11 December 2009

Blowing smoke up our tunnel?

California's Department of Water Resources has estimated that a "Peripheral Tunnel" that's 43 miles long, 150 feet underground and has a 15,000 cfs capacity will cost $10.6 billion to build.* That's only $1 billion more than the cost of a (surface) Canal.

I asked Steve Kasower, my comrade-in-bomb-throwing and author of a study that estimated this tunnel would cost $54 billion, what he thought of this estimate, i.e., "Steve, does this mean you were wrong or they are lying?"

Steve told me this:
I am not sure that it is a lie per se. I suspect that it is politically motivated since Lester Snow had already said it was about $10 billion in response to my paper and shortly thereafter. Upon what basis has the estimate been based? My inflation of the Chunnel was from their actual $21 Billion in 1994 cost. I inflated at 3% annually (very conservative through the time period from 1994). So if DWR could build the nearly identical tunnel for $10 billion in 2009 it suggests that the Brits and the French radically overpaid for their tunnel. DWR in today’s dollars is over half the cost of the Chunnel in 1994 dollars and less than 1/3 the cost in today’s dollars. How is it that DWR can be so economical? Let’s ask some basic questions:
  • Is there some radically different strata that will be encountered under the Delta compared to Chunnel that would account for such cost savings? Show us the studies.
  • Has the technology advanced so radically in tunnel boring that we can enjoy such a major savings? Show us the data.
  • Is the difference between the size of the Delta tunnel and the Chunnel contributing to such a major difference? Show us the incremental basis for this determination.
  • That is about all I can think of for radical cost differential. Certainly steel, concrete, labor, and energy all cost more today than in 1994 so what gives?
Steve's Bottom Line: It is a lie!**
DZ's Bottom Line: Steve was wrong -- see comments :)
* The Colorado River Aqueduct (CRA) has a 1,500 cfs capacity, which allows it to carry 1.3 maf of water/year from the river to SoCal.
** Section 3.3.1 of my dissertation:
MET [Metropolitan Water District of SoCal] based its demand calculation on the "habitable (not inhabited) area of the South Coastal Basin" and made no provision for the effect of prices. Milliman claims that MET engineers started with a CRA capacity of 1,500 cfs and projected a demand that would require that much water. After building the CRA, MET set prices to cover the cost of building and operating the CRA but ignored the possibility that demand would not exist at those prices. Given the average cost of Colorado River water, even on the basis of full capacity operation, is roughly three to five times the cost of existing water supplies and a reduction in demand due to high rainfall, actual sales missed projected sales by an order of magnitude.
This quotation explains how engineers gave a wrong, but political, answer to a question. The result (too much, and subsidized, water) drove SoCal's post WWII-sprawl.

10 comments:

Wes said...

I don't know whether to call the first item (re: chunnel) just another Snow job, or to raise it to the level of requiring us to refer to it as "ChunnelGate."

Anonymous said...

Kasower's estimate is bogus, the chunnel went 250 feet deep, through difficult strata on the French side, is 3 tunnels not 2 designed for human occupation, and in 2007 dollars was 10 billion pounds. It was built underwater and could only be approached from the ending points.
This is a detailed engineering estimate for a tunnel through clay, with multiple starting points so it is built in sections, with drilling machines in the middle of the common size range now used (the Chunnel was at the top end of the line at the time, with little experience). you can see the details on the BDCP website and can compare to the chunnel details easily on line. Go look it up before making fools of yourselves.

John Bass said...

The Swiss have only recently completed a twin-bore, 30+ foot diameter, 45-mile tunnel through the Alps that cost 7.2B USD. This tunnel is a fact. It is also almost exactly the same length and diameter as the proposed Delta tunnel.

Granted, the ground conditions that are being drilled through in the Delta are completely different from a mountain range. There is likely little ground water to deal with when drilling through mountains. The 150 foot depth of the Delta tunnel would not be through rock, but through alluvial and lacustrine sediments. Does it seem unreasonable to argue that water pumping v hard rock boring would result in a budgetary wash.

The chunnel is also a three-bore tunnel, not two, as are the Swiss tunnel and the proposed PC tunnel.

Anyway, I'm not sure that implying Lester Snow a liar contributes usefully to public discourse, esp. when there are projects that are not cost estimates or extrapolations from a 15-year-old project, but actual built things that have just been finished.

