The SOP specifically forbade the DLH from applying visual separation themselves. Pukindog is emphasising that pilots have to keep an eye out at all times to see and avoid, which is fair enough. This doesn’t override ATC being responsible for providing separation in Class B though, which is a specific criteria, usually more than just “make sure they don’t hit each other” - which is what you’re otherwise aiming to achieve with the see and avoid responsibility in VMC. See and avoid as a form of separation is what applies in Class G, the whole point of having higher classes of airspace is to provide protection abovejust see and avoid. If DLH want to ensure they’re given that additional protection at night, so be it.

Some posters were saying “well they should have taken more fuel for the inevitable delay”, again fair enough. However at no point that we know of did the DLH ever get an accurate delay. They were initially given “extended delays”, which apparently meant “at least 40 minutes”. We know they were told 10 minutes after about half an hour, then 14 minutes after that told “another 10-15 minutes”. How much extra fuel do you need to uplift for a 10 minute delay that extends by another 10 minutes every 15 minutes? How are you supposed to quantify that? If you’re not able to provide an accurate time to pilots, how on earth are pilots going to be able to make a definitive plan. A pilot asking for the delay in minutes shouldn’t need to be a conversation, it’s a question followed by an answer.

This isn’t an unpredictable weather scenario where the delay is fluid. If the delay is because you already have traffic sequenced back to SLC and you’re not willing to delay that sequence, then you know it’s going to be more than 10-15 minutes and you have to be honest and accurate. Whatever your stance on the justification for the delay, it doesn’t mesh with what the DLH was being given in terms of delay minutes. There’s obviously no gap in the sequence so why string them along?

Related or unrelated to the case as it might be,

If that is an accurate depiction, wouldn't that be a sign of a deep-rooted deficit in the specific operating culture? I would like to believe that that manager would be met with swear words as well.

Speaking of underlying causes, neither participant of the present exchange, nor any aviation stakeholder would probably mind if the narrow runway spacing at SFO was fixed. Outside players, that might be a different story ...

Request Orbit
Thanks for taking the time to correct my misunderstanding specifically about SOP.

About the multiple estimates of delay. I had wondered what could have led to ATC providing such a sequence of repeatedly unreliable estimates. Tempting though it will be, I'm not speculating about whether such unexpected, or even surprising, moving estimates of delay time can be traced back to controller staffing problems and fatigue issues. (The tempting part relates back to the report issued yesterday by FAA's National Airspace System Safety Review Team, organized in the context of FAA's Safety Summit earlier this year.)

Nope, that is exactly what ATC are there to do.

I'm looking in the FAA AIP for San Francisco. Full set of references to ILS on runways, with their identification codes. No reference to this not being available at certain times at all.

Is there a NOTAM saying this ILS will not be provided if requested ?

No, the LH SOP is not against visual approaches is about maintaining visual separation , 2 different things. LH SOP said only in daylight, so Capt has no discretion here but to follow his SOP.
again no,, no sizable degree of discretion , a controller has to accommodate a pilot legitim request, And he was initially planning to do so , but it escalated and in the end forced him to divert. hence this discussion . We are all trained to re-arrange sequence, we always do so after a go around for instance.
It may look that way but in fact this delegation of responsibility for separation is increasing workload as the number of aircraft close to each other has increased to the point where taking over in case of a mishap by one of the aircraft is almost nil. It is the number of aircraft you handle and have on the frequency that you have to monitor that is causing the workload and fatigue. Not the procedure in itself. If you are short on staff and fatigued, you want aircraft on published instrument arrivals well separated from each other , separated by a few NM, not laterally by hundreds of feet .
Plus as mentioned correctly by Request Orbit ,in class B airspace that procedure does not relieve the controller separation responsibilities , and since you are a lawyer, in case of a fatal mishap here, who do you think the Prosecutor will go after ?

@WHBM : .excellent question , but I guess we know the answer to that one.

Thanks for all the clarifications .... corrections of my own misunderstandings. Perhaps the diversion of the thread will help someone else who also actually wasn't ready to particpate intelligently.

I follow your logic (ATC W) about a potential prosecution on those facts, although if there's precedent for it in United States I admit I'm not aware of it.

Monday morning Armchair quarterbacking here purely in the interest of learning and avoiding a situation like this in future.

Most airline SOPS have a statement written within something along the lines of-The Captain has the discretion to amend/ignore/adjust the SOPS if she/he feels it is in the interest of safety.

I feel in this instance perhaps if they had just accepted the approach and landed and then submitted a Safety Report to their Company with the reasons why they did it-Much safer to land at a familiar airport, diverting is making a long day to begin with even longer and more fatiguing, zero chance of pax aggravation issues if they land at their destination...the list goes on.

At the end of the day LH would want their paying passengers in SFO not OAK. So, if it is safe-go against the SOPS (as we all have on rare occasions), file a report, so there is a written record, have those tea and biscuits with the Chief Pilot (he'll probably give you a minor bollocking, and then quietly thank you for landing at the filed destination), discuss changes to the Company's visual approach procedures, and move on to your next flight.

Just out of curiosity, when it was clear that a stalemate was occurring, would it not have been possible to divert to Oakland, while asking for a slot that met the sop and would not cause the crew to go beyond their hours, and screw up the arrivals and make the 20-30 min flight (including the approach) from across the bay?
Seems like that would have been a win-win

After putting on enough gas to complete the mission…

I am still trying to wrap my head around the crew duty day if this is about safety. Most airlines report 90 minutes prior for an international flight. 12.5 flight hours puts the day at 14. 2 hour delay leaving FRA and 2 hours on the ground at OAK puts them at 18 hours before departing OAK. How the pilot in command justified a departure at that point is hard to understand and illegal under under US rest rules.

Their rest rules are obviously different. Under 117, you could accomplish this with a 19 hr FDP if there are 4 pilots. It would be illegal with only 3 pilots though. 3 or 4 pilots, I'd have probably been done after setting the brakes in OAK.

Our safety people wouldn't care for that at all. If he'd exercised his authority after fuel became critical, that'd be fine, but to do it right off the bat would probably generate a phone call with some follow-up questions.

Since you’ve brought up the subject of rest, a bit of slightly off-topic background. Our standard is 6 on/4 off (6/3 for non-H24 units). The US standard is 5/2, however a lot of units are so short of staff they have a mandatory rostered overtime on day 6, so they’re effectively doing 6/1. This would break a whole heap of our rules in the way it’s rostered, but the biggest one is 3 periods of at least 54 consecutive hours off in any 30-day rolling period. Without going sick, there’s no way for them to even get a single 54-hour rest period, let alone 3 every 30 days. I’ve no idea if Norcal is on 6/1, and it doesn’t appear to directly impact this event, but I’d tend to be far more worried about how much rest the average US ATCO has had in the previous 12 months, than the DLH pilot going a couple of hours over in this case.

If you want I can breakdown all the different ways “the rattler” roster they get would break our rest rules (including finishing a morning shift at 3pm then back for a night shift the same day…)

If you're actively putting the commercial interests of your airline ahead of safety, and accepting that your actions will result in tea and biscuits with the Chief Pilot, you seriously need to consider how you're operating!

Without knowing any details of LH operations, if they prohibit visual approaches at night, I think credit is due to the operating crew for not being forced into an approach they didn't want to do. People may have an opinion that a commercial pilot should be able to shoot a visual, but that is neither here not there. A rushed, unbriefed approach, at night, after what I can assume is a long duty while possibly feeling the affects of fatigue is not where you want to be. So for me - credit to the flight deck for not doing something they were uncomfortable with.

In relation to how ATC handled it, some serious questions need to be asked. The LH crew made ATC aware very early on that they were requesting an ILS, so why wasn't this provided? Was the ILS unserviceable and on the ATIS? Surely a large airport like SFO has other approaches available? Why is a visual approach at night the only option to the crew particularly when the weather wasn't exactly "visual approach" weather? Why are ATC dictating what flight crew do with their aircraft?

This is my opinion - the FAA needs a complete overhaul of all its operating procedures from light aircraft, to RT, to ATC and aerodrome operations. How many catastrophic near misses has there been this year alone with runway incursions, aircraft trying to land on occupied runways in LVPs and general poor airmanship. I think it's only a matter of time, unfortunately.

Regarding FAA programming (or at least bureaucratic) response:

https://www.faa.gov/newsroom/faa-tak...ecommendations

Item from FAA website includes link to very recent National Airspace System Safety Review Team report, as well as other FAA initiatives. (Provided here without further comment, cynical or otherwise, about effectiveness, or even relevance, given Congressional and political dysfunction rampant in United States.)

I would have sent a tow truck to tow the plane by force :ugh:
Crossair 3597 has nothing to do with visual approaches at night.
The captain was someone know for his terrible technique of diving to the ground to get visual references. They were in IMC during this approach.
​​​​​​​

Just some added info for people unfamiliar with how most of these visual approaches are done:
There is the General aviation Visual Approach, were the center controller clears you direct to the airport, you see the field, inform ATC, he cancels ATC services, you find your own way to the airport, looking and avoiding other traffic on your own. That is not what SFO approach wanted from DLH.
For 121, scheduled passenger flights to big airports, everyone is vectored/altitude/speed controlled by ATC, until on a reasonable heading to intercept the straight in approach you will use to back-up the visual. You call the field in sight, and intercept the normal LOC/CRS followed by the GS/GP. The reason for the visual is things like reduced spacing,or less workload for controllers. In SFO it is because the parallel runways are quite close, so in order to use both simultaneously, without having to use PRM/SOIA approaches (more dangerous in my opinion), they use "visual" approaches. I firmly believe ATC should have worked with DLH better. But I am equally convinced DLH should have made an exception for the way these approaches are conducted at major US airports. When I flew in the EU, in night VMC we were not allowed to go below the MSA/MDA/MAA/MOCA/MEA/MVA. That makes sense. The DLH SOP does not.

EXACTLY! And this is what they ended up doing going into OAK. To complicate things, it would be an unfamiliar airport to add an additional threat to their already long day. And after landing now they have to deal with refueling, loadsheets, performance calculations and FM programming while being very fatigued.

Nope, I wrote, if it is safe-go against the SOPS (as we all have on rare occasions)

And the reason I wrote this is because it is far safer to shoot a very routine, radar vectored Approach onto an ILS that, at SFO for convenience, is referred to as a 'visual approach', than to divert to an unfamiliar airport, at night, with all the added stress and challenges that diversions always provide.

I would absolutely welcome tea and biscuits with the Boss, as it provides a great opportunity to bring to management's attention these sort of anomalies that show up from time to time on the line.

Let me asked you this: what would they have done differently because this was a visual approach as opposed to an ILS? The answer is nothing. Nothing different. They receive a very nice RV to intercept an ILS. They fly the ILS and then they land. How did NOT accepting this 'visual' approach at SFO make the operation SAFER?

Reminder noted, but I believe anyone who's flown in the US outside Class A -> STAR -> Class B/C -> SID -> Class A understands that there's a whole lotta aircraft out that aren't the merely ones ATC tells us about, or is even talking to, or even has a transponder or ADS-B that ATC or our TCAS can see. The concept is well-ingrained and why FAA regs require looking outside to try and visually acquire traffic anytime flight conditions permit.

I hope nothing I wrote implied that TCAS be used as a separation tool. That would be like running a canyon high speed, low-level, and blind using a weather radar in ground mapping mode. It won't end well. I don't know any pilot that isn't aware of TCAS's limitations and inaccuracies between display vs the visual reality outside; we use it every day and it's been around for 30 years.

Only RAs should result in maneuvering, and that's been the case since US carriers were equipped with them by 91/93. I'm not sure who you were lobbying hard against in the 90's. UK/Europe was a late TCAS adopter in 2000 after which, for awhile anyway, there was still debate and training differences as to whether RAs took precedence over conflicting ATC instructions as the 2002 midair exposed. Hopefully, everyone is on the same page now.

TCAS display as a substitute for visual acquisition...absolutely not, of course. I know my long post is eye-glazing, but that is exactly the point for much of it. TCAS was the first thing on the list I mentioned those things that aren't a substitutes to acquiring traffic visually, whether the aircraft has been pointed out by ATC or not.

What I did say was that it's an aid that assists visual acquisition by directing attention to sectors of sky for scanning. That is literally the stated purpose of having Proximate, Unidentified, and Traffic Advisory symbols with relative altitudes shown on the display, with the TA alert designed to reduce the startle factor and response time if an RA occurs. Mere mortals can't tirelessly focus on everything everywhere all at once and we must prioritize duties and attention according to the situation. In the case of SFO when they're conducting parallel visuals the traffic call from ATC will usually come between the vector for intercept and clearance to fly the published Visual approach and the before intercept. If that traffic call is the first time someone looks outside to attempt visual acquisition of the aircraft they'll pair up with...because you know you'll be pairing up with someone...he/she might want to reconsider his/her priorities.

The problem isn't that the other aircraft ATC wants you to see can't be seen (at that point on the SFO visual it easily can be). The problem is that ATC has called out 1 aircraft but there may be 3 or 4 aircraft easily in view from the cockpit window, and if the traffic is low perhaps some ground lights could also confuse. The process of the brain differentiating between things one sees and compensating begins when first looking outside, so maybe the best time to begin that process isn't at 180 kts converging with the other aircraft when someone also needs to be minding the store confirming that the intercept is happening when it should. Life gets a lot easier if one begins to get the SA picture beforehand and allow the brain more time to process and differentiate in order to avoid what could otherwise be an initially-confusing picture later.

Visually mis-identifying aircraft has always been an issue long before the advent of TCAS, which in my opinion has reduced the likelihood. Pilots are fine with azimuth but estimates vary greatly when it comes to judging altitude differences relative to one's own, especially over slant range distances. At night you're looking for aircraft lighting against what may be a background of lights with all sorts of brightness, colours, and flashiness. Or it could be a hazy day with low sun in your eyes and the vis from where one sits is far more restricted than anything being reported. So if ATC calls out 1 aircraft but TCAS is showing 3 in that sector in close proximity and minimum vertical spacing (perhaps just 500" if 1 is VFR) your brain already accepts that the first 1 you spot visually may not be the one ATC wants you to identify so you continue to scan.

Low aspect targets are more difficult to see until close-in and if they're hot there's less time to visually acquire. Constant bearing crossing traffic has little or no relative motion. When ATC gives a traffic position it's not BRA but rather a Clock Bearing relative to flight path, not relative to aircraft heading, so unless there's zero x-wind component the traffic can possibly be positioned 1 or 2 clock sectors off. In daily use, I doubt there's anyone who doesn't reference the no-conflict Proximate Traffic targets and try to spot them, and in this way through familiarity makes one better at visually judging relative altitudes and slant range distances air-to-air because you have something inside the cockpit displaying their altitude.

Until ADS-B w/CPDLC is universally adopted, all aircraft equipped, and always-working worldwide, the TCAS info displayed to the crew will always be used for situational awareness monitor proximate traffic and, as such, sometimes for positioning where surveillance and/or ATC is sketchy or non-existent; there's still wake turbulence at altitude that may need offsetting from, the occasional oncoming traffic that's uncomfortably at or near your altitude that should be monitored, or CBs that need deviating around along everyone else having the same idea.

While TCAS certainly has its limitations it's not your grandpa's Pong game either, and It's better to have a slightly imperfect picture of those around you than no picture at all. In those places where ATC does exist but adhere to a funny, aviation-variation of the "Less is More" philosophy that prevents them from advising you of overtaking and/or nearby traffic because "You won't be able to miss seeing it soon anyway" or perhaps the gov doesn't have the funds to go around replacing worn-out ATC PTT switches, this also applies.

"We have him on TCAS", yes, yes, absolutely agree. An entirely useless response to a traffic call that does nothing but create uncertainly with the Controller (and everyone else) as to whether the pilot is looking outside for the traffic or not. I'm glad you brought it up, and perhaps we can all agree now's the time to establish a worldwide system of fining any pilot $500 (or pro crew $500 per head) for uttering it on the radio, paid into a fund that will be used for a yearly Pilot/Controller bash somewhere on neutral ground; a chartered 250' Feadship party boat floating on the Caribbean, for instance.

