Power Plant Horror Stories

of 18
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Generator protection
  Power Plant “Horror Stories” Charles J. MozinaBeckwith Electric Co., Inc.6190-118th Avenue NorthLargo, FL 33773-3724 U.S.A.(727) 544-2326cmozina@beckwithelectric.comAbstract: Contrary to popular belief, generators can fail—not only from short circuits—but more fre-quently from abnormal electrical conditions such as overexcitation, overvoltage, loss-of-field, unbal-anced currents, and abnormal frequency. When subjected to these abnormal conditions, damage orcomplete failure of the generator can occur within seconds. In many cases, these failures can be pre-vented by proper generator protection. This paper relates a number of “horror stories” within the powerplant in the hopes that the “lessons learned” will help others to avoid the cases described. Introduction Generators are the most expensive piece of equipment in a power system. The cost of a major generatorfailure to a utility or IPP (Independent Power Producer) owner is not only the cost of repair or replace-ment of the damaged machine, but also the substantial cost of purchasing replacement power when theunit is out of service. An alert and skillful operator, at manned locations, can sometimes avoid remov-ing a generator from service by correcting an abnormal condition. In the vast majority of cases, how-ever, the event will occur too rapidly for the operator to react and automatic detection and isolation isrequired. Operators have also been known to make errors and create abnormal conditions where trip-ping to avoid damage is required. Inadvertent energization and overexcitation are examples of suchevents.Several power plant events within the last two to three years substantiate the premise that generatorscan, and do, sustain internal short circuits and abnormal operating conditions that require tripping. Thefollowing in-service events are described:ãMulti-phase generator faultsãStator ground faultsãAccidental off-line generator energizingsãOverexcitationãLoss-of-fieldãGenerator breaker failure (breaker flashover)In many cases, human error caused or contributed to the event. These events were captured on oscillo-graphs. This paper highlights the subtlety of analyzing non-fault events such as loss-of-field usingCOMTRADE format to convert current and voltage to R-X quantities to verify proper relay operation.The lessons learned in each event are also highlighted.  Multi-Phase Generator Faults When a generator multi-phase fault is detected by generator differential relaying, it is separated fromthe power system by tripping the generator breaker, field breaker and prime mover. The system contri-bution to the fault will immediately be removed when the generator breaker trips as illustrated in Fig. 1.The generator current, however, will continue to flow after the trip. The generator short circuit currentcannot be “turned off” instantaneously because of the stored energy in the rotating machine. This flowof damaging generator fault current will continue for several seconds after the generator has beentripped, making generator faults extremely damaging. Generator terminal leads are usually isolatedthrough isophase bus construction to minimize multi-phase terminal faults. Fig. 1Generator Terminal Fault Current  Fig. 2 is an oscillograph of a three-phase fault which occurred on a gas turbine when a connector failedat the generator lead connection to the generator breaker. The fault started as a line-to-ground fault, butafter five cycles, it evolved into a three-phase fault. The system currents (I A , I B , I C ) were interruptedwhen the generator breaker was opened by differential (87G) relaying in about three cycles. The gen-erator-side current (I a , I b , I c ) continued to flow after the unit was shut down. The oscillograph wasprogrammed to cut off six cycles after tripping, thereby preventing the display of the total length of fault current flow which is estimated to have continued for eight seconds after tripping.This extended flow of fault current is the reason that internal multi-phase generator faults typically damagethe unit to the point where it cannot be economically repaired. There is no means of “turning off” thegenerator current. This long decay time results in the vast majority (about 85%) of the damage occurringafter tripping. This is why every effort is made in generator and generator terminal design to make the onlycredible fault a ground fault. The generator is then grounded so as to substantially reduce ground current tominimize damage. If the fault is in the GSU transformer and the generator installation has no low-voltagebreaker, the long fault current decay can substantially damage the transformer. A significant number of thesetransformers have failed catastrophically with tank ruptures and oil fires.                            Stator Ground Faults The method of stator grounding used in a generator installation determines the generator’s performanceduring ground fault conditions. If the generator is solidly grounded (not usually the case), it will deliver avery high current to a SLG (single-line-to-ground) fault at its terminals, accompanied by a 58% reduction inthe phase-to-phase voltages involving the faulted phase and a modest neutral voltage shift. If the generator isungrounded (also not usually the case), it will deliver a negligible amount of current to a bolted SLG fault atits terminals, accompanied by no reduction in the phase-to-phase terminal voltages and a full neutral voltageshift. These represent the extremes in generator grounding with normal practice falling predictably in between.The high magnitude of fault current which results from solidly grounding a generator is unacceptable be-cause of the fault damage it can cause. Shutting down the generator through tripping the generator breaker,field, and prime mover does not cause the fault current to immediately go to zero. The flux trapped in thefield will result in the fault current slowly decaying over several seconds after the generator is tripped—substantially exacerbating damage. On the other hand, operating an ungrounded generator provides negli-gible fault current, but the line-to-ground voltages on the unfaulted phases can rise during arcing type faultsto dangerously high levels which could cause the failure of generation insulation. As a result, stator windingson major generators are grounded in a manner that will reduce fault current and overvoltages yet provide ameans of detecting the ground fault condition quickly enough to prevent iron burning. Fig. 2Multi-Phase Generator Fault Oscillograph Relay TripBreaker Open  Almost all large generators that are unit-connected are high-impedance grounded. High-impedancegenerator neutral grounding utilizes a distribution transformer with a primary voltage rating greaterthan or equal to the line-to-neutral voltage rating of the generator and a secondary rating of 120 V or240 V. The distribution transformer should have sufficient overvoltage capability so that it does notsaturate on SLG faults with the machine operating at 105% of rated voltage. The secondary resistor isusually selected so that for a SLG fault at the terminals of the generator, the power dissipated in theresistor is approximately equal to the reactive volt-amperes in the zero-sequence capacitive reactanceof the generator windings, its leads, and the windings of any transformer(s) connected to the generatorterminals. Using this grounding method, a SLG fault is generally limited to 3-5 primary amperes. As aresult, this level of fault current is not sufficient to operate generator differential relays.Fig. 3 illustrates a typical unit-connected high-impedance grounded generator. The most widely usedprotective scheme in high-impedance grounded systems is a time-delayed overvoltage relay (59N)connected across the grounding resistor to sense zero-sequence voltages as shown in Fig. 3. The relayused for this function is designed to be sensitive to fundamental frequency voltage and insensitive tothird-harmonic and other zero-sequence harmonic voltages that are present at the generator neutral.Since the grounding impedance is large compared to the generator impedance and other impedances in thecircuit, the full phase-to-neutral voltage will be impressed across the grounding device for a phase-to-groundfault at the generator terminals. The voltage at the relay is a function of the distribution transformer ratio andthe location of the fault. The voltage will be a maximum for a terminal fault and decreases in magnitude asthe fault location moves from the generator terminals toward the neutral. Typically, the overvoltage relay hasa minimum pickup setting of approximately 5 V. With this setting and typical distribution transformer ratios,this scheme is capable of detecting faults to within approximately 5% of the stator neutral. Third harmonicschemes (not described in this paper) are typically used to detect faults near the generator neutral. G       Fig. 3Unit-Connected High Impedance-Grounded Generator 
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks