Vertical path non-precision approach safety?

[Bumblebee]

the following is the straw man argument...

The "outcome" is not what determines the safety of the approach. If the outcome were the determining factor, you wouldn't be able to tell if an approach was safe until the aircraft was on the ground. If I fly an unstable approach and land 80% down the runway, but happen to be able to stop and keep it on the pavement, was that a safe approach/landing?

Here you say that if you don't do your job correctly, and throw out your SOP's and happen to keep it on the runway, is it a safe approach...well of course it's not.

Here is my assertion.


let's try to get real here...
which is safer?
If you do fly the approach within SOP and don't have to go missed...
or
A CANPA approach where you couldn't see the runway fast enough, and go missed, then divert or try again?

Now, if you fly the approach to SOP and then have to go missed at the VDP, does it have greater risk to do the missed from that point?

Or is the reason for all of this, (industry wide) a lack of competency? Is it more safe because We're not proficient?


Don't get me wrong, as I've explained multiple times now I fly step downs as a stabilized glide path until the MDA...so none of the power on, retrim, power off, retrim.

I just don't want to have it imposed when it is perfectly safe when performed correctly.
(oh and I won't because I'm not in the 121 environment BTW )

 

[PhilosopherPilot]

the following is the straw man argument...


Here you say that if you don't do your job correctly, and throw out your SOP's and happen to keep it on the runway, is it a safe approach...well of course it's not.

Here is my assertion.


let's try to get real here...
which is safer?
If you do fly the approach within SOP and don't have to go missed...
or
A CANPA approach where you couldn't see the runway fast enough, and go missed, then divert or try again?

Now, if you fly the approach to SOP and then have to go missed at the VDP, does it have greater risk to do the missed from that point?

Or is the reason for all of this, (industry wide) a lack of competency? Is it more safe because We're not proficient?


Don't get me wrong, as I've explained multiple times now I fly step downs as a stabilized glide path until the MDA...so none of the power on, retrim, power off, retrim.

I just don't want to have it imposed when it is perfectly safe when performed correctly.
(oh and I won't because I'm not in the 121 environment BTW )

The FAA Advisory Circular details the rationale behind why the risks are increased with dive and drive. There's a link to the AC in one of the posts above, if you're interested in reading that rationale.

You don't have to buy into it, certainly, but the industry standard way of thinking about safety, particularly from the airline management perspective, is to think of safety as managing risk. If the risks are determined to be higher for dive and drive, then the company has to decide if that risk is acceptable or not. It seems that most 121 carriers have decided to not accept that additional risk.

So the airlines aren't saying that dive and drive is "unsafe," they are merely saying that since the people who research this sort of thing agree that the risks are increased for dive and drive, they would prefer to remove that risk from the operation.

 

[Polar742]

do you get lower mins?

No, the MDA is treated as a DA in most circumstances (when there has been a terrain analysis, such as a runway w/ an ILS). In the rare circumstance, you add 50' to the MDA for your DA ensuring you don't go below the MDA in the case of a GA.

 

[PhilosopherPilot]

As far as the straw man argument goes, I was just pointing out that if you rely solely on the outcome to determine if an approach was safe or not, then there are plenty of things that can go on prior to the landing that increase the risk and make the approach unsafe that may not always result in a burning hunk of metal at the end of the runway.

That particular example I used was an actual approach that we investigated. In the report, the pilots said they landed "safely," meaning they didn't turn the airplane into a burning hunk of metal, but I'm not sure that most on here would have called it a safe landing if you saw the animation.

 

[Bumblebee]

As far as the straw man argument goes, I was just pointing out that if you rely solely on the outcome to determine if an approach was safe or not, then there are plenty of things that can go on prior to the landing that increase the risk and make the approach unsafe that may not always result in a burning hunk of metal at the end of the runway.

That particular example I used was an actual approach that we investigated. In the report, the pilots said they landed "safely." Meaning, they didn't turn the airplane into a burning hunk of metal, but I'm not sure that most on here would have called it a safe landing if you saw the animation.

right, thanks for the clarification...

I appreciate a good examination of things. Thanks for this as well...

The FAA Advisory Circular details the rationale behind why the risks are increased with dive and drive. There's a link to the AC in one of the posts above, if you're interested in reading that rationale.

You don't have to buy into it, certainly, but the industry standard way of thinking about safety, particularly from the airline management perspective, is to think of safety as managing risk. If the risks are determined to be higher for dive and drive, then the company has to decide if that risk is acceptable or not. It seems that most 121 carriers have decided to not accept that additional risk.

So the airlines aren't saying that dive and drive is "unsafe," they are merely saying that since the people who research this sort of thing agree that the risks are increased for dive and drive, they would prefer to remove that risk from the operation.

I agree that the risks are increased with dive and drive...never said they weren't...but when you go missed where you could have landed (when done properly) then you lose that risk mitigation because now you are flying (greater risk) rather than on the ground (less risk).


Mitigating risk is why they hired me...right?

any way thanks again for the thorough discussion...

 

[PhilosopherPilot]

right, thanks for the clarification...

I appreciate a good examination of things.

No prob.

 

[jtrain609]

right, thanks for the clarification...

I appreciate a good examination of things. Thanks for this as well...


I agree that the risks are increased with dive and drive...never said they weren't...but when you go missed where you could have landed (when done properly) then you lose that risk mitigation because now you are flying (greater risk) rather than on the ground (less risk).


Mitigating risk is why they hired me...right?

any way thanks again for the thorough discussion...

I can't disagree with this more. Taking a perfectly functioning aircraft around and going elsewhere is much safer than trying to do a dive and drive.

 

[Bumblebee]

they don't let you jet pilots circle any more either...

 

[PhilosopherPilot]

they don't let you jet pilots circle any more either...

You can get a type without the limitation, but it's not generally allowed in the airline ops specs. Same thing, increased risk.

Why fly around at low altitude in bad visibility with 150 passengers when you can just divert and wait for the weather to improve? (Assuming you can't get in for whatever reason without circling.)

 

[seagull]

Arguing if CANPA is safer than drive and drive is like arguing that landing at a modern airport is safer than an off-field landing. There is no question here, CANPA easily wins, and that debate is inane.

That said, a more important issue here is what was raised in the first place, and that is the following of a non-database vertical path. There are two kinds of FMS flight paths, those that are, literally, built with a database with approaches that are designed that way, and those where the vertical path is not available (like a LOC approach), and the system "builds" its own path. Now, there is a caveat here, that while Jepp does not code the LOCs, generally, LIDO does, so those operating with a LIDO database have a real, surveyed, database approach that can be chosen out of the FMS. Those without choose the ILS out of the database, and the FMS will build a path even though it is designed for an electronic glideslope. I would caution that while the airplane will fly it, you need to be REALLY careful about following the latter type of path. It LOOKS the same from the flight deck, but it is NOT the same!

It would be interesting to know if the UPS as well as the Asiana crew were following such a path. It would also be interesting to look at the surveys and see if the design of the approach procedure was correct in BHM. If I were with IPA, I would contract with Wally Roberts to take a hard look at it. In the Asiana case, it will be interesting to know if the A/T were in the mode where they can be armed but do not provide any sort of speed hold at all.

 

[Bumblebee]

An unnecessary missed approach is less safe than landing....and any argument to the contrary is inane.

Might as well stop doing those dangerous CAT 3 approaches too... I mean, switching the pilots so low to the ground! Crazy!

 

[DE727UPS]

It's also important to note that your FMC generated vertical path does not guarantee terrain protection below the MDA. From the MDA to the runway is a visual maneuver and at night only the VASI or PAPI is guaranteed (and flight tested) to give you terrain clearance.

 

[dasleben]

An unnecessary missed approach is less safe than landing....and any argument to the contrary is inane.

