Peak Oil is You

Donate Bitcoins ;-) or Paypal :-)

Page added on October 30, 2014

Bookmark and Share

Watson to Explore for Oil as Repsol Taps IBM’s Analytics Machine


Repsol SA (REP), Spain’s largest energy company, will work with IBM’s Watson data-crunching tool to try to make it cheaper and easier to find future caches of oil.

The companies will seek to develop applications to analyze data on geology and crude reserves, the first time Watson will be used in the energy industry, said Santiago Quesada, Repsol’s director for exploration and production technology.

For Repsol, Watson’s ability to analyze troves of data can help boost efficiency and accuracy as it tries to cheapen the development of oil fields, cut the cost of discovering new reserves and find bargains when buying assets. The partnership helps IBM enter a new industry as it tries to prove that its Jeopardy-winning Watson technology can be both useful for clients and make money.

“These technologies aim to extract as much as possible with as little effort and the lowest costs possible,” Quesada said in an interview. “It’s not so much about processing fast, it’s about finding the best solution.”

Working together at International Business Machines Corp.’s Watson headquarters in New York, the companies will each commit six to 10 employees to work on the applications, Quesada said.

Repsol is making an initial investment of $15 million to $20 million to develop two applications targeted for next year, he said. One app will be used for oil exploration and the other to help determine the most attractive oil and gas assets to buy.
Drilling Costs

James Sciales, a spokesman for IBM (IBM), declined to comment on the Repsol partnership and how much revenue it’s expected to bring in.

Cutting-edge computing technology has been key for oil companies pushing into unchartered territories, ranging from the Arctic ice to reserves trapped under 2-kilometer (1.2-mile) layers of salt below the Atlantic ocean. Reducing costs is a key part of cutting exploration risk.

Drilling a well in areas such as off-shore Brazil or Angola, where Repsol operates, can cost $150 million to $300 million and oil fields can typically have five or more wells. Dry wells are also frequent. Subsea wells can go as far as 7,000 meters (23,000 feet) below the water surface. Meanwhile oil prices are near four-year lows.

Watson can process questions asked by humans in natural language and sifts through information to respond with the most likely answers.

IBM has been trying to lure customers to its Watson tool as the company faces increasing pressure to grow newer businesses like data analytics. Last week, Chief Executive Officer Ginni Rometty abandoned a long-held earnings forecast for 2015 in order to focus on growth areas.

The company has won some new customers for Watson after starting a new business unit centered on the technology this year. This week, IBM announced a partnership with Twitter Inc. to analyze the social network’s data to glean information about what a company’s customers are saying about its products.

IBM has yet to disclose sales from the Watson business.


One Comment on "Watson to Explore for Oil as Repsol Taps IBM’s Analytics Machine"

  1. rockman on Thu, 30th Oct 2014 8:38 am 

    It always tickles me when I see implications of computers being better at finding oil/NG then a human. But make no mistake advances in computer tech have made a huge difference in the oil patch in the last 20 years. Especially true in seismic data analysis. On my 3d seismic work station I can produce more of a better product in one month then 3 geophysicists working 2d data in a year. But that’s due to how rapid the computer allows me to manipulate and displays the data as well as doing analysis that were once done with a desk top calculator. Some may still not believe it but the exploration tech advances have made it much easier to find oil/NG then when I started 40 years. Success rates are many times greater now than back then. The problem isn’t identifying places to drill. The problem is a lack of places to drill.

    But understand even with this huge advance in computational power the work station software doesn’t have any interpretation ability. Here’s a very short explanation of why computers can’t do interpretations: they function too logically. Our great strength as geologists: to think illogically. LOL. Some like to shine it up and say we just “think outside the box” but it’s the same thing. Even when they try to apply more “fuzzy logic” software it was still too logical. Think about it: the world is awash with geological and geophysical data. If the data has been around for a while and it clearly indicates a place to drill why hasn’t someone already drilled it? In some areas in the US one can actually see the effects of NG in the seismic data. There’s still a selection process but these seismic amplitude anomalies (“bright spots”) have pushed the success rate above 80% in some areas. But here’s the problem: he that gets that seismic data first wins. You spend $10 million buying data to hunt bright spots that has already been “mined” you just pissed away your money.

    I know my answer is too short to grasp a good sense of it but oil/NG exploration requires a rather illogical way of looking at the geology. One has to look for places to drill that, to a large degree, aren’t logical targets. And therein sits our real problem with fueling our future economy: most oil/NG has been trapped in areas in which one would have logically expected it to be found. But it also had to be technically and economically feasible to drill such areas. For decades there has been obvious reasons to expect the possibility of producible reservoirs in Deep Water areas such as the GOM as well as in the Arctic. About 35 years ago I saw DW GOM seismic data that indicated such potential. Didn’t matter at that time since the max water depth we could work in then was 600’. Obviously that’s no longer a restriction. I drilled and frac’d my first Texas shale formation 35 years ago. A 500,000# record breaking frac just like George Mitchell was doing at the same time. The first Eagle Ford Shale producers were drilled over 60 years ago. Yes: 60 years…not 6 years. The Monterey Shales was first drilled and frac’d decades ago. Again it was logical to expect oil/NG production from shale more than half a century ago. And again what was lacking was the tech and sufficient prices to do it.

    And other then the methane hydrates there’s really little reason left to expect a significant volume of oil/NG to exist in new areas that haven’t been considered. Lots of drilling left to do in the Arctic Basin and DW GOM, Africa, Brazil, etc. But those are known hydrocarbon plays. That’s it: no one in the oil patch is even speculating about other potential areas. But to a fair degree this was understood by the old hands when I started in 1975. Which is when I was first exposed to the idea of PO. For me personally this has been a 40 year long novel in which I knew the ending even though I might not be around to witness the last chapter.

    We’re surround and almost out of ammo. The cavalry (Air or otherwise) isn’t going to come to the rescue. We’re done for. The best you can hope for is that you don’t piss your pants and you go out with a bit of dignity. Yeah…right…that’s what we can expect from mankind. LOL.

Leave a Reply

Your email address will not be published. Required fields are marked *