Go-around after engine failure in light twin
Join Date: Jan 2001
Location: He's on the limb to nowhere
Posts: 1,981
Likes: 0
Received 0 Likes
on
0 Posts
WCollins,
This interesting NTSB Statistical Summary of General Aviation accidents includes accident rates for single and multi piston engine planes.
In 1997 (latest year available) the accident rate per 100,000 aircraft hours flown were
Single: 8.06
Multi: 6.36
So singles had more accidents, but for Fatal accidents the figures were
Single: 1.47
Multi: 1.92
So multis had more fatal accidents. Whether these are significantly different is not stated.
Yep, even if you are well trained you can get into a lot of trouble in those first fifteen seconds of flying a multi. If you are not well trained you can get into a lot more trouble in a multi than a single during the rest of the flight.
This interesting NTSB Statistical Summary of General Aviation accidents includes accident rates for single and multi piston engine planes.
In 1997 (latest year available) the accident rate per 100,000 aircraft hours flown were
Single: 8.06
Multi: 6.36
So singles had more accidents, but for Fatal accidents the figures were
Single: 1.47
Multi: 1.92
So multis had more fatal accidents. Whether these are significantly different is not stated.
Yep, even if you are well trained you can get into a lot of trouble in those first fifteen seconds of flying a multi. If you are not well trained you can get into a lot more trouble in a multi than a single during the rest of the flight.
Sub Judice Angel Lovegod
Join Date: Oct 2002
Location: London
Posts: 2,456
Likes: 0
Received 0 Likes
on
0 Posts
Slim_Slag
These are oft-repeated boring chesnuts, but, as I am sure you know, the mission profile for twins in terms of weather, terrain etc is considerably different from singles.
Most people who consistently fly in poor weather, over sea, at night, over mountains etc are sensible enough to do so in a twin.
If, heaven forbid, it was normal to do these things in SEPs the figures would be much different.
IMHO
W
These are oft-repeated boring chesnuts, but, as I am sure you know, the mission profile for twins in terms of weather, terrain etc is considerably different from singles.
Most people who consistently fly in poor weather, over sea, at night, over mountains etc are sensible enough to do so in a twin.
If, heaven forbid, it was normal to do these things in SEPs the figures would be much different.
IMHO
W
Join Date: Mar 2000
Location: TL487591
Posts: 1,639
Likes: 0
Received 0 Likes
on
0 Posts
On that self-same subject of boring old chesnuts, it can only be a matter of time before somebody suggests that engine failures in twins are more common than in singles.
Looking at that "fact", it has to be remembered that the pilot of a MEP is trained to shut an engine down and continue whenever he sees a condition such as low oil pressure or astromical oil temperature. In an SEP, your best option is to press on, and hope that nothing fails before you find somewhere to put down.
Looking at that "fact", it has to be remembered that the pilot of a MEP is trained to shut an engine down and continue whenever he sees a condition such as low oil pressure or astromical oil temperature. In an SEP, your best option is to press on, and hope that nothing fails before you find somewhere to put down.
Join Date: Jan 2001
Location: He's on the limb to nowhere
Posts: 1,981
Likes: 0
Received 0 Likes
on
0 Posts
The problem with "oft repeated boring chestnuts" is that they get repeated so often they pass into folklore, and nobody remembers whether they are based upon fact or just another story told over a case of beer at a hangar party 
Round here plenty (in fact the vast majority, twins are expensive to run) of people fly SEPs for the missions described above and the figures are what they are. Ok, lets get back to the numbers and look at total/partial engine failures due to mechanical causes - and I assume at the end of the day this is what causes SEPs to be a riskier proposition for flights than MEPs.
Around 6% of accidents are caused by total or partial engine failure of mechanical cause (Chart 8).
More importantly, 2.5% of fatal accidents are caused by total or partial engine failure of mechanical cause (Chart 9). In my opinion that's not much, and given the number of flights that are safely completed every day just goes to show how incredibly reliable engines are. I'd even bet that most of those failures are due to poor maintanance so have a human element.
Of course I'm now making the mistake of using selected figures in the report to back up my hypothesis that piston singles are no more dangerous than piston twins. If we are not careful it won't be long before that become another oft repeated chestnut.
