Non-Precision Approaches. What does your airline recommend?

[downfourgreen]

Well folks, one more crash and many fatalities during a non-precision approach and nobody knows when authorities will review that outdated procedure to set up a new safer standard in order to avoid future incidents and accidents during the 21st century .
Several airlines still recommend to fly exactly the approach chart profile view and its steps, ROD, etc. and everybody knows how hard and dangerous those techniques are for a low speed single engine airplane and I don’t need to tell you how demanding task is to align an MD-11 or any wide-body airplane just a few miles on final during a rainy and windy night at weather minimuns.
I know that some airlines have strict standard procedures for non-precision arrivals with many recommendation and rules to follow based on profile and constant-rate descents after FAF to the MDA or TCH, runway alignments, etc.
Can someone give me a feedback about those operation criteria, including training, wind and rate of descent limits, RNP, etc., beside the Jepp’s charts with a VNAV profile and VDPs?
Thank you fellows...
DN4GR

"While my guitar gently weeps..." G.Harrison

[ 27 November 2001: Message edited by: downfourgreen ]

[Hand Solo]

Don't know how Jepp charts depict NPAs but in my compnay the objective is to fly a constant 3 degree approach path (or the relevant slope). This generally means not levelling at intermediate step altitudes and possibly delaying the final descent beyond the initial approach fix in order to achieve a continous smooth descent through the FAF down to MDA. Training for NPAs is covered in the initial conversion course and recurrent sims, wind limititations are either the aircrafts limits or the (lower) limits of the Captains nerves! Descent rates for all approaches are restricted to less than 3000fpm when descending below MSA+3000, with type specific restrictions on autopilot descent modes to prevent high rates of descent & long engine spool up times close to the ground. Aircraft should be on the stable on the final approach profile by 1000 Radio or a go-around should be considered. If the aircraft is not stable on the profile by 500 radio a go-around is mandatory. On the Airbus we fly the final approach in track/flight path angle mode with the autopilot engaged in bad weather. The 'non-landing' pilot flies the approach with the 'landing pilot' calling out distance to go and height above or below the profile, plus any other pertinent information. The approach is terminated at the MDA if visual contact is not made. It is not permissilbe to fly level at MDA until the Missed Apporach Point is reached.

[411A]

Dive (within reason) and drive...has worked for me for over thirty years in several heavy jets.
The constant descent idea is OK...but sometimes it just does not work.
Some companies just do NOT get the message.
Training is the KEY.

[shakespeare]

411A.

I could not disagree with you more strongly. One sure fire method of upsetting a assymmetric non-precision approach is to use the "step down method". With every thrust change as the a/c levels off there has to be a corresponding input of rudder and then another pitch down to the next step = more power changes and rudder inputs and so it goes on. I have never seen an unstable approach flying a constant rate descent. Auto-pilot in or out!

JAR OPS requires this method to be taught in recurrent sim sessions, and if performed correctly, it is in my opinion by far the safest and most efficient way to fly a non-precision approach.

Irrespective of whether it has a 3 degree or 4 degree profile, VOR/DME, NDB/DME, constant rate works every time. Interestingly enough, if you take away the DME forcing the lads to use the tear drop procedure from overhead the navaid, they don't need to use step down at all. If it works here it should work on a straight in approach, and it does!

The trouble in many airlines is that pilots have not recieved thorough non-precision approach training, and as a result don't have a good understanding of flying the procedures correctly.

As you said, "TRAINING IS THE KEY".

[ 29 November 2001: Message edited by: shakespeare ]

[sabenapilot]

On aircraft with a FPA (Flight Path Angle) on the autopilot, European JAR-OPS requires that it is used iso the usual V/S mode.

This means you simply track the radial inbound and set the published FPA (for instance 3.7%). The plane will then descent on a virtual glideslope towards the runway once its horizontal path intersects with the virtual glidslope.
During the approach you'll see the autopilot will modulate the V/S in function of the (normally deceasing) GS and you'll skim the tops (if any) of the intermediate stepdowns.

