the 2003 blackout - pearson...

The 2003 blackout

In many developing countries, electricity may fail catastrophically in specific areas

because demand exceeds supply (supply may be unavailable or demand too great) or

because of trouble with the distribution system (these troubles may even be due to

sabotage in some cases). A way to deal with known supply-demand discrepancies is

institution of so-called “rolling blackouts,” which are scheduled outages, again to allow

supply and demand to get in sync. Some rolling blackouts were experienced in California

in the early 2000s, when deregulation, market manipulation, and increased demand led to

energy shortages. These occurred as small areas were cut off from the state grid to reduce

demand to what the generators could supply to prevent crashing the entire system (see

especially the extension Extension 8.4, Gouging? and the section on the California crisis

below for more detail). These rolling blackouts affected the small areas for a relatively

short time, and were deliberate, as occurs in developing countries. The entire imbroglio at

first somewhat embarrassed Californians, who considered the comparison unflattering to

California; later revelation of the deliberate exploitation of the crisis—even instigation of

the crisis—by greedy energy companies angered them and lessened the embarrassment.

Even so, the California governor who presided over the crisis was recalled by the voters in

large part because of the crisis and the pain it inflicted.

In August, 2003, a series of missteps led to a catastrophic blackout in a large portion of

the eastern and midwestern United States. That it occurred on a very hot day meant

maximum discomfort for many who lived through the experience, and was in part

responsible for the problem itself.

Energy, Ch. 4, extension 5 The 2003 blackout 2

Background

By fifty years ago, electricity was available in all but the most remote American regions.

The cities had been the first to embrace electricity—initially the largest, then smaller cities

as well. During the Depression, government programs brought rural electrification to

isolated farmsteads where private investment had failed.

The American continent was electrified, but in electric utility “islands.” Cleveland got its

electricity from the Cleveland Electric Illuminating Company, Columbus from Columbus-

Southern Ohio Electric Company, Pittsburgh from Duquesne Power, Newark, New

Jersey from the Public Service Electric and Gas Company, New York from Consolidated

Edison (known as Con Ed), and so forth. Most utilities were self-contained monopolies.

There was some long-distance transmission, especially in the West from the massive

Depression-era dam-building to the growing Western cities, and some adjacent utilities

were interconnected. The electrical distribution system was still basically local, and

engineers for the utilities could draw on seventy years of experience in keeping local

service uninterrupted. There were occasionally localized blackouts from transformer

malfunction or a truck hitting a telephone pole, but no blackouts over large geographical

areas.

Gradually, interconnections were extended. By the nineteen-sixties, there were many

more high-tension transmission lines, and so many more interconnections. The utilities

were still monopolies, but even in the East, electricity could be sent long distances. New

York City, for example, might be partly supplied by electricity originating in Canada.

Connecticut was connected to New York, to Rhode Island and Massachusetts, to

Vermont and New Hampshire.


Interconnections are good, because they bring stability to the system. Like Insull’s

example of the apartment house mentioned in Ch. 7, an interconnected system can reroute

energy where it is needed, protecting against problems when a generator is lost. At this

time, most of North America part way into Mexico is connected together in a gigantic

system. The American part of the system is broken into four regions, and within each of

the regions are subregions; the subregions are the domain of the Independent System

Operators (ISOs). These ISOs were created following an earlier widespread blackout (the

1965 blackout) to help assure reliability and make sure routing was done most efficiently.

However, the larger system also brings complexity. It’s similar to the difference between

an America’s Cup racer and an oil tanker—the racer can turn very quickly, the tanker will

have gone some kilometers before it has been able to turn appreciably. The added

complexity makes it difficult to respond to catastrophic failures, especially in the times

before computers were widely used. In sum, the basic advantage of the interconnected

system in normal operation can become a debility when generating capacity is not

sufficient for demand over a large area.

The 1965 blackout

This was the background for the first widespread blackout, the Northeast blackout of

November 9, 1965. New York City had had large but localized blackouts before, in 1936,

1957, 1959, and in 1961.(2) There had even been a smaller-scale blackout in the Midwest

in 1962 that affected parts of eight states, but its effects were much less alarming than

those of the 1965 wakeup call.

