Duke Energy has applied the Electric Power Research Institute (EPRI) -developed Power Plant Parameter Derivation (PPPD) software tool and methodology extensively to validate models for the generator and excitation system of their synchronous generation fleet.
By using the EPRI PPPD software and the knowledge from the PPPD User’s Group, Duke Energy has validated over 38 (some of which are sister units) generating units. The PPPD software has allowed the Duke staff to perform this work internally. The data captured and used with the PPPD tool was supplied from on-line voltage step tests. Thus, since the data is recorded on-line there is no interruption to commercial operation of the unit and so no costs associated with rescheduling power.
In addition, Duke Energy was able to validate the models on two black-start gas-turbines in a nuclear power plant using PPPD, which helped with Nuclear Regulatory Commission required studies for the plant. Another success was that Duke engineers were able to validate models for a power plant in a critical region of the network and do more accurate planning studies which helped avoid potentially costly system upgrades that may otherwise have been needed based on studies with the old, non-validated, models of the unit.
Keeping the Process In-House and Avoiding Unit Downtime
Previously, Duke Energy had to hire outside consultants to perform the necessary testing for model validation. This typically entails schedule time for performing “staged” testing of the unit during start-up after a scheduled maintenance outage of the unit. Staged testing includes performing a series of predefined tests that include small (~10 percent of the units rating) megawatt and megavar rejection tests and off-line and on-line voltage reference step tests. Units are required to be out of commercial service during such tests, and considerable time may need to be invested in planning for the tests. For example, the testing preparation for one plant required about six months of engineering/procedure writing to develop a nuclear test procedure. The time a unit is out of commercial service for testing can constitute a significant cost, particularly for large base-loaded units. Thus, the old testing approach costs significant time and money, on top of the cost of hiring of a consultant. In addition, Duke reported that based on their experience, the range of cost per generating unit for external consultants to do this work ranges between $15,000 to $30,000, depending on various factors. To be able to perform this work on several hundred units internally, using the PPPD tool and new approach, will result in significant savings over the past practice of taking the unit out of commercial service for staged testing.
What is the PPPD Software Tool?
The PPPD software tool is a simulation program that can be used for validation and parameter estimation of models for synchronous generators and their controls. This software significantly reduces engineering time needed for model validation. Input to the PPPD tool is a combination of DFR data in standard-formats and entry of machine performance information using a simple graphical user interface.
The PPPD tool exports validated model data in an easy to use format for planners that can be used to validate models. Tests of the PPPD tool have shown that reliable models can be developed using on-line recorded responses to system disturbances (e.g. from digital fault recorders), thus avoiding the need for removing generating units from service for testing.
Evolution of the PPPD Software
Work on the PPPD software tool began as a proof of concept software tool for automating parameter fitting using data collected from the staged testing of generators. This is recognized as the traditional method for performing this task. In 2008 and 2009, research was continued in which EPRI extended the software to be able to use on-line data, whether from disturbances or tests, captured by monitoring equipment such as digital fault recorders (DFR) to validate the power plant models. Duke was an early participant in this research. EPRI demonstrated the concept on several power plants including two in WECC and one in ERCOT. The software was completed and initially released in 2009. EPRI now has a software user’s group for utilities who are licensed and applying the tool. Duke Energy is one of many utilities that
is a member of that group.
Value for the Utility
“PPPD has filled a major hole in the process of validating models,” said Chris Schaeffer, senior engineer for Duke Energy. “The data needed for validation is fairly straight forward and can be captured during staged online/offline testing or from online disturbance data if units are equipped with monitoring equipment. The problem in the past was that the effort to develop the models, predict the response to the testing scenario, making the comparison and especially adjusting the model parameters to match test results was very specialized and few had the skills needed. This has required the use of consultants.”
With the PPPD software and the support provided by the PPPD User’s group, a technical specialist knowledgeable with excitation control equipment can quickly develop the needed skills and implement a validation program. “It is key to note that these efforts have required close collaboration of the generator owners and operators and our transmission planner,” added Schaeffer. “It is our intention to use this tool and the expertise we developed to support validation testing efforts on all our plants, which include Nuclear, Coal, Hydro and Gas plants in the Southeast (SERC Region) and Mid West (RFC Region).
Moving the Industry Forward
EPRI is committed to sharing what it has learned with the public and the industry through venues such as the PPPD User’s Group and several publications at conferences. The key benefit of this new and innovative tool is that through either automated monitoring of equipment or performing rather innocuous on-line voltage step tests, one can meet the imminent North American Electric Reliability Corporation (NERC) standards on generator model validation on a routine basis, without the need to take the generating facilities out of service and perform staged tests. Also, the tool is a rather simple and easy to learn software tool that with the appropriate training and experience allows utility staff to perform this type of work themselves and thus reduce internal costs.