(A serialized, expanded and updated version of this article can be
found at the Pterra Blog site:
Conventional wisdom says that the more motors connected to a feeder,
the faster voltage will collapse when there is a reactive deficiency.
This is true to the extent that voltages do drop faster, but the voltage
may not fall all the way — so a voltage collapse does not occur.
A different, and perhaps more problematic state is reached, when the
feeder is in equilibrium point at a low per unit voltage. This is
operation on the Voltage Ledge.
Stairway to the Voltage Ledge
preliminary steps to the Ledge are consistently recorded in major
voltage instability events, and has the following progression:
- High system demand conditions associated with a summer heat wave
require large amounts of reactive power being imported into a load
- Over time, the transient reactive supply provided by nearby or
local large synchronous generators is withdrawn by the same
generators (by operator action or through over-excitation limiters).
- More and more reactive power is then drawn from farther away
from the load pocket, increasing reactive transmission losses and
resulting in a gradual drop in transmission voltages.
- Step-down transformers between the transmission and distribution
level voltages adjust to the drop in transmission voltage by
automatically changing taps. Thus distribution loads see
“normal” voltage even as the transmission system voltage is
dropping. As a consequence the demand remains constant and the
load pocket’s reactive deficiency continues.
- As transmission voltages continue to drop, the step-down
transformers reach their tap limits. This would occur at about
0.9 p.u. voltage on the high side of the step-down transformers. The high
side may be at a subtransmission voltage level between 34.5 to 138
kV. From higher voltages, this would appear as higher per unit
values, up to 1.0 per unit at 345 kV and higher.
- When the step-downs reach their tap limits, the loads on the
distribution network begin to see a dropping voltage. As noted
at the start of this article, when there are more motors on a
feeder, the voltage drop is much faster. Motors, primarily
induction type air conditioning units, draw increasing reactive
power as voltage drops.
- As voltage drops, some motors may stall, drawing even more
reactive power. But some motors may tripout, which would tend
to give a slight increase in voltage. In addition,
non-motor feeder load would decrease reactive demand at voltages
below nominal resulting in an increase in feeder voltage.
- So we have a state where there are factors seeking to drop
voltage and others seeking to raise voltage.
- In addition, if voltage recovers, some motors may restart,
causing a drop in voltage. If the condition stays for several
minutes, some loads start to show a self-restoring quality — these
return their power demand to near normal levels even though terminal
voltage is below nominal.
- The combined effect of these various influences results in an
equilibrium state, characterized by low voltage, relatively steady
net demand and high line currents.
This is life on the Ledge. Loads, such as computers and
electronic devices, are exposed to a continuous shift in voltage leading
to insulation and other failures. Transformers, bushings and
insulation are exposed to high current that is peaky and filled with
harmonics. The voltage has held steady at the transmission level,
making operators unwary of the goings on in the distribution system.
Some actions intended to alleviate this state and their consequences
- Grid operators may succeed in starting up emergency generators
or obtain vars from a remote generator coming online. There
additional vars that reach the distribution system, but if
the amoutn is not
sufficient to relieve the Ledge state, it may just bring about a new
equilibrium state, another voltage Ledge.
- Operators may call for a voltage reduction which is a change in
the setpoint of the step-downs to a lower voltage. The
objective is to reduce demand. When the system is in the
voltage Ledge, this does not produce an operating change, since the
tap changers are already at their limit. If the voltage
reduction order is made before the feeders reach a voltage Ledge,
this might in fact hasten the entry into a voltage Ledge.
What actions help:
- If the system can hold on until the heat wave diminishes, the
lower demand will gradually ease the system out of a voltage Ledge.
- If a large load is shed, this would also allow some reprieve.
The system may remain on the Voltage Ledge for an extended period, up
to 2-3 hours as recorded in the 1999 East Coast event. If a
contingency occurs, such a line outage or generator trip, this significantly reduces reactive supply, the system
may enter into collapse. This time in earnest.
The following are voltage traces from previous voltage collapse and
near voltage collapse events.
- From July, 1999, near voltage collapse event in the PJM coastal
regions, a trace of 500 kV voltages going below 1.0 per unit for 2-3
From August, 2003 for Northeast Blackout event, a trace
of 500 kV voltages in PJM
From Tokyo Voltage Collapse event of August, 1987, trace
of EHV voltages
of the U.S. Department of Energy’s Power Outage Study Team, Findings
and Recommendations to Enhance Reliability from the Summer of 1999,”
Washington, DC, March 2000.
August 14, 2003 Final Report,” New York Independent System
Operator, February, 2005.
Report on the August 14, 2003 Blackout in the United States and
Canada: Causes and Recommendations,” U.S.-Canada Power System
Outage Task Force, April 2004.
Conference on August 14 Blackout,” North American Electric
Reliability Council, presentation by Shinicho Imai, Tokyo Electric
Power, January, 2004.
For questions, comments and further discussion, contact us at