But the 1965 blackout was different because it was huge—it extended from north of the


Canadian border in Ontario south as far as New York City and as far east as western New

Hampshire and Cape Cod—affecting four entire states, parts of four other states, and

parts of the province of Ontario for various lengths of time. About 30 million people

were affected. Figure E04.5.1 shows the blackout situation.

Fig. E04.5.1 Blackout area, 1965.(Exhibit 1-A, Federal Power Commission, Report to the president by the Federal Power Commission onthe power failure in the northeastern United States and the province of Ontario on November 9-10, 1965(Washington, DC: U.S. Government Printing Office, 9 December 1965))

The event started just north of the Ontario-New York border, near Niagara Falls at

Ontario Hydro’s Beck Power Station.(3) Beck sent five 230 kV high-tension lines toward

Toronto. Two more high-tension lines connected southward to the United States. At

17:16:11, a relay on one of the Toronto-bound transmission lines failed and tripped (a

circuit breaker opened the line, removing it from service). Ontario Hydro’s Lakeview

station was down with an emergency outage, and, since it was heading into winter,

demand was high in the Toronto area and the lines were near capacity. The trip spilled


demand onto the four remaining lines, which promptly tripped themselves. The

electricity then shifted onto the southward lines.

A New York utility, PASNY, has its St. Lawrence hydroelectricity plant on the other

side of the river from Ontario Hydro’s, and the southward surge was out of phase with

PASNY’s electricity, which eventually tripped the connection. This surge caused a

transient instability that destabilized the main transmission lines of New York State.

Within seconds, the Ontario grid was separated from the New York grid. The destabilized

transmission lines caused a cascade of further trips of both transmission lines and the

generators connected to them, again within seconds: New England was cut off (140 MW

demand at the time), downstate was cut off (400 MW demand). A few generators were

connected directly to large-scale users (for example, 11 of PASNY’s 15 generators at St.

Lawrence), and trips of some generators saved electricity supplied to these others.(3)

The cutoffs had unfortunate consequences, though. The rapid tripping broke the system

into islands within a few seconds of the original incident. Within these isolated systems

there were now either deficiencies that could no longer be made up or surpluses that now

had nowhere to go. This led to further failures, and within a short span of minutes these

islands lost their supply of electricity. Some 30 million people were affected, some for as

much as 13.5 hours.(2-4)

As the diagram shows, the New York City area experienced the longest outage. There was

a reason. Con Ed relied on “Big Allis,” a huge 1 GW generator made by Allis-Chalmers for

a substantial portion of its energy. Demand at this time of year was about 4 GW at the

time the lights went out. After some frantic scrambling on the part of operators at Con Ed

to find out what was going wrong, Con Ed’s automatic shutoffs tripped late during rush

hour, causing commuter chaos. With its load gone, “Big Allis” burned out its bearings, as


did other generators. The Federal Power Commission study pointed out that Con Ed’s

total reliance on fossil-fueled steam was responsible for some of the delay experienced by

New York in restoring service.(3)

New Yorkers were pretty good-natured in living through the blackout.(2,4) Even

commuters who couldn’t get home didn’t misbehave. The New York Times couldn’t

publish in the City (though the editorial staff worked by candlelight)(4) but they arranged

to have the paper printed out of town and delivered to the city and nearby areas. I lived in

Princeton at that time, got the paper for free, and was able to read all about the blackout

in a New York Times newspaper without advertisements. A telling anecdote from Ref. 4

shows the attitude behind the experience. When Central Hudson Gas and Electric was

able to restart a generator, they chose to send it to New York to supply the military

academy at West Point instead of the Ford plant in nearby Mahwah, New Jersey.

One of the consequences of the 1965 blackout as recommended in Ref. 3 was installation

of much more capable (and distributed) computer controls. Another was the formation of

the North American Electric Reliability Council (NERC), which unifies the ISOs

responsible for the various geographic regions.(5) NERC subsequently played an

important role in development of planning and operating standards for electricity

wheeling (moving electricity through transmission lines to sell it elsewhere than the region

in which it was generated). NERC now oversees and coordinates the 320,000 km of 230

kV or higher transmission lines in North America.

