«5°* : •^o 4 * <{,> o • o - *<& <0 4> <► ♦*TV ?V v^v ^^v v^y v^sv v~v °< V ** ^ • *bK ^6* • «o 5 s ^ ^0^ ••:<** .o* ..'.*!* *o. ^s> °c 4 9* °o, **7tT»* a oP ^ 4 9* 4? * ^m^m^m^Wiw^ ***** /4fe* * / ' BUREAU OF MINES INFORMATION CIRCULAR/1990 ¥9f The Bureau of Mines Ground-Fault Protection Research Program-A Summary By M. R. Yenchek and A. J. Hudson YEARS a ***u of ^ U.S. BUREAU OF MINES 1910-1990 THE MINERALS SOURCE Mission: As the Nation's principal conservation agency, the Department of the Interior has respon- sibility for most of our nationally-owned public lands and natural and cultural resources. This includes fostering wise use of our land and water resources, protecting our fish and wildlife, pre- serving the environmental and cultural values of our national parks and historical places, and pro- viding for the enjoyment of life through outdoor recreation. The Department assesses our energy and mineral resources and works to assure that their development is in the best interests of all our people. The Department also promotes the goals of the Take Pride in America campaign by encouraging stewardship and citizen responsibil- ity forthe public landsand promoting citizen par- ticipation in their care. The Department also has a major responsibility for American Indian reser- vation communities and for people who live in Island Territories under U.S. Administration. /information Circular 9260 The Bureau of Mines Ground-Fault Protection Research Program-A Summary By M. R. Yenchek and A. J. Hudson ii UNITED STATES DEPARTMENT OF THE INTERIOR Manuel Lujan, Jr., Secretary BUREAU OF MINES T S Ary, Director ILLUSTRATIONS - Continued Page 15. Sensitive CT design 8 16. Electronic relay and CT 9 17. Prototype sensitive GFR enclosure 9 18. Block diagram for prototype ac analog relay 9 19. Schematic diagram for prototype ac analog relay 10 20. Block diagram for prototype digital relay 10 21. Schematic diagram for prototype digital relay 11 22. Bruceton Mine power system 12 23. Power feed through in A-Butt, Bruceton Mine 12 24. Load center in 12-Room, Bruceton Mine 12 25. Sensitive GFR installation, Bruceton Mine 13 26. Totalizer amplifier circuits for digital and analog GFR's 14 27. Differential current relaying using saturable transformer 15 28. Saturable transformer current sensor 16 29. Saturable transformer output voltage versus dc fault current 16 30. Drop in sensor output versus frequency 16 31. Saturable transformer prototype 17 32. Dc relay block diagram 17 33. Dc relay timing diagram 17 34. Schematic diagram of dc relay prototype 18 35. Dc relay prototype 19 36. Typical high-voltage distribution circuit 19 37. Simplified diagram of ground-check monitoring system used in mine distribution systems 20 38. System block diagram 21 39. Block diagram of relay circuit 21 TABLE 1. Counter readings of GFR performance 13 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT A ampere kQ kilohm mV millivolt dB decibel kV kilovolt Q ohm op degree Fahrenheit lb pound pet percent ft foot mA milliampere s second h hour pF microfarad V volt hp horsepower fiU microhenry VA volt-ampere Hz hertz MQ megohm W watt in inch ms millisecond Wb/A-m weber per ampere-meter kHz kilohertz MS microsecond yd 3 cubic yard THE BUREAU OF MINES GROUND-FAULT PROTECTION RESEARCH PROGRAM-A SUMMARY By M. R. Yenchek 1 and A. J. Hudson 2 ABSTRACT The U.S. Bureau of Mines designed and constructed sensitive and coordination-free ground-fault relays (GFR's) for use on mine power systems. First, a list of GFR attributes for mine ac utilization applications was compiled. These practical guidelines specified design, construction, transient immunity, reliability, and operating criteria. The time-current characteristics of the ac and dc units, subsequently fabricated, were designed to be below the human electrocution threshold. The significant and highly variable capacitance of high-voltage distribution cables was found to preclude the sensing of ground currents in the milliampere range. However, the coordination-free system developed for high-voltage distribution should enhance safety by significantly decreasing response time to ground faults. Implementation of the sensitive GFR technology in the mining industry has the potential to eliminate the majority of injuries and nearly all the deaths resulting from contact with energized components. Electrical engineer. 2 Electronics technician. Pittsburgh Research Center, U.S. Bureau of Mines, Pittsburgh, PA. INTRODUCTION An analysis of U.S. Mine Safety and Health Ad- ministration (MSHA) statistics for electrical accidents at underground mines between 1980 and 1985 showed that electricians and mechanics suffered the highest incidence of injury and death. Specifically, there were 595 acci- dents associated with repair and maintenance of electri- cal apparatus, 29 of which were fatal. The resultant cost to the public and private sectors was estimated at $28 million (i). 3 These data are not surprising since electricians and mechanics have the greatest exposure to electrical hazards. Space on mining machines is extremely limited, with electrical control boxes crowded with parts. During troubleshooting of energized circuits there is danger that, through either inattention or an inadvertent slip, an elbow or arm will contact an energized component. Under such working conditions electrical safety can be improved through the use of effective protective devices for personnel. When a worker contacts an energized component, electric current flows through the worker's body and returns to the power source, either through the earth or via a ground wire. In this case the presence of a ground wire does not preclude or mitigate hazardous leakage current. The only available safeguard in such occurrences is a ground-fault detection system. Most ground-fault protection in use on underground mine power systems is inadequate from a shock prevention standpoint. Typical response levels are in the ampere range, significantly exceeding the electrocution threshold. Increasing device sensitivity results in undesirable nuisance tripping and unscheduled interruptions in production. However, a sensitive ground-fault relay (GFR) not only can identify and act to interrupt the small deadly ground currents that electrocute people, but can ignore spurious signals that may result when motors are started or circuit breakers switched. Recent U.S. Bureau of Mines research, aimed at virtually eliminating electrocutions resulting from direct contact, developed sensitive GFR's for ac and dc utilization circuits. In addition, coordination-free relays were devised for use on high-voltage distribution. Ap- plication of this technology in the mining industry could eliminate the majority of injuries and nearly all deaths resulting from contact with energized components. ELECTRIC-SHOCK ANALYSIS Although the prevention of sustained electric shock is an ideal goal for industry, it is usually impractical. The detection of such shocks and resultant body currents as low as 10 mA would likely be a complete impediment to production. Consequently, a more realistic goal is to design protection against electrocution, not electric shock. To devise effective personnel protection, it is first necessary to understand how electrical current can be lethal to humans. Ventricular fibrillation is by far the most common cause of death from accidental electric shock. This condition is induced when sufficient current flows through the chest and disrupts the nervous system impulses, internal to the heart, that synchronize normal heartbeat. The heart no longer acts as an efficient pump to circulate blood, and death is likely to occur within minutes. In light of this, safety can be enhanced if the potential current flow through the body can be minimized. This risk of electrocution is determined to a large ex- tent by power system configuration. To maximize safety, the recommended arrangement for ac systems features a direct- or derived-neutral point on the source trans- former secondary, tied to earth through a grounding resistor. Equipment frames are then grounded by a grounding conductor connection to the grounded side of the resistor. When a grounded worker accidentally contacts an energized conductor, the body current is limited in magnitude by the grounding resistor (fig. 1). Ventricular fibrillation is a function not only of current magnitude (I), but also of frequency, duration of expo- sure (t), and weight of the victim. The threshold was Italic numbers in parentheses refer to items in the list of references preceding the appendix at the end of this report. Figure 1 .