A320 Single Engine Missed Approach
 

Dear all,

I have a question regarding the missed approach procedure Engine Out. Specifically, when do the thrust levers go from TOGA to MCT.
Say i am coming in for the approach single engine. On the approach plate i learn that the missed approach gradient is a standard 2.5% and go around altitude 3000. I decide i want to follow the standard missed approach instead of a custom EOSID.
What i understand from some airlines i asked: at acceleration altitude (std 1500) the crew levels the aircraft for acceleration and cleaning the drag and once done, continue in open climb and promptly set thrust to MCT.
My reasoning says that once you do all that below missed approach altitude then you no longer can be sure of the approach climb gradient you have calculated in the performance calculations software of your choosing.

Any insights on the matter is most welcome, hopefully with a corresponding page in the FCOM, FCTM, ICAO Doc or a paraphrasing from an OMA. Only then i might be able to talk sense to someone in my company….


p.s.
Any idea where i can find the manual for the Flysmart Landing/Takeoff app?
Cheers

You are mixing up two things. The EOSID is the case where you have an engine failure at V1, on the ground, in a takeoff. It includes an acceleration altitude.

In a missed approach you are much higher a the start of the go-around and in many cases climbing over the runway. You can use the performance tool to check the single engine climb gradient, but there is no acceleration altitude specified. The acceleration altitude is either final go-around altitude, although some companies will allow acceleration once above MSA (and remaining in the vicinity of the airport).

If the single-engine go-around performance is limiting, some companies then specify you can always opt for the specific EOSID, but this is a little grey zone as you are mixing up two different aircraft configurations (TO vs go-around config). Ie on 737 takeoff can be flaps (1)/5/... and go-around could be flaps 15 single engine (engine failure on final).

There's no acceleration segment in missed approach procedure. Unless an operator has specifically checked out that acceleration at certain altitude can meet terrain clearance and gradient requirement you should be climbing to missed approach altitude before accelerating. And after level of once aircraft is cleaned and climb is resumed thrust is brought to MCT. Takeoff and missed approach paths are not exactly same. So EOSID may or may not be applicable. It will depend on the particular geographical situation.

BraceBrace Yes, i know i am mixing up the EOSID where it does not naturally belong, but as you said yourself:
And my company just loves grey zones. Here i am trying to figure out the hard truth.

vilas
Thats what I’ve been taught in my TR and so i started my crusade against all the TRE/TRI and chief pilot.

So can you two back this with any written proof?
I have tried the FCTM and FCOM and ICAO Doc 8168. I did not find any smoking gun. Just some vague things.

​​​​​​​cheers,
​​​​​​​

First of all, all PANS-OPS procedures are designed for all engines operative (8168 - 1.1.2)

Regulatory approach climb gradient (engine out) is 2,1% (2,5% for Cat 2/3) and not 2,5% as for all engine ops, so even if you follow published missed approach procedure (2,5%) there is a risk of busting obstacle clearance

As Vilas said, missed approach is designed without acceleration segment (there is a nice picture of it in 8168)

Taking all this into account, it would be unwise to accelerate before missed app alt (or MSA if lower), unless your company has specific performance calculations or if you are not 110% sure that you will be able to accelerate and still be above required gradient during acceleration :)

Ref. 8168 chapter 6.1

6.1.2 Phases of missed approach segment

In principle the missed approach segment starts at the MAPt and includes the following three phases (see Figure I-4-6-4):

a) initial phase — begins at the earliest MAPt, and extends until the Start of Climb (SOC);

b) intermediate phase — extends from the SOC to the point where 50 m (164 ft) (Cat H, 40 m (132 ft)) obstacle clearance is first obtained and can be maintained; and

c) final phase — extends to the point at which a new approach, holding or return to en-route flight is initiated. Turns may be carried out during this phase.