The 1977 (much smaller) blackout

The New York City region experienced another large-scale blackout over a decade later in

1977, this time because of lightning. It was a hot, muggy night and demand was high,


with air conditioners running full out all over the New York metropolitan area. A lightning

strike to a transmission line tower at 20:37 caused a connection of a transmission line to

ground.(6) Automatic circuit breakers then caused Indian Point 3 to trip. Then at 20:55

another lightning strike hit two other transmission lines, causing another generator, Con

Ed’s nuclear facility at Indian Point, to shut down. Just a few minutes later, at 21:19, a

transformer tripped out and 3 minutes later LILCO shut their connection when it

exceeded its emergency setting. Five minutes later a surge isolated the system. This was a

successful experience in that the experience of 1965 stopped the blackout from spreading.

But is was less successful for New York. Voltage reductions held off failure for a while.

Then a large generator tripped, and the system died at 21:36.(6)

The 1977 blackout was geographically much more limited than in 1965. This time

essentially only the City’s population and a few others, altogether about 9 million

people, were affected. New Yorkers were not generally happy with their second blackout

experience because of rampant looting. News reports from the time refer to a “festive”

atmosphere among looters and despite the looting, the mood was mostly not menacing.

But it was much less good-natured than 1965.

Many people were looting, wandering through stores and choosing things they wanted.

Reference 2 quotes an eyewitness as saying that “looters were looting other looters,” and

the police were overwhelmed. Over 3500 people were arrested during the time of the

emergency.(7) The lack of discipline exhibited by the populace during the blackout (in

contrast to the 1965 blackout), the whining self-justification of the looters, and the

disrespect they showed for others’ property left New York City’s image badly tattered

for years thereafter.

Reference 7 lists several lessons learned by Con Ed. One lesson of this blackout is that


the utilities must be more careful to make certain of the grounding. The utilities gas

turbines should be ready immediately (delays may well have contributed to the blackout).

Delays in restoration attempts did contribute to the overall system failure. The

communications problems experienced would contribute to the problems; future utilities

have needed to be better prepared with backup communications and centralized

dispatching for communications so that no information is lost when it counted in

preventing failure.

The California Blackouts of 2000-2001

The California blackouts were part of a deregulation crisis that gripped the state in 2000

and 2001 (see Extension 8.2, Deregulation in the 21st century and Extension 8.4,

Gouging?). The electricity generators apparently used every legal (if questionable) means

for removing generating capacity to increase the asking price of a kilowatthour, as well as

some illegal means, to charge Californians exorbitantly for electricity. Part of the political

fallout of this electricity crisis and the general drop in state revenues that was common

across the United States in the period between 2000 and 2003 was the recall election on

California governor Gray Davis, which featured 135 candidates—several politicians,

several actors, and many otherwise ordinary citizens—vying to replace him if he lost the

recall. One actor, Arnold Schwarzenegger, did become the governor. News analyses long

before the election painted an accomplished middle-of-the-road California governor with a

solid record sandbagged by the energy crisis.(8)

When conditions in Las Vegas echoing those that led to California’s blackout occurred in

2002, Gary Ackerman, the head of the Western Power Trading Forum was quoted as

saying “It’s not in anybody’s stomach to [allow a blackout] ... There’s just too much

sensitivity to the headline risk. ... FERC just won’t let it happen.”(9)

Fortunately,


Ackerman was right about Las Vegas. Unfortunately, he was wrong about FERC in the

larger sphere.

However, the California crisis did highlight the problem of “Path 15,” the single

transmission line connecting northern and southern California. Even during the crisis,

plans were being made to run a parallel line.

The Federal Energy Regulatory Commission (FERC) had been trying to develop

conditions for a truly national grid. However, it would not prevent the 2003 blackout.

