160 likes | 365 Views
Composite Load Model Implementation Update. Craig Quist, PacifiCorp January 23-25, 2013 TSS Meeting. Composite Load Model – Status Update….
E N D
Composite Load ModelImplementation Update Craig Quist, PacifiCorp January 23-25, 2013 TSS Meeting
Composite Load Model – Status Update… • Following the last TSS meeting (August 2012) a request was submitted to members to reevaluate the composite load model using an updated load model representation. • Responses from 11 entities were received by 1/15/2013 (due 11/15/2012) • Responses received fit into seven performance categories and two software/ data reevaluation categories, as noted in the table below: • Composite Load Model - status update meeting w/M&VWG held on 1-8-13 • A summary of responses received is noted on the following pages:
Composite Load Model – Status Update… • Public Service New Mexico (PNM): • No-disturbance simulation and two fault disturbances were evaluated. • A worst case analysis did not show stability criteria violations for the two disturbances with either the interim load model for the CMPLDW model. • The revised CMPLDW model data appears to remain in steady-state during no-disturbance simulation, unlike previous versions of the model. • System voltage and frequency response is similar between the interim load model and the revised CMPLDW model, with the exception of the long delay (14 seconds) in recovery of system voltage to a steady-state value.
Composite Load Model – Status Update… • San Diego Gas and Electric (SDGE): • We did not upload anything because we did not find any appreciable differences in our studies. • Snohomish PUD No.1 • The SNPD composite load model test results show that SNPD bus voltages and frequencies are within the limits of the WECC System Performance Criteria for loss of a 230/115kV transformer at a point of delivery from BPA to SNPD and a 500kV line Monroe – Custer or Custer – Ingledow. The test findings are summarized below: • The composite load models have more buses with large voltage dips during a fault. • The composite load models have slow voltage recovery. • Bus frequencies of the composite load models are less stable than those of the current 20% induction motor models and are slightly higher than initial frequencies.
Composite Load Model – Status Update… • Southern California Edison (SCE): • Although the results between the pre-CMPLDW and post-CMPLDW are not the same, no criteria violations were identified. • It should be noted that with the post-CMPLDW during the fault, the voltage dip tends to be lower than the pre-CMPLDW results and tends to settle at a higher frequency. • The studies identified that the post-transient voltage for the CMPLDW scenario settles at a higher voltage than the initial conditions. Typical voltage levels were about 1% to 8% higher. (This could be due to load be shed by the model.) • Perform further evaluations for more base case conditions is recommended.
Composite Load Model – Status Update… • Pacific Gas & Electric (PG&E): • Composite load model could increase transient voltage dip and transient frequency dip at some load and non-load buses. • Composite load model could cause increase violations of the existing WECC system performance criteria. Transient Frequency and Voltage Violations: The study results show that loss of the Round Mt – Table Mt #1 and #2 500kV lines could cause transient frequency dip below 59 Hz for more than 6 cycles at the following load buses in Area 40: Note: Klamath Falls generator data modeling issue!!
Composite Load Model – Status Update… • Pacific Gas & Electric (cont.) • Additionally, study results show that loss of two SONG generation units (G-2) could cause transient voltage dip more than 30% at the following buses in Area 22 and 24: • While PG&E recommends approval of the Phase 1 model, it is recommended that the composite load model for those buses identified above (Appendix A) should be re-evaluated.
Composite Load Model – Status Update… • PacifiCorp (PAC): • Only PSS/E user analyzing composite load model • Extensive analysis was performed: • August 8, 2008, St. George, Utah – disturbance Initial analysis performed, actual disturbance data limited • July 28, 2009, Mid Valley 138 kV – disturbance Post-transient voltages were higher than actual disturbance voltage trace → implying over tripping of loads in composite load model.) • Bridger West/Path C – northbound OTC Composite load model – phase 1, did not impact path ratings. • Path C – Southbound OTC Composite load model (1) terminated abruptly in PSS/E Version 32 or (2) simulations noted numeric instability for PSS/E Version 33. Further evaluation is underway.