Round here plenty (in fact the vast majority, twins are expensive to run) of people fly SEPs for the missions described above and the figures are what they are. Ok, lets get back to the numbers and look at total/partial engine failures due to mechanical causes - and I assume at the end of the day this is what causes SEPs to be a riskier proposition for flights than MEPs.
Around 6% of accidents are caused by total or partial engine failure of mechanical cause (Chart 8).
More importantly, 2.5% of fatal accidents are caused by total or partial engine failure of mechanical cause (Chart 9). In my opinion that's not much, and given the number of flights that are safely completed every day just goes to show how incredibly reliable engines are. I'd even bet that most of those failures are due to poor maintanance so have a human element.
Of course I'm now making the mistake of using selected figures in the report to back up my hypothesis that piston singles are no more dangerous than piston twins. If we are not careful it won't be long before that become another oft repeated chestnut.
Join Date: Aug 2002
Location: Surrey, UK.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
For my better understanding, do BFRs for multi-pilots under FAA regs have to be completed in multis?
I am trying to understand if, once I have my FAA MEP rating, I am ever/regularly compelled to be re-examined in a twin.
TIA
I am trying to understand if, once I have my FAA MEP rating, I am ever/regularly compelled to be re-examined in a twin.
TIA
Guest
Posts: n/a
W Collins/Slim Slag
There are lies, damned lies and statistics!
In risk evaluation, one needs to consider severity and probability as a system.
Thus whilst the probability of a problem may be higher in SEPs, we might also agree that the severity is lower, due for example to lower glide speeds, lower weights and easier diagnosis of cause/effect etc.
I fly as a pax on airliners regularly, knowing full well that severity could be very high, but probability is very low, not least due to the risk avoidance and mitigation embedded in airline SOPs. Does this mean that airliners are potentially more dangerous than light aircraft? (not in the view of the average person in the street I bet!)
So it all depends on how you view severity and probability
Until someone does the exercise fully and models the variations of mission, pilot ability etc and then calculates the standard deviation against a normal distribution curve, I suspect the SEP vs MEP safety argument will tend to reflect subjective views!
Interesting thread.
There are lies, damned lies and statistics!
In risk evaluation, one needs to consider severity and probability as a system.
Thus whilst the probability of a problem may be higher in SEPs, we might also agree that the severity is lower, due for example to lower glide speeds, lower weights and easier diagnosis of cause/effect etc.
I fly as a pax on airliners regularly, knowing full well that severity could be very high, but probability is very low, not least due to the risk avoidance and mitigation embedded in airline SOPs. Does this mean that airliners are potentially more dangerous than light aircraft? (not in the view of the average person in the street I bet!)
So it all depends on how you view severity and probability
Until someone does the exercise fully and models the variations of mission, pilot ability etc and then calculates the standard deviation against a normal distribution curve, I suspect the SEP vs MEP safety argument will tend to reflect subjective views!
Interesting thread.
Join Date: Mar 2000
Location: TL487591
Posts: 1,639
Likes: 0
Received 0 Likes
on
0 Posts
rustle asked:
The answer is no. A BFR conducted in a single will cover you for your MEP, unless it is large enough to require a type rating.
Interesting eh?
For my better understanding, do BFRs for multi-pilots under FAA regs have to be completed in multis?
Interesting eh?
Join Date: Aug 2002
Location: Surrey, UK.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
2Donkeys
The answer is no. A BFR conducted in a single will cover you for your MEP, unless it is large enough to require a type rating
Thanks - I had read that somewhere, but needed clarification.
The story I heard was that, because it's cheaper to hire an SEP than an MEP, non-owners are more likely to (FAA) BFR in an SEP...
We (CAA/JAA) don't have the choice, obviously, with the mandatory 12 month MEP renewal in an MEP.
So, for sake of statistics, we really need to compare CAA/JAA stats against the NTSB ones to see if there's a trend.
The answer is no. A BFR conducted in a single will cover you for your MEP, unless it is large enough to require a type rating
Thanks - I had read that somewhere, but needed clarification.
The story I heard was that, because it's cheaper to hire an SEP than an MEP, non-owners are more likely to (FAA) BFR in an SEP...
We (CAA/JAA) don't have the choice, obviously, with the mandatory 12 month MEP renewal in an MEP.
So, for sake of statistics, we really need to compare CAA/JAA stats against the NTSB ones to see if there's a trend.
