Building Big Things Better

Over the last year I have had the privilege of working on some of the largest tunneling projects in North America. Most recently, two of the first single-bore transit tunnels to be built here: Toronto’s Scarborough Subway Extension (technically, I have just been working on the stations, not the tunnel) and San Francisco’s Silicon Valley Extension.

Involvement in these megaprojects has led me to two questions:

• Is all this worth it?
• If so, surely we can build them better, how?

In one of my more recent pursuits of these questions, I have read two works that have improved my understanding of the problem (though still leading to more questions than answers):

How Big Things Get Done : a book by a by an economic geographer and professional planner named Bent Flyvbjerg and best selling author Dan Gardner. They look back over 30 years of Flyvbjerg’s career working on megaprojects and use a database Flyvbjerg has compiled of over 1,600 projects; with the goal of understanding what causes the successes and failures.

TransitCost.com’s executive summary : an executive summary of a more exhaustive report produced by a NY University research group that looks at the differences in high and low cost transit construction. They compare costs of transit projects in countries across the globe and try to summarize the reasons some nations build for much less money than others.

Here are some of the lessons I have learned:

High wages do not mean expensive construction

Sweden has the highest wages of the countries studied, but is in the low-cost group. High-cost countries tend to overstaff their projects with both blue-collar and white-collar jobs. The Transit Cost study found 40%-60% over-staffing in Boston and New York projects. I understand that NY and Boston are unique labor markets relative to most of the US, but from personal experience I could easily be convinced other projects are at least 30% overstaffed.

In addition to over-staffing, the cost of craft labor in the US is significantly increased by our overtime rules. Apparently, miners in Sweden do not get paid overtime, and French workers receive less than 1.5 times their normal pay. In some cases, weekend or night work in the US can be paid out as high as 2 times the base rate. And since shift selection is done by seniority, many of the more senior craft workers (read: high paid) will elect to work the most lucrative overtime shifts (compounding on the already high costs).

For white-collar workers, the problem is simply over-staffing. Since there is often little true coordination between stakeholders on a project, everyone will insist on having their own management and inspectors to oversee the work. In outlier cases like Massachusetts Green-Line-Extension, this led to 1 supervisor to every 1.8 craft workers.

Modularized projects succeed at high rates

Flyvbjerg found that solar power projects had mean cost overruns of 1% and only 2% of projects had cost overruns in excess of 50% of the baseline budget. This can be compared to tunnel projects at 37% and 28% respectively. There are several reasons for this, but his main justification is solar projects’ ability to modularize. Every part of these projects can be systematized in straight-forward ways: solar cells are assembled into solar arrays at one factor, another factory can make partially assembled mounting frames and hardware. Once everything is at the site, the Lego pieces can be assembled on a nice graded site with no overhead obstructions on foundations that are some form of small to medium diameter drilled or driven pile.

This is very different to most underground projects, only so much can be modularized in such a straight-forward manner. But, just because something is difficult does not mean it is not worth the effort. We could almost certain do more design standardization (according to Flyvbjerg, in the 90’s Madrid built 37 stations based on one modular design) so that procurement and learning curves can benefit from scale. Other ways to modularize would be some version of the the Boring Company’s Henry Ford approach of any tunnel diameter you want, as long as it’s 12 feet, though we might a few addition diameter options.

Using unit rate contracts make (inevitable) changes less litigious

This one might have been the most interesting lesson to me. In low-cost countries, the Transit Cost report finds that contracts are frequently unit price (they use the term “itemized”) and the prices are publicly available. In North America, typically little more than wage rates are public knowledge. Set rates allow for much quicker change orders: when a changed condition occurs, all the parties have to do is count up quantities since the rates are already set. In this way, change orders happen with much less disruption to the progress of the project.

The authors admit that in order for systems like this to work, the Owner has to have significant in-house expertise (both legal and technical). In the US it seems a bulk of the expertise is purchased by the Owner via consultants, which I think started in the 90’s with the Big Dig and Parsons Brinckerhoff. The push for small-government led to smaller budgets for full time staff, requiring consultants to be hired for new projects. The short term gain of lower payroll liability may have given way to a much higher long term costs of relying on outside help.

There are always more questions

The three lessons above are just the tip of this iceberg. If both texts are even partially correct, there are some pretty significant changes that would need to happen in the US tunnel industry to make real progress. And those changes are based on data from either other industries, other countries, or both; so it is unclear how well they will transfer.

These projects we work on are so large that no one person can understand their entire scope (though many times we feel it’s just all those people that don’t get it, I know exactly what is going on). And we simply will never find fully satisfying solutions to these problems. But in the words of Deng Xiaoping, we are “crossing the river by feeling the stones”. We can only know for sure what is right in front of us, the rest we have to figure out by iterating. Those iterations will be expensive, but not as expensive as waiting to cross.