The 2003 East-Midwest blackout

In the time between 1965 and 2003, there were some smaller blackouts (such as the one

mentioned above in 1977 and the California rolling blackouts). However, none matched

the scope of the 1965 blackout until 14 August 2003. The blackout affected about 50

million people in the northeast, the midwest, and Canada. Figure E04.5.2 shows the

progression of the blackout from the Cleveland area. Figure E04.5.3 shows the timeline of

the blackout in great detail.

Fig. E04.5.2 Progress of the loss of grid connection, adapted from Fig. 6.30 of Ref. 10. The arrowsrepresent the overall pattern of electricity flows.


Fig. E04.5.3 Timeline of the events culminating in the blackout.

This time, it was the Cleveland Plain Dealer and the Detroit Free Press that assembled

their papers by candlelight. The Plain Dealer published its abbreviated Friday paper from

the presses at the Akron Beacon-Journal, and the Free Press printed its paper from the

presses at the Battle Creek Enquirer. The Toronto Globe and Mail had presses in many

locations, as did the New York Times, and while shorter versions were published near

home, the “foreign” editions were full-sized.


As discussed in more detail in Ch. 7, electricity is not easily stored and is used essentially

as it is generated. This implies that there is a constant dance of supply and demand

Typically, a NERC region has one more generating facility operational than absolutely

needed to supply demand (this is known as the “N - 1 criterion”). This is insurance

against a random failure of a power plant through a trip or other outage, making certain

that demand can be met without overloading the systems. If such a trip does occur, the

region is expected to be back to the N -1 criterion within half an hour. A few important

and vulnerable areas run using an N - 2 criterion.

To give you some sense of the effect of the blackout in 2003, Table E04.5.1 lists the

largest blackouts in terms of size of population affected. Note that the August, 2003

blackout was an unprecedented event both in terms of the population affected and the

loss of load. Pictures from Air Force satellites, shown in Fig. E04.5.4, indicate the scope

of the blackout.

Table E04.5.1

Largest Blackouts in Terms of Size of Population Affected, 1965-2003 (Ref. 10)

Load PeopleDate Description Loss (MW) Affected

14 August 2003 Northeast-Ontario-Midwest Blackout 61,800 50 million

9 November 1965 Northeast Blackout 20,000 30 million13 July 1977 New York City Blackout 6,000 9 million10 August 1996 West Coast Blackout 28,000 7.5 million22 December 1982 West Coast Blackout 12,350 5 million2-3 July 1996 West Coast Blackout 11,850 2 million25 June 1998 Upper Midwest Blackout 950 152,000

In the years prior to the blackout, the FERC became concerned about the possibility of

troubles in the midwest, but this did not prevent what happened. Few local experts were


overly surprised that the system finally had failed.(11) Blackouts can happen anywhere, a

fact supported by the London blackout, which occurred on 28 August 2003 (just two

weeks after the American blackout), that took out 40% of the city’s electricity.(12) It was

found to have been caused by a bad transformer.(13) Another blackout that started in the

Italian-Swiss border and spread throughout Italy occurred in September 2003.(10)

a. b.

Fig. E04.5.4 The blackout region prior to blackout (a) and subsequent to blackout (b). Source:NASA/Chris Elvidge, U.S. Air Force.

NERC has specific operating procedures for its regions. The two governments affected,

the U.S. and Canada, formed a U.S.-Canada-NERC Investigation Team (hereafter I refer

to it as the Task Force) almost immediately after the event, and asked it to determine the

causes and make recommendations to prevent any recurrence.(14) Part of the reason for

the blackout, the Task Force identified, was a failure to follow NERC recommended

procedures. The Task Force final report (Ref. 10) made four general assessments of

reasons for the blackout. They are:

Group 1: FirstEnergy [FE] and ECAR [the East Central Area ReliabilityCoordination Agreement] failed to assess and understand the inadequaciesof FE’s system, particularly with respect to voltage instability and thevulnerability of the Cleveland-Akron area, and FE did not operate itssystem with appropriate voltage criteria.


The Task Force identified 5 specific problems in Group I.

