New Technology Magazine

SAGDo SAGDon’t

Trail-blazing in situ oilsands producer challenged by maturing assets

For a technology that wasn’t used on a large scale until 2001, steam assisted gravity drainage (SAGD) isn’t doing too badly. Last year, it produced more than 570,000 barrels of bitumen a day, making it Alberta’s second most important oil extraction method after bitumen mining, which produced 978,000 barrels a day in 2013.

Bitumen (both mined and from wells) made up 80 per cent of Alberta’s oil production last year, and SAGD accounted for 52 per cent of the bitumen produced from wells in 2013. Another 23 per cent came from wells using cyclic steam stimulation, (CSS) and 25 per cent was from wells that don’t use steam, according to the Alberta Energy Regulator’s 2013 reserves report, ST98-2014.

Because SAGD is Alberta’s newest oil extraction technology to be deployed on such a large scale, this issue of New Technology Magazine looks at how the first big project has performed during its first dozen years of operation.

Cenovus Energy Inc.’s Foster Creek oper­ation is Canada’s longest-running large-scale SAGD project. When it went commercial in 2001, the project in northeastern Alberta quickly established itself as the gold standard for the technology.

With output of 106,000 barrels a day, Foster Creek was Canada’s second-largest steam-assisted oil project in 2013, after Imperial Oil Limited’s Cold Lake project, which uses CSS not SAGD. (Foster Creek is operated and 50 per cent owned by Calgary-based Cenovus and 50 per cent owned by Houston-based ConocoPhillips Company, but all production figures used in this article include both com­panies’ shares and are before royalty deductions.)

Storm clouds?

But Foster Creek was the subject of some rare negative publicity last year.

In the third quarter, production fell by 22 per cent from the corresponding 2012 quarter. The full-year average of about 106,000 barrels a day was down from nearly 116,000 barrels a day in 2012 and 110,000 barrels a day in 2011.

Last year’s steam to oil ratio averaged 2.48, up from 2.16 in both 2012 and 2011. (The steam to oil ratio is a key measure of a SAGD project’s energy efficiency and economic performance. A steam to oil ratio of 2.48 means 2.48 barrels of water had to be converted to steam and injected into the reservoir to recover one barrel of bitumen.)

In March, investment analysts at Peters & Co. Limited issued a 12-page report titled Foster Creek Update—Analysis of Operating Challenges. It talked about problems with steam chamber coalescence and a high rate of well failures. “As one of the oldest SAGD operations, Cenovus has begun to encounter issues at Foster Creek that it did not fully anticipate,” the report warned. “Some of these challenges may be expected in the future by other operators unless pre-emptive measures are taken or the operational realities of a multi-pad SAGD project are incorporated into production forecasts.”

The ultimate indignity in the Peters report was a table titled “2013 SAGD Well Runtime Analysis.” It ranked Alberta’s 13 major SAGD projects by “average monthly well up-time.” (The report said “up-time” was calculated as total producing well-hours divided by total available well-hours.) At 88 per cent, Foster Creek was tied for last place with Nexen Energy ULC’s hapless Long Lake project, which has never produced more than about half of its design capacity. Peters’ industry average for well up-time was 96 per cent.

Is Alberta’s trail-blazing SAGD project faltering? No, not even close, says Harbir Chhina, Cenovus’s executive vice-president of oilsands and a SAGD veteran. In an interview with New Technology Magazine, Chhina addresses the questions raised about Foster Creek as well as the overall performance of SAGD during its longest-running commercial application.

SAGD versus CSS

One way to look at the performance of SAGD as a technology is to compare it to Alberta’s only other steam-assisted bitumen extraction technology that has significant production history under its belt: CSS.

On a CSS project, steam is injected into each well until the reservoir is heated, then oil is produced back through the same wellbores, and the cycle is repeated—which is why it’s called cyclic steam stimulation.

On a SAGD project, steam is continuously injected into a horizontal well to melt bitumen, which drains into a lower parallel horizontal well that continuously produces oil.

Canada’s biggest CSS project is also the country’s oldest large-scale steam-assisted bitumen recovery scheme—Imperial Oil’s Cold Lake project. And after more than four decades of production, Cold Lake in northeastern Alberta is still Canada’s biggest steam-assisted oil operation with average 2013 output of 153,000 barrels of oil a day.

