People told Neil Smith he was crazy.
For years, Crescent Point Energy Corp. had been testing a new technology that involved injecting water to free up "tight oil." It was an innovation that held the promise of squeezing out billions of extra barrels - but many in the oil patch were skeptical.
Mr. Smith, the company's vice-president of engineering and business development, remembers one industry veteran in particular.
"He had been almost 50 years with a large major international company and said, 'I predict that injecting water into tight will not work,'" Mr. Smith said. "And that was old school. It didn't work."
But after several years of testing, analysts now believe that the technique not only appears to work - it could radically boost the quantity of oil that companies can pull from Alberta and Saskatchewan, where a rush to develop new reservoirs has already nearly reversed long-standing declines in non-oil sands crude output.
"The ramifications are immense," Mr. Smith said.
The idea behind injecting water to help recover crude is not new - in fact, it's nearly a century old. When companies first drill into a reservoir, underground pressure brings oil to the surface, a bit like pop fizzing out of a shaken bottle. After a while, though, the pressure fades, and oil flow slows. But add water and the pressure builds again, pushing more oil to the surface.
That works well in traditional reservoirs, where the underground rock is relatively porous - think of how water passes through sand. Skeptics, however, have long doubted it would work in so-called "tight oil" plays, where the rock is more impenetrable, more like brick than sand.
The first tests indicated it wouldn't work in tight oil. "It was so tight they couldn't push the water into the ground," Mr. Smith said.
Yet in the past year, a series of new tests has begun to prove that, when technology is used to first fracture rock, water can "sweep" out the oil so effectively that it can bring dramatically greater volumes to surface.
In a research report published Tuesday, Dundee Securities analyst Travis Wood examined the results from a series of Crescent Point wells and concluded that early results bear out that water does work. While it remains in the "proving phase" - and some wells have experienced operational issues, such as a lightning strike, that have hurt production - early numbers look promising.
"We're fans of the water-flood so far. We think it's going to add a lot of value," Mr. Wood said.
Before the water-flooding, Crescent Point believed it could extract 19 per cent of the oil in place in the Bakken, a major new play that contains an estimated 4.6 billion barrels. With water, it expects to boost that to 31 per cent. That's a potential gain of roughly 500 million barrels in the Bakken alone - and water-flooding also has the advantage of bringing oil to surface much faster.
"In six years, we're getting out the oil that we would have got out in 50 years," Mr. Smith said. Crescent Point ended 2010 with 11 water injection wells. It plans to increase that number to 36 in 2011, although that remains far from the 700 it has drilled in the Bakken.
But the true impact of the technology lies in its reach. The Bakken is only one of a series of prolific tight oil plays where it could work, including the Lower Shaunovan, which contains 4.3 billion barrels; the Viking, which holds six billion; the Swan Hills trend, which has roughly seven billion; and the Cardium, which contains 10 billion barrels.
Others are testing the technology, too, including Legacy Oil + Gas Inc., which has one pilot under way and another coming this year. Privately held Manitoba company Tundra Oil & Gas Ltd. has also piloted water-floods in a Bakken-like play in 2007; it has seen recovery factors leap forward much like Crescent Point.
"We're reasonably confident that it's working," said Tundra chief executive officer Dan MacLean, who noted that water-flooded barrels are also cheaper to extract.
And water could be just the beginning. Some companies warn that some reservoirs are just too tight for water - and are now looking at other ways to obtain gains. Tundra is experimenting with carbon dioxide floods, while PetroBakken Energy Ltd. began injecting natural gas into an initial tight oil pilot well in mid-March; it plans to start four more pilots by year's end.
Without trying these new techniques, "we're just touching the surface in most cases," said Rene LaPrade, PetroBakken's senior vice-president of operations. "There's a lot of oil still be left to be taken out of the ground."