Dave_Phillips wrote:
Airborne Again – DA, PA, NPA etc. this is very, very complicated but we need to be careful not to compare apples with oranges. Unfortunately, you have done so in your statement above.
Procedure design, notably calculation of Missed Approach segments does not allow for any ‘bust’ below minima for an NPA whereas it does for a PA.
This could cause a safety problem if the MAPt is very close to the point where you go missed on a CDFA, so that you are still in the “undershoot” when you pass the worst case SOC expected in the procedure design.. This would seem to me a remote problem.
Do you agree with me that from as far as obstacle clearance is concerned, there is no problem with setting the DA=MDA(H) as long as you are not below the MDA(H) when passing the MAPt?
Cobalt, I don’t know about other operators but our procedures stipulate that we must add 50ft to the OCA when flying a CDFA and the OCA is a ‘hard base’. From the FAA:
Derived Decision Altitude (DDA). Pilots must not descend below the MDA when executing a missed approach from a CDFA. Operators should instruct their pilots to initiate the go-around at an altitude above the MDA (sometimes referred to as a DDA) to ensure the aircraft does not descend below the published MDA. Operators conducting approaches authorized by operations specification (OpSpec) C073, IFR Approach Procedures Using Vertical Navigation (VNAV), may use MDA as a DA.
Our job is flight calibration. 
Airborne, we crossed.
Answering you question, yes, I agree. To be pedantic, I would say that the aircraft needs to be climbing above MDA no later than 15 seconds after MAPt. I’m scratching my head to think of a practical scenario whereby you could get into trouble and my thoughts are it would probably be somewhere bizarre like the second segment of the Missed Approach if you are struggling to make the required climb gradient due to engine failure.
I’ll have a ponder with my favourite bedtime reading, ICAO Doc 8168 Vol II. 
Dave_Phillips wrote:
Cobalt, I don’t know about other operators but our procedures stipulate that we must add 50ft to the OCA when flying a CDFA and the OCA is a ‘hard base’. From the FAA:Derived Decision Altitude (DDA). Pilots must not descend below the MDA when executing a missed approach from a CDFA. Operators should instruct their pilots to initiate the go-around at an altitude above the MDA (sometimes referred to as a DDA) to ensure the aircraft does not descend below the published MDA. Operators conducting approaches authorized by operations specification (OpSpec) C073, IFR Approach Procedures Using Vertical Navigation (VNAV), may use MDA as a DA.
Our job is flight calibration.
Yes in FAA land you can’t sink go below the MDA even if treating it as a CDFA. C073 allows Pt91 operators to do it if you are on non-precision approaches with vertical guidance eg LNAV/VNAV (not LNAV+V). But you need procedures and an LoA to do this. While I think we can all rationalise 50ft prob isn’t going to make much difference, how about 100ft, 150ft? It is not a good idea to start ignoring these restrictions.
The relevant rule for determining the DA for a CDFA is NCO.OP.111 and it doesn’t mention any additional margin on top of the OCA(H).
I raised this with my ATO and they agreed this is correct, as the M is deleted in MDA for a CDFA NPA, and we have a DA, ie a Decision Altitude. However, isn’t it the case that for commercial operations the hard MDA principle still applies? Hence ATOs, including the one I raised the question with, still teach to take the ‘decision’ to go missed on a CDFA at MDA plus 50 feet. Arguably MDA plus 150 feet if you happen to be in a 747? The safety margin being based on inertia.
The CDFA should place you well before the MAPt, and you should have gone missed earlier if you are outside the DME altitude checks on the CDFA, hence the SOC design constraint in theory is less applicable for a CDFA?
Dave_Phillips wrote:
my favourite bedtime reading, ICAO Doc 8168 Vol II.
RobertL18C wrote:
The CDFA should place you well before the MAPt, and you should have gone missed earlier if you are outside the DME altitude checks on the CDFA, hence the SOC design constraint in theory is less applicable for a CDFA?
That’s would I would think, too. It would be weird cases like if there is a major reduction in headwind composant after passing the last check distance combined with an unusually early MAPt.
RobertL18C wrote:
Hence ATOs, including the one I raised the question with, still teach to take the ‘decision’ to go missed on a CDFA at MDA plus 50 feet.
We teach that solely for the fact that an examiner must fail you if you go below MDA by one single feet.The 50ft buffer is a rather conservative figure for a single or light twin. If you fly for yourself or for a commercial operator, you – or the operator – is free to define your own margin depending on type of aircraft flown and skill level.
what_next wrote:
We teach that solely for the fact that an examiner must fail you if you go below MDA by one single feet.
Yes, a MDA, but not a DA. When you fly using the CDFA technique, you don’t use a MDA, you use a DA…
I’m not splitting hairs here. It wouldn’t make sense to fail someone for deliberately doing something which is explicitly allowed by the OPS regulations.
Airborne_Again wrote:
Yes, a MDA, but not a DA. When you fly using the CDFA technique, you don’t use a MDA, you use a DA…
According to the (British) examiners which do my own exams I am not allowed to bust this CDFA “DA” by a single foot either. They emphasize this in every pre-checkride briefing. What regulation they base this upon I don’t know, I just make sure that I go-around early enough…
But this does not matter much with regard to this accident, as it was certainly not caused by adding anything to the published minima.