Join Date: Dec 2001
Location: europe
Posts: 546
Likes: 0
Received 0 Likes
on
0 Posts
we need to cmpare a lot more than just that.
The figures look very different if you consider the average GS of a single at 110kts and a twin at 160kts.
The accident rates then ratio at .0727 for singles
.0397 for twins
and fatals ratio at .0133 for singles
.012 for twins.
That assumes the fatals are crashes not victims.
Twins fly for distance, singles often fly circuits or very local area, often with mega hour instructors in them.
Singles often break landing gear due to low hour piloting, twins often have undercarriage problems.
In my experience engines are anything but reliable, and it is impossible to eliminate the manufacturers design weaknesses or poor maintenance from GA.
As always it depends what you want to prove with the statistics.
The figures look very different if you consider the average GS of a single at 110kts and a twin at 160kts.
The accident rates then ratio at .0727 for singles
.0397 for twins
and fatals ratio at .0133 for singles
.012 for twins.
That assumes the fatals are crashes not victims.
Twins fly for distance, singles often fly circuits or very local area, often with mega hour instructors in them.
Singles often break landing gear due to low hour piloting, twins often have undercarriage problems.
In my experience engines are anything but reliable, and it is impossible to eliminate the manufacturers design weaknesses or poor maintenance from GA.
As always it depends what you want to prove with the statistics.
Join Date: Aug 2002
Location: Surrey, UK.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
bluskis
As always it depends what you want to prove with the statistics
Exactly
I believe that 87% of statistics are utterly unreliable, 47.33% are actually very useful indicators, but 10 well trained monkeys might have a different view
My (only) point was that, even in the statistics game, comparing bananas against oranges isn't allowed
As always it depends what you want to prove with the statistics
Exactly
I believe that 87% of statistics are utterly unreliable, 47.33% are actually very useful indicators, but 10 well trained monkeys might have a different view
My (only) point was that, even in the statistics game, comparing bananas against oranges isn't allowed
Join Date: Jan 2001
Location: He's on the limb to nowhere
Posts: 1,981
Likes: 0
Received 0 Likes
on
0 Posts
Bluskis,
I agree totally, what do you want to prove? And I have no idea where you are going with your average groundspeed logic
To prove this point, I think engines are incredibly reliable, you don't. How do I prove this.
I say this because I don't know anybody who has had an engine failure recently in a single (I do in a twin, killed three people, but that was maintainace and a stupid pilot who killed a few others with himself). I know a lot crusty old crop sprayers and freight dogs who had one or two engine failures back in the old days, but not recently. I know a few owners of flying schools who put a lot of hours on their planes, and none have had an engine failure. One had a cracked cylinder head once, but that was caught in time. Give me a few beers and I could really make this sound convincing!
But this is all anecdotal and worth nothing, one of these old chestunts.
Lets have a look at the figures from the same report I cited earlier. In Highlights (Page 1) it says that light piston singles have 8.06 accidents per 100,000 hours, and 1.47 fatal accidents per 100,000 hours.
Same doc says 6% of accidents are caused my mechanical failure
So taking accidents:
8.06 x 0.06 = 0.48 which I round to 0.5 accidents per 100,000 hours caused by mech failure. Thats one accident per 200,000 flying hours. At 100 logged hours per year, you would have an accident every 2000 years.
Doc says 2.5% of fatalities are caused by mechanical failure.
So taking fatal accidents
1.47 x 0.025 = 0.036 FATAL accidents per 100,000 hours caused by mech failure. Thats one accident per 2.75 million hours flown. You would need to log 100 hours per year for close to 30,000 years to be killed by your engine.
I think that's reliable! Probably in the same league as flying an airline (but of course engine failures there are even less common).
And as the argument that singles are more dangerous than twins is based upon engine reliability, I think it's a bum argument.
So what do you think flying3greens, lies or statistics?
I agree totally, what do you want to prove? And I have no idea where you are going with your average groundspeed logic
To prove this point, I think engines are incredibly reliable, you don't. How do I prove this.
I say this because I don't know anybody who has had an engine failure recently in a single (I do in a twin, killed three people, but that was maintainace and a stupid pilot who killed a few others with himself). I know a lot crusty old crop sprayers and freight dogs who had one or two engine failures back in the old days, but not recently. I know a few owners of flying schools who put a lot of hours on their planes, and none have had an engine failure. One had a cracked cylinder head once, but that was caught in time. Give me a few beers and I could really make this sound convincing!
