Srrel | sce. Topics: EPRI CS/EL-4415 ESEARGH INSTITUTE Power system planning Volume 3 Operations research Projects 1681, 2153 On-line measurement systems. Interim Report Heat rate February 1986 Economic analysis Automatic control Power Plant Performance Monitoring and Improvement Volume 3: Power Plant Performance Instrumentation Systems Prepared by Potomac Electric Power Company Alexandria, VirginiaSUBJECTS TOPICS AUDIENCE BACKGROUND OBJECTIVES ‘APPROACH RESULTS REPORT SUMMARY Boilers and related auxiliaries / Plant systems and performance assessment / Power system planning / Power system operation Power system planning Heat rate Operations research Economic analysis On-line measurement systems Automatic control Generation managers / R&D engineers Power Plant Performance Monitoring and Improvement Volumes 1-3 To optimize power system planning, utilities need advanced instrumentation for analyzing plant performance. This report presents the initial results in a study to develop a total plant- monitoring system. Included are guidelines to determine com- bustion parameters, models to calculate the effect of deviations from design values on heat rate, and techniques to diagnose turbine degradation. Increasing fuel costs make it economically attractive to develop new technol- ‘gies to improve power plant performance and increase the efficiency of distribution networks. In 1982 EPRI started work to develop a monitoring system that would provide utility engineering, operations, and power-supply staffs with continuous detailed analyses of plant performance. Potomac Electric Power Company (PEPCO) is the host utility. + To develop advanced instrumentation and monitoring systems for Increasing heat rate and plant availability + To demonstrate how such systems can be used to optimize power system planning and power dispatch. ‘The project team is developing instrumentation and techniques to monitor and analyze plant and systems performance at the PEPCO Morgantown Unit 2 in Charles County, Maryland. Boiler-related studies are focusing on ‘excess-air minimization, back-end optimization, and power cycle testing. Systems studies are developing models to analyze heat rate sensitivity and algorithms to improve economic dispatch and unit commitment. The researchers are testing all sensors, monitoring devices, and computer hard: ware and software developed during the project at the utility's Pennsyivania— New Jersey-Maryland systems control center in Rockville, Maryland. This three-volume report describes the initial results of the project. Volume 1, on boiler optimization, discusses procedures to determine the optimum fire- side operating conditions in a coal-fired power plant and presents a com- puter program to optimize the effects of combustion parameters—such as EPRI CSIELA4tSs Vols. 1-3EPRI PERSPECTIVE PROJECTS: coal fineness, excess-air volume, and exit gas temperature—on heat rate. This volume also describes instrumentation and methods to elimi nate fuetair imbalances between burners, monitor tube wastage rates, assess the performance of regenerative air preheaters, calculate unit heat rate, and measure gas temperature, combustion products, and wall heat flux in the furnace region. Volume 2, on incremental heat rate sensitivity, presents a simulator code for determining the impact of changes in plant-operating states on fuel ‘consumption and system dispatch. It also describes studies to deter- mine how modeling errors affect incremental heat rate curves. Volume 3, on turbine cycle and instrumentation, presents techniques to measure the leakage flow of critical internal shaft (N2) packing from high-pressure to intermediate-pressure turbines and to determine control valve position, without the help of transducers, in very hot environments. ‘The other instrumentation described includes devices to monitor con: denser back pressure, to improve watthour metering systems, and to recalibrate condensate flow nozzles. A computerized data acquisition and analysis system is also presented. The techniques for monitoring and analyzing plant performance described in this report should help uttlites diagnose heat rate degrada- tlon in coal-fired power plants. The next phase of the project will explore advanced systems for continuous measurement of instantaneous coal flow and heating value, as well as on-line heat rate determination for system dispatch. Ultimately, a total plant-performance-monitoring system for improving heat rate, recovering unit capacity, and developing more effective maintenance programs will be devised. As the project, continues, workshops and video tapes will be produced to facilitate tech- nology transfer. Commercialization of specific products will also be studied. P1681, RPZI53 EPRI Project Managers: F. K. L. Wong; J. W. Lamont Coal Combustion Systems Division; Electrical Systems Division Contractor: Potomac Electric Power Company For further information on EPRI research programs, call EPRI Technical Information Specialists (415) 855-241Power Plant Performance Monitoring and Improvement Volume 3: Power Plant Performance Instrumentation ‘Systems CSI/EL-4415, Volume 3 Research Projects 1681, 2153 Interim Report, February 1986 Prepared by POTOMAC ELECTRIC POWER COMPANY 1400 North Royal Street Alexandria, Virginia 22314 Principal Investigators HG. Crim) J.C. Westcott R. W. de Mello, R. E. Brandon D. W. Parkinson J. 8. Czuba Prepared for Electric Power Research institute 3412 Hillview Avenue Palo Alto, California 94304 EPRI Project Managers E.K.L. Wong Performance and Advanced Technology Program Coal Combustion Systems Division J. Lamont Power System Planning and Operations Program Electrical Systems DivisionORDERING INFORMATION Requests for copies of this report should be directed to Research Reports Center (RRO), Box 50490, Palo Alto, CA 94303, (416) 965-4081. There is no charge for reports requested by EPRI member utilities and affiliates, U.S. utility associations, U.S. government. agencies (Tederal, state, and local), media, and foreign organizations with which EPRI has an information exchange agreement. On request, RAC will send a catalog of EPRI reports. ‘Cooyight © 1986 Electric ower Research Insite, Ine A gts reserved NoTICE “Ths roprt was prepare he ciganzations) named below a an account of wrk sponsored the Electric Power eacoren at, ne (EPRI. Net: EPA mambers of PAL, th organizations) med low, nr Any arson ating on bera fay of hem: (a) makes any warranty. expres or Imoiea wih reapect othe uso of ry Ssormation apparatus, method or process cioad inh report or hat uch ue may nt intinge ately ‘wnea rights: or) assumes any labibes wih espect tothe ue oor damages resting ar he ve {any mioration apparatus, meted or process dsloeesin ts rep Prepared Potomac Elects Powe: Company ‘Novanaia, VgaABSTRACT PEPCO's Morgantown Unit 2 and the PIM system control center are serving as the test facilities for this project. This first phase of the project utilizes currently (or s00n to be) available instrumentation for monitoring and analyzing plant and system performance on a continuous basis. The overall approach is to demonstrate in one facility all sensors, monitoring devices, and necessary computer hardware and software for on-line performance monitoring and dispatch purposes. Significant developments include turbine packing leakage measurenent, condenser back-pressure measurement, power cycle testing, and stuiies of the application of advanced instrumentation to system dispatch.EPRI FOREWORD Utilities have deferred the construction of new generating units because of lower ‘than expected load growth and high interest rates. The burden of meeting load demand, therefore, falls heavily on existing plants. To reduce fuel costs in these plants and to comply with government regulations, utilities are implementing measures to improve heat rate performance. In a recent EPRI survey, participating utilities estimated that an average improvement of more than 400 Btu/Kiwh in heat rate was possible. In many stations, however, efforts to achieve optimum efficiency and availability are hampered by the lack of instrumentation and techniques for accurate performance measurements. ‘This project, which started in October 1982, focuses on advanced instrumentation systens to monitor and analyze plant performance. ‘To help utilities improve unit heat rate, uit availability, and the economic dispatch of power, researchers are designing a monitoring system to provide on-line reporting of plant performance for immediate use in the dispatch center and plant control roam. They are also devising detailed procedures to enable utility engineering and power-supply staffs to use these data for long-term analysis. The Potomac Electric Power Company (PEPCO) is the host utility and prime contractor. Studies to evaluate all monitoring equipment and operating strategies, which are under periodic review by a project advisory committee composed of experienced performance engineers from 21 utilities, are being conducted at PEPOO Morgantown Unit 2, a 575-MW coal-fired plant in Charles County, Maryland. The work plan was drafted by EPRI staff and reviewed by the advisory committee.This three-volume report describes the scope and initial results of the project. Volume 1, Boiler Optimization, presents the instrumentation and the computer codes that are being developed to establish the relative contributions of coal fineness, excess air, and exit gas temperature to fuel and power costs. Volume 2, Incremental Heat Rate Sensitivity Analysis, describes the economic impacts of errors mde in measuring incremental heat rate on power dispatch. Volume 3, Power Plant Performance Instrumentation Systems, presents diagnostic techniques and instrunen- tation for establishing the efficiency of the turbine cycle. ‘The monitoring devices and analysis tectmiques described in these volumes can help utilities improve heat rate performance, recover unit capacity, optimize system @ispatch, and improve plant maintenance. The tools presented incluie procedures to evaluate the thermodynamic and economic effects of plant operating parameters on holler combustion efficiency, a critical internal shaft-packing (N2) leak monitor to measure the amount of steam leakage from the high-pressure to the intermediate pressure turbine, and an incremental heat rate sensitivity model to determine the performance ani cost impacts of deviations in key plant parameters from design values. Future work during this project will inclue developing advanced systens to measure continuously coal flow and heating value, determine the effects of unburned coal and coal-grind size in boiler combustion, and simulate optimal power flow dispatch. Frank Wong, Project Manager Coal Combustion Systems Division doi Lamont, Project Manager Electrical Systems DivisionSection conreNrs: INTRODUCTION Summary References WATTHOUR METERING SYSTEM original systen New Watthour Metering system Instrument ‘Transformers References ‘TURBINE EXHAUST PRESSURE MEASUREMENT Exhaust Pressure Measurement References PRESSURE RATIO MONITOR Description of systen Description of the Pressure Ratio System IMPLEMENTATION OF THE EXHAUST PRESSURE MONITOR, PRESSURE RATIO MONITOR AND WATTHOUR METER INTEREACE system Description Software Functional Description Feature Overview Structured Application Environment Object Oriented Architecture Nucleus 1/0 system 55 55 56 56 57Section 5 Task Organization Message Task Pressure Ratio Task Exhaust Pressure Task Analog Input ‘Task Maintenance Task 6 SYSTEM POR MEASUREMENT OF INTERNAL PACKING (2) LEAKAGE N2 Packing Flow Characteristics Discussion calculation of Py Test Procedure References 7 COMPUTER SYSTEM REQUIREMENTS Real-Time Cnputers Accuracy Requirenents Calibration and Scanning Requirements ‘Thermodynamic Properties of Steam and Water ‘Time-Sharing Conputer Data Storage and Retrieval summary 8 "TURBINE PERFORVANCE DIAGNOSTIC PROCEDURES Goals of Diagnostic Testing Onit overhaul Incremental Heat Rate viii 5-10 61 6-7 6-7 6-9 6-1 ell 12 14 15 1-6 mu 73 e1 a1 82 828 operator Controllables 83 Steam Temperature e3 Turbine Exhaust Pressure as Reguirenents for Canponent Condition Determination ea Flow Relationships 86 Historical Plots e-12 Shape Plots eas Prediction of Internal Conditions eis Example: Excessive Leakage Caused by Shaft Robbing a-18 Example: Solid Particle Erosion e-19 Example: Internal Danage 8-20 Example: Deposits e-22 Determining the Seriousness of Internal Conditions 28 Exated Conditions 8-26 Excessive Leakage 2-27 Internal Danage a-32 Deposits o34 Changing Historical Trends 234 sumary 8-35 9 conctustoNs otSection ‘APPENDIX A WAT'THOUR METERING SYSTEM Objectives ‘System Description Functional Specifications Hardware Description Potential Input Current Input Pulse Initiator outputs Analog Outputs Calibration Input Front Panel Individual Phase Accumulators Gross Electrical Output Pulse Accumulator Operate/Test Switch Reset Switch ‘Timing Interval Complete Indicator Power On Indicator Calibration ‘Technical Development of DWHM System Compliance to Functional Guidelines Reliability Stress Factor Printed-circuit Boards Integrated Circuits Sensitivity to Interference Page aL a2 a2 AS aT BT BT aT ae As ato alo All all all ael2 a2 a-l2 aa a3 als AldSection APPENDIX A Maintainability Camponents or Parts Inventory Level of Essential Parts Spare Parts Provisioning Spare Parts Lists Simplicity of Design Identification Extension Boards Maintenance Maintenance Instructions Manual ~ Normal Content Maintenance Instruction Manual - Special Contents ‘Compl iance Homan Error Climatic Conditions and Bnvironmental Control Requirements Minimum Rules of Thermal Design Equipment Power Supply AC Power Supply Required Information Grounding Equipment Safety References Page aa als ais ale ais a7 ae als alg a-20 20 aaa RD a2 23 4-23 a23 a24 a2 aA-248-2 83 a4 as, TLLOSTRATIONS New Watthour Meter Basket Tip ‘Tap Selector Schematic of Tap Selection Valve Stroke vs. First Stage Pressure First Stage Used nergy vs. Valve Stroke ‘Throttle Flow vs. /P/W at the First Stage ‘System Components System Tasks Message Task Flow Chart Pressure Ratio Flow Chart Exhaust Pressure Flow Chart Analog Input Flow Chart N2 Packing Measurenent System ‘Schematic of N2 Packing Leakage Flow Characteristics Pressure Drops Fran Packing Ring to Pressure Tap Typical Stage Flow Relationship for the First Stage Area vs. Throttle Flow for Conditions at Turbine Stop Valve First Stage AREA/FLOW Relationship at Full Load First Stage Pressure Hypothetical Plot of Reheat Bowl Conditions xiii, 46 5-4 58 5-1 s12 513 S14 2 6-5 6-8 6-10 as 89 8-10 en 813Figure 8-7 8-8 8-9 8-10 a1 a-12 8-13 ald Hypothetical Plot of IP Efficiency High Pressure Section Efficiency Effects of Shaft Rubbing typical Results of Solid Particle Erosion Damage Effects of Internal Damage Effects of Steam Path Deposits Approximate Effect of Nozzle Area Change on Stage Characteristics Heat Balance IP Section Apparent Efficiency Front Panel Input Leads Pulse Initiator Output Termination Analog