Cruise speed vs top of descent

[Atco EBBR]

Question for the pilots on the forum:

Say you're cruising at FL240 at a IAS of 290-310 kts. If you're asked by ATC to reduce to 250kts (hypothetically ), how does this affect your top of descent, if at all?

[Sabena320]

If we have to reduce and maintain this reduced speed during the complete descent the top of descent will come forward. We'll have to start the descent earlier as we cannot descend as steeply as when we still had 300kts. Of course, this is when you follow an optimum glide with engines in idle. You can always use the speed brakes to increase the rate again, but this breaks your efficiency

[Atco EBBR]

Thanks for the reply, I knew that a lower speed implied a lower rate of descent, but I didn't know that that was not compensated by the lower speed...

So basically, if I want to sequence using speed control instead of vectoring, I need to first descend with a good rate and then reduce the speed.

[Sabena320]

So basically, if I want to sequence using speed control instead of vectoring, I need to first descend with a good rate and then reduce the speed.

If aircraft have to reduce their speed below a speed that is 'normal' for descent it would indeed be better to let them descend a bit earlier so that they are below the altitude profile and then they can reduce the speed and is the less steeper descent more manageable. Of course the most efficient (excluding operational or time costs, but fuel wise) is to reduce the speed already just prior descent and have a continious descend to avoid a level segment at lower altitude. But of course in busy airspace or airports this is most of the time not feasible I assume

The most difficult is when they keep you high during vectoring, and then ask you to reduce to 210-220kts and descend at the same time. This is very difficult to manage as the speed brakes have very few effect with such a low speed. In that case you have to give priority to speed or to altitude.

[jan_olieslagers]

This makes for most interesting reading, thanks for sharing!

Myself limited to FL120 or so for lack of pressurisation/oxygen bottles, and to 100 kts Vne for lack of budget cannot really contribute to the facts, so sorry...

[737MAX]

Thanks for the reply, I knew that a lower speed implied a lower rate of descent, but I didn't know that that was not compensated by the lower speed...

So basically, if I want to sequence using speed control instead of vectoring, I need to first descend with a good rate and then reduce the speed.

Yep that's the way we like it!

That's until cost index n°6 airline arrives in the way with their 245kts as from ToD though...
Most airlines (unless the pilots are paid by the hour ) use a descent speed of about 280kts.

[737MAX]

This makes for most interesting reading, thanks for sharing!

Myself limited to FL120 or so for lack of pressurisation/oxygen bottles, and to 100 kts Vne for lack of budget cannot really contribute to the facts, so sorry...

You could! It's the same principle with your aircraft

The more you "deviate" from the best glide speed, the more drag you'll have -> higher RoD.
When pilots ask for high speed during descent, it's mostly to keep a high RoD (not to win time...).

[jan_olieslagers]

thank you!

If I wish to descend rapidly - which is standard practice under certain conditions - a sideslip is the standard answer. Which will indeed increase drag dramatically, while at the same time reducing wing efficiency, thus reducing lift. And I absolutely love side-slipping! But I can imagine it is one of those things one doesn't do to the unsuspecting innocent passengers of an airliner...

Reading over your words, I think I understand that, the bigger the plane, the finer the inputs should be; both on the controls and on the throttle(s). Many SEP pilots seem to treat their throttle like a kind of on/off switch, except in cruise...

[737MAX]

thank you!

If I wish to descend rapidly - which is standard practice under certain conditions - a sideslip is the standard answer. Which will indeed increase drag dramatically, while at the same time reducing wing efficiency, thus reducing lift. And I absolutely love side-slipping! But I can imagine it is one of those things one doesn't do to the unsuspecting innocent passengers of an airliner...

Hmmm... always dreamed to do that on short final too... but nope, not allowed

However, if you're still too high and/or too fast close to the airport, increasing speed doesn't make sense anymore (of course). It's time to slow down and increase drag with the landing gear & flaps.

To give an idea; the rate of descent at minimum clean speed (approx 210kts on a 737NG) is around 1000ft; increasing to easily 3000ft+/min with 300kts.

Descent planning (or energy management) is one of the most difficult thing to learn for ab-initio F/O's. It can be tricky as a descent is never the same twice; it depends on traffic, wind, speed, distance to be flown etc... On top of that, it also needs to be performed in an efficient way; that's where you can save fuel.