But this is all anecdotal and worth nothing, one of these old chestunts.
Lets have a look at the figures from the same report I cited earlier. In Highlights (Page 1) it says that light piston singles have 8.06 accidents per 100,000 hours, and 1.47 fatal accidents per 100,000 hours.
Same doc says 6% of accidents are caused my mechanical failure
So taking accidents:
8.06 x 0.06 = 0.48 which I round to 0.5 accidents per 100,000 hours caused by mech failure. Thats one accident per 200,000 flying hours. At 100 logged hours per year, you would have an accident every 2000 years.
Doc says 2.5% of fatalities are caused by mechanical failure.
So taking fatal accidents
1.47 x 0.025 = 0.036 FATAL accidents per 100,000 hours caused by mech failure. Thats one accident per 2.75 million hours flown. You would need to log 100 hours per year for close to 30,000 years to be killed by your engine.
I think that's reliable! Probably in the same league as flying an airline (but of course engine failures there are even less common).
And as the argument that singles are more dangerous than twins is based upon engine reliability, I think it's a bum argument.
So what do you think flying3greens, lies or statistics?
Join Date: Aug 2002
Location: Surrey, UK.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
Slim
And I have no idea where you are going with your average groundspeed logic
Guessing here, but I would imagine bluskis has converted the hours into distance, and then created stats based on miles flown, rather than hours flown... "Passenger miles"
And I have no idea where you are going with your average groundspeed logic
Guessing here, but I would imagine bluskis has converted the hours into distance, and then created stats based on miles flown, rather than hours flown... "Passenger miles"
Join Date: Dec 2001
Location: europe
Posts: 546
Likes: 0
Received 0 Likes
on
0 Posts
Russ
Good guess, sorry I did not make it clearer.
Slim
I was trying to prove
1. You can prove anything with stastics.
2. If you want anything useful from statistics you need to do a little work.
As far as engine reliability is concerned I have just run thru a list of occurrences 17 Sept01 to 12 Nov01. The dates are old because it is the list I have to hand.
In these two months there were 20 incidents of engine problems on G reg light aircraft. They resulted in safe landings, pan calls, maydays, forced landings, aircraft severely damaged, and a ditching.
In the last incident the pilot told me he did not intend to fly a single over water again.
I personally have experienced engine problems on average at least every 300 hours.
The only point I am trying to make with these statistics is that it is a disservice to those who may not yet have had engine failure to lead them to believe it won't happen.
It is better that they are aware that it can happen and prepare for when it does.
Good guess, sorry I did not make it clearer.
Slim
I was trying to prove
1. You can prove anything with stastics.
2. If you want anything useful from statistics you need to do a little work.
As far as engine reliability is concerned I have just run thru a list of occurrences 17 Sept01 to 12 Nov01. The dates are old because it is the list I have to hand.
In these two months there were 20 incidents of engine problems on G reg light aircraft. They resulted in safe landings, pan calls, maydays, forced landings, aircraft severely damaged, and a ditching.
In the last incident the pilot told me he did not intend to fly a single over water again.
I personally have experienced engine problems on average at least every 300 hours.
The only point I am trying to make with these statistics is that it is a disservice to those who may not yet have had engine failure to lead them to believe it won't happen.
It is better that they are aware that it can happen and prepare for when it does.
Guest
Posts: n/a
Slim Slag
As you asked
I think that your logic is based on a confidence level of 100% - i.e. you are 100% confident that you would be killed after logging 100 hrs pa for 30K years (which given the human lifespan is interesting!), but you don't consider the individual engine lifecycle in your argument.
The mean time between failure for a typical engine is not 100 x 30K, so what is a reasonable confidence level for a typical engine?
Let's just assume for a minute that a notional engine has a TBO of 2000 hrs
Let's say, for sake of argument, that the engine is most likely to fail at 2000hrs, but that it might fail at as little as 1000hrs - failure at more than 2000 is also possible.
I just made these numbers up BTW, to illustrate how probability works and in the real world the engine could fail at any point due to the factors that Bluskis mentions and others.