Output Termination 8-19 821 8-23 8-25 8-28 831 As‘This project was initiated in response to escalating fuel costs and the resultant increasing attention to thermal performance by the industry and the Regulatory PROJECT OVERVIEW authorities in particular. ‘The project's overall goal is to develop a "state of the art" performance moni tor- ing demonstration systen on a host utility unit where advanced measurenent devices and sensor instrumentation can be implemented with rithms which are computed and displayed on distributed, modular, camputer systems. The focus of the program is to provide the means to measure, calculated and dis- play on-line fossil power plant heat rate/performance as necessary to maximize im pact on the following: Operator controllable parameters ‘Thermal performance optimization of the boiler and its auxiliaries Turbine cycle performance optimization Incremental heat rate determination for accurate electrical system dispatch and costing of generation. Cammunication optimization between plants and dispatch Control Center. Documentation of official utility heat rate and fuel accounting record. Indication of maintenance needs and spare part requirements. Inpat to maintenance cost-benefit and scheduling. OL innovative performance algo-PROJECT SCOPE ‘The program detailed by the specific tasks in the Statement of Work is designed to accanplish the following: © Implement the cycle and boiler related instrumentation that is both necessary and available to measure the parameters required to ‘thoroughly quantify total unit heat rate and component contribution to it. © Establish a computer based performance system to read-out and util- ize the instrumentation installed. The system is building on that currently existing on the host unit. The system emphasizes accuracy, repeatability, and practicality. © The system design incorporates the peripherals, programs and a structure to accammolate software develomment’ (performance algo- rithm calculations) and testing of both the software and the in- strumentation upon which it is based. © The system is also designed to accammadate anticipated future con tract tasks including dynamic analysis for control optimization and heat rate penalty, operator guidance techniques and/or any other performance related power plant study which would require a fully instrumented test bed for its evaluation. © The boiler portion of the plant is receiving the same rigorous monitoring and analysis as the turbine cycle. In addition, the boiler system is being subjected to an in depth analysis of all the parameters which currently limit unit heat rate including physical, thermadynamic, econanic, environmental and industry tradition. ‘The most tangible of these limits will be extended by closer monitoring and thus a clearer understanding of their cause and effect such that lower heat rates can be economically reached and the methods of maintaining them can be established. A major goal is the op timization of fireside parameters such as excess air and exit gas temperature. © Unit dispatch based on on-line determination of incremental heat rate will be achieved by application of the heat rate characteristics determined by the system, The mechanics of ac- cumulating the proper heat rate data at the plant, camunicating it to the control center and the control center implenentation method will be the products of this effort. On-line determination and use of incremental heat rate for dispatch will be the focal point around which all other plant/control center camunication improvement efforts will revolve. © Tests under controlled conditions designed to provide specific Project needs are being conducted as necessary. The primary test and utilization of the system will hovever result fram the con- ‘tinuous operation of the data collection, storage and calculation system and the ongoing review, analysis and improvement of its results by the Potamac Electric Power Canpany project team and their respective contractor/consul tants.© ALI software developed during the course of this project will be done in conformance with ANSI FORTRAN 77 standards. ‘The developed codes will be on-line process canputer subroutines or modules that can be integrated with each other and with standard calling routines and will be complete with full instructions on how to handle input and output. Every reasonable attempt will be made to make the software as transportable as possible. PROJECT TEAM In early 1962 several utilities responded to EPRI's request for proposal. In July of 1982, Potanac Electric Power Company was selected as the host utility and proj- fect manager. Initial contract work began in October of 1982 and the current con tract runs through March of 1986. ‘The project team consists of the Potanac Electric Power Company personnel drawn fran their Central Diagnostic Team, their Process Camputer Section, the Morgantown Station Performance Engineers and the Dispatch Control Center engineering staff. ‘The primary subcontractors to Potomac Electric Power Canpany on the project are: Power Technologies, Incorporated in Schenectady, New York, relating to turbine cycle analysis, software develoment, instrumentation selection, data acqisition system develoment and dispatch systen studies: and tehigh University in Bethlehem, Pennsylvania relative to boiler optimization. Combustion Engineering in Windsor, Connecticut will be a critical source of boiler design and modeling information and consultant to the project relative to fireside and heat transfer analysis. 0-3General Electric Company's Turbine Performance Engineering Department is providing a critical review of turbine cycle test instrumentation, methods and procedures as well as turbine design information, Many other organizations are supplying equipment and services over the life of the project. Host UNIT DESCRIPTION The host unit, Morgantown No.2, is a coal fired, supercritical pressure (3500 PSIG, 1000%%/1000°F) Canbustion Engineering boiler, with a tandem canpound General Electric turbine rated at 575MM. The design includes a tjungstran air heater and an electrostatic precipitator. The unit uses eastern bituminous coals with approximately 6 percent moisture, 14 percent ash and 2 percent sulfur and a heating value of about 12,000 BTU/pound as fired, ‘The plant receives shipments of a large mmber of coals, sane of which are burned separately and others as mixtures. ‘The plant site is on U.S. Route 301 at the Potanac River in Charles County, Maryland approximately 50 miles south of Washington, D. ‘the plant consists of two very similar units which were placed in service in 1970 and 1971 respectively. Original design included a process monitoring canputer with a full set of performance calculations and associated instrumentation. the system accuracy level is above that of nomal plant instruments but less than ASME acceptance level. This existing system is being expanded and upgraded to the cur- rent "state of the art" practice.The project goals dealing with incremental dispatch and plant/control center cam munication will be implemented at the Potanac Blectric Power Company Dispatch Control Center located northwest of Washington, D.C. near Rockville, Maryland. ‘the Control Center consists of two Xerox Data Systens Signa IX computers with ex- panded active memory and data acquisition and control equipment housed in a spe~ cially designed building with full environmental and electrical power conditioning. ‘The Control Center was placed in service in 1976. The autanatic generation control system was separately designed and implemented on the Control Center camputers by Rockwell International, the original system vendor. Tt was placed in service in 1983, ‘The generation control system accepts econamic and regulation signals fran the Pennsylvania-New Jersey-Maryland Interconnection (PIM) and develops individual generating unit desired generation target values which are Aigitally transmitted to the generating stations. Individual unit target values are constrained by Limit values telemetered fran the generating stations. INETIAL PROJECT RESULTS Boiler Program - Exercise of a parametric moiel developed to characterize the boiler systen has indicated the relationships between the various system parane- ters and includes a furnace model by the boiler manufacturer. Field tests are being conducted to verify the moiel. Initial analysis and field test indicate a close coupling between excess air, unburned carbon and grind size of the pul~ verized coal. These results are influencing project emphasis. 0-5Inctemental Heat Rate - Madeling of the PEPCO dispatch system has characterized the sensitivity of production cost to performance measurement accuracy and pro- vides the project with guidance as to how accurately performance must be measured. Randan errors in the measurement of performance have a significant impact on sys~ tem operating cost at levels above two percent of rated input. Performance measurement bias has a significant impact on system operating cost at levels above one percent of rated input. The model also indicates the penalty imposed by the convexity requirement for positive sloping incremental curve shape which intro- duces a significant mojeling error into the input/output curve. This in turn re- sults in errors in dispatch, increases in production costs, and justifies research on alternative dispatch algorithns. Turbine cycle and Instrumentation - Initial effort in this area has concentrated on specification and procurement of the instrumentation, data acquisition and com Putational equipment to camplete the demonstration facility including data storage, retrieval and display capability. Systens received to date include: watthour metering, turbine exhaust pressure tap sequencing, turbine pressure ratio transmitter, turbine N2 packing flow metering, stack oxygen, duct velocity, air and flue gas humidity and flue gas sulfur trioxide. Computational system development has included implementation of packaged systems fort Fortran 77, statistics, graphics, cycle analysis and data base management. Project software development has included data storage and retrieval and standar- dization of existing performance calculations.Baseline tests of the unit have been conducted and analyzed to provide initial baseline condition for monitoring throughout the project life and to indicate areas of instrumentation need. REPORTS The first project anaual reports are currently being published and consist of a three Volune set as follows: Volume 1: Boiler Optimization 2: Incremental Heat Rate Sensitivity Analysis 3: Power Plant Performance Instrumentation Systems 0-7Section 1 INTRODUCTION ‘The monitoring of the performance of power plants is again becaming a high prior~ ity task for utilities. During the 1950's and part of the 1960's there was con siderable interest in the major power plant canponents performance canpared to guarantee. ‘These included the stean generator, steam turbine and condenser. Rarely was there mich concern about the performance of fans, pumps and heaters un- tess they Limited the maximum output capability of the unit. Manufacturers and sane utilities developed good instrunentation and capable engineering staffs for the sole purpose of determining the pover plants canponents performance. Mest of the peocedures and techniques were documented in the ASME Performance Test Codes. ‘There has been Little if any improvement in the accuracy of the results of these tests since 1960. Kratz and Westcott in a paper [1] presented at the American Poser Conference in 1955 stated the level of accuracy for turbine tests using manually read data, the only major improvement in accuracy since then is the ability to get a data sample which better represents the condition of a unit dur~ ing as steady state operation as is possible. A recent EPRI report [2] has reviewed what is considered camon performance monitoring practice by utilities. one conclusion wast "pespite a tenfold increase in fuel costs, only a few utilities have any significant effort under way to improve the measurement of plant performance". alThe use of real time canputers in power plants started in 1959. the major ap- plication of these systems was for data logging and alarm functions. Performance monitoring was considered to be a natural application. A survey by EPRI [3] was made in 1977 to ascertain the usefulness and acceptance of these systems in power Plants. The following is a quote concerning their use for performance monitoring. “the plant computer systems now operating were judged by almost all respondents to be useful. ‘The respondents identified the following functions as being the most important and helpful to the operator: im- Proved process monitoring, alarming and logging; better operational data; Performance calculations; and trending capability. However, there are problems of which the most often cited include inadequacies in the man/machine interface; performance calculation and turbine startup/control. ‘The last two are functions mast often discarded." However there were sane exceptions. King [4], in a paper presented to an EBI task force in 1965 and at the American Power Conference in 1966 presented the implemen— tation and results of a power plant computer and instrumentation system which was used to conduct an acceptance test on a turbine generator unit after startup in 1964. The paper stated: “Bor the first time, a turbine acceptance test which meets the require ments of the ASME Performance Test Code has been conducted on-line by a digital camputer system. There are many interesting results of this series of tests, sane of which will be discussed here. The acceptance test consisted of a series of 18 tests run at three dif— ferent valve points. The heat rate repeatability of 17 of these tests was within plus or minus 10 BTU/KWH. The heat rate of the other test strayed to a 15 BTU/WWH deviation." The results were considerably better than was stated in the paper. A total of nine tests at the valves-wide-open condition were run during an eight day period. ‘The standard error was only 2 BTU/KWH. 1-2After the leveling of technological advances in power plant canponents in the lat- ter 1960'S, utilities seemed to lose interest in performance monitoring. this loss of interest was reinforced by low fuel costs and the fact that fuel costs were not increasing at the same rate as other costs necessary for the production and delivery of electric energy. ‘the rapid escalation in fuel starting in 1973, generated considerable renewed interest in performance monitoring. Potanac Blectric Pover Company's Morgantom Unit 2 (and also Unit 1) was built with a computer and instrumentation system which had performance moni toring capabilities better than the average at that time. The system was designed so that periadic performance data could be obtained that would be useful in evaluat- ing conponent perfomance changes. Tt was not designed to cbtain data with the accuracy required for acceptance