Interconnection-Wide Oscillation Analysis: Baselining Oscillation Modes in the North American Power System Objective Purpose

Transcription

1 Interconnection-Wide Oscillation Analysis: Baselining Oscillation Modes in the North American Power System NERC Synchronized Measurement Subcommittee (SMS) Scope Document Objective The objective of the work task to be performed by the NERC Synchronized Measurement Subcommittee (SMS) is to better understand the inter-area modes in each of the interconnections (Eastern, Western, ERCOT, and Quebec). The goal is to identify the modal characteristics (mode shape, mode frequency, mode damping ratio) of the interconnected bulk power system using high-resolution, time-synchronized measurement data during major grid disturbances. Purpose Some interconnections such as the Western Interconnection have spent significant effort to understand the oscillatory modes of their respective interconnections, particularly due the small signal stability risks posed to them. However, other interconnections have not cohesively analyzed the oscillatory modes of the system using wide-area synchrophasor data from Phasor Measurement Units (PMUs) or other types of high resolution, time-synchronized Disturbance Monitoring Equipment (DME). With the proliferation of PMUs across all interconnections in North America, in conjunction with the formation of the NERC SMS, the electric utility industry is equipped with the measurements and capability to perform such an analysis to better understand the inter-area modes on the system. The purpose and goals of this task include: 1. Use synchronized measurements across the interconnection during grid disturbances or abnormalities to baseline the oscillatory performance of the interconnection. 2. Provide the electric utility industry with a better fundamental understanding of inter-area modes and forced oscillations on the bulk power system. 3. Enable better monitoring of system behavior and identify oscillatory conditions or anomalies on the system if and when they occur. 4. Use actual data measured during system events to compare the modal characteristics of the planning models used in transient stability studies (compare model vs. actual) as a component of system-wide model validation.

2 Applicability Synchronized phasor measurement data will be collected through the Regional Entities, requested from the following applicable entities: Reliability Coordinators (RC) NERC may also request data for selected events using the University of Tennessee Knoxville (UTK) GridEye/FNET system. Data Collection Options & Event Selection Criteria Wide-area time synchronized data is required for this analysis. Currently, there are two primary options for collecting Dynamic Disturbance Recording (DDR) data. The plan to use these data sources is: 1. University of Tennessee-Knoxville FNET/GridEye System: The FNET system captures timesynchronized phasor data including voltage phasor magnitude and angle and frequency for distribution-connected wall outlet devices. These devices report phasor quantities at 10 samples per second, which limits the frequency range for which oscillations can be analyzed. FNET data will be analyzed on a larger set of events to understand if data quality and resolution are sufficient for oscillation analysis. 2. Bulk Power System Phasor Measurement Units (PMUs): The proliferation of PMU coverage over the past 5-10 years has resulted in a wide-area coverage of PMUs across each of the North American interconnections. PMUs report voltage and current phasors and frequency at rates of generally samples per second. This would enable a wider range of frequencies to be considered and analyzed. PMU data will be used on a very selective basis for analyzing key events of interest meeting the requirements laid out below. Events will be judiciously 1 selected by the NERC SMS based on a technically justified set of criteria for analyzing inter-area oscillatory modes or other oscillatory anomalies on the system. The events of interest to the SMS for these purposes will include: 1. Underfrequency events where all or most of the generators within the interconnection are perturbed by the event. Generally, the following low frequency deviation thresholds will be used (although subject to change): a. Eastern Interconnection: Hz b. Western Interconnection: Hz c. ERCOT Interconnection: Hz d. Quebec Interconnection: Hz 1 The purpose of this data request is to gather data useful for performing analyses that improve the interconnection s understanding of oscillatory behavior. Therefore, not all events meeting the criteria outlined will be collected. It is expected that key events of interest meeting the criteria will be used to gather sufficient data to perform the analysis. 2

