Recommendations on How to Proceed with

Transient Voltage Performance Criteria

Dmitry Kosterev and John Undrill

June 20, 2013

I. SUMMARY

Members of WECC Modeling and Validation Work Group, because of their involvement in the development and use of the composite load model, have been asked to provide a technical opinion on the transient voltage dip criteria . This review covers how the evaluation of simulations can be impacted by the evolution of system modeling as typified by the new composite load model, and recognizes the actual phenomenon of parts of the system where behaviors such as Fault Induced Delayed Voltage Recovery (FIDVR) are observed.

Landscape:

-Table I of NERC TPL-001-4 Standard specifies the system performance requirements in terms of firm transmission service interruption and loss of consequential load for various categories of events

-NERC TPL-001-4 Standard requirement R5 states that each Transmission Planner and Planning Coordinator shall have criteria for acceptable System steady state voltage limits, post-Contingency voltage deviations, and transient voltage response for its System. For transient voltage response, the criteria shall at minimum specify a low voltage level and maximum length of time that transient voltages may remain below that level

-WECC RC is considering transient stability performance requirements in the operational planning horizon. It is desirable to have consistent performance requirements among planning and operational engineers

-Planning entities in WECC has been using transient voltage dip criteria as a measure of power system dynamic performance since early 1990s

-Actual events of Fault-Induced Delayed Voltage Recoveryhave been observed in parts of WECC. The phenomenon is due to stalling of single-phase compressor motors (such as in residential air-conditioners and refrigeration), that can occur during a normally cleared fault, and cannot be prevented. System performance during a FIDVR event can violate WECC transient voltage dip criterion.

-Actual events can include loss of load during a transmission fault because of end-use control and protection settings. WECC MVWG performed analysis of voltage sensitivities of various end-uses. Loss of such load must be treated as consequential to a fault.

Modeling expectations:

-Power system performance is evaluated using system models.

-Models can be tuned after the fact to reproduce an actual known disturbance event and to explain what happened during the event, as it was done for September 8 2011 Arizona – Southern California outage, or August 10 1996 outage

-The goal, however, is to predict the system performance during “planned for” events in the future.

-The power system model is expected to predict quantitatively system performance during events initiated by small disturbances of voltages (e.g. a single-phase fault, or no-fault)or frequency (e.g. a generator outage)

  • Unexpected control action is possible, e.g. La Rosita unit trip following Hassayampa – North Gila 500-k line tripping during September 8, 2011 outage

-The power system model is expected to predict the general phenomenon during the events initiated by large disturbances of voltages (e.g. a three-phase fault) or frequency (e.g. islanding)

-Modeling limitations need to be recognized when simulating large disturbances:

  • WECC has completed the development and in process of phased implementation of the composite load model. The new composite load model is a major improvement over the interim load model used in WECC since 2001. The model is capable of representing FIDVR phenomenon, but the details of the voltage recovery will vary depending on the assumptions on how much load tripped during a fault and how much of air-conditioning load stalls due to low voltages
  • NERC TPL-001-4 requires evaluation of power plant disturbance ride-through capabilities, and NERC PRC-024 provides voltage and frequency ride-through requirements for control and protection setting. There is no guarantee that a power plant will ride through a disturbance even within voltage or frequency envelope. Per NERC statistics, protection and controls of power generators and its auxiliaries can come into play during large disturbances resulting into unexpected plant trips and runbacks. It is not reasonable to expect to develop models that accurately predict such control responses for the “planned for” events.
  • In April 2007, a three-phase fault on a 500-kV line caused tripping of a wind power plant during a fault, a coal-fired generator a few seconds later, a gas-turbine plant and hydro generators
  • It is not practical to develop models that can predict whether a power plant trips during a disturbance. A judgment on whether the plant trips during a disturbance can be made based on proximity to NERC PRC-024 boundaries and by monitoring generator voltage, frequency, step-up transformer V/Hz, active and reactive power, generator field current relative to expected limits.

System Studies

-Initial studies with a composite load model show that the N-1 3-phase fault with normal clearing performs significantly worse that N-2 1-phase fault with delayed clearing. This is opposite to the existing criteria. The severity of the initiating disturbance matters

System Performance Expectation

Progress in system modeling over the last 4 years has made it clear that the present system performance criterion, which is stated as an allowable amount and duration of voltage dip, does not reasonably recognize the behavior of many parts of the power system in the wake of severe disturbances. Specifically, both satisfactory and unsatisfactory responses ‘fail’ when tested against the present criterion; this criterion thus fails to discriminate between satisfactory and unsatisfactory conditions.

We argue, in the light of the current system performance simulation capabilities, that performance should be evaluated against criteria that recognize the following:

  • High probability disturbances, such as 1-phase L-G faults, simple generator trips should be evaluated against criteria whose emphasis is continuity of supply to customers. The essence of the evaluation is that the power system must withstand these events without disadvantage to customers.
  • Low probability but severe events, such as three phase faults, should be evaluated against criteria whose emphasis is the immediate and sustained security of bulk transmission grid and generation. The essence of the evaluation is that the elements of the grid that are essential to the re-establishment of supply should take precedence over short term continuity of supply to all customers.
  • Sensitivity studies with respect to modeling uncertainty are required for events that are initiated by severe faults. Modeling uncertainty includes air-conditioner voltage stall thresholds, motor tripping, and power plant ride-through capabilities.

The nature of the criteria should be open to comprehensive review.A significant volume of impact studies is the essential next step in approaching this subject. We recommend assigning MVWG, including experts in load modeling, to produce such system impact studies for various regions with WECC.

1