Recent conversations with professionals in electric utility control centers reveal a sense of helplessness when it comes to blackouts. Almost universally, control center managers project the feeling that their utility could be next. Utilities need a decision tool to help determine when to alert operators when conditions are sufficiently serious to warrant special attention. This concern is legitimate. In a traditional control room, notification is provided to the operators when:

• A protective device operates in the field.

• A generator is tripped.

• A transmission line reaches its rated power transfer capability.

• A bus voltage reaches a high or low limit.

In addition, advanced applications are available that can be used to estimate values at unmonitored locations or to validate existing monitored values (state estimator).

Utilities have the capability to predict the transmission system's response to changes in load, generation or topology (operator power flow). They also can evaluate the effects of transmission and/or generation contingencies. Although these tools are sophisticated and considered adequate for operating today's power systems, a key step in the analysis process is missing: the ability to assess the risk of blackout caused by grid instability.

In fact, NERC's Policy 9 requires reliability coordinators to compute the “stability limits” for the current and next-day operations to foresee whether the transmission loading will progress beyond the operating reliability limit. Is this being done? The far-reaching Aug. 14, 2003, blackout event suggests this may not be the case — perhaps because detecting thermal and voltage violations is straightforward and can be executed on-line, whereas performing real-time stability assessment is a difficult proposition. Operating a power system without knowing its actual stability limit is like walking across thin ice. Because conventional stability methods cannot be used in real-time, a new way to define and solve the problem must be identified.

Understanding and Assessing Instability

Can a transmission service provider measure the risk of instability? Yes, but we need to define a “metric” for assessing the risk of blackout. Such a metric has been identified, and it's the distance to steady-state instability, also referred to as steady-state stability reserve. Can utilities analyze instability quickly enough to support on-line decision-making? Yes. In fact, tools that perform fast stability calculations have been around for quite some time.

Before proceeding, it is important to distinguish between transient stability, voltage stability and steady-state stability. Transient stability, according to NERC, is “the ability of an electric system to maintain synchronism between its parts when subjected to a disturbance of specified severity and to regain a state of equilibrium following that disturbance.” The “disturbance” is usually a short circuit. The transient stability limit (TSL) is the value, in MW, below which the system would be able to regain equilibrium after the occurrence of any power system disturbance, regardless of size or location. To determine the TSL, the analysis should evaluate all disturbances for a succession of increased MW levels and reduced voltages until at least one disturbance has caused transient instability.

Voltage stability can be considered a special case of steady-state stability. The sensitivities of system load to voltage will influence voltage stability and may cause a collapse of voltage in a load area caused by a small increase in power transfer. Voltage stability procedures are capable of finding the point of voltage collapse at individual buses by making certain assumptions about the nature of the load — but the process needs to be repeated to evaluate as many buses as possible or, at least, a minimum set of buses known a priori to be critical.

Steady-state stability aims at computing the Steady-State Stability Limit (SSSL), which is the amount of MW (internal generation plus imports) such that, for any loading smaller than SSSL, the system is stable in the sense of small signal stability.

Small signal stability refers to an occurrence of growing oscillations of system parameters during high power-transfer levels. These oscillations, typically in the 0.2- to 2-Hz range, can lead to loss of synchronism, or to line tripping and cascading outages caused by large power swings.

It has been theoretically proven that voltage stability (which is predominantly load stability) and steady state (or angle stability) are connected. The SSSL and voltage stability limit (VSL) are given by the same mathematical condition and depict the maximum loadability state.

Simply stated, below certain MW levels, the system is stable both during normal operation and in the presence of a disturbance, regardless how large. At higher load levels, certain large disturbances might cause instability — this is called transient instability. (Post-mortem analysis indicated that the 1965 Eastern United States blackout was caused by transient instability). At even higher loadings, the system might still be operating, but a large disturbance is not required for a problem to occur. The smallest load change or voltage reduction may result in instability. This is called steady-state instability or voltage collapse, and is apparently what caused the Aug. 14, 2003, blackout.

Transient Impact and Security Margin

The TSL is an elusive target. It does exist, although for practical purposes, it cannot be computed exactly. However, analysis and practical experience suggests SSSL and TSL are interrelated. They change in the same direction: If SSSL is high, TSL is also high, and vice versa. For a given set of relay settings, TSL depends on the same factors that affect SSSL, including topology and voltage levels. It is not known whether a mathematical formula relating TSL and SSSL has been or can be found, but the TSL/SSSL ratio can be approximated empirically. In other words, it is possible to determine a safe MW system loading, referred to as security margin, such that for any state with a steady-state stability reserve smaller than this value, no contingency — no matter how severe — would cause transient instability. The security margin is expressed as a percentage of the SSSL.

Therefore, an important enhancement to the control room tool kit is supplementing — with a computer program that's fast enough to be used both in real-time and for quick off-line simulations — the state estimation, operator power flow and contingency evaluation, which work both in real-time and study mode. It also is supplementing the transient and voltage stability, which are suitable only for off-line use. Such a solution must also be able to:

• Calculate the maximum transfer limit (the system MW just before SSSL and the distance to it, which is the steady-state stability reserve, or simply “stability reserve”).

• Determine the system MW for a user-defined security margin (for example, 20% below the critical state).

• Rank generators and tie-lines in order of their impact on stability.

• Calculate MW schedules that may improve stability, thus providing the ability to develop remedial action strategies.

We now have software programs, including QuickStab Professional, which are available from Energy Concepts International Corp. (ECI; New York, New York, U.S.) for use off-line and from the Web, and from Areva, for use in real-time and study-mode on Areva's e-terra SCADA/EMS platform. These are examples of applications that enable us to assess load and stability limits. Since the wheeling of power places greater demands on the power system, utilities need to tap into enhanced analysis tools if they are to safely operate the transmission network near its load limit.

Acknowledgments

The author would like to thank Dr. S.C. Savulescu of New York City (scs@ecieci.com) for his guidance and for making available a pre-publication copy of “Evaluation of the Stability Reserve of Transelectrica's Transmission System by Using QuickStab Professional,” a paper that was presented at the National Conference of Energy CNE 2004, June 13-17, Neptun, Romania.”

John Tweedy is a utility systems project manager for Power System Engineering, a consulting firm in Madison, Wisconsin. His 30-year career in the electric utility industry has involved experience with system planning, system operations, and energy management systems at investor owned utilities and electric cooperatives, as well as consulting experience in the area of bulk power, networks and automation. Tweedy has a BS degree in electrical engineering, an MS degree in systems engineering, an AAS degree in industrial machinery automation, and an AAS degree in visual basic programming. He is a registered professional engineer in multiple states.
tweedyj@powersystem.org

Getting a Grip on Security Margin

To illustrate the security margin concept, let's say that off-line studies indicate that for a given topology the system is stable for any disturbance when the total load is smaller than 800 MW, but it becomes unstable for at least one disturbance when the loading is larger than 800 MW. Let us also assume that for the same base case, the SSSL was computed to be 1000 MW. In this case, we would say that the “security margin” is ((1000-800)/1000)*100 = 20%. How do we use the “security margin” concept? Let's say that, for the same network running at 800 MW but having different operating voltages, we computed an SSSL in the amount of 1050 MW. Therefore, the security margin is 1050*0.8 = 840 MW. This says that transient instability does not happen below 840 MW but it may happen above. If we have artificially assumed that the security margin was, for example, 10% corresponding to a 945-MW loading, then we would have been induced to believe that 900 MW would be a safe operating level. In fact, at such a loading level, there will already be a risk of transient instability.