We have been developing heavy oil on the UK continental shelf for well over twenty years and the recovery factors that oil companies have achieved in the fields developed back in the nineties range from moderate (23%) to fantastic (74%), but have averaged a fairly decent 50%. The most recent spate of development projects are all envisaging recovery factors that are quite a bit lower, just 22% on average, if we ignore the Pilot steamflood project. Does this mean we have all lost our mojo, or that technology has gone backwards? It might seem so but when I analyse it in detail you will see that it is the nature of the reservoirs and the oils within them that determines the likely recovery factor.
The fields which make up the group I am analysing are all high quality Palaeocene and Eocene sandstone reservoirs deposited in a deep marine setting, so the rocks are quite similar; conversely, the viscosity of the oil in the reservoir varies considerably. Here is a chart of recovery factor vs viscosity, with viscosity plotted on a log scale. I have grouped the fields into producing (or produced) fields, the green squares, and discoveries and development projects, the orange diamonds. I have also plotted the Pilot field on the chart twice, once with the recovery factor a waterflood is expected to achieve (13%) and the average oil viscosity at reservoir temperature and again assuming a steamflood and the much reduced oil viscosity that would entail, the yellow triangles. The point of showing these two development options is that the recovery factor you get from your development is a function of what you have in the reservoir and what you do to the reservoir. It is not a fixed immutable point.
Well there is a pretty clear trend but there is also a pretty wide scatter, and knowing the viscosity doesn't give you the definitive guide to recovery factor that we might hope for. Let's add in another parameter and see if we can tighten up the scatter on this plot. Let's try plotting recovery factor against transmissibility. Transmissibility is permeability divided by viscosity, strictly speaking also divided by formation volume factor, but we can ignore that for now as the range is pretty small and I only know the value for a few fields.
Well, that has tightened things up a bit, there is a pretty clear relationship here, especially if we focus in on just the green squares, which are the producing fields where the recovery factors are mostly fact and only a little bit projection; by definition the recovery factors for the discoveries and developments are 100% projection. You might wish to substitute fiction for projection in the previous sentence but that would be a little too harsh on the reservoir engineering profession.
I would say that this analysis shows that most of the variation in recovery factor across these fields can be explained by just two parameters, one describing the reservoir and one describing the oil within it. Of course other things matter, such as well density, reservoir continuity and whether a field has bottom water or not, but just these two parameters help you estimate the likely range of recovery factor very quickly.
Of course, the key thing I would like you to remember is that even though the nature of the oil and the quality of the rock are fixed, the application of a little bit, or actually rather a lot, of steam can change the viscosity and boost recovery factor dramatically.