-Resistance-grounded power system with grounded worker. statistically defined by Dalziel and Lee (2) as that current through the chest that will cause ventricular fibrillation in 1 out of 200 people. For 110-lb individuals, the 60-Hz threshold was expressed as I = 116/ Ji, (1) where I = current, mA, and t = time, s. This relationship is shown graphically in figure 2. For brief exposures, relatively more current is needed to cause fibrillation. For longer durations, the limit decreases, down to 50 mA, below which fibrillation is unlikely no matter how long the exposure. Given these constraints, an optimum ohmic value can be determined for the neutral grounding resistor, R g , shown in figure 3. This value should be high enough to protect against ventricular fibrillation, yet sufficiently low to permit reliable ground-fault detection without nuisance tripping. Ideally, the resistor should limit current below the electrocution threshold for a direct contact shock. In such instances, the body resistance, R b , may be as low 350 4 5 Figure 2.-60-Hz fibrillation threshold for 110-lb individuals. as 500 (3). The equation defining the 60-Hz electro- cution threshold may be rewritten to determine the maximum nonfibrillating current for the total circuit clearing time as follows: where and I = 116/7(1! + t 2 ) , tj = operating time of GFR, s, (2) t 2 = operating time of molded-case circuit breaker, s. A relay operating time of 100 ms provides sufficient time delay to prevent nuisance tripping. A molded-case circuit breaker typically opens its contacts within 34 ms. Given a total operating time of 134 ms, the maximum current that will not result in fibrillation is 317 mA. This can now be used to define the ohmic value of the neutral grounding resistor, R (fig. 3): R g = (V in /I)-R b , where V- = line-to-neutral system voltage, V, (3) and I = maximum nonfibrillating current, A, R b = human body resistance, ft. For example, on a 480-V system, R g = 374 ft, and ground faults are limited to a maximum of (480/75) /374 = 740 mA. The human response to currents of varying frequency is shown in figure 4 (4). It is an unfortunate fact that humans are most sensitive to 60-Hz signals. The reason for this is that, physiologically, the muscles and the nerves of the body are most easily stimulated by changes in current magnitude. The ac sine wave is characterized by constantly changing magnitude, as opposed to pure dc where the only change occurs the instant the circuit is made or broken. Consequently, about 3.5 times more dc than ac is required to induce ventricular fibrillation (5). Conversely, the muscles and nerves do have a finite reaction time, such that with increasing frequency, the V|n© Rb Figure 3.-Electrical accident equivalent circuit stimulation of one alternation does not have time to elicit a response before it is annulled by the succeeding alternation. For dc power systems, a three-phase bridge rectifier arrangement is preferred (fig. 5). Shock hazards are reduced, not only by the presence of the grounding re- sistor, but also because line-to-ground voltage is one-half the line-to-line dc voltage. When a grounded worker inadvertently contacts an energized positive or negative conductor, the resultant current through the individual is half-wave rectified as shown in figure 6. Research con- ducted recently by Bernstein (<5) has established that the presence of an 18-pct ripple in a half-wave-rectified wave tends to reduce the threshold such that the following relationship applies (fig. 7): i = 348/yr, where I = current, mA, (4) and t = time, s. These electrocution threshold characteristics are an im- portant consideration for the design of sensitive GFR's. 600 10 100 1,000 10,000 FREQUENCY, Hz Figure 4.-Fibrillation threshold at currents of varying frequency. m ooo_ til Ripple frequency = 180 Hz TIME Figure 6.-Half-wave-rectified ground-fault current 900 Figure 5.-Three-phase resistance-grounded system feeding full-wave rectifier. "012345 Figure 7.-Half-wave-rectified dc ventricular fibrillation threshold. ALTERNATING CURRENT UTILIZATION Initial Bureau research concentrated on providing the ac utilization portion of a mine power system with sensitive ground-fault protection. The utilization system includes portable power cables, power centers, rectifiers, motors, and the associated protective devices. It is the most troublesome part of the power system in terms of safety and reliability because of its temporary nature. As mining progresses, the utilization system is stretched to its limit and repositioned, necessitating frequent handling of trailing cables and equipment repairs, a major source of electro- cutions underground (1). The presence of large motor loads strains the dependability of protection devices by introducing large voltage and current transients. Con- sequently, the application of effective personnel protection for utilization circuits would have a major impact on underground safety. Zero-sequence or balanced-flux relaying (fig. 8) is the most reliable and most common method employed for ground-fault relaying (7). As shown in figure 8, the phase conductors pass through the toroidal current transformer (CT) window. The sum of the three phase currents is the CT primary current and is proportional to the zero- sequence current (8). An unfaulted balanced system features little or no zero-sequence current, and the CT secondary current is approximately zero. However, when a ground fault occurs, the resultant secondary current is used to trip the relay. Zero-sequence relaying is unaffected by phase voltage fluctuations, and since only ground leakage is monitored, the relay can be made very sensitive. Consequently, it is the only practical technique capable of responding to low-level hazardous ground currents and the method of choice for the Bureau's sensitive GFR research program. BACKGROUND RESEARCH As a first step in the design process, a list of GFR attributes for ac mining applications was compiled. Next, test procedures were devised to ensure device compliance Phase A — B- C- ZN) sequence relay Ground wire Figure 8.