Officials patched and prayed while pressure built on dams
A freak March storm dumped 10 inches of rain on Houston in just six hours, causing widespread flooding. The vast reservoirs behind Addicks and Barker dams, already swollen from earlier downpours, filled with fresh runoff.
The earthen dams had been Houston’s main defense against catastrophic floods since the 1940s. Never before had they held back so much water.
The year was 1992, and the Ash Wednesday deluge put a scare into city and county leaders and the U.S. Army Corps of Engineers. The dams were not built for this. Development upstream was sending more runoff into the reservoirs. Development downstream limited how much could be released through the dams’ outlets. Something had to be done. The Army Corps studied the possibilities and, in 1995, listed 10 options. They included deepening the reservoirs, adding new reservoirs and buying out property owners. Which did the Corps recommend?
“Accept existing conditions and risk through No Action.”
Time and again over the years, the region’s public officials and Army Corps leaders — presented with evidence of the mounting pressure on Addicks and Barker — opted to play for time.
In a series of reports since the early 1990s, experts outlined the problems and identified potential remedies. Each time, decision makers concluded that the situation wasn’t quite bad enough to justify major corrective measures.
In August, Hurricane Harvey dumped more than 50 inches of rain on Houston, pushing Addicks and Barker to the brink. To prevent a breach, the Corps released billions of gallons into an already flooded city, inundating thousands of homes downstream. Upstream, thousands of other properties were drenched by the reservoirs.
Harvey’s overwhelming force was the immediate cause of the crisis. But decades of official inaction and procrastination played an important role, the Chronicle found after reviewing public records on the dams and interviewing experts and officials familiar with their history.
If the county, the state and the Corps had pursued even some of the remedies proposed over the years, the devastating releases of water during Harvey might not have been necessary — or might not have had to be so large.
“The floods come up. But then the sun comes out,” said U.S. Rep. Ted Poe, a Houston Republican whose district includes areas upstream and downstream of Addicks. “The
Addicks and Barker dams have protected Houston since the 1940s. Development and bigger storms have stressed the earthen structures. Hurricane Harvey pushed them to the brink — and exposed their vulnerabilities.
political rhetoric is that we’re going to do something, and then nothing ever happens.” ‘Accept the risk’
A year after the Army Corps opted for “No Action” in response to the Ash Wednesday flood, its local partner, the Harris County Flood Control District, outlined nine possible steps to lighten the load on the dams and protect communities near the reservoirs.
The 1996 study highlighted a “unique, once-in-a-lifetime opportunity” to construct a tunnel beneath Interstate 10 to carry water out of the reservoirs to a discharge point 12 miles east. The cost was estimated at $325 million to $400 million.
“The potential flood control problems are severe enough to consider this magnitude of project,” the report said.
The final option listed was: “Do nothing and accept the risk of flooding.”
That’s the one county officials ultimately chose.
“It’s easy to look back now and say we should have done this and should have done that,” Harris County Judge Ed Emmett told state lawmakers after Harvey. “We could have done better. I won’t even argue that point.”
The Army Corps took steps over the years to reinforce the dams after finding signs of deterioration. Today, a $105 million project is underway to replace the outlet tunnels and main spillways at both Addicks and Barker.
But addressing the dams’ fundamental problem would require action on a much bigger scale. Addicks and Barker are holding larger quantities of water for longer periods than their designers intended. That stresses the structures and reduces their margin of safety to handle intense storms.
The potential answers are no mystery. Relief canals would move water out of the reservoirs. A third reservoir would intercept storm runoff before it could reach Addicks and Barker. Buying out properties downstream of the dams would allow larger releases of water into Buffalo Bayou.
But Houston leaders were unable to forge consensus on such big-ticket projects, even as storms produced historically large reservoir pools again and again through the 1990s and the 2000s.
Edmund J. Russo Jr., deputy district engineer for programs and project management in the Army Corps’ Galveston district, said in a statement:
“Many of these proposals and subsequent decisions were made over a period of 70 years. Many simply were not authorized or funded. … What’s important now is that we continue to focus on working with city, county and state officials to propose solutions that reduce flood risk in the future. I think what we are seeing is a concerted and collective effort by our elected officials to make this a reality.”
Christopher W. Sallese served in the Corps for 27 years, including three as commander of the Galveston district, overseeing the dams. He said it would have been difficult to make the improvements outlined in the Corps’ 1995 report, given the many constituencies affected and the competition for federal money.