3 2. Wide-area forced oscillations 2 on the system that are impacting multiple Balancing Authority Areas. 3. Identification of poorly damped local or inter-area oscillations on the system that warrant further analysis using wide-area measurement data. Data will be requested for a given event. An event time will be provided in Universal Coordinated Time (UTC), and will likely include the local times as well for clarity (local Standard Time). In addition, the start and end time of data to be collected will be explicitly defined for clarity. The duration of data to be requested will loosely follow the guidelines set forth below: Large underfrequency events: 10 minutes pre-contingency data 15 minutes post-contingency data Forced oscillation events: Pre- forced oscillation data to capture normal grid dynamics (10 minutes) During and/or post- forced oscillation data to capture the event in question (15 minutes) Poorly damped oscillation events: 10 minutes pre-contingency data 15 minutes post-contingency data Data Collection Location Requirements All available 3 synchronized phasor measurement data should be provided that meets the following requirements: Time synchronization: The data must be time synchronized to UTC with expected accuracy within +/- 2 milliseconds. Voltage Phasors: Any one positive sequence bus voltage phasor (voltage magnitude and phase angle) at a bulk power system substation with nominal voltage greater than 230 kv. Frequency: Any one frequency signal (ideally derived from the voltage signal above) at a bulk power system substation with nominal voltage greater than 230 kv. Active and Reactive Power: Active (P) and reactive (Q) power measured at the high-side or low-side of the generator step-up transformer or single Point of Interconnection (POI) for each generating resource that meets either of the following criteria: 2 Forced oscillations are caused by an external input or driving force to the interconnected power system that results in abnormal or forced conditions. Examples include malfunctioning steam valve cycling, arc furnace dynamics, and other harmonics. Forced oscillations are typically undamped and persist until the malfunctioning device or condition is removed from the system. 3 There are no location requirements for this data request (i.e., data must be provided at identified locations). The location requirements only apply to currently installed synchrophasor monitoring locations to provide a wide-are view of the system. 3

4 Gross individual nameplate rating greater than or equal to 100 MVA. Gross individual nameplate rating greater than or equal to 300 MVA where the gross plant/facility aggregate nameplate rating is 1000 MVA. Aggregate nameplate rating greater than or equal to 100 MVA connected at a single Point of Interconnection (POI) to the bulk power system (i.e., variable energy facilities such as wind farms and solar PV plants). Data Collection Formatting Requirements For all submitted data, it should incorporate the following formatting and data quality requirements: Bus voltage magnitude should be reported in units of kv (i.e. 535 kv) Bus voltage phase angles should be reported in units of degrees (bounded by +/- 180 degrees), and should be the raw phase angle reported by the PMU (not a referenced phase angle difference) Frequency should be reported in units of Hertz (i.e Hz), with the highest resolution available (at least three trailing decimal places). Data should be submitted in COMTRADE (C37.111), revision C or later, or submitted in.csv format. Attachment A provides the file structure for a.csv file submission. C allows 4 different data types ASCII, 16-bit signed integer, 32-bit signed integer, and single precision floating point (32-bit). Single precision floating point is the preferred data format. C allows the.dat file to be split into multiple sections e.g. D000 through D999. It is advisable to keep the individual file sizes below a maximum of 50 MB. To take full advantage of information available in the synchrophasor data stream it is advisable to use the PSRC H8 Application of COMTRADE for Synchrophasor Data Schema for Phasor Data Using the COMTRADE File Standard5. Bad data or data drop out: At a minimum, any one of the C Status Word bits as non-zero is considered bad data. Any individual data point with bad or missing data should be reported as NAN in the submitted.csv file, formatted according to Attachment A. Data Request and Submission Requirements Data requests and submitted data will be completed within the following timeframes: Data requests for a particular event will be sent within 45 calendar days of the actual event. Requested data will be submitted within 45 calendar days of the data request. 4 IEEE. C IEEE Standard Common Format for Transient Data Exchange (COMTRADE) for Power Systems, Available:

5 Analytical Techniques Pre- and post-contingency PMU data will be used to estimate the modal properties of the dominant interarea modes in the system using ambient modal estimation methods. This includes an estimation of the mode frequency, mode damping ratio, mode energy, and mode shape for each of the dominant inter-area modes that can be observed in the PMU data. The analysis will provide insight on the presence and nature of the dominant system modes, and how they evolve over time. PMU data during system events will be used to analyze the dominant modal properties using ringdown analysis methods. Again, the modal properties of each mode including the mode frequency, mode damping ratio, mode energy, and mode shape will be estimated. These results will be compared with the respective estimations from ambient methods above for cross-validation. Depending on the nature and location of the disturbance, the ringdown analysis will provide information on which modes get excited by different contingencies. PMU data collected for the periods prior to and during forced oscillation events can be used to determine the source location and nature of forced oscillations. The conclusions will be reported to the utility providing the data for validation and verification. The names of specific generation facilities as well as participating utilities will be kept confidential and all the analysis results will be reported in an anonymous fashion. 5

