ATC instruction for ILS approach

[Marc Sieffert]

Hi, when flying to EGKK (Gatwick) yesterday I selected the arrival KUNA1 G and the approach ILS 08R. When I got closer to the approach phase, the ATC gave me some vectors to get me close to the ILS intercepting point FF08R. This point is located at 3000 feet (see below). What surprised me is that when I got really close to the intercepting point and did the last turn, ATC gave me the clearance for the ILS approach and asked me to take the Heading to FF08R and to descend to 2000 feet although at this stage I was at 3000 feet (the intercepting altitude) and at around 2-3 miles from the intercepting point. What is the right behavior for the pilot at this stage, maintain 3000 feet until the intercepting point or start to descend to 2000 feet before the intercepting point and if so how fast...

My fear is that if I descend below 3000 feet before the intercepting point the ILS will not be caught...

The same situation happened to me with an approach to Manchester...

 

Thanks for your help!!

[Simon Kelsey]

Hi Marc,

You should descend to 2,000 feet as instructed (obviously you don't have to go screaming down with idle thrust and the speedbrakes out, you can modulate your rate of descent so as to achieve a continuous descent and glidepath capture without levelling off). This is likely why you were desecended - to ensure that you didn't level off before capturing the glidepath.

The minimum radar vectoring altitude in the area is 2,000ft so this is perfectly fine (you can refer to the MRC chart if you want to cross-check any ATC cleared altitudes).

[Marc Sieffert]

37 minutes ago, Simon Kelsey said:

Hi Marc,

You should descend to 2,000 feet as instructed (obviously you don't have to go screaming down with idle thrust and the speedbrakes out, you can modulate your rate of descent so as to achieve a continuous descent and glidepath capture without levelling off). This is likely why you were desecended - to ensure that you didn't level off before capturing the glidepath.

The minimum radar vectoring altitude in the area is 2,000ft so this is perfectly fine (you can refer to the MRC chart if you want to cross-check any ATC cleared altitudes).

OK so you mean that ATC is allowed to give instructions until min 2000 feet. But if I am at 3000 feet already before the intercepting point and ATC asked me to descend to 2000 feet I need to stay at 3000 until the intercepting point and then ILS GS will anyway start the descend... right?

I should not descend further if I am already at 3000 feet before the intercepting point since the intercepting altitude is 3000 feet (see chart below), right?

 

Edited May 14, 2021 at 03:43 PM by Marc Sieffert

[Simon Kelsey]

No, I mean you should be descending towards 2000 ft as you have been cleared and instructed. But that can be a gentle descent that will result in you intercepting the glidepath at some point between 3000ft and 2000ft.

[Alistair Thomson]

As Simon says, you are safely cleared by ATC to the minimum radar vectoring altitude of 2000 ft, and you can trust that instruction. Setting a gentle descent rate is good, and it will get you on the glideslope before you get to 2000ft, whereupon you'd increase your descent rate etc. to suit. If your descent rate was high, you might reach 2000ft first and would need to level off.

However, ignoring your passengers' comfort, you could dive immediately to 2000ft when cleared, then proceed level until G/S intercept, but that's untidy flying! :)

So, the technical question then is, at what descent rate will you reach the glideslope at the same time as you reach 2000ft? And does anyone care? :)

[Kirk Christie]

7 hours ago, Marc Sieffert said:

My fear is that if I descend below 3000 feet before the intercepting point the ILS will not be caught...

This will not be a problem, the preference for most is to capture the Glide slope from below it. If you are below the glide slope, the Diamonds will come down as you meet up with the correct path, once the diamond is at the middle the glide slope will be captured, and you can descend on it.

[Alistair Thomson]

13 minutes ago, Kirk Christie said:

the preference for most is to capture the Glide slope from below it

:) Not just a preference! False glideslopes exist above the main one at higher angle of descent (multiples of 3 degrees, IIRC) and to avoid these, it's necessary to approach from below.

[Alistair Thomson]

On 5/14/2021 at 12:27 PM, Alistair Thomson said:

So, the technical question then is, at what descent rate will you reach the glideslope at the same time as you reach 2000ft? And does anyone care? :)

OK, so I had a go at answering my own question. (Yes, I need to get a life ;)

There doesn't seem to be any quick way of getting the answer, which seems to be 485 ft/min at a groundspeed of 160kts. Approach planners provide GS vs ROD tables for a reason!

The attached diagram shows the situation, with the optimum path having a slope of 182 ft/nm, and that becomes 485 ft/min at 160kts groundspeed.

I'll have an untroubled sleep tonight knowing that information...