Climb gradient in the final phase: The criteria of the intermediate phase apply. (i.e. 2.5% or whatever it is specified)

Bottom line: you will always have to maintain some sort of positive climb gradient up to the go around altitude or at least the MSA. Check Your OM-A Chapter 8 regarding one engine out procedures and look for the requirements set by your company regarding the missed approach OEI. It should tell you what to do and when.

Another pill:

PANS-OPS 3 approach charts were designed with an acceleration segment depicted in the missed approach procedure. If you happen to be flying one of those charts you can accelerate and clean the airplane as per the procedure. For the rest of approach plates that are PANS-OPS, there is no acceleration segment included, so if you level off and accelerate, you will be doing so without knowing exactly if you are hitting the ground clearance slope.

And, that ground clearance slope that you are required to maintain always, is calculated from the MDA or DA. If you do the go around below minimum, there is no approach gradient chart to check your perfomance with one engine, so you are required to follow the engine out procedure (even with all engines operating), to make sure that if you lose an engine you'll never hit the ground clearance slope.

There are airlines that require pilots to follow the engine out procedure when doing the go around above minimum and with all engines operating, if the OAT is above the flat rate temperature of the engine. The reasoning is that above that temperature, the engine doesn't produce TOGA thrust anymore even if the thrust levers are in TOGA (flat rated engines produce full TOGA thrust up to a certain OAT, after which the output thrust starts reducing) and therefore the performance doesn't comply with the requirements.

I believe all this is in the 8168 ICAO document.

Toga limitation
 

Thank you all for your time and effort.
I just wish Airbus had wrote something less confusing than:
”At the engine out acceleration altitude, apply the same technique as described earlier” FCTM-PR-AEP-ENG-ONE ENG INOP GA
It seems to throw people off in my company.

Also;
How can i be sure that i do not infringe the 10min TOGA limitation if i follow the published missed approach (and so i reduce thrust at missed approach altitude)?

Ten minutes should meet any MSA/ Missed Approach altitude.

You know by using a timer... But even with 10 minutes... try Geneva at maximum landing weight.

Just two remarks: the calculations we do are based on a set of rules, minimum requirements, etc... but they are never "reality". The EOSID is a calculation based on an engine failure close at V1. It is not an engine failure at rotation. It is not an engine failure in your first turn. It is not an engine failure at the current actual weight of the aircraft, your aircraft will give more than regulations in many cases. There is a difference between flight preparation regulations and SOP's.

Also - even though many are totally against this practice for very good reasons (negative training) - some instructors might ask you to "deviate" from what you would do in real life to save time in the simulator. Engine out practice is usually one of those cases, where, to save time, quick accelerations in a single engine go-around are accepted to avoid waisting too much time. In some companies, the amount of excercices that need to be carried out in the simulator is becoming really a sad burden for the training people.

Same problem here… nothing written in OM-A or SOPS and everyone accelerates at a different stage during OEI go around…

At a risk of stirring the pot somewhat and upsetting various people, we need to keep in mind that the "rules" are based on history and history-based "reasonable" assumptions. The rules do not provide an absolute guarantee that all will go well on every occasion.

So, what to do ?

The poor cousin end of the market has a tendency to suck in some air between clenched teeth and apply whatever protocols seem to fit, SOP or whatever and, often, without much independent thought being given to the problem. The sensibly risk aware operator (and pilot, if you don't have a serious operator behind you) will make sure that the dice are loaded your way by running appropriate calculations in the design office. This, really, is just basic corporate risk management.

If the obstacle profile for the runway considered to apply for the miss is very benign, then the normal SOP will not have a great problem in general. For many areas, this covers a large proportion of the extant runways.

However, as the obstacle problems increase, there is a sensible need to do some ops engineering type sums to make sure that you know what you might be facing for the dark night situation when it all goes awry. Do you have all the data you might like to have ? Generally, not. But all is not lost.

