A doomed Sumas dike failed as predicted. Many other levees could be next.

Photo by Peregrine Aerial Surveys

• Our comprehensive story on the challenges Fraser Valley dairy farmers are facing

• Our comprehensive story on the history of Sumas Prairie, Sumas Lake, and the Nooksack River

Unacceptable.

That was the verdict of engineers five years ago after they inspected a critical dike on Sumas Prairie. The 17-kilometre levee had a key job: stop the drained basin of what was once Sumas Lake from refilling with 10 feet of water. But inspectors warned the dike was two feet lower than it should be, and that it would be unable to stop flood waters if and when the Nooksack River breached its banks.

The engineers, who were hired by the province to review dikes all around the Lower Mainland, gave the crest (or height) of the Sumas dike the lowest possible grade—one out of four—and a corresponding label: “unacceptable.”

“Overtopping is expected during Nooksack River overflow,” they declared. “The dike design profile likely needs to be updated. The dike geometry is substandard, causing concern.”

The inspectors were right. The levee was never fixed. And last week, waters from the rain-swollen Nooksack spilled north and ran downhill into Sumas Prairie. By Tuesday, millions of gallons of silty water had pushed over, and then directly through, the so-called “interceptor” dike, forcing thousands from their homes, killing livestock, inundating Highway 1, and submerging one of the most fertile sections of land in the province.

Photos showed a hole in the dike the length of a soccer field. The breaches were so large that it took until Sunday to temporarily fix them.

The failure of the Sumas dike was not a shock for those familiar with the levee. And neither will it be surprising if and when dikes counterparts elsewhere in the Fraser Valley fail their next big test. Because that 2015 report shows the Sumas levee was not a victim of individual neglect; rather, the dike was one of many flawed pieces of a system that is known to be incapable of stopping a major flood—including one that could do far more damage than this year’s catastrophic disaster.

[story continues below]

Failure

The height of a dike is critical not only to stop water from going over it, but also to maintain its structural integrity. A dike is an earthen wall. Its sides are meant to absorb and block water. Its top is not, and most dike breaches occur as a direct result of floodwaters flowing over the crest of a dike (called overtopping). When water travels above the dike, the levee starts to erode, and any small flaw can quickly turn into a gaping hole.

When the Nooksack’s flood waters poured north into Canada, they encountered an area unprepared to withstand so much water. By Tuesday, as Abbotsford Mayor Henry Braun toured the area in a helicopter, he watched as water spilled over the dike across a vast length of multiple kilometres. Speaking to The Current around 3pm that day, he predicted the dike would be breached within hours.

It’s unclear when, exactly, the dikes fell, but by Thursday floodwaters had penetrated the Sumas dike in three different locations, including one estimated to be 100 metres long just west of Highway 1. In the west, it appeared that the flood was receding. But that was not the result of water emptying in the Fraser, but because millions of gallons were pouring into an area to the east that had once been a lake, before it was drained a century ago.

(Read our comprehensive look at the history of the lake, and how the Sumas Valley’s unique geography influences this flood here.)

Had Barrowtown Pump Station failed, the lake would have likely remained until the spring. As it is engineers, city workers, volunteers, and residents will be dealing with the consequences for years.

One inadequate dike among many

So why was the Sumas dike never fixed? One reason: its inadequacy is incredibly common across the Lower Mainland.

The Sumas dike is only mentioned a handful of times in the 233-page 2016 report that declared it unacceptable. If the report had dwelled too much on every faulty dike, it would have been far, far longer.

The report was contracted out by the province and intended to set the stage for a comprehensive flood management plan by the Fraser Basin Council and found that large parts of the region’s existing flood protections were faulty and doomed to fail. Instead of setting off red alarms, the declaration generated relatively little media or public attention. (This reporter did write about it and the Sumas dike’s low rating.) But for those who read it, the study showed the incredibly complex and costly task of adequately protecting the thousands of people living on floodplains across the Lower Mainland.

The study found more than 100km of dikes across the region that were too low to protect from floods. The eastern Vedder Canal dike protecting Greendale in Chilliwack was, and is, unacceptably low. As are the dikes protecting Dewdney. And those that protect Glen Valley. And those that protect Chilliwack from the north. And those that protect Fort Langley. Further west, a huge expanse of dikes protecting South Surrey and Delta from the Pacific Ocean are also substandard.

The Sumas dike was just the first to fail so spectacularly.

[story continues below]

An international problem

In recent years, the City of Abbotsford has focused more on the lingering Nooksack flood danger.

In 2016, the city’s then-Fire Chief Don Beer presented council with a new risk assessment of potential emergencies and disasters that could befall the city. Beer highlighted the threat from the south: “I watch the Nooksack way closer than I watch the Fraser,” he told council.

But, like flood protection efforts across the region, most of the work done in the last six years has involved more planning on paper than on-the-ground upgrades.

A long-dormant international task force on the Nooksack was set to finally reconvene in 2019, after having not sat since 2012. And that year, the City of Abbotsford applied for and received money to study how to prepare for and prevent a Nooksack flood. Cities frequently create such reports both to understand challenges and problems and to lobby senior levels of government for funding or other assistance. The Nooksack report was finished last year.

