Integrating "Large Data" into a Distribution System Digital Twin Model

Distribution system modeling has evolved over several decades from performing manual calculations to a software data-driven approach which is widely employed. The term 'Digital Twin' is the latest concept in the industry used to describe a data-driven and computerized approach to distribution system modeling. Although the definition for a 'Digital Twin Model' is still evolving it essentially describes an approach that consists of integrating real data from multiple information sources into a holistic computer model that accurately represents the behavior of a distribution system. It can be used as a tool to facilitate distribution system planning, asset management, system performance assessment, water quality evaluation, distribution system security/vulnerability assessments from outside actors, energy management and fire flow capacity assessment. Developing and maintaining a 'Digital Twin' distribution system model is a continuous process that relies on regular updates and enhancements as new technology and information become available. As the ability to capture additional and more precise information evolves with technological advances, the opportunity arises to integrate this information into a comprehensive system model. The City of Kansas City, Missouri (KC Water) maintains an all-pipes distribution system model that was constructed and calibrated in 2013 and was used for master planning efforts. This model consists of approximately 2,800 miles of pipes with approximately 170,000 service connections including multiple wholesale and large-use customers. Because the distribution system model was developed for planning purposes there were limitations in using the model for some of the 'what-if' scenarios that could support the decision-making process, especially since the model scenarios were tailored to an average day and maximum day/maximum hour scenario as is common for distribution system planning. KC Water desired the development of operational scenarios that represented more than just a design demand condition for planning; scenarios which considered the more typical everyday use during summer and winter conditions based on a data-driven approach and using observed data from the Supervisory Control and Data Acquisition (SCADA) system and from the Advanced Metering System (AMR). Additionally, training of the KC Water staff to use these operational modeling scenarios and to be able to adjust scenarios to perform ever day 'what-if' evaluations was also a priority. This presentation will detail the data-driven approach of analyzing large data and integrating this information into operational scenarios to represent typical winter day and summer day conditions. This approach included the development of dashboards to manage and review the large data being captured in SCADA and in the AMR system and to be able to quickly compare and illustrate the model results from various scenarios for users that might not have access to modeling software.