An asset-intensive business
The oil industry has to put more of its capital into physical assets, including drilling platforms, pipelines, pumps, and tanks, than many other industries. And these complex assets are often in remote and challenging environments. Repairs are expensive, and the business consequences for service interruption are significant.
New IoT implementations are helping oil field operations keep assets functioning, and, by guaranteeing uptime, allowing companies to reduce the amount of expensive backup equipment they need to keep in inventory. The money that is no longer tied up in replacement assets can be used more productively elsewhere.
In an industry stressed by low oil prices, these savings are significant. The old manufacturing term for getting the maximum out of assets is “sweating the assets”. The IoT helps the industry get the most out of assets, while keeping them in operation.
For a good example of how IoT-enabled asset management in the oil industry works in practice, it helps to look at something that might seem fairly mundane: pumps.
No more gushers
In old movies, an oil strike is always announced by a gusher, with an oil fountain high into the air. That remains the iconic image of oil drilling.
Gushers were always expensive disasters, injuring workers and destroying equipment, but aside from that, most new wells simply don’t have that kind of pressure, and what pressure there is drops as the well matures.
So getting the oil out requires some kind of artificial lift, in the form of pumps.
One of the most common pump types in current use is the electric submersible pump (ESP). According to GE Oil & Gas, there are more than 130,000 ESPs installed worldwide, and more than 60 percent of the world’s oil production comes through them. These pumps work in land and sea bed wells up to 12,000 feet deep, at temperatures up to 300° F, pumping oil at pressures up to 5,000 pounds per square inch. They are the workhorses of modern oil production.
The costs of downtime
Pumping crude oil is hard on equipment. Oil of variable viscosity comes combined with water, gas, sand, or scale, all of which affect pump performance and life. Because of the variety of challenges the pumps can face, ESP failure has been hard to analyze and predict.
And many wells are themselves are in remote and inhospitable locations, where asset replacement is difficult and expensive. Hilcorp Energy Company estimates that an ESP failure in one of its offshore pumps in Alaska can cost the company $100,000 to $300,000 a day in lost production.
The industry has been aware the knowledge of how ESPs fail can only be based on extensive data from all types of installations, so they have established a database for tracking those failures, ESP-RIFTS. These are mostly not failures tracked by sensors, but post-failure analyses of over 107,000 ESP failures worldwide, collected since 1999.
IoT sensors will increasingly provide real-time data on ESP operations and failures. But this already-collected historical data is invaluable for creating algorithms to predict and anticipate pump failure.
Giving visibility down the well
Installed ESPs no longer just vanish down a well until oil stops coming up and they need to be replaced. Sensors now monitor current leakage, discharge and intake pressures, intake and winding coil temperatures, and vibration. Each of these is associated with particular issues that can lead to damage and failure—and there are many different ways for such a pump to fail.
Even as sensors and data processing give operators a clearer view of the state of the pump, the pumps themselves are becoming more responsive and flexible. Changes in viscosity, temperature, and amount of solids can cause pump malfunctions, and eventually, failure. In response to sensor data, the pump can slowed to reduce vibration, sped up to move more cooling fluid past the motor, or even reversed to remove solids.
All of these can significantly increase both pump efficiency and life. Sensor networks also allow a single engineer to monitor vastly more pumps than before. Given the anticipated shortfall in trained personnel the oil industry is facing, this is badly needed.
And if an ESP is finally reaching the end of its useful life, tracking when it will fail allows for maintenance efficiencies, by scheduling maintenance and replacement of other parts of the well for the same maintenance visit that replaces the ESP.
Predicting and anticipating performance problems
Companies like Ayata have been working on implementations to monitor and improve ESP lifetime performance through data analytics. Each small increase in efficiency and pump life have a significant effect on overall production costs, allowing oil producers to compete even as prices are low.
Image by Paul Lowry on Flickr (CC by 2.0)