Apply this to a normal distribution curve and we can have 98% confidence that the engine will have failed by 2860 hours (since 98% confidence is represented by 2 standard deviations, or 860 hours)
However, looking the other way, we can also be 98% confident that the engine will not have failed by just under 1200hrs.
So the numbers say that if we don't fly a/c with engines in the 1200-3000 hour range, we are pretty safe (<2% confidence of failure.)
However, this is based on running 40,000 iterations of the model through a 'Monte Carlo' simulation engine, which smooths out the randomness of the event over a very large scale sample.
Also, if you apply a triangular distribution, for example because you have reason to believe that the engine will be likely to fail at the higher, rather than the lower end of the range, you will get different answers.
If you were a fleet manager running 40,000 SEPs, acquired with with brand new engines at the same time and kept until 1200hrs, then likely you wouldn't be much worried about the odd random event when you lost a hull, because over such a big sample, you'd lose 2% of your fleet and insurance would cover that! At 800 hulls lost per 48 million hours flown or 1 engine failure every 60 000 hours, you say that SEPs were pretty safe.
But in the real world, the experience of individuals will be very different because they are a sample of one, not statistically significant and much more vulnerable to the impact of random events (i.e. the '**** happens' experience), thus creating a broad range of different experiences. When a sample of one crashes that's pretty significant to the sample of one.
Bluskis says "I personally have experienced engine problems on average at least every 300 hours."
I flew a Pup one weekend back in the 90s and it suffered catastrophic engine failure on take off a few days after with another pilot who forced landed safely, but in my 220 hrs I have not directly experienced a failure.
Your experience is different again.
So to answer your reasonable question, your statistics are neither lies, nor damned lies, but rather data to consider very carefully. Do you feel lucky today?
On balance, it's hard to argue with Bluskis view that "It is better that they are aware that it can happen (engine failure) and prepare for when it does."
One thing for sure, me flying a MEP is much more dangerous than an SEP, due to lack of experience and currency - that is unless the light single kills me first!
As you asked
I think that your logic is based on a confidence level of 100% - i.e. you are 100% confident that you would be killed after logging 100 hrs pa for 30K years (which given the human lifespan is interesting!), but you don't consider the individual engine lifecycle in your argument.
The mean time between failure for a typical engine is not 100 x 30K, so what is a reasonable confidence level for a typical engine?
Let's just assume for a minute that a notional engine has a TBO of 2000 hrs
Let's say, for sake of argument, that the engine is most likely to fail at 2000hrs, but that it might fail at as little as 1000hrs - failure at more than 2000 is also possible.
I just made these numbers up BTW, to illustrate how probability works and in the real world the engine could fail at any point due to the factors that Bluskis mentions and others.
Apply this to a normal distribution curve and we can have 98% confidence that the engine will have failed by 2860 hours (since 98% confidence is represented by 2 standard deviations, or 860 hours)
However, looking the other way, we can also be 98% confident that the engine will not have failed by just under 1200hrs.
So the numbers say that if we don't fly a/c with engines in the 1200-3000 hour range, we are pretty safe (<2% confidence of failure.)
However, this is based on running 40,000 iterations of the model through a 'Monte Carlo' simulation engine, which smooths out the randomness of the event over a very large scale sample.
Also, if you apply a triangular distribution, for example because you have reason to believe that the engine will be likely to fail at the higher, rather than the lower end of the range, you will get different answers.
If you were a fleet manager running 40,000 SEPs, acquired with with brand new engines at the same time and kept until 1200hrs, then likely you wouldn't be much worried about the odd random event when you lost a hull, because over such a big sample, you'd lose 2% of your fleet and insurance would cover that! At 800 hulls lost per 48 million hours flown or 1 engine failure every 60 000 hours, you say that SEPs were pretty safe.
But in the real world, the experience of individuals will be very different because they are a sample of one, not statistically significant and much more vulnerable to the impact of random events (i.e. the '**** happens' experience), thus creating a broad range of different experiences. When a sample of one crashes that's pretty significant to the sample of one.
Bluskis says "I personally have experienced engine problems on average at least every 300 hours."
I flew a Pup one weekend back in the 90s and it suffered catastrophic engine failure on take off a few days after with another pilot who forced landed safely, but in my 220 hrs I have not directly experienced a failure.
Your experience is different again.
So to answer your reasonable question, your statistics are neither lies, nor damned lies, but rather data to consider very carefully. Do you feel lucky today?
