Therefore the Transmission Issues Subcommittee (TIS) provides the following definition:

Fault-Induced Delayed Voltage Recovery - a voltage condition initiated by a fault and characterized by:

  • Stalling of induction motors
  • Initial voltage recovery after the clearing of a fault to less than 90 percent of pre-contingency voltage
  • Slow voltage recovery of more than two seconds to expected post-contingency steady-state voltage levels

Although this phenomenon can occur on any voltage level, it becomes concerning when such events adversely impact the voltage recovery of the Bulk Electric System (BES). Further, if the BES voltage does not recover to 90 percent of the pre-contingency system voltage in a few seconds, the FIDVR can initiate further tripping of load and generation.

Longer periods of depressed voltage below such levels can cause damage to customer and electric system equipment.
Therefore, the TIS also provides the following additional definition for NERC:

BES Fault-Induced Delayed Voltage Recovery Event (BESFIDVRE) - Any fault-induced event on any voltage class which results in FIDVR on the BES.
FIDVR is caused by highly concentrated induction motor loads with constant torque which stall in response to low voltages associated with system faults. This results in an excessive draw of reactive power from the grid. FIDVR events become increasingly probable with the increased penetration of low-inertia air conditioner loads that lack compressor undervoltage protection.
FIDVR events can, and have, occurred following faults cleared in as few as three cycles! Both the frequency and impact of FIDVR events can be decreased, but the elimination of FIDVR events in the near term is unlikely.

Planning studies have not been able to replicate FIDVR events very accurately due to an inaccurate modeling of loads. Uncorrected, this modeling deficiency has a two-fold detrimental effect. First, it can result in studies that do not adequately identify potential FIDVR events.

Second, it can give false confidence in mitigation plans designed to prevent FIDVR events.
Fortunately, several groups of experts are actively developing better dynamic load models for aggregate induction motor load, utilizing results of extensive single-phase air conditioning performance tests, and detailing analyses of actual FIDVR events. This expertise has developed, by necessity, in response to dramatic, local FIDVR events.

The NERC Planning Committee (PC) can make an important and timely contribution to this education process. Six recurring items that can help educate utility planners and, in turn, help mitigate the risk to the BES:

  • The FIDVR phenomenon must be more universally broadcast and understood throughout the electric utility planning community
  • Dynamic load models adequate for FIDVR studies should be developed, communicated, and appropriately customized for local use by grid planners
  • Post Event Analysis has been vital in finding model deficiencies and implementing corresponding improvements and should be promoted
  • Understanding and proper planning of power system protection and control action is important in preventing FIDVR events, which are often initiated by single-phase-to-ground faults that progress to multi-phase-to-ground faults because of protection inadequacy or failure
  • The degree of urgency that should be assigned to FIDVR studies is directly related to the degree of air conditioning load penetration. Guidelines or standards should not be issued to require the same level of effort from Alaskan planners as from those in California, Texas, Arizona, or Florida
  • Unit level protection should be standardized for residential air conditioners to remove them from service for undervoltage conditions and lower FIDVR risk

What causes FIDVR?
FIDVR is caused by highly concentrated constant torque induction motor loads which stall in response to low voltages resulting from system faults. The stalled motors draw excessive reactive power from the grid and require five - six times their typical steady-state running current in this locked-rotor condition. Across many motors, this state can cause the system voltage to be significantly depressed for several seconds after the fault is cleared and lead to cascading system failure. An inability to adequately model dynamic loads has contributed to grid vulnerability due to FIDVR.
The delayed voltage recovery has an impact on both transmission and distribution protection.

Table 1 shows the Relay Loadability Rating of distance relays being decreased for voltage levels below 0.85 per unit voltage and angles greater than 30 degrees.

At the same time the increased current on distribution feeders as a result of the stalled induction motors may result in undesired operation of sensitive distribution feeder protection.

The Transmission Issues Subcommittee recommends that the Planning Committee:
1. Endorse this white paper for use in further educating the industry to the FIDVR phenomenon
2.  Support the WECC dynamic load model developed for PSLF and the SOCO dynamic load model developed for PSS/E by encouraging its general use with local customization for FIDVR studies unless an adequate dynamic load model is already in use
3.  Promote Event Analysis of all FIDVR events to ensure further data acquisition, understanding, and modeling expertise
4.  Tasks the TIS and SPCS jointly with developing a guideline for adequately addressing protection and control considerations in performance of FIDVR studies
5. Champion a local climate and load density-driven level of effort in FIDVR activities.
6.   Support the standardization of unit level solutions for FIDVR by A/C manufacturers

Industry experts assisting in the preparation of this technical reference paper:

Dmitry Kosterev

Bob Yinger
John Shaffer
Gary Bullock
Tom Gentile
Ian Grant
Lee Taylor
Bob Jones
Tom Cain
Robbie Bottoms Josh Shultz
Jim Mitsche
Jay Loock
Joe Eto
Gary Kobet
Bajarang (Baj) Agrawal

Protecting your electrical assets? today and tomorrow