In fact, since the Authorities everywhere aren't doing anything about it and in the interests of getting the biggest boat possible, there should be an entire list of useless, annoying, and offending transmissions and their associated fines, the amount for each commensurate to its annoyance level. Something like;

1) We have him on TCAS ($500)
2) F*** ($1)
3) You have a stuck Mic ($1500 If repeated, + $5000)
4) Fully ready ($480)
5) ____ on the meter ($975)
6) Charlie Charlie ($ 328)
7) You're on Guard!!! ($ 2,500 + 3 Anger Management classes + 5-year Party Boat ban)
8) Animal noises (Entire estate + Death)
9) etc.
10) etc.

My reading of the radio comms isn’t that LHA can’t do the visual at night, they just can’t do the separation with other traffic at night. I know that first up he says he can’t do the visual approach but I think that may be a language issue as he later says that SPECIFICALLY it’s the visual separation that’s forbidden. Granted he requests a standard ILS but maybe that’s just a way of saying don’t give me the dependent visual approaches.

There are two related but different parts to this:
1. Visual approach - can he/can’t he? I’d be surprised if LHA ban it at night (but maybe they have) as they still need to arrive into JFK, SFO etc
2. Maintaining visual separation with another aircraft - he says he definitely can’t.

My suspicion (will never know) is that what happened was:
1. LHA forgot to/didn’t give NORCAL enough notice that they needed special handling and were unclear about what the actual SOP prevented them from doing. From the radio it sounds like they were well into the arrival when it first came up
2. If they’d been clear that they couldn’t do the visual separation at night then NORCAL could have simply put them in the queue without the second aircraft on the parallel and all would have been fine - 1 landing slot lost
3. If they insisted they needed the ILS then with enough notice the sequence wouldn’t have been badly affected - maybe a total cost of 3 slots
4. By leaving it so late to advise, if the arrival queue was as long as it can be and 2-3 spots had to be found then that’s an unrealistic expectation for NORCAL to reorganise or slow the whole sequence or to take 2-3 aircraft out of the sequence to fit LHA in. Instead of holding, should have vectored him to the end of the sequence and been able give an accurate track miles figure instead of vague holding times.


3.

My understanding is that SFO and LHR have about the same number of average daily movements. (Is that correct?)

I have never heard a 'heavy' required to do any visual spacing at LHR, especially at night.

Why can they cope and SFO cannot?

One Other side of this story, which was not really discussed in detail, is How Exactly LH came to forbid pilots to take responsibility for visual separation (under certain conditions).

SFO has two tightly spaced runways. I have never had parallel approaches into LHR. Perhaps on the odd occassion it happens, but not regularly. At arrival peak there can be significant holding into LHR though.

Both places has great ATC working the traffic and they seem to the best with the cards they've been delt.

Reading through the later comments, my impression is that a few commenters have not watched the video in post #71, which provides an unofficial ATC explanation of this incident. According to that source, LH received an appropriate level of ATC service given LH's operational limitations (no visual sep) and routine priority.

Regarding whether ILS approaches were available, PAL104 requested and was given the ILS 28L approach. It appears that NorCal was able to accommodate this ILS request because spacing just happened to work out at that point in time relative to other arrivals in queue.

Regarding Euro carriers not accepting visual sep at night, Virgin 14R accepted maintaining visual sep to 28R.

Several commenters have stated that NorCal should have just “created” the necessary slot for LH, without specifying how this was to be done. From the post #71 video, appears two options were available to make this happen: (1) knock one or more aircraft out of the arrival path to create the required spacing for a LH ILS, or (2) take many aircraft off of the published arrival (i.e., vectors) to add track distance to create the space. The source in post #71 said neither of these was a a viable option given LH’s routine priority. What is not clear is why, after initially telling LH that his ILS request would cause “extended delays,” NorCal was unable to give LH a firm delay time to allow for fuel planning.

What this incident really highlights is the need for SFO to add capacity with a new, properly spaced runway (not green, ain’t gonna happen) or for the airlines to limit slots during peak times (hub & spoke, ain’t gonna happen). Thus, here we are.

