Air Conditioner Compressor Performance Model

TitleAir Conditioner Compressor Performance Model
Publication TypeReport
Year of Publication2008
AuthorsN. Lu, YL Xie, Zhenyu Huang
Date Published08/2008
InstitutionPacific Northwest National Laboratory
CityRichland, WA
TypePNNL
ISBN NumberPNNL-17796
KeywordsFIDVR, FIDVR-007, RTINA
Abstract

In recent years, phasor measurement units (PMUs) have recorded more and more slow voltage recovery (SVR) events in the electrical grid, especially in areas supplying heavy air-conditioning loads. The sustained low voltages have caused power quality issues and raised concerns about system voltage collapse. Meanwhile, the significant reduction in voltage stability margin may lead to a similar chain of events. For these reasons, it is critical that SVR events be modeled and prevented to assure that the system voltage recovers to its nominal value in a few seconds.

Fundamentally, SVR events are caused by stalled compressors inside single-phase air-conditioning (SPAC) units, which can not be simulated satisfactorily by either three-phase motor models or ZIP models. Therefore, a new modeling approach is needed. During the past years, the Western Electricity Coordinating Council (WECC) Load Modeling Task Force (LMTF) has led the effort to develop the new modeling approach. As part of this effort, the Bonneville Power Administration (BPA), Southern California Edison (SCE), and Electric Power Research Institute (EPRI) Solutions tested 27 residential air-conditioning units to assess their response to delayed voltage recovery transients. After completing these tests, different modeling approaches were proposed, among them a performance modeling approach that proved to be one of the three favored for its simplicity and ability to recreate different SVR events satisfactorily.

Funded by the California Energy Commission (CEC) via Lawrence Berkeley National Laboratory under its load modeling project, researchers at Pacific Northwest National Laboratory (PNNL) led the follow-on task to analyze the motor testing data to derive the parameters needed to develop a performance
model for the SPAC unit. To derive the performance model, PNNL researchers first used the motor voltage and frequency ramping test data to obtain the real (P) and reactive (Q) power versus voltage (V) and frequency (f) curves. Then, curve fitting was used to develop the P-V, Q-V, P-f, and Q-f relationships for motor running and stalling states. The resulting performance model ignores the dynamic response of the air-conditioning motor. Because the inertia of the air-conditioning motor is very small (H

URLhttp://www.pnl.gov/main/publications/external/technical_reports/pnnl-17796.pdf