Peak online TSAT Implementation Update Presented by Hongming Zhang, Manager EMS Network Apps

Agenda Transient Study Criteria Definition Peak TSAT Use Cases TSAT Visualization for Operation Awareness TSAT RAS Enhancement TSAT Model Accuracy & Solution Verification Proofs of the TSAT Simulation Results Findings from Recent Islanding Events Conclusion

Transient Study Criteria Definition Peak collaborated with CAISO et al entities to write a white paper about R-T transient study criteria: Adopt Transient Stability Criteria (new Peak SOL) Frequency criteria is set to 59.5 Hz for 0.1s at various PMU telemetered buses Unit ride through capability PRC-024 Minimal 3% damping ratio requirement

Peak TSAT Use Cases-Backup RTCA Initial TSAT use case is to monitor a subset of RTCA contingencies Normally those contingencies are unable to solve due to missing of dynamic or transient RAS modeling components. Such as Out of step relays Frequency rate of change Time delay protective actions

Peak TSAT Use Cases-Transfer Analysis TSAT can calculate power transfer limit like RT-VSA tool, but including dynamic models Power flow vs. time domain simulation Industrial motor load modeling Generic unit protection relays and load shedding Accurate RAS interaction Peak models COI, SDGE and SDGE/CEN Import IROL scenarios in TSAT to collect data for dynamic limit assessment

TSAT Use Cases- Frequency Response R-T Freq. A-C Measure in 7 day

Other TSAT Use Cases of Interest Dynamic RAS gen drop amount evaluation Composite load modeling impact study NWExport Study (to calculate the total export that can be achieved from BPA to the south and east Dynamic transfer limit calculation for Path 3 under major outages or islanding conditions Online Study for Loss of Solar Resources during Disturbances due to Inverter Settings

Visualization for Operation Awareness PI UI for TSAT Transfer Analysis TSAT Results Visualization in PI Processbook RTCA unsolved CTGs

TSAT RAS Enhancement: Requirements TSAT covers RTCA for certain contingencies associated with dynamic/transient RAS (e.g. Colstrip ATR and MATL OOS) Real time arming is needed for the RAS that Peak does not have the arming logic/look up table (for example: BC hydro RAS) For RAS that arming logic/look up table is known, discrepancy may exist in the real time arming and the look up results (for example: PDCI RAS)

TSAT RAS Enhancement: Implementation Export and transfer the EMS RAS Arming table e.g. SPMEAS every 5 min from EMS to TSAT server TSAT server applies the SPMEAS to the RAS file and solve the new case Cases are archived in TSAT output server Users can modify RAS Arming data via EMS study displays or TSAT UI as needed for offline study purpose

Proof of TSAT Model & Solution Accuracy Peak TSAT model and simulation results were validated against new system events e.g. PDCI loss, Colstrip tripping, and recent system islanding cases. The TSAT RAS model solution quality was verified and improved continuously Simulation w wrong RAS Gen Drop TSAT Simulated Frequency in the event PMU Frequency in a DC Event Simulation with correct RAS Gen Drop

Peak online TSAT vs RT-VSA Results

New Proof: Event on 10/xx/2017 Custer-Ingledow 500 kV #1 line was under planned outage. Around15:45 MST, Custer- Ingledow 500 kV #2 line tripped. Following the line trip, BC Hydro RAS C5L51_52 RAS opened Nelway-Boundary line and dropped MICA G05 as well as Arrow Lake G01. In addition, MATL local RAS tripped the MATL line. After these actions, AESO and BC Hydro system separated from the WECC system.

Real Time TSAT Results on 10/xx Last TSAT run before the event was from 15:41. TSAT gave secure result for the Custer-Ingledow CTG (MUC5L005 and BCT5L043 were essentially same). TSAT RAS model dropped Nelway-Boundary Line, MICA G05, and MATL line. Note at 15:41, Arrow Lake G01 was still not armed.

TSAT Prediction Results on 10/xx TSAT predicted separation of BC Hydro and that system was safe

RTCA Results on 10/xx RTCA results before the event showed unsolved for Custer – Ingledow contingency. MATL RAS is not able to be modelled accurately in RTCA and was not triggered on 10/16. This contributed to the unsolved results.

Findings from recent Islanding Events During Sept-xx Path 1,83 outages, the TSAT case becomes more sensitive to armed gen drop amount TSAT reports many “Insecure” cases for one week when the outages caused weak link between AESO and the rest system, and islanded AESO eventually More gen drop was armed after the outages. It’s verified by offline TSAT study that those “Insecure” cases solves fine with additional 100 MW gen drop Peak is working with relevant entity to review TSTA study results and evaluate if there’s any real issue

Conclusion In the last two years, Peak collaborated with the entities to validate V&R Peak ROSE and CAISO Bigwood RT-VSA tools, and end up achieved highly consistent VSA limits by two real-time tools Now Peak is well prepared to roll out online TSAT in Production for ROE daily watching and validation Peak look for productive collaboration with the entities as we did for RT-VSA tool validation, to make both Peak and CAISO’s online TSAT tools operational in Control rooms in next 12-18 months

Peak TSAT Production Rollout Plan and Internal Training for ROEs Presented by Stephanie Conn, Sr. DTS Engineer

Rollout Preparation Milestones Milestones (Peak Deployment) Finalize software (RAS arming, RAS models add PI displays) and TSAT model changes (Completed) ROE Training Delivery (10/19/2017) (Completed) Begin production use by ROEs (10/30/2017) ROE Validation Completion (5/31/2018) Coordination and sharing of results with TOPs (7/31/2018) Develop operating procedures with TOPs (11/30/2018) RCSO Training (Fall 2018 Training Cycle) Operational Use by RCSOs (2019)

ROE TSAT Training at Peak Initial training will be given to Real-time Operating Engineers to support model validation functions Initial system monitoring Data collection before and during FTL events to aid with model validation Later training will be given to provide a complete picture of the tool, its use and operating procedures Generator and RAS dynamic modeling Validation and mitigation of insecure results Tool troubleshooting TSAT training for RCSOs on tool and procedures prior to implementation for real-time use Implementation will follow a phased in approach: Initially, basic tasks to aid in model validation and familiarization with the tool As experience increases, the material gets more complex and will provide a complete picture of the tool and its use

ROE TSAT Training at Peak (cont’d) Initial training will cover Power system stability concepts Voltage stability Angle stability Frequency response characteristics Differences between Power flow and time-domain simulations TSAT modes of operation Basic study validation information Customization of Basecase & Study parameters Voltage stability PV Curves QV Curves Load characteristics Non-motor load Induction motor load Hot and cold load pickup Long term, classical and transient instability/fast voltage collapse Effect of AVR on stability Acceptable voltage criteria Angle stability Power transfer equation Factors affecting torque angle Equal area criteria

ROE TSAT Training at Peak (cont’d) Initial training will cover Building contingencies to mimic events Running contingencies and comparing to PMU data Storing data for further analysis Demo creating TSAT cases from STNET (EMS offline study tool) Voltage stability PV Curves QV Curves Load characteristics Non-motor load Induction motor load Hot and cold load pickup Long term, classical and transient instability/fast voltage collapse Effect of AVR on stability Acceptable voltage criteria Angle stability Power transfer equation Factors affecting torque angle Equal area criteria

Stephanie Conn, sconn@peakrc.com Hongming Zhang, hzhang@peakrc.com Stephanie Conn, sconn@peakrc.com