On meter data and analytics
- Aug 22, 2016 7:45 pm GMTAug 22, 2016 7:53 pm GMT
- 704 views
Co-author: Tom Anderson
Using analytics on your meter data can help detect and remedy non-technical loss (NTL). Some people associate non-technical loss with just theft, but that is only a part of NTL there is also other categories like poor equipment maintenance, faulty equipment, missing meter reads, or even administrative related mistakes in calculations and/or billing. All customers bear the cost of NTL whether theft of otherwise and NTL represents an avoidable financial loss for the utility. In this article we are going to discuss NTL and analytics related to non-theft/fraud especially since some utilities have different departments/groups who focus on different areas of NTL. One essential point to understand is that analytics applied to meter data of any kind, AMR or AMI, can help identify and remedy NTL and allow you to recover more revenue than you were able too prior to using analytics.
AMR provides one-way communication from the meter to the utility and typically has greater interval times between reads, while AMI or “smart meters” provide two-way communication and much more frequent interval data where every 15 minutes is quite typical. You will use different types of analytics depending on the nature/attributes of the data you have today and continue to adjust your analytic techniques in the future as you continue to gather new and additional data that will provide more insights to help you to reduce your overall NTL now and into the future.
Fig 2: Detail of AMR zero reads
In Fig 1 (Analyzing AMR data showing zero meter reads, see lead art for this story) and Fig 2 (directly above) are looking at AMR data of several months/years. Fig 1 shows different areas and different users and their electrical consumption. Fig 2 shows a detail of one customer where they had several zero reads from this meter. In this case a utility could use the average readings for this meter and go back and charge/recover this loss from this particular customer for the time period shown in Fig 2.
In Fig 3 (below) any blue line crossing either the lower control limit (LCL) or the upper control limit (UCL) shows possible NTL which should be investigated further perhaps by sending a technician out to the actual meter. As you start to roll out more AMI type of meters you will have the opportunity to configure these to send more information back to the utility as well as providing new services to the customers based on new insights from your analysis of this data.
Fig 3: Threshold analysis done over time where yellow = UCL, red = LCL, green = predicted value, and blue = actual reading
As you start to improve your overall NTL analysis and recovery of revenue the next step could be to feed this information into your overall load calculations which could then be used to improve your overall load forecasts for not only the short-term, but for the mid and long-term as well. As you can see applying analytics to any type of meter data has the potential to improve your overall bottom-line by reducing your NTL now and into the future. The sooner you start applying analytics to one part of your business the more use cases you will uncover that analytics can help improve your business and NTL is a great place for you to start focusing your use of analytics.