I get that and I’ve seen that bit of that video (as an aside, a 25 minute video for what looks like 2 notepad screens worth of text is a very painful experience), but people are saying it halves the landing rate. You definitely lose one gap, then you add on potentially an extra minute to each runway on top of that, but that’s massively different from halving the landing rate. If you were landing say 50/hr beforehand, you might now end up landing 47/hr at worst.

The bit that most stood out to me from that video was how it apparently adds risk to passengers and pilots to break them off an arrival to assign headings and speeds. It really is a different world…

Still my biggest issue having struggled to the end of that video: if it was that busy - and I’m sure it was - why was the DLH not just given a 60 minute delay at the outset so they could choose to divert straight away. To reiterate they first stated they required the ILS passing 13.5 at ~0342, it was 0422 when they were given the final “another 10-15 minutes”. Why were they being strung along for so long without any accurate delay information, on what are apparently risky vectors, rather than just let them make the obvious decision to divert to Oakland straight away?

My big takeaway is that ATC prefer LOTS of warning. The ‘charted visuals’ are broadcast on the ATIS, so DLH had lots of time to prepare/make their request. Yes - would have been better to sit up in the hold at 25K, back up the STAR. In fairness though, I wouldn’t expect them to know the ramifications of requesting the ILS. Nonetheless, a review of their decision as fuel was depleting might have prompted them to cancel the ILS request & fit in with everyone else. But there was likely a big dose of plan continuation bias, esp. after getting arsey with ATC.

It wasn’t just a request though, they required it, they were categorically not allowed to be responsible for visual separation at night as per their company SOPs. Whatever the merits or lack there of for that procedure, I’m sure if it wasn’t important and enforced they would have just done the visual.

The SOPs are a framework- they’re great 99% of the time but they can’t possibly account for every possible scenario. I wouldn’t have any problem defending to the CP why I willingly violated ‘no visual traffic separation at night’ to avoid a fuel emergency & diversion, especially when I couldn’t get a straight answer from ATC. Risk management is what they pay us the big bucks for.

We’re Monday Morning Quarterbacking this to death. Maybe there were other extenuating circumstances we don’t know about? Good discussion.

Question for anyone.
I'm wondering when the flight plan was filed by Lufthansa would they have filed for IFR complete with any ILS landing at SFO??

I have had somewhat different experiences, in that some places won’t let you descend until you’ve called visual. The coping mechanism (see my post earlier) is to call visual, even if you’re not, and follow the IFR profile, so at least you’re safe in that respect. I’ve often wondered what would happen if you called their bluff and went sailing over the airfield at 4/5,000’...
Agreed, but even though I’m not qualified as a controller, I think I could slot someone in given 15 minutes to find (or create) a gap. In this case it seems to be almost a “revenge” delay and unhelpfulness. Which begs the age old question: “am I up here because you’re down there, or are you down there because I’m up here?”...

You might be able to find a slot, but at the expense of another arrival who now takes delay vectors /hold to open up the spacing.

As to calling the field, I’m in SFO monthly and have never been pressured, either I do or I don’t have the field in sight. Respectfully, I suggest you rethink your decision to report the field in sight when it’s not.

I think you are not grasping that at SFO the runways are so close together that if shooting Instrument approaches it’s essentially a single runway operation. If doing visuals it’s a two runway operation. With 750 feet of lateral separation from the centerlines you don’t gain anything or have the ability to stagger traffic to increase arrival rate. Here is a quote from the airport.

SFO operates on two sets of parallel runways. On fair weather days, SFO can accommodate approximately 60 arrivals per hour. During periods of low visibility, current FAA safety regulations allow aircraft to arrive side-by-side only if runways are at least 4,300 feet apart. SFO’s runways are only 750 feet apart, so aircraft must arrive single-file. This reduces the airport’s arrival rate to approximately 30 per hour, requiring the FAA to implement a Ground Delay Program to meter flights into SFO.

Again, I’m in two minds over this one, but having operated a fair bit into SFO (not once a month, though) and in some less-than-perfect weather, I haven’t experienced delays even when everyone has to do the ILS due to the cloud ceiling/vis. That sort of hints that they could could do it “properly” if they wanted to but can’t be bothered.

Personally, I don’t call visual unless it fulfils the criteria I operate to; I was offering an explanation as to maybe that’s why I have experienced other aircraft calling visual when I, in the same position just before/after, see no possibility of doing so...

But only 1 aircraft is doing an ILS, the rest are still visual, unless there’s something that forces the aircraft 10 behind in the sequence to do the ILS just because someone else is? In this case it’s single runway for the DLH gap-pair, not every gap-pair for an hour? Even if they lose both the side-by-side gap for the DLH, and as it’s a heavy let’s say the gap-pair behind too, that’s only 3 gaps lost so you’re still landing 57/hr (NOT 30).

You wouldn’t experience a delay, because in those situations, the widebody coming in from FRA will be given the priority.

The 737 coming from Denver will be held on the ground for hours, or tactically vectored around the country to achieve the required spacing if already flying…which leads to a ton of passengers missing their connections to Asia, and upset airline managers calling ATC.

So when ATC makes a reasonable request, they expect and anticipate you’ll be able to comply and not have your hands tied by a paper pusher.

Have you seen the second video? It’s a bit long, but it clearly explains what ATC was dealing with.

Good story
Sometimes one should just "be the Captain"

​​​​​​​What on earth are you talking about??

FullWings said he hadn’t experienced delays when having to shoot the ILS into SFO, and suggested ATC was just being lazy on the day I’m question. I was simply explaining why he’d be less likely to see a delay.

His location is given as the U.K., so I’m assuming he’d be going into SFO from Europe.

And the 737 being held on the ground for hours? What's that all about?

If a U.S. crew operating into a major European or U.K. airport during peak times demanded an approach no one else was using that would slow the flow down then swore at the controller for the inevitable delay they received there would be quite a different reaction on the subject !

I’m assuming that it’s the one that explains it with a diagram like this. A normal 10-min traffic sequence would look like this:
.x….x….x….x….x…
..x….x….x….x….x.

To land the DLH would look like this:
.x….x….…x….x..
​..x…..……..x….x.

A halving of the landing rate would look like this:
.x….x….x….x….x..
​​​​​…………………………..

Or alternatively this;
.x……..x……..x…
…….x……..x……..

Doing full instrument approaches for every aircraft halves the landing rate, making an individual gap for an individual instrument aircraft should not halve the landing rate, and I didn’t see anything in the video to suggest otherwise. Sure, 3 gaps in a busy sequence is still a hit, but it’s infinitely different to drop to 57 planes in an hour, not the 30 planes in an hour that guy was claiming.

Wake turbulence separation is applied by time rather than distance. Minimum radar separation for non vortex pairs remains 2.5nm but the next step is to apply non vortex spacing via time also (possibly next year.)

LHR parallel runways are 1400 metres apart (so still not separated but much further apart than SFO), a landing rate of 50 is achievable if they are both used for landing ILS traffic staggered on parallel runways. 2 mile slant radar separation is the minimum and 2.5nm or vortex minimum (time) in trail.

If there’s a stream of arrivals on the left runway it’s efficient to vector arrivals for the right runway between the pairs of arrivals on the left.

If SFO are declaring 30 arrivals per runway that implies a spacing of about 5.8 miles per runway, should be enough room to get one on the ILS to the other runway in that gap. Even if you stretch it to 7 or 8 miles to accommodate the Lufthansa then you’ve still only lost a couple of miles.

But regardless of how easy it would have been to accommodate the DLH, you can’t give a delay of 10 minutes and then not update it until it’s expired only to then double the delay when queried.

**** is barely a swear word to Europeans whose second language is English. They are exposed to it so much in culture, music, Tv, films, etc. I’ve walked round supermarkets in Europe to the sound of “**** you” by CeLo Green, or been at kids sports days where the explicit version of a song in English is played, it’s completely normalised. If you watch F1 you’ll see a pattern of the European drivers saying ‘****’ almost all the time but the native English speakers never say it. It’s a cultural difference.
I still don’t think there’s any place for it on the RT but a German pilot would not consider it as strong a word as US or British crews and the DLH pilot was definitely not swearing at the controller.