When apporaching the MDA, no level flight till the MAP is made; instead the MDA must be considered as a DA and a go-around must be executed! This poses no big problem (for instance a too early immediate left turn), because of the design of the FPA mode, which will make sure the MDA will always be reached very shortly before (0.2NM at the max) or on the MAP itself (at the very latest).

Also note that in this FPA technique no VDP (Visual Descent Point) is to be used by the crew, but rather do they use the normal call outs and procedures as for a precision approach.

Quite clever from Airbus, isn't it?
BTW, does anybody know if you can do this FPA technique on the B737NG or B777 as well?

[moggie]

I have to agree with all of you - except 411A.

What do you fly, 411A, and who for? I for one would hate to be on it with your "dive and drive" approach. The bigger the aeroplane the less safe that style becomes - large adjustments to ROD near the ground are asking for trouble.

Hand Solo's SOP sounds just like the one I spend most of my time teaching pilots for a very big airline and I can tell you that the secret to a safe NPA is a stabilised descent rate and small adjustments - especially if dealing with aeroplanes at the lower-tech end of the market. It's all about situational awareness and mutual monitoring on NPA's and stepped approaches push up the workload, reducing SA and the ability of the Non-Handling Pilot to monitor what is going on.

Yes, a pilot should be able to handle a stepped approach, by why make life difficult.
They always say that the superior pilot uses his superior judgement to avoid getting into positions that require the use of his superior skill.

In addition, the airlines that I train for apply a factor to the MDA to turn it into a DA so that the same decision making and handling methods are used for precision and non-precision approaches. This formalises the procedure that you would have to go through anyway and does not reduce the likelihood of getting in.

[downfourgreen]

A nice debate.
In accordance with FAR and ICAO PANS OPS criteria, an airplane is not authorized to fly below MDA until reaching the required visual reference or begin a missed approach prior to the published missed approach point.
Therefore, based on your opinions I have some doubts and would like some help:
Almost everybody agrees that a constant-rate descent is safer and comfortable than a step down approach, but I’m little confused when, for example, an airplane has to reach a significant high MDA (e.g. 600ft AGL). During a profile descent, the pilots are waiting for visual cues to complete their approach, if the crew complies with JAR-OPS I understood that they should reach MDA at point where the airplane could complete a stabilized approach (3 degree, 800 fpm, at least), and in case of no visual contact or not stabilized they must start a go-around promptly, without stopping at MDA. However, if the MAP is over the VOR, some miles ahead, how and where do the pilots begin the missed approach maneuver? Don’t they need to keep the airplane at MDA until MAP (FAR and ICAO rules)?
How do you define the initial point to start a final descent based on a 3-degree profile angle until MDA when no information is provided on approach chart? Is it only an FMC calculation?
Thanks in advance

[FlapsOne]

In my company (737)- notional 3 degree (adjusted if necessary to make or not below step heights) flown in VS. MDA +50' is then used as Company Decision Alt. If no vis ref at Co DA then go around as if on precision approach.

We do not fly a 737 level at MDA to MAP. The thought of doing that with one donk out is horrifying!

[TechFly]

Tks to everybody for apporting their pilot techniques!

I think Moggie was giving two keywords: Situational Aw. & Monitoring.

As we all know sometimes it is possible to reach the same objective with different ways.

In this case I prefer to adhere for a constant descent (as close as possible to the std 3° angle) trying to figure out the VDP with reference to "Dist. from RWY" vs. "cloud base".

PNF will give the deviations from the published std profile (in terms of ft).

For what is concerning the MAPt I look at it in particular for the separation from obstacle in case of a GA with an engine failure.

Looking for your comments.

Cheers.

Fly safe & enjoy life.

[411A]

FlapsOne, better stick to the UK. If you came to the USA and could NOT fly level at MDA (and not the +50 nonsense)with an engine out, you would FAIL your FAA type ride.
Its called airmanship...and handling ability...and if you ain't got it, don't apply.

