Condition monitoring of electrical equipment in nuclear power plants

  • Published on

  • View

  • Download


<ul><li><p>IEEE Transactions on Energy Conversion, Vol. EC-1, No. 3, September 1986 1</p><p>CONDITION MONITORING OF ELECTRICAL EQUIPMENTIN NUCLEAR POWER PLANTS</p><p>A. Sugarman, Member, IEEE</p><p>NUTECH Engineers, Inc.San Jose, California</p><p>INTRODUCTION transmitters, terminal blocks and molded casecircuit breakers, the cost of CM must be less than</p><p>Condition monitoring (CM) is a subset of main- the costs associated with the purchase and instal-tenance testing. It is a quantitative, predictive lation of a new part for CM to be feasible. How-technique for assessing the effects of all types of ever, the costs associated with the unexpectedaging (environmental, cyclic, operational, etc) on failure of the equipment from increased down timethe "health" of the equipment. A difference during a period of high demand must also bebetween CM and maintenance testing is that the considered.latter is neither quantitative (i.e., measures therelative condition of the component or material as Feasible approaches to CM for less costlyopposed to merely verifying that its condition is equipment may be to utilize statistical analysis ofacceptable) nor predictive (i.e., makes judgments failures and testing on a sample of the equipmenton the ability of the component to perform at a and efficient utilization of the data recorded infuture time). the maintenance histories. Statistical evaluation</p><p>of the rate and type of failures may be partic-A common example of the principle of CM can be ularly effective for evaluating the condition of</p><p>illustrated with the automobile which has a life- equipment which exist in large quantities in atime that is small enough to observe all the nuclear generating station.periods (break in, random failure, wear out) thatoccur throughout aging. There are several weak The maintenance interval for much equipment inlink components in the car (e.g., water hoses, a nuclear plant is commonly m 1-1/2 years. Over acontacts in the distributor, generator, spark plug 40 year plant design life, this means a maximum ofcables, solenoid, etc) which if they fail will less than 28 inspections (including an inspectioncause failure of the automobile to either start or when first installed) can be performed if CM isrun. From the day the car is put on the road and begun from the very beginning. Since the clock onis subjected to heat and vibration, significant the design life generally begins well before theaging of these components occurs. Degradation in plant first goes into commercial operation, 28the water hoses, for example is manifested by the inspections may be optimistic. To spot trends itelastomeric casing becoming brittle and cracking. is therefore important to conduct CM as frequentlyFailure will occur when it can no longer hold its as possible so that a data base can be establishedpressure and the cracks propagate through the for comparison purposes.casing wall causing the hose to burst. Theapplication of CM to the hoses would entail the For certain safety related equipment, themeasuring of the physical condition of the hoses surveillance requirements have been established to(e.g., the outside diameter, hardness, etc) and an maintain the performance and certain operatingevaluation of the trend in the data so that an parameters within a narrow range so that signifi-unexpected failure could be prevented. The data cant trends are not permitted to occur. There arecould be used in (a) optimizing the maintenance parameters, however, such as the rate of change inintervals to permit the most economical maintenance drift, trends in alert levels and alarm levels forschedule for the equipment and (b) evaluating the the equipment and the evaluation and trending ofcause of failure in the equipment so that more parameters useful in diagnosing the health of theeffective maintenance can be performed. The latter equipment which can be used for does, not rely on CM but solely on maintenanceprocedures. Another condition which is imposed on a CM</p><p>program in a nuclear plant is the human factorsThere are many conditions imposed on the consideration (e.g., provisions of habitability,</p><p>maintenance of equipment in nuclear power plants high level of technical expertise required to runthat influence what CM can be performed. For the test and ambient conditions of temperature,small, relatively less costly equipment such as radiation and humidity). Some CM tests may require</p><p>that equipment and/or parts be brought to alaboratory where testing or analysis be performed.