That's why he doesn't get the delay...

That's clearly ridiculous. You obviously do have no idea. Pretty obvious though from your previous posts.

He most certainly was


So because the f word is used in music and TV it’s ok to use it to pressure and insult a controller with it if you’re not getting your way ?


What nonsense, almost as much as the idea Germans don’t know what the word means


A tip for this particular pilot in future that he seems to need is when asking for special consideration from a controller just be polite

You and I aren't getting anywhere with each other. Let's agree to disagree.

I’ve had a British pilot describe an RNAV approach as “!!!!” on the RT before. He wasn’t swearing AT me, not insulting me, just using it as an adverb.

I never said Germans don’t know what the F word means, just that it doesn’t carry nearly so much weight to someone who speaks English as a second language. I provided examples too and I’ve experienced it many times in Europe.

Im just trying to provide some context so views don’t get entrenched on the swearing issue. (Not sure how that’s going….😆)

To me, this looks like a fundamental Lufty flight planning weakness.

This mode of operation is pretty standard at SFO. If you know this, which the company does, plus that you'll be landing in darkness, you can understand sat in Munich way in advance, it's not so hard to put 2 and 2 together and identify a hazrad, ie. that your company SOPs risk a holding delay. It's entirely foreseeable. In that case, you brief this eventuality, agree what to do, and load fuel for holding just like you would if a weather delay was foreseeable.

This isn't a one-off charter landing in a rarely served field, but an exceptionally routine daily flight. The fact Lufty's SOPs and planning couldn't identify a basic hazard in an exceptionally routine situation (MUN-SFO daily scheduled), like a delay causing a night landing in a mode you can't fly, is not a good advert for them.

In terms of ATC, what are they meant to do if there is a queue of planes stretching over the Rockies? At the outset, ATC advised of a lengthy delay and they were right. ATC was accommodating the request, but that will naturally take time. All of that was predictable back in Munich.

Lufty's Karen outcry (by the male) was pretty pathetic and showed a lack of situational awareness and understanding about the airspace they were in. I get the crew may not be frequent visitors, but landing at a major global hub at 1900, that has a heavily restricted runway layout, and expecting anything other than what they got is naive at best. Plus, Lufty has been flying to the US for 70?years, they do know full well the US can do things differently.

This topic is becoming more ridiculous as time goes on. Of course the system has to be able to accommodate changes in the sequence. Are you seriously suggesting that the sequence was set in stone when LH458 left Germany? Local jets comply with the GDP and internationals fill in the holes that were deliberately planned-for. To suggest that an extra 2nm couldn't be inserted before the LH is just nonsense. It wasn't as though the LH was a new entrant at 40nm to run.

One would have thought that SFO, of ALL places, would have understood the value of ILS. One recent very near miss which had lined up on the taxyway during a visual night approach, one hit the boundary wall and was destroyed during a visual approach ... does nobody there care ? I wouldn't be surprised if the LH procedure was actually as a result of seeing such events.

Away with ICAO standard methods and SOPs then, here it's a FIFO procedure - "Fit In or F*** Off".

Now, don't be silly; of course the sequence isn't predictable, so what? The probability of this situation is predictable.

It is totally predictable that you are very likely picking up a nasty delay if landing during peak time at a busy operationally constrained airport when you know your company rules mean fitting in with ATC's standard procedure is a no no.

I don't know how familiar you are with SFO. It isn't only a matter of sticking in an extra 2nm on one stack, that should be easy enough. At SFO you basically lose an entire landing slot on the parallel runway, and no ATC is going along with that at peak landing time with aircraft already stacked up to capacity on both runways 100nm out. No ATC is going to send somebody around to accommodate this request, which is what would be needed in these predictable circumstances, at a constrained and busy airport during peak times.

Lufty's constraint is apparently about visual seperation of aircraft at night, and without losing a slot on the parallel runway, there is no way around this with SFOs set up.

Anticipate and plan for a predictable lengthy delay.