Might as well stop doing those dangerous CAT 3 approaches too... I mean, switching the pilots so low to the ground! Crazy!

I'm on my phone, but when I get home I'll be happy to put together a list of accidents caused by botched non-precision approaches, then a list of accidents caused by missed approaches. Put that argument right to rest.

Cat III approaches are, with limited exception, autoland operations with no control change.

 

[Polar742]

Arguing if CANPA is safer than drive and drive is like arguing that landing at a modern airport is safer than an off-field landing. There is no question here, CANPA easily wins, and that debate is inane.

Thank you

That said, a more important issue here is what was raised in the first place, and that is the following of a non-database vertical path. There are two kinds of FMS flight paths, those that are, literally, built with a database with approaches that are designed that way, and those where the vertical path is not available (like a LOC approach), and the system "builds" its own path. Now, there is a caveat here, that while Jepp does not code the LOCs, generally, LIDO does, so those operating with a LIDO database have a real, surveyed, database approach that can be chosen out of the FMS. Those without choose the ILS out of the database, and the FMS will build a path even though it is designed for an electronic glideslope. I would caution that while the airplane will fly it, you need to be REALLY careful about following the latter type of path. It LOOKS the same from the flight deck, but it is NOT the same!

Interesting. We run purely Jepp databases at our shop. The 748 runs the NG FMC with the greater memory, it has each approach coded and the appropriate glide path for those approaches in the database. The 744's archaic FMC still will run a differentiated glide path if the ILS & LOC g/p is different. I have yet to run across an approach loaded into the database that does not have a geometric GP loaded.

Not sure if that is along the same strain you are discussing or not.

In the Asiana case, it will be interesting to know if the A/T were in the mode where they can be armed but do not provide any sort of speed hold at all.

I agree. With the variety of speed modes and functionality, it could be easily confused if your eyes are outside, hand flying and the other hand on the throttles expecting them to move with any speed deviation from bug.

Thanks for your input to the thread.

 

[Zephyr]

I'm the author of the article referenced in the original post and currently do TERPS for a living. Hopefully I can answer some of your questions about obstacle protection below MDA.

CDFA/CANPA is certainly a good idea. However, the disconnect comes from pilots being conditioned to always follow a glidepath when available. That's obviously appropriate on an ILS/LPV/BaroVNAV when there is a real glidepath, but not when it's a derived glidepath on a non-precision approach. Usually our avionics display the derived glidepath information very similarly or identically to a real glidepath, which only adds to the confusion.

Long story short, if there is a real decision altitude for a procedure, you are protected from the DA to the runway. If there is an MDA, the procedure design does not guarantee protection between the MDA and the runway.

Non-precision approaches can effectively be protected between the MDA and the runway if there is a VGSI (with an angle equal to or less than the approach glidepath angle) or another set of minimums having a DA and the same glidepath angle for that runway. This is the spec that determines whether a derived DA is approved for that runway or not.

On a non-precision approach, if the VGSI is at a higher angle than the approach, the VGSI is protected, but the glidepath is not.

Visual areas are evaluated during procedure development, but are not required to be clear. Penetrations only result in increased visibility minimums or restrictions on the procedure. It should be noted that if the 20:1 visual area is penetrated by unlit obstacles (meaning, something is higher than you are within the visual area), the procedure will be NA at night, HOWEVER a VGSI may be used in lieu of obstruction lighting. This means that if the VGSI is at a higher angle than the approach glidepath and you wind up low on the VGSI, you're not protected.

Between misunderstanding of this topic among pilots and the opportunity for confusion between real and derived glidepaths as depicted on avionics, I think this will lead to an accident someday.

 

[Polar742]

@Zephyr. Excellent read. Other than crawling the TERPS, do you know of a document that discusses G/P and VGSI not coincident obstacle clearance?

Thanks

 

[derg]

An unnecessary missed approach is less safe than landing....and any argument to the contrary is inane.