How did the Swiss do this? New and better technology, possibly? Just being Swiss maybe?

Spencer said...

I agree with the above commenters that a water pipeline is not remotely comparable to the complexity and engineering demands of the Chunnel.

At 150 feet deep (and to preserve the ecosystem), California's probably not doing cut-and-cover, rather proposing digging using something like New Austrian Tunnel Method.

In DC, they're boring a tunnel (two, technically) for the Metro right now using NATM, which is a half-mile long and estimated at $85 million. Multiply that rate by 43 miles and that puts it at about $7.3 billion, which is pretty close to the Swiss example above. Even assuming some overruns (and not to mention needing only one bore w/o transportation improvements, though size of the hole may differ), the $10.6 billion estimate doesn't seem out of the ballpark by any means.

Which is not to say that spending billions on a new canal/tunnel is necessarily a good idea in the first place.

David Zetland said...

@Anon -- "Go look it up before making fools of yourselves."

I thank you and the other two commentators for your insights. I am not sure if you will call me a fool if I tell you that I learned something and agree with you, but if you do, then you are now in the company of fools :)

Anonymous said...

Not at all (I said before making a fool...) You did the right thing. You have my respect.

David Zetland said...

Oh good. Let's go get some beers :)

("before," btw is a good qualification, but I OFTEN state opinions based on what I know at the time, not exhaustive research; if I did that FIRST, then I'd die of overwork, and you'd die of boredom waiting :)

dfb said...

At this point, I do not think we know who is right or wrong. DWR is still too early in the spec phase to actually know what it will cost.

Aquafornia posted the presentation given to the eir/eis steering committee a couple weeks ago so you do not need to download the pdf. http://aquafornia.com/archives/16525 It includes the tunnel specs and comparisons to which anon referred.

I would like to point out slide 17. It gives construction pros and cons including risk levels for engineering and known/expected issues. Among other things, it says there is high risk due to unknown subsurface conditions. Is that what they say with all tunneling projects? Without more info, it suggests they have done only some initial exploration of the ground conditions. Perhaps, the tunnels will need to be 200 feet down. Who knows? How would such a development affect costs? Granted, I am not an engineer or gov't planner and not privy to what this sort of chart means. As a lay person, I've only seen and heard general information that suggests DWR has some detail but not enough to give a completely accurate prediction of costs.

Also, we commonly see this sort of politicized price tag with major public works projects and then are nickel and dimed with feature/scope creep, increased materials or labor costs, or increased borrowing costs. We saw that with the new bay bridge. I know they are different projects; however the political pressures are the same. I have a difficult time accepting that it will cost under $15-20 billion. I have a harder time believing folks in LA will not decide to use that kind of money on diversified local water reclamation and desalination efforts instead.

Anonymous said...

The costs may well go up as they learn more; they seldom go down. But there are several items left out of the cost analysis that all weigh to the tunnel approach:
1) time. The time for land acquistion is not included and is probably 3 to 5 years longer for the massive acreage to be acquired with the canal. That costs 1 MAF/year of water, and the lost opportunity cost of having no solution is about 1 to 1.5 billion added cost to the canal
2) time. the added time to acquire land also results in greater costs of construction just through inflation (count 4% per year).
3) land. the cost of land is low, based on the cost to other linear projects in the Delta, does not include severence issues, irrigation and drainage damage to the farmers... All more cost for the canal
4) liability. There is 50,000 af of water sitting behind those 45 mile levees. A break is going to be costly for the owner.
5) present worth of Operations and Maintenance. O&M is more expensive for a canal.
6) Environmental; now that the cost of the canal and tunnels are nearly the same (they are the same in terms of the accuracy levels), the tunnel wins the Clean Water Act 404 requirements, the canal fails it.

Next step, what size is really needed. Two tunnels gets about 6 MAF per year (up from the current 5, but no more than they used to get). One tunnel gets a tiny bit less at lower risk and much lower cost. Will they go as big as they want, as big as they can or as small as they need?

Josh said...

What's the comparison between French/Swiss ability to deficit spend, vs. California's inability to deficit spend (legally) plus looking at 21 Billion USD in the hole plus a comparison of each region's GDP vs. total spending for this project.

Perhaps this is a job for a federal project, but not a state.

Thanks for showing that it's ONLY 10 billion... a billion here, a billion there, soon we're talking real money.