Data Analytics An Answer To Faulty Pumps

Avoiding pump failure is a top equipment goal for water utilities because even minor pump problems can create big messes. But experts see hope for improvement in pump management by means of data analytics.

Officials at Aransas Pass wastewater treatment plant in Texas experienced pump problems first hand last month when a motor on a pump at the plant burned out. That sent 800,000 gallons of treated wastewater, including sludge, into Redfish Bay. The utility sent out a boil-water notice, which is now lifted, and replaced the pump.

“The trouble started at 4 a.m. when the motor on a pump at the plant burned out, sending some of the solid material to the next phase of treatment and into the bay. That material only comprised a small percentage of the total discharge,” KRISTV reported.

“State environmental officials determined that only a small amount of the leaked water contained raw sewage and wasn't enough to create a hazard, so the all clear was given,” KZTV reported.

Fernando Quintanilla, public works director for the city of Aransas Pass, described the extent of the damage. "Out of the 800,000 gallons that went out, two percent of that was actually the sludge, which was the solids," he told KRIS 6 News. "So it was the majority of the effluent water that was treated."

Aranas Pass is hardly alone. In December, heavy rain and pump failure “caused three sewer manholes to overflow with a combination of storm water and diluted wastewater” in Frisco, TX, according to the Frisco Enterprise.

Experts say data collection promises to improve management of both functioning and faulty pumps. Matt Wells, general manager of GE’s Automation Software business, pointed to ways for utilities to lower costs by improving the way they oversee pumps.

“If you have to move water with pumps rather than gravity feed, that’s an energy-intensive application,” he said, per Automation World. “One water plant in Australia, for example, schedules their pumping activity when spot prices for electricity are at their lowest.”

“[Analytics is key] because you’re able to see data within context,” Wells said, per the report. “Rather than looking at the data by individual asset, the power comes from being able to see the specific measurement points within a process that indicate when it’s performing sub-optimally. The key is defining which multi-variable data sets are most relevant for predicting a specific problem.”

For instance, vibration levels and outlet temperatures are often essential variables in spotting pump failure.

“By continuously pulling from a wide range of data sources, you can present information to operators that enables them to dig deeper to find out what’s really happening and why,” Wells adds. “As more of today’s workers retire and new ones without the domain expertise are hired, it will be essential to be able to quickly draw operator attention to what information they need to focus on most.”

For more on managing wastewater pumps, visit Water Online’s Pump Station Monitoring Solutions Center.