JAR and S.E. safety

[Bunnion]

We have all been saddened and concerned in recent weeks at the alarming number of accidents. My attention has been drawn to the two in the U.K., the Griffon in Canada, a Schweizer in CA (where a rear strut cracked..non fatal) and the posting by Nick Lappos of the Sea King with the T.R. failure.

Having flown in a few different countries with different perspectives I sometimes feel that the JAA obsession with engines failing is somewhat misdirected. Would like to hear others views...

For example in the U.S. the pilot is held responsible for determining that an aircraft is safe for flight regardless of who does the daily. I routinely saw pilots up at 6am opening cowlings on S61s and generally crawling all over them something you would never see here in the U.K. Starting up with a cowling open or tie down on was his/her responsibility alone, end of story. I remember a notice at one company emphasising the need to thoroughly look over your blades before and after every flight. On the other hand he would then get in and fly 19 pax Cat B... Often wonder who has the best approach....

[Helinut]

IMHO:

By any measure of objective risk management, regulatory obsession with engine failure in modern single engine helicopters is absurd. Just look at the accident/incident history and see how few incidents today are caused by single engine failure and how few of those that were lead to injuries. It is, however, so firmly fixed in the minds of the regulators (particularly our lot in the UK) that the chances of being able to move them from their obsession is pretty close to zero.

This is not just saying lay-off regulating helicopters altogether. It is rather saying that there are much more important hazards to control about helicopters that kill and injure people.

There are some historical reasons for the obsession with engine failure. In times gone by, engine failure rate was a fair bit higher than it should be in modern helicopters. The way that the UK Air Navigation Order is worded it identifies single engine failure (wrongly) as a uniquely important hazard. It therefore makes special provision for engine failure in SE aircraft in its Articles. In a way, the ANO is really a historical document, rather than what it ought to be a modern set of regulations for the management of safety in aircraft.

The creation of JAR OPS 3 was a unique opportunity to look at things afresh. Unfiortunately all that seems to have happened is that the national regulators have tried to perpetuate their previous regimes. A missed opportunity I am afraid

[Roundagain]

Agreed - CAA emphasis on SE limitations should be directed towards pilot training/experience and CRM issues. Bar the few mechanical failures that occur, these are the most obvious contributory factors in helicopter accidents that I read about every day. :(

[jbower]

If engines do not fail it would seem logical to conclude that learning and practicing autorotations may not be needed. I don't have any statistics but anecdotal evidence indicates more accidents and incidents occur during simulated engine failure than actual engine failure. Should the regulators ban practice auto's?

[paco]

Most companies in Canada now do power-on autos - examiners are looking for the correct entry into the auto, the correct flare point, then after that it's regarded as a coordination exercise. I personally feel happier doing a full-on to the runway, because then nobody needs to watch for the torque spike (on the jetbox, anyway) - that's an accident waiting to happen.

Phil

[Nick Lappos]

This thread matches my thoughts exactly. I believe the blind assurance of Cat A capability from everywhere in the flight is a throw-back to earlier days where the engine failure rate justified this attitude. Now, the regulation of pure Cat A is the easiest way to flex a regulatory muscle to appear to advance safety, without using the brain to actually attempt to make things safer.

The statistics speak for themselves. What gets us is generally labled "Pilot Error" but is many times the result of the inevitable scatter of human performance while doing difficult tasks. The more difficult the task, the greater liklihood for failure. A baseball pro can hit the ball about 1/4 of the time to be successful. An offshore pilot has to do a bit better!

In helicopter aviation, the environment and procedures that have worked in the past rely on a great deal of judgement and skill. The greater that reliance, the more likely a well trained, skilled human will fail.

I have always advocated that we should derive the growth of our regulations from what is broken, and not fix what is doing well. The crime in the OEI regulations that are being imposed is that they will spend money, weight and design effort where the problem is small, and this will divert such attention from the real issues, such as CFIT, maintenance error, and the like.

For an excellent source of accident data, see the attached pointer, which leads to a 300K PDF file from the Oil and Gas Producer's Organization, the summary of 2 years, 1,400 aircraft and 1.4 million flight hours. It was written by some great aviation professionals, and should serve as a rally point for these discussions.

http://www.ogp.org.uk/pubs/300.pdf

[t'aint natural]

I think the "responsible authorities" are perhaps more inclined towards our way of thinking than we might allow. When I started flying 18 years ago there was no requirement to study human performance and limitations at the PPL level, and only the most enlightened PPL instructors touched on the topic. But now there is an increasing emphasis on human rather than mechanical factors by those who enforce the rules.
The JAA and the CAA may not yet have the courage to address the nonsense of SE restrictions as currently practised in the UK, but there is some give in the system.
Incidentally, I have tried without success to find a case of engine failure in a Robinson R22 that was not down to mismanagement - running out of fuel or oil, or carb icing. With the exception of one that was caused by crimping of the fuel vent, I haven't been able to turn one up. Anyone got one?

[Helinut]

t'aint natural

I think it would be a bit optimistic to think that the CAA's interest in CRM and human factors indicates they are going to change their stance on SE.

Depends a bit what you mean by engine failure re: the R22 (and for that matter pretty much any other piston engne helicopter)

There were (and probably still are from time to time) a number of engine "events" caused by sticking exhaust valves. The severity of the effect on the engine varies but, as I recall, several lead to fatal accidents due to loss of RRPM or screwed up EOLs. I had one in an R22 which was sufficiently serious for both of the POB to think an auto and EOL were needed - fortunately without damage to aircraft or occupants in that case.

I also recall a couple of cases where the oil lines going to the oil cooler from the engine failed - not stricly engine failure I suppose but essentially lead to engine failure within a short space of time.

There was also another where an R22 had just been "rebuilt"/2000 hr overhauled. Can't remember the exact cause but something lose on the engine to do with the alternator I think lead to a piston/con rod poking out of the side of the engine. This fortunately occurred in the hover.

I am not trying to suggest an R22 engine failure is at all likely but with the vast number of total hours some events do occur. What does seem to be highly unlikely is a "dead-cut of a piston engine - other than due to lack of fuel of course!

If you really wanted to get a good handle on this you could ask the CAA Safety Data Analysis Unit for a set of MORs on the R22 - for the purposes of safety studies they will give you this info free. This only covers UK incidents (I think they may have changed their name recently).

[Avnx EO]

:( I'm not sure if people realize the half of it. Anyone who has been involved with a Cat A certification effort recently will have probably encountered the South West's interpretation for systems cert - that "Cat A" means the engine failure is "a given". So you get no credit for the failed engine in systems design. This means that systems must provide the same level of availability both before and AFTER the first engine has failed. The remaining engine itself doesn't provide more than 10e-5 (even though dual engine failure in IMC in part 29 is typically considered Catastrophic), and the remaining electrical system is close behind - but the regulatory agencies counter that they are willing to "make exceptions" in those areas, because "the technology isn't there." - But for flight and engine display systems (and for that matter, radios and everything else) "we can, and therefore we should in the interest of safety"

Now consider for a moment the EICAS (engine indicating and crew alerting system.) The criticality of the EICAS system actually INCREASES after the first engine failure (because if I mis-manage the reamining engine, I'm REALLY cooked.) With no credit for the engine failure itself, and with no other engine to perform cross-checking, I now have to provide a much more complex system, with a lot more redundancy in order to meet this new criticality. (triple Ng sensors, redundant processing paths, comparison monitors for multiple data paths, increased Caution messages to show loss of redundancy, miscompares, etc.)

This increased safety for systems is not free. It complicates things something awful, increases the size of generators, batteries, etc. and few - if any - of the current, sans-FADEC, steam-gauge cockpits could be certified today if this same interpretation were applied. The weight and cost devoted to over-designing these areas takes away from performance and other electronics. On one hand, I agree that for Cat A we need to assure that remaining systems are operational when they are most needed, but on the other hand, giving no credit for an engine failure is absurd. Most engines will give you 10e-4 or 10e-5.... Hell, if they just gave us 10e-2 credit for the sake of systems certification, you could assure more than reasonable system availability, given the fact that you should no be faced with this every day.

Anyone else been running ito this??