Or to tank a bit more fuel.

these utter muppets at lufthansa do even the radio the other way around.
I am so glad to see ATC showing them their place.
Saying f**** on the radio is just beyond any standards. maybe over Germany can say those words as it seems he was quite relaxed throwing that in a foreign FIR.
I wonder if his low modulated voice kept the same tone when sent to his utter embarrassment diversion.
Well done SFC ATC !! 👍

There is no room at SFO to add another appropriately spaced runway, being “ green” has nothing to do with it.

Exactly! People tend to forget that many of these airports have been there for 80 or 90 years. Out in the boondock's when they were built, but now completely surrounded by urban development making any runway expansion somewhere between horrendous expensive and impossible. When they added a runway at SeaTac (although technically a third runway, it was really just a second runway far enough away that they could use two runways for ILS at the same time), the literally had to extend the plateau that the airport was built on by a huge amount (and at tremendous cost) - rather controversial since there is now a big drop-off if you depart the new runway to the west so potentially turning what would have been an embarrassment into a catastrophe).
In Denver, the old Stapleton airport was well out of town in the 60's, but by the 1980's it was surrounded with no room to expand. They had to spend $billions to build DIA from scratch way east of Denver to provide room for expansion (much of the old Stapleton airport is now a shopping mall).

tdr, you are usually on point. But not this time. The two examples you cite are (were) land-locked airports, which certainly present expansion challenges. That is not the case with SFO as San Francisco Bay presents almost unlimited potential for expansion. In fact a significant portion (50%?) of the existing SFO footprint is built on bay fill. Oakland’s long runway is built on bay fill as are portions of the various military facilities along SF Bay.

The technology for such expansion is well established; e.g., HNL reef runway, Hong Kong, Chubu, etc. Plans for this type of expansion at SFO were floated in the 90s with the #1 obstacle being environmental groups. If the bay area governments can toss a couple $B at new sports stadiums, they can certainly pony up some dollars for a new SFO runway. What is lacking is the political will to buck the greenies.

More than greenies, there's already a large anti-noise group in place. Why should non-pilots want a new runway, which would only increase the noise due to more operations? I'm sure that most bay-area residents wouldn't mind if LH moved their operation to a less noise sensitive area -- like Fresno.

Bucking the greenies and other environmental groups is expensive - usually extremely expensive, and getting the needed environmental plans and permissions often takes decades - and even once approved, you can usually count on numerous more lawsuits and challenges - all of which cost time and money. Filling in sensitive wetlands to create new land has become an environmental mine field. This being the California Bay Area means the environment groups have outsized influence.
Concerns over tiny fish most people have never heard of have already shut down $billions in agricultural output by limiting the water available to grow food. Increasing the footprint of SFO so more carbon spewing aircraft can land in there is going to be a political nightmare in hugely 'progressive' California.

PukinDog (post #11):

Very succinctly explained; thank you!

The runways at LHR are too close (just!) for parallel approaches, so you won't get parallel approaches there. (If any 'appear' to be parallel, that is simply for other operational 'convenience' but the spacing will still be the same as if on the same approach.) The runways at LHR are about twice as far apart as those at SFO?

Approaches into LHR are on the ILS. Even when I flew into there in a Cessna 310 (with an 'Ambulance' callsign!) in CAVOK during the early morning 'rush', it was an ILS approach. I have flown into there many times in a 'regional jet' and every approach in all conditions was an ILS. Yes, there can be some holding and if I remember correctly, LHR required that 20 mins holding fuel should be available even if 'no delays' were expected. Why doesn't SFO stipulate the same?

I have also flown many approaches into AMS (before and after the new 'Voor Politieke Bedoel' -- the ILS ident! -- runway!) at all times of the day (and night!). It was always an ILS unless a visual approach was offered to us for our convenience (with us being 'considered local' operators) and all those that I remember were in good visibility and daytime.

(The controllers at LHR and AMS are the best that I have worked with.)

From what I read on here, SFO appears to be 'a large GA airfield' that has big airliners flying into it!

SFO and LHR/AMS aren't really comparable.

Parallel runway separation: SFO: 750' LRH: 4600' (about 6 X the distance) AMS: Practically separate airports.

I can see how from reading some of this that SFO seems like a "big GA airport with airliners flying into it but it's really not.