That wide-ranging report considered whether the dike should be upgraded. It looked at a range of options, including even possibly drilling a hole through Sumas Mountain. In the end, it said upgrading the dike wasn’t the best solution to the Nooksack threat. Upgrading the dike and suitably flood-proofing the region came with a $339 million price tag. It would be only slightly cheaper—$310 million—to build a brand new dike along the border. (Cost aside, that potential solution would likely have angered the Americans because it would have caused water to back up in the US, increasing damage in Sumas, Wash.)

There was a more practical third option: raising the levee along the Nooksack River itself and stopping a flood before it happened. The cost of that, Abbotsford council was told, might be as low as $29 million.

That would make the most fiscal and practical sense. It would prevent a disaster before it happened and help everyone. The only problem? Time, geography, and politics.

The river is in the United States, and the Americans would have to be convinced to do the work. And that, Vaughn Palmer of the Vancouver Sun reported, has historically been a fraught question. The Americans, Palmer wrote last week, “worry that raising the Nooksack levees near the border would simply push the floodwaters downstream and damage the more heavily populated communities near Bellingham.”

The task force met once in the last year, on Feb. 2, 2021. City officials said they were concerned about a lack of maintenance along the Nooksack channel, while the Americans said they were looking to build a floodway in Sumas to better channel the waters. Such a floodway may have been able to reduce the threat of a small flood by better funneling waters into the Sumas River. Whether a floodway would have been able to contain last week’s spillage from the Nooksack seems highly unlikely.

When the last major flood took place in 1990, George Ferguson was mayor of the District of Abbotsford. Ferguson, who also served as mayor after the districts of Abbotsford and Matsqui amalgamated, was a legendary character in the city and intimately familiar with the Nooksack threat. He also knew the area personally: he and his wife lived on Sumas Prairie. When Braun, an old friend of Ferguson’s, ran for mayor, the former mayor told him he needed to do two things to protect the prairie: 1) get a generator in place to power Barrowtown Pump Station in case the power failed for a long period of time; and 2) get Canadian and American officials together to tackle the Nooksack threat.

Last year, the city committed money for a Barrowtown generator. And in December, after receiving the report on the Nooksack threat, council told staff to put pressure on the Americans to get the ball rolling on their side on a long-term Nooksack fix by start doing a cost-benefit analysis on that new levee outside of Everson. Ideally, such an analysis could show the Americans that raising the levees would be worth it on their end. It could also provide a framework for determining how much the Canadians might pay for any such work.

If it seemed like progress, that’s because it was. But it was too late.

Ferguson died four years ago, but his wife continued living on Sumas Prairie. Last week, she was evacuated along with thousands others, as water from the Nooksack poured into her house and basement, where many of George’s papers and memorabilia remained.

A problem that will take years to solve

If the city had not chosen last year to wait for the Americans, and had instead chosen to construct a dike or a floodway itself, the project would still be years away from beginning construction. In fact, if the situation elsewhere is any indication, it would still be just one on a long list of desperately needed flood-protection projects that remain only words on pieces of paper.

Six years after that first report on the region’s dikes, the Fraser Basin Council has yet to publish its full flood management strategy. The most recent draft generated a range of feedback, including comments that it needed to speak more to the challenges facing the many First Nation communities along the Fraser. Some of those communities are in areas that aren’t currently protected by flood management structures.

And levees aren’t alone a solution.

Steve Litke, the Fraser Basin Council’s lead on the project, noted that flood protection structures like dikes are just one of a complex series of preparations that need to be considered. Another key is land use planning: building cities in ways that are less susceptible to flooding. That can mean locating key infrastructure off flood plains and increasing the resilience of buildings that are located on them. The environment and the effect on fish habitat must also be considered, as do a range of other factors, from transportation to housing to health care.

Flood management strategies have also evolved since many of the existing dikes were created. Narrowly constraining a river with dikes can squeeze water levels up. So setting back dikes from a river can be more effective. Doing so can also reduce the sorts of erosion problems that can be created when a river eats into an existing dike, as is happening in the north of Abbotsford. But our communities have grown around the existing dikes, and relocating them or building new ones can be a fraught question.

“That’s why flood planning is so complex,” Litke said. “It touches on all aspects of society and communities.”

That complexity means it will likely take a couple more years until the strategy is at a stage where it can even begin to be implemented.

And when that moment comes, it’s unclear who will co-ordinate all the various moving parts—and provide the money—to turn the strategy into action. The Fraser Basin Council is also not a governmental body: it’s a non-profit. And while the province has assigned it the task of developing a strategy, it has no money to actually implement it. Implementing any strategy would fall to the province.

A billion-dollar problem

And money is the key. Upgrading all the bad dikes in the Lower Mainland will cost billions of dollars. In Abbotsford alone, more than $400 million of upgrades are needed, and that figure doesn’t take into effect the full cost of Sumas upgrades or construction inflation. The flood management strategy is needed to determine the order of upgrades, and which projects get money first. But it won’t create money for dikes out of thin air. That’s a decisions politicians will have to make.