Group 2: Inadequate situational awareness at FirstEnergy. FE did notrecognize or understand the deteriorating condition of its system.

The Task Force identified 9 specific problems in Group II, including 2 NERC violations.

Group 3: FE failed to manage adequately tree growth in its transmissionrights-of-way.

The Task Force identified 2 specific problems in Group III.

Group 4: Failure of the interconnected grid’s reliability organizations toprovide effective real-time diagnostic support.

The Task Force identified 10 specific problems in Group IV, including 3 NERC

violations.

In addition, the Task Force found that both FirstEnergy and other units failed to act to

stop the developing blackout, for example, not informing the Midwest Independent

System Operator (MISO) and PJM Connection of the cascading failure, not acting after

the proximate cause, the failure of the Chamberlin-Harding 345 kV transmission line, lack

of load-shedding capability, etc. (see Table E04.5.2). The Task Force pointed out that a

major problem is that NERC is not independent of the electric power industry and it has

no authority to develop and enforce standards.(10)


Table E04.5.2

Violations of NERC Procedure identified by the Task Force Interim Report,19 November 2003

1. Following the outage of the Chamberlin-Harding 345 kV line, FirstEnergy did not takethe necessary actions to return the system to a safe operating state within 30 minutes(violation of NERC Operating Policy 2).2. FirstEnergy did not notify other systems of an impending system emergency (violationof NERC Operating Policy 5).3. FirstEnergy’s analysis tools were not used to effectively assess system conditions(violation of NERC Operating Policy 5).4. FirstEnergy operator training was inadequate for maintaining reliable conditions(violation of NERC Operating Policy 8).5. MISO did not notify other reliability coordinators of potential problems (violation ofNERC Operating Policy 9).

The Chamberlin-Harding line was the beginning. That failure came about because of the

proximity of untrimmed trees to the transmission lines; the high August temperatures

caused the wires to expand enough to touch the trees and trip the line. This trip did not

by itself cause the blackout, but started a rearrangement of the electricity flows that

sloshed unstably for almost an hour before the system finally failed. Additionally, other

transmission line-tree contact trips occurred, for the same reason, along the way to the

disaster. A very disturbing conclusion of the Task Force is that the lessons of preceding

blackouts had not been learned, and the errors that caused those were repeated in August,

2003.

The Task Force criticizes this lack of understanding of its system on the part of

FirstEnergy as the root cause of the blackout. Inadequate system awareness by

FirstEnergy was another cause. FirstEnergy was not even aware that the Chamberlin-

Harding line had failed. And, as noted, FirstEnergy neglected tree trimming along its

transmission lines. FirstEnergy has to bear the primary responsibility for the largest

blackout in American history. In fact, the Task Force decided that the larger blackout


could have been avoided had the FirstEnergy system been isolated early in the sequence.

Table E04.5.3

Final Recommendations of the U.S.-Canada Power System Outage Task Force (Ref. 10).

1. Correct the direct causes of the August 14, 2003 blackout.2. Strengthen the NERC Compliance Enforcement Program.3. Initiate control area and reliability coordinator reliability readiness audits.4. Evaluate vegetation management procedures and results.5. Establish a program to track implementation of recommendations.6. Improve operator and reliability coordinator training.7. Evaluate reactive power and voltage control practices.8. Improve system protection to slow or limit the spread of future cascading outages.9. Clarify reliability coordinator and control area functions, responsibilities, capabilitiesand authorities.10. Establish guidelines for real-time operating tools.

The U.S.-Canada-NERC Investigation Team made a total of 46 specific recommendations.

The categories of recommendations made by the Task Force are shown in Table E04.5.3.

It considered its most important recommendation was that compliance with reliability

standards be made mandatory and enforceable legislatively, that is, that there be penalties

for noncompliance.

In the aftermath of the blackout, there have been calls for reengineering the grid. The grid

is aged, and it is a stepchild of the utility industry. Investments flow into other parts of

the system, but the profits on wheeling are too small and the opposition to new high-

tension transmission lines from the people who would have to live near it are too great.

