Oil in place, gas cap and aquifer
A Havlena–Odeh fit for oil in place and gas-cap size over the surveys before an aquifer arrives, then whatever is left over as water influx.
- UC-MBE-003
- UC-MBE-007
Engineering use cases UC-MBE-003 (Estimate OOIP with Havlena-Odeh) and UC-MBE-007 (Estimate Gas Cap Size). For the aquifer fitted rather than left over, see mbe-fetkovich-aquifer; for the drive indices of this field, mbe-drive-indices.
Havlena–Odeh writes the material balance as a straight line: withdrawal F against the oil and gas-cap expansions Eo and Eg. With no aquifer the fit gives N, the oil in place, and m, the gas cap's size as a fraction of the oil zone.
What the fitted N and m do not explain is water influx. A We that grows steadily with time is an aquifer; one that scatters around zero means there is none and the fit stands. Here it is zero for three surveys and then grows, back to the influx the data was built with.
Which surveys go into the fit decides the answer. Over all six, this data gives 175 MMSTB and a negative gas cap: the aquifer's water is read as oil. A result like that is the sign to look for an aquifer, not to report the number.
F, Eo and Eg are in reservoir barrels and rb/STB, from the PVT at each survey pressure, so the quality of the answer is the quality of the PVT.
Source
using System;
using System.Linq;
using Team76.Engineering.MBE.OilMBE;
public class S : ICalcScript
{
public ScriptResult Run(ICalcContext ctx)
{
// Six yearly surveys: withdrawal F (RB) and the oil and gas-cap expansion terms Eo, Eg (rb/STB), for 50 MMSTB with
// a gas cap a third the oil zone. An aquifer reaches the field after the third survey and adds the last term to F.
double[] years = [1, 2, 3, 4, 5, 6];
double[] Eo = [0.010, 0.022, 0.035, 0.049, 0.064, 0.080];
double[] Eg = [0.015, 0.031, 0.048, 0.066, 0.085, 0.105];
double[] aquifer = [0, 0, 0, 0.3e6, 0.8e6, 1.5e6];
var F = Eo.Select((o, i) => 50e6 * (o + 0.33 * Eg[i]) + aquifer[i]).ToArray();
// Fit the surveys from before the aquifer arrives, so its water is not read as oil.
var (n, m) = HavlenaOdeh.EstimateNAndM(F[..3], Eo[..3], Eg[..3]);
// Whatever the oil and gas cap don't explain, at any survey, is water influx.
var et = Eo.Zip(Eg, (o, g) => o + m * g).ToArray();
var we = F.Zip(et, (f, e) => HavlenaOdeh.InferWaterInflux(f, n, e)).ToArray();
// Fitted over all six, the same data reads the aquifer as oil and gas cap.
var (allN, allM) = HavlenaOdeh.EstimateNAndM(F, Eo, Eg);
return ctx.Result()
.Set("Water influx", new ScriptChart(years, we.Select(w => w / 1e6).ToArray(),
title: "Water influx the fit leaves over", xLabel: "Year", yLabel: "We (MMrb)"))
.Set("N, first 3 surveys (MMSTB)", Math.Round(n / 1e6, 2))
.Set("m, first 3 surveys", Math.Round(m, 3))
.Set("N, all 6 surveys (MMSTB)", Math.Round(allN / 1e6, 2))
.Set("m, all 6 surveys", Math.Round(allM, 3))
.Set("We at year 6 (MMrb)", Math.Round(we[^1] / 1e6, 2));
}
}