-Zero-sequence relaying. with the desired operating characteristics. These practical guidelines (9), addressing GFR design and construction, transient immunity, reliability, and time-current char- acteristics, are summarized below. Proper Design and Construction A relay system of suitable design and construction does not pose a personnel hazard, reduces the amount of down- time caused by GFR failures, and facilitates acceptance of the GFR as a useful safety device. Electronic instruments designed and constructed for military use must comply with Military Standard 454 (10). The key portions of that standard, which can be applied to GFR's in underground mining, are related to safety and accessibility: The design shall incorporate methods to protect personnel from accidental contact with voltages greater than 30 V root- mean-square (RMS) or dc during normal operation. All external surfaces shall be at ground potential during nor- mal operation. All terminals shall be corrosion resistant. Sharp external projections shall be avoided. Suitable ac- cess shall be provided for adjustments, testing, and routine maintenance. No unsoldering shall be necessary to remove the front cover for troubleshooting. Mine Worthiness Underground, GFR's are located inside metal-clad load centers, so both the relay and CT must have metal mount- ing lugs. Terminal strips should be sized for No. 12 AWG wire. In addition, the relay case should be moisture and dust resistant. Size Limitations Space is typically limited inside mine power equipment compartments. Since several GFR's may be used in a sin- gle power center to protect all outgoing circuits, they must not be much larger than present GFR's. Thus, the relay components should be mounted in a compact enclosure not exceeding 3 by 6 by 6 in. To minimize flux leakage, the CT window should only be large enough to accommo- date the encircled power conductors. The outside dia- meter of a 4/0 single-conductor cable is 0.807 in (11). Three such cables fit snugly through a 1.750-in-diameter window. For ease of installation of cables with terminals, the window diameter should be increased to 2.100 in. Present ground-fault CT's in use underground have outside diameters smaller than 4 in. Since they are placed between the molded-case circuit breaker and the load- center coupler, they are no more than 3 in wide. Electrocution Prevention The primary reason for employing sensitive GFR's in mining is to prevent accidental electrocutions. For 60-Hz circuits, the desired region for GFR operation is below and to the left of the threshold shown in figure 2. To define GFR time-current characteristics, a variable 60-Hz voltage source in series with a 50-fl, 225-W fixed resistance is used to inject current through the CT primary as shown in figure 9. A double-pole, single-throw switch initiates the test and triggers the storage oscilloscope, which monitors relay contact activity. Test currents are varied from to 1,000 mA. Power Harmonics GFR CT, should not falsely activate the relay. In addition, they should not damage the relay control circuitry. An impulse generator, constructed in accordance with Underwriters Laboratories (UL) Standard 943 (13), is em- ployed to simulate transient overvoltages as they would occur on residential and industrial power systems. The test circuit consists of a relay switch and resonant circuit, shown schematically in figures 11 and 12. The generated waveform exhibits the following characteristics under no load: The filtering for GFR's must be designed so as to pre- clude false tripping by any harmonics superimposed on the power conductors. However, attenuation of these higher frequencies must not be so severe that hazardous currents above the electrocution threshold (fig. 4) are permitted. In testing GFR frequency response (fig. 10), an audio oscillator and power amplifier provide high- frequency currents from 60 Hz to 10 kHz. For each frequency, the voltage is slowly increased until the GFR activates. Voltage Surge Immunity Mine power systems frequently experience voltage surges when circuit breakers and switches are opened or closed. Although the duration of these transients is less than 50 ms, past research indicates their magnitude can reach five times the utilization voltage (12). These surges, when present on the power conductors encircled by the Resistor — WV L — -To oscilloscope v. trigger A (£) Voltage V- / source ..Current probe to oscilloscope channel 2 JT T_ To channel 1 Relay Figure 9.-Test setup to determine 60-Hz tripping characteristics. 1. Initial rise time of 0.5 ps between 10 and 90 pet of peak amplitude; 2. Period of following oscillatory wave, 10 /is; and 3. Amplitude of each successive peak, 60 pet of the preceding peak. The amplitude of the first peak is fully adjustable from to 9,000 V. In the first part of the test, 10 successive 5- kV surges are imposed on the power conductors encircled by the CT while the relay contacts are observed. Next, ten 1-kV impulses are applied in parallel with the 120-V ac control voltage, and at random with respect to its phase. Afterward, the relay is operated at 60 Hz to detect possible damage to circuitry. Line output Relay control rrn Cathode-ray oscilloscope trigger output Neutral ground Figure 11. -Impulse generator circuit CT Frequency counter Frequency generator \®^n =z 7^. <£> Ammeter Figure 1 0.-Frequency response test circuit Light I20V 60 Hz II, 50 V A lOkil, IW D l -w* ►Hr — t 32 M F, 250 V 11* >' CR| relay Cathode-ray oscilloscope gate input R 3 I kfl, '/feW I M.a, 7 2 W rm Figure 12.-Relay control circuit for impulse generator. Common-Mode Transients Safe Failure Modes Sensitive GFR's must be unaffected by large transient currents occurring simultaneously on all phases of an ac utilization circuit. Such currents may briefly exceed six times full-load motor rating during starting or under heavy intermittent load. The maximum short-circuit settings listed in 30 CFR 75.601 (14) effectively limit balanced three-phase loading to 2,500 A. Nevertheless, balanced currents up to 2,500 A should be tolerated for up to 5 s without activation of the GFR. A three-phase high-power source is used to variably supply balanced three-phase currents through a shorted trailing cable encircled by the GFR CT. The voltage is increased until the relay activates or the 2,500-A ceiling is reached. Tripping thresholds are confirmed through repeated tests. Current Withstand The molded-case circuit breakers used on low-voltage ac mine power circuits typically have an interrupting rating of 30,000 A. Currents of this magnitude are quite possible during three-phase faults. Since the GFR CT is a part of the power system, it too should withstand up to 30,000 A for the time it takes the breaker to clear (a few hertz). The withstand test is conducted with a high-current circuit breaker tester as shown in figure 13. The tester is equipped with an initiate switch that can be jogged to reasonably control the test duration. Current magnitudes are recorded on a storage oscilloscope connected across a 400-A, 100-mV shunt. The CT secondary is shorted to preclude high secondary voltages. The CT is subjected to 30,000 A for approximately 4 Hz. The 60-Hz current ratio and winding resistance are measured before and after the withstand test to detect any degradation of the CT. Quality Assurance For dependability underground, all devices from the same manufacturer should be consistent in electrical and mechanical performance over a reasonable service life. In addition, each GFR should be equipped with a means to test its operation. In the event of failure of the GFR's internal circuitry, it is vital that the unit activate its associated circuit breaker to remove power and prevent a false sense of security. Two common failure modes are the loss of 120-V control power to the GFR and an opening of the CT winding. PROTOTYPE DESIGN AND CONSTRUCTION Sensitive GFR's are available commercially for use in British mines and in the U.S. irrigation industry. Three such brands were evaluated in-house for potential appli- cability to U.S. mining using the above criteria. None was found suitable without design modifications (15). How- ever, these tests served as the basis for the development of a mine-worthy sensitive ground-fault protection system consisting of a CT and an electronic relay. Current Transformer The CT for a sensitive ground-fault system has the duty of precisely sensing the existence of small ground-fault currents on three-phase line conductors feeding mine ma- chinery. It must be able to distinguish these faults within the complex electromagnetic environment that exists in mine power equipment. Specifically, the CT must not send erroneous signals to the relay circuitry when in the pres- ence of external magnetic fields and when observing high common-mode line currents. This precise device must also be able to physically, electrically, and magnetically with- stand the mine environment. The output of the CT secondary was anticipated to be in the millivolt range when sensed currents were in milliamperes. The following expression is used to pre- dict open-circuit output voltage for a given ground-fault current (16): E = = hNuflplnCR^Rj) (5) where E = RMS output voltage of secondary volt- age, V, h = toroid core width, m, To storage oscilloscope Mine duty circuit breaker Secondary shorted Figure 1 3.-Current withstand test circuit and N = number of turns on secondary winding, u = core permeability, Wb/A-m-turns, f = frequency, Hz, I p = RMS value of ground-fault current, A, R Q = outer toroidal radius, m, Rj = inner toroidal radius, m. Equation 5 is valuable in designing a current sensor for a sensitive ground-fault protection system. It shows that CT output voltage may be affected by various electrical and mechanical quantities. In addition, consideration must also be given the CT burden, or external load impedance on its secondary. The burden impedance cannot be so low as to reduce the sec- ondary voltage to an unusable value. Voltage input to the relay must be sufficiently high so that amplification and noise concerns are minimal. Since core dimensions are limited by available space, voltage output can be maxi- mized by using a high-permeability core and a large number of secondary turns (see equation 5). Also, a high- permeability core minimizes flux leakage inherent with window-type CT's. Cores are usually constructed of thin laminations to reduce eddy current losses. For the GFR toroid, the core was made by continuously winding a 0.006- by 0.75-in tape of magnetic material starting with a 2.5-in ID. After winding to an outer diameter of 3.5 in, the core was annealed to remove strains and impurities from the tape. An 80 pet nickel-iron alloy core was able to meet the accuracy requirements of sensitive ground-fault relaying. However, one problem with nickel-iron cores is that then- excellent magnetic properties can be degraded by winding stresses and pressures. A nonmetallic or phenolic box (fig. 14) around the core provided protection against mechanical damage as well as insulation for the secondary winding. The effects of vibration and shock are mitigated by cushioning the core in a silicon casing. The CT must be able to operate within significant mag- netic fields generated by nearby power transformers or the monitored power conductors themselves. To minimize any volts-per-turn imbalance requires a regressive winding technique (fig. 15) (17). One-half of the 250 secondary turns of No. 22 AWG magnet wire were wound around the core in one direction. The second half were wound around the core in the opposite direction across the first half so that each second-half turn falls between a pair of first-half turns within the window crossing on the CT sides. A separate single turn was wound on the core for primary injection testing. As stated previously, power conductor common-mode currents can produce local core saturation, noncancelling voltages across the CT secondary, and nuisance tripping of the relay. It was found (18) that a 2.1-in-ID, 0.1-in-thick, concentric, low-permeability iron buffer adjacent to the power conductors tended to distribute local flux saturation effects. The core, windings, and buffer were potted in epoxy and enclosed in a metallic housing that has convenient holes for mounting the assembly inside mine power equip- ment. The output cable contains the four conductors for the CT secondary and the test winding. Electronic Relay Solid-state sensitive GFR's were designed to operate with the CT (figs. 16-17). Two versions were conceptu- alized, one based upon analog techniques, the other using Nonmetallic insert Nonmetallic case Insert cushioning Tape- wound core Figure 14.- Nonmetallic core box construction. Toroid core, 2.5-in-ID, 3.5-in-OD, 0.75-in-high, tape-wound with 0.006-in nickel- iron alloy, cased and phenolic boxed. 250-turn, regressively wound CT secondary Transformer secondary leads Test winding leads -2.l-in-ID low-permeability steel pipe with single wrap of 0.00l-in nickel- iron alloy on outer surface. Figure 1 5.-Sensitive CT design. digital (19). Both incorporate the same control power supply and electromechanical relay trip circuitry mounted on interchangeable cards. Analog Version A block diagram of the analog relay is shown in fig- ure 18 and a schematic is shown in figure 19. Since the sum of the three line currents that form the primary cur- rent in the CT is nearly zero under normal operating con- ditions, normal phase currents should not trip the relay. However, when a ground fault occurs, a voltage will be induced across the secondary in proportion to the fault magnitude. The CT (T 2 ) secondary was connected across a burden resistor (Rj) and an inverting operational ampli- fier QA : (fig. 19). Diodes D 1 and D 2 were included to limit short-duration transients at the GFR input. A first-order, low-pass filter was utilized to match the relay frequency response to that of humans (fig. 4). This filter consisted of operational amplifier QA 2 , resistors, R 5 ♦ 12 3 i 1 L_J Figure 16.- Electronic relay and CT. h 2 - 32 "H 0.596" typical 0.86" typical |— \ -E5B 0.2l9-in-diam mounting holes (4) Figure 17.-Prototype sensitive GFR enclosure. R 6 , R 7 , and capacitor Q. It had a 3-dB cutoff frequency of 2 kHz and a gain of 2. Full-wave rectification was accomplished by the combination of the two operational amplifiers QA 3 and QA 4 . Diode D s , resistor R 14 , and capacitor C 5 provided a smoothed dc output signal comparable to the peak value of the input sinusoidal voltage at QA 3 . The smoothed signal was then applied to the noninverting input of the comparator QAj. This stage was designed such that its output state changes abruptly from zero to a high voltage when the input reaches an adjustable pickup reference voltage. QAi Amplifier QA 2 Filter QA 3 , QA 4 Full- wave rectifier R-C Integrator QA 5 Level detector R-C, Q, Definite- time delay Open-CT trip signal Qz Trigger -ojp- Reset Target Trip relay :N.C :N.O. Figure 18.