Now, I've put the sunglasses on
which means (a) I'm confident of what I'm about to say, though with the caveat that it applies to UK CAA/JAA operators and the requirements placed thereon, and (b) if you disagree, your glaring red faces won't upset me too much.....
First, may I say I'm astounded at the lack of knowledge and awareness displayed relating to this issue. Nothing personal, but Crossunder's post above illustrates my point very clearly.
In general terms, an aircraft commander must know he can make a flight safely. It is only in those terms that Crossunder's comments make sense. However, the regulatory authorities are responsible, again in very general terms, for creating the information environment, and they have not seen fit to provide or require sufficient data to make other than an extremely cursory analysis of the terrain around airports.
First FACT: If there is no emergency turn, then the NTOFP takes you safely straight ahead off the runway to 1500ft and 25nm. In those 25nm you will normally plan to make MSA, and achieve this. Perf A guarantees that the Nett aircraft will be safe.
Second FACT: No aircraft performance data (that I have seen or worked with) analyses SIDs for terrain. Why should they? SIDs are written to get an aircraft from one place to another expeditously, in a manner which suits local ATC requirements, and minimises noise nuisance. SIDs are not of interest to most performance planners, and many seem to pay no heed to them at all. Indeed, one performance provider often says 'Follow the SID' for engine failure in cases where the SID takes you sraight towards the mountains (the performance provider's name is hard to spell, and the airfields involved are coastal Mediterranean). If you doubt this, ask yourself what you would do if you arrive at work and there is a temporary change to a SID - and your performance data does not reflect this change, or when (almost everyday) your departure instructions are to deviate from the SID track and fly a heading. Can you not, then take off, if you cannot fly the precious SID?
The reason for the lack of terrain analysis in SIDs is that (i) the terrain information is not available other than in some instances at great cost and in a format which makes manipulation difficult and (ii) by the time you start to throw in turn performance, raw data navigational accuracy, wind effects, and so on, the calculations become too difficult and vague to be of value, and thus would often be very limiting.
So, that is the performance-related case for following the EOSID or emergency turn (basically the same thing only different), or going straight ahead where no other procedure exists. Of course, as commander, you may do what you like, but if you clang into the hill, you'll carry the can.....
Now, why shouldn't you follow the SID? Well, first, we have seen that there's no assurance of terrain separation. Second, some SIDs involve turns very soon after take-off. Even in an acceptably modern aircraft, engaging the autopilots and executing ENG OUT in the FMC will leave you with drills to do and high workload while you monitor what the aircraft does. In the small Seattle product, do you want to be flying a forty degree turn manually at five hundred feet after take off with an engine on fire and failing while trying to carry out the appropriate drills and keep an eye on your colleague? I don't believe that's a sensible option.
Finally, there is some confusion in relation to ICAO document 8168 and what it says on this issue. Without my copy to hand, I recall that this document says you should follow the 'departure route' following a malfunction. This does NOT mean the 'Standard Instrument Departure', but relates to the route you plan to fly.
From the ATC point of view, once you have said 'Pan' or 'Mayday', you tell the controller what you're doing and he'll help you as much as he can. Here, we meet the issue of airspace, and again I cannot see the logic in, for example, deciding to follow a SID which takes you towards very busy airspace, in the event of a problem. So, brief staright ahead until at a safe altitude (not necessarily MSA; the radar controller will have a much, much lower, safe vectoring altitude which he can tell you about).
Remember the KISS principle: Keep It Simple and Safe.
I do believe, however, that this thread should be compulsory reading for performance departments and Chief Pilots. It illustrates how many are operating in the dark......