Composite Load Model – Status Update… Note: Potential numeric instability!!
Composite Load Model – Status Update… • Basin Electric Power (BEPC):(Matt Stoltz): We took a look at this but the data is all in GE format. Therefore we don’t have the ability to use it. If there is PTI data available, let us know. • Public Service of Colorado (PSC): (Sirisha Tanneeru): Is it mandatory for every utility to verify the models? We perform studies using PSSE, although we build cases using PSLF. Since the composite load model is not accommodated in PSSE yet, we will not be able to review the study results. Please let me know if that is ok. • (WECC – Enoch Davies): There is no PSS/E data available for the whole case. PacifiCorp did their own conversion and created a version of the case in PTI that had the composite load model for PacifiCorp’s area. I am sure that Craig was closer to that than I was so, I will leave it to him to comment further. • (WECC – Enoch Davies): We certainly understand that there is not a version of the case in PSS/E yet. We cannot expect you to be performing testing on the case if you don’t have the case available. I hope that we will have an equivalent case in PSS/E soon that people in your situation can perform some testing but it is not available yet. • Availability of PSLF to PSS/E data conversion is a concern…
Shifting Gears… Transition from: • Composite Load Model discussion to: • PSLF-PSS/E Software Implementation Issue Discussion
PSLF, PSS/E, PW Software Implementation Issues PTI PSS/E S/W Revisions • Version 32.1.1 [9-16-11] • Composite Load model released (issues with array size) [Wind V5] • PacifiCorp performed first level of testing (reduced composite load model db: AB, BC, NW, CAL, NV) • Version 32.2.0 [9-27-12] • Modified to accommodate 5000 composite load models [Wind V6] • PAC – (GSN-FCK) ) disturbance terminated • Version 33.3 [9-25-12] • PAC – (GSN-FCK) disturbance, voltage instability • PSLF (18.0.1) to PSS/E (33.3) power flow conversion issues resolved relative to SVD models (type 5, 6 & 7) • PAC has reinitiated the PSLF to PSS/E dynamic data conversion process • WECC Staff currently trained on PSS/E Version 32 PF and Version 30 TS data conversion GE PSLF S/W Revisions • Version 18.0_00 [8-4-11] • Three new SVD models (type 5, 6 and 7), composite load “long identifiers” were expanded from 12 to 32 characters, & updated to read PSS/E Version 32 Raw data. (Note: Does not write composite load model data to *.dyd file for PSS/E use. [4-23-12]) • Version 18.0_01 [10-24-11] • PSS/E Raw import and export was updated, and some dynamic model issues were addressed. • Version 18.0_01b [11-15-11] • Corrections to the PSS/E raw data were applied. • SVD model data was not in WECC base cases until early 2012. • Version 18.1_00 and 18.1_01 • Have both been released, but WECC Staff has not been using either version for base case production. • Version 18.1 will include updated composite load model data interface
PSLF, PSS/E, PW Software Implementation Issues • How to address the PSLF-PSS/E-PW software implementation issues? • Pause software advancement until all packages are ready to move ahead? • This is probably a non-starter due to continual development. • Recommended PSLF-PSS/E-PW software coordination procedure: • Advance notification (2-3 months) ahead of PSLF release • Notification would include details of all database and model changes • Off-ramp/halt release advancement if errors are encountered • Other software related issues • Software conversion training for the WECC Staff • When will PSS/E software to generate composite load models be available?
Where do we go from here? • Reevaluation of specific issues identified, including numeric instability… • Address WECC/NERC Planning Standard/ Criteria implications? • Resolve the PSLF-PSS/E-PW software implement-ation issues
Composite Load Model – Target Date… • Based on remaining efforts, including software implementation issues, what is the current target for moving ahead with the composite load model (Phase 1) implementation? • August 2013 – TSS Meeting