Hetch Hetchy is expendable, new tool finds
By Tom Philp — Bee Associate Editor
Published August 29, 2004

Can a computer see things that conventional wisdom overlooks?

A University of California, Davis, graduate student named Sarah Null took a new computer model that analyzes water management and asked the computer a century-old question: Does San Francisco really need to rely on a dam in Yosemite National Park? In 1913, Congress said yes. It allowed the city to construct the dam in one of Yosemite’s two glacial valleys, the one known as Hetch Hetchy. The dam is on the Tuolumne River. Since 1923, Hetch Hetchy has been the city’s primary water supply.

Today, however, three other major reservoirs occupy this river. Null and her faculty adviser, Jay Lund, the inventor of the computer model known as CALVIN, did what John Muir could only dream of. They assumed the dam at Hetch Hetchy wasn’t there. Left intact were the other three reservoirs. They plugged in more than 70 years of historical river flow data. They made minor changes in the plumbing. And they calculated how much water the different system could deliver compared with the existing one. It was nearly the same. That surprised Null and Lund at first. But the closer they looked, the more it made sense. Null and Lund explain:

BEE: How did you get the idea of analyzing Hetch Hetchy and the impact on San Francisco, Modesto and Turlock if this reservoir no longer existed?

LUND: The Hetch Hetchy system is a classic example. You go out and protect a watershed. You pipe it in from a high elevation. You make hydropower. You have essentially no operating costs for energy. You get very good quality water. And you don’t need a lot of technology. So in the early 1900s, from an engineering decision, it was beautiful. Absolutely beautiful. But a lot of time has gone by. If you were to do it today, you would do it differently.

NULL: You wouldn’t have a dam in a national park …

LUND: … Because you would have this other storage on the system. Hetch Hetchy, for California standards, is not a large dam. Hetch Hetchy is only 360,000 acre-feet. And downstream is a big reservoir. Two million acre-feet.

BEE: Going into this, did you have sentiments, personal sentiments, the longing to restore Hetch Hetchy, or was this more of an academic curiosity?

NULL: Probably mostly an academic curiosity. I have somewhat of an environmental lean, but not to the extent of taking out water supplies for large cities.

LUND: I thought it was a good opportunity to see how you might be able to modernize the operation of a significant part of the California water system.

BEE: Briefly describe the Tuolumne water system.

LUND: Basically, you have this huge reservoir, 2 million acre-feet, New Don Pedro. Upstream, there is Hetch Hetchy at 360,000 acre-feet. And Cherry and Eleanor, which combined are about 300,000 acre-feet. So that whole system has more than 2.5 million acre-feet of storage on a river that has about 1.8 million acre-feet of water as an annual flow. So already you have a system that is not poor for storage. It may be even a little bit wealthy on storage.

NULL: And there is local San Francisco storage.

LUND: Plus there is 100,000 acre-feet of additional storage in the San Francisco area.

BEE: The ownership of these reservoirs up there. San Francisco owns and runs Hetch Hetchy. Who owns and runs the others – the two other high-country reservoirs, Eleanor and Cherry – and who runs the big one downstream, New Don Pedro?

NULL: San Francisco owns Cherry and Eleanor, the two other high ones. And Turlock and Modesto irrigation districts own and operate New Don Pedro.

BEE: Could you describe what CALVIN is?

LUND: CALVIN is a model that does all the water balance accounting. So it makes sure the water flows and agrees with the laws of physics. It limits the operation of the system within the physical constraints that you have, the capacity of reservoirs, pipelines, things like that. And then it operates within that range of feasible operations to maximize economic benefits.

BEE: So in this case, it knows how much historical rainfall there is, how much storage the system has, how much you’re taking out if you take out Hetch Hetchy and how much it can then deliver?

NULL: Right.

BEE: Other than assume there was no Hetch Hetchy Dam, what other plumbing changes did CALVIN make?

LUND: For us, the only real creative thing we did was we added this inter-tie between New Don Pedro and the Hetch Hetchy aqueduct (to San Francisco).

NULL: If you assume that there is an inter-tie between New Don Pedro and the San Francisco aqueduct, according to CALVIN, there is very little difference in water supply.

