ARRA Interconnection Planning - Load Modeling Activities

The projects described in this report are the result of a multi-year effort by the U.S. Department of Energy to assist the Western Electricity Coordinating Council’s (WECC’s) Load Modeling Task Force (LMTF) as they attempt to address these modeling challenges. Although there are many tasks funded under the American Recovery and Reinvestment Act (ARRA) Interconnection Planning activities, this report updates Pacific Northwest National Laboratory’s (PNNL’s) progress on the following seven aspects: (1) Load Composition Model (LCM), (2) WECC support, (3) upgrade of commercial load composition; (4) aggregate motor protection model; (5) protection model of single-phase motors; (6) software tools to support WECC Joint Synchronized Information Subcommittee (JSIS) and Modeling and Validation Work Group (MVWG) activities; and (7) motor response and protection in commercial buildings.

Section 2: Load Composition Model: PNNL developed the LCM and created a MATLAB version. The LCM is based on a component-based approach. The component-based load modeling approach is a bottom-up technique that consists of aggregating distribution loads according to standard load classes, specifying the types of devices comprised by those load classes, and then assigning the appropriate composition of the aggregated load to the various components of a suitable load model structure.Section 3: WECC support: PNNL supported the WECC LMTF through the Phase 1 CMPLDW implementation. We were able to collect data from Bonneville Power Administration (BPA), the Salt River Project, Puget Sound Energy, Pacific Gas and Electric, and IdahoPower. This task included load composition and load shape analysis for these WECC members. We also developed a database for LCMs and utility data.Section 4: Upgrade of commercial load composition: PNNL revised load composition data sets for commercial buildings that are temperature sensitive. A total of 96 simulations in EnergyPlus software have been done for eight commercial building types in 12 climate zones. Commercial building data from the California Commercial End-Use Survey was upgraded by adding the HVAC load temperature sensitivity.Section 5: Aggregate load protection model: PNNL developed an aggregated motor protection model for studying FIDVR events. The model considers motor behaviors in several different stages in sequence: pre-fault, fault, clear, reclose, recovery/trip, and restart. Different states, On, Off, Start, Stall, and Trip state, are used to describe motor responses in the voltage event. During FIDVR events, some fraction of the motors in each state transition to other states. The number of motors in each state changes depending on voltage level and the time duration in the fault. As a result, the aggregate power demand dynamically changes as the time increases.Section 6: Protection model of single-phase motor: PNNL improved a single-phase motor model with motor protective device action. The model mimics contactor and thermal protection responses of motors, and captures the sensitivities of motor real and reactive power requirements to voltage and frequency.Section 7: Software tools to support WECC, JSIS, and MVWG activities: PNNL has developed a Load Model Data Tool, Frequency Response Analysis Tool, and Power Plant Model Validation Tool in cooperation with BPA to support the WECC JSIS and MVWG.Section 8: Motor response and protection in commercial buildings: PNNL developed building motor response tables for a given set of commercial building types with voltage variances and time frames of interest. We categorized motor protection and control responses for these voltage variances and times along with energy management system control logic restart times and motor protection trip delays. We identified “make and break” times of relays and contactors for the given voltage variances and recovery times. The results are shown in a set of tables giving our estimated motor response for the type of motor load, type of building, and level and duration of the voltage variance.