testing. ‘The first phase of this project relative to the unit and its monitoring system is designed to: Upgrade the present instrumentation to an improved accuracy level necessary for evaluation of new performance monitoring procedures « © Utilize new instrumentation techniques to refine the information necessary for predicting both cause and effect of equipment perfor- mance changes. Significant accomplishments of the first phase include: © Development of a technique to measure the critical internal shaft packing (camonly called the N2 packing) leakage flow. © Development of a technique to determine control valve position without the problems associated with position transducers operating in a hostile (very hot) environment. © Upgrading of instrunentation including a new watthour metering system, recalibration of the existing condensate flow nozzle and planning the improvenents in instrumentation accuracy necessary for the projected tasks in the next phases of the project.SUMMARY Section 2 of this report discusses the need for accuracy in measuring generation and describes the watthour meter that was chasen for the project. The specifica- ‘tion of the watthour metering systen is reproduced in Appendix A. The methad of measuring turbine exhaust pressure is discussed in Section 3, An accurate means of determining control valve position by using the ratio of first stage pressure tomain steam pressure is described in Section 4. Section 5 describes the hard- ware and software of the micro-campater system used to receive the output of the new watthour meter, and implement the exhaust pressure monitor and pressure ratio monitor previously described. Steam leakage in the packing between the high pressure and intermediate pressure turbines has the effect of reducing output and increasing heat rate. Tt also cat Plicates the calculation of efficiency of the intermediate pressure turbine. Section 6 of this report describes a system of measuring the steam leaking through this packing. Implementation of the system and its use to measure leakage flow are future tasks of this project. Requirements of conputers in power plants are quite varied. The canputer require ments of this project, and the means taken to satisfy those requirenents, are described in Section 7. One of the basic goals of this project is to develop procedures for monitoring Power plant canponent performance in a way which allows an analytical evaluation ‘of the camponents' separable characteristics, ‘The steam turbine's various sec- tions will be given considerable emphasis. It is necessary to be able to predict areas of performance degradation so that better maintenance scheduling and cost effectiveness can be attained. Section 8 is a discussion of a procedure used for turbine performance diagnosis. As the continuing phases of this program are implemented these procedures will be expanded and refined. 1-4Section 9 draws conclusions from the initial efforts of this project. REFERENCES ry B.M, Kratz, J.C, Westcott, “Procedures for Testing Large Steam ‘Turbine-Generators in Central Stations," Transactions of the ASME, Volume 79, No. 2 (CURRENT FOSSIL FUEL POWER PLANT PERFORMANCE MONITORING, EPRI #1,-3339, Voumel, ef OSOSstC—CSsSsSsSM DOCUMENTATION OF UTILITY EXPERIENCE WITH PROCESS COMPUTERS IN POWER PLANTS, EPRI NP-290, 1977 G.L.H. King, "High Accuracy Turbine Performance Test Application to On-Line Digital Process Camputer," EEI Task Force on Computer Applications to Utility Engineering, Boston, 1965 15Section 2 WAPTHOUR METERING SYSTEM ‘The measurement of electrical energy is critical to the accurate and repeatable monitoring of the turbine-cycle heat rate, low pressure turbine efficiency, com Genser performance and heat input to the steam generator calculated by the ab- solute heat loss method. It also has a direct influence on the determination of increnental heat rate if it is used to determine net output. An error of 14 in the measurement of gross generation will result in an error of 1 in gross turbine eat rate, The only other measurement having a similar effect on gross turbine heat rate is that of the primary flow: condensate flow or feedwater flow. ‘The importance of power measurement is well recognized and PIC-6 [1] states that "A single-phase watthour meter or indicating wattmeter of the 0.25 percent ac A wat- curacy class shall be used in each phase to determine the power output. thour meter per phase avoids errors introduced by unbalanced phase power or by operation at non-unity power factor since separate calibration of watthour meters, potential transformers and current transformers permits correction factors to be developed and applied per-phase. In recognition of the need for an accurate power measurement a new watthour meter- ing system was implemented and PEPCO agreed to make the B phase current trans- former available for the new watthour metering system. Current transformers on A and C phases and potential transformers on all three phases are presently used to measure generation. The new watthour meter will provide the measurement of electrical generation to performance calculations developed for the project. The 241existing watthour meter, with its Links to the dispatch center, will be retained. ORIGINAL SYSTEM ‘The original watthour metering system on Morgantown unit 2 is a normal rotating polyphase watthour meter utilizing a cycloneter for visual readings and a pulse initiator output to the plant computer. This watthour meter is a two rotor type (sw63) using the A and C phase current transformers and the A, B, and C phase potential transformers. this is commonly referred to as a two and a half watt- meter measurement method, This watthour meter was not calibrated using a proce dure resulting in a correction curve which could be used by the camputer. ‘The two and a half wattmeter method is not the most accurate procedure for measur- ing power since individual factors cannot be applied for the metering transformer ratio correction or phase angle correction. Indeed it is reported in PIC-6-REPORT [2] that metering systems of this type have an uncertainty of 0.5% when calibrated in a three phase configuration. Uncertainty can be as high as 1.08 if each phase of the meter is calibrated separately. NEW WATTHOUR METERING SYSTEM The stated objective of the new watthour metering systen is to be able to measure the gross electrical energy output of the generating unit with an error contribu- tion by the watthour instrunentation of less than 0.05%. This is required so that the total error including metering transformers, watthour meters, and timing and quantitization will be less than 0.1% at the lowest power output of the unit.A specification was written and proposals fran three vendors were obtained. The successful bidder was Scientific Columbus and the final product was delivered within the buigetary estimate. The specification appears as Appendix A of this report. A picture of the watthour metering system is provided in Figure 2-1. Figure 2-1. New Watthour Meter Significant features of this system are: © The high resolution pulsed output minimizes quantization errors.© ‘The solid state design has no moving parts except for display registers. © Single phase and total 3-phase energy (KWH) are accumulated and displayed. ‘The 3-phase display register cannot be reset. © The system places a low burden on the instrument transformers. It therefore has negligible contribution to phase angle shift. © Pulsed output (1 pulse = 50 KW-hours), and analog outputs (4-20 m” and 0-1 mA) are provided per phase. © Provision has been made for convenient interchecks among the three individual meters. The pulse outputs fran the new watthour metering system will go to the microprocessor, described in Section 5, that was developed for the Pressure Ratio Monitor and the Exhaust Pressure Monitor. ‘The microprocessor will accumulate the watthour meter's pulses and transmit 2minute and 10-minute generation data to the PRIME canputer system where they will be used for the periadic calculation of performance. INSPRUMENT ‘TRANSFORMERS Errors in power measurement, in addition to those inherent in the watthour meter, are introduced by the potential transformers and current transformers through: © Transformer ratio variations: calibration of the instrument trans formers allows a ratio correction factor to be determined. © Phase angle displacement between primary and secondary windings of the instrument transformer: calibration allows a phase angle cor- rection factor to be determined. Both these effects are caused by the magnitude of the current exciting the transfomer, fraction of rated voltage or current, power factor, and impedance (or burden) of the instrumentation connected to the secondary winding. It is planned to have the potential transformers recalibrated at several burdens during the major outage in early 1985, It would be desireable to actually measure the burden but if this measurement is impractical, the potential transformers will also be 24calibrated at zero burden so that the potential transformer ratio correction fac- tor and phase angle shift can be calculated for any burden. REFERENCES. PIC 6-1976, Performance Test Codes, STEAM TURBIN: St., New York, NY 10017 ASME, 345 East 47th PIC 6-REPORT - 1969, Performance Test Codes, "Guidance for the Evaluation of Measurement Uncertainty in Performance Tests of Steam Turbines," ASME, 345 East 47th St., New York, NY 10017Section 3 ‘TURBINE EXHAUST PRESSURE MEASUREMENT ‘The exhaust pressure of a steam turbine is not controllable in the sane sense that the inlet steam conditions are. Units using mechanical draft cooling towers can have controlled exhaust pressure by changing the number of fans running. Morgantowm 2 uses cooling water directly fran the Potanac River which means that there is no real-time control of exhaust pressure. The output of the steam tur- bine is affected by the exhaust pressure which, in turn, influences the input-output characteristics used for econanic dispatch. The accurate measurement of this pressure is very necessary for accurate and reliable performance monitoring. EXHAUST PRESSURE MEASUREMENT ‘The Morgantown 2 turbine has an exhaust hood design referred to by General Blectric as a Herzoy hood, named after the designer J. Herzog. ‘this hood design was a dramatic improvenent in reducing the stean pressure loss fron the last rotating stage of the turbine to the condenser inlet. These hoais, however have peculiar flow distributions requiring pressure to be measured in four (4) places at each single flow exhaust flange area for a total of sixteen (16) pressure measurements. The positions for these measurements have been determined by exten sive laboratory and field tests conducted by General Electric. Because of the chactic and high velocity flow patterns at the turbine exhaust a device referred to as a basket tip, Figure 31, is used to cbtain a measurement of static pressure. coHa EXTRA HEAVY Pipe (stancess Sreeu) 0-116" HOLES 1/4" PITCH SCREEN (STAMLESS STEEL) ST oes oa wane WE TUE 0. Figure 3-1. Rasket Tip A basket tip is formed by wrapping a wire mesh around the open end of a pipe. The Pipe and mesh are then inserted into the LP turbine exhaust flow strean with the ‘open end of the pipe facing downstream at an angle of 45° with respect to the flow path. This configuration was reported by Kratz 2] in 1954 to be a very reliable method for measuring static pressure at the turbine exhaust. The wire mesh serves to dissipate much of the local flow stream manentum resulting in a static pressure measurement which is almost independent of the Flow stream velocity as illustrated in Figure 3-1 which was taken fran Kratz [2]. For this reason, a basket tip is 32one of only two devices recomended by PIC-6 [1] for measuring steam turbine ex- haust static pressure. Basket tips are installed at the exhaust flange of the low pressure turbine. Pressure at the exhaust flange is not the same as the pressure in the condenser. there are feedwater heaters, and perhaps other obstacles, Located in the space be- tween the exhaust flange and the condenser causing additional pressure drop. A future task will investigate this effect on the measurement and prediction of con denser performance. ‘the pressure at individual basket tips is not necessary for the calculation of performance; only the average is required. However the pressures must be measured individually. When basket tips ave manifolded together flows are induced in the connecting piping due to pressure differences between pairs of basket tips. Pressure measured in a manifolded system will not be the average of the individual basket tip pressures. It will be influenced most by the two basket tips closest to the pressure transducer and may be canpletely insensitive to the pressure of the basket tips farthest from the transducer. In addition, since the pressure is sub-atmospheric and usually less than five inches of mercury absolute (about 2.46 PSIA), the basket tips must be purged in order to prevent a water column buildup which would result in an erroneous (lower) turbine exhaust pressure reading. ‘The Exhaust Pressure Monitor provides the necessary hardware and software to. con trol the purging of the lines and basket tips and to read the individual pressures. ‘the system will calculate the average turbine exhaust pressure in inches of mercury absolute corrected to 32P. Both hardware and software are in- corporated in the microprocessor system described in Section 5.Specifically, the Exhaust Pressure Monitor is comprised of Four, 4-tap controlled purge tap selectors. Tap selectors are manifolded together in pairs to provide two 8-tap systems. Figure 342 shows one of the tap selectors; a diagram of the tap selector is shown in Figure 3-3, Figure 3-2. Tap Selector ‘wo analog inputs (4-20 mA) to read pressures detected by pressure transducers connected to the tap selector systems. Sixteen digital outputs to control the three-way solenoid valves used to connect the basket tips alternately to the purging gas or to the pressure transducer. 3-4BASKET TIP PRESSURE TRANSDUCER PURGE SUPPLY Figure 3-3. Schematic of Tap Selection ‘The sixteen basket tips are divided into two groups of eight, one group for each condenser section. within each group the pressure of the basket tips are measured one at a time. the basket tips not being measured are purged to prevent the buil- dup of a water column, At any instant seven of the basket tips in each group will be purged while the pressure of the remaining one is measured. As each pressure measurement is made the average pressure at the turbine exhaust is updated. 