Pumps & Systems, October 2007
With highly reliable electrical systems, protective relays may be called upon to operate very infrequently. However, the effects of faults and abnormal conditions can be severe and protective relay systems must be designed carefully to protect against the worst possible fault conditions.
This article briefly describes the basic goals and philosophies behind relay system design and the types of protection that are applied in water and wastewater treatment facilities. As motors for pumping applications are particularly critical to water and wastewater facility operations, the major faults and abnormal conditions that affect motors are also covered. The relay schemes discussed here are typically applied to systems with operating voltages greater than 1000-V.
Purposes of Protective Relaying
An electrical fault is the establishment of an unintentional conducting path. On a three-phase system, the unintentional path may be between two- or three-phase conductors, or between one or more phases and a metal enclosure or the earth. A fault can be established through:
- Insulation failure due to age
- Overheating
- Exposure to the elements
- A lightning strike
- Mechanical failure of equipment
- Misapplication of equipment
- Accidental forced contact between conductors, such as from a maintenance error, vehicle accidents, or animal contact
Abnormal conditions may exist with or without an actual failure, but may lead to a failure if not corrected. Abnormal conditions include:
- Overloading
- Low voltage
- High voltage
- Incorrect frequency
- Unbalanced current
- Unbalanced voltage
A complete protective relaying system consists of all the components necessary to detect faults or abnormal conditions and operate the appropriate switching devices, such as circuit breakers or automatic switches. Proper operation requires integrating a variety of electrical and electronic technologies at both high and low power levels. Major components are as follows:
- Current transformers (CTs) and voltage transformers (VTs) to reduce the voltage and currents of the electrical power system to levels suitable for relay inputs. Secondary wiring and disconnects are associated with the current and voltage transformers.
- The protective relays and associated auxiliary relays, selector switches, control circuit disconnects, indicating lamps, and control wiring to the circuit breakers. The system may also include separate relays for primary and backup protection to avoid losing protection if a relay fails.
- The circuit breakers or automatic switches that will perform the switching at power circuit voltage and current levels. These devices consist of switching mechanisms, interrupter assemblies, trip and close coils and control circuits.
- A reliable source of control power for the relays and circuit breakers. Control power must be available even if the power system is faulted or otherwise unavailable. While various control power arrangements have been used, the most reliable is a substation battery with a charger.
Basic Design Goals
Regardless of the complexity, any protective relay system design is governed by a few basic goals and philosophies. These goals include speed, selectivity and reliability.
Speed
High-speed operation of a protective relay system is necessary to limit the effects of a fault, which can include equipment damage, process upset, and hazards to personnel.
A fault typically causes an "overcurrent" condition, with the current exceeding the rating of the line conductors, switches and transformers that must carry the current. This may cause a violent arc at the fault location. The fault may also cause abnormally high or low voltage.
By definition of a fault, the equipment is damaged or destroyed, but the overcurrent, arcing and abnormal voltage may also damage equipment at other locations on the system. The risk of more widespread damage increases if the fault is allowed to persist.
Even if damage does not occur, the abnormal conditions may result in tripping other devices and upsetting the processes supplied by the system. High-speed operation of the protective relay system is essential in limiting the equipment damage and risk of a wider system disturbance due to a fault.
Beyond risk of equipment damage and process upset, the failed insulation and arcing associated with the fault presents hazards to personnel who might be in the vicinity. Electric shock hazards are well-known. Becoming more formally recognized are the hazards associated with an arc flash, including flash burns, hearing loss from the blast, vision loss from the flash, and injury due to the impact of particles expelled by the blast.
The National Electrical Code (NEC) requires equipment to be marked with an arc flash warning if it's likely to require servicing while energized. OSHA enforces requirements for arc flash hazard analysis and personal protective equipment found in NFPA 70E-2004, Standard for Electrical Safety in the Workplace. The energy received by a person exposed to an arc flash depends on the current magnitude and the time it persists. High-speed protective relaying can dramatically reduce arc flash energy and hazards to personnel.
Selectivity
Selectivity is the ability of a protective relay system to isolate the smallest portion of a system necessary to isolate a fault or abnormal condition.
Tripping a switchgear main breaker for a fault on a motor circuit supplied from a feeder breaker and downstream motor control center (MCC) is obviously a major process upset. While limiting damage and system upset demands high speed, selectivity often demands some delay to allow time for protective relays closest to the disturbance to operate.