But doing so would have left Houston better prepared for Harvey, he said.
“If you could have had the synergy to bring all the pieces together, the private residents, the local government, the state and federal pieces … and done things you wanted to do, the losses would have been less,” Sallese said. “I believe that.” ‘Darn lucky’
The price of postponement was put in high relief eight years before Harvey. Addicks and Barker, in 2009, were placed in the highest-risk category of federally operated dams: DSAC I, signifying “very high urgency.”
Army Corps experts had discovered voids in the embankments and under the main spillways, elements in a “progression toward failure.”
The Corps made interim repairs to fill the voids. The $105 million project to replace the dams’ outlet works — the system of discharge tunnels, spillways and gates — began in 2015. Instead of five concrete tunnels at each dam, three steel pipes will discharge water, a more structurally sound method, said Richard Long, one of the Corps officials overseeing the work.
The project is scheduled to be completed in 2020. Even then, the dams likely will retain their high-risk rating. In part, that’s because other problems have yet to be addressed, including potential weaknesses in the dams’ emergency spillways.
Of more than 700 dams operated by the Corps, only 13 others are in the same high-risk category, according to a Corps presentation. None of them protects an area as populous as Houston.
The emergency spillways are designed to serve as a safety valve in an extreme storm, an escape route for water that otherwise could flow over the dams’ crests. Inspections in 2012 found that the spillways’ concrete slabs were cracking and separating at the joints.
Harvey exposed other vulnerabilities. For the first time in the history of Addicks, water rose high enough to escape around the dam’s northern end, where the embankment slopes down to meet the natural terrain. Unchecked, that cascade could have caused serious damage to adjacent neighborhoods and the embankment itself.
Army engineers have studied dozens of scenarios for how the dams could fail, and several involve water coursing around the northern end of Addicks or over the emergency spillways. The Corps opened the dam’s flood gates during Harvey in part to prevent either from happening.
The Corps has been seeking funding since 2013 to study both problems.
“I think we were very fortunate and lucky that those two reservoirs didn’t breach,” said U.S. Rep. Michael McCaul, a Republican who represents the Northwest Houston area and who chairs the House Committee on Homeland Security. “We’re darn lucky that something that was built — what, 70 years ago — was able to hold up to this.”
Long said that despite the dams’ problems, they withstood Harvey’s battering with only “limited issues.” Asked if they could stand up to the same punishment again, he answered without hesitation.
“Will it still have a large impact to the community? Yes, it will,” he said. “But we expect the project to perform as well as it did this time.” ‘Devastating scenario’
Addicks and Barker made modern Houston possible. Few pieces of U.S. infrastructure are so vital to a major metropolitan area. The dams safeguard a population of more than 1 million, the Texas Medical Center and petrochemical facilities that form the epicenter of the U.S. energy economy.
Failure scenarios developed by the Army Corps spell out, in granular detail, the consequences if either of the dams gave way. The projected toll in lives varies according to assumptions about the water level in the reservoirs, how many people would heed evacuation orders and how fast they would move to higher ground, among other factors.
A 2010 scenario for a collapse of the outlet works at Addicks projected 2,573 deaths in a “bestcase” scenario.
In the worst case, 6,928 would perish, according to the analysis.
More than 120,000 structures would be flooded. Property damage would total $22.7 billion.
The long-term effects would include “the reduction of a community’s image as a dependable and viable entity, and a reduction in its ability to provide basic services,” the analysis said.
Yet, many Houstonians give little thought to the dams, a testament to how well they’ve performed for more than 70 years.
On a recent cool afternoon at Briarbend Park, 8 miles east of the dams, Lynn Tan sat on a bench, writing in her journal as she often does before or after her shift at a nearby grocery store.
Tan, 30, likes the mix of tranquility and bustle: the joggers, dog walkers and children practicing soccer. Buffalo Bayou flows along the park’s northern boundary. She went out on the water once in a canoe.
It doesn’t cross her mind that the park and the neighborhood around it — the Trader Joe’s, The Buckingham retirement community, Second Baptist Church, Mark White Elementary School and hundreds of homes and businesses — depend on the dams for their survival. If Addicks and Barker failed, all of this and more could be wiped out.
The surge of water likely would devastate everything between the dams and downtown: the Galleria, Montrose, River Oaks, the Fourth Ward.