- Block diagram for prototype ac analog relay. During normal operation, the current from the 24-V supply flows through the coil of the normally closed elec- tromechanical relay K ls resistor R^, and the green light- emitting diode (LED) D 14 , because the thyristor, Q 2 , is off. When a ground fault occurs above the pickup setting, the voltage signal from the comparator starts to charge capaci- tor C 6 through the adjustable resistance R^ and fixed re- sistor R^. This continues until the unijunction transistor (UJT) turns on and discharges C 6 . The GFR operating time is determined by the time it takes to turn on the UJT. To activate the GFR, a trip signal greater than the trip level setting must still be present after the desired delay has occurred. The output of the UJT initializes the gate of the thyristor, Q 2 . Once the thyristor is fired, it latches on. This deenergizes the electromechanical relay and turns on the red LED D 13 . Several fail-safe characteristics were incorporated into the relay design. For example, a loss or significant drop of control power deenergizes the electromechanical relay coil. In addition, if the secondary winding of the CT disconnects from the analog circuitry, the relay will initiate a tripping signal. An internal power supply provides regulated ±15-V power for the relay's active integrated circuits, as well as an unregulated supply of approximately + 24 V for the re- lay K, and its associated network. The 120-V primary of the power supply transformer is fuse protected and in- cludes transient suppression. A test circuit, consisting of a test winding on the CT T 2 , momentary switch S 2 , and a current-limiting resistor R 33 is provided on the relay. When the switch is closed, a test current greater than the relay pickup setting is primary injected into the CT, and the relay activates. Digital Version A block diagram of the digital prototype is shown in figure 20 and a schematic in figure 21. As with the analog version, the CT T 2 is connected across a 1,000-Q bur- den R 2 . The ac output voltage of the CT is amplified by the inverting operational amplifier QA 1B . The output 10 I k,Q + 15 V R20 10 Mil R|5 R22 22kft IOkil : -W\l — — + I5V' K-@TP 3 Time ± R23 delay ■ 500 kU adjustment -@TP, / \ D| 3 D| 4 (Red) (Green) X C10 0.1 uF ■= 35 V "~ ""-15 V MC79LI5CP Figure 19.-Schematic diagram for prototype ac analog relay. OAia 0pen-CT detector QA IB Zero- sequence amplifier QA 2A Filter QA 2B Level detector Time delay detector Qi Trigger 4 i -0J0- Reset N.O. Target Trip relay Figure 20.-Block diagram for prototype digital relay. of QA 1B is then fed into a first-order, low-pass filter with a 3-dB cutoff of 1.6 kHz. The purpose of this filter is to attenuate high-frequency signals such that false tripping due to harmonic currents or high-frequency, long-duration transients may be avoided. This attenuation should not be too steep, because high-frequency currents can cause elec- trical fatalities. The ac output of the filter network is fed to the com- parator QA 2B through a dc-blocking network consisting of Q, R 12 , and C 4 . Consequently, an ac-filtered trip signal will be delivered to the negative input terminal of this dif- ferential amplifier. The voltage divider consisting of R 13 and R 14 provides a positive reference voltage to the posi- tive input terminal of the differential amplifier. When the sinusoidal trip signal at the negative amplifier input is less than the trip level setting at the positive input terminal, the output of the comparator is at its positive saturation value of about + 13 V dc. At any instant when the trip signal is greater than the trip level setting, the comparator output goes to its negative saturation value of -13 V dc. During this time, diode CR 5 clamps the inputs to the CD4528 retriggerable one-shot and the 555 timer to ap- proximately V. The retriggerable one-shot remains in its stable state of unity logic output as long as a l-to-0 transition does not occur at its input. However, when such a transition does occur (corresponding to a trip signal), the retriggerable one-shot goes to its unstable state of zero logic output for as long as the trip signal is present. Should the input switch back to a high state before the GFR activates, the one-shot will return to its stable state and GFR operation will be blocked. 11 Test © winding Figure 21 .-Schematic diagram for prototype digital relay. The GFR time delay circuitry includes the 555 timer, the IQu NOR gate, and the CD4528 retriggerable one- shot. As long as no l-to-0 transitions occur at the timer input, its output remains in a stable zero state. When a l-to-0 transition does occur (corresponding to a trip ini- tiation), the 555 timer output becomes unstable (1) for 100 ms, the GFR intentional delay time. The IC^ NOR gate isolates the timer output from the CD4528 one-shot input. The relay firing circuitry includes the IC 5A NOR gate, the 2N1595 thyristor, Q u the electromechanical relay K 1} and associated components. The relay tripping signal, a high NOR output, is developed only if the outputs of both CD4528 one-shots are zero. This occurs only if the relay delay time has expired and the trip signal exceeds the trip level setting. The remainder of the firing circuit is similar to the analog version, as are the power supply and test circuit. The digital version also has open-CT and loss-of- control-power safety features. Four analog and four digital prototypes were con- structed under phase 1 of Bureau contract J0134025 (19). Their ability to detect low-level faults was confirmed in the laboratory. What remained to be demonstrated was satis- factory operation in an underground mining environment. FIELD AND LABORATORY EVALUATIONS To ensure an unbiased evaluation, the Bureau assumed the responsibility of testing the units at a cooperative commercial mine. This demonstration would have a two- fold purpose: 1. To expose the prototype GFR's to the transients and anomalies associated with a mine power system to gauge circuit durability. 2. To measure the number of tripping events and de- termine if false or nuisance tripping is of concern when GFR's are actually providing ground-fault protection. Bruceton Mine Initially, the economically depressed coal industry was unresponsive to published solicitations for coopera- tors. Consequently, the devices were first installed at the 12 Bureau's Bruceton Mine near Pittsburgh, PA, to establish some performance history. Bruceton Mine Power System Electrical power at a potential of 7,200 V is delivered to the mine substation on the surface, as shown in fig- ure 22. The three-phase, 480-V utilization voltage for the mine is derived via a delta-wye configuration. Ground- fault currents are limited to 15 A by an 18.5-0 resistor connected between the secondary neutral point and the earth grounding bed. From the substation, power feeds into the surface control building (building 07) where it divides to supply both the Experimental Mine and the Safety Research Coal Mine (SRCM). Overcurrent, short- circuit, undervoltage, and grounded-phase protection are provided at the beginning of these branch circuits, located within a wall panel in building 07. The sole ground-fault protective device (GFR 2) for the Experimental Mine is in this panel. From building 07, power for the SRCM is transmitted underground through a borehole. A 480-V, three-phase service disconnect is situated underground at the bottom of this borehole. From this disconnect, power is fed into A-Butt, where three-phase auxiliary and single- phase lighting power is tapped (fig. 23). The auxiliary circuits, infrequently used to maneuver equipment in and out of the mine, are equipped with grounded-phase pro- tection (GFR's 7 and 8). In 12-Room, the 480-V power connects to a load center (fig. 24) for the portable mining equipment. Grounded-phase protection is incorporated on each of the four outgoing circuits. In summary, as shown in figure 22, ground-fault protection is provided at eight locations within the Bruceton Mine power system. Toroidal transformers 7,200/480 V r~ _Buildlna _07 Surface substation ©) (D) Safety Research Coal Mine -Load center 3-phase, 480-V 12-Room A-Butt !