I assume ATC were trying to help, but it looks like it would have been easier for you to simply have continued at 3000 ft until the glideslope was encountered!

 

[Tobias Dammers]

Capturing from above is somewhat problematic for another reason.

If you capture from below, then you will be in level flight, or in a stable shallow descent; capturing the glideslope, then, will have you nose down, which will temporarily increase your airspeed, and the autothrottle then reacts by gently retarding to bring you back to your selected approach speed. By contrast, if you capture from above, you will also be in a stable descent, but at a fairly steep angle, so probably at a low power setting (if not idle), and when you capture the glideslope, the autopilot will pull up. Your airspeed temporarily decreases, and the autothrottle will counter this by advancing - but it takes a while for this to happen, and then a bit more for the fans to spin up, and during that time, you are flying slower than you should be. Now if you remember that airspeed is life insurance, it should be immediately obvious that capturing from below is much safer.

And another reason: you're in a critical, fairly high-workload flight phase, during which a lot of things can go wrong. Now consider what happens when, for whichever reason (technical, human, doesn't matter), you fail to capture the glideslope. If you're capturing from below, then the default response, doing nothing, results in a relatively safe level overflight of the airfield along the extended runway centerline. Not ideal, and you probably want to tell ATC about it ASAP, but you have plenty of time to live and figure out how to deal with it. If you're capturing from above, then you are in a steep descent that intersects this frightful thing known as "the ground", and the do-nothing default will get you down a bit sooner than you had hoped. Say you're descending 2000 fpm, and you miss your 2000 ft AGL glideslope capture - you now have 1 minute (less if the terrain isn't flat and free of obstacles) to realize that something went wrong, and correctly deal with the problem. Both can be done of course, but I know what I'd prefer.

[Simon Kelsey]

3 hours ago, Alistair Thomson said:

I assume ATC were trying to help, but it looks like it would have been easier for you to simply have continued at 3000 ft until the glideslope was encountered!

Good effort with the calculation! 🙂

The problem however, occurs if, for instance, the vector provided was likely to result in localiser intercept inside D8.6 (which may be a slight error, or may be intentional for spacing etc). In this situation, as we shouldn't be descending on the glidepath before we are established on the localiser, we end up in an awkward situation where we perhaps slightly miss the GS and then end up having to intercept from above - in such a situation there are four possible outcomes:

1) A proactive pilot will notice this developing and request a lower altitude
2) A proactive controller will be aware of the potential complication and issue a lower altitude to avoid the situation arising (as I suspect was the case here)
3) Neither of the above occurs but the pilot then recognises the problem when establishing on the localiser and executes a glideslope capture from above (not ideal, but possible and there are approved ways of doing this safely)
4) A go around because the situation was not managed in advance and any mitigation attempt (option (3)) has failed/is not executed properly.

There are certain places (Glasgow is one) where it is quite common in real life (partly as a result of the surrounding terrain and associated minimum vectoring altitudes etc) to be vectored in such a way that an intercept from above is required. The procedure for doing so will vary between aircraft but it is worth becoming familiar -- for the A320 the steps are:

1) Ensure you have LOC green (i.e. active/captured) and (very, very important) GS blue (armed);
2) Push for managed speed (this will help reduce the chance of an overspeed)
3) Select the FCU altitude above the current altitude (because we don't want to get in to an ALT capture situation which causes us to level off before we have intercepted the glide)
4) Pull V/S and select a suitable rate of descent (about -1500fpm is a good starting point)
5) Monitor the descent and ensure that the glideslope is captured correctly
6) When GS mode becomes active set the missed approach altitude

All the above works best in Conf 2/Gear Down to avoid the speed running away. Obviously it is very important to a) cross-check the rate of descent and altitude vs a DME fix on the chart to ensure that the correct glideslope has been captured and not a false lobe, as Alistair rightly highlights as a potential threat above; also, ideally we will have set and briefed a bottom line for throwing it away (i.e. the stable approach 'gate' of fully configured and on speed/glidepath etc by the 1000 ft callout would be a good position!).

Done correctly and with the appropriate mitigation strategies in place it is a perfectly safe procedure; but Tobias is exactly right to highlight the potential threats if the procedure is not correctly executed and/or monitored (see https://avherald.com/h?article=4ae84b8a for an example of what can happen if not executed correctly - in this instance, doing so before the localiser was captured).

