From $250M to $6.5B: The Bay Bridge Cost Overrun
citylab.com
citylab.com
One thing I see time and again, in projects I've either been involved with directly, or indirectly, is a failure to do any substantive geo-technical assessment in the early planning phases. Not understanding your environmental invariants properly (when they are usually diverse due to the physical scale of these projects) always comes back to bite you later on. It's usually (eventually) a critical path item on any schedule since the structural design is so dependent on them.
"In April 2006, a consortium involving American Bridge and Fluor won the tower contract. It was built in China to save money—a decision that carried its own costs when inspectors later found poor welding and busted bolts at key points that required fixing. Frick says the current $6.5 billion total is a rough estimate, and that it doesn’t include interest or financing costs."
A mistake most larger EPCM's made in those days. The horror stories regarding the quality of Chinese steel and fabrication back then are very real. It is unthinkable now to let any fabrication of that nature happen there without adequate on-the-floor supervision and oversight.
Why let it ever happen "there" again? What hope is there for domestic manufacturing if China is the go-to source of everything?
As one of our great presidents once said:
Fool me once, shame on ... shame on you.
Fool me... You can't get fooled again!
Because when researched and managed properly you can still get a quality product at much lower costs. It just requires more effort on your part (a decent supply chain department that does their research and manages the risks, mobilizing an appropriate team in country and all the costs associated with that, which may or may not end up working out more).
I'm not from the US but we (Australia) have similar issues here. At the moment I work on projects with fabrication of big steel structures with high quality requirements, design once, build three or four slightly the same. So maybe take what I say next about manufacturing (design once, build millions) with a grain of salt. Chinese manufactures can do things that no one else can. There was a blog around 12 months ago where a group of MIT students went to Shenzhen. This quote stood out to me and really summed up the situation for me.
"In an interesting twist, the factory boss suggested that we could build the precision molding tools in China and then send these tools to a US shop for running production. Due to our requirement for clean-room processing, he thought it would be cheaper to run production in the US -- but the US shops didn't have the expertise or capability that his shop in China had to produce the tools; and even if they did, they couldn't touch his cost for such value-added services." http://joi.ito.com/weblog/2014/09/01/shenzhen-trip-r.html
https://research.stlouisfed.org/fred2/graph/fredgraph.png?g=...
Note that the bridge isn't unsafe, the modern span is much, much safer than the old WW2 era span. The new bridge won't collapse in an earthquake. However, because of design failures and shoddy construction, it will last much shorter than planned.
http://www.sfgate.com/bayarea/article/Bay-Bridge-s-troubles-...
"You only get paid if the project comes in under budget" might just only give you the lower-tier players who have no better project to work on -- maybe the more professional people would say 'no way' to an arrangement like that, especially given the % of these projects that wind up running over.
Then you wind up over budget with a sub-par contractor in charge of the project.. what do you do, change horses mid-bay? No way, turns out the threat doesn't even have teeth.
After a constant stream of cost overruns the US military tried to do fixed price contracts for awhile. What happens is the big players don't bid, or they do it through independent subsidiaries.
Then when the project costs balloon the contractor comes to you and says "Look, you have two choices - either you can modify the contract so we can make money, or we'll declare bankruptcy and you'll have to write off every nickel you've spent so far."
Incompetent planning, political self-aggrandizement, aesthetic confusion, diffusion of responsibility, corruption, not-invented-here, trade protectionism, failure to learn from experience, anti-corruption measures that actually promote corruption like closed competitive bid-plus overrun processes, and a general lack of interest among the public in electing officials that even promise to confront past failures and improve are just the beginning. The Augean Stables of American infrastructure spending are deep in muck and we have no Heracles on the horizon.
The Second Avenue Subway, the eventual Hudson tunnel, the Bay Bridge, insanely expensive BART to SFO, Seattle's tunnel, CA High Speed Rail, and the like are just a taste of a very expensive and gridlocked future if we don't change our ways.
[0]https://pedestrianobservations.wordpress.com/2013/06/03/comp...
https://pedestrianobservations.wordpress.com/2012/10/28/cons...
https://chrisgagne.com/1255/mary-poppendiecks-the-tyranny-of...
And there's a better telling of it in her excellent book:
http://www.amazon.com/Leading-Lean-Software-Development-Resu...
This makes a ton of sense to me; the American culture around plans is thoroughly fucked.
In software, otherwise reasonable people expect me to give them numbers at the drop of a hat. When will this be done? What will it cost? What will our revenues be for the next 5 years? When I tell them I have no data for numbers like that, they will ask again, as if they thought I hadn't quite heard the question.