There is no easy way out of the dilemma - you need the obstacle data, you need the AFM, and you need the technical skills and time to run some analyses to figure some weight limits and flight paths appropriate for the situation. Presuming you already have assessed the obstacles and have run any appropriate OEI escape plans for the runway, you should have enough obstacle data available to run the sums on a quasi takeoff process using the available AFM performance data and, if necessary, some simple flight test data to fill in any missing bits for the office work. It's not rocket science but it does take a bit of knowledge and understanding and the will to spend some money on the work up to provide the crew with useful and sensible data for the miss planning.

Some of us take considerable pains to get this stuff sensibly right, others tend to rely on motherhood statements and winging things on the fly. I know which approach I favour.

A few personal thoughts on previous post comments.

when do the thrust levers go from TOGA to MCT

If it's a planned for situation, much the same as for takeoff and you plan the story to meet the AFM engine limitations. If you are in some nasty emergency situation where pre-planning couldn't have been done, I don't think you need to worry too much about running a bit over the time limits if you have a reasonable need to do so. Maintenance might not include you on their Christmas card list for the year but that's the least of your worries.

I decide i want to follow the standard missed approach instead of a custom EOSID.

Bit late, methinks, to be making those sort of decisions in the heat of the moment. That is the sort of stuff you need to think about at pre-flight planning or, if you are in an emergency divert situation, during the divert cruise.

In a missed approach you are much higher a the start of the go-around

Perhaps, perhaps not. How about if you are faced with a miss from the flare or very late final ? And you still have to figure the distance needed to reconfigure to a takeoff situation where you can better fit the performance to the obstacle profile(s). Always keeping in mind the main mantra - "don't crash".

There's no acceleration segment in missed approach procedure

Which is why you need to run an obstacle analysis. While this can be done with gradients, and with some mathematical pushing and shoving, it is much easier to use a discrete obstacle analysis ie where you know the obstacle x, y, z data.

So can you two back this with any written proof ?

I doubt it very much. There is just too much variation to countenance the motherhood approach.

If you happen to be flying one of those charts you can accelerate and clean the airplane as per the procedure.

Not really. There is no "one size fits all" in the reconfiguration. Perhaps you might like to compare several Types for third segment distance data. Do include the DC9 in that selection ...


Now, has anyone thought much about this on a day where the airspace is filled to the brim with turbulence ?

Some points to note:

• 10 minutes TOGA//takeoff-thrust is a hard limit for OEI, you can damage a good engine if you exceed. It is not wise even by a few minutes.
• If you encounter an engine failure on approach there is no EOSID, you have to follow the missed approach procedure. The question is good one, does it take into account the climb gradient to ensure obstacle clearance?

Rest I agree with you John, great analysis. Great question also from the poster very curious about this one.

​​​​​​​If ten minutes is going to be an issue then shouldn't something planned before approach or takeoff itself?

Exactly.
Flysmart offers the possibility to amend the GA Gradient as required and to amend the target altitude till which said gradient must be maintained, although this last function has to be activated upon request by the operator.

Surely you understand why using the timer is not a viable option. (Stop calling me Shirley!)

What you say is almost the full story, but not quite, it arises out of §33.27.... and that is the way the world has been. Until now... the modification I am doing an STC on large engines, coming out of the TFE731 tests gives max rated TO thrust levels while the engine is below MCT RPM, EGT and fuel flow. (Turns out, there are nice things that can be done with aero).

Can you run the engine at full rating in an emergency for longer than 5 minute or 10 minutes? The qualification of the engine required demonstration to show there is zero damage in such a case, and if there was any damage, then the engine has to operate for another 5 minutes after the time that damage occurs. For the demonstration, it is not known and is not required to be known that the engine was run for more, so at that time limit, it can only be said that it is the required test point, after which, you are on your own. However, the test point is not at the RPM and EGT and thrust of TO/GA, it is at 120% RPM related to that point, and that is not a trivial matter; that is, your 100% N1 limit, the engine has to do the test at 120% N1, and that then means the thrust output is between 1.4-1.44x the rated thrust, the EGT will increase by the relationship it has to N2 (N3) and so has fuel flow, that is, your EGT will be far higher than the limit EGT that the engine has been tested to.

