Lufty at SFO

[ATC Watcher]

PukinDog

Most likely some ,if not all, of the LH crew had flown the very same Visuals during the daytime (when they normally arrive when on schedule) and were in no way surprised they were in use. They were handcuffed by their SOPs. Bummer.

Yes that is what my (FRA) LH contacts on the 747s are saying too... and when they exceptionally came at night it has never been an issue before to get an ILS. I suppose the same is valid on the 350s fleet out of MUC. Looks like a combination of wrong timing ,plus fatigue and a bit of wrong egos on both sides of the mike..

[maxrpm]

The LH flight comes in at 13:00 local time. This particular flight was very much delayed thus arriving a night

[YRP]

Originally Posted by ATC Watcher

Yes that is what my (FRA) LH contacts on the 747s are saying too... and when they exceptionally came at night it has never been an issue before to get an ILS. I suppose the same is valid on the 350s fleet out of MUC. Looks like a combination of wrong timing ,plus fatigue and a bit of wrong egos on both sides of the mike..

Any word on whether they are going to change / clarify the SOPs to avoid this issue in the future?

[PukinDog]

Originally Posted by ATC Watcher

PukinDog
Yes that is what my (FRA) LH contacts on the 747s are saying too... and when they exceptionally came at night it has never been an issue before to get an ILS. I suppose the same is valid on the 350s fleet out of MUC. Looks like a combination of wrong timing ,plus fatigue and a bit of wrong egos on both sides of the mike..

Agreed, and since the discussion has a branch that's gone something like: LH diverts ---> LH off-schedule ---> Timing issue ---> Inability to work LH in for an ILS ---> LHR vs SFO traffic number debate.....it follows there could be a further, natural progression into: ---> IATA Level 3 slot-governing for all aircraft vs Level 2 voluntary, airline schedule-facilitating airports ----> Fundamental differences in Air Traffic Management philosophy and why...,..which ultimately leads us to...---> Which side of the Atlantic has the better Philosophy?

Now, although I was highly-impressed and pleased that Mustang was able to so casually work an apropos John Locke quote into this thread (which unfortunately been removed, it seems) , in the interests of avoiding upping the ante from LHR/SFO traffic count numbers to full-blown, epistemological willy-waving, I won't bring the IATA thing into it.

(Of course, if someone really wants to, I'm game).

[Request Orbit]

As I said at the time when stating the numbers, I don’t really feel it’s relevant to the point at all for the most part. What is of more relevance, at least to me, was the absolute belief the two original posters had that they were correct. Even when stated the numbers from the places they’d requested, they still refused to believe them or alter their position. This then leads me to take anything else they say with an absolute mines worth of salt, which is a pity as it appears at least one of them had a lot of first-hand, relevant info to add to the discussion. There may be an interesting tangential discussion to be had about why they’re so convinced SFO is much busier.

Originally Posted by ATC Watcher

Yes that is what my (FRA) LH contacts on the 747s are saying too... and when they exceptionally came at night it has never been an issue before to get an ILS.

That is of far more interest to me. It explains why the DLH pilot was so baffled that he was still being vectored around 40 minutes after first confirming they were unable to apply visual separation. If it was never an issue before why was it this time?

[PukinDog]

Originally Posted by Request Orbit

As I said at the time when stating the numbers, I don’t really feel it’s relevant to the point at all for the most part. What is of more relevance, at least to me, was the absolute belief the two original posters had that they were correct. Even when stated the numbers from the places they’d requested, they still refused to believe them or alter their position. This then leads me to take anything else they say with an absolute mines worth of salt, which is a pity as it appears at least one of them had a lot of first-hand, relevant info to add to the discussion. There may be an interesting tangential discussion to be had about why they’re so convinced SFO is much busier.



That is of far more interest to me. It explains why the DLH pilot was so baffled that he was still being vectored around 40 minutes after first confirming they were unable to apply visual separation. If it was never an issue before why was it this time?

No telling, but the traffic situation can and does change daily/nightly. There's nothing in the video that paints the larger picture but it's 100% certain it differed from previous LH flights because even the previous night would have been different. There could've been a stream as far east as Salt Lake City, or perhaps the other LH aircraft gave far more advance notice, or arrived at a different time after sundown, or a different day of the week.