35REFERENCES ASME, 345 Hast 47th St., PIC 6-1976, Perfomance Test Codes, STEAM TURBINE New York, NY 10017 2, E.M. Keatz, “Experience in Testing Large Steam Turbine-Generators in Central Stations," ASME paper 54-A-258, 1954 26Section 4 PRESSURE RATIO MONITOR ‘The measurement of steam turbine control valve position, or valve stroke, is cam plicated by the fact that steam turbines are equipped with a series of valves, not just one valve as the term "valve position" would imply. Except during startup these valves are operated sequentially (partial arc steam admission) with sane overlap so that each successive valve begins to open prior to the previous valve reaching its wide open position. This is done to attempt to make the steam flow a Linear function of valve stroke. Mechanical indications of valve stroke currently in use provide neither an accurate nor a reliable indication of valve position. When working such systems may be in error by as much as 10% or more. Currently there seems to be Little need to know valve position except for purposes of testing. Present dispatch techniques do not handle the canplication of Loading units in a manner that optimizes turbine efficiency by selecting optimim valve positions, often called valve point loading. ‘There is, however, great interest in valve point loading as a means of improving overall efficiency of operation. In addition to providing the information needed to accanplish valve point loading an accutate knowledge of valve position would provide: @ Confirmation that a requested change in load has been initiated in timely fashion; © Accurate knowledge of remaining capacity; © Ability to adjust valve position accurately so that turbine tests can be performed at known valve positions.A method utilizing the reasonably close relationship between stean flow, initial pressure, first stage pressure and control valve position has been developed and will be demenstrated at Morgantom, Measurements male upstrean of the valves (throttle pressure and temperature) are used to determine control valve inlet conditions. Downstream of the control valves the measurement of first stage pres- sure and a canbination of thermadynamic relationships and steam turbine design in- formation is used to determine conditions at the control valve outlet. Steam flow and valve position can then be determined. DESCRIPTION OF SYSTEM The Pressure Ratio Monitor consists of a microprocessor system with Analog-to-Digital conversion equipment to measure: © Throttle temperature; © Throttle pressure; © First stage pressure; © Pressure at the second extraction of the LP turbine; and @ Temperature at the second extraction of the LP turbine. DESCRIPTION OP THE PRESSURE RATIO SYSTEM ‘The following iterative step by step procedure identifies the method utilized to determine valve position from measurements of throttle pressure, throttle tempera ture and first stage pressure, STEP 1: Measure throttle pressure (PT), throttle temperature (TT) and first stage pressure (Pl). Fran the measurements calculate throttle en— thalpy (HT), throttle specific volume (VT) and square root of the ratio of throttle pressure to throttle specific volume SORT (PT/VT) « smmp 2: Determine an approximate equivalent valve position fran the near Broportionality between flow rate and first stage shell pressure. Figure 4.1 presents the relationship between valve position and first stage pressure for Morgantown unit 2 by plotting data taken from heat balances made by the manufacturer. These data are tabu- lated below:FIRST STAGE VALVE ‘TaROrTLE PRESSURE POSITION FLOW 2632.6 100.0 3,964,071 1935.6 75.0 2,974,053, 1523.4 59.7 2,367,904 1142.8 45.0 1,783,832 882.6 35.0 1,387,475 751.1 29.8 1,181,236 100 m 50 VALVE STROKE 750 1750 2780 FIRST STAGE PRESSURE (PSIA) Figure 4-1. Valve Stroke vs. First Stage Pressure History generally indicates that large steam turbines will have a flow capacity approximately 3% greater than that sham on "guarantee" heat balances. For that reason, it is suggested that 3% be added to each of the throttle flow rates tabulated above. Actual flow rates may increase for a given valve position, and/or first stage shell pressure, if erosion or rubbing causes local area increases in the early high pressure stages. A similar problem in the opposite direction might occur if internal damage causes a reduction in area in the turbine's first or second stage. These problems are sufficiently likely that an added check is suggested to help recognize their existence.sTep 3: The calculated "valve stroke" is a hypothetical value that assumes the control valves can be manipulated to achieve a linear change in flow in proportion to the signal. Tt provides a useful reference in the analysis outlined here, but it should be recognized that ac~ tual control valve position as a function of calculated valve stroke will be established by test at the plant. # Calculate an approximate first stage enthalpy (Hl). For standard throttle conditions (3500 PSIG, 1000°F) the work done in the first stage (the used energy, UE) is a function of control valve Position. ‘The relationship is shom in Figure 4-2, and ignores the effects of partially opened valves. These data, also fram heat balances made by the manufacturer, are tabulated below as well: VALVE FIRST STAGE POSITION USED ENERGY 100.0 27.5 75.0 50.2 59.7 60.0 45.0 68.9 35.0 76.3 23.8 76.4 100 i w> go 5, 3 50 Fe a g 0 0 so 100 VALVE STROKE (2) Figure 4-2, Tirst Stage Used Rnergy vs. Valve Stroke a4sTEp 5 sree 4 ‘The used energy (UE) of the first stage actually droops sig- nificantly between valve points as the next higher control valve opens. Flow passing through the partially open valve experiences a high pressure loss which causes the first stage efficiency to drop. ‘The magnitude of the drop off is difficult to establish since stean conditions at individual nozzle arcs cannot be measured. Fortunately the need for accuracy is important for the Pressure Ratio Monitor only in those areas where the individual control valves are approaching the open position and before the next valve starts to open. In this zone, the error caused by ignoring the ef- fect of partially opened valves is small. ‘The first stage shell enthalpy (Hl) is determined by subtracting the first stage used energy fran throttle enthalpy: mL HT - UE Determine the square root of the ratio of pressure to specific volune at the first stage: SQRT(P1/V1). Use first stage enthalpy, Hl (calculated) and first stage pressure, Pl (measured) to find specific volume at the first stage, V1. Determine the mass flow rate (or valve stroke) as as a function of SORT (P1/V1) « Fran the manufacturer's heat balance design values the following, also shown in Figure 4-3, are known: ‘THROTTLE. soRr(P1N1) FLOW (%) 100.193 100.0 75.163 75.0 59.644 59.7 34,100 35.0 Correct the calculated valve stroke for non-standard initial conditions. Prior to this point, the calculated valve stroke and the throttle flow (WI) were considered identical. Where the throttle conditions deviate fran design values, however, the con- trol valves will have to be opened to a different value to achieve agiven floy rate. At 50% initial pressure, for example, the valves would require almost a 100% stroke in order to accamplish a 50% throttle flow. ‘The most effective correction for non-standard conditions employs the square root of the ratio of pressure to specific volume as shown below: VALVE (8) = WI(8) * 130.157 / SORT (PT/VT) where VALVE(8) is the corrected valve stroke in percent, WI(8) is throttle flow in percent and 130.157 is the square root of the ratio of throttle pressure to throttle specific volune evaluated at standard conditions of 3500 PSIG, 100%.sre 7: 100 THROTTLE FLOW (2) 0 50 100 SQUARE ROOT IPI/V1) Figure 4-3. Throttle Flow vs. /P/WV at the First stage Recalculate the first stage used energy (UE). Using the corrected valve stroke cbtained in Step 6 above, a new estimate of UE should be cbtained. The same relationship of valve stroke and UE employed in Step 3 can be used. It is recognized that large deviations fran standard conditions do modify the used energy of the first stage. The effect is fairly small and the error introduced in the ‘SgRT(P1/V1) is even smaller. For example, with a throttle pressure of one-half standard (normal temperature), the UE might increase by about 3 BTU. Samewhat larger errors will exist due to the neglect of throttling losses through partially open valves. These errors - say 6 BIU can cause about 1% error in specific volune and about. 0.58 in SoRT(PI/V1). ‘This error appears acceptable. It is rela~ tively small, but ‘also occurs at flow conditions where precise Knowledge of the valve position is not essential. Fram the improved estimate of UB, determine Hl, VI and SORT(P1/V1). Té the value of SQRT(PI/V1) has Changed by more than 0.5% fran that calculated in Step 4, repeat Steps 5, 6 and 7. Convergence will be reached in 2 or 3 iterations.tthe above steps provide a systen to determine a valve stroke fran measured values of PT, 1] and Pl, the value of this project will be further enhanced by methods that show individual control valve positions as a function of the valve stroke. ‘This will require actual valve position measurements as a function of either throttle flow or the pressure and temperature measurements listed here. Such measurements have been taken at Morgantown and will be added to this analysis. ‘The accuracy of the measurements and analyses may becane camplicated by erosion, rubbing or internal steam path damage to the early stages of the high pressure turbine, To identify that such problems exist or to monitor the origin of such problems, additional measurements of pressure and temperature at the second ex- ‘traction of the LP turbine will be made. A comparison will be made of the ag- parent percent of valve-wide-open (WO) throttle flow at this Location with that deduced at the first stage shell, ‘This is of course the ratio of actual to design ‘SoRT(P/V) calculated for both the first stage and the second extraction of the LP turbine. the stage following the second extraction of the LP turbine should be sufficiently far Eran erosion and insensitive to rubbing that it will provide a fairly reliable check. For example, if the SQRT(P/V) for the LP turbine, second extraction shows 80% WO flow and the first stage shell suggests 758, one might be warned that the earlier stages may be eroded or rubbed. A physical check of the valve position would also be called for.Section 5 IMPLEMENTATION OF THE EXHAUST PRESSURE MONITOR, PRESSURE RATIO MONITOR AND WATTHOUR METER INTERFACE ‘The canputer originally installed at Morgantom, a G2-PAC 4020, scans over 800 in- struments that measure the various pressures, temperatures, flows, speeds and levels of units 1 and 2, In addition it monitors the state of a Like nunber of Limit switches. The ability to install new functions or expand the nunber of monitored points is Limited by the lack of manory and free time. Furthermore, ex- pansion is hanpered by the difficulty of programing in the machine's assenbly language. Both the pressure ratio monitor (Section 4) and exhaust pressure moni- tor (Section 3) require additional data acquisition and computation capability. In addition, the new watthour meter (Section 2) provides a pulsed output per phase with each pulse representing 50 KWH of generation. Additional data acquisition is needed to accommodate the outputs of the new watthour meter. Rather than upgrade the existing plant computer or wait for a new plant canputer system to supplement the existing camputer, it was decided that a microprocessor system would provide the additional data acquisition and camputational, capabilities that were needed, Furthermore, it was felt that the value of the ex- aust pressure monitor and pressure ratio monitor could be best exploited in the industry if they were included in a small dedicated system, One that could be in- stalled in a power plant where no camputer was available or the computer could not be easily expanded.SYSTEM DESCRIPTION An Intel single board computer (iSBC 88/40) was chosen for the system, other boards were added to provide additional memory: random access memory (RAM), erasable programmable read only memory (EPROM) and bubble memory. The single board canputer has follaving features: © High perfomance 5 mil iAPX 88/10 8-bit INOS processor © 12-bit, 20 ki A/D converter ¢ 16 differential analog input channels © Four EPROM sockets for up to 32k bytes of storage © isc 337 miltimeiule numeric data processor © sx 351 programmable serial 1/0 multimodule board © 24 parallel 1/0 lines ‘There are two-mamory expansion boards. One is an Intel iSBC 028A that provides 128K bytes of RAM with parity. The other is a Control Data B1014 128K byte EPROM board. An Intel iSBC 254-S1 bubble memory board provides 128k bytes of alterable non-volatile bubble menory. ALL input/output signals are terminated with heavy duty terminal blocks which are campatible with utility usage. These are wired to signal conditioning panel: Intel iCS 910 for termination of analog inputs and Intel ics 930 for termination of digital outputs. ‘The multibus boards are housed in an Intel iCS-80-kit 640 chassis which provides all of the necessary voltages and cooling. ‘The iCS-80-kit 640 chassis, signal conditioning and termination are housed in a 19-inch RETMA cabinet.ALL inpat and output signals are protected fran electrical transients through the use of MOV devices. The system requires one 120 VDC, 15 AMP, 60 HZ power input. Figure 5-1 illustrates the hardware configuration. SOPIWARE, ALL software runs under Intel's iR¥K-86 real time multitasking operating system. ‘The iRMK-86 operating system a comprehensive mul tiprogramming software system for Intel iSRC 86 and isaC 88 single board canputers and other iAPK86 and iAPK@8-based microcomputers. the iRMK-86 operating system extends the architecture of the un- derlying processor by providing a collection of new operations that act on the operating systen dbjects provided by the system or user created extensions. Each layer of the operating system simplifies user access to the underlying hard- ware by taking advantage of the mechanisms provided by the layers below. Each layer of the iRMK-86 operating system is configurable to allow applications to custanize the system to particular needs. FUNCTIONAL DESCRIPTION Services provided by the iRMK-86 operating system include facilities for executing programs concurrently, sharing resources and information, servicing asynchronous events, and interactively controlling system resources and utilities. In addition, the iRMK-86 operating system provides all major real time facilities in- cluding priority-based system resource allocation, means for concurrently moni tor- ing and controlling multiple external events, real time clock control, interrupt management, and task dispatching. The iRM%-86 operating system contains the fol~ lowing modules: 5-3squouodung usqsks *T-5 ombty© An object-oriented Nucleus © Device independent basic and extended 1/0 systems © Temminal handler © Bootstrap and application loaders © Human interface with camplete cammand line interpreter @ An interactive, object-oriented debugger. Because the maiules and services provided by the operating system are user selectable, application specific operating systens can be created by imx-36 users. The iRW-86 operating system therefore eliminates the need for custan operating system design, thereby reducing development time, cost, effort, and risk. FEATURE OVERVIEW ‘The iRMK-86 operating system provides users with simple, easy-to-use, quality software tools for creating a wide range of application systems, Same important features are explained below. Structured Application Bavironnent The imMx-86 operating system provides a consistent structure fron application to application, and CPU to CPU, thus allowing experience gained on one system to be easily transferred to others. Often entire programs can be ported fran one ap- plication to another. 