Will the engine fail at 5:01 at MAX TO/GA thrust? No. Will it need to be inspected, yes, it exceeded a limit. Some engines have a designated time exceedance inspection program, some do not. Coming off the stop is desirable though as you have one engine between your beer and being a glider.

watch the video on www.delta-burn.com which shows a very simple but repeatable test. The ground test matches the flight test quite accurately.

So, in this case, the time limit for the thrust level which is predicated on the margins that arise from the § 33.27 (a) Turbine, compressor, fan, and turbosupercharger rotor overspeed. requirement. From that, the question is what can you run a modified engine at... as modified, we are always around 10% lower RPM and 140C lower EGT, and 30% TSFC better than when the engine came out of the factory. If that sounds odd, know that for the pleasure of paying for MSP GOLD, the engine in this video came back after overhaul with a ZERO EGT margin, thanks! In the video, it is putting out more thrust than the standard engine... so... what is it's max TO thrust time limit? It never gets to Max EGT, (not because of the efforts of an MSP GOLD MPI). If run at normal target RPM for the rated thrust, which is say, 98.1%, 3,700lbf, it gets 5,250lbf out of it, which is a nice bump but not for normal operations. When pulling the RPM back equal thrust occurs with a >9% lower N1, which is lower that normal cruise RPM, an EGT that drops by 140C, 740C vs 885C and so what is the time limit on the engine? Outcome? This engine gets sea level TO thrust to 9000' sea level max CLB to 16,000' and above that it has 2 times the excess thrust that the standard engine gives. Alternatively, and as they are my engines, they run slower, colder and cleaner, and with lower fuel burn, and the range of the aircraft goes up. 50% up.

The STC program for the first engines has started up, this is straightforward on the engine, it is an irritant for Boeing architecture, but mainly simple for the Airbus, and the EPR target aircraft. N1 target is fine for manual thrust systems, otherwise G/A can be interesting.

For a normal case, 33.27 gives the time constraints and the qualification that has to be met, it just happens that the engine STC that I am doing kind of messes with the mind somewhat.:)

OK fantastic summary, thank you for taking the time to explain it at such a level! I won't argue with that.

Extremely insightful!

Now I'm curious. So it makes sure you reach the target altitude & accelerate within the 10 minutes before setting MCT or it gives you a warning you need a better climb gradient? I'm just curious to know how you make sure everything is done within 10 minutes.

On the Boeing OPT there is the standard MACG check where we can adapt the MACG to check single engine performance on the missed approach. But there is nowhere a check that verifies how long it takes to fly the full procedure up to clean-up. I might be missing something...

I might be missing something...

If you run it as a piecemeal analysis, you end up with the time to wherever and then you can iterate the process to figure what you can and can't do. No different to a normal T/O analysis, just a bit more involved in the detail.

Can you assume a minimum climb rate and ground speed then determine the time taken to reach a specified altitude or distance?

That's the whole point I'm trying to make: performance calculations won't cover you in every aspect. If people claim it's covered, I presume these are things done and checked in the simulator by a standard set of procedures.

Take Geneva 22 with a 3,4% climb gradient starting from 1500ft up to 7000ft. Standard 2,5% is 400-500fpm, (a performance check that we used to include in single engine go-arounds). 3,4% requires more like 600-700fpm. If with 10' thrust, you have to climb 5500ft you know it is going to take you already 8 minutes and there is 2 minutes left to accelerate. We have a 5' TOGA limit, but the performance tool we use is never going to warn us, it is up to the pilots to know and adapt and not think "it is covered".

At 66T GW for OAT 40°C and PA2000ft. the OEI GA gradient works out to 4•7%. So it should take even less than 8mts. That's why I said 10mts should take care off most if not all missed approach altitudes.

