Driving asset reliability and minimizing downtime

The client

A major oil producer on Alaska’s North Slope

The situation

The client’s Anchorage-based Asset Reliability and Maintenance team monitors the status of equipment – which ranges from turbines to pumps to motors and pipes – used to extract oil on Alaska’s North Slope.  The key challenge was that equipment efficiency values were calculated and transmitted in different ways: manual entry by the operator, the operator read the value to a member of the team over the phone or the asset did the calculation. The result was that there was no single, accurate data source that the team could depend on to monitor the reliability and availability of all assets deployed in the field.

Because the data was inconsistent, it was challenging to get an accurate view of each machine’s status. This made it difficult for the team to schedule preventive maintenance activities which would help increase the reliability of the assets being managed. Furthermore, the Asset Reliability and Maintenance team did not have the expertise needed to develop a solution. Not only did the knowledge required span several departments including operations, maintenance, IT and engineering, but the team also did not have deep knowledge of the monitoring systems in place.

The solution

We assembled a team to take on this challenge comprising two automation engineers, two software developers, one software architect and a project manager. A key foundational philosophy that guided our approach was that calculations should be performed in as close physical proximity to the equipment as possible. Because our team is experienced with the legacy systems in place, we modified the control system’s code to support the Overall Equipment Efficiency (OEE) calculation. This ensured that the calculation occurred at the same level as the stop or start event. The next task was to make the data visible on the process control network by integrating legacy equipment.

Finally, we had to overcome the challenge that data is gathered at different frequencies: millisecond to seconds at the automation layer, every 30 seconds at the network level and daily/weekly/monthly at the reporting level. We installed a data concentrator that interfaces with the process control network and the process information network. This allows us to collect real-time data and historize it for reporting purposes. Reports are made available via email and the Machine Train Operational Status (MTOS) website.

Impact

By automating OEE calculations, we have empowered the Asset Reliability and Maintenance team to monitor the status of equipment accurately. This allows them to schedule preventative maintenance in a much more efficient way. The real benefit is that the team is able to minimize downtime and achieve reliability targets of 98%. And the operators on the Slope are freed up from manual data entry and reporting. By un-stranding legacy systems in place, we have also expanded the use of predictive analytics by bringing data that was previously inaccessible into GE SmartSignal.

Extending value

Based on the success of the solution, our team has replicated this approach to other processes including monitoring the state of safety equipment in the field.  In accordance with regulations, safety equipment must be tested periodically to confirm that it delivers protections as designed. To test safety equipment, other assets have to be shut down. However, when equipment is shut down for other reasons, safety equipment is often triggered. Because we are collecting information about what triggers safety equipment during a shutdown, we are able to count this as a valid test of the safety functions, thereby eliminating the need for a separate shut down.