55Object-Oriented Architecture The iRMK-86 operating system extends the capability of the underlying CPU by ad- ding a nunber of new data structures (ebjects) and a number of functions to oper- ate on these objects. ‘This architecture provides a simple, symmetric, and easy to learn interface to a camprehensive system. The heart of the operating system, the Nucleus, provides the means to create, manipulate, and delete the basic objects necessary for any application. It also provides a mechanism for users to create their om object definitions and use them as part of the basic operating system. Bach of the outer layers of the system add to the list of available cbjects by using this same extension mechanism. Nucleus ‘The Nucleus of the iRMX-86 operating system provides the foundation upon which a variety of applications systens can be built. It includes the facilities to manage the basic dbjects of the system necessary to perform multiprograming, multitasking, critical section managenent, and extensive task-to-task camunica- tion and control. Dnbedied in the Nucleus are the Facilities to support concurrent program execution and handling of simultaneous asynchronous events. ‘These facilities alloy inter- rupts caning fran specialized peripheral devices to be serviced in an efficient manner. The R1X-86 operating system allows the CPU hardware to be used by mul- tiple applications, thus reducing the overall system size, camplexity and cost. 5-61/0 systen ‘The iRMK-86 1/0 system provides an extensive facility for device independent 1/0 through a series of supplied device drivers, or any number of user supplied device drivers that can be configured to operate at any 1/0 port address. The Basic 1/0 System (8105) implenents an asynchronous interface to the device drivers allowing users to explicitly overlap 1/0 functions with other operations. The Extended 1/0 system (EI0S) perfoms all of the synchronization necessary to do read-ahead and vwrite-behind buffering autmatically, and to reference files with logical names. By configuring the appropriate interface, applications can develop an 1/0 subsys~ tem with the optimum degree of device control while requiring a minimun of design time and effort. Furthermore, the device-independent nature of the systen allows use of different devices without redesign. ‘TASK ORGANIZATION Programs written for the pressure ratio monitor, exhaust pressure moniter and wat~ thour meter interface reside in the system as a number of independent programs ot tasks. these tasks are managed by the Nucleus of iRe-86. Figure 5-2 shows a block diagram of the task structure of the system. Nessaye Task ‘the Message Task is responsible for receiving messages fran the camunication port that connects the microprocessor to a terminal or to another canputer. when a data request is received, the message task hegins execution and decodes the mes- sage received. It then sends a semaphore to the task to which the request was directed, There are three messages for the system. ‘The first is a request for pressure ratio task to perform its calculation and output the results over the same cammmication port. The second is a request for most recent exhaust pressure data to be output to the camunication port. Finally, the third is a request to S71go into maintenance maie. Figure 5-3 is a flow diagram which shows the operation of the message task. Pressure Ratio Task The pressure ratio task's function is to calculate the pressure ratio and control valve position. It gets its inputs fron the analog input task and send the results, via the R5232 cammunication channel, to whanever requested it. ‘This function is executed whenever a request for pressure ratio has been received. Figure 5-4 is a flow diagram of this task. Exhaust Pressure Task ‘the exhaust pressure task performs the function of basket tip selection, reading pressures, averaging and transmission of data when requested. The exhaust. pres- sure task reads one pressure fran each condenser section at a time and puts the value in a table, The task then selects the next pair of basket tips using the the parallel output port. Figure 5-5 is a flow diagram of this task. Analog Input Task ‘The analog input task has the job of reading analog inputs and converting then to engineering units. The analog input task responds to requests by formatting a message to be sent back to the requesting task. 59Messages sent between the analog input task and other tasks are contained within segments, contiguous memory areas maintained by the operating system, Figure 5-6 is a flow chart of the analog input task. Maintenance Task The maintenance task is an interactive task which camunicates via the serial port and provides the user with the ability to change and view conversion constants, system constants, hardware address and to control storage of this data in the Dubble memory. The maintenance task allows the user to view and change input/output signals for maintenance. 5-10INITIALIZE MAIT FOR A MESSAGE PR MESSAGE SIGNAL EXHAUST PRESSURE TASK To OUTPUT ITS DATA SIGNAL PRESSURE RATIO TASK To ovTPUT TTS DATA SUSPEND EXHAUST PRESSURE TASK AND START Maintenance [>] TASK WAIT FOR NATNTENANCE f= TASK TO FINISH Figure 5-3. Message Task Flow Chart SINITIALIZE \{— WAIT FOR A SIGNAL FROM THE MESSAGE TASK SCAN INPUTS AND CALCULATE VALVE POSITION i SEND OUTPUT MESSAGE Figure 5-4. Pressure Ratio Flow Chart 5-12INITIALIZE SCAN PRESSURE AT SELECTED IBASKET TIPS AN UPDATE AVERAGE SUSPEND SELF; WAIT TO BE RESTARTED ‘SUSPEND MESSAGE, SEND OUTPUT NESSAGE - (CONDENSER PRESSURE) SELECT NEXT SET OF BASKET TIPS AND DELAY 5 SECONDS Figure 5-5. Exhaust Pressure Flow Chart 5-13© INITIALIZE + WAIT FOR A MESSAGE FROM ANOTHER TASK SCAN INPUT, CONVERT COUNTS TO ENGINEERING UNITS SCAN INPUT ~ RETAIN COUNTS READ FROM A/D CONVERTER SEND MESSAGE CONTAINING ENGINEERING UNITS TO REQUESTING TASK SEND MESSAGE CONTAINING COUNTS TO REQUESTING TASK Figure 5-6. Analog Input Flow Chart 514Section 6 SYSTEM FOR MEASUREMENT OF INTERNAL PACKING (N2) LEAKAGE A large nutber of steam turbines are arranged with opposed flow high pressure (HP) and intermediate pressure (IP) sections. Such an arrangement places the high pressure steam seen by the turbine's first stage (HP turbine section) in close proximity to the lover pressure stean at the reheat boil (IP turbine section). Leakage of stean along the turbine shaft from the first stage to the reheat bowl is inevitable and a packing, carmonly called the N2 packing, is used to minimize this leakage. Such packing is critical to the performance of the turbine, since Leakage passing through the packing © Bypasses all high pressure turbine stages except the first stage. © Bypasses the reheater and, © Reduces the mixed enthalpy of the steam entering the reheat turbine. the three effects reduce both efficiency and capacity. Cycle losses of 1.08 fran this single cause have been experienced. Units the size of Morgantown unit 2 may experience a capacity loss of several megawatts. Determining the clearance of this packing will serve several valuable purposes. ¢ Improve the knowledge of actual turbine canponent: performance. ‘¢ Improve maintenance planning. © Help identify the conditions which cause shaft rubbing so that they may be avoided.© Identify whether thermal distortion is causing leakage losses be- yond those determined fran physical clearance measurenents. HPAP wring alow j 2 packing ty setae Seats! Sputot ‘ave Hanger ang concrete Wa water [9 to condenser Spry Pigure 6-1. N2 Packing Measurement System A method of determining the N2 packing flov and clearance using the packing's emergency blowlom system has been devised. the blawown system is an safety fea ture that connects the packing to the condenser when the turbine is tripped. Following a trip the blowiown prevents the steam that is bottled up in the HP tur- bine section fram leaking into the IP section where it may contribute to turbine overspeed; steam fram the HP turbine section is channeled through the blowiown Piping to the condenser instead. During normal operation the blowiown valve closed and no steam flovs through the blowlown piping to the condenser.Modifications to the blowdown system have been made and instrumentation installed so that stean flow through the blowiown system may be controlled and measured. In addition, pressure and temperature of the steam will be measured. These are shown in Figure 6-1. ‘The blowowm system will be used to measure an effective N2 packing clearance. Special tests will be conducted during which a controlled flow will be allowed to pass through the blawlom system to the condenser. Measurements of blowlown ten perature and flow along with measurements of pressure at the first stage, blowiom annulus and reheat bosl will be used to calculate the clearance of the N2 packing. The tests will be conducted periodically so that changes in packing clearance may be monitored. ‘once the clearance of the N2 packing is known the steam flow through the packing can be calculated under normal operating conditions (no blowdown flow) fran measurements of pressure at the first stage and reheat bowl. N2 PACKING The N2 packing of Morgantam unit 2 is shown schematically in Figure 6-2. Pressures and flows sham in the figure, and variables used in the equations that follow are listed below. Leakage area. Area is calculated as show in By. 6-1. By: A function of the number of teeth and pressure ratio across the Portion of the N2 packing between the first stage and the blowown annulus, see By. 6-2, There are 16 teeth in that section of the packing. Ba A function of the nunber of teeth and pressure ratio across the Portion of the N2 packing between the blowdown annulus and the RH turbine section, see By. 6-3. There are 30 teeth in that section of the packing. 63By? A function of the number of teeth and pressure ratio across the en- ‘tire N2 packing between the first stage and the reheat bowl, see Bq. 6-4. There are 46 teeth in the N2 packing. Ct Packing clearance in inches. Dz Packing diameter, 25.0 inches for the Morgantown N2 packing. kz A factor for the packing type and condition [1]. For the Morgantown N2 packing k is taken to be 54.0. First stage pressure Pressure at the reheat bowl Pressure at the blowown annulus Specific volume of steam at the First stage Specific volune of steam at the blowdown annulus Leakage flow fran the HP turbine section into the N2 packing Leakage flow fran the N2 packing into the reheat bowl Steam flow through the blowown system A=a7De (6-1) 1.0 - (P,/P,) 16.0 = log, (P,/P,) om 2 1.0 - (P)/P,) (6-3) 30.0 - log, (P2/P,)2 1.0 = (P,/P,) 8 : 46.0 - log, (P)/P) (6-4) REHEAT BOWL A gq FIRST STAGE y Me BLOMDOKN 1 ia C RANULUS P. re ae Me SHAFT CENTER LINE Pigure 6-2. Schematic of N2 Packing As shown in Pigure 6-2 Morgantom unit 2 employs an N2 packing which separates the steam discharged fram the high pressure first stage fran that of the reheat tur- bine first stage. ‘The rated pressure drop on the N2 packing is approximately 1900 Pst, with a design leakage flow rate of over 47000 18/HR. At valves wide open, the pressure and flow rate would be about St greater. If rubbing were to cause a doubling of this leakage, the reduced output of the high pressure turbine would be about 1800 KW. 65Martin's Formula [1,2] and a conservation of mass relationship permits expressions for the flows Wy, Wz and Wy to be developed. these are given beloy in Eqs. 6-5, 6-6 and 6-7. = 25.0 a 8 | gt (6-5) x (6-6) x (6-7) Pressure at the blowdown annulus, P,, may be determined by calculation of the pressure drop between the blowlown annulus and the pressure tap which has been 1o- cated as close to the turbine shell as practical. Assuming the clearance is uni- form throughout the packing (the sane for all rings), the following can be used to determine the effective clearance: — 6-8} 25 1k D (8, VPI - 8, VPLV,) aaFLOW CHARACTERISTICS For given packing clearance and steam conditions packing leakage flows and blow~ down flow can be calculated as a function of blowiown annulus pressure P,. In the example below this is done with steam conditions taken fran the manufacturer's valves wide open (WWO) heat balance. Stean conditions ar Vy=0.3902 FT3/tB, P7=469.5 PSIA. A packing clearance of 0.035 inches is assumed for the example. Since packing clearance is assumed, the pressure at the blowiown annulus, Py, is fall that is needed to determine packing flows Wy and W2 using Bys. 6-5 and 6-6. As blowdown flow increases P, will decrease. Results are show in Table 6-1 and plotted in Figure 6-3. 800 58,092 15,745 42,347 700 59,230 12,643 46,587 600 60,177 9,116 51,001 DIScUSsrON At zero blowlom flow 81 (Bq. 6-2) calculated for the first 16 packing teeth must be equal to By (eq. 6-4) calculated for all 46 teeth of the packing. Since Py and Pp are know P, can then be determined. For the example above Py is 1601 PSIA, 2 pressure lover than first stage pressure by 238 of the pressure drop between the first stage and the reheat bowl. 6770 60 50 (KLB/HR) 40 30 20 LEAKAGE FLOW 10 0 2 4 6 CeO iced iG PRESSURE AT BLOHOGHN ANNULUS (KPSIA) Figure 6-3. Leakage Flow Characteristics The blowlown pipe pressure at zero bypass flow should be 1601 PSIA at standard WHO conditions. For other zero bypass conditions the pressure should also be lover than first stage shell pressure by 23% of the pressure difference between the first stage shell and the reheat bowl. I€ it is not close to this level it should bbe suspected that the clearance is not uniform throughout the packing. This might be caused by broken teeth (the tooth pressure drop is greatest at the law pressure end of the packing), or by non-uniform temperature of the rotor or packing ing holder. ‘The actual deviation fram 238 can be used to determine the probable ratio Of clearance before to clearance after the blowiown annulus. 