AGI Seminar Anderson Nov 19

Distributed Architecture for T&D Data Integration in a Multi-Cloud Testbed
With rapid increases in grid edge devices beginning to affect the bulk power system, transmission utilities are seeking ways to gain visibility of the behavior of distributed energy resources (DER) in aggregate. The amount of data associated with these devices can quickly overwhelm traditional centralized architectures and point-to-point communications between utilities (such as ICCP links). An alternative approach is to implement a hierarchical approach, extending the concept of the IEC 61968-1 Enterprise Service Bus to the grid edge. A reference implementation based on the GridAPPS-D and VOLTTRON platforms running on 28 Raspberry PI edge nodes are used to provide data integration for transmission and distribution utilities in Vermont. The Common Information Model (CIM) is used to create a single-source-of-truth data environment for mapping streaming SCADA and AMI data polled via Azure, AWS, and Google Cloud environments to the power flow model. The combined model is used to develop novel adaptive under-frequency load shedding (UFLS) and volt-var optimization (VVO) apps.