[Softons_Mum]

Using the continuous 3 degree descent procedure, u treat the MDA as a DA - if the required visual reference is not there you go around. In terms of if the missed approach point if it is some distance ahead, you continue the tracking as required to reach the missed approach point whilst climbing away, then follow the standard missed approach when reaching the specified point...or as directed by ATC ....

[Softons_Mum]

in terms of 411as method of flying level - it may well be practical (or required - some procedures require a visual level segment after reaching the missed approach fix) - however in many cases it is not practical - why fly level along the MDA to the missed approach point if you know that you cant descend safely to the runway with appropriate rates of descent and speeds when you reach that point? in most companys i would say that it would be viewed very poorly doing greater than 1000fpm below 1000 feet.......

[LevelFive]

DN4GR,
For non-precision approaches – ceiling less than 1,000feet or visibility less than 3 statute miles, the first officer must fly using the autopilot while the captain seeks visual reference with the runway, or the captain can manually fly using the HGS system. FLAPS must be 40 by the final approach fix. Max descent rate 1,200 fpm. Minimum descent rate 1,000 fpm. The captain does the landing. Sim check is get down, drive in and descend as soon as runway is in sight. Wind limits – 70 kts peak gust headwind, max crosswind 35 kts, max tailwind 10/15 kts depending on model.

[2daddies]

411A,
Mate, I promise you that I can fly at MDA + 50 - 0 (and don't give me any of that "I can maintain an altitude to within 10 feet" rubbish - you bloody well CANNOT and you know it!) with one-engine inoperative and without falling into a slanging match about our respective country's standards I'm sure you'll find that if I can pass an Australian Multi-Engine Command Instrument Rating test AND a base-check with my airline I can pass an FAA equivalent - it's what I get paid to do.

My point though is: why dive and drive if you don't have to? It's can be uncomfortable for passengers (descent rates/markedly changing engine note) , it makes speed, energy and configuration management difficult, it places unnecessary power demands on the engines and quite frankly the closer I get to the ground the more reluctant I am to dive towards it - distractions and mistakes happen and if you make an error in a nice constant 3 degree descent chances are you'll have A LOT more time to fix it than if you're driving down to the next step at 3000 fpm (not to mention avoiding triggering the GPWS).

Airmpanship encompasses many elements of flying - but basically don't do something if there's an easier and SAFER way of getting the job done!

[ 30 November 2001: Message edited by: 2daddies ]

[ 30 November 2001: Message edited by: 2daddies ]

[411A]

You missed the point entirely 2daddies.... the FAA wants to see a NP approach flown to the missed approach point, NOT a miss upon reaching MDA. If you cannot do this, level at MDA, with an engine out, it's pink slip time, simple as that. It's called maximum time at MDA to be able to see the runway or approach lights. Diving at 3000 ft/min is certainly not required, 800-1000 generally works fine. If an airline wants to impose additional restrictions, that is up to them, I mentioned the FAA requirement for the rating. Does not necessarly apply to everyday plain vanilla line ops. Those of us trained long ago use the dive&drive method because it works, for us anyway. For the more "junior" guys, new to line ops, the constant descent method may well be better suited. Different strokes for different folks

[CaptA320]

411A I agree that in the old days, in fact not so long ago at all the method of flying a NPA was the dive and drive, most of us were trained in this method throughout our initial training. However this has changed as the constant rate of descent procedure has been found to be both safer and more comfortable for pax. When I did my intitial pilot training there was no such thing as CRM...get my point.
The issue here is not if a pilot has the ability to fly MDA+50, despite the fact that most of us had to display that in order to get our ATP tickets, line flying has little to do with an FAA check ride. Major airlines around the world have recognized that the "stablized" approach not only increases situational awareness but also is safer.
The constant rate of descent method is no so widely recognized that Jepps decided to publish on their charts the dme crossing altitudes for the portion beyond the FAF.
The way we train pilots is two fold.
If the chart has the published dme/altitude profile then it is recommended to be used in conjunction with the published gradient which should be converted to FPA.
If the chart does not have published dme altitudes then the crew should calculate the profile of descent by first looking at the VOR (lets say), calculating its displacement from threshold (when on threshlod what should my dme read) then knowing we want to cross the threshold at 50 feet work back to platform altitude based of the gradient.
So if the VOR is 1 mile before the threshold and threshold elevation is 60 and platform altitude for the approach is 2000ft with a gradient of 5 it would be as follows.

At threshold (-1DME) we should be at 110ft at the VOR we should be at 410ft, at 1DME 710ft, at 2DME 1010ft, at 3DME 1310, at 4DME 1610ft, at 5DME 1910, at 5.3DME 2000ft.
Now this profile must be compared with the chart to ensure it complies with any DME/Altitude restrictions, it also complies with the FAF (FAF cannot be less that 5.3DME) and at MDA the pilot must have obviously visual contact to continue.

[sabenapilot]

411A,
you say: the FAA wants to see a NPA flown to the missed approach point, NOT a miss upon reaching MDA.
In other words: If you reach the MDA before your MAP, you have to fly level at MDA till the MAP before starting the missed approach.


First of all:
purely technically speaking (based on European JAR-OPS); after descending to the MDA on a NPA, you do NOT have to stay level at MDA till the MAP! An MDA is nothing more then a MINIMUM, so there is no rule wharsoever preventing you from already starting to climb up again. However, you are correct on that you have to stay on the appraoch track till the MAP. Only then are you allowed to initiate any turn (if the missed approach calls for that of course...)

Second remark:
I do not know what plane you fly, but it clearly is no airbus with FPA (Flight Path Angle), because then you'd know this discussion of yours is purely theoretical...
FPA allows you to fly a continuous descent NPA, skimming all tops of any intermediate stepdowns (when used correctly), and reaching your MDA at any selected point!
If you use your MAP for this selected point, then your MDA is nothing more then a virtual DH and a go around MUST be flown upon reaching the altitude.
If you work out and use a VDP however, then you may indeed continue your approach by flying level to the MAP, but as the subsequent descent from that MDA at MAP is no longer on your smooth and continuous descent profile, you will easily end up with all kinds of unpleasant audio warnings like: SINK RATE, TOO LOW... TERRAIN etc.. Not really something I'd like to hear on a NPA appoach in poor visibility (only RWY lights in sight, no RWY itself for instance...)
BTW- If Airbus stongly advises against any level flight at MDA on a NPA approach for all these reasons, then why try to be sharper then these highly qualified guys?

[411A]

Well sabenapilot, the level flight required at MDA is a REQUIREMENT for the FAA rating, what the airline ops specs says about this (in normal line flying) is quite another. If the individual airline wants to treat an MDA as a DA, that is their option. Seems quite unnecessary to me. And, so is the +50' business added to MDA...more European nonsense. Either you can fly accurately, or you cannot, rather simple really...

[RatherBeFlying]

In these wonderful days of FMCs, you can project a straight slope back from the screen height at the threshold to stay above all the step-down fixes and convert the MDA to a DH -- just make sure you key it in right and monitor those step-down fixes.

The first instrument approaches were non-precision approaches. NDBs were new fangled technology and the criteria were designed around the DC-3 -- and still are except that the circling minima were given widened bands as approach speeds increased.

On my very first IFR flight test, as the stopwatch ran out the time on an NDB approach, I lifted my hood at the MAP to find the runway half a mile off to my left after what looked to me like a good tracking of the course -- and passed! I'd really want to know very well the guy who swung the compass before doing an NDB for real -- today you take along a GPS.

[411A]

Yes, and those GPS approaches can be very accurate...why is it that this very useful aid has not found wider acceptance in the EU?

[ 01 December 2001: Message edited by: 411A ]