Wheeling is an important source of cost savings. The U.S. Department of Energy

estimates that consumers save $13 billion dollars per year by buying energy from distant

suppliers.(10)

The grid was built over a long period of time, and many of the parts of the grid are quite


antiquated. Restructuring the grid is needed.(15) The blackout may lead eventually to a

willingness to allow the new transmission capabilities that a twenty-first century grid

needs. This will need to be integrated with the new reliability regulations recommended

by the Task Force.(16)

Reactive power, and its part in the blackout

As discussed in Ch. 3, power is measured in watts. Electrical engineers work with circuits

that have three basically different sorts of behaviors. Resistors reduce the flow of electric

current, and use power. For example, a toaster uses electric current to make the element

inside glow and cook toast. This sort of power is often called “active power.” Active

power is the eletrotechnological term for power that is dissipated in resistances and the

power used in machines.

Other devices store electric charge or electric current; they are called capacitors and

inductors, respectively. In these devices, the current’s phase is changed. This phase

change is often described mathematically by use of complex numbers. A complex number

has a real part and an imaginary part. It is usual to write a complex number C as

C = some real number + i (some real number).

This is known as Euler notation. In electrotechnology, it is useful to characterize power

as “active,” doing work or generating thermal energy, and as “reactive,” meaning that it

differs in phase (leading or lagging the “active” power).

The “apparent power” is the square root of the sum of the squares of “active” and

“reactive” power. It is common in electrotechnology to write electric current as a complex

number in Euler notation. In this formalism, “active” power is the real part of the

(complex) power, the “reactive” power is the imaginary part of the (complex) power, and


the “apparent power” is the magnitude of the (complex) power. Thus, we write the

apparent power as

S = P + i Q.

Here, P denotes active (sometimes called “real”) power, and Q denotes reactive power.

The symbol S in electrotechnology as actually used means |P + i Q|, the magnitude of the

complex power.

Electrical engineers are convinced that power comes in different “boxes,” some watts are

different from other watts (reactive power is different from active power), while

physicists think that power is, simply, power. As a result the engineers define the unit of

reactive power (the watt) as a new unit, the var, which stands for volt-amperes-reactive.

This expresses their view of the importance of the reactive power in engineering

generation and transmission systems.

Reactive power is crucially important for AC circuits, and in this regard you will recall

that transformers work only for AC. To reach the high “tensions” (potential differences)

in transmission lines, AC had to be used. (It is now possible to increase the potential

difference in DC circuits as well, and some transmission lines are DC.(1) One suggestion

to reduce the possibility of system crashes is to interconnect regions using DC, so energy

is transferred, but not the AC fluctuations.) Indeed, some of the Task Force

recommendations focus on reactive power. Reference 10 lists one of the seven key

concepts in operating transmission lines: “Balance reactive power supply and demand to

maintain scheduled voltages.”

Any AC transmission line has a capacitance and an inductance, as well as a resistance, per

unit length. So it is impossible to banish reactive power in any case. The total reactive

power comes from these “lumped” circuit elements and the reactive power of the


generators themselves. In addition, capacitor banks are maintained at generating stations

for the express purpose of supplying reactive power as needed. The key here is the

balance needed between active and reactive power in the transmission line, which is

normally maintained. As the Task Force puts it,(17)

Reactive power is particularly important for equipment that relies onmagnetic fields for the production of induced electric currents (e.g., motors,transformers, pumps, and air conditioning.) Transmission lines both consumeand produce reactive power. At light loads they are net producers, and atheavy loads, they are heavy consumers. Reactive power consumption bythese facilities or devices tends to depress transmission voltage, while itsproduction (by generators) or injection (from storage devices such ascapacitors) tends to support voltage. Reactive power can be transmitted onlyover relatively short distances during heavy load conditions. If reactive powercannot be supplied promptly and in sufficient quantity, voltages decay, and inextreme cases a “voltage collapse” may result. ...