68‘the enthalgy of the bypass steam should be used to determine the specific volume of the steam at both calculation points (W and W). ‘This enthalpy also has sane value in determining steam conditions at the discharge of the first stage, al- though it should be recognized that it might be sanevhat high due to rotation loss, conduction fran nozzle boxes, etc. A significant additional value of the N2 packing test is the correct determination of the leakage steam enthalpy. ‘The enthalpy is needed, along with the flow rate, to properly calculate the IP section efficiency. Tt should also be noted that the enthalpy of the leakage flow will fluctuate not only with throttle enthalpy and valve position, but with leakage clearance, first stage reaction, erosion on the first stage nozzle, root seal clearance and second stage flow area, CALCULATION OF Py Figure 6-4 shows the estimated loss of pressure for steam being drawn fran the packing area into the blawown pipe and bypass system. ‘the total loss is about 4,5 velocity heads to the location of the pressure tap. ‘the velocity will be low during the planned normal pressure operation tests so great accuracy in determin~ ing the local velocity is not critical. when the unit is open for inspection, measurements of flow areas will be made, In the meantime it will be assumed that local velocities at full load with W, = 30,000 L8/HR will be about 65 feet per second.‘ollection and Guidance ‘to holes: 1.5 velocity heads Passage through holes: 1.5 velocity heads Enter Pipe and flow to Pressure tap: 1.5 velocity heads Pressure Drops From Packing Ring to Pressure Tap Tf serious doubts arise concerning the accuracy of the calculation for P,, a trial and error system can be formulated to provide a check. When packing clearance increases, the bypass flow needs to be larger in order to Provide adequate sensitivity in regard to a significant change in Py. For example, with an effective clearance of 70 mils the change in P, with 30,000 LB/iR bypass flow will be about 408 of that obtained at 35 mils. Still, this is about 240 PSI and should provide a satisfactory accuracy when the relatively small in- strument and calculation errors are considered. 6-10TEST PROCEDURE ‘A WWO test should include two rates of bypass flow; 15,000 LB/HR and 30,000 B/HR are suggested. Pressure in the bypass line should also be recorded with zero bypass flow. A lower load test (about 35% is suggested) is also desirable, with the bypass mass flow rate correspondingly lover as well. Tests at all 4 valve points are sug- gested for base line information but the low bypass flow rate need not be used at second and third valves. ‘the calculated clearance should be the same for all tests. If the calculated clearance varies significantly, added test conditions will be necessary to determine what variable (such as main steam temperature, re- heat temperature, first stage shell temperature, etc.) might explain the deviation. It is essential that all parts of the system are in thermal equilibrium prior to recording data. The turbine should also be maintained at a fixed valve position. ‘This is necessary to minimize errors due to temperature fluctuation. The bypass flow should be set and sufficient time allowed for steady state conditions to be~ cane established. Thirty minutes is suggested although careful observation for temperature stabilization may indicate when stability actually occurs. REFERENCES: 1, Spencer, R.C., Cannon, C.N., Cotton, K.C., "A METHOD FOR PREDICTING THE PERFORMANCE OF STEAM TURBINE-GENERATORS 16,500 KW AND LARGER," ASME 62-WA-209, 1962, Revised 1974. 2. Salisbury, J.K, STEAM TURBINES AND THEIR CYCLES, Robert E. Krieger Publishing Company, Huntington, N.Y., 1974, euSection 7 COMPUTER SYSTEM REQUIREMENTS ‘The basic process of a power plant performance program is the acquisition of data and its analysis. Monitoring power plant performance, pinpointing causes of unit deterioration, identifying solutions to deterioration of plant performance, and enhancing the econanic dispatch of generating units all require an autonated system, Computer requirenents fall into two broad categories: (i) the real-time requirenents of monitoring instrumentation, alarming abnomal conditions and dis~ playing the current state of the plant; and (ii) the general computing require- ments of a research project where “new ideas" can be tested quickly and efficiently. the first category falls within the traditional boundaries of event-driven, real-time pover plant computer systens. The second category is typical of multi-user, time-sharing computers with the added requirenent that data monitored by the real-time computers must be collected and saved for subsequent analysis. ‘The simaltaneous requirenents of real-time computer operation and a time-sharing environment are difficult to implenent ina single computer system. Indeed, sev- eral computer systens are being used to meet the various needs of the project. ‘These incluie the canputer originally installed at Morgantom, a GE-PAC 4020, a custan microprocessor system, a distributed canputer systen to be installed in the future, and a general purpose time-sharing minicamputer, a PRIME 750.REAL-TIME COMPUTERS The project requires computers to: (i) monitor the plant's instrumentation; (41) provide the canputational capability needed to implanent the monitors and Giagnostic programs developed during the project; (iii) provide the man/machine interface between the plant and the operators and performance engineers at the Plant; and (iv) provide Links to remote canputers for the exchange of information. ‘The canputer originally installed at Morgantown, a GE-PAC 4020, scans over 400 ine struments that measure the various pressures, temperatures, flows, speeds and levels on each of the two units in the plant. In addition it monitors the state ©f approximately 700 limit switches on each unit. The man/machine interface of the computer is crude by today's standards. Furthermore, the ability to install ew functions or expand the number of monitored points is Limited by the lack of memory and free time, It was decided to install a computer to canputer Link be- ‘tween the 4020 and PEPCo's PRIME 750 camputer to take advantage of the existing instrumentation, Honeywell was contracted to install the data link in the 4020 and provide the necessary programming changes to make it work. The project how- ever required a new computer system, referred to as the prototype systen, to pro- vide the expansion capabilities that are lacking in the 4020. Initial requirements of the prototype system will permit implementation of new instrumentation, data transfer to and fram PEPCo's PRIME 750 camputer, and provide operator and engineer process interfaces using CRT and keyboard. ‘The prototype ‘system, though, needs to be expandable to allow implementation of results of the Project. A fully distributed system best fits these needs. A distributed system is one that provides a functionally divided solution to a specific application. The functions, assigned to relatively autonamous hardware canponents or "boxes; could be divided as follows:© input/output: All input and output functions are provided in one unit. Por instance, for analog inputs, the function would be the acquisition of raw counts, correction, conversion to engineering units, alarm processing (not necessarily alarm reporting) and transmission on the data highway. © Logging: Logging functions should include the gathering of data to be logged, processing of data, archival of data in non-volatile memory and printing of logs on output devices. © Operator Displays: Display functions should include the storage of all displays, processing of displays, alarm reporting, operator in- terface and acquisition of data fran the data highway. ‘his functional division provides the ability to design the system using a build- ing block approach where individual functional units, or "boxes," are included as required. In addition, expansion in the field is very easy and allows for an al- most bumpless transition between an old system and a new expanded systen. A distributed data highway system eliminates the problems of the past associated with central computer based systems where the camputer was the bottleneck of the system. Any expandability was determined by the canputer systen's free time. In a distributed data highway system, the focal point of the system is the data high- way and its ability to pass data rapidly. If the band width and the communication scheme cannot handle large amounts of data then the data highway becanes the Limiting item, It was of utmest concern to the project that the data highway not becane a bottleneck in the system. The specification of the prototype system re~ quired that the data highway have enough capacity to accamadate double the ini- tial complement of input, output, logging and displays. The data highway of the prototype system procured for the project operates at 10 megabaud (10 million bits per second).‘The prototype system is being supplied by the Process Systems Division of Camputer Products Inc. Installation is expected in 1985. Major canponents, apart fran the data highway, include: © Operator's Station: ‘The operator's station is used to monitor the Plant and generate reports. It consists of a microprocessor, a 40 megabyte Winchester disk, a video generator, printer, black and white CRT, two color CRTs and various keyboards. * Engineer's Station: The engineer's station includes all the func~ tions of the operator's station and permits modification of the data base describing the points being monitored. In addition, logs and graphic displays are generated at this station. The station is configured like the operator's station with the addition of mag- netic tape and high speed printer. © Input/Output Boxes: Bach input/output box controls the scanning, Limit checking, alarm checking, conversion to engineering units and calibration correction of a set of points. Use of fiber optic Links permit flexibility in the placement of these boxes. © Gateway Box: ‘The gateway box consists of a microprocessor and an interface canpatible with the PRIME 750 computer. Key requirements of the prototype systen inclue the accuracy of the input hard ware and its scanning and calibration capabilities. Requirements for loyging, alarm processing, color graphic display capability and so on were also an impor- tant part of the prototype systen's specification. Since most modern systens pro vide sophisticated and canprehensive functions to meet these later requirements they are not considered further. Accuracy Requirements The accurate measurement of the plant's perfomance is central to this project. ‘The accuracy of the input hardware therefore is of the utmost importance. Clearly the measurement of performance requires accurate and high quality instrumentation. Tt also requires a data acquisition system whose inherent errors are smaller than those of the instrumentation. For these reasons the prototype system was required to have an analog to digital converter with 14 bit (13 bits plus sign) resolution. Furthermore, overall accuracy of the input including multiplexing hardware, analog 74to digital converter and associated amplifiers was required to be less than 0.05% of full scale. ‘The accuracy of information made available by the data acquisition system can be adversely affected by the method used to scan the instrumentation. Sone of the systens available today scan their input channels quite frequently but go through the process of conversion to engineering units, alam checking and transmission of a point over the data highway only when there has been a significant change in the point's value. Such systems negate the benefits of high accuracy input hardware. Resolution of analog inputs for such systens is determined not by the capabilities of the hardware but by the deadband corresponding to the “significant change." ‘The prototype systan was required to go through the full process of calibration correction, conversion to engineering units, alarm processing and transmission each time a point was scanned. Calibration and Scanning Requirenents The prototype system must be capable of providing a unique calibration, in addi— ‘tion to the normal conversion to engineering units, for each analog input. Furthermore, each analog input must be scanned at least every two seconds. The camplete process of calibration, conversion, alarm checking and transmission on the data highway must be done with each scan of each point. Many systems of the past, particularly those incorporating a centralized canpater, Aividea analog inputs into scan classes of varying frequency to evenly distribute canputer loading. Such scan classes are unnecessary in a distributed systen. They are, in addition, undesirable since they impose an engineering design burden on those who mast decide when and how often points are scanned. With the cost of canputation falling rapidly and the cost of engineering rising it makes little sense to use engineering talent to optimize computations. Engineering talent 15should be devoted instead to analyzing the results of those cnputations. The ex- istence of scan classes also makes adding and deleting instrumentation more dif- ficult because of the possibility that a small change in instrumentation will necessitate a reassignment of scan classes for a large muber of points. In a Project such as this, where frequent changes in instrumentation are anticipated, the burdens associated with scan classes are unacceptable. Thermaiynanic Properties of Steam and Water Functions that calculate the thermodynamic properties of stean and water are, of course, a necessary part of any program that monitors the performance of a steam turbine, stean generator and associated auxiliaries. It is necessary, therefore, that these functions be included in the prototype systen. There are three impor- tant requirenents of these functions: accuracy of the functions, their range of validity and their continuity. & bias in thermodynamic property functions can be tolerated if the goals of the Performance monitoring program are simply to detect and identify deteriorating efficiencies. If hasever the performance of several units must be canpared with one another, as is the case in the econanic dispatch of generation, absolute ac~ curacy is necessary. The specification of the prototype system required accuracy consistent with those of the 1967 formulations (Thermodynamic and Transport Properties of Steam, ASME, 1967) regardless of the camputational means used to implement the functions. 