There are oil fields in Saudi Arabia with many, long-lived wells and not enough inspection equipment to frequently check all the pipes for corrosion damage.
“With a limited number of corrosion logging tools, we cannot cover all of them in 1 or 2 years,” said Mohammed AlAjmi, a production engineer for Saudi Aramco, who estimated that it would take 5 years with the available equipment in a presentation at the SPE Digital Energy Conference and Exhibition in The Woodlands, Texas.
To reduce the risk of undetected corrosion damage inside and outside the casing, AlAjmi developed a method to predict which wells were at the highest risk of damage using data-driven models and artificial intelligence methods.
Each well’s location, depth, normal thickness, and age is inputted into an artificial neural network model, which based on its training, develops ways to predict how much steel is left in the walls of the casing by recognizing correlations among the multiple inputs. The strongest connection was based on location in a field, where certain parts of the formation are more corrosive, and by depth, since shallower aquifers tend to cause greater damage.
There were 131 data sets for training and testing the model used to predict spots with the highest risk of pipe corrosion damage. The paper (SPE 173422) said there was a good correlation between what the data sets predicted and the damage detected using logging. The highest correlation (R value of 0.93) was in wells with the greatest level of metal loss, which represent the greatest risk, the paper said.
Stephen Rassenfoss, JPT Emerging Technology Senior Editor
01 May 2015
Seeq’s Focus on Time-Series Data Draws in Chevron, Shell, and Pioneer
The 5-year-old software startup is getting noticed by the oil and gas industry for its ability to accelerate analytics projects by taking care of all the tedious work involved with data wrangling.
BP and Startup Beyond Limits Try To Prove That Cognitive AI Is Ready for Oil and Gas
BP has invested more than $100 million into nine different startup companies in the past 2 years—but only one of them wants to turn your brain into a piece of its software.
Simulation Algorithm Benefits by Connecting Geostatistics With Unsupervised Learning
A new geostatistics modeling methodology that connects geostatistics and machine-learning methodologies, uses nonlinear topological mapping to reduce the original high-dimensional data space, and uses unsupervised-learning algorithms to bypass problems with supervised-learning algorithms.
Don't miss out on the latest technology delivered to your email weekly. Sign up for the JPT newsletter. If you are not logged in, you will receive a confirmation email that you will need to click on to confirm you want to receive the newsletter.
09 October 2018
04 October 2018
11 October 2018