[CAPTION] Eastmain residents will be safe and dry on the hill just south of town if the Opinaca reservoir bursts, according to this 1988 Hydro-Quebec map which has never been shown to the community. But Eastmain itself would be flooded by 8 metres (25 feet) of water. Residents would have 12 hours and 5 minutes to evacuate. (See pp. 12,14 and 15 for other maps.)

October 13, 1983. It was a stormy night at the LG-2 reservoir. A violent storm was blowing in from the north-east. In front of the main dam, the open water stretched out to the horizon and beyond. Furious winds blew in at up to 60 km/h, sending huge waves thundering into the dam and nearby dikes.

The storm was the first big test for Hydro-Quebec’s new dams in James Bay, the world’s biggest hydro-electric complex. It was a test Hydro-Quebec wasn’t prepared for. At no time in its history would the utility ever come this close to an unthinkable disaster.

For 10 hours, the waves battered the dam and dikes. Made up of massive 700-kilogram rocks – known as “riprap” – the dam and dikes were designed to withstand winds of 100 km/h or more.

Or so thought Hydro-Quebec. In fact. Hydro’s underestimated how big the waves could get in the reservoir. And the riprap suffered from apparent design and construction flaws.

During construction, small stones and sand had somehow got mixed in with the larger rocks. It was a bad combination, like
oil and water. The waves easily dislodged the small stones, causing the riprap to slip out of place. The result was landslides all along the main dam and two of the biggest dikes in Hydro-Quebec’s system. The storm was so strong it even washed away rocks on the down-river side of the installations, the side that faced away from the waves.

One Hydro employee who was at LG-2 during the 1983 storm looks back at that night and thinks of the disaster that might have been.

“It was not dangerous. But if the storm had lasted a week it would have been dangerous,” said the employee, who requested anonymity. He said the waves would have punched a hole in the dam or dikes. “The reservoir would have emptied completely.”

At the time, H-Q didn’t yet have any emergency studies for the Cree communities downriver, according to Hydro’s documents.

Today, Hydro-Quebec downplays the storm. “At no point were the dams in any danger. For us it’s become ancient history,” said spokeswoman Sylvie Tremblay. “It’s not like it was a spectacular storm.”

Spectacular or not, Hydro admits the storm still caused millions of dollars in damage to its prized La Grande dam complex. The utility spent the next 14 years rebuilding, redesigning and strengthening its facilities at a cost of $100 million, according to Gérard Verzeni, Hydro’s director of dam security.

Hydro-Quebec never disclosed the events of October 13,1983, to the public, and the incident has never been reported before. No Cree official we interviewed had heard of the storm damage to the La Grande complex.

The incident was discovered as part of a two-year investigation of Hydro’s emergency planning. The Nation has obtained thousands of pages of internal Hydro files, including most recently four reports that tell the story of the October 1983 tempest.

Last summer, we reported on other Hydro documents that show the utility routinely failed to perform required inspections and maintenance at its northern dams. Year after year, Hydro violated its own regulations governing inspections between 1991 and 1996. In 1991, for example, LG-2 saw only half the required inspections; at LG-3, only 23 per cent of required inspections were done.

The latest reports covering the 1983 storm suggest Hydro-Quebec’s dams and dikes weren’t designed or built properly, and this caused the premature erosion.

[CAPTION] Kuujjuaq would be under 20 metres of water if the Caniapiscau dam bursts. It has 30 hours to evacuate, according to this 1989 H-Q map which was never released to the community.

Somehow, “fine particles,” such as small stones, got mixed in with the larger rocks that made up the riprap, according to the reports. These particles shouldn’t have been there and caused much of the damage, said the reports, which total 40 pages.

The worst storm damage happened at a three-kilometre dike known as CH-21, about 20 km south of the main LG-2 dam. This dike suffered extensive erosion due to “an inherent default,” according to one of the reports. “The dike should have normally resisted winds of 100 km/h according to the design.”

The report said the dike was likely not built properly. “The damage was caused by the presence of fine particles (smaller than 30 cm) mixed with the riprap. This suggests the design criteria weren’t respected.”

Another report mentions a second “violent storm” in August 1983. This earlier storm caused major damage to a dike in the Opinaca reservoir, OA-02, directly south of the LG-2 reservoir. The report blasts the dike’s designers for not giving it a decent riprap. “Most of the damage to the riprap of this dike is probably due to this choice, which was doubtful at best,” according to the reports.

The main dam of the Opinaca reservoir, OA-05, also suffered from badly designed riprap, the reports said. This dam first suffered major erosion in the fall of 1992, “and this, even though the east winds aren’t very strong at this time.”

The reports call for an overhaul of the security of Hydro installations. They also say Hydro’s faulty calculations underestimated the size of waves in the reservoir.

The reports also show that, after initial emergency repairs in 1983, it took Hydro 10 years to finish a redesign of its dams and dikes. It wasn’t until 1996 and 1997 that the storm rebuilding was finally completed, according to Verzeni, Hydro’s dam security director.

In an interview, Verzeni said the dams and dikes are now much safer than they were in 1983. The crests of the dams and dikes were raised by 1 metre and bigger rocks were used in the riprap. “They (the rocks) were already enormous,” he said. “But we realized they needed to be bigger still.”

But Verzeni admitted that Hydro still hasn’t heeded a warning about the lack of filters in the riprap. The lack of filters was cited as a major reason for the 1983 erosion.

Verzeni also disagreed with the allegation in the Hydro reports that the design and construction of the riprap was faulty.
He dismissed this allegation as merely “an opinion of an engineer.” He said, “There’s going to be a debate among engineers. Sometimes we have a muscular discussion.”

But Mai Phat, an engineer in Hydro-Quebec’s dam security department, acknowledged the utility underestimated the size of the rocks needed for the riprap and misjudged the size of the waves that could hit its facilities. “After five to 10 years, we realized these rules were insufficient. We used new rules to rebuild the dikes,” he said.

Gilles Marinier was a vice-president of engineering at Hydro, working on construction of the James Bay project from 1979 to 1984. He said he failed to see The Nation’s interest in the possibly faulty dams.

“It’s a technical affair, not a journalistic issue,” said Marinier, who was also a consultant to the provincial inquiry that investigated Hydro’s role in the ice storm (see News, p. 7). “Maybe there was an error of conception or construction (in the dams),” he conceded. “So what? Why write about that?”