Release date: 17 September 2019
What can you do in one minute? Check your phone, write an email, scan the news?
In that time, BP can get the lowdown on each of its wells, thanks to the work of petrophysicist Sarita Salunke. She has built an algorithm and trained a machine to accurately predict the type of rock drillers might encounter in a completely new well.
Understanding the subsurface is vital to oil and gas exploration, but getting to know what lies beneath usually takes a lot of time and money.
Salunke, who has a PhD in physics, has sped up that process and saved a big, but highly confidential, amount of money.
Existing data from five wells supplied by BP’s onshore US team was fed into the machine. The data provides detailed information about what is in the subsurface.
She then tested the machine’s predictions against the readings gathered over 10 days from ‘down-hole tools’ analysing the well. “It was a good blind test,” she says, “and gave the team confidence that we should start using this technology more widely.”
Salunke is now working with teams in Oman, Azerbaijan and the Gulf of Mexico (GoM) to implement the technology in their operations. She hopes to create a version of the software that will send data back from the wells to the drilling team in real time.
Sarita Salunke, petrophysicist
“We currently have about a four-hour lag on the data,” she explains, “but to be able to have a system that tells us how the rock is going to behave and prepare for that as we drill would be invaluable.”