Coordination studies attempt to determine settings for time-overcurrent devices that balance the opposing demands of speed and selectivity. Time-overcurrent coordination generally requires longer time delays for devices closer to the supply point of a system. As discussed later, differential relaying can be used to avoid excessive time delays at certain locations.
Reliability
Reliability of a protective relay system is defined in terms of two components: dependability and security. Dependability is the ability of the relaying system to always operate correctly for a fault or abnormal condition. Security is the ability of the relaying system not to operate when there is no abnormal condition present, or if the condition is temporary or should be isolated by relays in another part of the system.
As with speed and selectivity, dependability and security are opposing demands. Improving dependability tends to reduce security and vice versa. All protective relay system designs attempt to balance the competing requirements of dependability and security. Lowering the overcurrent pickup setting of a relay is a simple way to increase dependability. Adding time delay is a simple way to increase security.
Also, relay designs often include features that enhance both dependability and security. For example, restraint elements may be a security feature for a differential relay so the relay avoids tripping for faults that should be cleared by other devices. The same differential relay may also have an unrestrained overcurrent element with a high setting to maintain dependability.
A motor protective relay may have firmware routines that recognize when the motor is started and restrain tripping on the high starting current during a programmed interval. An overcurrent setting can be closer to the motor starting inrush current for better dependability while the restraint feature maintains security.
Zones of Protection
Visualizing an electrical power system as divided into zones of protection allows protective relays to be logically applied to achieve the goals of speed, selectivity and reliability. The concept is simple: each component is a separate zone requiring specific protection.
In a water or wastewater treatment facility, the protective zones might be the main transformer, medium voltage switchgear, medium voltage distribution line, unit substation transformer, low voltage switchgear, low voltage feeder, motor control center and motor. The zones are separated by switching devices that will operate to selectively isolate a faulted zone from the rest of the system.
Zones of protection are defined by the location of switching devices and the current and voltage transformers that provide inputs to the protective relays; therefore, the zones should overlap. For example, a main transformer supplies medium voltage switchgear through a main breaker. The transformer zone should include the medium voltage switchgear main breaker. The switchgear zone also should include this main breaker. Transformer relaying and switchgear relaying can both respond to faults at the switchgear main breaker.
While it may seem obvious, including the correct logic and wiring to trip all switching devices necessary to isolate a faulted zone is important. For example, if a transformer is equipped with high and low side circuit breakers, both breakers must be tripped by the transformer protective relays. Bus protective relays must trip the main, tie, and feeder breakers for the protected bus section. Auxiliary relays trip the required breakers and may provide a lockout function so that equipment is not inadvertently reenergized by operators after a trip.
Historically, individual protective relays have been designed to detect specific conditions: one relay detects overcurrent for fault protection, one relay detects undervoltage, one relay detects motor overloading, etc. and a package of individual relays provided the protection for a particular zone. This was a natural design because specific types of electromechanical elements or analog electronic circuits were required to perform the different types of measurements.
Modern protective relays are now microprocessor-based, while storing the various protection measurements and logic decisions in relay firmware. This means once the currents and voltages are converted to digital form, one microprocessor-based instrument running multiple algorithms can perform multiple protective functions and provide all the protection for a particular zone.
Types of Protection
Since the zone of protection consists of a specific type of component, the appropriate relay protection for each zone must be chosen. Here are the most commonly used types of protective relays and some ways they are designed and applied to balance dependability and security:
Overcurrent Protection
Overcurrent protection is perhaps the most basic form of protection and can be applied to any power system component.
As current transformers supply the current to be measured, if the current is above the pickup setting for the set time delay, the relay produces a trip signal. The zone protected by an overcurrent relay depends on the pickup setting. A lower pickup setting extends the zone and allows the relay to respond to faults farther away from its location. Since zones of protection should overlap, an overcurrent relay should be able to respond to faults in at least the next adjacent zone.
Time delay is used to allow the relays in the next zone to operate first and achieve selectivity. Overcurrent relays typically provide a settable time delay and an instantaneous function to trip with no intentional time delay. To maintain selectivity, the instantaneous pickup setting must be high enough so that it does not respond to faults in other zones.
In a three-phase system, separate overcurrent protection is provided for faults involving only the phase conductors and faults involving one or more phase conductors and ground. The ground current measurement can be made indirectly with current transformers connected to form the sum of the phase currents. This sum equals the ground current.
