DRAFT

ERCOT Dynamic Voltage Recovery Criteria

For, NERC Category A Contingencies

NERC Category B Contingencies

NERC Category C Contingencies

NERCCategory D Contingencies that a transmission planning entityhas determined should meet NERC Category C requirements. (See note 1)

  1. Do not allow a voltage collapse
  2. For systems with undervoltage load shedding (UVLS) schemes, allow no UVLS load to be shed.
  3. Generator terminal voltages must recover to 90% of nominal voltage within 10 seconds after the terminal voltage falls below 90% of nominal voltage.

For, NERC Category D Contingenciesthat a transmission planning entity has determined to be of credible concern. (See note 1)

  1. For systems without an UVLSscheme, do not allow a voltage collapse
  2. For systems with UVLS schemes, allow no more than 1250 MW of UVLS load to be shed

or 300 MW (DOE reportable load shed)

or % of system load

or MW size of largest unit in the affected system

or % of load on UVLS

other

  1. Generator terminal voltages must recover to 90% of nominal voltage within 10 seconds after the terminal voltage falls below 90% of nominal voltage.

1Category D contingencies considered under this standard should be those that have been demonstrated to have a relatively high probability of occurrence relative to other category D contingencies. This determination should be supported by statistically significant outage history data for the subject transmission system.

Steady-stateReactive Adequacy

Design for sufficient steady-state resources is the industry accepted first step in providing adequate reactive support. Steady-state adequacy can be evaluated using either PV analysis or QV analysis. Voltage collapse is caused by reactive demand exceeding reactive supply. There are essentially two mechanisms that result in the steady-statereactive demand not being met:

  1. The reactive demand exceeds the reactive capability of the reactive sources.
  2. The network impedance limits the amount of reactive that can be delivered to the load to a value less than the demand.

Voltage collapse is initially a local phenomenon. Voltage collapse can occur at one bus or a limited number of buses and can cascade to large areas if not arrested. It is possible and desirable to identify these groups of buses that act as a coherent bus group during voltage collapse. Similarly, the sources that can fully deliver reactive power to a coherent bus group will not be very distant (roughly less than 50 miles) from the coherent bus group. To properly analyze the first voltage collapse mechanism mentioned above, it is necessary to identify the dynamic reactive sources that can deliver reactive power to each coherent bus group.

The second mechanism describes the situation commonly observed where reactive reserves of nearby sources become exhausted and reactive reserves on remaining sources must be delivered across the impedance of the supply system. Reactive reserves on the remaining sources cannot be fully delivered before collapse occurs. As the reactive demand increases, the voltage drop across the supply impedance increases and eventually reaches the collapse point in the area of high demand.

Both of these mechanisms come into play in most power grids. They are addressed in slightly different ways by the following steady-state criteria.

ERCOT Steady-statePV Criteria

This method directly addresses the combined collapse mechanism which has been cited as the typical voltage collapse scenario for most power systems.

From ERCOT Operating Guides Section 5.1.4:

Voltage stability margin shall be sufficient to maintain post-transient voltage stability within a defined importing (Load) area under the following study conditions:

  • Peak Load conditions, with import to the area increased by five percent (5%) of the forecasted area Load, and NERC Category A or B operating conditions (see NERC Table I in ERCOT Planning Criteria); and
  • Peak Load conditions, with import to the area increased by two and one half percent (2.5%) of the forecasted area Load, and NERC Category C operating conditions

ERCOT Steady-state QV guidelines

For a given coherent bus group the reactive margin calculated using QV should be at least equal to the greater of:

  • The reactive capability of the largest plant or dynamic reactive source contained in the coherent bus group. The reactive capability of the largest plant is defined as the sum of the Qmax of all on line dynamic reactive sources (generators, STATCOM’s, etc.) at a station connected to the transmission system at the same voltage; or
  • The increased reactive demand following any Category A, B, or C contingency. Category D contingencies or other contingencies having a probability of occurrence comparable to a Category C contingency, may be treated as a Category C contingency for voltage stability/reactive margin purposes

How are coherent bus groups determined? One method is to perform modal analysis on the transmission system. Another method is to first note all on-line dynamic reactive sources that are not at their Qmax. Then, for all load buses of interest, perform QV analysis. For each bus, note the dynamic reactive sources that have reached their Qmax at the QV knee point. Coherent buses are all the buses that had the same dynamic reactive sources reach their Qmax at the knee point. (Put another way, the knee point represents the largest stable reactive load for that bus. The dynamic reactive sources that have reached their Qmax when the QV analysis reached the knee point are the dynamic reactive sources that can supply reactive power to that bus load. Buses that are supplied reactively by the same dynamic reactive sources will experience voltage collapse together because their reactive supply (the same dynamic reactive sources) cannot supply any more reactive power.)