Imperial uses CSS at Cold Lake because thin shale barriers impede vertical permea­bility, ruling out a lower-pressure, gravity-flow process such as SAGD. So Imperial uses high-pressure CSS to fracture the shale bar­riers and force steam into the bitumen-bearing sands. In reservoirs with good vertical permeability—such as Foster Creek’s McMurray Formation—operators can use SAGD.

In the best parts of Imperial’s Cold Lake reservoir, which is in the Clearwater sands, the company expects to ultimately recover about 60 per cent of the original oil in place.

While CSS has been in large-scale commercial use for about three times as long as SAGD, Chhina points out that SAGD—at least at Foster Creek—is ahead of CSS in one important respect. While Cold Lake is expecting to ultimately recover about 60 per cent of the oil in place, some wells at Foster Creek have already recovered that much.

“Twenty to 25 per cent of our [well] pads have already reached 60 or 70 per cent recovery,” Chhina says of Foster Creek. So while each reservoir still has to be judged on its own merits, no one can seriously suggest the jury is still out on SAGD as a technology. Given that nearly three quarters of the oil in place has already been recovered over 20–25 per cent of the Foster Creek field, “we’ve learned most of our lessons that we have to learn,” Chhina says, adding, “Our target 10–15 years ago was to recover 60 per cent to 70 per cent. And we’re already there.”

He was referring to six well pads at Foster Creek. (Each pad has six to eight SAGD well pairs.) But even though they’ve recovered 60–70 per cent of the original oil in place, those wells still have combined production of about 23,000 barrels a day, which implies even higher ultimate recoveries.

Those high recoveries are from wells that have been on production the longest. What about the remaining three quarters of the field?   “We haven’t seen anything to date that tells us we won’t get that recovery factor,” Chhina says. “What we have seen is that the wells will perform differently. Some wells will have a higher [oil production] rate and low [steam to oil ratios]. Some wells will have a lower oil rate and a slightly higher [steam to oil ratio]. But [given] what we’ve seen with the wells and pads we’ve drilled to date, we do not expect the recovery factors to be an issue at Foster Creek.”

Production

So why did Foster Creek’s production fall last year?

In the third quarter of 2012, production averaged 126,000 barrels a day—6,000 barrels a day above nameplate capacity. But this was due to one-time factors such as a large number of new wells that had come on stream and the resulting increase in reservoir pressure. Given the fantastic production levels, Cenovus decided to delay taking wells offline for routine maintenance.

But deferring maintenance has a price. To run SAGD, you need temperature and pressure sensors downhole. By the start of 2013, only half of Foster Creek’s instrumentation strings were working. Not doing main­tenance in 2012 created a backlog in 2013, and the failed downhole pressure and temperature gauges had to be repaired or replaced. Each of those wells had to be taken offline while a coiled tubing rig retrieved the failed sensors then ran new or repaired ones back into the hole.

By the end of 2013, 95 per cent of the instrumentation was working, production was back to normal and the company had learned a lesson. “You can’t compromise on the instrumentation to cut costs,” Chhina says. “And so we will never do that again. We will make sure we always have proper instrumentation so that we’re not running the wells blind.”

What’s a realistic expectation for Foster Creek production? With the current phases running at full steam, the project can produce 120,000 barrels a day. But given that mechan­ical maintenance has to be done periodically, it is assumed the project will actually produce at 90–95 per cent of that capacity. In other words, Foster Creek should run at anywhere between 108,000 and 114,000 barrels a day, says Chhina. In the first quarter of this year, output averaged 110,000 barrels a day. So measured by production performance, Foster Creek is doing exactly what it was designed to do.

Well productivity

According to the Alberta Energy Regulator’s 2013 reserves report, the province’s 976 SAGD wells had average production of 596 barrels of bitumen a day in 2013. That compares with average output of only 64 barrels a day per well for Alberta’s 3,987 CSS wells and 37 barrels a day for the 7,437 non-thermal bitumen wells (which could include primary production, waterflooding and polymer flooding).

