Indian experts find AI-based technique that can help in oil and gas exploration
Scientists and experts at the Wadia Institute of Himalayan Geology (WIHG) have come up with a new artificial intelligence (AI) based technique to analyse data from seismic waves (natural or induced by explosive material) to ascertain the type of rock formation and geological features beneath the surface which could help in exploring hydrocarbons like oil and natural gas in less time with high efficiency. The technique was first tested by WIHG scientists in the Taranaki Basin of New Zealand.
The research on the new technique was conducted by WIHG director Kalachand Sain and WIHG’s other research associate Priyadarshi Chinmoy Kumar. Their research on the technique named ‘Machine learning tool for interpretation of Mass Transport Deposits from seismic data’ was published in the noted journal Nature Scientific Report on August 24.
Terming the new technique a very significant one, the researchers claimed that they have received several requests from foreign countries like Norway, UK, Australia and others to get access to it. Similarly, some India-based oil and natural gas exploration companies have also contacted them to share the research on the new technique.
Director Sain said the new technique interprets various aspects of a sub-surface (beneath the surface) area using the data acquired from the surface in ‘very less time but with high accuracy and efficiency’.
“In the existing technique available for the study of various aspects of the sub-surface area like rock formation, geometry or architecture of rocks, shifting of tectonic plates and others, is mainly done manually which is a very time consuming and laborious process. This new AI-based technique just needs to be fed surface data acquired, to study the sub-surface area and give highly efficient and accurate results in a few seconds,” said Sain.
He said that usually the volume of data acquired from the surface through seismic waves is huge in size and takes many hours and even days to interpret the sub-surface area manually.
“Conventionally, the interpretation is done manually, in a phase-by-phase manner which sometimes takes days to complete that too with chances of human errors which can prove very costly. This machine learning tool or AI-based technique developed by us, interprets the data with minimum human involvement in few seconds with almost zero room for error,” said Sain.
“This technique is very significant in the exploration of oil or natural gas in the sub-surface area and analysis of complex sub-surface features of any area by producing the end result in the form of a 3D image.”
Sain’s co-researcher in finding the technique, Priyadarshi Chinmoy Kumar while revealing more about it said, “the research was being done since about last five years during which they first tested it in Taranaki Basin of New Zealand.’
“During the test, it produced very accurate results. We compared the time taken by it for giving the result with that of conventional one with human involvement, in which we found that it took just 21 seconds while the latter took about 40 hours,” said Kumar.
Explaining the significance of the technique Kumar said, “It is one of the most sought-after techniques in oil and gas exploration industry around the world and in the field of geoscience.”
He informed that in the process of analysing any geo-surface area, there are three phases including acquiring data from seismic waves, its processing and then interpretation. The new AI-based technique becomes important in the interpretation phase.
“In the whole analytical process, the interpretation of processed data is very important. A mistake in it would reflect in the end result which could result in heavy loss of resources, especially in oil and gas exploration. If the end result is faulty, then the people involved in drilling will drill at the wrong place which would result in a significant amount of money and time going to waste. Our technique produces reliable, accurate and efficient interpretation results, almost ruling out the above possibility,” he said.
“The technique could be also very helpful in studying any area having earthquakes so as to take effective measures to curb human life and property loss,” said Kumar.