Geneva has standard GA gradient only thing is the missed approach altitude is high. What happens mostly is you climb at gradient much steeper than the required. So you may get sufficiently distanced from the obstacle clearance plane. In that case it may be possible to accelerate, clean up and resume climb at MCT without penetrating the plane. But that needs to be studied by the operator. More restrictive are high GA gradient go arounds like hongkong which is 7%plane.

Depends which minimum you use. CAT III/II and the lowest CAT I minimum require a climb gradient of 3,4% up to 4.500ft, 3,7% to 4.500ft to remain inside controlled airspace (good idea, there is a lot of VFR traffic around). On an A320 that is not limiting, on an A319 it can be, but choosing a higher minimum can assure the minimum required climb gradient, but will not keep the aircraft in controlled airspace which means that pilots might not even get traffic information, let alone separation and might have to deviate from the published missed approach to avoid unpowered VFR traffic, not ideal in an OEI situation close to terrain.

At least flysmart lets us check the OEI G/A climb gradient up to a given altitude, and that can sometimes be surprising at an otherwise benign airport like FRA where a climb gradient of at least 5% (up to 2.000ft) has to be met on 25R, which meant we could nearly never use that runway with an A319 and rarely with an A321, but nearly always with an A320. All subject to the power rating the airline buys/uses.

So let me put this straight, you're literally saying that the OMB of every single airline in Europe (or most of them) is... Wrong?

Mate, that's like shooting to your own foot.

You have a nice thingy on the OMB saying what to put on the perf go around page, published miss approach altitude and that's it, that complies with everything and your airplane does meet the requirements ( hey, that's why you're doing the approach after all, right!?).

​​​​​​​Very nice attitude towards the TRE/TRI and chief pilot, sounds like someone is gonna look for a job soon.

I disagree with both those points.

First the 10 min TOGA limit. 3 issues with this really. First, is it 10 mins total? Is 10 mins continuous? If I use 8 mins during the first go around do I only have 2 mins left? Second, it's actually an EGT limit that's been converted to a time limit because us pilots are too stupid to look at the EGT gauge. If I climb to altitude TOGA, MCT, and Climb Thrust are all the same. Does my 10 min limit count now? Thirdly, and this is the most important, if I need it I'm going to use it. 10 min limit or no 10 min limit. However, it's in the manual that it's a 10 min limit, and so it should be respected where possible. But people need a deeper understanding of why the limit is there and what it's secondly trying to achieve, rather than 10 mins and that's it.

The next point I also disagree with. On an SE missed approach you can follow the EOSID, or the EFATO procedure, or the normal missed approach, or even the normal SID, or any combination of the above. YOU are the captain of your aircraft, its safety is YOUR responsibility, not the manuals, not ATCs, not managements, YOURS.

You are mixing up things. The 10' limit is a hard limit that should be respected at all times, unless your life depends on it. From the moment you apply the rating, you can only keep it applied for consecutive 10 minutes. The idea that pilots need a "deeper understanding" is very dangerous as the story is usually not that simple and contains a wide series of domain, from insurance to airworthiness and maintenance. Claiming you "know" is usually an indication you don't know the full story and the consequences of your actions, and you're going to have to do a lot of explanations to cover those consequences. We've recently seen a very poor example with an A330 diverting with a fuel leak and keeping the engines running because they had different ideas than the procedures. The most standard black/white clear procedure of all procedures are limitations. We can all do what you say, easy. It's not about what we can or how much we know about it. The reason why is much less important than the possible consequences of not adhering to the limitation.

I think the topic makes clear that there are limits to what can be prepared and verified, and those limits provide a very good level of safety. However, that does not cover every single situation (every operator is very clear on this topic as well), and it is your job to be prepared and adapt as appropriatly.