On balance, it's hard to argue with Bluskis view that "It is better that they are aware that it can happen (engine failure) and prepare for when it does."
One thing for sure, me flying a MEP is much more dangerous than an SEP, due to lack of experience and currency - that is unless the light single kills me first!
Last edited by Final 3 Greens; 10th Jan 2003 at 12:18.
Join Date: Jan 2001
Location: He's on the limb to nowhere
Posts: 1,981
Likes: 0
Received 0 Likes
on
0 Posts
3greens,
Thankyou for your excellent answer to my question, it certainly demonstrates an understanding of the use of statistics - and my question
I don't think statistics can lie except if the data or question being asked is faulty, but you cannot blame stats for that. Statistics will actually tell you if the data being worked on is good enough to answer the question being asked. I think if you have sufficient data, you can explain everything with stats. Quantum mechanics scares me!
Yes, training can turn a fatality into an accident, an accident into an incident, and incidents into a safe flight. The NTSB figures show fatalities due to engine failure are dropping, but don't suggest a cause. One can only surmise that training has a lot to do with it.
As for engines. Make sure you have enough oil and fuel, understand the use of carb heat and mixture and you can save a lot of death from engine failures. There you go, simple eh? For engine failures in multis around take off, make sure you have high quality recurrent training in a sim. Not so simple, but it's what the airlines do. Most engine failures are human caused, engines are very very reliable (as my simple non statistical analysis proved). Not so for humans.
So how does bluskis manage to encounter a problem "on average" every 300 hours. Depends what a "problem" is of course, but are you flying a homebuilt with a subaru conversion? I think a problem "on average" every 300 hours is a problem itself (but stats will answer the question definitively). Actually, if it happened to me "on average", I would be concerned.
I have found fouled plugs on the ground which make me taxi back far more frequently than that, and it is training helps me recognise it. I also don't even elevate that to incident or problem. It's also "probably" training related, the person before needs to understand use of mixture, or a mechanic did something wrong.
From the data I cited, and it is a very large sample from an excellent source, I think you can get a "p" number which will indicate whether singles are more dangerous than a multi. I'd do it, but cannot remember how. Somebody has done this and claims that statistics prove singles are safer.
Aviation Safety Study
I've emailed a request for the report. Doubtless there are experts here who have already crunched the numbers and already know the answer. Insurance company quotes would appear to suggest high performance singles are safer than twins, and lots of multi time and training in a respected environment is the only thing you can do to get the quote down. Which luckily takes us back to the subject of the thread.
Regards
Thankyou for your excellent answer to my question, it certainly demonstrates an understanding of the use of statistics - and my question
I don't think statistics can lie except if the data or question being asked is faulty, but you cannot blame stats for that. Statistics will actually tell you if the data being worked on is good enough to answer the question being asked. I think if you have sufficient data, you can explain everything with stats. Quantum mechanics scares me!
Yes, training can turn a fatality into an accident, an accident into an incident, and incidents into a safe flight. The NTSB figures show fatalities due to engine failure are dropping, but don't suggest a cause. One can only surmise that training has a lot to do with it.
As for engines. Make sure you have enough oil and fuel, understand the use of carb heat and mixture and you can save a lot of death from engine failures. There you go, simple eh? For engine failures in multis around take off, make sure you have high quality recurrent training in a sim. Not so simple, but it's what the airlines do. Most engine failures are human caused, engines are very very reliable (as my simple non statistical analysis proved). Not so for humans.
So how does bluskis manage to encounter a problem "on average" every 300 hours. Depends what a "problem" is of course, but are you flying a homebuilt with a subaru conversion? I think a problem "on average" every 300 hours is a problem itself (but stats will answer the question definitively). Actually, if it happened to me "on average", I would be concerned.
I have found fouled plugs on the ground which make me taxi back far more frequently than that, and it is training helps me recognise it. I also don't even elevate that to incident or problem. It's also "probably" training related, the person before needs to understand use of mixture, or a mechanic did something wrong.
From the data I cited, and it is a very large sample from an excellent source, I think you can get a "p" number which will indicate whether singles are more dangerous than a multi. I'd do it, but cannot remember how. Somebody has done this and claims that statistics prove singles are safer.