®;)@WW) ! [-1-1-1 — I_J -L I Roof Loader bolter Cutter Auxiliary 3-phase power Experimental Mine Auxiliary fan encircling the power conductors are connected to small socket-type relays with sensitivities of 6 A. Ground-Fault Relay Installation To preclude extensive tripping of the power, it was in- tended that the prototype sensitive GFR's operate event counters in lieu of activating the circuit breakers each time leakage current exceeded 50 mA. The existing ground- fault protection and grounding resistor remained in service. Simply connecting the unit's control power to the mine power system and exposing it to anomalies and transients constitutes one test of mine worthiness. Eight prototype GFR's were available for the demon- stration. Since their durability was unknown at the outset, it was decided to utilize only six, keeping two as spares. Consequently, no sensitive GFR's were installed on the rarely utilized auxiliary circuits in A-Butt. Two analog and two digital versions were installed underground in the load center; two digital units were placed in building 07. Figure 23.-Power feedthrough in A-Butt, Bruceton Mine. Figure 22. -Bruceton Mine power system. Circled numbers indicate GFR's. Figure 24.-Load center in 12-Room, Bruceton Mine. 13 All prototype GFR's installed underground were ad- justed for a definite time delay of 100 ms, the recom- mended setting for shock protection. To incorporate a degree of coordination, the upstream devices in build- ing 07 were fixed with a delay of 250 ms. As shown in figure 25, the CT's of the sensitive GFR's were installed around the three 480-V phase conductors. The single-phase control power for the electronic relays was obtained using 480/120-V control transformers. The electromechanical counters were rated at 120 V and 6 W. Upon installation, the test circuit of each GFR was ex- ercised to verify operation. At each location (building 07 or the 12-Room load center) all relays tripped when the test button of any one was activated. The following were investigated as possible sources of this false tripping: • Physical orientation of the relays (because of space limitations the units were installed on their sides in the load center), • Induced currents in parallel CT secondary leads, • Low control voltage, • Voltage transient from the counter activation. The GFR's were tested in the laboratory with the relays upright, on their sides, upside down, etc.; the sensitivity and time delay were unaffected. Next, two GFR's were energized with their CT secondary leads at various dis- tances and orientations with respect to each other. No differences in performance were detected. Voltage mea- surements of the control transformer underground showed it remained constant at 120 V when the test buttons were pushed. Further tests in the laboratory revealed that the relays would not trip until the control voltage dropped to 65 V. Finally, the counters were disconnected, and the relays operated properly without interfering with each other. These counters consist of an electromagnetic coil that, upon loss of signal, counts by disengaging a ratchet wheel. The switching of this inductive load created a voltage transient that activated the other units. Several methods of transient suppression were tried in the laboratory without success. Metal-oxide varistors (MOV's) were connected across the input of the relay electronics; capacitors were placed across, and ferrite-core inductors in series with, the relay power supply. Only the placement of 0.33-^F capacitors across the counter coil solved the problem. It was felt that GFR transient immunity was not com- promised by the tripping associated with the electro- magnetic counters. The voltage transient created by the inductive coil switching was severe, but extremely localized in the power system. The switching of large motor loads should not give rise to similar problems because of the damping effect of the intervening cable impedance. Ground Fault Relay Performance The relays were then connected to the mine power sys- tem, with mine personnel periodically inspecting the units and resetting those that tripped. Table 1 lists the GFR performance at the end of 30 days. Table 1. -Counter readings of GFR performance Relay location 1 1 2 3 4 5 6 After 30 days 3 1 9 3 5 3 1 2 1 n After additional 2 weeks . . . 'See figure 22. 480-V, 3-phase Figure 25.-Sensltlve GFR Installation, Bruceton Mine. At this time all units were removed from service and examined in the laboratory. Both digital units installed in building 07 were malfunctioning. GFR 1, protecting the SRCM, could not be reset and remained in a trip mode. The GFR for the Experimental Mine would not trip when tested. After troubleshooting, the following components (fig. 21) were found defective and repaired: GFR 1: Digital IC 5 , CD4001. GFR 2: Thyristor Q^ 2N1595; reset switch S,; in addi- tion, the main power transformer Tj was resoldered. Before the evaluation was continued underground, the prototypes were thermally stressed at 120° F while ener- gized in an environmental chamber for 1 week. No addi- tional malfunctions were uncovered. The units were then reinstalled at their original locations on the mine power system and monitored for 2 weeks by mine personnel. The largest motor fed from the load center, the 60-hp, 480-V pump motor for the cutting machine, was started repeat- edly to determine if the resultant current inrush would affect GFR 3. No false trips were observed. The ab- sence of tripping when the pump motor was started is encouraging from the standpoint of immunity of transient common-mode currents. At the end of this time, digital GFR 5 was found inoperative because of defective IC 5 , CD4001. In addition, it was observed that when the relay electronics cards were inserted into the relay enclosures, they did not always contact the rear terminals without being jiggled. 14 The prototype ac GFR's were installed on the Bruceton Mine power system for a total of 6 weeks. During this period they were exposed to anomalies and transients as- sociated with the system. The count history revealed a high number of trips for the GFR protecting the Experimental Mine (GFR 2, table 1). It may be specu- lated that the relays in building 07 were more accessible than those underground and were reset with a higher frequency by mine personnel. Also, it may be speculated that the malfunctions of the digital relays in building 07 were the result of a design flaw or their greater exposure to upstream power system transients, especially lightning. Additional studies were necessary to pinpoint the causes of these malfunctions. Ground Fault Relay Modifications The 6-week evaluation at the Bruceton Mine pointed out several deficiencies in the test setup and the GFR design, which were addressed in followup laboratory work. First, the use of 120-V electromechanical counters to monitor relay activation was judged to be of limited utility, since mine personnel were relied upon to check the units visually. An unattended GFR may count one trip when in reality it could have tripped many times with prompt reset. Consequently, electronic totalizers were procured as replacements for the electromechanical counters. Char- acterized by high input impedance, the totalizers featured an internal battery power supply and were sensitive to 6-V pulses. These monitoring devices were connected to the GFR's internal circuitry via amplifier circuits shown in figure 26. The totalizers monitor the presence of tripping signals, internal to the GFR, that are present any time leakage currents through the CT exceed 50 mA for the prescribed time delay. Since the internal tripping signals are unaffected by relay status, trip events can be monitored without reliance on mine personnel. As shown in figure 21, outputs 3 and 11 of NOR gate ICjg were grounded. However, through inversion these