Seattle's new airport terminal can’t fit planes it was designed to handle | Hacker News Reader

Seattle's new airport terminal can’t fit planes it was designed to handle

paddleyourownkanoo.com

111 pointshnburnsy67 comments

Whether those particular gates were actually designed for widebody aircraft is exactly the point of contention. Clark Construction says the design specifications for those gates were for narrowbody aircraft, and suggests that this is but one of many SeaTac flip-flops that caused tremendous cost inflation, resulting in an earlier lawsuit initiated by Clark for unpaid labor costs.

Given the acrimony it would seem rather difficult to know as an onlooker what transpired and why.

Seattle Times has a better article: https://www.seattletimes.com/business/boeing-aerospace/sea-t...

> Clark’s latest legal brief, filed in late February, asserts that the contract documents defining the work “specifically state that Clark was to design the Project to accommodate narrowbody aircraft at Gates A6 and A8, not widebody aircraft.”

"Given the acrimony it would seem rather difficult to know as an onlooker what transpired and why."

It's very plausible that the people writing the requirements took the cheap/ignorant route and it's very plausible that Clark knew it was a mistake but it would be a worse mistake to fight.

This seems like a common problem not just in construction but in all parts of society. It very likely that plenty of engineers looked at the design of this and said that it doesn't work this way and that they were silenced along the way.

Germany built massive transit hub train terminals in a matter of a year during the Nazi period(I forgot the name, but it was a transit hub for military equipment between Hamburg and Dresden I think), while today, this story here reads exactly like Stuttgart 21, where a decade long construction projects ends up with less capacity than the previous train hub.

My dad was telling me about when he learned construction where there was the same guy responsible for building the stairs and sometimes that person would diverge from the plan because his experience would tell him that the design was wrong.

I managed part of a construction project when I was working in Government in Germany and many of the subcontractors told me exactly the same thing, the guys designing the network, AFA and data center access control(Bosch), as well as the network cabling, and even the carpenter and the elevator company all told me the same thing. The architect sent them wrong specs and these guys said that, were they to change any of the specs they would be liable for any upcoming issue. The elevator company was the only one that looked at the specs and told his worker to change it and then in front of me told his worker that he would take responsibility for the change.

It went beyond that, because the Architecture firm running the construction project actually told these subcontractors not to send us issues that they see, so until I started actively poking around they told me that they were told not to inform us directly.

For all other building services engineering parts I had to get personally involved, finding the issues, then corroborate them with a second contracting firm only to then face an uphill battle to convince leadership of the decades of upcoming problems if we don't change them.

This is the same reason people hire Accenture and EY we read in the other thread. As long as leadership doesn't know they're not going to be liable for all the crap that was installed under their watch. And because the contracting firms know that they will building things EXACTLY according to spec(or this case the construction tender).

https://en.wikipedia.org/wiki/Stuttgart_21

I'm seeing some weird parallels with the byzantine generals problem/distributed consensus and the architect/contractors in your story. Did the architecture firm eventually sign off on the changes the subcontractors made?

They had to, we hired them. But it's more interesting that the leadership was really upset at having to tell them to change plans. After the changes to the construction project they heavily limited what information I have access to.

Fundamentally I didn't have any authority to change these things on my own, so what caused the leadership to change course was basically an email in the following format:

"Here's the current state we have ...

Here's what will what happen if you move forward with it ...

I would heavily recommend against proceeding with the current course.

Ultimately it is your own decision and you will have to bear the consequences of that decision"

> Germany built massive transit hub train terminals in a matter of a year during the Nazi period

Well, things are always easier easier to build when you have access to slaves, but they still won't solve the project management issues in debacles auch as Berlin Airport. Do you have any concrete transit projects built under Nazi rule in mind? Everyone knows of the Autobahn which was drastically expanded during Nazi rule, but I can't find anything about other transit projects outside of the regular railways that were built for the army or deportations.