</p><p>85 WM 245-6 A paper reconnended and approvedby the IEEE Nuclear Power Engineering Committee of OBJECTIVE AND APPROACHthe Power Engineering Society for presentation atthe IEEE/PES 1985 Winter Meeting, New York, New This paper is based on work that was supportedYork, February 3 - 8, 1985. Manuscript submitted by the Electric Power Research Institute (1). TheSeptember 11, 1984; made available for printing objective of this work was to identify and reviewJanuary 4, 1985. what CM techniques are applicable and/or being</p><p>performed for Class 1E equipment in nuclear powergenerating stations. This objective wasaccomplished by:</p><p>(a) Conducting preliminary manual and computerliterature searches on condition monitor-ing techniques for electrical equipment.These searches were primarily for test</p><p>0885-8969/86/0009-0001$O1 .00 1986 IEEE</p></li><li><p>2</p><p>techniques for evaluating the condition ofinsulation systems in representative Table 1equipment items because these systems aresusceptible to degradation and the CATEGORIZATION OF CLASS 1E EU(IPMENmaterials are representive of those in R)R CONDITION MONITORINGweak link components.</p><p>1. Equipment for Which Condition Monitoring May Not Be Feasible(b) Conducting a telephone survey of eight Teinal Blocks Signal Converters</p><p>nuclear generating stations and visitingfour of them to discuss the engineering Manual Switches Recordersand surveillance procedures performed by Amplifiers Metersthem for their Class 1E equipment.</p><p>2. Equipment for Which Condition Monitoring is FeasibleEight utilities (Duke, Boston Gas and Electric,</p><p>Commonwealth Edison, Washington Public Power Supply Batteries *SolenoidsSystems, Georgia Power, Southern California Edison, Batter Chargers RelayNortheast Utility and Tennessee Valley Authority)were selected for a telephone survey on the status Generators Limit Switchesof their surveillance and monitoring programs, what</p><p>Transformers *Circuit BreaJcerscondition monitoring is being performed on Class 1Eequipment and what links there are between the *Cables *Transmitterssurveillance programs and their environmentalqualification program. Four utilities (BG&amp;E, CECo, Motors (for Com-SCE and NU) were visited and detailed discussions pressors, punps,etc.with engineering and maintenance staff members wereconducted. Motor Starters</p><p>The utility inputs were sought because of their *Motor Operatorsexperience with the operation and maintenance of 3. Equipmnt for Which Sas Condition Monitoring May Be Feasibletheir equipment. During the discussions with theutilities, an attempt was made to categorize Class Penetrations1E equipment into three broad categories relating Power Supplies (Inverter, Rectifier)to the feasibility of doing condition monitoring.The categories are: equipment for which condition Teuperature Sensorsmonitoring may not be feasible, equipment for which Radiation Monitorcondition monitoring is feasible and equipment forwhich some condition monitoring may be feasible. * Selected for in-depth evaluation in Reference 1.The category with the largest number of items isthat for which CM is feasible (13 out of a total of23). applicability to CM. CM of equipment in the first</p><p>group would not be feasible because in many of theOther considerations for proper classification items, the functional components are electronic and</p><p>included the complexity of the equipment and what tend to fail suddenly and without warning.age-degradable components exist in the equipmentwhich if monitored would provide the status of the Amplifiers and signal converters (e.g., analogequipment. to digital, square root extractor, logarithmic to</p><p>linear, etc.) are assemblies of electronicOf the list of equipment for which condition components which are difficult to monitor for</p><p>monitoring was considered feasible, five items were degradation because they tend to fail withoutselected for an "in depth" evaluation. These items warning. Gross, drift measurement can be made andwere selected because in addition to being consid- if the received signal falls outside the allowableered good candidates for condition monitoring they band, removal and replacement of the proper circuitappear as components in many different items. The board can be performed. Monitoring the performanceitems selected were: cables, motor operators, of individual circuit boards or components (e.g.,solenoids (in valve actuators), circuit breakers transistors, resistors, capacitors, etc.) may notand transmitters. always be feasible in the field. This must</p><p>ordinarily be done by the manufacturer who has theDISCUSSION proper test equipment and who knows the performance</p><p>requirements of the test pieces.Not all equipment is equally suitable for being</p><p>monitored for age-related degradation. Electronic Some equipment falls into a category for whichequipment, for example, tends to fail suddenly CM is perceived to be feasible. It is out of thisrather than degrade in service. Large numbers of category that five items (denoted by asterisks)components are necessary to obtain statistically were selected for in-depth evaluations of CMsignificant failure data. Electronic equipment is techniques. Many of the CM techniques discussed intherefore impossible to monitor with the small this evaluation are applicable to other icems innumber of components used in a nuclear generating this group. For example the CM techniques for thestation. Other equipment (e.g., terminal blocks, insulation in the stator of the motor operator maymanual switches, and recorders) may not be suitable be used for generators, other motors and trains-for condition monitoring because the low cost of formers. Similarly, the performance tests andthe replacement parts may not make it feasible to insulation tests on circuit breakers are applicabledevote the effort required for condition to monitoring relays.monitoring.