One of the major issues in the utility space that directly affects the bottom line is data accuracy and timeliness. Estimates I have heard from utilities of the reliability of their records database (comprised of as-builts), typically range between 40% and 70%. There is a wide consensus that to get the five 9's reliability that is the number one motivation for the smart grid in the US, data accuracy and timeliness have to improve. As Geoff Cameron at AGSI has said, "100% accurate and real-time."
Data volume
In addition the volume of data is increasing, partly because of intelligent devices, like smart meters that can report energy usage every hour or even every 15 minute. Bill Mintz of Alabama Power/Southern Company in his talk said that he is seeing 10,000X growth in the volume of data they have to deal with. Alabama Power has deployed over a million smart meters.
Data quality
Becky Harrison of Progress Energy, soon to be part of Duke Energy, gave a valuable perspective on the process by which an organization recognizes the business value of data and designes processes to maintain and increase that value. Smart grid is a major impetus enabling this to happen. A major initiative at Progress is distribution demand response (DSDR) a key component of which is load flow analysis. Before embarking on this program, load flow analysis for the distribution network was done every couple of months, and I suspect required a significant effort to collect and validate the data that was required by the analysis. To enable DSDR, load flow analyses have to be done every 15 minutes, which meant that business processes had to be changed to ensure high data reliability all the time, not just very two months. One of the tangible business benefits is that DSDR reduced peak demand and Progress did not have to build and maintain two additional gas combustion generating plants.
Part of the process change was involving operations people in the field as part of the data quality initiative. When you look at the information flow, it seems obvious that the field staff should be involved because they work with the outside facilities every day, and if they are encouraged to provide updates and corrections, the records database becomes more reliable. Reliable records help field staff do their jobs better by reducing repeats or returns, increasing their productivity and utlimately improving the bottom line. Geolocation is a critical part of the data that a utility needs to ensure is reliable.
Data latency
Retha Hunsicker of Duke Energy identified data latency as a major challenge. What she means by latency is the time between when raw data is acquired and when it gets converted into actionable information. With data volumes increasing almost exponentially, this means that instead of having to process one meter read every year or every month, every meter is capable of almost 3,000 reads per month. Duke is now able to process meter reads and display the result to customers within 24 hours.
Subscribe
Comments