

How does this factor in fatal RV accidents? Of the 31 fatal accidents since 2001 examined, the average total pilot hours was over 1,000 -- 1,674 to be exact. Included in that average, are 14 pilots who had over 1,000 hours, a testament, I think, to the number of commercial and ATP certificated pilots who are attracted to the RV line. But almost exclusively, the RV aircraft is not the first airplane an RVator has flown; just the first one built. And isn't it also possible that the building process helps the pilot know more about the construction, limitations, and flight characteristics of his or her aircraft?

But even if you took all of those pilots out of the total, the average will still be remarkably high. Only three pilots has total flying time of fewer than 400 hours. One accident did not record the total flight time.

So, at least where RVs are concerned, and at least in the accidents for which data is available, the "Killing Zone" theory does not apply.

What may apply, although it's difficult to say since the records are not consistently kept, is the theory of Richard L. Collins' son. The time actually spent in the accident airplane. But, even then, the total hours average is surprisingly high. There are 16 investigations in which total time was indicated, for an average of almost 180 hours. Still, that's quite a few hours in an individual aircraft. Might the average be skewed by a few high-time accidents? Possibly. There were 6 fatals of less than 40 hours, but seven of more than 200 hours. It sure would be helpful if the NTSB/FAA were more diligent in recording this statistic.

And finally, purposes of this conversation, the flight phase where most accidents occurred is telling. Hands down -- and not surprisingly -- it's the landing phase.

The NTSB/FAA is even less diligent in recording time flown in the accident aircraft in the preceding 90 days. Only 10 such records exist -- again, back to 2001 -- and the average time is 17 hours over the preceding 90 days.

So what does this all mean? These thoughts, keep in mind, are almost exclusively anecdotal. For me, they're part of a longer research that I'm only now beginning. But as several threads in the last week have shown, the RV is an easy airplane to fly, and while accidents happen and risks exist, the accident statistics speak well of the machine, and the pilots.

Return to RV Builders' Hotline home | Subscribe to the RV Builder's Hotline (free!)