Craig Huitt, TransAlta, Dmitry Kosterev, BPA, and John Undrill, GE
Grid planning and operating decisions rely on simulations of the dynamic behavior of the power system. Both technical and commercial segments of the industry must be confident that the dynamic simulation models and database are accurate and up to date. Optimistic models can result in unsafe operating conditions (such occurred in summer of 1996 in the Western Interconnection), while pessimistic models and assumptions can result in underutilizing transmission capacity-thereby inhibiting most economic generation dispatch.
Model validation studies of the August 10, 1996 Western Interconnection outage could not adequately reproduce the actual system behavior. To restore confidence in simulations and the database, the Western System Coordinating Council (WSCC) instituted a requirement that all (except small) generating units be tested to confirm the data used to represent them in grid simulations.
It became clear that successful testing depends strongly on the availability of fully qualified engineering staff. Confidence in the fundamental data has been restored, improvements in simulation practice have been developed, and first-pass correspondence between new simulations and reference events has been much improved. The effort has also revealed an important new issue: that continual effort is needed to ensure that the database is kept current. WSCC policy required periodic 5-year re-testing of generators for model validation. This requirement was later adopted in “NERC Planning Standards.”
Repeating the original testing is possible. Direct costs of testing have been significant but manageable. But simply repeating original tests, which are made in specially staged conditions, is not the only alternative and does not detect changes occurring between test sessions. Generation equipment testing is not the goal, but rather one of the several vehicles available for validating model data. Bonneville Power Administration (BPA) has been using disturbance recordings collected from its Wide Area Measurement System for power plant performance monitoring. To date, high-speed continuous data monitors have been installed at interconnection points of major power plants totaling 20,000 MW capacity. Power plant monitoring can be economically attractive and a technically viable alternative to repetitive testing.
BPA partnered with TransAlta Centralia Power Project and General Electric Company to demonstrate the performance monitoring approach for generator model validation. Centralia coal-fired power plant is located in western Washington between Seattle and Portland metropolitan areas (see photo). Two generators are rated at 870 MVA each have 750 MW turbines. Because of its size and location, the plant is an important contributor to grid reliability. Centralia generators are providing critical voltage support to loads in western Washington and to power transfers from the Pacific Northwest to California.
The baseline generator testing was performed by BPA, Centralia and GE engineers. The test plan was tailored to address complex operational needs of a large coal-fired generator, yet provided the most informative model data. A combination of off-line and on-line testing was used to establish baseline generator, exciter and control data. All tests were performed with no impact on power production.
One of the BPA monitors was located at Centralia point of interconnection, measuring bus voltages and currents in Centralia powerhouse lines. Several significant disturbances were captured. Recorded voltage and frequency were “played in” as model inputs and powerhouse real and reactive power were “measures of success.” Comparison of recorded and simulated results confirmed the validity of the baseline model for grid voltage and frequency disturbances.
Click here to enlarge image
Having full time dynamic monitors in place has been shown to reduce costs and improve quality in both initial (baseline) testing and in the vital task of continual revalidation. Costs of monitoring are moderate; continual revalidation based on monitoring of natural grid events does not require costly special plant operations or prolongation of outages for testing. The inherent ability of full time monitoring to key model validation to real and varied grid conditions is a major contributor to quality improvement.
Thus there is a sound basis for extension of testing standards to:
“- retain present requirements for initial baseline testing.
“- acknowledge the benefit of revalidation based on monitoring and accept it as an effective alternative to repetitive retesting.
Huitt is an electrical engineer with TransAlta and can be reached at email@example.com. Kosterev is an electrical engineer with the Bonneville Power Administration and can be contacted at firstname.lastname@example.org. Undrill is a principal consultant with General Electric. He can be reached at email@example.com.