Basic HTML Version
OH – ET – VA - LL: Analysis of Dynamic Data in Shale Gas Reservoirs – Part 1 – Version 2 (December 2010)
p 2/24
2 - Methodology used in this document
In Part 1 we use NO real data. The point here is not to draw a straight line on real data, or
match an analytical model, or a numerical model, and get results. As we see in Part 2 anything
will fit and prove nothing. We get some results, and then what? Even rolling dice or picking
random numbers would give us result.
In this document we will list our hypotheses, run the “exact” simulations according to these
hypotheses, and then apply different techniques to assess their pros and cons.
This document is structured as follows:
Presentation of the different techniques and modelling tools: early time (§3) and late time
(§4) straight line analyses, analytical models (§5) and numerical models (§6);
Validation of the reference nonlinear numerical model (§7);
Simulation and application of the different techniques to a single fracture (§8) and a
fractured horizontal well (§9);
Integration of desorption (§10) and non-consolidation (§11);
Material balance plot on the SRV with sensitivity to fracture density and spacing (§12);
Influence of real data signature was removed and transferred to Part 2;
Conclusions were also removed and transferred to Part 2;
List of additional theoretical developments to be done (§13).
We illustrate all this by simulating a reference case that has to be representative of what we
are likely to encounter, say, when fracing a horizontal well in Barnett shales. We use the
following parameters:
reservoir permeability
1E-2 ;
1E-4 md
; 1E-6
horizontal well length
4000 ft
reservoir thickness
100 ft
horizontal well position
centered
reservoir porosity
3%
horizontal well radius
0.32 ft
(connate) water saturation
30%
horizontal well skin factor
0
reservoir initial pressure
5000 psia
number of fractures
10
reservoir temperature
200 °F
fractures ½ length
200 ft
gas specific gravity
0.65
fractures conductivity
infinite
percent CO2
0%
fractures penetration
full
percent H2S
0%
fractures position
centered
percent N2
0%
production duration
4 years data
20 years forecast
Adsorbed gas
120 scf/ton
well flowing pressure
500 psia
In most cases an “average” permeability value of 1E-4 md is considered. When needed, this
base case is complemented by simulation runs on lower (1E-6 md) and upper (1E-2 md)
permeability. This will be sufficient to illustrate the strengths and pitfalls of the various
techniques presented here.