Having not seen the contract, but having worked in construction, I think it's likely the contract did not specify clearances and/or turning radii, or maybe even dimensions. Clark was probably hoping the Port would cave and pay for change orders. It's even possible that individuals at the Port were aware of deficiencies, but let it slide...for reasons.

>specifically state that Clark was to design the Project to accommodate narrowbody aircraft at Gates A6 and A8,

So sounds like it's a design-build contract.

What many don't realize about this is that in contrast to other contract types (eg design-bid-build), the "design and build" in design-build are both contracted to one entity.

This reduces what would be typical conflicts between the architect and contractor, because it eliminates incentive to play blame game when plans have internal contradictions or issues with constructibility/ buildibility (specs are impossible to build). This process has other advantages, like shorter lead and construction times and fewer delays, since the builder doesn't need to wait on all design phases to be complete before starting work.

But it also effectively eliminates a party of review.

To all the other comments suggesting VR/ or software simulation as a solution: 1) on a project this size, I guarantee 3D models were made and presented for both schematics and visuals, and 2) all the modeling in the world doesn't fix the original specs being wrong.

Even if 3D lifelike visuals aren't a part the deliverables, the contractor has incentives to make them, including because of the case at hand.

And (for those without construction experience) they have much greater incentive to point out errors in the (owner-provided) specs than to ignore obvious blunders.

Change-Orders are the bread and butter for most companies doing this type of work, because the contract guarantees an inflated profit margin on the value of changes made. Which aligns incentives for both parties.

It sounds to me like the owner/rep for seatac just wasn't paying any attention or doing their job.

Which happens a lot on public pet projects when politicians want to be hands on and not outsource owner/rep responsibility to a more experienced party. Many states have departments that specifically will do this for any public entity in the state, and others allow their DOT to handle construction responsibilities for other departments, on-demand.

> $1 Billion […] After more than 400 million labor hours

Or, an average of $2.50/hr of labor?

> [much later] Oregon Live reports that Clark Construction sued the Port of Seattle last December for $100 million, claiming it had been underpaid labor costs

Oh, well, I suppose that checks out.

… you get what you pay for, Seattle?

(I do wish journalists would ask the obvious followups when presented with what is hopefully a wrong number.)

Ironically city of Seatac was one of the first with $15/hr min wage and now has $19 - highest in the US.

that 400 million number sounded awfully high to me.

> Originally earmarked to cost no more than $968 million, the Port authority was forced to pump an additional $18 million into the project in 2021, but now there are fears an additional $78 million will need to be found to address the gate layout design.

Wait, they're complaining about a potential *10%* cost overrun? They should drop by California, I've got a lovely HSR system to sell them. It won't cost a dollar over $33 billion, I swear.

the Port of Seattle claimed that the shortfall of available gates could cause damages “in the tens or hundreds of millions of dollars over the expected life of this project.”

That’s a pretty big range of damages, spending $78M (which probably means $100M by the time it’s done) to fix something that could only cost $50M in lost damages over the 30 (?) year lifespan of the project doesn’t seem like a good return on investment. It’s not like the gates are completely unusable, they just need to be used by smaller planes. How much would it cost to make those 4 gates completely unusable over what is sure to be a multi year construction project?

It’s an international terminal, so the plane jet bridges exit out to immigration, customs. I’m not sure if they can edit out into the airport, though I suppose it is probably possible since that’s how they are loaded. If used for international flights, there are simply not many narrow body planes to consider…Canada would be the only place close by, but they mostly do US immigration/customs in Canada via special arrangement.

737 max can do 4000+ miles in on paper. That puts all of Mexico, most of Central America, and a good chunk of the Caribbean in range I believe.

Ya, but Seattle is still a mid sized city, I don’t think most of those routes could work profitably in narrow bodies. Alaska has only four Mexican destinations, all the other airlines probably find transfers more profitable.