While diking is technically a municipal responsibility, the costs are almost impossible to cover for most cities (there are some cities, like Richmond, that have thrown their own money at the problem). In the case of Abbotsford, politicians have said the city cannot cover the immense cost of upgrading dikes while also providing standard city services.

The federal and provincial governments do allocate flood-protection funding to cities. But much of that money has been aimed either at keeping existing dikes from crumbling or commissioning reports to further study the flood risk.

On Matsqui Prairie, at the northern tip of Abbotsford, the Fraser River is eating into its banks and threatening the dike. The province gave the municipality $10 million to try to fix the problem and divert the water. But that money is only aimed at maintaining the current dike. (Its elevation has been deemed to be “fair”—one of the better assessments in the valley.) Last year, just a month before council heard about the danger posed by the Nooksack River, city staff revealed that a sinkhole in the Matsqui dike would take $1.5 million to fix. This February, the province funded half the cost of that project.

Chilliwack has received money to upgrade dikes, including $2 million in 2015 to improve one along Young Road. But that money is just a drop in a bucket of what is needed, and Chilliwack was also required to pay one-third the cost of that dike. Requiring substantial investments by municipalities can be an obstruction, because many dikes protect rural areas with relatively few people—and taxpayers—to foot the bill.

The Current was able to find just one funding contribution from the province that paid for improvements along the Sumas dike: in 2008, the BC government gave Abbotsford money to remove two “failing drainage culverts” and replace them with a “larger capacity culvert to help carry storm water away from homes and businesses.” The province spent $40,000 on the project.

Most other projects funded by the province have been paperwork exercises, even if many of those are intended to lay the foundation for actual dike upgrades. Flood protection money must also compete with other disaster preparation projects, including preparing southeast BC from an earthquake and better fireproofing Interior communities. But even as climate change increases the frequency and scale of floods, there has been little sign that governments are willing to spend the hundreds of millions of dollars that are necessary to adequately protect residents and their property.

Braun has said it’s time to stop studying the dikes and start fixing them. He said so again Sunday in an interview with The Current.

“Enough studies,” he said. “This has been studied to death. We know what has to be done, and we’ve got to do it.”

A drop in the bucket

In August, the provincial government bragged it had spent $36 million on structural flood mitigation projects since 2017. (Just before losing power, the BC Liberals spent $80 million on flood projects.) In February, the province’s parliamentary secretary for emergency preparedness declared: “This program has gone a long way toward increasing British Columbia’s resiliency in the face of potential flooding.” But the program had not gone a long way to increasing the flood protection. All the estimates of the scale of the funding needed suggests the work has barely begun.

The last four years of structural flood prevention commitments—$36 million—have amounted to less than one tenth of what is needed not across all of British Columbia, but in a single municipality: Abbotsford.

On Friday, The Current asked Public Safety Minister Mike Farnworth if the province was prepared to spend the billions needed to upgrade BC’s dikes.

“We have in this province ongoing dike maintenance and dike improvement projects,” Farnworth said. “But events like this show very clearly that climate change is going to impact us in ways that it never did in the past. And a part of that is ensuring that our diking system is as strong as it can be. And that means that yes, the province and the federal government are going to have to be paying and investing in the years to come in our diking system.”

The Current asked the Province what it had done since it heard the Sumas River dike was unacceptable and whether it had provided any money to upgrade the dike since 2008. No answer was received. In response to a request for a flooding document, a spokesperson wrote that “responsibility for repairs and maintenance of dikes falls to local diking authorities.” That answer ignored the issue of dike improvements, which the province boasts of financing, on those rare occasions when it does so.

Yet, whoever is on the hook for protecting flood-prone communities, the recent flood has demonstrated that the cost of not stopping a major Lower Mainland flood is almost inconceivably high.

The Nooksack has flooded a relatively small portion of the Lower Mainland, but it has taken out key infrastructure, shut down transportation along vital corridors, and done huge damage to the province’s agriculture industry. Damages seem likely to exceed $1 billion. And yet every year carries the threat of a far worse flood.

The big one

Officials have warned for years that a major Fraser River flood could cause up to $30 billion of damage. It could swamp tens of thousands of homes, not to mention highways, hospitals, water plants, and other vital pieces of infrastructure. Unless a large earthquake hits first, such a flood would be the costliest disaster in Canadian history.

And yet, the same 2015 study that revealed Sumas Prairie’s dike would likely fail revealed flood protections against a major Fraser flood were just as rickety. Indeed, that was the main point of that study: a flood vulnerability study that was also conducted at the time ignored the Nooksack and other non-Fraser floods.

Litke said the Nooksack flooding could provide momentum to actually addressing the Lower Mainland’ susceptibility to a major flood. But it is unclear if that urgency will come in time. The region is in a race against time. A clock is ticking, but no one knows just when the buzzer will sound and the next flood arrives.

“It’s a question: does the next big one come next spring?” Litke told The Current. “It might not be for five or 10 or 20 years. In the Lower Mainland there have been a number of close calls and bullets that have been dodged. In this most-recent event, we did not dodge the bullet.”