Author: Alexander Apostolov, OMICRON electronics, USA
The predominant system condition addressed by IEEE C37.117 involves the use of protective relays for underfrequency shedding of connected load in the event of insufficient generation or transmission capacity within a power system. It is performed in order to minimize the risk of a further uncontrolled system separation, loss of generation, or system shutdown. If sufficient load is shed to preserve interconnections and keep generators on line, the system can be restored rapidly. If the system collapses, a prolonged outage will result. Fig 1
Underfrequency load shedding plans are based on studies of a system's dynamic performance, given the greatest probable imbalance between load and generation. Underfrequency relaying can also be utilized to sense disturbances and separate power systems by opening system ties, as well as to separate non-utility generation from the utility power system. Plans should be coordinated between interconnected power systems as well as with underfrequency isolation of generating units, tripping of shunt capacitors, and other automatic actions which occur in the system under abnormal frequency, voltage, or power flow conditions.
After an underfrequency load shedding event, frequency relays can be utilized to automatically restore or supervise the restoration of load to a power system. Sufficient time delay should be employed to assure that the power system is stable prior to initiating load restoration.
In areas where isolation of a large surplus of generation relative to connected load can be anticipated, automatic overfrequency tripping of generation may be considered to prevent excessive high frequency and resultant uncontrolled generator tripping and equipment damage.
Underfrequency load shedding and restoration philosophies are applied to restore the system frequency to an acceptable level following a major system emergency that can cause a generation deficiency and prevent a total system collapse, as well as help achieve fast restoration of all affected loads.
Underfrequency load shedding must be performed quickly to arrest power system frequency decline by decreasing power system load to match available generating capacity. Severe frequency decline can occur within seconds. Manual or operator/SCADA (supervisory control and data acquisition system) initiated underfrequency load shedding generally cannot be accomplished fast enough to prevent partial or complete system collapse. Automatic schemes, employing frequency-sensing relays, are therefore employed to shed individual loads or blocks of load at discrete underfrequency set points or at specific frequency rates of decline. These set points are predetermined based on guidelines created by power pools covering a wide geographic area.
The load shedding and load restoration schemes must be designed to work together with the protection and control schemes that trip and close the line or feeder breakers. The load restoration scheme must reset any lockouts operated by the load shedding scheme, or otherwise create a permissive condition to allow manual/SCADA-controlled breaker closing. As with load shedding, microprocessor-based relays used with a communication system can allow supervisory load restoration schemes to be modified to adjust for variations in system conditions.
There are three basic types of underfrequency relays available for application in load shedding schemes. They are electromechanical relays, solid-state (or static) relays, and digital (microprocessor) relays. The different types of relays can use different operating principles. Load shedding traditionally uses underfrequency relays designed to operate on the instantaneous value of system frequency that operate any time the frequency drops below the set point of the relay.
While the system frequency is a final result of the power deficiency, its rate of change (df/dt) is an instantaneous indicator of power deficiency and can enable incipient recognition of MW imbalance. The relay's rate of change of frequency (df/dt) measurement is an "instantaneous" one, in line with the definition of derivative of a function. Monitoring only the instantaneous value can be misleading sometimes, since the rate of change in frequency may be non-linear also. Hence some abnormal frequency monitoring relays provide an element for monitoring the average rate of change of frequency. By monitoring the frequency change trend, a more secure decision can be made during contingencies.
In order to increase the security and selectivity of the underfrequency load shedding schemes, the underfrequency load shedding element is used in a scheme and may be supervised by a voltage, current, directional power or rate-of-frequency change element. The design of a load shedding protection scheme should be both dependable and secure to prevent unnecessary outages. Effects of voltage change on frequency load shedding should also be considered. The guide also includes information gathered for regional coordinating councils and several additional entities in the U.S., France, Ireland, and Nordel, which coordinates operations in Denmark, Finland, Iceland, Norway, and Sweden.
Load shedding is usually performed over several stages with the total amount of load assigned over all shedding stages based on a credible, but worst-case scenario of maximum loss of generation. Since the actual amount of generation lost is never known, the number of stages and the load shed in each stage must be properly assigned to avoid undesirable consequences.
The effectiveness of a stage of load shedding depends on what proportion of the power deficiency it represents. Each stage of load shed should contribute to the survival of the system during a contingency and needs to be properly set.
All protection functions and control logic affecting the power system operation require verification during commissioning and normal periodic maintenance to ensure reliability. Examples and Bibliography are available at the end of the Guide.
IEEE C37.117 can be purchased from store.ihs.com