[Atco EBBR]

Of course the most efficient (excluding operational or time costs, but fuel wise) is to reduce the speed already just prior descent and have a continious descend to avoid a level segment at lower altitude. But of course in busy airspace or airports this is most of the time not feasible I assume

Yes indeed, top of descent is in most cases much earlier than Brussels Acc. It will require very performant, interconnected ATC systems to do a pre-sequencing using speed control starting at cruise level. Something to aim for

The most difficult is when they keep you high during vectoring, and then ask you to reduce to 210-220kts and descend at the same time. This is very difficult to manage as the speed brakes have very few effect with such a low speed. In that case you have to give priority to speed or to altitude.

If you encounter such a situation, you can be sure that the controller has screwed him/herself . If you have a lot of traffic that needs to be vectored for a sequence, you need to get them down and speed reduced asap. If you reduce while they're still high on profile, they'll stay too high too long, and any benefit from the speed reduction will be lost (with the same airspeed (IAS), groundspeed will increase by +/- 6-10 kts per 1000 feet)

That's until cost index n°6 airline arrives in the way with their 245kts as from ToD though...
Most airlines (unless the pilots are paid by the hour ) use a descent speed of about 280kts.

No idea who you're talking about But rest assured, these 245kts airliners very often get slapped with a speed of 300kts on first contact . The pilots are usually very happy with that, they often ask spontanuously if there is any speed control... If I can, my reply is 'negative, free speed', which would make all other pilots happy...

jan_olieslagers wrote: ↑Yesterday, 17:53
:If I wish to descend rapidly - which is standard practice under certain conditions - a sideslip is the standard answer. Which will indeed increase drag dramatically, while at the same time reducing wing efficiency, thus reducing lift. And I absolutely love side-slipping! But I can imagine it is one of those things one doesn't do to the unsuspecting innocent passengers of an airliner...

Hmmm... always dreamed to do that on short final too... but nope, not allowed

Unless you're in the Gimli glider, that is

However, if you're still too high and/or too fast close to the airport, increasing speed doesn't make sense anymore (of course). It's time to slow down and increase drag with the landing gear & flaps

Funny thing is, I recently had a talk with a colleague from approach who had a plane (E190 iirc) at FL80 15nm final. I was surprised to hear that in order to catch the glide they ask to reduce instead of increase. That's not a thing that would happen in an ACC environment What would be a typical maximum speed to extent the gear?

Descent planning (or energy management) is one of the most difficult thing to learn for ab-initio F/O's. It can be tricky as a descent is never the same twice; it depends on traffic, wind, speed, distance to be flown etc... On top of that, it also needs to be performed in an efficient way; that's where you can save fuel.

Probably for the same reason, sequencing arrivals is on of the most difficult things to learn for trainee controllers. You really need to grow experience on what speeds and levels are 'right' and what is too fast/slow, too high/low...

[HQ_BRU_Lover]

thank you!

If I wish to descend rapidly - which is standard practice under certain conditions - a sideslip is the standard answer. Which will indeed increase drag dramatically, while at the same time reducing wing efficiency, thus reducing lift. And I absolutely love side-slipping! But I can imagine it is one of those things one doesn't do to the unsuspecting innocent passengers of an airliner...

Hmmm... always dreamed to do that on short final too... but nope, not allowed

However, if you're still too high and/or too fast close to the airport, increasing speed doesn't make sense anymore (of course). It's time to slow down and increase drag with the landing gear & flaps.

To give an idea; the rate of descent at minimum clean speed (approx 210kts on a 737NG) is around 1000ft; increasing to easily 3000ft+/min with 300kts.

Descent planning (or energy management) is one of the most difficult thing to learn for ab-initio F/O's. It can be tricky as a descent is never the same twice; it depends on traffic, wind, speed, distance to be flown etc... On top of that, it also needs to be performed in an efficient way; that's where you can save fuel.

Very interesting read guys, let's continue this way.

@737MAX: can you give an example of descent planning, f.e. ARVOL STAR 25L in EBBR? Or giving an example is impossible as winds, traffic, weight are not known?

[737MAX]

Yes indeed, top of descent is in most cases much earlier than Brussels Acc. It will require very performant, interconnected ATC systems to do a pre-sequencing using speed control starting at cruise level. Something to aim for

Yep... isn't that called the "single European sky" project?
Our politicians are probably too busy with "how to tax aviation" rather than "how to save fuel by working together and optimizing airspace regulations"...