“That would be pretty bad,” said Tan, who moved to Houston a decade ago from Pittsburgh. “That’s the first time I ever heard of that.”
Outside a Kroger a few blocks away, Rachel Green, 26, was packing the trunk of her white Chevy sedan with the makings of a nachos dinner for herself and her mother.
Green, a software engineer, grew up in Beaumont and knows about Houston’s dams. But she wasn’t aware of their high-risk rating or what could happen if they failed.
“Goodness,” Green said. “If there are steps we can take to remedy this, that devastating scenario, then I want to know.” Taming the bayou
Addicks and Barker grew out of a disaster: the 1935 Great Flood of Houston, which laid waste the city’s business district and killed seven people. Buffalo Bayou rose 10 inches an hour for 24 hours, cresting at 52 feet above sea level at the Capitol Street Bridge, nearly nine times its normal level.
Beginning in 1942, the Army Corps dug up 8.4 million cubic yards of prairie and sculpted huge embankments to intercept storm runoff from across the Buffalo Bayou watershed. The government acquired 25,000 acres of woods and wetland north and west of the dams, an area nearly twice the size of Manhattan, to serve as reservoirs.
At the time, the dams were among the longest earth-filled barriers in the world. Addicks extended 11.6 miles, Barker 13.6.
Barker was finished in 1945. As Addicks was nearing completion in 1948, Houston magazine proclaimed in capital letters: “DAMS WILL END CITY FLOODS.”
Political and financial pressures would undercut that promise.
The dams were supposed to be the centerpiece of a broader flood control system, including a third dam and reservoir on White Oak Bayou and two canals to carry water to the San Jacinto River and Galveston Bay.
Those other components never materialized. “Rising land costs and rapid development” made them “impracticable,” according to a Corps history.
A planned levee south of Cypress Creek, to hold back storm water headed for the Addicks watershed, also was shelved. Instead, the Corps increased Addicks’ storage capacity slightly.
Under the protection of the dams, the city marched north and west. In 1950, Houston was the 14th largest city in the country, with nearly 600,000 people. By 1990, it was the fourth largest, with 1.6 million.
The very growth they made possible put the dams in a vise.
The original design called for four of the five discharge tunnels at each dam to remain open all the time. Water would be stored
behind the dams only for brief periods.
But a free flow impinged on the increasingly dense development along Buffalo Bayou. Property owners complained, and the Corps installed gates on two of the four open conduits at each dam to control releases. In 1963, the remaining two were gated.
The bayou’s limited carrying capacity was an additional constraint. The original plan for Addicks and Barker called for enlarging and straightening 29 miles of the waterway. The project had gotten through 7 miles when the Corps shifted its attention to Brays and White Oak Bayous. Opposition from environmentalists brought the Buffalo Bayou project to an end in 1971.
Eventually, the permitted flow out of the dams was reduced to one-eighth what their designers envisioned.
While less water was leaving the reservoirs, more was coming in. Prairie and ranchland that had once served as a natural sink were gradually covered with subdivisions and strip malls. More runoff flowed downstream, toward the dams.
The larger, longer-lasting pools caused water to seep into the embankments, creating pockets of soft earth. The Corps reinforced the structures in the late 1970s and early 1980s.
Then came the Ash Wednesday storm of March 4-5, 1992, the last in a four-month series of downpours. The reservoirs filled to record levels. It took months for them to drain. Flooding the fringe
The storm highlighted the risk to suburban neighborhoods on the western edges of the reservoirs.
In the 1940s, the government bought enough land to contain a 70-year flood in Barker and a 100-year deluge in Addicks. But the dams were capable of holding back more water than that. If enough accumulated, the reservoir pools would spill beyond the government-owned area. The dams’ designers didn’t think that would ever happen, but in 1992 it nearly did.
By then, there were thousands of homes and businesses in what the Corps called the “fringe” area between government property and the maximum reservoir pools. After the Ash Wednesday storm, Army engineers considered what could be done about the problem.
They laid out the choices in a “Special Report on Flooding” issued in 1992. The reservoir bottoms could be excavated further to increase their capacity. The government could buy out property owners. It could find ways to detain more water upstream.
Or it could do nothing. The engineers spelled out what that would entail.