Peak time? Hardly.
https://cimg2.ibsrv.net/gimg/pprune....d0541c2bde.png

I guess SFO can't cope with a missed approach or two and slotting them back in to the sequence...

What a useless place.

To be fair to NORCAL I think they could cope but according to some folks on here asking for an ILS and saying no can do to maintaining visual separation at night is being awkward!

I must have been lucky all the times I’ve been there and flown the ILS without any drama. Only time I’ve held was in quite atrocious weather landing easterly. It was on TAF so we had the fuel. I think there’s much more to this episode than any of us can know.

Probably a good idea. Repeatedly discounting and deflecting away from the particulars of the approach cited by those who have routinely flown it, and instead continue to push the idea you believe it's inherently unsafe because of what you see on a chart and assume goes on, or to impugn the ability ATC to "handle traffic" as if there's no reason for this approach-type's existence, seems rooted in a starting position of never having actually flown it.

Seemed that way from the beginning with your ":rolleyes:" at my statement mentioning that TCAS assists with spotting the aircraft you will eventually be pairing-up with on the parallel visual approach before ATC calls it out and asks to verify in sight and maintain separation. It helps identify the stream for the parallel runway and/or gaps into which someone is being vectored that will eventually be on an opposing dogleg, base, or combination and wind up about 500' off your wingtip. If you deny that TCAS helps with SA as it relates to one's place in the stream relative to others and picking them up visually, then I don't know what to say to remedy that, and it's not as if SFO doesn't share the airspace with Oakland and San Jose either.

Everyone routinely flying in and out of SFO knows the form-up into pairs on final is the critical juncture due to the possibility of an overshoot by either aircraft, and eyes need to be kept on the traffic you've verified to ATC is in sight. Everyone knows to tune the respective ILS freq with raw data displayed as a SA back-up and for eventual use within the last few miles when lining-up from the slight offset. Everyone who's done it knows there is no fancy flying required for this visual, it's benign as it get in that respect. What everyone knows about SFO with centerlnes 750' apart, is vigilance with respect to other aircraft when parallel approaches are being conducted no matter the approach type.

Long before they form inbounds up in pairs on final, the in-trail spacing in each stream has already been established by ATC using assigned speeds, The "maintain visual separation" instruction does not cancel one's speed assignment. If it did and everyone's speed control was left to their own devices the entire exercise would be pointless as streams would become giant Slinkies. Speed assignments are usually maintained until 5 miles out when aircraft can dirty up and be stabilised by 1000'. Again, routine speed control stuff that mirrors what you'll find on an ILSs worldwide.

Those familiar also realize the speed-controlled visual approach pairings with slight in-trail & vertical stagger instead of the 1 to 1.5 miles spacing using parallel ILSs facilitate more arrivals and create the gaps that allow those pesky little things called "departures" to take place with greater frequency. Also, if the traffic you're following blows his speed assignment by slowing too early TCAS is the first indicator this will be easily be observed and, depending on the aircraft types and wind conditions, whether an adjustment of glide path to avoid wake turbulence is advisable.

And to reiterate: You can fly to the US and refuse the Visuals because you believe them unsafe and wait to be worked-in for an ILS. Swell. Good for you, but if they're conducting visual approaches you're in VMC conditions, and thus the FAA mandate that "vigilance shall be maintained" by the pilot in order to see and avoid other aircraft applies. That's a "shall", and while this may come as a surprise and create despair for those from where every minute spent on an IFR fight plan flight is logged as "Instrument Time" regardless of flight conditions, in FAA-land flight conditions matter when it.comes the pilot's responsibility to look-for and separate oneself from other traffic by avoiding if necessary. TCAS/ADS-B/IFR fight plan/established on a STAR/being vectored/ inside Class B airspace/established an ILS approach..none of the aforementioned negates the rule to maintain vigilance to see and avoid, The fact that ATC is responsible for traffic separation of all traffic in Class B doesn't erase the pilot's mandated responsibility in VMC. Neither does being cleared for an instrument approach.

