Smarter Grids with Richer Data for Better Response to Natural Disasters
By Dr. Zhaoyu Wang
Natural disasters can threaten lives, disable communities, and devastate electric generation, transmission, and distribution systems. Keeping the lights on is central to the community recovery from paralyzing disasters. Recently, these extreme weather events have been more frequent and disruptive due to climatic changes. The hurricanes in East Coast, wildfires in Pacific Coast, as well as tornados and winter storms in Midwest in past three years are the most recent examples indicating significant vulnerability of power grids to climatic hazards. A resilient power grid in general should anticipate disruptive events, rapidly recover, and absorb lessons to adapt its operation and structure to mitigate the impact of future events. With such a grid, people will experience fewer and shorter outages in extreme events. One key challenge in enhancing grid resilience is poor information about extreme events and their impacts on power grids. Fortunately, smart metering devices can provide richer data to make utilities more informed. The large amount of field data together with advanced data analytics, distribution automation, and distributed energy resources can help power grids better respond to natural disasters. This article will introduce challenges and opportunities of leveraging data-driven techniques to enhance distribution grid resilience against extreme weather events.
In the past three years, our group has been supported by the U.S. Department of Energy, the U.S. National Science Foundation, and Iowa Economic Development Agency for research on power grid resilience modeling and enhancement. Our work has a unique feature of being driven by real data. We have collected a large amount of outage data and field measurements from utility partners such as Alliant Energy, City of Bloomfield, Maquoketa Valley Electric Cooperative, Algona Municipal Utilities, and Cedar Falls Utilities. While most of existing research focuses on building detailed mathematical models for outages, we believe it is important to extract actionable information from real historic data and develop high-level statistical models to quantify outage occurrence, propagation and recovery. We are using stochastic processes such as Poisson process and Markov chain to model outages, which will help utilities predict their occurrence and costs. Furthermore, these statistical models will provide guidance on optimizing resilience investments and proactive preparation for upcoming events. A major challenge in modeling outages is the sparsity of event data. Therefore, it is necessary to study both small, medium and large outages and identify their common features. A near future direction is to investigate exogenous variables’ impacts on outage models. For example, it is interesting to quantify the relations between weather event severities and outage probabilities and sizes; and to study how different distributed energy resource (DER) penetration levels affect outages.
Besides using historic data to model outages, another challenge is to enhance the real-time situational awareness during events. Lacking observability has always been a main obstacle in restoration. Currently, most utilities still rely on customer phone calls and field inspection to assess damages and identify outages. Recent studies have shown that smart meters and social media have the potential to significantly improve this process. For example, smart meters can send the last gasp before outages and people may discuss or report outages in Twitter. Moreover, system fragility models and weather information can be utilized to estimate damages. The challenge, however, is to integrate these different information sources. A possible solution is to use probabilistic fusion or machine learning techniques to develop a multisource data driven outage assessment, thus offering utilities real-time enhanced situational awareness during events.
In addition to big data analytics, the deployment of distributed energy resources, microgrids, automatic switches and other smart grid technologies provides another opportunity to improve grid’s response to natural disasters. For example, a number of useful engineering and research efforts have explored using topology reconfiguration, automatic microgrid formation, renewable generation and energy storage to deal with extreme weather events. However, fundamental challenges still inhibit effective resilience enhancement. Firstly, these new resources have not been fully utilized and systematically coordinated in disaster response. Secondly, what useful information should be extracted from data and how to use it to guide resilience enhancement needs more investigation. Last but not least, current resilience enhancement efforts are uncoordinated and split into long-term hardening, short-term proactive preparation, and post-event recovery without considering their interactions. We argue a comprehensive resilience that systematically integrates and optimizes various resilience-enhancing phases. Formulating and explaining integrated portfolios of resilience enhancement will make these efforts more effective. Considering most utilities still rely on experiences and heuristics to make decisions in disaster response, it is necessary to study how to take advantage of the richer data to design situation-aware optimization models to help utilities better prepare for and quickly recover from large-scale outages. The models should co-optimize all resources including networked microgrids, distributed energy resources, energy storage, automatic switches and repair crews. Further, long-term investments, pre-event preparation and post-event recovery should be jointly optimized to systematically enhance grid resilience at different time scales. The above optimization will lead to large-scale stochastic mixed integer programs, thus requiring efficient solution algorithms to be designed.
The data-driven analytics and advanced optimization will significantly enhance power grid resilience. However, this requires investments in new metering devices and system upgrades. For the work to be applied to the nation’s power grids, it needs to be communicated effectively. This is a challenge because people are reluctant to invest in future disasters, even when it is highly cost effective. Communicating benefits, costs, and risks to policy makers, engineers, and the public is essential to convince people to invest in resilience when the sun is shining and the sky is blue. Moreover, complex mathematical models cannot be easily accepted by practitioners, policy makers, or the public. It is therefore particularly important to make the investments accountable so that their benefits to both utilities’ and people’s concerns and way of life are clear and quantified. In addition, the resilience benefits of upgrades should be coordinated with their benefits in normal operation, to make a comprehensive case for their value.
Zhaoyu Wang (S’13-M’15) is the Harpole-Pentair Assistant Professor with Iowa State University. He received the B.S. and M.S. degrees in electrical engineering from Shanghai Jiaotong University in 2009 and 2012, respectively, and the M.S. and Ph.D. degrees in electrical and computer engineering from Georgia Institute of Technology in 2012 and 2015, respectively. He was a Research Aid at Argonne National Laboratory in 2013 and an Electrical Engineer Intern at Corning Inc. in 2014. His research interests include power distribution systems, microgrids, renewable integration, power system resilience, and data-driven system modeling. He is the Principal Investigator for a multitude of projects focused on these topics and funded by the National Science Foundation, the Department of Energy, National Laboratories, PSERC, and Iowa Energy Center. Dr. Wang received the IEEE PES General Meeting Best Paper Award in 2017 and the IEEE Industrial Application Society Prize Paper Award in 2016. Dr. Wang is the Secretary of IEEE Power and Energy Society Award Subcommittee. He is an editor of IEEE Transactions on Power Systems, IEEE Transactions on Smart Grid and IEEE PES Letters, and an associate editor of IET Smart Grid.