Thanks to DB for providing his excellent presentation; it is a good for a number of reasons, not-the-least-of-which is that it provides the basis for further debate. However, there has been little comment since it was made available so it might be a good time to look at some of the underlying conditions of offshore approaches so that we can explore what we might, or might not, be able to do to improve procedures.
If I repeat some points/assumptions from earlier posts, I apologise but it is necessary, in the interest of understanding the issues, to draw a number of previously stated points together.
Firstly, the ARA is a non-precision approach which has more in common with NDB procedures than the ILS. It also has less risk (with obstacle clearance) than onshore procedures. Some points about the ARA:
- Before the procedure is flown it is mandatory to ensure that there are no ‘targets’ (radar returns) in the final approach and missed approach segments. This ensures that there are no obstacles (apart from the surface) and those at the destination to consider.
- Onshore procedures constrain vertical descent (and DH/MDH) to ensure that obstacles are cleared vertically; the approach segment is also oriented so that obstacles are avoided laterally (it’s built into the profile).
- Offshore procedures are designed in the knowledge that the vertical extent of any obstacle may not be known; furthermore, it is likely that the highest obstacle in the area is the destination (hence the point about absence of targets in the final and missed approach segments).
- Offshore procedures rely entirely on avoiding obstacles laterally; for this reason, from the OIP, the heading/track has to be divergent to provide lateral separation of 3-400m to one side of the destination structure (even if the missed approach heading is applied later than at the MAPt).
- The MDH is no different to those onshore where, descent limitations should be plus 100ft minus 0 (based upon the RADALT – the setting of the baralt to RADALT is a wise option but only if checked as for normal procedures); these limits should be maintained until visual reference (see below) has been achieved or a go around, at the MAPt or before, has been commenced.
- The approach should be oriented into wind and flown at 60/90kts with a ground speed of no more than 70kts.
- The instrument procedure ends at the point where the pilot can proceed visually (unless there is a go-around at, or before, the MAPt).
- Visual reference has a specific meaning in most procedures; if one goes to Appendix 1 to JAR-OPS 3.430, and looks at the paragraphs that describe this for each type of approach (non-precision; CAT 1; CAT 2, CAT 3 and ARA) it can be seen that this is an important part of the procedure. For the ARA it says “No pilot may continue an approach beyond Decision Range or below MDH/MDA unless he is visual with the destination”. Without further guidance, that statement demonstrates a weakness in the regulation.
So far this procedure (as described) has been non-decelerative within chosen airspeed boundaries. For fixed wing it remains (almost) non-decelerative, the visual segment (in a stable approach) is a continuation of the approach path (be it on ILS or provided by the FMS or another device). This can be achieved because the glide path is continuous to the threshold and fixed wing land-and-stop as opposed to stop-and-land. There are other benefits of fixed wing procedures: the obstacle environment clears at the runway approaches; the aircraft is inherently stable; the field of view remains (reasonably) constant throughout the landing manoeuvre; either pilot can see and complete the landing manoeuvre from anywhere on the approach path.
None of these necessarily apply to helicopters making approaches to an offshore installation.
Because we are approaching the highest obstacle in the area, there can never be a stable approach in the sense of fixed wing, and obstacle clearance has to be achieved laterally; unless the MDH is raised extensively (almost to the level of the height of the installation), there will always be a level segment. This level segment includes a ‘margin of error’ (at the OIP and MAPt) which dictates the location of both of those points; improving the accuracy of positioning of the MAPt (or using a fly-by offset procedure) can move it closer to the destination.
In my view, the actual cloud break procedure is not the key issue – safety benefits from having vertical guidance are nice to have but not critical. The stability of the approach from (when visual) before, or at MAPt is the key issue (if it is just a matter of keeping a speed and height, it is easy to achieve and monitor). The key issue is in having both crew members briefed and/or aware of the envelope that is to be flown (and flying it). The level segment terminates at the point where the slope to the landing manoeuvre is intercepted. The flight path conditions at this entry point (or gate) will be dependent upon the height (distance) at which it occurs.
Let’s construct a flight path: Using a 3° glide would put the helicopter 220m back from the platform at 40ft above the deck (from which point the level segment of the standard landing manoeuvre could be initiated). Assuming a 150ft deck (which is 50ft above the mean deck height of the North Sea) distances would be:
For a 150ft deck and 3° glide, the distances back from the helideck would be:
- 40ft above the deck (190ft amsl) would be achieved at 220m
- 150 above the deck (300ft amsl) would be achieved at 872m
- 250 above the deck (400ft amsl) would be achieved at 1,454m
- 350 above the deck (500ft amsl) would be achieved at 2,034m
For a 150ft deck and 6° glide, the distances back from the helideck would be:
- 40ft above the deck (190ft amsl) would be achieved at 100m
- 150 above the deck (300ft amsl) would be achieved at 435m
- 250 above the deck (400ft amsl) would be achieved at 725m
- 350 above the deck (500ft amsl) would be achieved at 1,015m
The steeper the approach the longer the visual segment (my experience is that pilots are more likely to fly a 6° approach than 3°). I would also like to see what deck ovality is achieved with these approach angles – intuitively my feeling is that the shape on the presentation for ‘too low’ might be 3°, or even 6°.
Up to the slope intercept, we should have stable conditions – i.e. level flight, and a set speed (ground speed no greater than 70kts is recommended in the procedure) and monitoring and crew interaction is relatively simple. If the cloud base is 300ft at night, the intercept point will be 872m with a 3° slope and 435m with a 6°. Either of these intercepts might require a(n announced) deceleration (depending upon wind) but at least it would be in level flight – i.e. monitoring could be relatively easy.
Once on the slope, the NHP will see less and less of the deck until, by 40ft, nothing at all will be seen and further monitoring could be quite difficult. The speed the approach is flown will be dependent upon the wind speed; with light winds, there might not even be an indication of airspeed once the helicopter has decelerated below 40kts. Any monitoring will therefore to be based upon a floor of deck-height plus (say) 50ft and the (minimum) speed that has been declared for the approach (with a probable minimum of 35kts). If either element of this combination is called by the NHP a response would be required from the HP – no response would indicate a loss of control situation and an escalation in the procedure.
In my view, the key to safety is stability; on an ARA the aircraft should be stabilised (height, airspeed and heading) before the MAPt. If there is to be a circling manoeuvre, the aircraft must be stabilised before MAPt (at least 500ft, airspeed and heading); following on from that, night shuttling should be stabilised (at least 500ft, airspeed and heading) at the equivalent of the MAPt – which means any necessary turning must have been carried out and the aircraft stabilised before that point (as there is a requirement for 5km visibility, this should not cause any control/stability problems). Monitoring (up to the slope intercept) should be based upon a declared height of the level plus or minus 100ft and the declared airspeed plus or minus 20kts; any departure outside this envelope should be called by the NHP a response would be required from the HP – no response would indicate a loss of control situation and an escalation in the procedure. In the level segment, any bank of more than 20° or any undeclared departure of more than 20° from the heading should be called and invoke a similar procedure to that above.
I totally agree with DB that once a stable level segment has been set, there should be no descent until the slope is intercepted. There is no necessity for an offset to be applied with night shuttling as the destination will always be in sight and no missed approach is envisaged. A go-around should be performed at any time when either pilot considers that control is being lost or there are insufficient visual cues (the go-around should be discussed extensively and practiced – from the blind side - on periodic training). The use of automatics should be encouraged for all segments where stability is required.
Comments are anticipated.
Jim