Another way to measure ground fault current on insulated cable circuits is to pass the insulated conductors through a specialized current transformer designed to accommodate the conductors in its window. With this arrangement, the summation is performed magnetically. Measurement can also be made directly with a current transformer installed at the location to which ground fault current returns in the circuit - typically a transformer or generator neutral terminal.
The advantage of including both phase and ground overcurrent protection is to create greater sensitivity. Overcurrent pickup settings for phase faults must allow for normal and emergency load current in the phase conductors. Under normal conditions, current in the neutral or ground is very small and may approach zero.
Ground overcurrent pickup settings do not have to accommodate load current and can be much lower (more sensitive) than phase overcurrent pickup settings. Ground faults typically have higher impedance in the circuit path than phase faults and the circuit may have current-limiting impedance intentionally added. The result is that for most locations on a system, ground fault current is less than phase fault current. Since most faults (80 percent or more) involve ground, sensitive ground overcurrent protection separate from phase overcurrent protection is desirable.
Many water and wastewater electrical systems are designed as radial systems, where a source at one end of the system supplies the distribution network and loads connect to the other end of the system. In radial systems, the path for fault current is readily defined. For those faults, the current does not flow through the paths between the fault and the loads, except for transient current from stored energy in rotating motors.
For improved reliability, or to increase load-carrying capacity, some systems are designed with multiple sources connected together. For example, two or more utility lines may be connected to a common substation bus or the facility may operate a methane-fueled cogeneration plant connected to the distribution system that is also supplied by a utility service. For such systems, fault current might flow through a protective zone for faults on either the source side or load side of that zone. Overcurrent relays may respond to faults in many zones of the system or even faults on the utility system.
It's typically not possible to develop settings that will provide the necessary sensitivity and selectivity for all fault locations on such a system. Directional overcurrent relays determine the location, upstream or downstream, of the fault relative to the zone that detects the fault current.
The directional measurement is made by comparing the measured current to a reference quantity, usually a voltage that does not change with fault location. Settings for a directional overcurrent relay are then made to meet sensitivity and selectivity requirements for faults in the desired direction.
If a water or wastewater facility operates generators in parallel with the utility system, directional overcurrent relays may be required at the service point to prevent the local generators from inadvertently energizing the utility system. If a fault occurs on the utility system, the local generation would continue to energize the fault, creating safety hazards for the general public and for utility crews making repairs.
In one example, the designation 52 is the IEEE Std. C37.2-1996 for a circuit breaker. The phase-time overcurrent relays are designated 51, and the ground time-overcurrent relay is designated 51N. If instantaneous phase or ground overcurrent protection were applied, the designations 50 and 50N would be added to the relay symbols.
There are three phase overcurrent relays and three CTs. The dotted line from the relays to the circuit breaker denotes the relays are wired to trip the circuit breaker on an overcurrent condition.
Voltage and Current Balance Protection
Three-phase power systems are designed and operated to achieve the most balanced voltage and current conditions possible.
In a balanced three-phase system, the voltages are of equal magnitude and their sinusoidal waveforms are displaced from each other by 120 electrical degrees, or one-third of a power cycle. The sinusoidal voltages also reach their peak values in a known repeated sequence. For example, the sequence might be phase a, phase b or phase c. With balanced voltages and balanced loads, the currents will also be balanced.
Unbalance of either voltage or current can be caused by a fault, an abnormal condition (such as unintended load imbalance or an open conductor), or reversed phase connections. Deviation of the magnitudes or phase displacement angles indicates an unbalanced condition.
Voltage and current unbalance beyond a small tolerance is particularly detrimental to rotating machinery such as motors and generators, causing damage by overheating even when an overcurrent condition does not exist. Voltage and current balance relays measure the three-phase voltages or currents and calculate the degree of unbalance. Pickup settings usually are expressed in percent of the rated voltage or current and a time delay is typically used for security to allow an unbalanced condition to be corrected by fault protection relays in other zones.
Voltage balance relays can detect unbalance between the source and the point where the relay is applied, but not between the relay location and downstream loads. Typical applications for generators and motor circuits are at utility service points, at switchgear buses to control the automatic transfer between alternate sources, and to motor control centers or switchgear buses that supply multiple motors. Current balance relays are typically used to detect current unbalance for individual circuits.
Voltage-Based Protection
Several types of protective relays measure voltage to perform specific protective functions. Overvoltage and undervoltage relays operate when voltage is higher or lower than a settable level. The settings must allow for normal voltage variation associated with changes in load and utility service voltage tolerance.
