Reactive Power and the Blackout
- Nov 4, 2003 12:00 pm GMT
- 2913 views
Voltage drops related to reactive power caused blackouts on the Pacific Coast in 1996 and in France in 1978. PJM itself came close to a blackout due to reactive power problems in 1999, avoided it, and took corrective steps. Yet, by having rigorous regional monitoring of reactive power and rules for its operation and compensation, PJM is unusual within the electric industry. Even though reactive power is essential to electricity reliability, many areas dont actively manage it - under-reacting to a long-known problem.
For example, no restructured transmission operator with real authority like PJM exists in the area where the blackout started, and no regional rules governing reactive power or its monitoring existed there on August 14 or today. If inadequate reactive power was a main element of the blackout, the balkanized agglomeration of individual utility control areas, the lack of any restructured regional coordination, and the absence of mandatory regional rules explain the existence of reactive power problems.
What is reactive power?
Generation and transmission cannot do anything without reactive power, a poorly understood but essential part of electricity. Often electricity is thought of simply as the electrons that flow through the wires. But there are a few component parts with familiar names - volts, amps and watts - that are not well understood.
Electric current, measured in amperes or amps, is the stream of electrons. The current cant go anywhere without voltage - the force that pushes the current through the wires. To make an analogy to water, the current would be the drops of water, and the voltage would be the water pressure that pushes it through the pipes.
Working together, amps and volts become the watts, or megawatts (MW) in larger volume, that define how much work the electricity can do. But both amps and volts travel in waves and only produce watts when the waves are in phase with each other.
Reactive power, measured in voltage ampere reactive (VAR), is when the amps and volts are not in phase and is, to some extent, an unavoidable by-product of producing megawatts. If these additional out-of-phase waves travel down the transmission line, they decrease the amount of megawatts (the in-phase waves) that can travel through the line.
The VARs wave can either be leading ahead of the in-phase wave or lagging behind it. If there is a VARs wave leading, then you want to produce some equivalent VARs waves that are lagging such that the leading and lagging VARs waves cancel each other out and leave only the in-phase wave. In this way the maximum amount of megawatts can travel through the line and voltage is maintained. If theres not enough voltage, the current cannot be pushed through the power lines, leading to a voltage collapse. Power plants and transmission lines are designed to shut down when there is inadequate voltage to prevent damage to the equipment.
How reactive power may have contributed to the blackout
Problems with reactive power did occur on August 14. The likely, though not conclusive, scenario is that several hours before the blackout, FirstEnergy noticed low voltage on its system, a sign of insufficient reactive power, and increased VAR production at nine of its plants. The Eastlake plant, northeast of Cleveland, experienced problems, and operators made several adjustments, but low voltage caused the plant to shut down, creating an even bigger deficit of reactive power.
About two hours before the blackout, a brush fire in southwest Ohio knocked out a power line, redirecting power onto the rest of the system, changing and/or increasing the need for reactive power on other lines. Then about one hour before the blackout, power lines between Cleveland and southern Ohio failed. And a few minutes before the blackout, all links between northern Ohio, known for some time as a weak spot for reactive power, and southern Ohio shut down.
How reactive power is managed
In the past, a single monopoly utility owned and controlled all of the generation and transmission lines in its own control area and had to provide reactive power just as it had to provide sufficient generation and voltage. Local monopoly control areas were part of a regional reliability council, but these councils were not responsible for dispatching generation or operating the grid.
In northern Ohio, in the East Central Area Reliability Council (ECAR), thats still the story. According to Tom Kraynak, Manager of Operations and Resources, ECAR has no guidelines on the provision of reactive power, leaving it up to the individual utilities.
However, the individual control areas east of the Rockies, north of Texas, and north into Canada, operate as part of an interconnected transmission grid. When electricity is transmitted between control areas, which happens all of the time, there has to be communication to properly operate the system, including adjustments to reactive power. The Midwest Independent System Operator (MISO) has not yet defined any system rules concerning reactive power, also leaving it up to the individual control areas.
To put it bluntly, many places do not really manage reactive power. This isnt news to the electric industry. Yet its an abdication of responsibility that a system to provide reliable reactive power has not been implemented everywhere.
Reactive power in PJM
PJM, on the other hand, does actively manage reactive power, and not just within its own system.
Twice in July 1999, PJM experienced an extraordinary and unexpected peak load, reaching 51,600 MW, far above the forecasted peak load of 47,570 MW. On these two days, PJM saw voltages dip to dangerous levels, particularly in northern New Jersey.
PJM implemented emergency procedures to successfully avert major problems, but this severe stress on the PJM system revealed a number of unexpected vulnerabilities that PJM staff and stakeholders viewed as a wake-up call. PJM made some immediate procedural changes, but also undertook what Bob Hinkel, PJM General Manager of Strategic Integration, describes as a painful root-cause analysis. PJM reviewed all operating procedures and produced a 61-page report identifying 25 primary root causes of the low voltage problem and 20 recommendations for improvement.
One of the key conclusions was that low voltage on these two days occurred because reactive demand exceeded reactive supply. PJM found that the actual capability of its generation units to produce reactive power was much less than their reported capability 54 generation units in eastern PJM had actual maximum output of only 70 percent to 72 percent of what PJM thought they could produce. The unexpected shortfall in reactive power capability, along with some offline capacitor facilities, contributed to a near voltage collapse in eastern PJM.
PJM also concluded that it had good emergency plans and planning processes to deal with a shortage of megawatts, but inadequate plans for dealing with a shortage of reactive power. In many cases, generation owners within PJM had no contractual obligation to provide reactive power. Some generators also have to produce fewer megawatts in order to produce more reactive power. At the time, however, generation owners were paid for megawatts but not for reactive power, thus provided with no incentive to produce additional reactive power.
PJM adopted key changes to fix these and the other problems identified, and continues to assess and improve reactive power operating procedures and rules.
- PJM accelerated the implementation of a new Energy Management System that it had been testing, providing PJM with more accurate real-time information concerning voltage levels and the need for reactive power in different parts of the system.
- PJM adopted limits on the volume of megawatts it will allow to flow from one point to another in the system without local generation being present to produce VAR. Just like pumping to keep up the pressure in water pipes over a distance, the voltage that pushes current through the transmission wires and the reactive power that supports it must be produced and delivered at more points in the system than the current.
- Generators now receive lost opportunity payments to compensate for any lost energy revenue because they must provide additional reactive power.
- While the lost opportunity payments remove the financial disincentive to produce VAR, since November 2002, PJM also has been including specific VAR obligations and penalties for non-compliance in each new interconnection service agreement with a generator.
- On August 13, the day before the blackout, PJM announced plans to deploy an enhanced Voltage Stability Application, a software system designed to further increase PJMs ability to monitor the need for reactive power.
Restructuring, reactive power, and the blackout
Reactive power problems alone cant explain what caused the blackout. Whether managed solely by an individual utility control area or by a larger regional transmission organization, effective system management requires sufficient monitoring and adjustments to keep the power system stable. If intervention is too slow or not done properly, voltage on power lines can collapse suddenly, and the lights go out. The task is more difficult when there are multiple control areas that need to communicate and coordinate management. The task is simpler and more effective when a single RTO has all of the necessary information and rules, as well as the authority, to get the job done.
Reactive power and restructuring
Some have suggested that restructuring makes the problems of dealing with reactive power more difficult, highlighting two issues: merchant (non-utility) generation and related financial incentives; and transmitting power over longer distances with multiple transactions.
PJM has thousands of megawatts of merchant generation and the most sophisticated competitive restructured market in the country (perhaps the world), yet doesnt experience reactive power deficiencies. Reactive power must be produced and paid for, whether the generator is owned by the former monopolist or is a merchant generator. All it takes is a system of rules that provides incentives and/or requires sufficient reactive power.
It also makes no sense to suggest that a sharp rise in long distance power transactions is the problem. It is true that reactive power doesnt travel very far and that VAR must be produced locally to some extent, and at times more of it must be produced. Power can travel over dozens of paths at once, in complex, varying patterns. However, load is load, and the system must be managed to serve the load.
If a system isnt producing enough reactive power to maintain sufficient voltage to serve load, its because the system doesnt have the information infrastructure or the necessary rules and communications to do so effectively. The remnants of balkanized, smaller control areas make it difficult - a fully restructured RTO is best able to effectively maintain sufficient reactive power.
The blackout proves the point. Reactive power problems in non-RTO control areas were the problem. On August 14, PJM experienced conditions worse than those experienced back in 1999, yet the PJM system held together well with minimal loss of load. The lesson to be learned is to focus on the need for MISO to get up and running, with rules that require sufficient reactive power, pay for it to be produced, and implement the sophisticated information and communication systems necessary to operate the system reliably.