Smart meter data detective work in Denmark

Performance tuned data model, powerful new system console and much more

Utilities can learn a great deal from smart meter data. Analytics software is a powerful tool to turn data into value.

In Denmark, even before deploying analytics for their installation of the Siemens EnergyIP smart grid platform, the utility NRGi has moved beyond their initial goal of enabling meter-to-cash benefits. In the last year, NRGi has also started using smart meter data to diagnose and resolve outages and other grid stress.

"We chose flexible smart meters because we wanted options to measure much more than just consumption data," said NRGi Project Manager Poul Berthelsen. He noted that smart meters can record more data than what a utility may routinely collect, which can be gathered only as needed. Also, many smart meters can be reconfigured by remote, on the fly and on an individual basis, to monitor new values in response to situations that arise in the field.

Beyond measuring kWh consumption (measured and in 15-minute intervals, transmitted four times per day), NRGi's smart meters also record maximum and minimum voltage, calculated and reported once daily. (On a three-phase European electrical system, this totals six additional values.) However, these meters still retain a week's worth of hourly current and voltage data, which NRGi can request by remote as needed. The meter's internal memory is divided into three partitions to support this strategy.

"With the current and voltage data, we get to monitor the exact status of the grid," said Berthelsen. "If we notice that voltage is low in one area, we can do something about it, even before customers notice that there's a problem. Analyzing data allows us to generate a ticket and start an investigation. More meter data can help us identify the source of the problem, such as a bad electrical service connection or overloading in that part of the grid."

If consumption and hourly voltage data indicate a problem, and those values alone aren't sufficient to indicate the cause of the problem, NRGi can reconfigure meters to learn more. For instance, voltage data can be collected in five-minute intervals, rather than hourly. "We can reconfigure the meters, gather the new data, and after a day we can start analyzing," said Berthelsen.

This approach helps NRGi address some common problems that save considerable money and time by reducing repair truck rolls. For example, when routine meter data indicated deviations in voltage on a specific feeder line, Berthelsen was able to gather and analyze more data. This determined which street cabinet needed maintenance.

"Previously, we'd have to send a repair person out to check all cabinets in that neighborhood one at a time, measuring this and that. Eventually we'd find the problem -- but that could take a couple of days," he said. "Now we know where the problem is on the grid, before anyone goes anywhere. Now we just send the repair person to a specific cabinet to do a specific job. That's much easier for everyone."

Voltage is a common concern in Europe, where most areas are densely populated. "In Denmark we have a lot of customers on each transformer, up to 900 on each. We need to ensure that voltage levels stay within very tight tolerances -- otherwise a lot of customers can have problems," Berthelsen explained. "But depending on length of the cable, voltage can drop as it goes downstream. The smart meter data helps us decide where to locate and how to adjust transformer to ensure quality and consistency of service. We have all the data we need to do that."

Smart meter data also helps NRGi accommodate renewables. "About 5500 of our 215,000 customers have rooftop solar panels installed. That can increase voltage on the grid -- and we also need to consider that in regulating our transformers," Berthelsen observed.

NRGi's smart meters communicate via a powerline communication network (rather than a radio network, which is more common in less densely populated nations). This offers the added advantage of using the meter communication network itself as a diagnostic too. "When load gets high and we notice that something changes in how a meters is communicating, that can indicate a bad connection," he said. "Also, powerline communications can help identify which phase of the electrical connection might be a problem. Combining different data -- the meter values, and from the communication network -- really helps."

NRGi plans to implement EnergyIP analytics software soon -- but for now, they're realizing operational benefits simply through using spreadsheets with meter data. Berthelsen is eager to implement analytics software. "Right now, I'm really good at spreadsheets. But when you have this much data, you can spend days getting answers out of it," he said. "When you have too much data, it's hard to understand what's going on. It's important to be precise, to only look at the data you really need."