The second bullet of the ERCOT Steady-state QV guidelines can be calculated by summing the pre-contingency Qmax and Qgen of the dynamic reactive sources supplying reactive power to the coherent bus group. The difference between the Qmax sum and Qgen sum is the pre-contingency reactive capability. Apply the contingency and solve the load flow case. If the case does not solve, the pre-contingency reactive capability is inadequate. If it does solve, sum the dynamic reactive source Qmax and Qgen and take the difference. This is the post-contingency reactive capability. The difference between the pre-contingency and post-contingency capability is the needed reactive margin for this contingency.

Dynamic Reactive Adequacy

The following discusses methods for determining critical contingencies, reactive margins of dynamic resources, and/or the need for additional dynamic reactive sources. Adherence to the ERCOT Dynamic Voltage Recovery Criteria can be evaluated by the following three methods:

Method 1

For a given coherent bus group the reactive margin calculated using QV should be at least equal to:

  • Six times the steady-state motor reactive requirement for buses whose voltage drops below 70%or 80% or 60%of nominal during a fault. The reactive capability of generators supplying reactive power to the coherent bus group shall be X (X to be determined) times the generator’s Qmax value.

Unless more specific information is available, consider 75% of the load flow reactive load to be motor load in summer cases, and 20% motor load in all other seasons. Industrial loads should be considered 75% motor load regardless of season.

Note: it is recognized that the above involves numerous simplifying assumptions about the distribution network and load. The above is considered a reasonable compromise between network detail and computational efficiency.

The following discussion gives some of the rational for the guidelines.

Dynamic simulations can be very time consuming, when and a large number of contingencies must be evaluated. The above guidelines are intended to be applied to load flow cases. (Determining[MEC1] buses that drop below 70%or 80% or 60%will require dynamic simulation.) Once the critical contingencies are determined, it is recommended that detailed dynamic simulations be performed to verify the margin determined above is adequate.

For any contingency, it is possible to calculate the steady-state reactive needs of coherent bus groups. Dynamic reactive margin means maintaining sufficient pre-contingency dynamic reactive resources so that should a contingency occur, sufficient reactive power can be delivered to the load (coherent bus group) to meet the short term dynamic reactive demand and allow transition back to acceptable steady-state conditions.

The six times motor reactive in bullet three approximates motor inrush reactive demand and the x% voltage recognizes that motor inrush will not occur at higher voltages. The multiplier increasing the generator reactive capability recognizes the fact that generator exciters have a short time overload capability to help meet the motor inrush demand.

How are coherent bus groups determined? One method is to perform modal analysis on the transmission system. Another method is to first note all on-line dynamic reactive sources that are not at their Qmax. Then, for all load buses of interest, perform Q-V analysis. For each bus, note the dynamic reactive sources that have reached their Qmax at the Q-V knee point. Coherent buses are all the buses that had the same dynamic reactive sources reach their Qmax at the knee point. (Put another way, the knee point represents the largest stable reactive load for that bus. The dynamic reactive sources that have reached their Qmax when the Q-V analysis reached the knee point are the dynamic reactive sources that can supply reactive power to that bus load. Buses that are supplied reactively by the same dynamic reactive sources will experience voltage collapse together because their reactive supply (the same dynamic reactive sources) cannot supply any more reactive power.

Method 2

Some industry experts maintain that most results of long-term simulation for voltage and reactive studies can be accomplished manually using a series of quasi-dynamic studies. This class of analysis can be performed with a standard power flow program.

The first step in the analysis is to model generators as controlling low side terminal voltage. Regular power flow studies typically model generators as controlling the transmission voltage when in reality the automatic voltage regulators almost always use the generator terminal voltage as the setpoint. Regulation of the transmission voltage is typically achieved by occasional manual adjustment of the terminal voltage setpoint by the generator operator. Quasi-dynamic studies assume that no manual adjustment takes place during the short-term post-contingency period (0 to 3 minutes). This can be modeled in the power flow by adding explicit elements that represents the generator main power transformer if they are not already included. The base case is solved with the generators regulating the transmission voltage to the desired level and the generator terminal voltages are recorded. The generator data is then modified to have the generators regulate their terminal voltage to the level that was present in the base solution. This now becomes the new base case for contingency analysis.

The second step is to lock taps to model short time response of system. The automatic tap changers will not have time to fully respond in the short term post contingency state. For conservative results in quasi-dynamic studies the taps should be locked at their base case positions in the power flow contingency solution. Contingency analysis and PV analysis are then performed with these modeling assumptions.

This method is used to differentiate the need and amount of dynamic reactive resources from static reactive sources. However, it is still recommended that detailed dynamic simulation be performed to verify the results.

Method 3

If the transmission planner has developed the dynamic models needed for a dynamic voltage analysis, then they can screen the transmission system using dynamic simulation to determine the critical contingencies resulting in slow transient voltage recovery. Generally this screening can be applied to a set of contingencies that is known to be of interest due to steady-state results or actual experience, rather than be applied to an exhaustive list of contingencies. Appendix 1 contains a report compiled by the ERCOT Dynamics Working Group in 2003 which gives some basic parameters to test the system. Note that the document discusses ERCOT Transient Voltage Security Criteria. Just to avoid confusion, the TVS (Transient Voltage Stability) criterion was adopted and the TVD (Transient Voltage Dip) criterion was not adopted for ERCOT. A number of consultants screen for critical contingencies using the same basic method. Essentially, the transmission planner applies a fault at a bus for some length of time and trips either one or two elements and then after several cycles or seconds of simulation determines if a bus or a number of buses are below a particular voltage.A full dynamic voltage analysiswould then be performed on the critical contingencies identified by the screening process.