Typical applications are at utility service points, at switchgear used to connect generators to a system, and at switchgear buses to control the automatic transfer between alternate sources. Overvoltage relays also are used to detect ground faults on generators and systems designed with grounding impedance to limit ground fault current to very low values.
Overfrequency and underfrequency relays measure voltage to determine the system frequency. Deviation from the nominal frequency (60-Hz in North America) might be caused by the failure of a generator governor, but usually indicates a mismatch between the power input to system generators and the power consumed by the system loads and losses. Operation at off-nominal frequency causes fatigue and cumulative damage to generator steam turbine blades and frequency may be applied for steam turbine protection.
On a system level, off-nominal frequency may be accompanied by abnormal voltage and could indicate the system is becoming unstable. Frequency relays may be applied for system protection to avoid a total collapse by disconnecting certain loads or tripping certain circuit breakers to separate the system into islands. In this way, stability can more likely be maintained or recovery more easily managed.
Frequency relays may be applied at the utility service point for water and wastewater facilities having local generation. Often the utility requires such protection as part of the relaying package intended to prevent the local generation from energizing a fault on the utility system. For water and wastewater facilities with local generation, frequency relays may also allow the facility to separate from the utility and continue operating at least some of the processes in the event of a utility system fault and disconnection of service.
Synchronism check relays measure the magnitude and phase angle of the voltage across an open point where two sources are to be connected. These relays are not protective relays, but serve a supervisory role in controlling the operation of circuit breakers. The two sources may be utility services from separate substations. Often, one source is the utility service and the other source is a local generator.
To connect the sources, the three-phase voltages must be nearly equal, the frequencies must be nearly identical, and the phase sequence of the sources must be the same. Generator paralleling controls automatically adjust governors and exciters to achieve synchronism before paralleling the generator with a utility system. Connecting two sources that are not in synchronism may result in damage to rotating machinery because of torsional stress to the shafts, and damage to the tie circuit breaker and switchgear with associated risks to personnel operating the equipment.
Defining a zone of protection for voltage and frequency relays that may respond to faults or abnormal conditions for any location on a system is difficult. Voltage and frequency relays are usually applied as a last line of defense to separate systems at interface points. Time delays are used to allow isolation of a fault or abnormal condition by relays in a specific zone.
Directional Power Protection
Power directional relays use voltage and current inputs to measure power. Either real power (watts) or reactive power (vars) can be measured, but for most applications, real power is the desired quantity.
Since power is the product of voltage and current, there is directionality associated with the measurement. Direction is determined by the phase angle between the voltage and current waveforms and power is transmitted (or "flows") from one point to another.
For example, power flows from a source to a load, into or out of a switchgear bus and into or out of a generator. For systems with multiple sources, power can flow in more than one direction. A power directional relay operates when power flow is higher or lower than a set pickup in a specified direction. The pickup and directional settings are selected to indicate an abnormal or fault condition.
At a water or wastewater facility with local generation, loss of the prime mover power input to the generator allows power to flow from the system into the generator and the generator acts as a synchronous motor. This condition mainly creates the risk of damaging the prime mover, so a power directional relay is applied to detect power flow into a generator on loss of the prime mover.
If a wastewater facility operates a methane-fueled cogeneration plant, the service agreement with the utility company may or may not provide for the utility to buy surplus power from the facility. If there is no agreement for the utility to purchase power, a directional power relay may be required at the service point to prevent transmission of power from the cogeneration plant to the utility system. This relay may be part of the required protection at the service point to prevent the cogeneration plant from inadvertently energizing the utility system.
If two transformers supplied by separate utility feeders are connected to a common bus at a water or wastewater facility, a fault on one utility feeder will allow current and power to flow from the other transformer through the common bus and back through the first transformer to the fault. Depending on the fault type, location and transformer winding connection, the current may be nearly zero and the only power flow may be that associated with the transformer losses. The direction of the flow will still indicate an abnormal condition.
Differential Protection
Differential relays operate on one of the fundamental principles governing behavior of electrical circuits - Kirchhoff's Current Law, which states that the sum of all currents flowing into an electrical node is zero. Stated another way, the current flowing into a node must equal the current flowing out of the node.
A differential relay measures the sum of all currents connected to whatever node is being protected. If the sum is not zero, there is an unintentional path for current flow that creates an unbalance between current entering and current leaving the node. The unintentional path is a fault between phases or between one or more phases and ground.
