Skip to Content

Harnessing AMI to Improve Demand Forecasting

For utilities, metering technology affects a wide range of business functions, from billing and spot purchasing to demand forecasting and scenario planning. New technologies such as advanced metering infrastructure (AMI) facilitate automatic metering and ongoing communication of electrical usage patterns. Smart meters make it possible to access usage information at the point of delivery, giving utilities more granular data than ever before. How a utility uses this new data can make the difference between valuable insight and wasted opportunity.

Demand forecasting is critical to a utility’s success, driving a host of important operating decisions – including power generation, power purchasing, and planning service interruptions. The financial stakes are equally high: even a modest improvement in forecasting accuracy can increase revenue by millions of dollars. It’s little wonder that utilities are looking for analytic applications that can help them make the most of smart-meter data.

To incorporate more granular, more frequent data, utilities need demand forecasting models that can handle enormous, information-rich AMI data sets. Such models should also capture demand dynamics and key drivers such as weather. A utility that can use this complex data to derive multiple usage patterns can gain insights into customer behaviors and better manage supply.

AMI data can be clustered based on time (such as day of week or season), location, and demographics to reflect customer usage patterns. Clustering increases statistical significance and maximizes the predictive value of the model. Using meaningful clusters of AMI data, utilities can apply common forecasting algorithms and produce more accurate and detailed load demand forecasts. Clustered data can also be aggregated for more precise and accurate forecasting, as well as insight into customer segmentation.

This insight allows a utility to:

  • Understand seasonal behaviors and consumption patterns at more granular levels
  • Improve overall decision making
  • Improve prediction of future loads to increase revenue
  • Produce short-term demand forecasts at different levels in the network
  • Reduce fuel and operational losses due to unneeded generation

How SAP Can Help

Performance and Insight Optimization Services from SAP uses scientific and technology expertise to help utilities leverage advanced analytics and AMI – and to better understand customer behavior. Using a mathematical methodology, SAP can help model and forecast energy demand at the customer or meter level. The model is based on data clustering of meters that share commonalities, strengthening model quality while increasing statistical significance. Clustering also reduces data dimensionality while preserving the granularity of information contained in the data. By finding the optimal aggregation point where the data is most powerful, clustering enables better decision making based on customer level forecasts and predictable load usage. With predictive analytics, a utility gains insight into specific customer behaviors that can improve demand forecasting – and, ultimately, the bottom line.

For more information on Performance Insight and Optimization Services from SAP, please visit www.sap.com/PredictiveAnalytics.

To report this post you need to login first.

Be the first to leave a comment

You must be Logged on to comment or reply to a post.

Leave a Reply