Pretend SFO happily works you in and clears you for an ILS 28R on a VMC night you refused the visual approach. Your responsibility under FAA regs remains the same; to see and maintain the surveillance of an aircraft 500' to your right and ahead 1 mile on the parallel approach as well as the traffic you're in-trail of to the same runway. If the traffic ahead happens to be a Lufthansa A380 flown by a Captain who decides to let his party animal side hang out by accepting the Visual approach, are you not going to monitor your spacing in terms of distance and vertical path?

If you feel safe in your ILS bubble and erroneously believe looking outside for and at traffic is too difficult or beyond your scope of responsibility, I give you Cathay 892 at SFO in 2019 in VMC conditions, already at an assigned altitude and speed being vectored on a left downwind for Rwy 28L, planning to intercept the LOC. When Cathay was given a base turn heading 010 and speed 180 they acknowledged, but 30 seconds later they still hadn't begun their turn and slowed only 10 knots. When ATC queried this re the turn Cathay acknowledged but requested the heading assignment again because someone must not have bugged 010 the first time they read it back. Due to Cathay's non-turn, ATC had to sharpen if off downwind and assigned them left to 330. 30 seconds after ATC assigned 330 they turned them left further to a heading of 310 degrees and cleared them to intercept the 28L LOC. Cathay acknowledged and read back the heading assignment and clearance to intercept.

30 seconds after receiving the clearance but prior to the intercept ATC called-out to Cathay the traffic on approach to the parallel runway 28R; United at 3 o'clock, 2 miles, 800' higher and asked Cathay if they had them in sight. Cathay responded they had United in sight. ATC then asked Cathay to maintain visual separation and descend to 3000', which Cathay needed repeating. When repeated, Cathay read back the full clearance including the directive to maintain visual separation from United. At this point the flight paths of the 2 aircraft were gently converging. When they were 1 mile apart with Cathay yet to intercept the LOC, United was to their right, 500' higher which put them in clear view from Cathay's cockpit window. In clear view, that is, if anyone had still been keeping an eye on them as they had acknowledged to ATC mere seconds before they would, and as required by FAA regs.

But despite the clearance to intercept the LOC, and despite their acknowledgement that their traffic was in sight, and despite the associated acknowledgement to maintain visual surveillance in order to ensure separation, Cathay still blew through their LOC course before overtaking to pass 700' underneath United who, by then, had arrested their decent at 3900' and confirmed the traffic call by ATC when it came. United levelled-off and maintained their altitude until Cathay, 1000 below them, had turned left to point their nose at the correct runway.

A couple points; Notice that Cathy's clearance to intercept the LOC and their acknowledgment to do so did absolutely nothing to ensure they didn't blow through it. Also note; despite verifying their traffic was in sight and acknowledgment of the subsequent instruction to maintain visual separation from that traffic and in contravention of 91.113 re maintaining vigilance to see and avoid plus the right of way rule pertaining to overtaking another aircraft, they wound up right underneath United and following an RA. Whooda thunk it.

Now I don't know how many sets of eyes and ears were in that Cathay cockpit, 2 for sure, probably more. After they verified United was in sight, at no time would it have ever not been in sight unless nobody was looking outside. What is patently obvious is that nobody was, even for traffic they knew they were converging on. That requires some serious heads-down flying. Even if they tuned the wrong LOC freq they should have never gotten to the point they wound up under United, and they wouldn't have had they been complying with their clearance and regs in VMC conditions.

As for moaning about SFO ATC not being flexible, the Controller with the patience of that guy who's famous for being patient, instead of right then sending Cathay to go enjoy the view of fabulous Oakland, instead asked him their intentions. When that wasn't forthcoming ATC offered Cathay a visual approach, which they accepted. The Controller..in slow, clear language...then cleared Cathay for the visual approach to 28L, directed them to reduce speed ASAP to the slowest practical, and (again) to maintain visual separation behind (that same United) traffic ahead with the reminder that United was going for 28R and they were cleared for the Left. There was nothing left on the pallet for him to try and paint for Cathay in words the clearest possible SA picture they hadn't put together for themselves up until that point. Cathay accepted the clearance to continue on a visual for 28L. Only 20 seconds later, to give Cathay even more opportunity to salvage their cluster of an approach and hand him off, the Controller approved them for an S-turn if needed to stay behind United still high to their right and directed them to contact the Tower. Meanwhile, United had already been handed-off to Tower but was still maintaining 1000' above Cathay.