Foster Creek’s 170 SAGD well pairs and 67 wedge wells produced 106,000 barrels of oil a day last year, or about 450 barrels per well. (Wedge wells, a Cenovus innovation, are single horizontal wells drilled between two SAGD well pairs to drain oil the SAGD wells couldn’t reach.) Production from individual well pairs at Foster Creek and Christina Lake can range between 300 barrels and 3,000 barrels a day, says Chhina.

For Cenovus, the sooner it can abandon the wells and move the facilities to the next pad, the better the economics.

“If I could get all the [recoverable] oil out in 12–15 years—I’m actually very happy with that. And that is actually the goal,” says Chhina. “What I don’t want to do is keep those facilities for 25–30 years and keep producing five barrels a day.”

Inter-well communication

While inter-well communication, or “steam chamber coalescence,” at Foster Creek has had some publicity recently, the phenomenon is nothing new.

In thermal recovery schemes, the steam chamber, or heated zone, expands as steam penetrates more of the reservoir, and the wells eventually become interconnected. How long it takes this inter-well communication to occur can vary dramatically with reservoir characteristics.

At Foster Creek, the SAGD well pairs operate in isolation from each other for the first three or four years. After four to seven years, inter-well communication occurs between the well pairs, which are 100 metres apart. After seven to 10 years, all the steam chambers within an individual well pad have merged. And after about 10 years, depending on their locations, Foster Creek’s well pads (each with about eight well pairs) start to become interconnected.

In contrast, the wells at Cenovus’s Christina Lake project “communicate basically from day one,” says Chhina, even though Christina Lake is one of Canada’s top-­performing thermal oil projects. The reason is Christina Lake has bottom water and top gas.

Wellbore conformance

Another issue that has received some attention lately is Foster Creek’s wellbore conformance—the even distribution of steam along the length of each well.

Here too is an example of the dramatic differences that can exist between two high-quality reservoirs with different characteristics. Almost every SAGD project begins by circulating steam into both wellbores to heat the reservoir and melt the bitumen. Once the reservoir is warm enough that bitumen will flow, the lower of the two parallel horizontal wells is converted to oil production.

At Christina Lake, as with most SAGD projects, steam is injected in both wells of each well pair for the first three months. Then the producer well is converted to oil production.

Foster Creek was the exception.

At Foster Creek, Cenovus was able to get the wells started by just injecting a small amount of steam into a producer well, or both the producer and injector, for a week or two. “I worked on 26 different pilots in my career. Foster Creek in the Athabasca oilsands was the only reservoir that we could produce from primary production,” Chhina says, meaning some oil could move without heat. “Because it’s the southern-most project of the Athabasca oilsands, it’s also at the greatest depth.” He says this means it has lighter oil—nine to 11 degrees API gravity—than most projects in the Athabasca oilsands region.

But because steam is injected into both wells for three months before oil production begins, Christina Lake has better wellbore conformance.

Chhina displays 4-D, or time-lapse, seismic maps of the Foster Creek and Christina Lake reservoirs. On the Christina Lake map, nearly the entire area where steam injection is ramped up is shown in red, meaning it is heated. About 95 per cent of its wells are working as intended. But at Foster Creek, there are plenty of “bald spots”—only about 75 or 80 per cent of the reservoir is heated.

To improve conformance at Foster Creek, Cenovus will now put new well pairs on steam circulation for up to three months before oil production begins. The goal is to raise well conformance at Foster Creek to 95 per cent.

This means Foster Creek will have a higher steam to oil ratio in the short term since new wells will be on steam injection without oil production for up to three months. But in the long term it should mean a lower steam to oil ratio as oil production rises due to better wellbore conformance.

 But even if steam circulation does nothing, “Foster Creek is performing perfectly well right now,” Chhina says. “It’s right where [it’s] expected to perform.” He says about 10 per cent of the wells have reached a cumulative steam to oil ratio of two.

And that’s with conformance of only about 75 or 80 per cent. Given that 20­–25 per cent of the wells have already produced 60–70 per cent of the oil in place—and are still producing about 23,000 barrels a day—Chhina expects the steam to oil ratio and the recovery factor to continue to improve. “This is fine-tuning the last 10–15 per cent,” he says. “Foster Creek is running 85 per cent to the way it was designed. We’re trying to get it to 95 per cent.”