The entire comparison between LHR and SFO or any other European airport as it pertains to "the ability to handle traffic" just because they have parallel runways is silly because LHR is slot-governed for all aircraft arrivals like almost every other European airport is slot-governed for all aircraft arrivals. Unlike the US, the airways of Europe are also essentially slot-governed. The US has only 1 slot-governed airport that any international flight has to deal with, JFK, and that's only the 9th busiest airport in the US and world. At their busiest, LHR is comparable to JFK (movements, widely-spaced parallels, IATA Level 3). The US doesn't even cap the traffic for airports 1-8, so they don't for SFO either. Point is, the level of congestion can change a great deal day-t0-day, week-to-week, month-to-month.

Because of this, you certainly can't take previous LH aircrafts' ability to get worked-in for an ILS as a historical guide to assume there should have been an easy way to do so that night.

[Request Orbit]

The ability to make a gap to work them into the sequence for an ILS is a very similar task to fitting in a go around or dealing with an emergency - or at least it should be. Unless you’re also suggesting there would have been no capacity to deal with either of those events that night, I don’t personally feel that stands up.

[Del Prado]

I still struggle to see why they couldn’t have given the Lufthansa a 3 mile separation ‘bubble’ by increasing the gap on one runway by a mile and vectoring the Lufthansa in the middle of that gap for the other runway.

Tower would have lost a departure gap but after (what would have become) a 20 minute delay for Lufthansa that would seem like a fair compromise.

Having a “queue of aircraft stretching all the way back to the rockies” doesn’t cut it. There must be some flexibility to fit in go arounds at least.

[Request Orbit]

Originally Posted by Del Prado

There must be some flexibility to fit in go arounds at least.

Especially since according to FAA 7110.308E (12,f) if one of a pair goes around then situationally both have to go-around, they should be used to dealing with go-arounds in pairs. If there’s a queue to the Rockies do they both just get sent to join the back?

[safetypee]

Not "how do we do this"; Ask "should we be doing this"

Re Normalisation of Deviance
Whom or what is deviant; - a range of viewpoints, each assessing deviation from the others', or that the overall operational system deviates from best practice - everyone is deviant by acceptance.
ATC, Airport, using procedures where the level of risk is higher than the industry norm, particularly at night.
The Operator with specific procedures to reduce risk in particular situations, but which conflicts with local norms.
The regulator who approves and oversees the risks in operations.

Consider 'what if', after the accident:-
Visual procedures, but with FMS waypoints.
Visual self separation;
The daytime 'follow the aircraft type/operator' assumes that all operators have common knowledge, aircraft recognition, even that the colour scheme matches the locally known ATC information of operator/call sign.
At night, all that might be seen are the aircraft lights, against a bright cityscape background; thus how is 'the aircraft' visually identified as the one to follow (and for the ACAS lovers, the emphasis is on visual, not being headdown multitasking reducing lookout time).
What is the procedure if the lead aircraft goes around - your clearance was to follow the aircraft; any alternative assumes good communications / specific procedures.
The risk of suffering the black-hole illusion could be higher; bright cityscape background with the approach over water.

… ‘well-intentioned people in dysfunctional organisations’ (systems) …

The emergence of confused consensus
The art of muddling through
Law of continuous bodges

[WillowRun 6-3]

Well, in the context of apparent conclusions to be drawn from an understanding of the occurrence, will it not be the case that some level of administrative authority within FAA will either want to address, or be required to address even if it doesn't want to do so, reluctance on the part of the ATCO primarily involved to deal with the requests made by the Lufthansa flight?

I'm asserting (despite mere SLF/atty guest-status here) that an accurate and professionally-spoken estimate of delay time should have been given. Why did this not occur? Was the controller primarily involved having an understandable slip from otherwise usual and expected proficiency?; (and many reasons why someone might not score 100 out of 100 in a given instance - notice, I'm not speculating as to any reason specifically). If not, then, why was what the apparent consensus here holds was a nominal estimate of delay time not given?