BEE: So this inter-tie is basically a new pipe that connects this big downstream reservoir, New Don Pedro, to the San Francisco system. At the moment, San Francisco does not take water directly from New Don Pedro.

LUND: That’s correct. There is not a physical ability to make that.

BEE: So how far away are these pipes that supply San Francisco from New Don Pedro?

NULL: They cross New Don Pedro.

BEE: So physically, it is not a huge challenge to connect the San Francisco water system to New Don Pedro.

LUND: It would probably require a pipeline and a small pumping plant.

BEE: So CALVIN assumed that if you no longer had a dam in Hetch Hetchy, San Francisco would have to take at least some of its supply directly from this inter-tie in New Don Pedro.

NULL: Right. Essentially, CALVIN just moves the storage of Hetch Hetchy down to New Don Pedro. And because New Don Pedro is such a large reservoir, there aren’t many differences.

BEE: Were you surprised at all at what CALVIN found?

LUND: Yeah, actually.

NULL: I was, too.

BEE: It seems hard to grasp that you can remove a dam, a water supply dam, and not really impact water supply abilities very much. Can you explain how you can get rid of this dam without significantly impacting the water supplies for Modesto, Turlock and San Francisco?

NULL: This is a unique case.

LUND: Storage is not water. If you have a system which is rich in storage, if you take some of that storage away, you’re not taking away any of the water. In the case of the Tuolumne River system – including Hetch Hetchy, New Don Pedro, Cherry and Eleanor – it is well off in storage. So if you take away some of the storage, it doesn’t affect the water supply deliveries very much.

BEE: Did CALVIN assume that San Francisco would still divert water into its pipelines and tunnels that exist just below the Hetch Hetchy dam as the water flows down the river?

LUND: Yes.

NULL: During the spring runoff, the Hetch Hetchy aqueduct would still be full in most years.

BEE: So you are still diverting water from the same pipes just below the valley. Based on CALVIN, you’re just not impounding the water in the valley.

NULL: Right.

BEE: Has San Francisco contacted you to explore your findings?

LUND: Nope.

BEE: These findings don’t exactly follow conventional wisdom in San Francisco.

LUND: There is an impression that there is only Hetch Hetchy and its 360,000 (acre-feet) of storage for San Francisco. Where, in effect, there is potential for almost 3 million acre-feet of storage.

BEE: Your CALVIN computer seems to be leaving out a little bit of the politics.

NULL: All of the politics.

BEE: The current political arrangements allow San Francisco to run its high-country dams – Hetch Hetchy, Eleanor and Cherry – and downstream, for Modesto and Turlock to run their dam. CALVIN is suggesting that, in order to eliminate the dam in the national park, San Francisco, Modesto and Turlock would have to essentially join forces, cooperate together and manage together the remaining three reservoirs.

LUND: They would all have to enter the same room, talk and leave happy.

BEE: But CALVIN is suggesting that it is less a question about adequate supply and more of a question about new political arrangements to keep everybody whole.

LUND: Yes. There seems to be enough water in the system. And there seems to be enough storage in the system if the parties can come to agreement to re-operate and reallocate the benefits in the system.

BEE: If you’re looking at this from the Modesto or Turlock perspective, what’s in it for them?

LUND: The farmers are certainly in a good negotiating position to have quite a favorable contract if they chose to go that route.

BEE: What about global warming? CALVIN used historical rainfall data. What if the future portends more rain and less snow? Would the system, if it didn’t have the dam in Yosemite, run short?

LUND: I think there is some risk in that. We have not run climate change studies. I think that would be something that would merit further examination.

BEE: Did you explore its impact on flood control? Would you have to use flood control space in New Don Pedro in order to capture the water?

LUND: We used the same flood control space as the Corps of Engineers currently has for New Don Pedro.

BEE: So what should the outside world take from your computer’s advice, about the real-world feasibility of restructuring water supplies and restoring this Yosemite valley?

NULL: People have talked about it for a long time, but there was very little done quantitatively, looking at the numbers, seeing what could be done, what was possible.

