Utilities Cashing In On Smart Technologies

Prodded by renewables to embrace smart technologies, utilities are discovering ways sensors, data, and connectivity can save them millions of dollars.
Michael Reid, a general manager of technical programs for Duke Energy, said Wednesday the company saved $122.3 million in avoided costs over five years after installing 30,000 sensors in 10,000 pieces of equipment at its power plants. The sensors track temperature, pressure, vibration, and other factors, and applications alert staff to any anomalies.
"It’s really a game changer, and we’re really starting to expand these efforts," Reid said in a webinar hosted by the Utility Analytics Institute.
For example, on one occasion the sensors noticed subtle drops in vibration inside a steam turbine. "We usually think, 'Well, if you have lower vibration, that's a good thing,'" Reid said, but the software flagged the issue with an alert, and technicians decided to investigate.
"When we looked inside the device we saw that there was blade damage occurring. Pieces were breaking off of blades and it was actually getting worse and worse. This was a significant find because if we hadn't shut down the machine we could have had a significant failure which have resulted in several months of downtime and significant repair costs."
The effort has changed the way Duke performs maintenance, Reid said, from time-based maintenance, in which parts are changed at a certain age, to condition-based maintenance, in which parts are changed when they need to be changed. The latter approach is cheaper, but it requires "knowing about the health of your equipment."
"It's cheaper to repair the damage early on than wait for a significant failure," he said. "If you have a transformer fire or a steam turbine that has a catastrophic failure, that's a significant impact on the fleet, and it could also have an environmental impact, often resulting in an outage for several months at significant cost."
The system detects problems early, but technicians hope to perfect the process so it becomes prescriptive—noticing the potential for damage before any damage occurs. Nonetheless, not everyone at Duke has been convinced by the millions saved, Reid said. The system is winning support in other ways.
"That alone wasn’t enough to convince everybody, and there are probably people today who still aren’t convinced, but when you start to change people’s lives—the way they do their work—like the maintenance specialists at the stations, that’s when you really start to get more recognition. And that rings true with our plant managers and executives when they see that we’re changing the way work is done, when they see there’s less manual collection of data and more automation."
CPS Energy Products, the San Antonio utility, reported modest savings just from consolidating its data collection systems and building new dashboards that combine relevant kinds of data. In 2014, the company streamlined its data collection platforms. In one year it spent about $1 million on the effort, said Amelia Badders, CPS's director of commercial analytics and pricing, and saved $1.3 million because of it.
For example, the new platform can overlay wind speed, weather, and market-performance analytics to more precisely determine what factors influence the price of wind energy, Badders said, so the company is better positioned to market wind resources.
"In one month alone we increased the value on those by about a half million dollars," she said, adding that renewables like wind prompted CPS to improve its data collection and analysis. With support and prodding from the Obama Administration, some utilities invested in sensors, data and analytics in recent years to prepare for renewables that are likely to be intermittent and distributed.
“It is in large part the need to deal with intermittent resources," Badders said, "that has highlighted the critical need for timely, responsive and reliable information.”