Deep Steam

I wrote earlier about why 3,000' is considered the conventional limit for steam floods and this is a follow up note exploring some ways that the industry could increase that limit.

It has some significance; for the UK sector of the North Sea at least, we have estimated that, if you could push the limit down to 4,500', the incremental recovery possible could be more than 2 billion barrels of oil. Almost as much oil as in Johan Sverdrup, if that isn't worth thinking about what is?

UKCS heavy oil fields plotted showing reservoir depth and oil viscosity at reservoir temperature; bubble size is proportional to oil in place; produced (dark blue) remaining reserves in producing field (blue); reserves in fields under development (green); reserves in discovered fields (mod orange); potential additional reserves from steam flooding (dark orange) 

UKCS heavy oil fields plotted showing reservoir depth and oil viscosity at reservoir temperature; bubble size is proportional to oil in place; produced (dark blue) remaining reserves in producing field (blue); reserves in fields under development (green); reserves in discovered fields (mod orange); potential additional reserves from steam flooding (dark orange) 

In fact, people have been thinking about how to inject steam into deeper reservoirs for a long time. In 1977 the Department of Energy in the USA set up a five year programme, called "Project Deep Steam" to develop downhole steam generators,  thermal packers and insulated injection strings. That was all good stuff, but mostly focussed on reducing heat loss in the wellbore. However, if you can drill and complete wells into which you can inject around about 10,000 bbls/day of cold water equivalent of steam, then heat loss in the wellbore is a much smaller problem than you (or we) might have imagined.

Given that most commentators see heat loss in the wellbore as the biggest hurdle to steam injection into deeper reservoirs, it may be that this realisation alone is reason enough to move the limit deeper for fields with high permeability reservoirs.  In which case Statoil, Xcite et al should maybe skip the rest of this article and fire up their thermal reservoir simulators.

However there is another problem and that is the steam temperature, which increases with pressure. This plot shows the energy in steam versus pressure, you can also work out the water phase and by interpolating between the red isotherms, the temperature. Most steam floods aim to inject a mixture of steam and water with an enthalpy of between 2,000 and 2,500 kJ/kg. At 3000', for a normally pressured reservoir, that equates to 85 bar and 300ºC. In fact it is equivalent to the top of the green box in the enthalpy diagram.

300ºC is hot, but why does 300ºC become a temperature limit? Well it doesn't have to be, but in sourcing potential completion equipment for the Pilot project we have found that there isn't a lot of downhole equipment available at temperatures this high. Some components have been developed for the geothermal industry and have quite high temperature ratings but others lag behind. So far the highest rated packer we could find could cope with 540ºC, but the best sliding sleeve was rated to 315ºC.

Looking at the enthalpy chart it seems as if temperature climbs very rapidly as pressure increases but while that is true to some extent, the effect is magnified by the fact that on this chart pressure is on a log scale. At 4,500' steam temperature for a normally pressured reservoir would be 338ºC. Given that Schlumberger offer expansion joints and packers rated to 343ºC it may well be that extending the depth limit is just a matter of trying it out with the right equipment in the wells and confirming that the temperature ratings are valid.

Steam temperature vs Reservoir Depth

Steam temperature vs Reservoir Depth

However, as we go deeper, the pressure increases, and the enthalpy of condensation decreases, what that means is that in the reservoir more of the swept zone will be occupied by, admittedly very hot, water and less by steam. The residual oil saturation for steam is less than that for hot water at the same temperature so we would lose some of the possible benefit of steam flooding. That might turn out to be a marginal effect, but it would probably be better if we could find a way to keep more of the fluid we inject into the reservoir in the gas phase.

Co-injecting a non-condensible gas would do just that and as that would also reduce the partial pressure of the steam it would also reduce the steam temperature. But perhaps a better alternative is to co-inject a fluid which is miscible at reservoir conditions. That seems to be a promising approach in its own right: adding about 5% propane or butane to the injected steam can accelerate oil production and reduce the amount of steam which has to be injected by about 30%. There are many trials of this technique underway in Canada.

In fact there seem to be multiple avenues for investigation of what could be the best steam injection strategy for deeper reservoirs, and all of them seem quite promising. Of course not all deeper heavy oil fields will be perfect for steam flood, some have bottom water which can act as a thief to the heat one is trying to inject into the reservoir. But with the advent of horizontal wells with steam injection valves along the wellbore, much more precision in the placement of steam is possible than in the past. That gives us many more tools to control where the heat goes so steam injection could well be feasible in many more fields than people have realised.

If a steam flood is feasible it is best done right from the start of production; all the time and energy expended in performing a conventional waterflood before doing a steam flood is wasted, so best to crank up that reservoir model now. Alternatively, as a first step, come and talk with us about proving that steam flood can work offshore on the one North Sea field that is shallow enough to pass all the conventional screening criteria.