Additional Discussion

Applying the guidelines discussed in Method 1 or 2 would require a minimal amount of dynamic simulation. Meeting the reactive margins as discussed would meet the dynamic reactive adequacy requirement, but as it was stated, detailed dynamic simulations are recommended to verify the dynamic reactive margins are adequate. Likewise, if these methods indicate that no reactive margin exists or is in fact in deficit for a critical contingency, dynamic simulation is required for determining the amount and type of dynamic reactive resources required to ensure the transmission system meets the ERCOT Dynamic Voltage Recovery criteria.

When performing full dynamic simulations, it is recommended that the transmission planner include a dynamic load model that takes into account the effect of small motor (such as residential air conditioners) and large motor characteristics on voltage recovery.

1

This document contains proprietary information and shall not be reproduced in whole or in part without prior written permission of ERCOT

Appendix 1


ERCOT Transient Voltage Security criteria DEVELOPMENT (part i)

prEpared for ercot REliability Operations SUBCOMMITTEE

BY ERCOT DYNAMICS WORKING GROUP

September 9, 2003

Table of Contents:

Acknowledgments

Transient Voltage Security Subgroup

Disclaimer

1. Executive Summary and Recommendations

2. Scope of Work and Voltage Stability Problem

2.1 Scope of Work

2.2 Introduction

2.3 Transient Voltage Security

3. Development of Transient Voltage Dip Acceptability Criteria

3.1 Absolute Voltage Magnitude vs. Percent of the Initial Value

3.2 ERCOT TVD Criteria

4. Development of Transient Voltage Stability Criteria

4.1 Factors Involved in TVS and Load Modeling requirements

4.2 Analytical Techniques and Dynamic Simulation

4.3 Margin to Voltage Instability

4.4 ERCOT TVS criteria

5. Conclusions and Further Investigations

6. Bibliography

Appendix A. ERCOT Transient Voltage Security Criteria

Appendix B. A Sample System Transient Voltage Stability Analysis

Acknowledgments

The ERCOT DWG members thank Mike Connolly of Center Point Energy, and Biju Mathew, and James Armke of Austin Energy for their support, reviews and suggestions.

Transient Voltage Security Subgroup:

From 2003 ERCOT Dynamics Working Group

Tom Bao /
Lower Colorado River Authority
Vance Beauregard / American Electric Power
Roy Boyer / Oncor
Jose Conto / ERCOT System Planning
Reza Ebrahimian* / Austin Energy
John Moore / South Texas Electric Cooperative
Yan Ou / ERCOT Operations
Juan Santos / ERCOT System Planning
Wesley Woitt / Center Point Energy

* 2003 Chair of DWG

Disclaimer

The Electric Reliability Council of Texas (ERCOT) Dynamics Working Group prepared this document. Conclusions reached in this report are a “snapshot in time” that can change with the addition (or elimination) of plans for new generation, transmission facilities, equipment, or loads.

ERCOT AND ITS CONTRIBUTING MEMBER COMPANIES DISCLAIM ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING ANY WARRANTY OF MERCHANTABILITY OR FITNESS FOR ANY PARTICULAR PURPOSE WHATSOEVER WITH RESPECT TO THE INFORMATION BEING PROVIDED IN THIS REPORT.

The use of this information in any manner constitutes an agreement to hold harmless and indemnify ERCOT, its Member Companies, employees and/or representatives from all claims of any damages. In no event shall ERCOT, its Member Companies, employees and/or representatives be liable for actual, indirect, special or consequential damages in connection with the use of this data. Users are advised to verify the accuracy of this information with the original source of the data.

ERCOT is the corporation that administers the state's power grid. ERCOT serves approximately 85 percent of the state's electric load and oversees the operation of approximately 70,000 megawatts of generation and over 37,000 miles of transmission lines. Its members include retail consumers, investor and municipally owned electric utilities, rural electric co-ops, river authorities, independent generators, power marketers, and retail electric providers.

ERCOT is one of ten regional reliability councils in North America operating under the reliability and safety standards set by the North American Electric Reliability Council (NERC). As a NERC member, ERCOT's primary responsibility is to facilitate reliable power grid operations in the ERCOT region by working with the area's electric utility industry organizations. The Public Utility Council of Texas (PUCT) has primary jurisdictional authority over ERCOT to ensure the adequacy and reliability of electricity across the state's main interconnected power grid. An independent Board of Directors comprised of electric utility Market Participants governs ERCOT.

1

This document contains proprietary information and shall not be reproduced in whole or in part without prior written permission of ERCOT