outputs became high and should be floated. Cor- rection of this error eliminated the overheating of ICjb (CD4001) in the digital GFR models. In addition to these basic changes, further modifications were recommended. Troubleshooting in the laboratory was hampered by the lack of readily identifiable test points on the printed circuit board. Key circuit locations could be made more accessible by the installation of color-coded test points on the side of the board. Also, to permit measurements while the GFR is in service, extender boards could be fabricated. Extracting the printed circuit board from the metallic enclosure opens the CT secondary. During service, high 2N2222 To k ,Q rose "^-0|— HMrl |l«- DO !2g CJ* jjjznfi m — — I a >• a u •o "5 E a o> B ■o o a E £ i il "\A^-{gpH 19 The components of the prototype are mounted on an interchangeable board within a metallic housing as shown in figure 35. This compact design measures 8-3/4 in high by 5-1/2 in deep by 2 in wide. LABORATORY EVALUATION The sensitive dc GFR prototypes were evaluated for immunity against voltage and current transients in a manner similar to that used for the ac prototypes. In all tests, the relays did not falsely activate nor did they become damaged by the simulated power system surges. The dc GFR's are available for installation in an underground mine having dc face equipment fed from a three-phase full-wave rectifier. ALTERNATING CURRENT DISTRIBUTION The final phase of the Bureau's ground-fault research program focused on the ac distribution portion of coal mine power systems. These high-voltage circuits feature a series of switchhouses, each with inherent ground-fault protection, distributed along the way from the surface to the section load centers. The operation of the electro- mechanical GFR's, installed in these switchhouses, is coordinated by large, intentional time delays to ensure that faults are cleared without interrupting power to sound, upstream circuit portions. These additive delays may in practice average 2 s per relay. As a result, upstream relays respond slowly to hazardous ground currents and workers are exposed to fire and burn hazards. In response to these problems, the Bureau conducted research to design and construct a coordination-free relaying system that reacts to a ground fault 50 to 100 times faster than present pro- tective systems. To understand the concept of coordination-free re- laying, a series of switchhouses must be visualized installed along a high-voltage distribution line (fig. 36). If a relay in an upstream switchhouse detects a ground fault and a downstream relay does not, the fault must be in the zone joining them. Consequently the upstream relay will acti- vate the local circuit breaker. If both upstream and down- stream relays detect faults, the fault must be downstream from the zone and neither relay will activate. BACKGROUND RESEARCH Initial research centered on determining the highest practical ohmic values for high-voltage grounding resistors. These components, connected between the transformer- secondary neutral point and the earth grounding medium, restrict ground-fault current magnitude. To provide pro- tection against electrocutions, this limitation should be in the milliampere range. Underground high-voltage cable is required to have metallic shielding around each power conductor. Con- sequently, this cable exhibits significant distributed Figure 35. -Dc relay prototype. SH I SH 2 SH 3 Figure 36.-Typical high-voltage distribution circuit 20 capacitance not only among power conductors, but also between power and ground conductors. In addition, the use of discrete capacitors for surge and power-factor correction adds to the circuit capacitance. Thus, dis- tribution circuits feature charging currents of up to 10 A. Through a computer-assisted transient evaluation of mine distribution circuits, it was concluded that instability results if ground faults are limited below capacitive charging currents. Thus, it was recommended that present ground- fault-limiting levels of 25 A continue to be utilized, precluding complete personnel shock protection. With coordination-free relaying, detected-fault status is continuously transmitted from the downstream to the up- stream relay. Utilizing dependable blocking logic, the signal from the downstream relay acts to prevent tripping. The most practical communication link available appears to be the ground-check conductors already present in almost all high-voltage cables used with pilot ground-check monitors (GCM's). Ground-check monitoring systems for high-voltage dis- tribution usually monitor the integrity of the grounding conductor connected between switchhouses (fig. 37). Separate, independent GCM units are installed in each switchhouse. GCM's commonly circulate low-voltage 60-Hz signals through the pilot-ground circuit. Con- sequently, a transmission frequency of 1,000 Hz was chosen for the coordination-free ground-fault relay (CF-GFR) blocking signal. Active electronic filters were used to couple and decouple the superposed signals. Two intentional delays were proposed for the ex- perimental relays: 50 ms for primary and relay-racing protection and 150 ms for backup protection. Primary protection must include time for fault signal detection and a safety factor. Relay racing occurs if a blocking signal is not received in time to prevent false tripping upstream. The backup delay is desirable should the downstream breaker fail to open during a ground fault. It was recommended that backup be graded in 150-ms increments, with the most downstream breaker in a series path having a 50-ms delay as primary protection and no backup relay. The next outby breaker would have a 50-ms delay for primary protection and a total delay of 200 ms in backup. The next relay would exhibit delays of 50 and 350 ms, respectively. PROTOTYPE DESIGN AND CONSTRUCTION Once design criteria were established, a CF-GFR was designed and three prototypes were constructed. Block diagrams for the overall system and each relay are shown in figures 38 and 39, respectively (27). The CF-GFR design has been fully documented through schematics, wir- ing diagrams, and component listings (22). The blocking signal, tuned to 1,000 Hz, is coupled to the pilot and ground wires through a switch controlled by a logic device. When a ground fault is sensed locally, the coupling is accomplished. The purpose of the attenuator is to avoid a very low impedance path between the pilot and ground wires. Such a path would severely reduce the blocking signal amplitude and prevent its propagation to the upstream switchhouse and relay. The upstream detector circuit checks for proper ampli- tude and frequency in the incoming downstream blocking signal. The 60-Hz GCM signal is attenuated by the band- pass filter. When the blocking signal is detected, a block- ing logic signal is sent to the trip element actuator to in- hibit primary protection tripping. Local zero-sequence ground faults are sensed by the toroidal CT. When this ground current is greater than the pickup level of 5 A, a trip signal is sent to the trip element actuator after the appropriate delay interval. However, primary protection is inhibited if a blocking signal has been detected from the downstream relay. Backup delay time is then sensed. The trip element is then actuated at the end of the backup delay interval if the locally sensed ground fault is still greater than the pickup setting. Additionally, if the local fault-sensing CT becomes open, then, irrespective of any other signals, a trip is Pilot conductor 2 Grounding conductor GCM J? Pilot conductor Grounding conductor GCM J) — Pilot conductor t Grounding conductor SH 1 SH 2 Power flow O Figure 37,-Simplified diagram of ground-check monitoring system used in mine distribution systems. GFR signal generator Switch Pilot wire Ground wire I, , Primary delay circuit Blocking logic Trip element actuator Backup delay circuit a Circuit breaker Attenuator Machine frame 60-Hz GCM 21 GFR detector Logic device Band-pass filter Local fault sense circuit X Phase conductor GCM relay Pilot wire Ground wire Figure 38.-System block diagram. Phase conductor Sense and amplify Open-coil detection Filter Backup delay generator circuit T Logic for trip signal generation Adjust pickup level Comparator Adjust delay Primary delay generation circuit Local trip signal generator Logic for trip signal generation Trip element Downstream inhibit signal Figure 39.-Block diagram of relay circuit 22 activated instantaneously. Also, if an increase in imped- ance of the pilot- and ground-wire circuits is sensed by the GCM, an instantaneous trip is activated. LABORATORY EVALUATION Laboratory tests with a 12-V commercial GCM and the CF-GFR prototypes verified that both safety devices will function without interfering with each other. The operat- ing range of the GCM's, whose voltage is compatible with the GFR filters, is limited to 4,000 ft by the substantial impedance of No. 8 AWG pilot conductors. Nevertheless, this should not preclude the direct appli- cation of CF GFR's on most underground distribution circuits. CONCLUSIONS Sensitive and coordination-free ground-fault protec- tion has been designed for use on resistance-grounded mine power systems. To facilitate commercial manu- facture, the designs are fully documented by detailed schematics, assembly drawings, and component listings. Prototype units have been tested in the laboratory and are available for installation in underground mines. Implementation of this practical protection would not require extensive alterations to mine power systems. Existing GFR's would simply be replaced with solid-state units. The sensitive GFR's, when installed on ac and dc mine utilization circuits, can prevent nearly all the poten- tial electrocutions on these low-voltage power systems. CF GFR's, installed on ac distribution circuits in coal mines, can reduce the incidence of fires associated with high-voltage power systems by significantly decreasing response time to faults. REFERENCES 1. Oyier, A. Electrical Accidents in Mining (1980-85). Fatal and Nonfatal Accidents Underground and on the Surface at Underground Coal and Metal-Nonmetal Mines. BuMines IC 9259, 1990, in press. 2. Dalziel, C. F., and W. R. Lee. Reevaluation of Lethal Electric Current. IEEE Trans, and Ind. Gen. Appl., v. IGA-4, No. 5, 1968, p. 467. 3. Lee, R L. Electrical Safety in Industrial Plants. IEEE Trans. Ind. and Gen. Appl., v. IGA-7, No. 1, 1971, p. 11. 4. Geddes, L. A., and L. E. Baker. Response to Passage of Electric Current Through the Body. J. Assoc. Adv. Med. J - f rum., v. 5, No. 1, 1971, p. 233. 5. Knickerbocker, G. C. Fibrillating Parameters of Direct and Alternating 20 Hz Currents Separately and in Combination-An Experimental Study. IEEE Trans. Commun., v. COM-21, No. 9, 1973, p. 1017. 6. Bernstein, T. Safety Criteria for Intended or Expected Non- Lethal Electrical Shock. Proceedings of the First International Sympo- sium on Electrical Shock Safety Criteria, Toronto, Canada. Pergamon Press, Inc., 1985, p. 283. 7. Morley, L. A. Mine Power Systems (contract J0155009, PA State Univ.). Volume II. BuMines OFR 178(2)-82, 1982, p. 79; NTIS PB 83- 120386. 8. Wagner, C. F. Symmetrical Components. McGraw-Hill, 1933, 437 pp. 9. Morley, L. A., F. C. Trutt, and D. J. Rufft. Electrical-Shock Prevention (contract J0113009, PA State Univ.). Volume II-Ground- Fault Interrupting Devices. BuMines OFR 177(2)-83, 1982, 110 pp.; NTIS PB 84-102953. 10. U.S. Navy. U. S. Standard Requirements for Electrical Equip- ment. Mil. Stand. 454J, Sec. 1 and 36, Apr. 1984, pp. 1-1-1-12 and 36-1-36-2. 11. Cablec Corp. (Indianapolis, IN). Mining Cable Engineering Handbook. 1987, 168 pp. 12. Stanek, E. K., W. Vilcheck, and A. Kunjara. Mine Electrical Power Systems. Transients Protection, Reliability Investigation, and Safety Testing of Mine Electrical Power Systems (grant G0144137, WV Univ.). Volume I-Transients in Mine Electrical Power Systems. BuMines OFR 6(1)-81, 1979, 169 pp.; NTIS PB 81-166761. 13. Underwriters Laboratories (Melville, NY). Ground-Fault Circuit Interrupters. Stand. 943, 1980, 14 pp. 14. U.S. Code of Federal Regulations. Title 30-Mineral Resources; Chapter I-Mine Safety and Health Administration, Department of Labor; Subchapter 0-Coal Mine Safety and Health; Part 75-Mandatory Safety Standards-Underground Coal Mines, Subpart G-Trailing Cables; Sec. 75.601. July 1, 1988. 15. Yenchek, M. R, and M. N. Ackerman. Evaluation of Sensitive Ground-Fault Interrupters for Coal Mines. BuMines IC 9057, 1985, 15 pp. 16. Magnetics (Butler, PA). Design Manual Featuring Tape- Wound Cores. Publ. TWC-300R 1980, 28 pp. 17. Steen, F. L. Supertoroids With 'Zero' External Fields Made With Regressive Windings. Electron. Des., v. 18, Sept. 1976, p. 45. 18. Dolinar, K. D. Improved Ground-Fault Protection System for Low- and Medium-Voltage Trailing Cables. Conference Record of the 1980 IEEE-Industry Applications Society Annual Meeting, Cincinnati, OH. IEEE, 1980, p. 594. 19. Morley, L. A, F. C. Trutt, and T. Novak. Sensitive Ground-Fault Protection for Mines. Phase I-Alternating Current Utilization, (contract J0134025, PA State Univ.). BuMines OFR 26-85, 1984, 89 pp.; NTIS PB 85-185767. 20. Kohler, J. L., F. C. Trutt, and L. A. Morley. Sensitive Ground- Fault Protection for Mines. Phase II-Direct Current Utilization (contract J0134025, PA State Univ.). BuMines OFR 29-87, 1986, 81 pp.; NTIS PB 87-203410. 21. Trutt, F. C, H. G. Rotithor, and J. L. Kohler. A Coordination- Free Relay for AC Mine Distribution Systems. Conference Record of the 1988 IEEE-Industry Applications Society Annual Meeting, Pittsburgh, PA IEEE, 1988, p. 1727. 22. Trutt, F. G, J. L. Kohler, H. G. Rotithor, L. A. Morley, and T. Novak. Sensitive Ground-Fault Protection for Mines. Phase III— Coordination-Free GFR's for AC Distribution (contract J0134025, PA State Univ.). BuMines OFR 2-89, 1988, 95 pp.; NTIS PB 89-143382. 23 APPENDIX [.-ABBREVIATIONS AND SYMBOLS ac alternating current PW pilot wire AWG American wire gauge Q thyristor C capacitor QA operational amplifier CF-GFR coordination-free ground-fault relay R resistor CR relay Rt body resistance CT current transformer R s grounding resistor D diode R, inner toroidal radius dc direct current Rl load resistance E RMS output voltage of transformer secondary R outer toroidal radius f frequency R s source resistance F fuse RMS root-mean-square GCM ground-check monitor S switch GFR ground-fault relay SCR silicon-controlled rectifier (thyristor) h toroid core width SH switchhouse I current t time *p RMS value of ground-fault current ti opera^g time of GFR IC integrated circuit t 2 operating time of molded-case circuit breaker K electromagnetic relay T transformer L inductor TP test point LED light- emitting diode u core permeability MOV metal-oxide varistor U digital logic component N number of turns on secondary < winding UJT unijunction transistor N p primary turns v tt control voltage N s secondary turns v* dc voltage N.C. normally closed v ta line-to-neutral system voltage N.O. normally open z Zener diode INT.BU.OF MINES,PGH.,PA 29177 > z m O c > i- o "0 ~o O 3 m O < m 3) 413= 90 Jp*fc. ^"•7m*" <** **<& o. • _ . /\ #H' /% -jBR : /\ l %9) J*x -•' • d>*^ - =5^ 1 >N^5^^^5^T?^ * ^ ^ «4