In the quest to continually reduce the cost of asset maintenance, new technologies offer realistic and tangible near-term benefits, but not without navigating some hurdles. Considering the broad array of assets for which utilities are responsible — transformers, pumps, vehicles, miles of pipes and cables — and the risks associated with asset failure, a reliable maintenance strategy is a priority for the industry.
Utilities have progressed from primarily a reactive, break and fix “repair” approach towards a preventive maintenance approach. Parts are replaced as they age or on a schedule recommended by the manufacturer. This approach has been facilitated by asset-tracking capabilities in ERP (enterprise resource planning) and by maintenance systems developed for the purpose.
While scheduled preventive maintenance represents progress, it isn’t the ultimate goal. Prescriptive maintenance is the next big step forward in the evolution of asset management. This means moving from planned preventive maintenance to a state where required maintenance is predicted by systems, and a course of action is prescribed. We can call it preventive maintenance with built-in intelligence.
There are clear gains to moving in this direction. Some advantages are more direct, such as getting ahead of equipment failures to hopefully avoid major breakdowns or outages, or avoiding replacing perfectly good parts when not actually required. Other benefits come from having a greater understanding of what components are deployed in the field. For instance, when a mechanic is assigned to repair a component, he knows which parts and tools to bring to the job. Business intelligence of this caliber provides the knowledge to drastically improve repair time and effectiveness.
However, to get to this level of sophistication requires some groundwork that will make the assets themselves better able to communicate the information operators and mechanics need to know.
For this, utilities can leverage M2M (machine to machine) technology and advanced analytics systems. M2M solutions continuously transmit performance information and diagnostic data from field assets to a control center. Analytics tools help extract insights from the data to predict when equipment needs maintenance, repair or replacement before an actual outage occurs.
Moving Towards Intelligent Prevention
M2M uses sensors attached to assets to take preventive maintenance a step further by making it predictive. Sensors continuously generate data about those assets and transmit it to a control center, where it is converted from data to meaningful information in the form of a dashboard or notifications, where a dispatcher monitors and responds to the activity.
For example, a pump component might start vibrating more than normal or the sound of a machine changes in pitch, indicating something unusual is occurring. The sensor picks up on that change and emits an alert to the dashboard — or to a smartphone or other device — warning that the asset is projected to fail in the next 36 hours. This gives the dispatcher or operator a chance to schedule a repair to prevent costly downtime of the equipment. By employing this predictive approach, utilities can maximize their equipment uptime.
With today’s data collection capabilities, we can compile statistics on age, failure rates, hours of use, environmental conditions and inspections for each asset, and even compare them against a whole fleet of like assets. Patterns and trends will emerge that can guide the organization in developing an effective prescriptive maintenance plan. As a result, maintenance is performed according to what the machines tell you, not the calendar or a predetermined metric. We can now repair or replace only the components that need it, when they need it.
Hurdles, the Role of Modernization and the Way Forward
Installing sensors on existing assets may not be possible or realistic. The actual assets in the field may be outdated; the data formats used may not be suitable for sending data. Though there is a push for data standards there are still several proprietary standards in use and no formal standard as of today. Manufacturers may not allow a sensor to be installed, or the cost of doing so is impractical.
Modernizing assets is capital intensive, and for the utilities industry, regulatory and compliance requirements may either accelerate or hinder asset replacement. The business case for early replacement of an asset with a more modern intelligent version has to be weighed against keeping the old model for a longer period. This is where the industry (utilities, manufacturers, technology partners and regulators) can work together to prioritize the most beneficial use cases.
A new era is emerging in asset management with the advancement of technology. New algorithms are being released with greater accuracy, sensors are becoming smaller and less expensive, and the cost of computer hardware continues to drop significantly, all contributing to more innovative solutions. The future of asset maintenance is prescriptive, and each step towards this goal helps companies reduce costs, ensure equipment availability and up-time, increase service reliability, and most important improve safety.