When I make it clear that although I'm glad to do real estimates I won't just make up numbers, they get mad. They need numbers! [No, they don't. They want them, often for political reasons.] Other people have given them numbers! [I am not other people.] This is how everybody does it! [Thanks, I've seen the stats on how well that works.] I should just give some ballpark numbers! Nobody will take them seriously! [Except they always, always, always turn into real numbers.] It makes me crazy.
For years, I kept thinking that if I just explained it well, people would come around. But I've come to believe that the root problem here is a managerialist culture, where power matters more than expertise or professional responsibility. Managers want numbers, so managers get numbers. They don't actually help, but managers get to feel in control. If it all happens to work, they're brilliant managers. And if it doesn't, the blame falls upon the people who gave them the numbers, or the people who failed to execute according to the always-impossible plan.
Now I just look for people and work cultures that are focused on actual results and are comfortable with bottom-up power structures.
Just as an example, the NJ Turnpike recently finished a project widening it from 6 to 12 lanes for a 35 mile stretch, and came in on schedule and $200M under budget: http://www.njturnpikewidening.com/
https://en.wikipedia.org/wiki/List_of_transport_megaprojects
Polish minister goes to France to meet his counterpart. They meet in an amazing office and he asks his french college - How did you get money to build this? - Can you see the bridge outside the window? - Yes. - 500 mln on paper, built it for 250. Voilà.
After a year they meet again in Poland in even bigger an more magnificent building. French minister ask: - How did you get money? - Can you see the bridge outside the window? - No. - Voilà.
Straussian interpretation: "Pole can do it" resourcefulness of polish people do due to limitation of communist state. No customer support, no warranty - something breaks you have to fix it yourself. Inefficient economy causes lack of services and products which require people to hack stuff.
As a Pole I'd like that part of economy back, thank you. It's better than current "if it breaks, throw away and get a new one" combined with "if we can make it break just after the warranty period, they'll throw it away and buy a new one" mindset of western manufacturers.
1. The general public is uninformed, half-educated, and outraged easily.
2. Money in the beginning stages of a project is relatively expensive compared to the remainder of the project.
When the bid was selected, the project wouldn't be completed for another ~10 years -- Buying steel / concrete hedges and currency futures to guarantee prices would tie up tens of millions (maybe more) for ten years with interest before the bridge was operational. If the prices of steel and concrete and returned to their historical norms, the public would likely be blaming the firm for wasting all that money on worthless financial engineering. Could you imagine the protests if it came to light that the bridge consortium paid Goldman Sachs something like 10%-15% of the cost of the bridge with nothing in return?
"The cost increase was the result of detailed engineering studies conducted during the year or so after the initial estimate was released. Among other things, soil testing in the Bay had revealed that bridge pilings would need to be anchored “deeper into bedrock than expected,” she writes."
Now hindsight is 20/20 and I am not an engineer in this field, but it seems that if you're floating an estimate that isn't informed by the engineering studies necessary to give an accurate estimate then you probably shouldn't have given that initial estimate in the first place? Or at least should have given the initial estimate as a range and/or with a huge disclaimer that you might get into researching the bridge and the estimate could cost multiples more?
This industry is so old yet the innovation is really behind. Technology (?lame word?) should be speeding things up and making things more accurate and less error prone, but instead it seems to be delaying the time work gets completed.
Some promise is held in Information Modeling like BIM or CIM, but they are not trickling into the industry at the pace required. And further, many of the present day engineers are not in a position to understand this stuff.
These companies run on legacy networks, so I can tell first hand - developing your own software tools is only possible at private companies. Any of the major public companies have such old networks that no engineer can innovate. Suffolk Construction in Boston is a great example of the future of the industry.
My estimate though is that proper use of BIM (not just design & engineering, but also accounting and time management... 5D) would save anywhere between 40-70% on their projects.
So, why isn't it happening? Because the construction industry is always the last industry to adopt new technologies. "If it ain't broke, don't fix it" is the mentality. Also, if you're 40-65 years old and been building forever, it's not so fun to listen to 25-30 year olds tell you what's what. But that's what needs to happen.
The older guys in firms don't listen (the ones that do save a bunch of money...), and so the young talented people leave for better industries like software engineering, where they are actually respected and valued.
We are started to see SOME private companies, like Suffolk, really adopting new stuff. The big guys like Turner and AECOM are trying to adopt, but like I said earlier, their technology infrastructure is a JOKE! Try writing BIM automation scripts when your network admin won't even let you have root access. What a sad, sad joke it is.
For example, if the architect can design the building, then the engineering optimization is automated (plumbing is well-understood...), and the accounting, BOMs, and fabrication documentation are all automated. This idea removes the time-consuming and pointless conversations between architects, engineers, and the construction team.
This is the potential of BIM that I see.
Because if that's true, who cares about old companies - start a new company ,raise money(there's plenty for good opportunities ), train some people, and start offering everybody 30%-40% discount on their building. You'd kill the market and make ton of money , right ?