Cathay didn't switch to Tower, however, because it seems the cockpit fog still hadn't cleared despite moments earlier accepting the new visual approach clearance. Upon being asked to switch to Tower freq Cathay responded by asking Appch to "confirm we're cleared for the ILS 28L?" Wow, after all that mucking-about that guy's still thinking "ILS, must have the ILS".

At that juncture, realizing that acknowledgements and readbacks by this particular gang were as reliable as one-night stand promises in Wan Chai, the Controller finally punched Cathay's AMF button and grabbed big binos to scan for the smoke trail being left by all the helmet fires going on inside that A350 cockpit. Meanwhile on the RH side tower freq United was still at 3100', 1500' above Cathay and by now really wanting descend if not for the RA he received, asking Tower what Cathay is doing behind him rather than immediately accepting Tower's landing clearance. Tower points him out at his 8:30 low position and then advises United that Cathay is being broken off his approach.

What I'm wondering is how many on that crew were dead-set against visual approaches, deeming them unsafe, although I'm not sure even the word "Visual" would nudge them into accepting the fact that windscreens are made clear for a reason other than to look up at 200'. .

I have no experience with European or Oz SOPs, but I've never come across a well-written set that doesn't set forth in the preamble some reference to there being an expectation of exercising good airmanship at all times, and in this way good airmanship becomes part of the SOP. Maintaining vigilance to see and avoid other aircraft when conditions permit falls squarely into that directive in most people's thinking, and I can think of more places than just Dubai or Hong Kong where at times it seems few F's are given during arrival/approach sequencing re wake turbulence avoidance in the vertical where they should know better: Descend that Triple down through my altitude? No worries mate don't trouble yourself to point him out. I saw him and..whoa hey look at the size of that giant cargo ship hauling Happy Meal toys to America.

And last time I checked, IFR flight plans and Instrument approaches are just about useless as forms of insurance against un-announced airborne surprises if one is flying an approach into Mopti, departing out of Khartoum, or even just trucking down a high altitude airway between Cairo and Luanda etc etc ad infinitum. And while your snarky little emojis probably play well in certain circles, if you aren't using TCAS to help pick out airborne traffic more easily so as to allow you more time to pick up targets of the type that may not show up, then good luck. But if you didn't like that, you're gonna hate this: I maintain that FLIR/HUD is also a useful tool to help picking out traffic visually, especially at night...how many :rolleyes:s does that earn me? No matter, I'll just keep using the tools at my disposal until someone give me a good reason I shouldn't. Still waiting.

Which reminds me, you've been asked how much you've flown into SFO and experienced these approaches, most likely because you seem to feel so strongly about them and express yourself with such authority. But I'm not seeing much that indicates familiarity. Any answer yet?

As an aside, this occurrence reminded me of something I encountered app 12 years ago.

Back then, I was on leave from my regular outfit, and flew in Dubai for 2 years.
Traveling as pax on Lufthansa from FRA to DXB, as we approached DXB we got put in a hold (as happens every single time approaching DXB at that time of day). Didn't think much of it until the CDR came on the PA stating we would be diverting to AUH (Abu Dhabi) and refuel....

I thought, wow... LH flies every day, same time of day, FRA-DXB, and being an A340-600 on a 5.5 hr flight, why not bring x amount of extra fuel to accommodate for the "guaranteed" hold arriving into DXB??? I am not pointing fingers at LH, but to me it is strange that you don't bring extra fuel into high density airports where you are familiar, thus knowing there are going to be delays..

Sorry for the drift..

As someone who operates regularly into SFO; in my experience they control you onto base leg then... "cleared for the visual RW xx"...call visual and treat it as an ILS. The only thing that will change is the go around.

Anyone know what spacing they use at SFO?