There are more than 25 international destinations served by nonstop routes from SeaTac[0]. In my experience (more than a dozen international trips from SeaTac), the majority of international routes from SeaTac are either non-stop or transfer internationally rather than domestically -- most flights to Europe and Africa, for instance, feature a transfer at Amsterdam or Frankfurt.

But, of course, those routes are generally not less than 4000 miles. SEA->LHR is one of the shorter routes to Europe, and even that is >4700 miles.

0. https://www.portseattle.org/page/non-stop-international-rout...

A lot of routes to Asia, which are still pretty far (10-12 hours to Tokyo). Nothing to Russia far east, which would be a bit closer.

I’ve been in plenty of international trips where I transferred in SF or even Chicago or New York. It is a bit weird flying down to SF just to go back up to asia.

That's the reality of the US: it has lost its capacity of building anything under budget and correctly. Engineering skills have been replaced by blind management guided by computer software. This is the result.

Disagree. Engineering skills are better than they have ever been, and tools are better. What management has figured out is that governments have more to lose. Do you want a multi-year lawsuit and delay with election season coming, or do you want your bridge built eventually for just a hundred million more?

The reasons why this is happening don't matter. It is exactly because of this line of thinking your expressed that engineering skills are going down the toilet. What matters is the result for the country, and the results are the worst possible.

Clark Construction's lawsuit is Case No 22-2-20747-7 SEA filed with the King County Superior Court[1].

That being said, it never ceases to amaze me how much friction there is in attempting to access public records in our country's court systems; at least King County Superior Court uses Daily Journal tech, which isn't the grift-by-undesign dumpster fire that the Western District of Washington's PACER interface is, but you'll still need to register a username to dive deeper.

[1] https://dja-prd-ecexap1.kingcounty.gov/?q=node/420/3327963/F...

Only after the facility was built did the port authority realize that widebody aircraft like the Airbus A330, Boeing 777 and 787 Dreamliner, which are regularly used on long-haul international flights, couldn’t fit into four of the gates at the facility.

Four out of 20 gates! That is ludicrously sloppy.

I enjoy that their recourse for such a failure in design so far has been to "write a strongly worded letter". Is there some kind of protection for the builder that is preventing them from being sued? Is the airport body governing this actually the one that failed? What a bizarre situation

Probably a Design-Build bid, based on how other large infra has worked here in Seattle (See: Hw 99 Tunnel).

It appears the 4/20 work with smaller planes, so merely may not have been need until now.

Local reporting has a few other issues mentioned about specific gates, which to me sounds like a "Our contract says you have 30 days to reply to our letter informing of issues before xyz happens". So purely posturing before mediation before lawsuits. But we'll see.

For anyone interested, I found the local media coverage was higher quality: https://www.seattletimes.com/business/boeing-aerospace/sea-t...

As written in the article:

> ... the Port is counter-suing Clark to recoup costs involved in fixing the gate layout fiasco.

usually design documents are stamped by a "professional engineer" (PE) licensed by the state. They're names on the documents and they're supposed to check them.

I used to be a civil engineer and for the earthworks the PE always had us checking things. especially grades in the areas around the perimeter. Our initials were on it but the PE stamped it. We did catch a few things where the plans called for a "wall" because if we did our 3/1 grading we'd meet existing over the property line.

Not sure how the engineering works for airports.

It works exactly the same way. There is an engineer of record associated with all the plans. The design-build contractor says they were told to build it this way in the contract. The Port of Seattle disagrees. This dispute is over what the Port actually contracted for. It's not reall a matter of shoddy engineering.

Some entity must have been responsible for ensuring that the technical drawings satisfied the technical requirements. If this entity exists, they were at fault here. If the entity doesn't exist, the entity responsible for assigning that responsibility is at fault.

The argument is that the technical requirements given to the construction company do not match the requirements currently being stated by the Port:

> “The Port is now identifying modifications to gate design based on operational considerations that were not contemplated in the program requirements in our contract,” [a Clark Construction representative] wrote.

https://www.seattletimes.com/business/boeing-aerospace/sea-t...