We are basically talking about applying AI cost functions to the AEC industry (optimize design, time, cost).
If you're building a cookie-cutter high rise residential building, it's kind of well known and easy. The price has already been reduced to about where it is going to be. Typical floors mean typical documentation.
However, as soon as you start talking about complex building structures and services, BIM easily pays itself off. Think automated finite element analysis. The key is to plan far ahead of time, and design far ahead of time. There is no automation today, it is all done by hand. But I believe someday, BIM automation will dominant the market.
A few architects out there use procedurally generated shapes for architecture. A few engineers out there use AutoCAD plugins to streamline engineering services. BIM automation is already on the way...
You aren't suggesting that such root access would be any less of a joke? Maybe it's just me, but deploying scripts should be authorized for one or more separate roles, none of which are "root" or managed by the network guy.
It's a similar problem to why people keep using Excel instead of expensive database products - because the latter are too constrained and inflexible and not really up to handling the actual work that needs to be done.
Oh, you want the chocolate sauce on top? That will be an extra few billion dollars, please.
Please stop this ride.
BTW the golden gate was built mostly with private funds from bonds ($35M in 1930's dollars)
Also a record of "only" 11 people killed building it.
Immediately following the Loma Prieta quake, the
GGBHTD engaged a team of consultants to conduct a
vulnerability study. The conclusion of the study was that
under a Richter magnitude 7.0 or greater earthquake
with an epicenter near the Bridge, it would experience
severe damage that could close this important
transportation link for an extended period. If a Richter
magnitude 8.0 or greater earthquake centered near the
Bridge, there would be a substantial risk of impending
collapse of the San Francisco and Marin Approach
Viaducts and the Fort Point Arch, and extensive damage
to the remaining Bridge structures, including the Main
Suspension Bridge.
Also, a Richter 8.0 is a pretty damn hefty quake.
Those seem like relatively minor differences, but (assuming the new span actually performs like designed) the Bay Bridge should easily survive an earthquake that is 3.1x larger and releases over 5x more energy than one that will "extensively damage" the Golden Gate Bridge.
So presumably an 8.0 is plausible but an 8.5 may not be.
http://www.smithsonianmag.com/science-nature/what-will-reall...
When the people who are not knowledgeable about the actual details of the project require their expectations to be met, other expectations of theirs will not be met (the classic "fast, cheap, or good, choose two" joke). I like to say to people making unreasonable demands "Do you want to be disappointed now, or later?"
If the first estimate had been $4 billion (assuming we're taking $6.5 billion in 2015 dollars and working back to 1998 dollars), the project never would have gotten off the ground. The government would have said "fuck no" and asked for another bid. It would have been mired in discussions, argument, etc for years before eventually settling on a $1 billion price tag -- that will eventually balloon to $7 billion or so anyway, because the winning bidder intentionally underbid because it was the only way it would get approved.
The only way to build large public projects like this is to take advantage of the sunk cost fallacy (or "bait and switch".) Government contractors will get their cut, and the regulatory tack-ons added by local governments to put their stamp on it (and get some operating budget!) also add money.
The other way to build large public works is to figure out why they are so expensive and actually fix it, and then build things for less than 24x the original estimate in the first place.
And if that can't be done, maybe it actually shouldn't be built.
(I mean this in general, not about this specific case. But it is always something. And that is indicative of a problem on it's own.)
You simply can't build something like the Bay Bridge for $250 million. And the reason public works projects are so expensive is usually because they're big, things go wrong, and there's a political element. You can't predict everything that will go wrong, and you'll never remove politics from it.
The problem isn't that huge public works projects cost a lot of money, it's that the estimation process is more about politics than reality. People building the Bay Bridge knew they could go find comparable projects and build an estimate that way -- they didn't do this because they were trying to come in at a certain price point, which means throwing a defendable methodology out the window.
I'm not saying this is the way it SHOULD be done, but it's the way things are done. And if you try to do it differently, someone else will just underbid you and raise the price later.
The estimates were about $4 billion. I believe the project never went anywhere because the states didn't want to put in their chunk of that $4 billion.
Also, don't underestimate the effect inflation can have on large public works projects. Inflation over the course of a 10 year project can be as high as 35% - even if inflation stays under Fed's targets. When stating dollar numbers, projects almost always do so in current, nominal dollars because that's what gets you the lowest number. And all of this is before taking into account the financing costs of a large public works project (municipal bonds are tax-exempt, but they aren't free).
It would be hard to stick to such a pledge, of course.
$6 billion overrun could have fed, clothed and health insured million of people
Because you still need a bridge that won't collapse in an earthquake?
The Bay Bridge is the 2nd busiest bridge in the US... $6B seems like a lot until you look at how much disruption and damage a collapse would cause to the $600B GDP of the SF Bay area.