No idea who you're talking about But rest assured, these 245kts airliners very often get slapped with a speed of 300kts on first contact . The pilots are usually very happy with that, they often ask spontanuously if there is any speed control... If I can, my reply is 'negative, free speed', which would make all other pilots happy...

In Belgium most of the time indeed. Free speed for most belgian airlines mean 280kts or more, but for some (Cost Index 6 is the airline where you have to pay for your hand luggage, you know...) the standard descent speed is +/- 245kts and has to be strictly respected by the company unless they become too high on descent profile and increase speed to catch up. Cost index, for those who don't know, is a relation between the aircraft cost per hour and the fuel cost, basically. The lower the index is, the lower the speed will be and vice versa. This index is therefore different for every airline.

Funny thing is, I recently had a talk with a colleague from approach who had a plane (E190 iirc) at FL80 15nm final. I was surprised to hear that in order to catch the glide they ask to reduce instead of increase. That's not a thing that would happen in an ACC environment What would be a typical maximum speed to extent the gear?

On the 737, the maximum gear extension speed is 270kts (you can increase further to maximum speed once the gear is extended). However an extension at 270kts is very noisy and thus not comfortable for our pax. That's an option we use when there is no other choice. But indeed, when you're so close to the glide you have to reduce speed and increase drag with flaps & gear, otherwise you'll never make it.

Also, some airlines allow visual approaches (which are excellent for fuel & time saving), some don't. That E190 was very likely one of those airlines where pilots perform visual approaches when traffic/wx permits. That plane is known to become a stone with the gear extended

Probably for the same reason, sequencing arrivals is on of the most difficult things to learn for trainee controllers. You really need to grow experience on what speeds and levels are 'right' and what is too fast/slow, too high/low...

I can imagine! Add to that annoying pilots who want to fly different speeds all the time -)
I have to say that in Belgium, ATCO's are quite well trained for efficient vectoring.

[737MAX]

@737MAX: can you give an example of descent planning, f.e. ARVOL STAR 25L in EBBR? Or giving an example is impossible as winds, traffic, weight are not known?

For a 737NG, you have to multiply the distance to be flown by 3 to get the altitude at which you should be to be on the optimum descent profile (example; if you have 60NM to fly, you should be at approx FL180 or 18000ft).

You also have to take into the following into account:

1. Approx 10 nm extra required to reduce speed from 300kts to 210kts (typical flap extension speed)
2. Add 2nm per 10kts tailwind (or reduce by 2nm for 10kts headwind)
3. Weight also has an effect
4. The aircraft should ideally be configured with Flaps 1 or 5 on glide interception (3000ft - 10NM from the runway), otherwise speed will increase even with idle thrust without extra drag (such as gear).

The FMC calculates the ideal profile according to the weight, forecasted winds, cost index, eventual altitude restrictions etc... Using VNAV (a specific autopilot mode) is the best way to follow an optimum descent, but the thing is that it's only possible when you fly what you expected to fly and that's very often not the case. It's then up to the pilots to adapt their speed/configuration to the new profile.

Arriving via ARVOL 25L in Brussels means FL240 at the boundary (which is too low) and you will have to maintain FL80 until you are abeam with RWY25R north of Brussels as departing traffic is cleared to FL60/FL70 below you. You become then too high if you are all by yourself, and possibly too low if ATC requires extra miles due to traffic. When possible for ATC, you'll be given the track miles they expect you to fly so you can adapt accordingly (ironically, that is usually not given when there is too much traffic and that's when you would need it the most...).

[Poiu]

In fact it is the angle of descent and not the rate which is what you are after.
Aircraft have an optimum glide speed, which depending upon weight, is around 200-220kts for a 737/320.
The closer to that speed, the shallower the descent, so reducing the speed will bring the top of descent forward.
The highest AOD is achieved with gear and flaps extended, that’s why the E190, in your example, asked to reduce in order extend as much flaps as possible.
The 245ers appreciate a clear instruction:eg “maintain 300kts”, free speed means 245 according company procedures.
What would be extremely useful on the arrivals from the west is the track miles at the moment descent below FL80 is given.

[HQ_BRU_Lover]

737MAX & Poiu: thanks a lot for these very useful reactions. I've used to fly on the sim with the 737NG (PMDG package) which came close to reality what concerns graphics/cockpit features/... But reading these kind of replies on this forum shows to be very aware that flying a sim is not flying in real life (yes, I know that a lot longer - but just to confirm the great value of both posts).