Aviation Safety Study
I've emailed a request for the report. Doubtless there are experts here who have already crunched the numbers and already know the answer. Insurance company quotes would appear to suggest high performance singles are safer than twins, and lots of multi time and training in a respected environment is the only thing you can do to get the quote down. Which luckily takes us back to the subject of the thread.
Regards
Guest
Posts: n/a
Slim Slag
I'll give you a better reply than this tomorrow, but regarding stats, there are two factors you should consider.... (a) accuracy and (b) validity.
Accuracy is a given - inaccruate stats are of no use to anyone.
However validity is a bit tricky.... accurate numbers can be invalid if used inappropriately, so we must always think carefully about the stats we're using and what they mean.
I'll give you a better reply than this tomorrow, but regarding stats, there are two factors you should consider.... (a) accuracy and (b) validity.
Accuracy is a given - inaccruate stats are of no use to anyone.
However validity is a bit tricky.... accurate numbers can be invalid if used inappropriately, so we must always think carefully about the stats we're using and what they mean.
Sub Judice Angel Lovegod
Join Date: Oct 2002
Location: London
Posts: 2,456
Likes: 0
Received 0 Likes
on
0 Posts
2, Rustle
It turns out that there was a third explanation
The Aztec is in for C of A and apparently the left engine had zero compression on two pots!
There is always an answer.
W
It turns out that there was a third explanation
The Aztec is in for C of A and apparently the left engine had zero compression on two pots!
There is always an answer.
W
Join Date: Dec 2001
Location: europe
Posts: 546
Likes: 0
Received 0 Likes
on
0 Posts
SS
To put your mind at rest as to the accuracy of my personal stats.
Con rod through the crank case over water, engine definitely stopped. Single engine, landed sucessfully on terra firma.
Engine coughing and low on power after flight over 250 miles of sea, ignition system problems. Single.
Engine bleeding massive amount of oil, engine shut down and diversion to nearest field. Twin.
Increase of engine power unobtainable on approach, approach continued on one engine. Throttle cable anchorage bracket adrift following engine overhaul and CofA. Twin.
The figures of engine based occourances you can obtain regularly together with Flight Safety magazine, and the CAA will probably have figures of total hours flown by GA. When you have done the arithmetic I am sure you will come up with a small figure.
I am now not sure what point you are trying to make, but to assert that engines, together with the mandatory maintenance, are reliable does not stand up.
Right now there are considerable numbers of GA aircraft grounded by AN's concerning engine components.
Please note I have not voiced my opinion on this thread of the relative safety of singles and twins, merely tried to correct what I believe to be a dangerously misleading opinion on engine reliability.
The posters who have pointed out the necessity of relating small probability and severity of consequence have it right.
To put your mind at rest as to the accuracy of my personal stats.
Con rod through the crank case over water, engine definitely stopped. Single engine, landed sucessfully on terra firma.
Engine coughing and low on power after flight over 250 miles of sea, ignition system problems. Single.
Engine bleeding massive amount of oil, engine shut down and diversion to nearest field. Twin.
Increase of engine power unobtainable on approach, approach continued on one engine. Throttle cable anchorage bracket adrift following engine overhaul and CofA. Twin.
The figures of engine based occourances you can obtain regularly together with Flight Safety magazine, and the CAA will probably have figures of total hours flown by GA. When you have done the arithmetic I am sure you will come up with a small figure.
I am now not sure what point you are trying to make, but to assert that engines, together with the mandatory maintenance, are reliable does not stand up.
Right now there are considerable numbers of GA aircraft grounded by AN's concerning engine components.
Please note I have not voiced my opinion on this thread of the relative safety of singles and twins, merely tried to correct what I believe to be a dangerously misleading opinion on engine reliability.
The posters who have pointed out the necessity of relating small probability and severity of consequence have it right.
Join Date: Aug 2002
Location: Surrey, UK.
Posts: 0
Likes: 0
Received 0 Likes
on
0 Posts
WCollins
"...apparently the left engine had zero compression on two pots"
Um, unless the engineer was referring to "left" when viewed from the pointy end as opposed to the traditional "left", that doesn't aid my understanding.
We failed the (real) left
"...apparently the left engine had zero compression on two pots"
Um, unless the engineer was referring to "left" when viewed from the pointy end as opposed to the traditional "left", that doesn't aid my understanding.
We failed the (real) left