7-6Iterative techniques are often used to calculate perfomance parameters. An ex- ample of this is the calculation of wet extraction flows and LP turbine exhaust enthalpy. Iterative techniques can “wander about" before settling dom to a solution. During the course of such wanderings the thermodynamic functions may be asked to calculate various properties in regions of pressure and temperature not encountered in the actual process. It is important therefore that the range of the functions not be Limited simply to the regions that contain the cycle under consideration but be valid over all possible regions fran vacuum to supercritical and beyond including subcooled and saturated Liquid, dry and saturated steam and saturated steam/water mixture. Iterative techniques can also be sensitive to discontinuities, In fact discon- tinuities can cause an iterative procedure to "hang" while successive iterations jump fran one side of the discontinuity to the other never reaching a solution. When thermodynamic properties are implemented with multisection curve fits, it is important that the curves be continuous from one section to the next and not take a small jump as the boundary of a section is crossed. ‘TIME-SHARING COMPUTER A PRIME 750 computer, previously purchased by PEFCo, is being used to provide gen- eral canputing resources to the project's engineers. The camputer is configured with: © Two 600 megabyte disks; © One 80 megabyte disk; @ Two and one-half megabytes of main memory; © ‘Twenty-four asynchronous ports to accamodate user's terminals;© Five low speed line printers; © One high speed electrostatic printer/plotter; © ‘Three asynchronous ports to serve as links to other computers; and © Magnetic tape for backup and archival purposes. ‘The PRIME 750 is being used for the develoment and testing of various algorithms, generation of reports and presentation graphics and various engineering analysis Programs. It is also being used for Data Storage and Retrieval which is described below. ‘the requirements of a time-sharing computer such as the PRIME 750 are quite different than those of the real-time systems that are being used for data acquisition, Where real-time computer systens are characterized by many small tasks working in unison while sharing the memory and other resources of the camputer, time-sharing systens are characterized by many independent tasks that must be insulated fran one another. A time-sharing systen must not allow one user's program to interfere with the operation of another user's program, even if the first is riddled with programing errors. The task to task camnication facilities that real-time systems use to coordinate various tasks are not generally available to users of time-sharing systens; indeed they are rarely needed and potentially dangerous. Thus the first requirement of the time-sharing system is that it not allow one person's programing mistakes to interrupt the operation of another person's program. ‘Time-sharing systems are almost always used in an interactive mode: its users converse with the conputer using CRT and keyboard. The systen must therefore sug port a large nunber of interactive terminals simultaneously. Larger and more ex- pensive resources such as printers and magnetic tape drives can be shared among the users of the systen. In addition to the ability to support a large number and variety of terminal equipment, the follaving characteristics are essential for the efficient support of interactive programs.‘The 1/0 system must be able to accept user responses and data records in “free format" (i.e., without being sensitive to column position). Lack of this feature results in interactive procedures that are cumbersane to use and error prone; productivity suffers as the user's attention is diverted fran his problen to "counting blanks." During a single work session, user input streams, both conversational dialog and data, may cane from multiple sources and program output may be directed to a vati- ety of destinations. Usually such selections are made during program execution. ‘thas, the file syste must support dynamic file assignments; that is a progran must be able to access existing data files and create new data files "on the fly" during program execution. the user should not be required to decide which data files are to be used prior to initiating a program, The creation, extension, truncation and deletion of data files under program control is essential. Also essential is the randam access of data within these files; that is, the ability to read data fram (or write data to) the middle of a file without reading (or rewriting) everything fron the beginning of the file to the point of interest. Since the PRIME 750 will be used for software development a "friendly" programming environment is needed. Three of the most important requirements for friendliness are virtual memory, the availability of an interactive text editor and the availa~ bility of high level programming languages. Virtual manory allows programs to be written as if "infinite" mamory were available. The canputer's memory management systen relieves the programmer of the camplexities of cae and data overlaying. Virtual memory also gives flexibility in adding new features as programs are refined and evolve (and inevitably grow) «An interactive text editor is an essential element of the friendly programming enviroment. The text editor is the tool that engineers and progranmers use to write their programs. It should have the ability to access an entire source file ‘at one time regardless of the size of the file. Text editors that view program source files in "chunks," one chunk at a time, are not acceptable. Also essential is the ability to "jump around" within the source file making changes near the beginning of the file, in the middle and at the end of the file in random order. Text editors that require their users to make updates systematically, and in order, from beginning to end are intolerable, There are many other features of good text editors, the number and power of their editing camands for example. Tt is important to realize however that a good text editor can enhance the produc tivity of the people developing and testing programs; a bad text editor is an ob- stacle that must be overcane during program develogment and testing. High level programming languages, in particular FORTRAN 77, both improve the productivity of people implementing functions with canputer grograms and enhance the portability of the program. FORTRAN has long been the programming language of choice for scientific applications because of its powerful abilities to express mathematical relationships. Indeed the very name of the language cones fran the words “formila translation." FORTRAN has, in the past, lacked sane of the data types and logical control structures that make program development: convenient. FORTRAN 77 hovever eliminates many of the deficiencies of the past. FORTRAN is Row a fairly convenient language to use. Moreover, the standatd that defines the Language, ANSI X3.9-1978, is one of the most canplete of any programming language. This eases the movement of prograns fram one canputer to another canputer, Provided both have FORTRAN canpilers that adhere to the standard.DATA STORAGE AND RETRIEVAL one of the goals of the project is to identify the information that should be col- lected and saved to be able to detect deteriorating performance and identify the causes. Puture parts of the project, the on-line monitoring of perfomance or in- cremental heat rate for example, will benefit fran being able to use historical as well as on-line data during develogment and testing of algorithms. Enormous amounts of data will be monitored throughout the course of the project and it is impossible to select at the outset the subset of these data that will be needed as the project progresses. therefore a systematic means of saving all data is required. In addition, an interactive method of searching the saved data and selecting and retrieving relevant portions is required, Data Storage and Retrieval (DSR) programs have been developed to meet these requirements. The collection of data is an autanatic function: once started data are collected and saved without the need for human intervention. It is estimated that measure— ments fran as many as 1000 instruments must be saved each minute of each day for the duration of the project. At this rate about 80 megabytes of information will be collected each week. Saving all data has the advantage that nothing is lost. Tt has the disadvantage however that it is difficult to locate a particular datum because of the volune of information available. ata canpression techniques can be applied to reduce the volume of information collected. the value of a monitored point may be saved only when it changes significantly. Alternatively a 10-minute or 30-minute average may be saved. For sane points, points unimportant to the performance monitoring of the plant, noth- ing may be saved. Searching and sorting the reduced data set is much easier. ‘Same data however are inevitably lost. TuDSR uses both techniques and saves everything in one file while saving a cam pressed set of data in another file. thus the compressed set of data can be quickly searched for events of interest and used as an index to the full set of data. Retrieving data for review and/or analysis is a highly interactive process where an engineer, using CRT and keyboard, instructs DSR. An engineer may, for example, want to plot a histogram of turbine bearing vibration versus turbine speed during startup. A startup must be located in the volumes of information available and selected portions extracted for plotting. DSR provides the ability to interactively: ® Select a data file (history file) to examine. This may be data that was gathered only manents previously or many weeks or months ago. © Select a subset of points to examine or output: DSR initially con- tain over 900 points. Tt would be cumbersane if they all had to be viewed regardless of the the current interest. © Search for events: For example, find the next instance of throttle Pressure greater than 3200 PSIA. © — Select time intervals of interest. © Calculate simple statistics: average, standard deviation, 958 con- fidence interval of the mean, maximum value, minimun value and the number of samples in a given’ time interval. © Generate plots of points against time. © — Generate plots of points against one another. © Generate data files for external analysis programs or presentation graphics programs. © Run simple performance calculations such as HP turbine efficiency or IP turbine efficiency. yaSUMARY The canputer requirenents of project RP 1681/2153 are being met with a canbination of existing camputers and new computers all of which are linked together so that monitored information can be collected in one place for review and analysis. Needed computers fall into two cateyories. the first is typified by real-time computer systens that perform data acquisition functions among others. Measurement accuracy and ease of configuration (and reconfiguration) are important requirements for these systems. ‘The second category is typified by time-sharing minicamputers. The most important requirenent is that it offer a convenient environment for program development and interactive program use. ‘The special requirements of collecting, storing and recalling measurements made by the various data acquisition camputers resulted in the development of an interac— tive Data Storage and Retrieval system for the time-sharing computer. 7-13Section 8 TURBINE PERFORMANCE DIAGNOSTIC PROCEDURES Over a periad of years, diagnostic techniques for predicting internal steam tur— bine problems have been developed. The resulting knowledge of the specific sources of damage and the degree of seriousness is a critical input to performance evaluation and effective outage planning. ‘The presence of one or more of the four major sources of steam path damage: ex- cessive leakage, solid particle erosion, foreign material damage and deposits can be identified and differentiated. These techniques are described with a canbina— tion of actual and hypothetical examples used for illustration. Future tasks in this project will canputerize and autanate many of the tests and diagnostic techniques described in the sections that follow. GOALS OF DIAGNOSTIC TESTING Goals of turbine diagnostic testing include many of the overall project goals. In addition to establishing the efficiency of the turbine cycle the goals include: © Determination of overhaul scope, schedule, and replacement parts that have a long procurement time. © Determination of incremental heat rate curve shape for system dis— patch and generation costing.© Determination of the cycle condition's impact on operator control- lable parameters. Unit Overhaul Cycle evaluation at the level of canponent condition can allow review and adjust= ment of unit overhaul schedules. Tf heat rate degradation is significant or ma- chine safety is questionable, the increased cost of overhauling earlier than otherwise scheduled may be justified. If, as a result of on-line evaluation, it can be determined that overhaul delay is permissible major overhaul expense can be Gelayed or saved. Overhaul schedule acceleration is degendent upon the avai labil- ity of parts. the host unit, for example, was scheduled for overhaul in the Spring of 1985. Tests conducted in April of 1983 indicated that overhaul ac celetation could be justified but lack of a long lead time critical part precluded the change. In order to be most beneficial the ideal test program should be con- ducted on a continuing basis to achieve the followin * Post-Overhaul: Evaluate the degree of performance restoration achieved by the overhaul. Correlate the as-left observed condition to test results for refinement of the evaluation process. © Mid-term: Evaluate the option of overhaul acceleration or delay. Order Long lead time parts. © Pre-Overhaul: Influence the overhaul scope. Correlate tested re- sults with as-found condition for refinenent of the evaluation process. Incremental Heat Rate Tt is possible to determine turbine cycle heat rate within about 0.35% using ASME PIC-6 equivalent procedures. This requires division by a boiler efficiency and correction for auxiliary power to yield a unit heat rate. Boiler efficiency can- ot be measured as accurately as turbine cycle heat rate using any currently known methals. Given the uncertainties of coal floy and heat content measurement it is more accurate to test the boiler and turbine cycle simultaneously so that unit 8-2heat rate and turbine cycle heat rate can be compared on a relative basis over the load range. ‘The more accurate turbine cycle curve can help to explain variations in the unit heat rate curve, Performance deviations may be evaluated by recogniz— ing known changes in perfomance over the load range. This is particularly true for steam generator efficiency for which performance chang can be estimated fran stack loss changes due to temperature, excess oxygen or unburned carbon. The in- cremental heat rate is the first derivative of the firing rate equation and thus, for dispatch, shape assumes greater importance than absolute level. Operator Controllables an Operator Controllable parameter program does not require a fully instrunented monitoring and test capability but the condition of several cycle components directly influences the ability to maintain design values. The following discus- sion is offered to indicate the value of a test program to a better understanding of operator controllables. Steam Temperature. Stean tenperature is a function of boiler type, firing rate, superheater/reheater surface ratio, cold reheat temperature, control methad (tilt, spray or dangers), sootbloving procedures and superheater/reheater duty ratio amang others. As