How does an oil and gas (O&G) company effectively analyze sensor data being collected on 25,000 wellheads in the West Texas oil fields? If you have 70,000 miles of pipeline, how can you proactively monitor operating data to detect and fix problems before disruption occurs? If you are managing a fleet of vehicles that operate in remote regions, how do you use operating data to increase profitability? What’s the most efficient way for refineries to maximize asset utilization of five to eight million parts?
For all of these questions, the answer is prescriptive maintenance. Prescriptive maintenance uses data aggregation for better decision support. In enables symptoms analysis by gathering information in real time from sensors on and around machinery to automate decision support and the scheduling of corrective action. This proactive measure enables equipment to be refurbished before it fails; to decrease repair time, reduce downtime, improve productivity, and ultimately, overall utilization. This blog identifies some key learnings to convince you to consider a transition to using prescriptive maintenance for your operations.
Why change now?
Maintenance practices have evolved significantly over the past 50 years. We’ve gone from run to failure, to calendar-based maintenance, to maintenance based on usage, to condition-based maintenance, and to reliability centered maintenance. These practices have served us well, but with advances in sensors, machine learning, machine-to-machine (M2M) communication, data analytics, and the Internet of Things (IoT), there are better ways.
Automated, integrated, intelligence-based solutions enable O&G companies to transform maintenance practices and increase process efficiency in unprecedented ways. Process efficiency reduces cost and yields higher output, which translates into higher revenue and greater profitability.
Despite these benefits, many companies are still performing maintenance as they have for the past 30 years. They’re still tracking equipment on spreadsheets and looking at operations on a plant-by-plant basis, or in upstream, side-by-side. The time has come to put an end to this outdated approach.
Companies today want to rationalize solutions globally and establish operational standards. Headquarters wants to know how each unit is performing and if they’re meeting standards and driving performance and productivity. Beyond maintenance, they want systems in place that can keep track of personnel, safeguard operations, and demonstrate compliance. Paper-based, siloed approaches do not support these types of initiatives.
Making the transition to prescriptive maintenance
SAP offers a solution to help O&G companies embrace prescriptive maintenance. The solution, the result of a partnership between Mtell, Rolta, and SAP, enables early action to avoid or reduce the effects of machine degradation and failure. Built on the SAP HANA platform, the solution brings together Mtell’s machine learning technology and Rolta’s analytics and visualization software.
Mtell’s IOT, M2M neural network-based solution tracks operating conditions and correlates them to failure patterns. When a problem is detected, the system runs a series of diagnostics to pinpoint the problem, predict time to fail, and make recommendations for correction. It is ISO 15926-based and has API notification standards built in from American Petroleum Institute.
The solution links directly with asset management systems, such as SAP Plant Maintenance, so failure alerts trigger a notification that includes the failure code with a full description for action at that point in time. You know exactly what’s wrong and what’s required to fix it.
Integration with SAP HANA allows for the real-time analysis of large volumes of O&G asset data. This analysis is then delivered to the Rolta OneView software solution, where it is displayed on one of the 30+ job-specific dashboards designed for O&G operations personnel. It tracks over 400 key performance indicators that are predefined using ISO 15926. Rolta OneView works with SAP BusinessObjects Business Intelligence and the SAP HANA platform, so companies can choose whatever environment is best for them. It’s built for everyday people to quickly use and master for fast time to value.
Previous approaches to maintenance have been based on anomalies detected by data scientists that analyze operating data. There’s a lot of work involved in that approach and a lot of latencies. More importantly, anomaly detection is not as accurate as precise pattern recognition. But for pattern recognition to be effective, you need to have trusted data. That’s why it’s essential to establish prescriptive maintenance programs based on proven technology. The technologies from Mtell, Rolta, and SAP provide this peace of mind.
To learn more read this blog: Fixing problems before they arise in oil and gas.
Feel free to share your views with me about this topic in the comments below.