Continuous Descent Final Approach NPAs

[212man]

Regarding CDFA Non Precision Approaches, how does the operator arrive at the artificial 'DA' of (typically) MDA plus 30-50 ft? Does this margin really prevent the aircraft from descending below MDA during a missed approach from the 'DA'?

[411A]

Yes, it works OK using 50 feet above MDA.
Then again, I have always thought that this idea was a bit daft...after all, dive and drive has worked for years...for me anyway, even in heavy jet transports.
One might say...'if you can't stand the heat, stay out of the kitchen'.

With credit to Harry Truman...

[Old Smokey]

I agree with 411A on 50 feet being quite adequate for prevention of 'busting' the MDA. If you want hard proof (and runways are pretty hard), on the B777 we use 20 feet as the Cat IIIB minimum, and missed approaches from 20 feet are regularly achievable without runway contact.

That's for an automated Go-Around of course, add a further 30 feet 'fudge factor' for the human manipulated Go-Around, and 50 feet additive to the published MDA is quite adequate, I've seen it numerous times in the simulator without MDA 'busts'.

Regards,

Old Smokey

[212man]

Thanks guys, much appreciated.

[Thridle Op Des]

Hi 212 Man,

One consideration is that the 50' has been used as an assessment for the loss of altitude involved in the dynamic change in velocity vector for a typically high mass fixed wing with a critical engine inoperative. An S92, which is I guess where you are coming from will require a lot less to change its vector even on the now more common single engine consideration. Another aspect that increases the demands on the G/A at MDH+50 for something like a 777 or 330 is the large change in yaw input with the changing thrust on one engine, something the 92 won't have.

Regards

TOD

[dartman]

I recall reading somewhere recently, that you could indeed fly a NPA down to (or 'just' above) the MDA, much like DA on a precision approach, then initiate the G/A? Anything further on that?

dartman...

[alf5071h]

The following text IIRC, originates from JAR-OPS 1 NPA 41 (01/03/06), I do not know if this has been adopted or is included in Subpart E. Anyone cross check please.

Issues related to DA/H and MDA/H
The CDFA (Continuous Descent Final Approach) concept requires all NPA operations to be flown with a decision altitude/height (DA/H). When determining the applicable DA/H, the operator must take account of the missed approach point (MAPt) and the minimum descent altitude/height (MDA/H). While it is quite clear that a missed approach must be initiated not later than at the MAPt, the question of MDA/H is a different matter.
Quite a few operators use MDA/H as the DA/H with no height add-on, in fact this is the case with a majority of the largest European operators. This modus operandii has raised concern that the unavoidable height loss below the MDA/H (during a go-around) might introduce a safety risk, even if the height loss can be minimised by the use of appropriate operational procedures (call-outs, high degree of on-speed/on-path discipline, training). In order to evaluate the safety of the use of MDA/H as DA/H, the AWOSG has compared the obstacle protection for this type of approach with the obstacle protection for ‘traditionally’ flown non-precision approaches as well as with the protection for approaches with vertical guidance (APV) using the criteria contained within ICAO PANS OPS. The
AWOSG is convinced that using the MDA/H as a DA/H offers adequate obstacle protection. Another comparison between the CDFA technique and the ‘traditionally’ flown non-precision approaches indicates that the latter involve several safety traps, such as:
• Early descent with a prolonged flight close to obstacles;
• Multiple step-downs possibly inside the FAF;
• An approach which is, by definition, destabilised;
• Temptation to make a late and steep descent from MDA/H towards the threshold;
• Risk of descending early from the MDA/H;
While there are no records of accidents related to the CDFA type of approaches, there are several accidents attributable to the risks listed above.
The matter of using the MDA/H as a DA/H is progressing in the ICAO OCP and OPSP. Until there exists a final result of the work in ICAO, it must be left to the discretion of each Authority to make decisions on the matter. Since the benefits of the CDFA are generally acknowledged, the decision is typically whether to require an add-on to the MDA/H to ensure that the height loss does not lead to flight below the related MDA/H during a go-around, based on formal or other reasons. It is not the intention of this NPA to imply that an add-on to MDA/H must be required, but rather an acknowledgement that it is an option.

[212man]

Alf, thanks; very interesting and in fact very sensible: by definition the CDFA will result in a go-around from a position much closer to the runway than the MDA will have been based on.

TOD, you are correct in your assumptions, but my question was not so much based on the inertial response of the aircraft, but the actual autopilot reponse to the g-around button being pressed. Having seen it in action (on the type we are talking of) it is a little 'leisurely'!

[Capn Bloggs]

Alf5071h (nice engine, BTW),

Could you please decode AWOSG for me?

