The most important tool we have to allow for renewable energy growth and decarbonization of our power grid is flexibility. Flexibility could be provided by a customer who can shift their consuming load (demand response), or by a grid-connected battery. The problem is that even when these devices are smart and responsive, they don't have a trusted, open signal from the grid to respond to. We created a data service application for three specific types of User who want to help green the grid. BYODevice Demand Response data service forecasts the load by substation to indicate when the grid will be stressed. It also includes marginal CO2 emissions by hour, and energy market pricing as KPIs: Grid Operator - S/he wants to be able to see crucial grid performance data and share it with their demand response participants. Device Portal - Devices can be set up to automatically call this data service via API to get suggestions for when to schedule their consumption or generation (in the case of a battery). When the device tells the service how far in advance it is planning, the service gives the device back a ranking by hour for when to consume (or generate). Data Scientist Portal - For data scientists and curious folks, the data scientist portal of the app shows how the loads were forecast for substations and feeders, based all on the Predix weather and load data from ESB.

Key Features: load forecasting, Predix data queries, Predix WebApp, PowerBI, Azure Machine Learning Workbench.

Share this project: