Water Saturation (Sw)

In REP, you enter Sw (water saturation) and not Shc (hydrocarbon saturation). The program will use (1-Sw) in the calculations. But in the probability distribution you enter, the P90 will correspond to a higher reserves case than the P10. This confuses some people, but only at first.

There is often considerable uncertainty in Sw and serious error in what is entered into REP.

Many thousands of petrophysical man-hours have been used up calculating Sw by any number of equations. From the point of view of hydrocarbons-in-place the differences between the answers are usually small compared with other uncertainties - gross rock volume, for example. Doubling Sw from 15% to 30% is a reduction in hydrocarbon saturation from 85% to 70% - a decrease in volume of only 18%. (There may be an equal or greater effect on recovery factor, however.)

So very often the calculation (or estimation) of a particular saturation is not a big problem. But when you ask a petrophysicist what is the likely or observed water saturation in a field the answer will often be the residual saturation above any transition zone. If there is a only a short transition zone, or if there is considerable structural closure, then this is a good value to guide you. But where this is not the case (and as reservoirs become smaller and more complex this is increasingly common) then you need to be rather careful.

You should also remember that in many traps the trap volume does not increase linearly with depth - in such cases far more of the reservoir volume is closer to the hydrocarbon contact than to the structural crest.

So here is the first question you should answer: is there an observed or likely transition zone and if so will it have an effect on the reserves calculation at any possible spill point or fluid contact? If not, you will confer with the petrophysicist and enter a likely range in Sw; the pert distribution is recommended. But if there is, you have two options:

1.Define a dependency (See 'Variable Dependencies') between Sw and spill point (or water contact). This will be a negative dependency; as spill point increases, Sw decreases. You may need to do some back-of-the-envelope calculations to work out suitable ranges in Sw, always remembering that you are entering average Sw over the whole reservoir volume.

2.Use the Sw-height (See 'Water Saturation Height Functions') option for Sw. Although this may mask some of the primarily petrophysical uncertainties (though you can use the Sw-Ht error function to model these) it will account far better for transition zone effects.

The effect of using an Sw-height function is to decrease reserve when the hydrocarbon column is small - decreasing your mean value, and significantly decreasing your P90 value. But better to be forewarned of this than have a nasty surprise when you drill the well.

It is noted in the literature that overestimation of small prospects and overestimation of the downside in all prospects is very common; and this commonly ascribed to natural bias and a desire by geoscientists not to be working on prospects which will never be drilled. However, there is little doubt that mishandling Sw is also a contributing factor.