Oldest wells winding down

Just as Foster Creek was the first large-scale SAGD project to come on stream, it will also to be the first big one to have a large number of wells decommissioned. This means the Alberta Energy Regulator will face a precedent-setting decision: how much of the in-place bitumen should a SAGD producer be required to recover before wells are abandoned?

The first of Foster Creek’s 25 well pads went on full blowdown in the first quarter of 2013. Full blowdown is the final stage of production where steam injection stops entirely and is replaced by either methane or air injection. This reduces operating costs and lowers the project’s overall steam to oil ratio.

When this article was written, only one Foster Creek well pad was on full blowdown. Two more pads—about 15 well pairs in total—were on steam-and-methane co-injection, which prepares them for full blowdown. (One reason for co-injecting a small amount of non-condensable gas into SAGD wells near the end of their productive life is to maintain reservoir pressure. As steam condenses, reservoir pressure falls. Methane doesn’t condense, so reservoir pressure is maintained.)

And Cenovus has received regulatory approval to put two more Foster Creek well pads—F and G—on co-injection. On two wells the company will test air co-injection; the rest will co-inject methane. The reason for testing air is that it’s cheaper than methane.

Air will be injected at low pressure. Air injected into an oil reservoir at high pressure can cause ignition, but on the test wells air will be injected at rates too low to generate the high temperatures needed for combustion.

Also, Cenovus was preparing to file six more regulatory applications to put six more pads on co-injection. Pending approval, the company will continue to put well pads on co-injection as it deems appropriate.

Meanwhile, the Alberta Energy Regulator has to rule on what it considers an acceptable ultimate recovery rate for SAGD wells.

“We would prefer to start to put wells on blowdown after we recover 50–60 per cent, but today we have a lot of wells sitting above 60 [per cent recovery] that still aren’t on blowdown,” says Chhina. This was partly due to the company’s optimism during Foster Creek’s abnormally high production levels in 2012, but it is also the subject of discussions with the regulator.

Part of the Alberta Energy Regulator’s mandate is to ensure no economically recoverable oil or gas is left in the ground when wells are abandoned. Companies, on the other hand, want projects that provide the best rate of return, which doesn’t always favour low-productivity wells.

For the past four years, Cenovus has been talking to the regulator about recoveries that can ultimately be expected from Foster Creek and when it is appropriate to put wells on blowdown. Chhina says the company and the regulator are “on the same page.”

To calculate the province’s economically recoverable resources, meanwhile, the regulator assumes a 50 per cent recovery factor for SAGD versus only 25 per cent for CSS at Cold Lake, according to ST98-2014, which was released in late May. By comparison, it estimates the bitumen recovery factors for waterflooding and polymer flooding at 10 per cent and primary recovery schemes at five per cent.

There is one Alberta policy Cenovus would like to see changed. The company would prefer to be allowed to file one blowdown application for the entire field, rather than the current policy of having to submit a separate application for each individual well pad.

Once pads are on full blowdown, steam that would have gone into those wells is re­directed to new pads. Cenovus plans to accelerate the drilling of sustaining well pads—wells that will offset declines in production from mature wells.

So far, Cenovus hasn’t said how long it expects blowdown to last or how much additional bitumen might be recovered during the process. Chhina says, “We’ve already proven we can reach [recovery rates of] 60–70 per cent, and the wells are still producing oil, so we should be able to exceed 70 per cent recovery factors, we feel, but we have to prove it.”

Recovery factors can vary across a field. But post-steam core analysis offers a glimpse at how good recoveries can be. In the best parts of the Foster Creek and Christina Lake reservoirs, Chhina says, bitumen saturation has been reduced to as little as six per cent from the original 80 per cent.

“So that confirms that we are recovering this oil just like we expected to,” he says. “So everything is working according to plan—in terms of the drainage, the residual oil and the recovery of the oil.”

Learn more on SAGD technology and project updates at the Canadian Heavy Oil Conference being held on Nov. 3-4, 2014. The Canadian Heavy Oil Conference will feature a technology component with an enitre program track dedicated to subsurface issues in the heavy oil and oilsands industry. For more information please visit heavyoilconference.ca, or view related program sessions here.

• Exclusive