Yes, despite mere SLF/atty guest-status here I'm seeing an apparent consensus that (i) working the LH flight into the stream of arrivals could indeed have been done without screaming or freaking out, and (ii) although as PukinDog has articulated keeping visual contact with identified traffic is a nominal requirement, LH has perhaps peculiar content in its SOPs, but this doesn't justify fumbling delay time estimation. I get it, f-bomb is unprofessional but an unnecessary diversion apparently was not justified by one word or one R/T with unpro tone. Blast holes in my statements here, if they deserve it. I'm only, only trying to direct a bright light on what FAA has to do NOT to apologize, just to do better goin' forward.

[txl]

Just stumbled on this thread after this little kerfuffle is now making the rounds of German aviation outlets. Loved the story about this LH crew that refused to pass the broken signal even if instructed to do so by ATC. What do you expect from people who'll stop at a red pedestrian light at 2 AM with no traffic in sight anywhere? This is Germany. There are rules.

But I was wondering if Lufthansa (or any operator's) SOP is cast in stone or did LH458 have some wiggle room to bend the rules a bit? "Ja, ja, visual separation is verboten, but we'll just pretend we didn't do it." I suppose the beancounters back home really appreciate diverting to Oakland and having to ferry the ship to SFO later. Will it be tea and biscuits for the LH crew anyway? Or would bending the SOPs a litte get them in serious trouble?

[lederhosen]

Good post from WillowRun! I suspect the LH captain had an easier job than the controller explaining to management how he adhered to SOPs. If ATC was ultimately held responsible for a material financial loss it would be interesting to know what ramifications that would have. Probably unlikely, but it is an interesting thought experiment.

[PukinDog]

Originally Posted by Request Orbit

Especially since according to FAA 7110.308E (12,f) if one of a pair goes around then situationally both have to go-around, they should be used to dealing with go-arounds in pairs. If there’s a queue to the Rockies do they both just get sent to join the back?

What you've linked isn't applicable to the night LH diverted, plus you've assumed dual go-arounds for paired aircraft are common at SFO which they are absolutely not, and ignored the reduced congestion levels that will be encountered by SFO inbounds when it's IMC conducting Simultaneous Dependent Approaches due to SFO-bound aircraft elsewhere being held on the ground, thus ensuring spacing is available for go-arounds to be worked back in within a reasonable time.

The document in the link you provided applies to Simultaneous Dependent Approaches conducted during IMC (where the min Lead/Trail stagger is 1.0 mile min, if winds permit) not to the simultaneous, slight-stagger Visual approaches that were being conducted that night.

SFO "used to dealing with go-arounds in pairs"? No. If they're using Simultaneous Dependent Approaches they aren't going sequence aircraft in a way that results in possible dual go-around scenarios. That scenario is driven by Wake Category/Weight Class differences between a Lead aircraft of a pair and its trailer, and only if the Lead aircraft is the one to go around, and only if the Lead is a higher Wake/Weight than its trailer. At SFO, the Lead aircraft must be Cat F (737/A320 size) or lower, which precludes them from triggering a go-around for their trailer unless the trailer is a Regional Jet, most corporate jets, turboprop, light aircraft etc. Solution: sequence the RJs, Falcons, etc. to Lead that pair and the possibility of a dual go-around goes away. At SFO, Supers and Heavies will always be a trailer, or un-paired.

Dual Visuals result in the aircraft in slightly-staggered pairs, practically side-by-side, so wake in the event one goes-around is no factor to the other.

I don't know if these posters are the ones you're choosing to ignore due to the LHR/SFO numbers thing, but those by Sailvi767 ( 60 & 216) and Check Airman (131 & 203) have already explained exactly what occurs when SFO's arrival capacity due to prevailing IMC (or other) conditions is reduced and why they have space to re-sequence missed-approach/go-around traffic without undue delay; SFO-destined aircraft are ground-held at their departure airports and/or flights cancelled so the IFR arrival/approach (not the VFR) capacity of the SFO airspace is not exceeded. That IFR capacity limit factors-in spacing for events like go-arounds.missed approaches.