A landing rate of 30 per hour per runway would imply an average of 5 and half mile spacing. 6 or 7 miles would be enough to fit in aircraft that couldn’t accept visual separation.
What’s wrong with adding a mile or two on one gap to accommodate the DLH? You wouldn’t lose a whole movement on the other runway although both sequences would be a mile or two further back.
Just remember to position the DLH in the middle of the gap rather than on the shoulder of the preceding so there maybe one departure movement lost.

Eg. An A380 followed by an A320 both on 27R. They probably give 6 or 7 miles there for the vortex. Vector the DLH on 27L in that gap and you’ll have over 3 miles radar separation from them both. No loss of movement or increase in delays.

It was 10 years ago but I remember a pilot bulletin to the effect that Norcal would not be issuing KSFO visual approaches to foreign carriers for the time being, so expect delay

Overreaction, but the disaster occurred in perfect day VFR conditions to a 11,000 ft runway

Great post Pukin :ok:

Hear,hear to megan concerning the post from PukinDog. :ok:

Sounds like everyone is entrenched.in its own beliefs and no-one is going to convince the other side. Signs of the times,
I normally here always try to defend ATC and give background on the why, but here, it is a bit the other way around, an ATPL from any airline , and even more an established one with a quite good safety record as LH, has to follow its employer SOPs. So he had to refuse to maintain visual separation at night. if his SOP said so

ATC is a before all, a service to pilots, to keep them safe according to established standards not a commercial entity pushed to increase numbers to make profit for the airlines or the airports. . Accommodating pilots requests is our job. ATC can propose never force a pilot to do something out of the standard. Here ATC should have accommodated him and not force him to divert.
On the : delaying 12 others to accommodate one : It is still all over the world for ATC : " first come first served". We wanted years ago to have this changed to : " best equipped, best served" but this was fiercely opposed by ..the USA under the pressure from its large domestic airlines that were still operating Jurassic jets. and opposed to any retrofit costs. I do not think this has changed.
So you get these own separation visuals out of the standards instead of RNPs...

Last remark from PunkinDog on TCAS :
Both IFALPA and us fought hard on the ICAO FANS in the 90s.to keep TCAS where it belongs , a last minute anti collision system . It is not a separation tool and due to its poor azimuth definition ( +/- 11 degrees) it cannot be used for visual Aquisition. In my days when we issued a conditional clearance involving another aircraft and passed traffic info and the pilot replied, " we've got it on TCAS" we always replied, " fine ,but I need you to identifying visually this one ,by looking out of the window.
Remember that the one you see on TCAS might not be the one I am talking about.

Bending the rules because it seems to work is really not sound safety management.
Just read the transcript of the NTSB Chairwoman , Jennifer Homendy on the US senate ,relating to ATC understaffing and fatigue , Not pretty . An not a good idea to bend the rules when you are in this situation .

But ATC Watcher, isn't one of PukinDog's main points that the Luft. captain could have complied with the visual separation task despite the SOP against visual approach?

I'm saying this just from point-of-view of the traveling public: while the ATC system does indeed provide service to flights, in this situation, the Luft. captain had responsibility for just his flight operation, while the ATCOs had a goodly number of flight operations arranged and in sequence for approach and landing onto runways presenting some challenge due to their lateral proximity. To the traveling public, it looks like a situation in which controllers had some sizable degree of discretion. So - even if they could have re-arranged the line-up of flight operations in sequence, that does not mean they were required to do such re-arrangimg, or even that they should have done so. I wonder, were the inaccuracies of the estimated delay times a reflection of heavy workload, other stressors in the traffic? . . . it does not sound like the changing estimates of delay time were at all deliberate. But the larger point is, doesn't ATC have discretion about whether to "work someone into the existing sequence" or not. (I'm not apologizing for my ignorance of proper terminology, although as SLF/attorney I obviously recognize it indeed is ignorance on my part.)

By the way, NTSB Chair tells a Senate Committee to worry all over again about inadequate controller staffing levels and prevalence of fatigue issues. Does that not support the need to acknowledge significant controller discretion in handling exactly this kind of situation?