An ILS is the best place to have an engine failure in a light twin. You are already configured, using low power and planning to land. In general you want to think very carefully before going around in real life ie I would need a really good reason to ever go around OEI in a piston twin.
Peter wrote:
and if you get one while flying an ILS then unless you are really on top of your game you are going to get killed, either right away or when doing the go-around.
Hmmm, incorrect.
Firstly, an engine shutdown in a DA42/62 is a very simple affair. Anecdotally, I’ve had four in-flight shut-downs in PA31s and one in Diamonds. Hours are pretty much even between the two.
Anyway, I can’t think of a better place to have an engine failure other than when established on the glide. Plenty of time to identify and get yourself sorted for the (inevitable) go-around. 
and if you get one while flying an ILS then unless you are really on top of your game you are going to get killed, either right away or when doing the go-around.
Do you have any first-hand experience with doing OEI approaches or go-arounds?
Peter wrote:
if you get one while flying an ILS then unless you are really on top of your game you are going to get killed, either right away or when doing the go-around.
At low power settings and relatively high speed, it should be relatively benign (if noticed in time). If not noticed on the go around, that’s a totally different thing. I was told (possible speculation) that the twin crash at Hawarden had been a loss of control in the go-around (possibly started because of the said engine failure)
Peter wrote:
anecdotally it is a lot higher than 2x, especially, historically, on the DA42 on which most private owners I know have had an in-flight shutdown
Engine failures on a same plane are far from independent. There is certainly some positive correlation, due to many factors:
- bad maintenance: likely to have been done the same way on both engines
- design flaw: idem
- fuel contamination
- icing
- pilot mishandling engine control (or previous pilot)
- overheating
- etc etc
Still, the probability of having no engine working on a twin engine is obviously MUCH lower than having no engine on a single engine.
How much does he have to know about go/no-go criteria for real-world flying to pass the checkride? Granted, a VFR student might not be better, but then all he really has to worry about is cloudbase and visibility. For IFR, it is much more complex than that.
Exactly right. A RHS gets lots on-the-job training from an experienced pilot, which is not the case a PPL/IR flying single-pilot aircraft. So the PPL/IR needs to gradually gain experience, test the limits. Not unlike a newly minted PPL, except that you can really get in over your head in IFR.
But a new VFR pilot may have his share of challenges too. A bumpy approach, a landing in strongish crosswind, flight into degrading visibility, a challenging nav…
The risk of an engine failure when flying a 3 deg glideslope is very very low. Low power, a small time window, and the engine has proved itself over the whole preceeding flight.
The risk of an engine failure in a twin is 2x that of a single (anecdotally it is a lot higher than 2x, especially, historically, on the DA42 on which most private owners I know have had an in-flight shutdown) and if you get one while flying an ILS then unless you are really on top of your game you are going to get killed, either right away or when doing the go-around.
EuroFlyer wrote:
200’ minima in an SEP means slight problems with survival in case of an engine out
Correct, but of course the associated risk is proportional to your exposed time. If such conditions are widespread along the whole route, I would not launch in an SEP. If they exist only at the destinaton (very common – think low coastal stratus or advection fog), the risk of an engine failure during the few minutes on the approach with a low power setting is generally acceptable to me.
Peter wrote:
Actually somebody who just passed their IR checkride is probably a very good instrument pilot
That depends heavily on the training. I dare to say even more so than on the PPL-VFR level, at which a minimum standard is ensured e.g. with the required solo cross-country flying. An instructor signing a student off on that knows that it happens on his own ticket and makes sure the student is able to cope with at least a certain amount of unexpected stuff. An IR student will never fly solo and training might consist of the same approaches over and over at the few suitable airports in reasonable distance…
The newly-minted IR holder PROB90 is very good at flying the needles and the procedural part of IFR flying (approach briefings, cockpit flows…). So he is just as prepared to execute an approach down to the DA/MDA as a more experienced pilot – because that’s what he was trained for and what he had to demonstrate on the checkride. Minimum visibility is a totally different thing, as this cannot be effectively simulated in the plane and IMHO is more difficult to tackle than cloudbase.
But is this what IFR flying is about? Is keeping the needles centered the demanding part of IFR flying? How much does he have to know about go/no-go criteria for real-world flying to pass the checkride? Granted, a VFR student might not be better, but then all he really has to worry about is cloudbase and visibility. For IFR, it is much more complex than that.
Why not?
Because of the risk of an engine failure. Same debate as whether one should fly at night or not.