Thanks once again both.

[Atco EBBR]

Thanks to everybody for replying, very interesting reading.

Arriving via ARVOL 25L in Brussels means FL240 at the boundary (which is too low) and you will have to maintain FL80 until you are abeam with RWY25R north of Brussels as departing traffic is cleared to FL60/FL70 below you. You become then too high if you are all by yourself, and possibly too low if ATC requires extra miles due to traffic. When possible for ATC, you'll be given the track miles they expect you to fly so you can adapt accordingly (ironically, that is usually not given when there is too much traffic and that's when you would need it the most...).

It's strange that you say FL240 at the boundary is too low. Whenever possible, I use 'when ready descent...' and almost never I see an aircraft leveling of at FL240 to start the descent later. Coming from KOK yes, then quite often descent starts 10 nm or more past KOK...

The problem with the track miles is that you have to calculate this case by case, the system doesn't know this. Say it takes around 5 seconds to calculate and you'll know why you can get that information only when it's calm... And we do realize it's very useful information when it's busy. A good arrival managing system should be able to handle that, hopefully that'll come with next ATC system (2022-2025?)

Probably for the same reason, sequencing arrivals is on of the most difficult things to learn for trainee controllers. You really need to grow experience on what speeds and levels are 'right' and what is too fast/slow, too high/low...

I can imagine! Add to that annoying pilots who want to fly different speeds all the time -)
I have to say that in Belgium, ATCO's are quite well trained for efficient vectoring.

The trick is not to allow pilots to choose their speed
And thanks for the compliment

Another question for the pilots here: when we vector, you are given a heading. Could a modern aircraft fly a track, just as easily? That would be easier and more precise...

[jan_olieslagers]

The trick is not to allow pilots to choose their speed

WHAAhahahaha!

Could a modern aircraft fly a track, just as easily?

I cannot speak about modern planes, but from a programmer's point of view it would be a lot easier. I have over and again insisted that, in the 21st century, magnetic information should be used only in extreme cases - after all, who uses a compass, today? Using only "true" headings would, for one example, get us rid of the changing of runway ID's every so many years. And true headings are not that difficult to calculate with trigoniometric difference calculations between (gps-derived) coordinates. Adjusting for magvar* adds a lot of complexity, since magvar changes both with location and with time. There are solutions to that, but the problem shouldn't exist.

*magvar: Magnetic Variation, the discrepance between true heading as measured on a map or chart, and what a magnetic compass indicates. For example, runway id's such as 11-29 at EBAW Antwerp are based on the runway's magnetic heading, which changes with the years - I remember when that same runway was called 12-30. And no, they did not install a pivot under it It would be easier and clearer for all concerned to base them upon "true" heading, as measured on a chart or map. And the ID's would remain constant, saving paint and effort and changes to the AIP and other publications.
Here in Western Europe, magvar is negligible, making the point rather moot; but I seem to remember it can be some 30 degrees in California, and even more as one gets nearer to the poles, either North or South.

[DIBO]

....that's not talking about wind drift, but doing a topic drift

[jan_olieslagers]

Guilty, your honour.

[737MAX]

It's strange that you say FL240 at the boundary is too low. Whenever possible, I use 'when ready descent...' and almost never I see an aircraft leveling of at FL240 to start the descent later. Coming from KOK yes, then quite often descent starts 10 nm or more past KOK...

The problem with the track miles is that you have to calculate this case by case, the system doesn't know this. Say it takes around 5 seconds to calculate and you'll know why you can get that information only when it's calm... And we do realize it's very useful information when it's busy. A good arrival managing system should be able to handle that, hopefully that'll come with next ATC system (2022-2025?)

How many track miles from the boundary to the ILS25L for a usual approach into BRU?
If you already are in descent, some continue but with a lower rate instead of leveling off for a short while. I rather meant that your descent starts too early if you have to be a crossing the boundary at FL240.

The trick is not to allow pilots to choose their speed
And thanks for the compliment

I like that!

Another question for the pilots here: when we vector, you are given a heading. Could a modern aircraft fly a track, just as easily? That would be easier and more precise...

On the 737 you can only select a heading on the autopilot, you would need to adapt the heading by yourself to continue on the requested track. But for modern planes ( ) like the 787, you can switch that autopilot mode between heading and track easily.