Phasor measurements at BPA

The Bonneville Power Administration (BPA) has been involved in dynamic system measurements since the 1970s.  Telemetered signals from MW, MVAR, & kV transducers were used for AGC as well as system monitoring.  These signals were recorded when a disturbance was discovered and used to analyze power system dynamic response.  WAMS was first used to describe a system of computers that recorded locally signals from transducers at substations.It included simple data file retrieval and merging.  As phasor measurements evolved, the original system has been mostly replaced by the current real-time phasor measurement system described in the following sections.
BPA began its long involvement with phasor measurements with a research contract with Virginia Tech in 1986.  Under this and a complimentary contract from American Electric Power, the first PMUs were developed under the direction of Professor Arun Phadke.  BPA received 2 units for test on the real power system.  As a specialist in measurements in the BPA laboratory, I was assigned the project of designing a suitable test, carrying it out, and analyzing the results.  Little did I realize this would become a whole career.  Installing anything on an operational power system, even for "test", is a challenge in itself.  After many months of obtaining communications, producing design documents and configuring the test units, we were ready.  We built a data concentrator that would receive and record the phasor data and in parallel the analog telemetry that we were already using for system measurement.  The test was set up to see if PMUs could produce comparable measurements of voltage and power as well as reliable phase angle and frequency.  The results showed a high accuracy measurement with very low noise compared with the telemetry.  The dynamic response was sharper and quite comparable.  We also learned a number of things about operation of a digital system, such as requiring a validity check for control commands, like a CRC. Otherwise one can expect odd operation caused by random noise.
However the test did not prove accuracy or response in any quantifiable way, so we initiated a series of tests using a digital model power line.  Our 3-phase signal generators would only produce constant signals or changes with limited dynamic characteristics.  With the digital power line we could produce any signal that we could represent mathematically.  We found that the PMU was not only very accurate but would track any dynamic signal within the conversion passband (we used the 720 sample/second calculation rate most of the dynamic tests). Pretty neat.  All we needed were more PMUs, some improvements in the software, and of course, some money.  It turned out we also needed an application that someone could use.  We got a connection to SCADA where we input our one phase angle measurement.  We also packaged up a portable unit that we used in dynamic voltage collapse and generator excitation boosting tests.  So the project went on standby while we went on to other things, particularly developing a GPS receiver configuration for traveling wave fault location.
 In 1993 EPRI initiated a project using PMUs installed at several utilities in the WECC (then called WSCC) for a wide area control action.  Several of the participating utilities, including BPA, successfully installed these units by the summer of 1996.  BPA recorded data during the August 10, 1996 blackout that showed the growing oscillation as the system became unstable.  [Figure 2]

The measurements from throughout the day also showed the increasing phase angle as the system weakened.  In the aftermath of the event and subsequent analysis, BPA implemented many system improvements, one of which was improving the real-time phasor measurement system. The system at that time consisted of PMUs at 4 substations with modems to provide real-time data communications to the control center.  The data concentrator did not work reliably.  It did not provide adequate data management or recording.  There was no means for system management.  There were no real-time applications.  To rectify these shortcomings, we decided to build a new phasor data concentrator (PDC) with the following features:

• Expandability for a much larger PMU network
• Low latency (delay) throughput
• Error checking of input data
• Flexibility for accepting multiple protocols
• A function to detect event triggers and save files of data
• Recording of all PMU and PDC errors to facilitate management and maintenance
• Development of suitable protocols to accomplish the objectives

To simplify programming and facilitate expansion, we also decided on a distributed processing approach.  The PDC would have multiple processors, each operating independently and delivering data to a table for output.  As the system grew, more processors could be added as needed.  Functions available in other units could be utilized.  The PMUs included trigger functions for event detection, so the PDC did not need to detect system events.  Displays and other applications would be on separate machines doing their own data processing.  Data is the unifying medium, and it is readily distributed using with modern data communications.
In addition to the PDC, we developed several applications to aid in system operations and data utilization. The StreamReader application which displays data in real time and records it continuously with automatic old file removal [Figure 3]. 

 The PhasorFile application will read and display the file data for analysis [Figure 4].

The status monitor application displays system status and records performance statistics. [Figure 5].

The guiding philosophy was to make the system reliable so data was always available, and make the data as accessible as possible.  These applications run on standard PCs with data distribution by network using the IP protocol.  UDP was used rather than TCP in most cases to simplify distribution as well as minimize latency for control applications in the future.  This approach also simplified application and data distribution to anyone who needed it.