Load power factor is a measure of the relative magnitudes of real power andreactive power consumed by the load connected to a power system. Resistiveload, such as electric space heaters or incandescent lights, consumes only realpower and no reactive power and has a load power factor of 1.0. Inductionmotors, which are widely used in manufacturing processes, mining, and homes(e.g., air-conditioners, fan motors in forced-air furnaces, and washingmachines) consume both real power and reactive power. Their load powerfactors are typically in the range of 0.7 to 0.9 during steady-state operation.Single-phase small induction motors (e.g., household items) generally haveload power factors in the lower range.

The lower the load power factor, the more reactive power is consumed by theload. For example, a 100 MW load with a load power factor of 0.92 consumes43 Mvar of reactive power, while the same 100 MW of load with a loadpower factor of 0.88 consumes 54 Mvar of reactive power. Under depressedvoltage conditions, the induction motors used in air-conditioning units andrefrigerators, which are used more heavily on hot and humid days, draw evenmore reactive power than under normal voltage conditions.

In addition to end-user loads, transmission elements such as transformers andtransmission lines consume reactive power. Reactive power compensation isrequired at various locations in the network to support the transmission ofreal power. Reactive power is consumed within transmission lines inproportion to the square of the electric current shipped, so a 10% increase ofpower transfer will require a 21% increase in reactive power generation tosupport the power transfer.

Reactive power problems did play a role in the August, 2003 blackout, but indirectly. It

was the management of the reactive power that was the problem, as noted above. Air

conditioners use induction motors, which consume reactive power, and the week leading


up to 14 August 2003 was muggy and hot in the Midwest. This demanded that the

utilities pay attention to reactive power management, for which FirstEnergy was faulted

in the Task Force report. Its reactive power reserves were inadequate. The prolonged heat

wave was taking its toll. The Task Force examined FirstEnergy records and found that

“[a]ctual measured voltage levels at the Star bus and others on FE’s transmission system

on August 14 were below 100% starting early in the day.” (10) As the Task Force notes,

this raises problems.(17)

Reactive power does not travel far, especially under heavy load conditions,and so must be generated close to its point of consumption. This is whyurban load centers with summer peaking loads are generally moresusceptible to voltage instability than those with winter peaking loads.

The loss of the transmission lines as the afternoon progressed led to more and more

difficulty in maintenance of reactive power demand. While FirstEnergy asserted that the

load factors were typical of a warm summer’s day, the Task Force faulted FirstEnergy’s

training program and its internal standards as inadequate. It said that “The team

conducted extensive voltage stability studies..., concluding that FE’s 90% minimum

voltage level was not only far less stringent than nearby interconnected systems (most of

which set the pre-contingency minimum voltage criteria at 95%), but was not adequate for

secure system operations.” The Task Force further noted that Ohio Edison, FirstEnergy’s

predecessor company, used the common 95% criterion of its neighbors. As a result, the

reactive load reserves were gone by 4 o’clock, and the overextended system soon crashed,

dragging the rest of the region along with it.

A spate of news articles accompanied the one-year anniversary of the blackout.(18,19)

Most say that progress in training has been made, and the complex system is better able

to address the future (we all hope). FirstEnergy has trimmed trees. But the problem of

relays, key to transfers among regions, has not yet been adequately addressed.(19) These


relays sense even small power surges and shut down automatically in “cascades,” carrying

large swaths of the grid along. This is both good and bad; good because it isolates the

systems’ generators, transmission lines, and substations in small failures; bad because it

exacerbates the crash, shutting down even if the relay is far removed from the actual

problem, in large ones. Balkanization of the regions can leave some without enough

power, some with too much, both bad for stability of the complex systems.

We see that human frailties, translated into lack of attention to crucial details, played a

significant role in the blackout. This fact, the apparent inability of load management to

learn lessons from earlier blackouts, and the complexity of the system itself all make

another blackout inevitable someday. This is especially true of management of reactive

power.

Work on new transmissiuon lines have still not begun. And Congress during the

intervening year failed to pass legislation to make the regulations mandatory because of

political disagreements that have nothing to do with the blackouts. If we do really learn

the lessons of 14 August 2003, we will be better prepared to make it a localized and short

one rather than a regional or national problem.

the 2003 blackout - pearson...

Documents