When we are presenting our plans to fellow professionals, we often get asked, "Surely steam flooding is an EOR technique, why steam from the start?" It is a good question, as generating steam is not cheap, of course we have a good answer to the question and the more we think about it the better the answer gets (For the time-challenged: the answer's at the end in bold).
Let me start by putting steam flooding in context. It is often conflated with other EOR techniques and with SAGD operations. That's not always helpful, so let me try and make some distinctions so that you can understand how to think about what steam flooding is really doing. In general what you are trying to do in EOR is one of two things: you are either improving sweep efficiency (the Heineken effect – refreshing the parts other beers cannot reach), or driving down the residual oil saturation (washing that oil right out of your hair – is that the South Pacific effect?)
Polymer floods are doing the first, surfactant and fresh water floods the second, steam does both. But each technique has a range of applicability, you can't summarise all the screening criteria in one chart but the figure below does it better than most.
Steam techniques really only work well on viscous oils in relatively shallow reservoirs. But not every steam technique is equally applicable in all circumstances. They all benefit from the heat of the steam reducing the viscosity of the oil and the fact that steam drives down residual oil saturation, but there are different methods for different circumstances and each method has its own challenges. Let's start with SAGD, that's steam assisted gravity drainage for the uninitiated. Though if you have got this far without being a petroleum engineer I am delighted to have held your attention. This is for the really viscous stuff, the Athabasca tar sands, as they used to be known. There nothing flows without some heat and designing how to warm the wells up and to start production is half the battle.
Cyclic steam stimulation, CSS or colloquially "Huff and Puff", can be used for these types of oils as well but this technique tends to work better if there is at least some oil mobility in the reservoir. Steam flooding really needs an oil that can flow cold. However, there is a lot of overlap between these two techniques, CSS tends to have a lower steam oil ratio (more bang for your buck) but also a lower recovery factor (fewer bangs and hence fewer bucks). So it takes a bit of economic analysis to work out which technique is best. In the North Sea where infrastructure costs are high and scale matters, steam flooding is the right answer; onshore Venezuela where there is more oil than you can shake a stick at, and adding another well is easy, CSS is the way to go.
So what have we got in the North Sea. Well up to now there hasn't been much call for steam, the "heavy" oil fields that have been developed to date weren't really great candidates for steam flooding, but there are three or four fields coming up, which all lie in the sweet spot for this extraordinarily efficient EOR technique. They all have very high quality rocks, a prerequisite for a successful steam flood. In similar reservoirs onshore, such as Kern River and Duri, Chevron has seen recovery factors of 60% to 80%. Taking steam offshore is not a new idea, we have been doing that since 1844 at least, though of course we would be generating steam on a bigger scale, but scale is just engineering; and the economics work because you not only recover much more oil, it's produced over a much shorter time scale.
Here is that chart again with most of the UK's heavy oilfields marked on the plot.
Of the projects that fall in the steam flood category only the Heimdal reservoir in the Mariner field is actually past the point of no return (Final Investment Decision, or FID), Bentley, Bressay and our two fields, Pilot & Elke, are still approaching that point.
Why do I call FID a point of no return? That's because for a developed offshore field retrofitting steam flooding would be a nightmare, what costs a little bit extra on the drawing board before anything is built, costs an arm and a leg when done as a modification to an existing plant. Getting the concept right at the start is critical.
Still, surely EOR is for later, you know, at the end of field life, once the easy oil is gone. Well that brings me back to the question I started with. "Why steam from the start?" The answer is simple, unlike light oil reservoirs, there isn't very much easy oil in a heavy oil development.
This chart is a cartoon, but that doesn't mean the point isn't valid. With a light oil, primary recovery (what you get out of the ground without putting anything back) and secondary recovery (what you get if you pump water in) will get you about half of the oil in place. The UKCS average is 46%.
EOR can still improve the recovery factor for light oils, but it is an increment, not a transformation.
For heavy oils the picture is different. Primary and secondary recovery is limited, whilst the tertiary contribution can be huge. Steam flooding can triple even the secondary recovery factor, and the ultimate recovery factor can be just as high, if not higher, than you might get with a light oil reservoir.
So our answer to the original question "Why steam from the start?" goes thuswise:
"Why waste thirty years pushing water through a reservoir, when you get triple the oil in half the time with steam?"