● AI · Unconventionals · CBM ANN ANFIS Modern Applied Science · 2017 Open Access · CC BY 4.0

CBM reservoir property
prediction from
open-hole logs

ANN and ANFIS pipelines trained to infer three key coalbed methane reservoir properties — gas content, Langmuir parameters, and permeability — directly from open-hole log data, benchmarked across six training algorithms on 315 field-scale CBM datasets from lignite to bituminous.

315
Datasets · lignite–bituminous
6
Training algorithms benchmarked
7
Key CBM properties predicted
0.032
ANFIS NRMSE · gas content

The three key properties of CBM reservoirs

CBM differs from conventional reservoirs in a fundamental way: it acts simultaneously as source rock and reservoir rock. Gas is stored in the coal matrix through adsorption, not as free gas in pore space. Three properties govern its behavior:

  • Gas content — volume of gas per tonne of coal (scf/ton); determines IGIP
  • Langmuir volume (VL) — maximum gas adsorption capacity at infinite pressure
  • Langmuir pressure (PL) — pressure at which half of VL is adsorbed; governs desorption dynamics
  • Permeability — cleat system flow capacity; rate-limiting factor for production economics

The measurement problem

Direct measurement requires core samples and specialized laboratory analysis — a significant budget item not available for every well.

Open-hole logs, however, are routinely acquired in every well. If a reliable predictive relationship can be established between log responses and these properties, CBM characterization cost drops significantly across an asset.

This research covers CH₄ and CO₂ Langmuir parameters — the two dominant gas compositions in the study area.

Seven log inputs → six CBM property outputs

/ Inputs — Open-hole log data

  • Depth
  • (C−B)/B — caliper deviation
  • Gamma Ray (GR)
  • Shallow Resistivity
  • Deep Resistivity (log scale)
  • Neutron (NPHI)
  • Density (RHOB)

/ Outputs — Three key properties

  • Gas Content (scf/ton)
  • VL CH₄ — Langmuir volume
  • PL CH₄ — Langmuir pressure
  • VL CO₂ — Langmuir volume
  • PL CO₂ — Langmuir pressure
  • Permeability (md)

Resistivity data converted to log scale before normalization. All inputs normalized to range [−1, 1]. All outputs converted to logarithmic form to avoid negative prediction artifacts. Total: 315 datasets from lignite (young) to bituminous (mature) coal rank.

ANN training algorithms benchmarked

Network architecture: 4-16-1. Six training algorithms tested, with two generalization techniques (early stopping and Bayesian regularization):

  • ANN_GDX — Gradient descent + adaptive learning rate
  • ANN_LM — Levenberg-Marquardt (fastest for function approximation)
  • ANN_SCG — Scaled conjugate gradient
  • ANN_RP — Resilient backpropagation
  • ANN_BR — Bayesian regularization (uses all data, no validation split needed)

ANFIS configuration

Network: 16-4-1. Cluster radius 0.35. Hybrid optimization combining backpropagation and least squares. Early stopping technique for generalization.

Validation dataset: 4% of total data (training + validation split). Process run to maximum 1,000 iterations. NRMSE used as performance metric throughout.

Process flow: Data prep → Pattern/function identification → Generate prediction networks with generalization → Select best technique per property.

ANN_LM, ANFIS, and ANN_BR dominate across all properties

Property ANN_GDX ANN_LM ANN_RP ANN_SCG ANN_BR ANFIS Best
Gas Content 0.160 0.040 0.137 0.134 0.241 0.032 ANFIS
VL CH₄ 0.165 0.127 0.158 0.164 0.169 0.106 ANFIS
PL CH₄ 0.073 0.032 0.060 0.055 0.034 27.192 ANN_LM
VL CO₂ 0.220 0.150 0.172 0.235 0.196 0.462 ANN_LM
PL CO₂ 0.088 0.015 0.094 0.072 0.091 0.120 ANN_LM
Permeability 0.071 0.054 0.079 0.046 0.018 0.182 ANN_BR

Rather than selecting a single best technique, the study recommends the three best techniques per property to capture the range of prediction uncertainty:

Gas Content
ANFIS · ANN_LM · ANN_SCG
VL CH₄
ANFIS · ANN_LM · ANN_RP
PL CH₄
ANN_LM · ANN_BR · ANN_SCG
VL CO₂
ANN_LM · ANN_RP · ANN_BR
PL CO₂
ANN_LM · ANN_SCG · ANN_GDX
Permeability
ANN_BR · ANN_SCG · ANN_LM

Conclusion: Both ANN and ANFIS successfully identify the function between open-hole log data and the three key CBM properties. ANN_LM, ANFIS, and ANN_BR are the best overall algorithms. Using multiple techniques per property provides a range of uncertainty for CBM reservoir evaluation — a more honest representation of prediction confidence than a single-model output.

/ Journal paper · Open access
Hadad, A.F., Rachmat, S., Ariadji, T. & Sidarto, K.A. (2017).
The Prediction of Three Key Properties on Coalbed Methane Reservoir Using Artificial Intelligence.
Modern Applied Science, 11(8), 57.
DOI: 10.5539/mas.v11n8p57 · Published by Canadian Center of Science and Education
License: Creative Commons Attribution 4.0 (CC BY 4.0)