The need for slot-type governing for specific airports is determined and initiated by the ATCSCC in Washington DC.
https://www.faa.gov/about/office_org...nas_ops/atcscc.

The ATCSCC is tied in with all the air carrier Ops and the weather-guessers nationally, because most ground delay/stop restrictions are due to Sigmet-level weather which occurs with great regularity for 4 seasons in the CONUS. Due to the extra-close runways at SFO, however, the weather doesn't need to be that significant; just IMC combined with peak-hour traffic can do it.

If you doubt the veracity of those mentioned who've already cited the ground stop/cancellations as a system for metering traffic into specific airports and airspace when necessary and you're actually interested, here: https://www.faa.gov/air_traffic/publ...ection_10.html.

When those ground delays/cancellations/slot-times are in effect for a specific airport (usually due wx conditions during peak times), the affected, normally-IATA Level 2 US airport essentially becomes just like a UK/European always-IATA Level 3 airport. Not coincidently, with that transformation comes the flexibility LHR, AMS, and a hundred other European airports enjoy every day of the week by always keeping traffic on the ground elsewhere (even when VMC prevails). It's planned that way and is, in fact, the entire point.

And in no way is SFO unique for these ground-stop measures, Eastern US airports commonly experience this, usually due to large areas of convective weather associated with low-pressure cold fronts. When lines of Level 5/6 CBs stretching from Canada to Georgia are crawling east and starting to pinch all the airspace from BOS, JFK, LGA, EWR, PHL, IAD CLT, ATL, up against the coast, why do you suppose there's not a bunch of UK/Euro-inbounds holding out over the Atlantic or long streams of domestic inbounds continuing to flow into the ever-shrinking available airspace even when IMC conditions don't yet prevail? It's because they (ATCSCC) sees what's going to happen and they've initiated ground-stop/cancellation IATA 3-like slotting for those airports to ensure airspace is not saturated with the short-range stuff when they arrive.

The crew and pax of those International arrivals are oblivious to the dozens or hundreds of aircraft with thousands of domestic pax that are going to run sometimes hours late, miss their connections, or cancelled and re-booked. Alternatively, it's the presence of those aircraft that explains why an airport that's operating at it's VFR capacity can't just automatically slot-in an aircraft as if it's an IMC day...the airspace/approach streams are already more congested than would happen if actually IMC.

Why does the US not adopt a UK/Euro - style system of forcing all aircraft to obtain landing slots at every major airport on the continent? Because in their view that would under-utilize what the system is capable of, significantly so when you consider what it handles by reserving the airport slot-governing to where-and-when as needed instead of blanket restrictions across the continent. That's why the US had the top 9 busiest airports in the world last year, based on movements, which is not out of the ordinary. LHR and AMS cap their slots at about 500,000 movements per year and last summer for months LHR even capped their traffic based on passengers, not movements, which drove arrivals down.

The US and Eurocontrol have a similar airspace size (Us is a little smaller), but very different types of weather that can affect the airspace use. Here's a comparison of usage compiled by the Agencies for the US vs EC in 2017 (pre-COVID). A couple excerpts first:

Airspace (square km, millions). US: 10.4. EC: 11.5
Average daily flights: US - 41,874 EC: 28,475
IFR GA as a % of daily. US: 19%. EC: 3.5%.
Airports with ATC services: US: 517 EC: 406
Approach Control Facilities: US: 26. EC: 16
Enroute facilities: US: 20 EC: 62.

https://www.eurocontrol.int/sites/de...mance-2017.pdf

Here's a suggestion for that LH flight since their SOPs are written the way they are. If they're running late and know they're going to arrive after sundown and the forecast shows VMC conditions likely, make sure a message is passed to ATC a lot sooner than they did with NORCAL Appch, especially since it is obviously a Company operational restriction that could very well affect that flight.

Since it could be an operational restriction at the Destination for a late-running flight arriving into an airport they know conducts visuals during VMC, how about LH noting the restriction on the flight plan under "Other Information" in the Block 18. Perhaps a simple plain-language: "RMK/ UNABLE VIS APPCH would have helped, and back that up by letting the first appropriate controller, rather than the last, know of the restriction and need for an Instrument approach as noted on the Flight Plan.