Although the 757's are the largest allowed, for the sake of PPRuNe its good they don't have Int'l flights into DCA. Someone might have to fly the River Visual rwy 19 then comment about the maneuvering while skirting a couple prohibited areas along the way. If this SFO thing is an indicator, the ensuing uproar would melt this website down within 24 hours.

The river visual is heaps of fun :)

I’ve been trying (unsuccessfully) to understand the obsession with flying an ILS on a VFR night…

SFO has about 1300 arrivals a day. LHR about 650. That’s not about the same. It’s substantially different.

I’ve watched that video, and disagree wholeheartedly that they were given an appropriate level of service, whatever it claims. It was not treated with “routine priority”, it was dumped down the order behind aircraft that turned up 30 minutes+ after it, and basically treated as a Cat. Z with zero priority. It made a perfectly reasonable request to fly a published instrument approach, that so far as I can tell wasn’t notified anywhere as being unavailable. They knew it was going to be inconvenient and were expecting some delay.

No one has stated how they’d create the gapbecause it’s such a routine part of the job. If there was a go-around they’re not going to delay it for another hour because “the sequence is already built”, likewise if the DLH had actually declared an emergency you’d like to think it would have been given a fairly expedient approach. One of the comments in the video you reference was: “We aren’t going to vector 40 aircraft and take them off the arrival, and assign speeds and headings and introduce more risk to those passengers and pilots in order to prioritize a special request.”

Firstly, no ATCO I know would routinely describe any heading or speed to an aircraft as inherently dangerous/risky. Assigning safe headings and speeds is the job! Anyone with that attitude wouldn’t be given a radar license. Secondly, it highlights that a request for a published instrument approach is deemed a “special request”. Having to apply standard separation in Class B should never need to be deemed a special request. Ref. the PAL104: if I only accommodated requests that didn’t inconvenience me in any way, I wouldn’t expect to keep my job for long. If the ILS requirement is so special and the sequence is so set-in-stone you must know exactly where they can eventually fit and be able to give an accurate EAT. The reluctance to do this suggests to me they knew they should be fitting it in, but for whatever reason - be it personal capacity, airspace, ability or anything else - it didn’t happen. For that video to claim the unit was happy with how the controller handled it, given the complete butchering of keeping the delay updated, suggests this is culturally accepted.

I’m not a pilot so I can’t comment on whether the DLH SOP is sensible or required in the first place, which seems to be where most of the taking sides against the DLH seems to come from, nor add anything to that side of the debate. Once the requirement was stated (and as an aside, I wouldn’t consider passing 14000ft particularly late notice) I feel more qualified to contribute. I would add that usually with these situations, pilots take the side of the pilot and ATC take the side of ATC. From my reading of the comments, pretty much everyone on the ATC side is asking questions about how ATC handled it, and specifically why such a meal was made of something that should be straightforward.

My 'siding' is merely an observation having operated in the US for a couple of years (SoCal and a little NorCal), the national airspace system is very flexible (especially for GA) but not good for anyone that underperforms. Some ATC could be assholes and I think it came down to workload and being close to maxed out. Many a times you switched freq but there was no opportunity to check in, just wait until being called. I can sympathize with ATC not wanting to explain everything on air. (Though as a pilot I would be very glad for an expected delay, even if unnecessairly conservative.)

Then there is the "maintain visual separation" which is poorly understood by those who haven't operated there. On IFR practice runs we often had to cancel IFR to relieve ATC of some separation responsibilities, yet they would treat us as IFR. It was just good cooperation. Had I insisted on staying IFR we wouldn't get the job done. In this case Lufthansa wouldn't exactly cancel IFR, but still help ATC out with separation.

My use of “sides” is far too black and white for something that is obviously a lot more nuanced so appreciate that sentiment. Separation is a pretty defining characteristic of air traffic, and when it comes to IFR traffic is very much an ATC responsibility inside controlled airspace. If you can delegate that to pilots to keep things moving at times, great. If the whole thing falls apart if a single pilot says no, not great. It is not the pilots responsibility. If saying no results in that situation, it’s bad airspace/sectorisation/training/procedure design. We’re going to declare the airspace Class B and publish instrument procedures, but don’t dare actually ask us to provide the level of service that implies.