Intelligent prediction of stable isotope geochemistry of coalbed gas for its origin

Mohammad Asif, Lei Wang, Randy Doyle Hazlett, Yu‐Shu Wu, Keka Ojha

2021

Closed access Metadata (CSL-JSON)
DOI 10.1002/essoar.10508760.1
Data confidence 70 medium
How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

Stable isotope analysis gives the criteria to define the characteristics of the coalbed gas. In this paper machine learning approach was applied for this purpose. There are two fundamental origins of coalbed gas, i.e., thermogenic and biogenic gas. Stable isotope analysis is the primary method to evaluate the origin of coalbed gas.Samples were collected from the previous literatures and artificial neural network (ANN) was developed for calculating stable isotopes of CH 4 , CO 2 and H 2 . First model was trained with around 300 samples and then cross validated with 40 samples. Hydrocarbon, CO 2 -CH 4 (CDMI), Dryness index, Depth and vitrinite reflectance (R o ) have been used as input parameters and stable isotope of three gases were determined. Feed forward back-propagation was extensively used as the optimum network for the effective results. Before feeding into the network, data was scaled down between 0 to 1 using linear normalization. The learning process used 80% of the data, 10% were used for validation and 10 % for testing. for all the process was above 0.9 and overall, it was observed that R 2 =0.97762. The stable isotope of coalbed gas was achieved through this method, viz. , and . Based on the stable isotope of coalbed gas, the coalbed gas was characterized and different results were obtained. The Bernard and CD diagrams were also plotted for the coalbed gas characterization. The validation of predicted values by actual values was also shown in the paper. The current research has vast application on coalbed methane fields as well as conventional and other unconventional gas resources. This is the first kind of research which provides the stable isotopes of coalbed gas using machine learning. As stable isotope of coalbed is necessary to recognize the types of coalbed gas. Stable isotope geochemistry of coalbed gas also gives the primary knowledge to management to assess the secondary recovery of methane. This study provides the thorough knowledge on stable isotope geochemistry of coalbed gas for judging the sweet spot for the secondary recovery of coalbed methane.