LUND: A lot of time the political discussions and the public discussions are not so well-informed on that score. There is a lot of fear that any change in a water system would be bad. Sometimes that is not the case.

Published August 29, 2004

Seventeen years ago, Interior Department Secretary Donald Hodel had a provocative idea for Hetch Hetchy, Yosemite Valley’s smaller twin:

Dismantle the dam that has kept the valley underwater since 1923, thus restoring the granite peaks and signature waterfalls to the national park system and the American public.

President Reagan’s appointee met a reaction as swift and mighty as a wall of water unleashed by a storm. He didn’t have a sound alternative for replacing San Francisco’s water supply, which Hetch Hetchy largely provides. It was no surprise that his plan for Hetch Hetchy soon died.

What Hodel needed to make his case didn’t exist then, but it does today. That ally is CALVIN, a new, water-modeling computer program also known as the California Value Integrated Network.

With a blissful ignorance of politics and conventional wisdom, CALVIN concerns itself largely with two questions: How much water can be delivered, and with what plumbing?

Using state and federal dollars, the University of California, Davis, invented CALVIN in 2001 to calculate how changes would affect a water system. It has come in handy in other California water quandaries thanks to its dispassionate, outside-the-box view of the world.

Last year, the minds behind CALVIN tried an interesting exercise. They programmed CALVIN to consider Hodel’s idea. CALVIN punched a virtual hole in a virtual Hetch Hetchy dam. It added a virtual pipe and a virtual pump downstream. CALVIN then calculated whether San Francisco would be short of water.

The results surprised its human operators. CALVIN found minimal impact. Hetch Hetchy’s dam, CALVIN announced, is expendable.

How could that be? CALVIN examined the flow of the river, the Tuolumne. It examined its four dams and, based on the river’s typical flow, concluded that the other three dams could do the job.

Besides Hetch Hetchy, the Tuolumne’s flow is interrupted by the Cherry, Eleanor and New Don Pedro dams. San Francisco owns Hetch Hetchy, Cherry and Eleanor. Hetch Hetchy provides nearly 85 percent of the city’s water and a large portion of the water for Alameda, Santa Clara and San Mateo counties. Irrigation districts for the Central Valley communities of Modesto and Turlock own New Don Pedro, which can store 5.6 times the water Hetch Hetchy can.

New Don Pedro rests alongside San Francisco’s existing pipeline system from the Sierra, but they are not connected. CALVIN, applying a computer’s cold-eyed logic to the situation, connected them.

They aren’t connected today because of politics. Legal agreements meticulously divide the Tuolumne River’s water among Modesto, Turlock and the Bay Area. Since 1913, when Congress allowed San Francisco to build the dam in Yosemite National Park, four legal agreements have governed the water distribution. Draining Hetch Hetchy would require a fifth agreement. It would need to allow San Francisco to draw its supply downstream and outside the park, from New Don Pedro instead of Hetch Hetchy.

Computers don’t write legal agreements. Lawyers do, ones hired by water district leaders. These lawyers are a risk-averse breed. They crave certainty. They trust concrete.

Their instincts serve them well in many cases, but not in all. San Francisco is planning to replace a local reservoir in the East Bay’s Calaveras hills with one that has potentially more capacity than Hetch Hetchy. New Don Pedro has the potential to be raised slightly to add even more storage.

The prospect of “new storage” in exchange for eliminating some “old storage” at Hetch Hetchy offers a kind of balance at a time when California continues to weigh the competing interests of the environment and development. CALVIN wouldn’t appreciate the symmetry in the least. It deems the proposed East Bay dam unnecessary. But CALVIN wouldn’t have the last word. It has done its job, which is to reveal whether a river system is flexible enough for change. This one is.

Secretary Hodel’s idea seemed like folly back in 1987. Today, CALVIN reports that his wasn’t an outlandish proposal after all. A Yosemite National Park with two spectacular valleys wide open for the public? Twin valleys reunited? Hetch Hetchy regained?

Imagine the possibilities. Donald Hodel did in 1987, though unsure of how to make them a reality. Californians can imagine them again today, with the knowledge that they are within reach.