You have made my point quite well. Its not a hard limit in that if you exceed it not much will instantly happen. It is however a limit. Like the MLM, or VNE, or all the other limits we have on aircraft.
Use it for 10 mins, but don’t crash into a hill because your 10 mins are up.

And here’s a thought with regards to the 10 min limit. If I use 10 mins for TOGA, what has to happen before I can use another 10 mins?

Having a deeper understanding of your aircraft is not a dangerous thing, it worries me that you think it would be. It stops you thinking you know better, but it also allows you to understand what the actual limits are and why they are there. If you do have a better understanding of your aircraft you are less likely to start arbitrarily disregarding set procedures, but you are more likely to chose the right procedure for your situation. You’re also far more likely to understand the outcome of any given procedure and its overall effect on your aircraft before you commence the procedure and therefore make decisions in a timely manner.

Do you notice all the nonsense you're talking? Like for real, aren't you worried about it? "OH THE 10 MIN TOGA IS NOT ACTUALL 10 MIN, WHAT IF WHAT IF..." 10min and then MCT, it's not about agreeing or not, IT IS LAW. You dont do the 10 min? Good, prepare your ass, if everything goes good you'd better have a very good explanation, if something goes SLIGHTLY wrong... Say goodbye to your license, job and maybe life ( jail time, remember not following procedures and the result is someone being injured could cost your free will ).

Also, you have a very nice app which calculates gradients for the go around/sid etc and if not, you'll most likely have several thousands of sheets on your ipad saying which gradient you'll have on an EFATO/EOSID/EOGA or EVEN your airline has a very VERY NICE study in which all or most part of the airports you fly to, with their procedures are withing the envelope of the aircraft. Might have to check again those ATPL subjects or some ICAO documents.

I cant believe you go flying not knowing that you aircraft complies with the gradients not to mention that you guys believe that the airline does that and you just follow blindly. That goes for maverick as well, the initial pilot warrior that started the post.

Oh dear.

Let me ask you a practical question, engine failure on approach to destination, high enough to go around and not low enough that you would continue. Can you tell me the exact EOSID? No you can not, hence you will follow the missed approach procedure.

This is why the question is relevant and we are not considering the EOSID.

The question shouldn’t come up during the go-around, but during the landing performance calculation in cruise and the subsequent approach briefing and TEM. At which point the EOSID would be briefed and prepared if required. Have done that, informed ATC about it (ZRH in that case) and flown the approach, alternative would have been to divert, as the OEI go-around performance has to be within limits.

That's what the TRE next to me has said when he saw your comment about " respect a limitation where possible". Not worth

Im really not sure why this is controversial.
Just to highlight two very mundane times aircraft limitations cannot be respected. Overspeed during turbulence in the cruise, this happens on a regular basis all over the world. Second would be an overweight landing. There's even a check list for this one! Ive done it at least 3 times after shutting engines down inflight. Wasn't sacked, jailed, or even asked for a "very good explanation". Was thanked for my efforts on every occasion.
Just to go back to the OEI 10 min TOGA limit, should it be respected where possible? Yes, of course it should. And if you have TOGA selected and your EGT is red lining, then you need to keep a VERY close eye on it and 10 mins might be stretching it. But also, if I need to use 11 mins of TOGA to ensure the safety of my aircraft then I'm going to do it. I would hope every pilot would.

Just to reiterate my point.
Have a look at the A321 (CEO, NEO is far worse) gear down OEI performance. It’s a bit of an eye opener, especially at average take off weights. Anything over about 84/85T and you’re on your way down, and that’s vastly better than it used to be on the non sharklet aircraft. In such a situation I might just forego the 10 min TOGA limit if I have to.

I fly to many destinations where the missed approach climb gradient is in excess of 8%. In some cases as high as 12%. If you are making an approach and lose an engine you are not doing the standard missed approach. What you’re probably going to do is fly the engine failure procedure. And as Denti states, this will have been calculated and briefed during the cruise.

