A novel solution to use Big Data technologies and improve demand response program in aggregated residential houses

Demand Response (DR) is a great way to save energy and revenue, it could be beneficial to both consumers and the utility companies. Aggregating residential houses and then analyzing DR on these can help understand and manage DR better. Aggregation can also help the utilities understand the patterns of electrical consumption for a geographical region, area, community. This pattern learning process could further help utility companies to plan the resources appropriately and save revenue while purchasing power and help the power producers to commission new power plants appropriately. Big Data technologies can further make things easier and give granular control over data that is generated from various electrical measuring instruments and climate control equipment installed at a residential building. In this research work, a group of residential houses of particular area were simulated on GridLAB-D to generate power consumption data and indoor temperature data for every minute during the month of July. Big Data technologies like Apache Spark and Apache Cassandra were used to analyze the data and run real time jobs to calculate DR times and unit level DR lift off times to meet user specified comfort levels and compliance standards. Only heating ventilation and air conditioning (HVAC) systems on the remote houses are turned off during the DR times, and two way smart thermostats are considered to be present in the units to pick up the DR signals.