The growing danger of a Delta levee failure is reviving the highly controversial Peripheral Canal as a way to protect a drinking water source for more than 22 million Californians.

After a rare summer flood dramatically illustrated how a break in one of the region's fragile levees can imperil water supplies throughout the state, a pair of leading scientists say the canal plan, which was bitterly opposed in Contra Costa two decades ago, ought to be considered.

Few policymakers are ready to propose building a peripheral canal -- the catchall term for any structure that takes river water upstream of the Delta and carries it to water users in a way that bypasses the Delta.

It was, after all, one of the most controversial engineering proposals in state history.

"The politics of the Peripheral Canal were so poisonous it was taken off the table," said Jeffrey Mount, a UC Davis geologist and an authority on California rivers and levees. "I've never heard a rational explanation for why."

The problem with relying on the Delta for drinking water is that many of the region's sunken islands sit 10 to 20 feet below the surrounding water. And many of the levees holding back occasionally brackish water were built in the 19th century. Already old and fragile, the earthen berms are now considered even weaker than previously thought, Mount said.

When those levees fail, as they did at Jones Tract this summer, the breach can gulp so much water that seawater can wash into drinking water supplies, a very difficult and expensive problem to fix.

Circumventing the Delta and the levees with a canal could protect drinking water for 22 million Californians, from the South Bay to Southern California.

Mount and his UC Davis colleague, fisheries biologist Peter Moyle, said they are not necessarily proposing that a canal be built, only that a modified version of it be considered.

"It's the obvious solution to the problem," said Moyle. "My concern is that if you get this major collapse of Delta levees, it would cost billions of dollars to fix. And, realizing (the fix) would be on a temporary basis, suddenly the economics of the Peripheral Canal look really good."

It is unclear how much it would cost to build a conduit around the Delta, but the last official estimate before voters defeated the Peripheral Canal in 1982 was $2.5 billion, according to historian Norris Hundley's book, "The Great Thirst."

By comparison, raising the levees to modern safety standards would cost about $1 billion, according to a 1999 study that almost certainly underestimates what the cost would be today.

But that fix would simply enlarge the levees. It would do nothing to shore them up against major earthquakes and little to address the higher sea levels and bigger floods that are expected in a changing climate, said Mount. Fortifying the levees against earthquakes would require replacing substandard Gold Rush-era foundations with engineered structures, an extremely expensive endeavor.

A peripheral canal has other potential benefits, Moyle said. For example, the West Coast's largest estuary could be managed less as a drinking water source and more like a natural estuary with more dynamic flows and changing salt levels that could serve as habitat for native species. It is unclear how well it could be restored, since the Delta has been so dramatically altered over the years.

And any plan to build a canal faces huge political obstacles and re-ignites passions that have been largely dormant, especially in Contra Costa County, where 500,000 people get all their tap water from the Delta.

"It's come up, it's come up, it's come up. But as far as them whipping us on it, I don't see it," said Joseph Campbell, a member of the anti-Peripheral Canal steering committee in 1982 and now president of the Contra Costa Water District. "I hate it."

Rare flood offers warning

June 3 was a reminder of why the levees matter and a close brush with a serious statewide emergency. Under the pressure of an unusually high tide that day, a levee in the southern Delta that was believed to be in good shape crumbled. So much water gushed so fast onto the Jones Tract, eventually flooding 11,000 acres of farmland and displacing migrant workers, that salt water from the Bay began to creep near Delta pumps.

Operators were forced to crank down the massive pumps near Tracy for six days to prevent salt from fouling the water supply. Cut off from their Delta source of water, San Joaquin Valley farmers and Southern California cities had to rely instead on groundwater and Southern California reservoirs -- backup sources inadequate to meet sustained demand.

The summer flood was made more complicated because flooding on one island raises the risk to other nearby levees and islands. The day after the Jones Tract levee broke, for instance, crews rushed to shore up a second levee protecting tens of thousands of acres of additional farmland. Leaks and seeps sprouted on adjacent islands.

One of the gravest fears of water operators -- a cascade of levee failures -- appeared possible.

"As the flooding of Jones Tract demonstrated, there are many known and unknown risks that can reduce our ability to import water from the Delta," Tom Clark, general manager of the Kern County Water Agency, wrote recently to the editor of his local paper.

Clark, who is retiring, endorsed a plan for a scaled-down peripheral canal to augment existing plumbing. That plan has been shelved since 2000, but is due to be reconsidered by state water officials as early as 2007.

Clark did not respond to a request for comment.

Water-grab fears

In Northern California, there is concern that any new infrastructure with the potential to move more water south will eventually do just that, and in Southern California there is reluctance to stoke those fears.

The largest urban water agency in the state, for example, has studiously backed away from any discussion of a new canal.

"You're going to have to drag my agency kicking and screaming, because we're so afraid of what it's going to do to North-South relations," said Tim Quinn, vice president of the Metropolitan Water District of Southern California.

Anglers worry about the negative effect on fisheries, while environmentalists are concerned that Delta ecosystems would be further damaged.

There are other concerns, as well.

Without the South's reliance on the Delta for drinking water, statewide support could dwindle for maintaining the levees that protect a half-million Contra Costa residents' water supply, which would probably continue to come from the south Delta.

Delta farms, which predate the projects that move water to Southern California, would face the same problem.

Bad odds for Delta

Over the last year, Mount's research has shown the pressure on levees growing as the islands continue to sink and the sea level rises. Without corrective action, the pressure will continue to build.

Moreover, a major earthquake might lead to abrupt, multiple levee failures that are likely to overwhelm any emergency response efforts.

Mount calculated a two-in-three chance that sometime in the next 50 years Delta levees will fail and lead to significant flooding.

His work has not been formally reviewed by other scientists, but it is generating interest in scientific and water policy circles and comes at a time when California is increasing, or at least cementing, its reliance on the Delta.

The U.S. Bureau of Reclamation, for example, is in the process of renewing more than 100 25-year contracts for nearly 2 trillion gallons a year from the Delta system. The proposed contracts are very similar to the contracts they will replace and do little or nothing to reduce the state's reliance on the Delta.

At the same time, the California Department of Water Resources is expected in the next several months to move ahead with a plan to ramp up its pumping capacity by 27 percent to increase the reliability of water supplies.

Canal repackaged

The original Peripheral Canal was envisioned as part of the State Water Project, the massive plumbing system that includes a dam in Oroville, pumps at Tracy and an aqueduct that carries water 444 miles to Riverside County.

As originally designed, the canal would have diverted water from the Sacramento River just south of the state capital into a new channel large enough to carry the entire river in the dry fall months.

The water would have traveled in an arc around the east side of the Delta, arriving at pumping stations at Tracy cleaner than water that travels through the Delta. At various points along the way, water would be released from the canal into rivers feeding the Delta to address environmental concerns.

After the state Legislature and then-Gov. Jerry Brown approved the canal in 1980, incensed opponents put a referendum on the ballot.

Former Contra Costa County Supervisor Sunne McPeak was a co-chairwoman for the opposition group. McPeak, now a Cabinet secretary in Gov. Arnold Schwarzenegger's administration, did not respond to requests to her office for an interview.

Delta farmers, environmentalists and Northern California residents strongly opposed the canal, and the referendum killed the proposal in 1982.

But the idea did not go away.

In the early '90s, conditions in the Delta were reaching a crisis. Fisheries were collapsing because of overpumping. Federal regulators were threatening to curtail water deliveries south of the Delta unless something was done. A broad-based effort called CalFed was begun to fix the Delta.

One proposed solution -- with the new, less threatening name "isolated conveyance facility" -- would be to build a pipe smaller than the old proposed canal.

The final CalFed decision in 2000 gave the pipe strong consideration. But the "conveyance" was ultimately "rejected as infeasible due to social and technical considerations, based in large part due to the contentiousness" it would foster.

But CalFed added that a new canal proposal could not be ruled out entirely, and might be reconsidered as early as 2007. In the meantime, CalFed would maintain and upgrade levees, improve water quality and restore ecosystems and increase Delta pumping capacity. If that doesn't improve fisheries, water quality and water supply reliability, the "conveyance" will come back into play.

"After that first stage, we're going to have to evaluate: ... Is the through-Delta system the right way to go?" said Patrick Wright, who leads CalFed as executive director of the California Bay-Delta Authority.

Wright added a cautionary note.

"I think people need some time to digest this material (from Mount) before they start leaping to conclusions about plumbing."

Canal alternatives

Dealing with the Delta's instability does not necessarily mean that a new water delivery system must be built. Maybe there's a third way.

"I think it's sad that the only thing people want to talk about is a peripheral canal," said John Cain, a restoration ecologist at the Natural Heritage Institute. "We don't think about Delta fragility or the seismicity of the Delta because of the P-word."

Cain suggested that new levees could be built to divide islands up like pie slices. That way, if there is a flood, only a portion of the land would be flooded.

Another idea is to start rebuilding the islands to higher elevations by planting tules or using rice straw bales. Unfortunately, filling the islands with dirt or rock will not work, because that material is heavy enough to crush the peat and cause further sinking.

"I don't see the Peripheral Canal as a solution to this problem," Cain added. "I see the Peripheral Canal being a walking-away from this problem."

Still, Wright, Cain, the UC Davis scientists and others say it would be a good idea to begin coming up with a plan for dealing with the levees now, instead of waiting for the calamity that Mount has forecast.

"We think it makes sense to start this dialogue now, rather than waiting for the Delta to deteriorate," said Jonas Minton, a former deputy director at the California Department of Water Resources and now a water analyst for the Planning and Conservation League.

"We think it's the sort of issue where Californians should not stick their heads in the peat," he said.