It is perhaps easiest to visualize differential protection applied to a substation bus, since a bus exactly fits a simple definition of an electrical node: a junction with current-carrying branches connected. For relaying purposes, other power system components also meet the definition of a node.
For example, a transformer can be considered as a node with high and low side conductors carrying current to and from the node. Motor and generator windings have leads on each end that carry current into and out of the windings. Transmission lines have two or more terminals where currents flow in and out. For water or wastewater facilities, differential relaying is typically applied to buses, transformers and rotating machinery.
A basic form of differential relaying consists of multiple current transformers with their secondary terminals connected in parallel. The parallel connection sums the currents from the individual current transformers. An overcurrent relay is connected to the CT secondary junction to measure the differential current.
The differential connection provides significant advantages over simple overcurrent relaying in terms of selectivity, sensitivity, and speed. Differential current transformer connections define the zone of protection. With a properly designed system, differential relays will not detect normal load current nor will they operate for faults outside the defined zone.
Since a differential relay does not respond to load current, the pickup can be very low to provide sensitive protection. Since the protection is inherently selective, the relay can therefore be allowed to operate with no time delay.
Such ideal characteristics depend on perfect current transformers, all perfectly matched. Practical current transformers have limits beyond which the measured current will not accurately represent the actual current. Current transformer errors will result in a false differential current in the relay.
For example, a fault might occur just outside the protected zone, but one or more current transformers may be unable to accurately reproduce the primary current in the secondary circuit. The sum of all the secondary currents will not equal zero and the differential relay may trip incorrectly, resulting in a loss of security and selectivity.
Differential relays and relay schemes are designed to accommodate current transformer error without loss of reliability. Several types of differential relay systems have been developed to maintain reliability when current transformer performance is not perfect and each type has its own operating principles and application requirements. Two of the most common systems that are used at water or wastewater facilities are the multi-restraint system and the high impedance system.
Multi-restraint differential relaying is based on the principle that current transformer error is likely to increase as the primary current increases. A high CT secondary current is more likely to be inaccurate than a low current. Currents from individual circuits connected to a protected zone are used to restrain operation of the relay. Higher individual currents restrain relay operation so that a higher differential current is required to produce a trip signal.
To obtain this characteristic, differential relays calculate a differential quantity and a restraint quantity.
Id = | I1 + I2 + ...|
Ir = | I1 | + | I2 | + ...
Where Id = differential current
Ir = restraint current
I1, I2, ... = individual measured current for each circuit connected to the differential zone
For some relays, the restraint quantity is calculated as Ir = Maximum of | I1 |, | I2 |, ...
The symbol |*| indicates the magnitude of a phasor quantity that has both magnitude and phase angle. The differential current is the phasor sum of the individual currents. The restraint current is the sum of the individual current magnitudes or the maximum of any individual current.
A typical differential relay requires the current Id to be some fraction of Ir to produce a trip signal. If the fraction is constant for any value of Ir, the relay provides a constant percentage characteristic. If the fraction changes for different values of Ir, the relay is said to provide a variable percentage characteristic.
A lower percentage setting provides higher sensitivity for lower currents where current transformer error is expected to be small. A higher percentage setting provides better security for high fault current where current transformer error is likely to be significant and high sensitivity is not as important. Modern microprocessor-based differential relays usually provide settings for minimum pickup and two percentage settings for different ranges of Ir.
High impedance differential relaying uses the fact that a current transformer must develop a voltage to produce its secondary current. The current transformer characteristics place a limit on the voltage that can be developed.
For high impedance differential relays, the current transformer secondary terminals are all connected in parallel to perform the summation. Instead of measuring the differential current, the relay places a high impedance voltage measuring element at the junction of current transformer secondary connections. If no fault is present, the currents sum to zero at the junction and the voltage is zero. If a fault is present, the currents will not sum to zero and the current transformers will attempt to force current through the high impedance relay.
Rather than measuring the current, the relay measures the voltage produced by the current transformers. The voltage produced depends on the magnitude of the fault current. For a severe fault inside the zone, the voltage tends to approach the maximum the current transformers can produce. Because the measured quantity is voltage instead of current, this type of relaying is sometimes referred to as high impedance voltage differential relaying.
Current transformer error can result in a significant voltage produced for faults outside the protected zone. The voltage setting must be low enough for good sensitivity and high enough to avoid operation on the error voltage for faults outside the protected zone.
Next month we'll explore the types of protection applied to specific equipment installed at water and wastewater facilities and some typical criteria used to develop protective settings.
