Spe-192002-ms case Study Applied Machine Learning to Optimise pcp completion Design in a cbm field

Figure 1
Figure 2
Figure 3
Downloaded from http://onepetro.org/SPEAPOG/proceedings-pdf/18APOG/2-18APOG/D021S016R002/1220497/spe-192002-ms.pdf/1 by Vedanta Limited - Cairn Oil & Gas user on 28 June 2023

SPE-192002-MS
9
Figure 4
Conclusions
Production periods within CBM wells can effectively be converted into stationary representations that describe a wells state. These stationary representations are in fact necessary to reduce the residual model error to a point where there is a manifold that can be represented by a functional estimator such as a Gaussian process.
Gaussian processes can be used effectively in scenarios such as well failure where the data is expensive to capture and observations are few. Their application can offer benefits in both design of future experiments as well as exploit designs that offer a stronger certainty to running longer.
For practical purposes it maybe relevant to consider censored data where equipment has not failed in order to get more observations being mindful that heteroskedacity with be introduced and the model will be closer to a lower bound for runlife.
The methods and transformations described in this paper offer robust techniques in allowing CBM
operators to improve their completion designs by recommending configurations that have not been tried before and also exploiting configurations that are known to perform better.
References
1.
D.M. Smith, FL. Williams, T. Sato, Y.H. Lee and RD. Malone "The Methane Content of San
Juan Basin Coal Implications for Coalbed Methane Resource Assessments"
2.
R. Dhir, RR. Dern Jr. and M.J. Mavor: "Economic and Reserve Evaluation of Coalbed Methane
Reservoirs"
3.
C.A. Mora, RA. Wattenbarger: "Comparison of Computation Methods for CBM Performance"
4.
G.R. King "Material Balance Techniques for Coal Seam and Devonian Shale Gas Reservoirs"
5.
S. Moghadam, O. Jeje, L. Mattar Advanced Gas Material Balance, in Simplified Format"
6.
J.Q. Shi and S. Durucan A Model for Changes in Coalbed Permeability During Primary and
Enhanced Methane Recovery"
7.
A. Kantzas, Y. Ham "Measurement of Relative Permeability of Coal to Gas and Water"
8.
C.R. Clarkson, MR. Rahmanian, A. Kantzas and K. Morad: "Relative Permeability of CBM
Reservoirs: Controls on Curve Shape"
9.
W. Q. Meeker and LA. Escobar (1998), Statistical Methods for Reliability Data, John Wiley Sons, Inc.
Downloaded from http://onepetro.org/SPEAPOG/proceedings-pdf/18APOG/2-18APOG/D021S016R002/1220497/spe-192002-ms.pdf/1 by Vedanta Limited - Cairn Oil & Gas user on 28 June 2023

10
SPE-192002-MS
10. CE. Rasmussen & CK. I. Williams (2006), Gaussian Processes for Machine Learning, the MIT
Press.
11. KP. Murphy (2012), Machine learning a probabilistic perspective, the MIT Press.
Downloaded from http://onepetro.org/SPEAPOG/proceedings-pdf/18APOG/2-18APOG/D021S016R002/1220497/spe-192002-ms.pdf/1 by Vedanta Limited - Cairn Oil & Gas user on 28 June 2023