</p><p>Some equipment falls into a category in whichIn Table 1, the equipment types are categorized CM may be feasible and cannot be placed in either</p><p>into three groups according to the perceived of the first two categories. For example the</p></li><li><p>3</p><p>detector in radiation monitors may operate with When the reading exceeds the limits of the innerhigh voltage and leakage or dielectric breakdown band, the instrument is recalibrated and when thismay occur because of degradation in the insula- happens several times, the instrument is repairedtion. On the other hand it is in the electronics or replaced.(preamplifier, amplifier and counter) where failurewould ordinarily be expected. Another plant reported that whenever an</p><p>instrument drifts beyond a predetermined amount,Non-destructive (NDE) techniques for monitoring the surveillance interval is shortened.</p><p>age-related degradation are often desirable becausethey are fast and easy to perform and may often be Most plants are in the process of bringing onperformed in-situ on the equipment. In general, line computerized maintenance management systems.destructive (DE) techniques that measure the A demonstration was given at one plant in whichchanges in mechanical properties (e.g., elasticity, multicolored graphics were used to identify thetensile strength) are commonly used in monitoring maintenance history of their equipment. Eightthermal degradation. NDE techniques which measure thousand files out of 100,000 had been entered intodielectric loss have had some success. An overview the system. The data base will eventually containof potential condition monitoring techniques which data on 40,000 components identified by the A/E.will be discussed in more detail in the nextsection are given in Table 2. The computerized data bases being brought on</p><p>line can be a valuable tool in implementing CM.They are able to store large quantities of testdata over long periods of time, facilitating the</p><p>Table 2 records keeping and data processing for trending.</p><p>POTENTIAL CONDITION MONITORING TECHNIQUES CM Techniques for Motor Operators</p><p>CORRELATION TYPE OF Motor Operated Values (MOVs) can be monitoredPRtOPERTY WITH AGE MEASUREMENT by testing the insulation with ac tests or dc tests</p><p>and performance tests. Example of the use of someof these tests in monitoring degradation in equip-</p><p>o Equipment Performnce Low NDE ment will be discussed below. A discussion of theo Electrical Lose in Insulation Moderate NDE theory and other potential CM techniques is giveno Electrical Resistance of in Reference 1.</p><p>Insulation Low HDEo Excitation Current in Kadotani et al. (2) examined the breakdown voltage</p><p>Transformrs ?ME in individual coils for the windings in twenty-nineo Cable Tests (Bridge Methods, 3 kV motors which had operating periods of 15 to 16</p><p>Radar, Tracer Tests) LOW DME years. By examining 10 to 12 coils in each of theo Visual Inepection Low HOE machines, they found that a correlation existedo Mechanical Tests of Insulation High DE between the age of the machine and the breakdown</p><p>.o Thermal Analyses of Insulation: TGA Possible DE voltage (average and minimum) which is shown inOTA ? DE Figure 1. Two of the machines, A and B, showed</p><p>o Insulation Hardness HNDE severe degradation. The breakdown voltage (BDV) ofo Hig Potential Test LOw DE severely degraded windings from two machines are</p><p>also shown in the ficure. B similar correlationbetween the ratio B with working years</p><p>A.verage BDVfor the motors was aso8 ta 1.ned.</p><p>HOE Mon Destructive Evaluation CKadotani et al (2) divided the insulationDE Destructive Evaluation measurements they evaluated into three</p><p>categories. They are:TGA Dfereogrntitric Analysis0Th Differential Thermal analysis (a) Measurements which relate the condition of</p><p>the insulation (e.g., degree of polymeri-zation, carbonization, impurity content,</p><p>The results of the utility survey showed that moisture absorption). The measurementsthe maintenance procedures used in the plants were are; dissipation factor (Tan u), capaci-generally obtained from the manufacturer's recom- tance at low voltage and insulationmendations. Some plants stated that plant exper- resistance.ience, industry experience, 'gut feeling" and E.Q.inputs were major considerations in the testing (b) Measurements which relate total voidprocedures. content in insulation. The measurements</p><p>are; A Tan 6 (the difference betweenThree plants reported that they will be com- Tan 6 at the rated voltage and Tan 6 at a</p><p>mitted to, or already do, post-mortem examinations low voltage), Ac/c0 (difference betweenon failed equipment such as solenoid valves. Four capacitance at a high voltage andplants reported that they use the Nuclear Plant capacitance at a low voltage,...</p></li></ul>


View more >