Basic HTML Version
OH – ST - ET: Analysis of Dynamic Data in Shale Gas Reservoirs – Part 2
p 2/18
History plot, log-log plot and Blasingame plot
Using different methods we match pressures and rates with the best possible parameters, then
we forecast the production and ultimate recovery. To do this we will use and compare the
methods developed in Part 1.
We start with the simplest method, then we progressively use more sophisticated models in
order to take into account the “real” geometry and the “real” diffusion of the problem. In this
workflow we use the result of the previous analysis as a starting point for the next one. We see
how using a more complex model affects the results and our production forecast. At a stage
we may go back and eliminate sophistications that do not affect the end result.
We assume that gas flows from the reservoir to the wellbore through the fractures only. Only
dry gas flow is considered, although a residual water saturation may be added in the reservoir
model. Water flow back was noted during the clean-up and the first hundred hours of
production. So we may expect a pseudo-skin from the early time, but this will not impact the
description of the general model.
0
10000
20000
Gas rate [Mscf/D]
0
400
800
1200
Liquid rate [STB/D]
0
5E+8
1E+9
1.5E+9
Gas volume [scf]
qg
qw
Qg
0
1000
2000
3000
4000
5000
6000
Time [hr]
4000
9000
Pressure [psia]
Pi
p
Production history plot (Gas rate [Mscf/D], Pressure [psia] vs Time [hr])
1
10
100
1000
Time [hr]
1E+5
1E+6
Loglog plot: Int[(m(pi)-m(p))/q]/te and d[Int[(m(pi)-m(p))/q]/te]/dln(te) [psi2/cp] v s te [hr]
10
100
1000
10000
Time [hr]
1E-7
1E-6
Gas potential -1 [[psi2/cp]-1]
Blasingame plot: q/(m(pi)-m(p)), Int[q/(m(pi)-m(p))]/te and d[Int[q/(m(pi)-m(p))]/te]/dlnte [[psi2/cp]-1] vs te [hr]