Just as an aside to this, no airline I have flown for in over 25 years has has EOSIDs in the FMS.

Again, these are 2 diffferent cases from a pilot's perspective. The overspeed and also overweight landing when a checklist says "land at nearest suitable airport" end up in techlog entries and at the time deemed necessary end up in maintenance actions and are covered. The overspeed is a nice one. On my aircraft you can go 30 kts into overspeed and maintenance can still sign you off without any inspection. It is unprepared, it happens. Is that a reason to say "oh, what the hell, I'll just let in increase speed, it'll be fine"? No it isn't, it remains a limit. Boeing knows this happens, you are covered. The overweight is one where the situation should justify it and even for the overweight landing, the books explain pretty clear these situations.

However for the 10' limit, the books are also very clear in that it is a hard limit and you shouldn't exceed it unless "to save your ass". And you seem to have a very wide interpretation of that phrase. You seem to interpret this as any other limitation "oh, I know what I'm doing so I can go for 11 minutes, just like a go into overspeed". If that ended up being the case, you f$cked up pretty hard during your preparation as these are very remote situations. If everything is prepared well, you shouldn't end up in this situation.

The background knowledge you have as a pilot is one that should be comforting, so that you feel at ease the engine will help you out as required for 10 minutes. It's can't be a reason to think "with the necessary personal background limitations, I know I can even bend the 10 minutes limitation".

I haven’t found anywhere in any of the manuals where the 10 mins TOGA limit is deemed more or less limiting than any of the other aircraft limitations. Nowhere does it say “to save your ass”. The 10 min TOGA limit is a limitation, and should be respected as such. I don’t have a “wide” interpretation of that phrase because it doesn’t exist.

What I am saying here is that the 10 min limit can be exceeded, but only in very limited circumstances where it is used to ensure the safety of the aircraft. Yes these are very unlikely situations but if you need it you need it. my example of the gear stuck down and single engine operation (that actually happened to a colleague, fortunately in an A320) would be one such example of where the 10 min limit would be exceeded.

What I find very interesting is that everyone knows the 10 min limit, it says it in the book. What is that limit based on? How long is it before I can use another 10 mins? Does the engine need maintenance action before another 10 min TOGA timing? Is it cumulative over the entire flight? These are all questions that pilots should have answers too. Understand your aircraft. Not just spout the book answer. Anyone can do that.

Jonty: I appreciate where you are coming from. In the late '70s when I was flying the 707 my then company paid to use a procedure which had been devised by PanAm which allowed the use of MORE than Max Take Off Thrust out of Nairobi. So, the expected EPR was increased and the only "hard" limitation applied was to the max N1 rpm. This allowed an increase in both the field limit weight and also the second segment climb. We understood that effectively engine life was being "sold" at any rate I never had any engines give out using this procedure.
With more modern engines I suppose that the only time real TOGA is used is for the initial stage of a go-around and certainly seldom used for take-off.

As a guy who has worked as an engineer in an engine workshop, and currently flying a B737...

1) The overspeed and overweight limitations and use are clearly explained in the FCTM. The books never say "to save your ass", that is my phrase. The books will always state the PIC is allowed to deviate in the interest of safety. However these situations are pretty clear and as said, the deviation should not be an excuse for lack of preparation.

2) your question on where the limit comes from. It comes from certification, it is related to airworthiness and maintenance plans and can be found in the AFM. If 20 years ago you would have asked me (the engineer) the question, my answer would be easy: the why is unimportant, it is a result of certification tests following a laid out plan that gives you a certainty it will work for 10', and continues to work for the remainder of the flight in MCT thrust situations. If you decide to deviate and bust limitations, you are entering a grey zone where engineers might not have valid test data, and basically nobody will certify your engine will continue to work as expected until you're back on the ground after a diversion.

So you might feel comfortable doing so based on your knowledge, the engineer will be very curious to look at your engine to answer some questions he didn't have the budget for in the testing phase.