“In this case, no action would mean accepting the risk that substantial numbers of houses will be damaged by rare, severe flood events,” the report said. “Residents will be forced to evacuate and remain in temporary housing for long periods. Health and safety risks will be perpetuated. The government will continue to be subject to potential claims for monetary losses, and the Corps will be faced with a continuing adverse public image.”
The Corps studied the options further in preparing the 1995 report on Addicks and Barker. Any of the alternatives — other than “no action” — would have protected homeowners in the fringe area and put the dams in a better position to cope with Harvey.
But the Corps wasn’t looking that far ahead.
It subjected each proposal to a rigorous cost-benefit analysis. For example, deepening even a small portion of Barker reservoir was estimated to cost $11 million. Doing nothing would result in only $78,000 a year in property damage, barring an extreme storm, the 1995 report said. So the proposal was deemed “nonviable … from the standpoint of economics.”
The report concluded that “flooding problems associated with the reservoirs” are “not yet as severe as first thought.” The study of possible solutions would “be terminated because of insufficient economic benefits to justify project modification.”
In the years ahead, thousands more homes would be built in the reservoir fringe, and many of them flooded during Harvey. Hundreds of homeowners have sued the Army Corps, contending that the flooding was an illegal seizure of their property.
In 2000, the Harris County Flood Control District hired the engineering firm Costello Inc. to assess ways of protecting communities on the reservoirs’ western border. ‘A certain complacency’
The study looked at a by-thenfamiliar list of answers: buy out homes threatened by the reservoir pools, excavate the reservoirs, connect them via a tunnel that could balance their water levels, build pipes to carry water from Addicks and Barker to White Oak Bayou or the Brazos River.
The report included cost estimates for various proposals: $123 million to deepen the reservoirs, $588 million to buy out adjacent properties.
The conclusion: None of the six options studied was “warranted.”
In 2011, the flood control district began examining how Cypress Creek was affecting Addicks. During heavy storms, massive amounts of water overflowed from the creek into the Addicks watershed. The problem was expected to get worse. The population of northwest Harris County, then about 340,000, was projected to double over the next 50 years.
The district, in 2015, issued a report on possible remedies. It ran more than 1,000 pages. The two preferred solutions centered on building new reservoirs upstream of Addicks to collect Cypress Creek’s overflow.
The county has not moved forward on the proposals.
Matthew Zeve, director of operations for the flood control district, said early cost estimates ranged from $300 million to $500 million, 3 to 4 times the district’s annual budget.
Referring to the 2015 report and earlier studies, Zeve said, “Many of the ideas mentioned ... may have merit. They didn’t move forward due to funding concerns and/or lack of authorization for the Corps to carry out their processes.” He said the flood control district “supports the Corps, but it is up to Congress to authorize them.”
Russ Poppe, executive director of the district, said the aversion to large-scale remedies has to change post-Harvey.
“A ‘do nothing’ alternative is not sustainable,” he said. “The fact is it took 70 years since the reservoirs were built before we experienced this event, and we believe a certain complacency had developed since the reservoirs were doing their job and doing it well.” ‘Patch and pray’
Since Harvey, political leaders have resurrected some of the same proposals that languished for years. They vow to get them done this time.
A majority of Harris County Commissioners Court has signaled support for a bond issue, possibly $1 billion or more, that could help pay for a third reservoir and other projects. Voters would have to approve.
City and county officials have called on Gov. Greg Abbott to tap the state’s $10 billion “rainy day” fund to finance the third reservoir. State leaders are looking to Washington for money. Congress has approved several installments of disaster aid, but none included funds for major infrastructure improvements in the Houston region.
Abbott has asked Congress for an additional $61 billion for hurricane recovery across Texas. The outlook is uncertain at best.
None of this surprises Robert G. Bea, a professor emeritus of engineering at the University of California-Berkeley. A former engineer for the Army Corps and Shell Oil, Bea lived in Houston for 11 years, in a Memorial-area home near Buffalo Bayou — one that flooded during Harvey.
Bea investigated the failure of New Orleans levees and the nearfailure of California’s Oroville Dam earlier this year. He said the story of Addicks and Barker — of missed opportunities and remedies deferred — reflects a “patch and pray” approach to maintaining vital infrastructure.
“We do understand it could fail, but we’ve chosen to normalize that potential of failure,” Bea said. “We’ll tolerate Harvey, we’re all upset about it for a while, until it kind of dies down, and then we go back to business as usual until we have the next one.”