Let’s make sure we are all talking about the same thing here.
IIRC the RNP approaches into CS from the south are AR approaches, only available to VA, JQ and QF due to advanced aircraft FMS upgrades, extra pilot training/checking and recurrency in the sim. There are no ‘bog standard’ older style GNSS/RNAV straight in approaches from the south. The RNP AR’s due to much lower RNP values and the ability to follow multiple curved flight paths (at the lower RNP values) have paths that weave around obstacles such hills and also avoid settlements for NAP. These approaches are constructed from multiple ‘Radius to Fix’ turns (arcs) hence the use of the term ‘arc’. But I believe that the references above to the ‘arc’ from some posters is the DME arc whereby the ‘arc’ is followed around a single DME value, such as the 24 mile arc, to intercept the final approach LOC through a waypoint, in this case HENDO, the whole procedure to the runway following the arc, FAC and step down procedures as you go. If flown manually, or even with basic autopilot functions, in actual IMC these approaches are a **** fight, as many have identified and I agree having done it many times.
But (correct me if I’m wrong) the problem with the VA and QF incidents is that when the crews were programming the FMC with the arrival STAR and the RNP AR approach, they didn’t select HENDO as a STAR transition. If you do not do this then it doesn’t enter the crossing height at HENDO as 6800A, it just calculates the constant descent path between the previous and next waypoints, leaving you too low at HENDO, hence the ATC warnings. This also happened a few times about 10 years ago when this approach was first introduced and my company went to great pains to reinforce the need for the transition selection for the obvious reason of terrain avoidance.
So, yes, the approaches to RWY33CS, no matter how you actually fly them, are pretty ****ty. If flown in basic modes you’ve gotta be aware of where you are and what’s around\underneath you. If flown in more advanced modes then in has to be entered and checked in the FMC, then monitored throughout to ensure correct compliance with the design approach. BUT having said all that, if there’s hills there, then the RNP AR should be constructed/coded correctly to avoid this situation……I agree, code it for 6800A regardless. But (not saying this was actually what happened) all they had to do was push another button and it would have been all good. Not being critical here just factual (I have no first hand knowledge of the actual events or the circumstances leading up to them)…..I’ve had my fair share of stuff ups as well so not throwing stones.
Not sure they would have got closer and closer, the AR’s do recognise a below path situation and will level off to correct it when the next segment path angle comes into the equation. But, like I said don’t know the specifics, if both crews had grossly incorrect QNH set then the aircraft will fly the entire approach at the incorrect altitude (10 hpa high QNH means 300ft low on path
) and all the indications such as crossing heights on your altimeter match the charted heights but you are actually 300 foot lower in space though.
so hope this helps anyone trying to wrap their head around it. But I think we all agree the arrivals could be better designed.