System-wide Energy Savings Analysis Based on CVR Deployment in Sample Feeders | IEEE Conference Publication | IEEE Xplore

System-wide Energy Savings Analysis Based on CVR Deployment in Sample Feeders


Abstract:

The estimation of energy savings analysis based on Conservation Voltage Reduction (CVR) deployment is fundamental to establish its effectiveness. Historically, utilities ...Show More

Abstract:

The estimation of energy savings analysis based on Conservation Voltage Reduction (CVR) deployment is fundamental to establish its effectiveness. Historically, utilities conduct pilot CVR projects on a limited number of feeders where the time-series (TS) voltage and power data are measured, preprocessed and utilized in a mathematical model for energy saving calculation for each individual CVR deployed feeder. The challenge is establishing energy savings in large-scale projects where CVR is running on thousands of feeders and feeder by feeder data processing and savings analysis could be cumbersome. This paper proposes a methodology to estimate system-wide energy savings by creating sub-clusters of the CVR feeders. Initially, a combined approach based on random sampling and K-means clustering is introduced to determine the representative sample feeders required to perform CVR out of a large feeder population. Following the results of the proposed clustering approach, a mapping technique based on Euclidian distance methodology is developed to link each non-sample feeder to a sample feeder and create sub-clusters of non-sample feeders. In addition, the proposed methodology benefits from a developed extrapolation algorithm where savings of the sample feeders are estimated even though the CVR data is unavailable during some periods of the analysis. Next, energy savings of sample feeders are utilized to estimate the energy savings of the non-sample feeders, and eventually, to estimate the energy savings of the entire sub-clusters. Finally, the total system-wide energy savings is calculated by summing up the energy savings of all the sub-clusters. The proposed approach is tested using real field data from a large-scale CVR program within a utility service territory.
Date of Conference: 29 September 2019 - 03 October 2019
Date Added to IEEE Xplore: 28 November 2019
ISBN Information:

ISSN Information:

Conference Location: Baltimore, MD, USA

References

References is not available for this document.