Right. And per my reply, whomever’s responsibility it was to validate that — or to assign someone to validate that — is at fault.

Not really, you said the technical drawings had to be validated against the technical requirements. By Clark's account, those two were in sync.

What wasn't in sync was the ports "general ideas" about what they wanted from the transaction and the technical requirements they produced attempting to encode those "general ideas". Unfortunately there's not a great way to validate those, at least that I see. Pay for a better engineer to sit in on your years of meetings rambling about what the "future of the port" will be?

It's hard to sue when you ordered four narrow body Gates and then have buyers remorse later. This is the case that the Builder is making

Who had the incentives at the Port Authority to make these decisions and what were their motivations or failure of motivation is the question...

Maybe it was on purpose, as by cutting the size of four gates to handle all but the largest aircraft, net-net you get a whole extra gate or two. I was about to snark about project anti-patterns and blame weak parenting for raising generations of worms, but I'd be willing to bet instead of 18 total gates, you get 20 gates by giving up some of the largest ones.

Accepting the economics of maybe 10-15% of aircraft as a design constraint on all gates seems unwise, and maybe some climate people figured the same.

France had a similar debacle with ordering the wrong size trains [0]. Supposedly that blunder cost $20 billion.

[0] https://www.bbc.com/news/world-europe-27497727

Edit: Apparently $20 billion was sticker price for all trains. Presumably something was worked out so they were not out the full bill.

Because many train platforms were older and didn't leave as much room for the train. Given the cost I wonder if they could have just done remedial work on all those platforms as the trains came online. After all, track and bridge maintenance has to happen quite regularly.

It was just the station platforms, and yeah, it turned out to be cheaper to fix the platforms.

So, nobody built a scale model or 3D simulation before they built it IRL? Or is the article leaving something out?

Other articles are pointing out that the construction company claims that the contract specifies 16 widebody gates and 4 narrowbody gates, exactly what was delivered.

Of course they did. It's a common part of large-scale construction. You don't just model the finished product; you model the stages of construction. This is called BIM (Building Information Management), Seattle wrote BIM deliverables into the contract, and Clark used Bentley software to project-manage the construction. Bentley is using this as a PR case study: https://www.bentley.com/wp-content/uploads/2022/05/CS-Clark-...

Stuff like this used to be designed at drafting tables, by hand. Now we have computers that facilitate sophisticated modelling during every stage of the design process.

So how is it even possible to fuck up this bad?

Checks and balances failed. People phoned in the work. I can imagine it.

When something gets easier we get sloppier

Well today we have lower iq scores.

Someone probably decided it was not worth it to pay for the right software that could detect it

Because machines are never the substitute for thinking and checking.

Tools need to be used with intention and not by coincidence.

Even with a simple calculator, just by getting a result after mashing buttons it doesn't mean that the result is valid for the problem you were seeking to solve. The same goes for more complex tools and new things like LLM expose the lack of critical thinking by humans even further.

"The system told me so" is the laziest lack of ownership and appeal to an authority that is not only not above reproach but usually extremely buggy for different reasons.

There is no real engineering in charge of this, it's all managers making decisions and engineers are asked just to do the boring parts. How would managers know that gates need to be dimensioned for airplanes?

Since the construction company claims that the contract only called for 16 widebody gates, I don’t think that there was a failure during the design process. I’d be willing to bet that the project management changed their minds but failed to actually amend the contract.

In my city they built an underground train station and forgot that diesel locomotives are still in service: the venting system was designed for electric trains only!

> The Port of Seattle says it could cost as much as $78 million to fix an apparent design flaw

A good detailed simulation that would have caught this would be a small fraction of that cost. Release it as part of flight sim games and let the public explore it. Offer rewards for suggestions that lead to design changes. With such an expensive risk of mistakes it should be only a matter of time before detailed explorable sims of major projects become routine.