In Australia, the MAPt on virtually all NPAs is at the aid, which means that you can comfortably fly down final, stabilised on a CDFA on automatics (3° slope approach from the database), and have the aircraft level off at the MDA (set in the ALT SEL and in the FMS). If you get Visual early enough, all well and good, but if you don't, the jet simply levels off at the MDA and then you start the Go Around when you get to the MAPt. This has the added advantage of allowing a circle to the opposite runway if the bad weather was only localised at one end of the airfield.

However, it would be very much neater though to be able to go to the MDA still on the 3° slope then hit the G/A buttons immediately if not Visual, like an ILS, because you get 50-odd ft closer to the ground, still stabilised, before you have to G/A.

Watching this thread and the official rules re this with interest...

[411A]

Oddly enough, telstar, it ain't just the good 'ole days...still flying the trimotor, and even more strange (for perhaps some anyway) the First Officers I fly with now can do dive and drive just as well as I can.
It just takes a bit of practice...of course the new guys might not get this experience very much, now.
What with map shift and all, I'm surprised many can even find the airport.

[212man]

Captn Bloggs,
AWOSG: All Weather Operations Steering Group.

NPA 41 Progress: AWO CRD and final text proposals for final review at RST 06-3 Oct 06 before NPA goes to JAAC 06-4 Nov 06 for adoption

[Brian Abraham]

A couple of articles by John Deakin on "Dive and Drive"

http://www.avweb.com/news/pelican/182091-1.html

http://www.avweb.com/news/pelican/182092-1.html

His final line is "Sometimes the old ways are still the best!" so I guess he qualifies as a dinosaur and probably drinks from the same swamp as 411A. Not entering the debate as its outside my area of expertise but look forward to the discussion.

[john_tullamarine]

Could develop into an interesting flight standards discussion ... only thing to note for some, though, is the "play the ball, not the player" rule ...

[212man]

I started reading that John Deakin article but had to give up I like the way he illustrates it with MS Flight Sim screen shots of cockpits!

[alf5071h]

212man, et al,
Here is a view from India where operators there apparently use an altitude addition.

Another, perhaps obvious point, is that the vertical path should be constructed to the 50ft threshold ’waypoint’. If the artificial DA (DA+50) is used this may not provide a constant flight path, possibly becoming steeper after DA.

Another reminder is that the European view is biased toward ICAO PANS-OPS where procedures created under these rules probably have different safety margins from those procedures created by TERPS – am I correct in thinking that TERPS DAs are lower thus there is more risk if you descend below DA, i.e. crews should add 50ft margin for TERPS?

A further caution is on the need to consider temperature correction; see information from DoT Canada. Similarly SOPs for altimeter pressure setting should be reviewed. Waypoints are defined using baro alt, thus a baro setting error or FMS error is doubly hazardous as it could affect both the required flight path and the crew cross-check of approach accuracy when using an altitude – range table.

For 411A and other non beliers, here is a selection of the range of human errors which can occur on NPAs and so called night visual approaches – see TAWS Saves. Be aware that many of the ‘mistakes’ could apply equally to VNAV operations.

NPA 41 has some interesting (erroneous) views on workload and dubious justifications for some of the reductions in minima, eg:-
Re NDB (NPAs) “… this type of approach involves a considerable degree of imprecision and pilot work load. If the NDB is combined with a DME, the control of the vertical flight path is improved and for that reason the minima can be lower.”
This is really only effective if the DME is referenced to the runway, better to use RNAV dist, but – see ‘TAWS saves’.

Re proposed wording for SOPs, “To assure the appropriate descent path is flown, the pilot not-flying should announce crossing altitudes as published fixes and other designated points are crossed, giving the appropriate altitude or height for the appropriate range as depicted on the chart”.
Such a procedure is open to the same error as having a mis-set altimeter due to a single source of information, i.e. if the PNF is in error (chart, altimeter, mistake) then so is the PF. Each crew member crew should have their own approach chart to provide an independent cross-check, again see ‘TAWS Saves’ …. “During the descent briefing most crews rigorously check that their chart dates agree, but how many crews cross check if their understandings of the approach agree?”

Re the calculation of RVR (table 6). The visibility along the line of sight is depicted as RVR; this is incorrect, it is slant visual range (SVR). Although small angle theory equates them geometrically, in reality, fog/low visibility does not have consistent visibility with altitude; thus more generally than not, SVR is less than RVR and the crew will not see such a large visual segment.Thus some of the reductions in minima may not be justified.

I hope that these ‘muddles’ are not progressed to the point of applying them universally across all low vis operations including VNAV.

[411A]

OK, lets look at a few possible examples which, oddly enough, have come up in the last few years of personally flying the trimotor.
Several airports that I have operated to in Africa lately require non-precision approaches, and with limited visibility due to rising sand, it is vitally important to get down to MDA reasonably quickly, so as to have the maximum TIME to actually look for the airport, and thereafter have a reasonably stable flight path, to the end of the runway, especially if circling to land, which is sometimes the case.
Dive and drive works very nicely in these cases, IF done properly.
Now, I can quite understand some airlines restricting their pilots to continuous descent approaches, simply because many NPA's are not flown all that often, so the less demanding procedure is more appropriate for these carriers, especially as (generally speaking) handling practise by pilots (hand flying) certainly seems to be on the decrease...autopilot use is up, hand flying skills go out the window.
Now as to the 'dinosaur' issue.
Forget NPA's and look at aircraft handling.
Not all that long ago, a jet transport having just departed from a New York area airport had its vertical fin depart, simply because...the airline in question, and ESPECIALLY those in a training capacity therein, had failed to understand what was learned so very long ago.
Namely, the rudder is absolutely NOT to be used grossly at higher speeds for ANY type of upset scenario.
To do so can be fatal.
Of couse, the airline in question found out the hard way.
Dinosaurs learned long ago, and operate safely accordingly.
Some newer folks?
Perhaps they never learned properly in the first place.
It's called....training deficiency, and some airlines should be ashamed of themselves.

[Carnage Matey!]

I seem to remember that the change in rudder use was the first thing drummed into me when I changed from props to jets, and I am very much new school. I think its extremely tenuous to try to make any connection between CD NPAs and that rudder incident. Its about as relevant as me saying that because Delta lost a tristar to windshear that all 'dive and drive' practitioners are prone to coming a cropper in microbursts. The link just ain't there.

[Capn Bloggs]

Mike,

So, in that case, you are operating to higher minima & planning to maintain level flight for the circling part of the approach. Completely different criteria & operating parameters.

Not so. Our environment sometimes has Cu showers which are very localised. It is quite legal (and safe) to pop out in sky clear weather, at the Straight In Approach minima after doing a CDFA NPA to a runway, and then circle not below 400ft AGL (Cat C) to land on the opposite runway. Provided the procedure is briefed by the PF, there is no drama.

[SIDSTAR]

I can see no reason why we shouldn'y add the 50 ft to all MDAs when flying NPAs. The only possible drawback is that it might occasionally result in a G/A where using the actual MDA might not. In any case while not decrying 411's expertise in his big bus, there is no doubt in my mind that the constant descent angle method is far preferable to the dive and drive. In over 15 years of simulator instruction in everything from basic dinosaur jets to modern FMS a/c, I have seen the quality of NPAs improve out of all recognition. Just think about it. Crews routinely fly ILSs all day and are accustomed to this kind of CDA aproach. The CDA feels natural to them as does executing a G/A at MDA+50 (or MDA if that's you authority's view.) I leave it up to the experts in ICAO, EASA and FAA to decide for me whether or not I need that extra 50ft. Until I'm informed otherwise, I believe it's much better to err on the side of caution. As an aside, to watch a good crew execute a Constant Descent Angle Non Precision Approach from the back of the sim is one of the best examples I can think of in terms of good CRM at work. PF flies the a/c using V/S (or TRK/FPA if he's lucky enough to have it), the PNF calls out the ideal heights/altitudes for each mile along the approach. If no official DME, then FMS GPS distance to the threshold can be used as it's always going to be better than eyeballing it on a timed approach.

[alf5071h]

Originally Posted by Mike Jenvey

I briefed my partner on the CDFA profile & how he should call DME distances versus given altitudes as per the Jepp plate.

Mike, a pedantic, but potentially life saving point is that during the approach, altitude should be the dominant parameter. Always check altitude before (in preference to) range.

Incident number 2 in TAWS Saves is an example of the hazards of allowing range to dominate our thinking.

There is a similar example in a previous version of the paper published by the FSF (European Safety Seminar in Athens). This involved an incident (in Europe) where there were two (combined) altitude–range tables on the same approach chart, one for an ILS/DME the other for an offset VOR/DME (TACAN). The approach may also have been approved for VNAV, I will see if I can provide a link to a copy.
Your use of an altitude–range check is an excellent safety example, but be aware of the potential hazard of single sourcing the information due to the NFP/PM calling the check. The briefing for the approach must provide a mental picture of what is expected – one or two significant check points so the PF can crosscheck the information from the other pilot.

Altitude–range tables are a potential life saver; in this incident - “1200 ft below glideslope”, use of a table would have aided the crew in detecting the error earlier. Although this was an ILS, the error could just as easy apply to a VNAV ‘electronic’ glideslope.