Using geographic information systems to assess potential of targeted demand side management programs
Predictive analytics using geographic information systems (GIS) is a tool that can identify potential customers for utility demand side management (DSM) programs. The insights from this class of analytics come from developing statistical relationships between customer attributes and geographical information. In a previous blog post we highlighted how this can work for energy efficiency programs. In this post we discuss how geospatial analytics for a targeted DSM approach can inform a customer program recruiting strategy to meet overall goals for the program while focusing on areas of distribution systems with capacity constraints.
Utilities can use targeted DSM as a tool to reduce load on specific circuits. Targeted DSM can be a cost-effective alternative to building new infrastructure on certain overloaded circuits, or on circuits where the utilities expect load growth on the circuit to exceed the rated loading for the distribution system lines. An example application is calling demand response events only on overloaded circuits when needed. This has a benefit of not having to disrupt customers in other parts of the utility’s system when there is no need to reduce their load.
To implement targeted DSM, planners need to know where to target; GIS allows utilities to identify customers who are more likely to participate in their programs than others within areas of high need from a system perspective. To make use of GIS in this application, customer attributes data, as well as customer premise-to-circuit-to-substation topological information needs to first be geo-coded. Next, using GIS, analysts develop statistical relationships between the two sets of data, accounting for the multiple potential benefits. Analysts then translate these modeled results into an actionable recruiting plan for DSM program managers. In a study for Louisville Gas & Electric and Kentucky Utilities, we used GIS to determine where the highest potential was for customer engagement in dynamic pricing. Figure 1 shows the likelihood by zip code from this analysis. In that study we analyzed potential market size using the likelihood of participation and census urban-rural classifications.
To take the data from that study and incorporate it for targeted DSM, the next step would be integrating the topography of distribution systems. These distribution systems would include a risk rating, which could be the frequency with which the circuit is overloaded, or the difference between the expected load and capacity. Systems where the highest participation potential intersects with the highest risk potential for targeted DSM would indicate where the utility should focus their efforts on recruiting.
Bringing together DSM and distribution planning to collect and analyze company data can help develop solutions for dealing with specific capacity constraints in a more cost-effective way. GIS provides a tool to combine these different data sources into a common analysis.
 LG&E and KU Smart Meter Business Case Assessment, DNV KEMA, December, 2013. https://lge-ku.com/sites/default/files/documents/lge_ku_dsm_ee_app_011714.pdf. Accessed November, 2014.