6 Analysis Strategy & Process Phase 1: Characterization of Modal Characteristics Phase 1 is the primary focus of this work task of the SMS. The following process will be followed for accomplishing this task: 1. Events meeting the criteria defined herein will be considered by the SMS. 2. The SMS will select events for data collection as they occur and judiciously execute data requests to gather PMU data for those respective events. 3. Data will be provided by Reliability Coordinators (RC) and University of Tennessee-Knoxville (UTK) 6 within the appropriate timeframes as defined herein. 4. All submitted data will be kept confidential. 5. Tools and expertise will be utilized by those willing and able to participate. This will likely include engaging with commercial vendors and subject matter experts. Non-disclosure agreements (NDAs) between each party supporting the special assessment and NERC will be put in place to ensure data is kept confidential and secure. 6. Collected data will be used for determining the modal characteristics of the interconnection. 7. Through this process, tools and analyses will be compared with each other to identify 1) modal characteristics and other attributes, and 2) considerations for improvement for the tools used. 8. Results will be explored by the SMS members as a whole; actual data will only be provided to the sub-team of analysis members who have signed appropriate NDAs. 9. Results will be published in a manner that only provides modal characteristic behavior of the bulk power system. Actual data and entities will be kept completely anonymous. The data will provide characteristic behavior only. 10. If anomalous data or other data of interest (i.e., forced oscillation events) NERC will work with the effected entities to explore identification of those oscillations for the purposes of improving reliability. Phase 2: Integrated Utilization and Further Development of Modal Analysis Tools Phase 2 is the expansion of this work, once modal characteristics have been identified, and is not the primary focus of this work task. Phase 2 is a longer-term vision in which this type of analysis can become a regular business practice for NERC and other operating entities. This includes: Comparison of tool performance, after initial utilization, to identify areas for improved results, usability, visualization, automation, etc. 6 UTK is not subject to comply with the 45 submittal. However, NERC will collaborate with UTK to get necessary data for select events in a timely manner. 6

7 Automate the identification of system modal characteristics using tools selected. Automatically detect anomalous oscillations (forced and inter-area) such that preemptive situational awareness can be achieved. Deliverables The following deliverables will be provided throughout the task: Analytical results on modal analysis using ambient data, ringdown analysis, and/or forced oscillation analysis; Description(s) of modal analysis techniques used to analyze interconnection-wide or regional oscillations; Characterization of inter-area modes in each interconnection analyzed; and A periodic report culminating in a description of the modal characteristics of the interconnection(s). 7

8 Attachment A -.CSV File Submission Formatting The.csv file will follow the Guideline for Data Format Used in Engineering Analysis Applications of Disturbance and Simulated Data 7. This simplified.csv (comma separated value) file structure effectively accommodates the extraction and compilation of synchrophasor data for disturbance analysis. Headers Lines The simplified header structure consists of four (4) header rows, as shown in Figure 1. The simplified requirements for these header rows are described here: Line 1: Signal Names Signal names should provide enough detail to discern the name of the substation or location where the measurement is located, as most PMU naming structures currently do. Signal names can contain any characters except commas. Special characters other than underscore ( _ ) should be avoided. Line 2: Signal Type Acceptable signal types include: Signal Type Voltage Magnitude Phase Angle Frequency Active Power Reactive Power Line 3: Units Acceptable units include: Signal Time Voltage Magnitude Phase Angle Frequency Active Power Reactive Power Line 4: Description Acceptable TYPE VM VA F P Q Acceptable UNITS SEC (Seconds) KV (Kilovolts) DEG (Degrees) HZ (Hertz) MW (Megawatts) MVAR (Megavars) Description of the signal can contain any alphanumerical values (except commas). The Description should clearly explain the signal and location of measurement. 7 WECC Joint Synchronized Information Subcommittee (JSIS). Guideline for Data Format Used in Engineering Analysis Applications of Disturbance and Simulated Data. August Salt Lake City, UT. Available: 8

9 The headers lines should be formatted to the following requirements: Cells A1:A3 will not change. Cell A4 should include the event start time, in UTC. See included.csv file data submission template. An example.csv file with dummy data is shown below. Figure 1: Sample.csv file format (opened in Microsoft Excel for clarity). Data Lines The data lines should be formatted to the following requirements: Times will monotonically increase, with no missing times (i.e., don t skip time stamps ). Times will be equally spaced, with 5 trailing decimal places (e.g., , , etc.). Frequency measurements should have at least 3 trailing decimal places; 5 trailing decimal places is preferable. 9