Might as well stop doing those dangerous CAT 3 approaches too... I mean, switching the pilots so low to the ground! Crazy!

Granted, it depends on the airplane.

A missed approach procedure in a 757/767 is "Go around thrust, Flaps 20" (positive rate) "Gear up" and select a roll mode at 400 and the thing will literally take care of itself to MAA. Smooth, magical and all braindead "Airbussy" in it's execution. The hardest thing about doing a missed approach in a 757/767 is not touching anything and letting the airplane handle it's business.

The CANPA marries in together with automated aircraft very well.

I think the missing point that we're dancing around is that not every airplane in our fleet is CANPA-authorized, nor should they be.

As much as it makes sense to "dive and drive" in a Metroliner or DC-9 is as much as it makes sense to use the "electronic crutches" in a highly-automated aircraft.

Certain fleets and certain operators are authorized to do this and we don't even have the ability to 'dive and drive' in my fleet as it's not authorized and we wouldn't have a leg to stand on when the FAA POI wondered what you were doing trucking around at 400' AGL at final approach speed looking for an airport in a 767-300.

I think two important things that we've missed in this discussion are:

VDP's, which if you're before it, we weren't allowed to descent and if we're after it, we're not allowed to descend at all.

Plus, if we're past our MAP, we can't necessarily conduct a MA procedure, we're actually into the realm of "rejected landings" so you'd better be familiar with TERPS, "Green Pages" (as my airline calls them), two-engine 10-9 procedures or, if you're unlucky, EO 10-9 procedures.

So in our operation, the CANPA is the best method of keeping it simple, using our airplane's automation package the way it was meant and negating all of the legal manuevering involved with before the VDP, after the VDP, and all of the madness that's involved with flying past your missed approach point and trying to figure out if you're supposed to fly a MA procedure or rejected landing procedure.

TERPS is a helluva drug.

 

[BobDDuck]

With the variety of speed modes and functionality, it could be easily confused if your eyes are outside, hand flying and the other hand on the throttles expecting them to move with any speed deviation from bug.

That's been one of the toughest things for me adapting to a a plane with autothrottles. I don't know if it's a Boeing thing or a holdover from Douglas (or maybe even MD) but when you are in IDLE CLAMP as a descent mode and then turn off the autopilot and slow your descent rate (which was based on throttles idle and bugged speed) the engines will not spool up to hold whatever airspeed was bugged unless you tell the throttles to come out of idle clamp. The IOE captain let me get pretty damn slow (still safe though) to remind me never to do that again.

 

[derg]

That's been one of the toughest things for me adapting to a a plane with autothrottles. I don't know if it's a Boeing thing or a holdover from Douglas (or maybe even MD) but when you are in IDLE CLAMP as a descent mode and then turn off the autopilot and slow your descent rate (which was based on throttles idle and bugged speed) the engines will not spool up to hold whatever airspeed was bugged unless you tell the throttles to come out of idle clamp. The IOE captain let me get pretty damn slow (still safe though) to remind me never to do that again.

My biggest point of advice on Douglas equipment like the DC-9 series, of course I've never flown the 717, but if it's anything like the rest of the DC-9 variants, consider the auto throttles as "Santa's Little Helper", but you are Santa.

Follow them though with your hand, watch them like a hawk and if they're not moving when they should, move them where they need to be.

One of my biggest problems when I transitioned to the "80's Modern" 75/76 was keeping my hands off the throttle during climbs, descents and the early parts of the approach phase.

 

[PhilosopherPilot]

Cat III approaches are, with limited exception, autoland operations with no control change.

Depends on the aircraft and operator. On our E190s it's a captain maneuver down to 50' RA mins. There's no control change. The captain hand flies the whole maneuver turning off the autopilot by 800 feet using the HUD.

Forgive me, but I forget if it's Cat IIIa, b or c. It's whichever is not auto land, and down to 50'.