Do you believe LH bears any responsibility for their situation? Could they have done nothing themselves to have prevented ATC's last minute notice? It seems that you're assuming they did, not unlike the assumption that SFO ATC would be used to conducting dual go-arounds when In fact they plan and sequence so they don't.

[Request Orbit]

For the first part of your post, I don’t say I didn’t believe a word they’d said (and you got one of the two wrong, although it appears there’s been either some self-deletion or moderation since). I said you can’t guarantee it’s accurate, which is very different to saying it’s all incorrect. I don’t doubt for one minute there’s ground stops when it’s IMC. I very much doubt they’re regularly seeing dual go arounds, but if they’ve not all had to deal with them in training at some point, I’d be surprised.

Originally Posted by PukinDog

Do you believe LH bears any responsibility for their situation? Could they have done nothing themselves to have prevented ATC's last minute notice? It seems that you're assuming they did, not unlike the assumption that SFO ATC would be used to conducting dual go-arounds when In fact they plan and sequence so they don't.

As both Del Prado and I have said on multiple occasions, and as the DLH pilot made clear on the RT when initially informed of the delay (“if that’s the case, that’s the case”), you’d expect some delay. By the time the pilot is mentioning fuel on the RT when he says an emergency would **** the sequence, it’s almost 40 minutes since they first mentioned being unable to provide visual separation. 40 minutes is absolutely, categorically, not last minute. By the end of the last delay estimate given, it will be almost an hour since they first asked for the ILS. If an approach controller needs an hour to make a plan, that’s not planning.

They have to make a single gap for the lost paired aircraft, and potentially an extra mile on top of that behind the DLH. That’s almost exactly what they’d need to do for a single aircraft going around, give or take a mile. Do you think a 40-minute additional delay if you went around would be acceptable?

[InSoMnIaC]

Bring back lockdowns. That should bring ATC workload to a more manageable level.

[ATC Watcher]

LH SOPs

Originally Posted by YRP

Any word on whether they are going to change / clarify the SOPs to avoid this issue in the future?

That would be a decision taken at Chief pilot level, but frankly I doubt it as this SOP is most likely based on previous safety issues, not necessarily from this airport. LH goes to quite some funny places.

@ Txl

would bending the SOPs a little ?

Well , LH is not Spantax. . Plus there were 3 in that cockpit and you do not know who was the other Capt on the jumpseat.

[Uplinker]

Originally Posted by txl

..........But I was wondering if Lufthansa (or any operator's) SOP is cast in stone or did LH458 have some wiggle room to bend the rules a bit? "Ja, ja, visual separation is verboten, but we'll just pretend we didn't do it." I suppose the beancounters back home really appreciate diverting to Oakland and having to ferry the ship to SFO later. Will it be tea and biscuits for the LH crew anyway? Or would bending the SOPs a litte get them in serious trouble?

I would think that going against SOPs that were created for safety reasons, when there wasn't a flight emergency situation, might be viewed by LH bosses as wilfully and deliberately endangering an aircraft, and lead to dismissal or demotion?

I have to say it does seem odd that SFO ATC seemed to have no plan or ability to slot in a different arrival in their approach sequence. Had there been an actual emergency requiring an ILS, or a blocked runway; would ATC have coped or would there have been chaos in the air?

UK ATC uses published holds for Heathrow as buffers to organise the very busy approach sequence, does SFO have a similar facility?, I can't remember.

And is it really, truly safe for SFO to require visual approaches at night, just to get the volume of traffic in? That seems a bit like the tail wagging the dog to me.

[MarcK]

As I remember from the first few posts, LH was, in fact, offered an ILS, but refused it because it required "maintain visual separation". I think Air Force 1 gets that kind of service, but not ordinary carriers.

[Chronic Snoozer]

Being German and likely a non-native English speaker, the captain probably meant “…phuck your sequence” in which case it’s OK. Mind you, he probably thought controller’s name was “Rick” with a silent P.