Our world we be so much better if the people responsible for mega projects actually thought this way! But, I suspect that a person needs to have a certain hubris to believe themselves worthy of heading a huge project, and that this belief militates against seeking input from outsiders that might catch mistakes.

They didn’t need a detailed simulation all they needed was to import 2d dxf’s of the planes into the terminal drawing and see if they fit. I bet requirements changed and no one reviewed those last 4 gates before construction docs were printed.

Wouldn't it have sufficed to check the planes wingspan?

They had a good detailed simulation. Any competent construction project does, and most of the time, agencies are federally mandated to require BIM both for the construction phase and the maintenance phase.

No amount of detailed simulation can help if your contractor thinks you wanted 16 widebody bays and 4 narrowbody bays. There was nothing wrong with the execution; the disagreement is about what was to be built.

Would love to get more insight into what happened in the design phase? Besides the construction contractor who was the architect, why aren’t they being sued?

Was wondering the same. I worked for an architect as a CAD draftsman in college and know that what gets drawn vs what actually gets built can be two very different things. Unless the architect is at the site every other day checking things the contractor treats construction plans as a basic guide at best.

Project ran for 2x as long as planned and 3x the budget. And we're left with finger-pointing and lawsuits

This is absolutely something that they should have caught before building the terminal.

In the old days where the executive who’s pet was this type of construction project, it’d have been painfully obvious in the obligatory scale model across the room.

The executive would be playing “park the planes”, like a six year old plays with toys, and suddenly notice that his shiny new 787 scale model doesn’t fit at the gate.

If only Meta had nailed VR in time, this disaster could have been avoided.

Measure never; build again and again and ....

People at the top have a mindset to get things done first and they can ask for forgiveness later.....

This sort of stuff combined with the train situation really does feel like the start of an Ayn Rand novel.

I will say this is a remarkably small overspend for this sort of project though "Originally earmarked to cost no more than $968 million, the Port authority was forced to pump an additional $18 million into the project in 2021, but now there are fears an additional $78 million will need to be found to address the gate layout design."

Normally these projects at least double.

What do you mean by “the train situation”? I get the subject matter connection with Rand, but I’m not seeing a connection with SeaTac which has had a light rail connection for many years now.

I meant more broadly the train crashes in the US, although the light rail specifically has also had problems

https://www.seattletimes.com/seattle-news/transportation/sou...

Interestingly, reporting at the time made no mention of adding narrow-body plane gates.

Example from 2020: https://www.airlinereporter.com/2020/01/sea-tacs-internation... "The new security corridor along the face of Concourse A will provide eight international wide-body aircraft gates with direct access to the new arrivals facility."

The Port of Seattle's own deck from 2014 implies only seven: https://meetings.portseattle.org/portmeetings/attachments/20... (page 3)

By 2018 the Port has decided eight are coming: https://www.portseattle.org/sites/default/files/2018-03/Sea-...

Maybe they opened a gate in the meantime?

This 2014 article talks about an increase of nine: https://airlinegeeks.com/2014/11/16/seatac-airport/ "It will also increase the number of wide-body gates from 11 to 20."

So does this 2016 article: https://www.bizjournals.com/seattle/news/2016/09/29/fly-thro... "The new facility will allow airlines to add international flights to and from Sea-Tac by increasing the airport's international wide-body aircraft gates from 11 to 20..."

This 2015 one sticks to the "add seven" plant: https://informedinfrastructure.com/15920/port-of-seattle-rev... "Nearly double the number of existing gates capable of serving international wide-body aircraft from 11 to 18 gates"

By 2019 they're talking about adding eight, but not starting from 11: https://seattlebusinessmag.com/aerospace/more-international-... "... a multilevel complex that will expand the number of gates at Sea-Tac that can accommodate international wide-body airplanes from 12 to 20."

So I think they squeezed in a gate before construction started.

It's fascinating that over the years apparently nobody knew what the hell the plan was. I guess it's no shock that the end result was surprising.

Reply on news.ycombinator.com