Search

• Deep Neural Networks applied to estimate solar panel size, tilt and azimuth. • Solar panel size estimate with mean percentage error of 3.98%. • Tilt and azimuth estimated with absolute error of 2.55 and 4.71 degrees respectively. • Net load data resolution does not heavily impact accuracy. • Proposed approach is robust to mislabeled training data. There is an ever-growing number of photovoltaic (PV) installations in the US and worldwide. Many utilities do not have complete or up-to-date information of the PVs present within their grids. This research presents a deep neural network approach for estimating PV size, tilt, and azimuth using only behind-the-meter data. It is found that the proposed deep neural network (DNN) method can estimate PV size with an error of 2.09% in a data set with fixed tilt and azimuth values and 3.98% in a data set with varying tilt and azimuths. This is a lower error than the benchmark linear regression approach. A net load data resolution of 1 min provides the lowest error when estimating the PV size. The proposed DNN is also reasonably robust to erroneous training data. When applied to estimate PV tilt and azimuth, the proposed method achieves a mean absolute percentage error of 10.1% and 2.8% respectively. These error metrics are 2.0 × and 3.7 × lower, respectively, than the benchmark linear regression achieves. It was observed that a higher data resolution (1 min) does not provide significant gains in accuracy. It is recommended that a data resolution of 60 min is used to reduce the effects of phenomena such as cloud enhancements. The proposed deep neural network approach is also highly robust, maintaining reasonable accuracy with high levels of mislabeled training data. [ABSTRACT FROM AUTHOR]

In this work, we develop a mathematical model and framework to represent rolling-horizon unit commitment (UC) processes with multiple periodicities. In control center operations, UC is solved repeatedly to adjust device commands based on new information about load, generation availability, renewable energy production, and other aspects of system state as uncertain conditions are realized. We develop a three-level model including 24-h UC, rolling-horizon UC (RHUC), and economic dispatch (ED) and give formulations for the three problems including interdependencies. This framework allows for evaluation of, among other things, different periodicities of the rolling horizon problem and the benefits of more accurate forecasting information. Experimental results are shown for a 6-bus system and a 3012-bus system with wind generation in two wind scenarios. Although the generation costs are very similar, the deviation between RHUC schedules and actual deployment is noted to be superior for a 20-min periodicity compared to a 30-min periodicity. Additionally, less reserve is deployed in the 20-min RHUC case. [ABSTRACT FROM AUTHOR]

A method for formulating and solving a decentralized unit commitment problem is presented in this work. The method, which extends the alternating direction method of multipliers (ADMM), is presented along with several heuristics and refinements to mitigate oscillations and traps in local optimality that result from the nonconvexity of unit commitment. We present and discuss the promising results from testing the method on large-scale systems of more than 3000 buses. The scalability observed so far suggests that this method is a practical option for use with large systems and may provide a significant benefit for computational speed. [ABSTRACT FROM AUTHOR]

This paper provides a survey of big data analytics applications and associated implementation issues. The emphasis is placed on applications that are novel and have demonstrated value to the industry, as illustrated using field data and practical applications. The paper reflects on the lessons learned from initial implementations, as well as ideas that are yet to be explored. The various data science trends treated in the literature are outlined, while experiences from applying them in the electricity grid setting are emphasized to pave the way for future applications. The paper ends with opportunities and challenges, as well as implementation goals and strategies for achieving impactful outcomes. [ABSTRACT FROM AUTHOR]

The recent concurrence of electrical grid failure events in time with extreme temperatures is compounding the population health risks of extreme weather episodes. Here, we combine simulated heat exposure data during historical heat wave events in three large U.S. cities to assess the degree to which heat-related mortality and morbidity change in response to a concurrent electrical grid failure event. We develop a novel approach to estimating individually experienced temperature to approximate how personal-level heat exposure changes on an hourly basis, accounting for both outdoor and building-interior exposures. We find the concurrence of a multiday blackout event with heat wave conditions to more than double the estimated rate of heat-related mortality across all three cities, and to require medical attention for between 3% (Atlanta) and more than 50% (Phoenix) of the total urban population in present and future time periods. Our results highlight the need for enhanced electrical grid resilience and support a more spatially expansive use of tree canopy and high albedo roofing materials to lessen heat exposures during compound climate and infrastructure failure events.