Edited May 15, 2021 at 11:53 PM by Simon Kelsey

[Alistair Thomson]

1 hour ago, Simon Kelsey said:

Good effort with the calculation! 🙂

The problem however, occurs if, for instance, the vector provided was likely to result in localiser intercept inside D8.6 (which may be a slight error, or may be intentional for spacing etc). In this situation, as we shouldn't be descending on the glidepath before we are established on the localiser, we end up in an awkward situation where we perhaps slightly miss the GS and then end up having to intercept from above - in such a situation there are four possible outcomes:

1) A proactive pilot will notice this developing and request a lower altitude
2) A proactive controller will be aware of the potential complication and issue a lower altitude to avoid the situation arising (as I suspect was the case here)
3) Neither of the above occurs but the pilot then recognises the problem when establishing on the localiser and executes a glideslope capture from above (not ideal, but possible and there are approved ways of doing this safely)
4) A go around because the situation was not managed in advance and any mitigation attempt (option (3)) has failed/is not executed properly.

There are certain places (Glasgow is one) where it is quite common in real life (partly as a result of the surrounding terrain and associated minimum vectoring altitudes etc) to be vectored in such a way that an intercept from above is required. The procedure for doing so will vary between aircraft but it is worth becoming familiar -- for the A320 the steps are:

1) Ensure you have LOC green (i.e. active/captured) and (very, very important) GS blue (armed);
2) Push for managed speed (this will help reduce the chance of an overspeed)
3) Select the FCU altitude above the current altitude (because we don't want to get in to an ALT capture situation which causes us to level off before we have intercepted the glide)
4) Pull V/S and select a suitable rate of descent (about -1500fpm is a good starting point)
5) Monitor the descent and ensure that the glideslope is captured correctly
6) When GS mode becomes active set the missed approach altitude

All the above works best in Conf 2/Gear Down to avoid the speed running away. Obviously it is very important to a) cross-check the rate of descent and altitude vs a DME fix on the chart to ensure that the correct glideslope has been captured and not a false lobe, as Alistair rightly highlights as a potential threat above; also, ideally we will have set and briefed a bottom line for throwing it away (i.e. the stable approach 'gate' of fully configured and on speed/glidepath etc by the 1000 ft callout would be a good position!).

Done correctly and with the appropriate mitigation strategies in place it is a perfectly safe procedure; but Tobias is exactly right to highlight the potential threats if the procedure is not correctly executed and/or monitored (see https://avherald.com/h?article=4ae84b8a for an example of what can happen if not executed correctly - in this instance, doing so before the localiser was captured).

Outstanding analysis! I made my assumptions based on the OP's declaration that he was 2 - 3 nm from I-GG on the localiser and tracking 078 plus or minus pilot error when instructed to descend to 2000 ft. But of course who can tell if that was entirely correct, if not ATC (your outcome 2 above)? So that descent instruction was probably an intervention by a proactive ATC.

You have explained exactly how a clued-in pilot could recover from that situation. That recovery is based on knowledge and understanding from the pilot's perspective, and includes awareness of the true glideslope - but we're not talking about being 3 degrees high at 3000 ft (hopefully)!

It would be nice to know that VATSIM pilots could be as clued-in as that ATCO, whoever he/she was.

[Tobias Dammers]

7 hours ago, Alistair Thomson said:

we're not talking about being 3 degrees high at 3000 ft (hopefully)!

Indeed - being 3 degrees high at 3000 means you'd be on a 6° angle from the runway, which means that you're at a point where you should be down to 1496 ft on the correct glideslope. At this point, you're 4.7 miles out, which means you have passed the outer marker a good while ago, and that alone should have been a red flag. If the visibility isn't lousy, you should also have the field in sight by now, and it should look weird, because you're looking at it from way too high.

[Alistair Thomson]

3 hours ago, Tobias Dammers said:

you should also have the field in sight by now, and it should look weird, because you're looking at it from way too high

Yep. Probably not recoverable in a jet, and a light single VFR might also have real problems. A good landing comes from a good approach, and that position ought to be bad enough to declare a go-around. But I've never been in that position in a jet, so I defer to superior expertise on that :)

13 hours ago, Simon Kelsey said:

Good effort with the calculation!

I finally remembered my trusty Airtour CRP-5, so I whipped it out of my tatty old flight bag and turned it over to the circular slide rule, which verified my calculation. Even with the CRP-5 I think it would have been difficult to derive, then set up, the required ROD before hitting the glideslope at 3000 ft even at 160 kts groundspeed.

So you're correct - Marc should start a descent towards 2000 ft when instructed, at basically any comfortable rate since the optimum is unobtainable in the time available, then re-adapt when either 2000 ft or the glideslope is reached, whichever comes first.