high pressure turbine efficiency decays over time between over- hauls its exhaust temperature (cold reheat) will increase thereby increasing hot reheat temperature lacking any control. After design hot reheat temperature is achieved, control will be imposed to its limit at which point superheat tenpera- ture may be affected in order to hold reheat within design limits. ‘The above im plies that the operator's window of control opportunity may be severely limited by the current repair condition of the turbine and achievable goals may require adjustment.tn machines having a single outer shell enclosing both the high and intermediate pressure turbines, a further major impact results fran the internal shaft packing between these sections. Cycle heat rate and unit capacity are affected by wear of this packing and intermediate pressure turbine efficiency measurenent capability is complicated. teakage through this packing bypasses the high pressure machine and the reheater thereby reducing reheater flow and increasing hot reheat tenpera- ture with similar impact on operator controllable range. Turbine Exhaust Pressure. In a manner similar to steam temperature, achievable back pressure is a function of turbine and condenser condition. As turbines wear their flow passing capability may increase, efficiency decreases and the relation ships of flou, generation and exhaust enthalpy change such that the heat rejected by the condenser increases and achievable exhaust pressure for a given load and anbient condition will increase. ‘This will increase heat rate and will reduce the maxima capability of the unit. All of the above implies that a clearer knowledge of turbine condition is useful input to an intelligent controllable parameters program. REQUIREMENTS FOR COMPONENT CONDITION DETERMINATION In order to estimate machine condition at the canponent level it is necessary to determine significantly more than apparent “enthalpy drop" section efficiency. Major value accrues fran determination of change in flow passing capability of the notmal steam path as well as the leakage paths. Confident usage of component con- dition predictions requires the very best effort be made to validate all critical data. Calibration of sensor and readout system is essential and assumed for pur- Pose of this discussion. ‘The applicable physical and thermaiynanic relationships can be utilized to display the degree of confidence which is proper for a given set of data. Examples of these relationships are: a4© Pressure flow relationships for all turbine stages except first and last are linear and must intersect zero. © ALL turbine stages except first and last operate at essentially ‘constant pressure ratio and therefore as constant volume flow machines over the flow range. © Stage cross sectional area is constant over the flow range for a given set of test data but can change with time between tests. © Flow per unit area at each stage is a function of the square root of state pressure divided by specific volume: /P7 © Crass sectional area is a function of flow divided by flow per unit area: W//P7T © Stage and section efficiencies except first and last are essen- tially constant over the flow range. © High pressure section efficiencies and turbine cycle heat rates for partial arc admission machines have predictable shape within a valve loop. ‘The degree to which raw data or calculated values conform to the above relation ships is an excellent measure of their validity. In order to benefit fran these validation techniques it is necessary to test over the flow range. ALL or sane of the following should be available to the diagnosticiai © Steam pressures and temperatures at turbine inlets, turbine extrac- tions and first stage shell. First stage shell temperature is usu- ally not available. © Main steam flow, feedwater flow, condensate flow Turbine section efficiency (superheated sections) © Shaft end packing leakoff flows © Valve position @ Generator kilowatt output © Turbine exhaust pressure © Turbine bearing vibration and angle (if available) © ‘Thrust bearing temperature© Unit performance history ‘The information of greatest interest will be the most recent, but as will be seen, periodic measurements and reports enable the diagnostician to determine when se- rious changes have occurred and whether they were gradual or sudden changes. This will greatly enhance the value of the conclusions and recanmendations.. Flow Relationships Several flow relationships will provide imediate information regarding the inter~ nal conditions of the turbine, ‘The basic flow equation, By. 1, shom below is correct for circumstances where local stage pressure ratios are constant. If erosion, danage or other problens cause changes to local areas, the equation is no Longer accurate. Still it can be used to recognize that a change has occurred and to help estimate the magnitude of change. w=KaJG (8-1) In By. 8-1 W is the mass flow to the following stage; K is a constant of proportionality; A is the flow area of the stage; P is the absolute pressure; and V is specific volume,Equation 8-1 can be rearranged to solve for flow area: EA (8-2) When the Elow to the following stage, pressure and temperature (W, P and T) are known it is very easy to determine whether changes in area have occurred. For most circunstances the proportionality constant, K, can be ignored. For any given stage after the first stage, effective area renains essentially con stant at all flows and conditions. Fran manufacturer's heat balances, local ef- fective area plots can be mae for subsequent comparison with actual measured values. Figure 8-1 shows a typical area versus flow relationships for the first stage. For a given throttle flow rate, if initial pressure or temperature deviates fran standard, a different effective area must result as the valves open or close to accanmalate the nonstandard condition. Changes are shom as dashed curves in Figure 8-1. the above approach assumes that valve settings are reasonably close to the design settings. Figure 8-2 is a plot of effective area versus Flow for a first stage at standard throttle conditions. the plot is taken fran a real case. By camparing design data with measured data it is apparent that the first valve are atea has increased by about 448, At full load the area increase over design appears to be about 138. Tt should be recognized that the stage flow passing ability is effected not only by area but also by the stage pressure ratio. The stage flow coefficient is also effected by off-angle flow conditions, changed reaction, damage to profiles, etc.WAPTT 0 20 40 60 80 100 120 THROTTLE FLOW (%) Figure 8-1. Typical Stage Flow Relationship for the First Stage If the flow has increased as a result of greater first stage nozzle area, the first stage shell pressure will also be increased. Below the acoustic pressure ratio (approximately 0.55) this will not effect the flow per unit area. But at higher valve openings and pressure ratios the increase in first stage shell pres- sure will greatly restrict the flow increase. As show in Figure 8-3, it may re- quire 2 108 change in area to achieve only a 3% change in flow a8WPT © Design X Measured 0 20 40 60 a0 100 120 THROTTLE FLOW (2) Figure 8-2. Area vs. Throttle Flow for Conditions at Turbine Stop Valve For the full load case shown in Figure 8-2, it can be roughly calculated that to achieve a 13% higher flow capacity by means of increased first stage area alone, an area increase close to 65% would be required. Considering the full load data of Figure 8-2, the following conclusions are possible: © If no other changes have occurred the nozzle area probably has in- creased by about 65%. This is sanewhat unlikely since area in- creases of this magnitude are usually caused by erosion. Such at- tacks are mast often observed to enlarge the first and second valve ports more severely than the other ports. 8-9Change in Throttte| Flow (%) -10 10 Change in First Stage Nozzle Area (%) Figure 8-3. First Stage AREA/FIOW Relationship at Full Load © The areas of the second, and perhaps subsequent stages, may be en- larged so that the first stage shell pressure is lower than normal. If so, the first stage flow can have increased without the neces— sity for such large area differences in the first stage nozzle. © ‘The Leakage area (packings, spillstrips, pressure taps, intershell seals, etc.) may have enlarged to either bypass the first stage or to further lover first stage shell pressure, Which of these conditions is most likely, can be determined by examining pressure/flow relationships ahead of the second stage. Figure 8-4 shows a cam parison of design and measured pressure for various throttle flow conditions. Pressure is plotted instead of effective area since specific volune can rarely be determined for this turbine location. 10FIRST STAGE SHELL PRESSURE \e 0 20 40 60 80 100 120 THROTTLE FLOW (%) Figure 8-4, First Stage Pressure It can be seen in Figure 8-4 that the first stage shell pressure is low (for a given throttle flow) by 17%. This easily proves that a massive first stage nozzle area increase (beyond the 11% known to have occurred) does not exist. The low pressure cannot be explained by an increase in second stage area alone nor by packing rubs alone, since neither can have that much effect. It should be suspected that both are enlarged, and further, main steam inlet seals should be suspected of leakage as well. e1lIn Like fashion, all extraction and inlet locations should be plotted. Bffective area (W/P7Y) is usually the preferred plot where W, the floy to the following stage, is knom. Where flow is not known or is suspect, a substitute plot of /E77 or P versus fi can be utilized instead. Plots of PWV reflect flow per unit area. Where a change has occurred in the area, it will cause an apparent effect in the flow per unit area. Note again, hvever, that while all turbine stages between the first and last normally operate at close to a constant pressure ratio across the Load range, large local area changes will alter the proportionality of /PWV or W/V/P7V with physical area so that the actual area change will not be proportional to the measured W//E7T or VP7V. ‘The VP7V can be plotted against #7 output for comparison with either guaranteed values or with historical ones. If heat rate has deteriorated, for a given iY output, the flow rate will have increased a proportionate amount and VPA should have risen in direct ratio with the flow unless area changes have occurred. So far all the comparisons have been between measured and design values. In most cases, the turbine manufacturers can accurately predict expected internal condi: tions and consequently such plots have considerable diagnostic value. Historical Plots Historical plots, where the necessary data are available, have even greater value. Figures 8-5 and 8-6 depict a history of serious erosion and repair cycles. In the ninth year a sudden loss of area and efficiency indicates an incident of internal damage. The history also shows that with the exception of the first inspection at year 2, each repair cycle failed to recover the performance level existing after the prior inspection. a-12Second Inspection 90 First Inspection = Se a & C a Initial ee Startup 70 0 2 4 6 8 10 12 YEARS AFTER STARTUP Figure 8-5. Hypothetical Plot of Reheat Bowl Conditions Shape Plots Plots made across the load range can contribute to awareness of trouble locations. Figure 8-7 shows expected versus measured high pressure section efficiency as a function of control valve position with measurements taken at valve points. Since the apparent loss at 35% valve position is much greater than at 100% valve position, it can usually be inferred that the first stage is the most damaged Location as opposed to the later high pressure stages. e41392 90 Design IP EFFICIENCY YEARS AFTER STARTUP Figure 8-6. Hypothetical Plot of IP Efficiency PREDICTION OF INTERNAL CONDITIONS Four major cannon areas of trouble can be expected to afflict large steam turbines. These are: © Excessive leakage © Brosion (solid particle) © Internal damage © DepositsHP TURBINE EFFICIENCY (VALVE POINTS) 0 20 40 60 80 100 120 CONTROL VALVE POSITION (%) Figure 8-7. High Pressure Section Efficiency Bach of these potential troubles has characteristics that will enable the diagnos~ tician to differentiate, using the available monitoring results, which or in what cawbination they are present. Table 8-1 is a sumary of the symptans normally present for each of the cammon problens. ‘this brief Listing will help diagnose turbine internal conditions. B15‘suous) peusyean {syxONq papeoyIEA0 szefueg Teyeds “moTg Se WOTADaayp OURS OUT oq Few @seorouy ysn3yy fpeot IYBTT 3e ysaqeaz6 oq Aeu aqueIEN 09 pexeduoo ssoT ADUTOTIIa PEOT aUBIT 32 3899e—I6 aq Aeu s3D0339 AIA sy0933a adeys “uOTANGHAASTP MoT] pebueyD fuoTINGITISTP A/G pebuew asa pebueID 8399339 @prs *eaTes 38217 U3 Aq pay FTUT aya 7e 4SI0M eq Kew opn3rubew uorsore ‘ebeys asayy ye fsyoTUr ouqang ye ATTeNsN 399330 3870M Aouaror339 seseez00q Ayyoedeo moqy saseaz0ur 909339 16907 Aqtenpexb szeadde Aqyens souereaddy 50 apow uoysoxd o[o|a4ed PITOS szebued [eqoads *(sanTes eanTosqe) T ueyy aeyeo26 Atyensn sj (hots UT afueyD queored) 03 (Kous}o1339 Jo sbueyo jusoIed) Jo OFVea 00334 edeus *yen71 uo 3D9gJ9 STITT sey ATTeUION -UOTyeborbes amnzeredusy MoTZ suEsIOy 199554 opts “sayun #oT3-pesoddo uo 1=ybyy aeedde oy Aousyor332 dozp Adyeyzue dI esneo ey “sebe4s MOTs ouNfoa MoT UO ySIOM — ADUsTOTI Ie WoT{D8S SeseeIDeq “UoTIONS gH ayy UF ISOYBTY ST 39379 STyL “Agyoedeo MoTZ SaSEaIOUT S990 [e007 dnqae3s 48373 & uo Ajexy] ozou “Aquappns suaddeq ‘souvzesddy 30 POH Bupjoea poe sdyaastyds uo feud Gurqqna Sse SSoT Jo TED opsOUBETG 18 orc 8-16seanqon33s TeOTUeYDeu PvasCOT JO peUayEay, sze6ueg Teyoods ‘uxe4360 303357509 ON ‘saegsd edeus "Fenny pebuegs fsuoyanqaastp A/G/ pebuey fuoyyEIqFA PeseOIoUL 8390334 @pIs Aayoedeo moty pesvoxsep {Koustorg3e peseezoed '$30033 TeO01 ‘Suoaduts juanbasqns eney Aew - 3dnage ATTensn ‘aouezeeddy Jo @poH ‘ebeued yeusequr asnaug eapsseoxg saebueg Teyoeds sebueyo 9sn7yL THOTS JO UOFIDSITP oy 0} ay750ddo oq Aew *kytoedeo Moly se Yonw se souyy p-£ esvazop Kew Aouayoryze uoyADeg sqoessa_ adeus: uoranqraastp A/a pebuew #38nay3 pebuew 30933 Ops. ‘Wayoedes MOTy peseeissp {Aousyoq3J@ posearseq “sqoey3a Te001 “fuyas eanqesodupy a0(eu 30 UNOpynys © BuTROT{O} SseaID9p 03 aeade ‘Kew {20y323 eseezouy 30u vai ‘pny qubew GuryyuRT-JTes e Yoeor Kew !yenpex6 ATTeNsA souezeaddy Jo 3poH ‘sa FS0deq) (ponuzauos) T-8 orden e-17Example: Excessive Leakage Caused By Shaft Rubbing The effects of excessive shaft leakage, as shom in Figure 8-8, are to decrease measured HP turbine efficiency. IP turbine efficiency will appear to increase on units with opposed flow HP-IP turbines and decrease on units with separate IP sections. The effects will appear suddenly. > . Measured 2 2 [Opposed Flow Type Units es Design gS BE TP TTT TT are Separate IP Sections Sz Sz EE Measured Ee TIME ttm Shaft Shaft Rub Rub Figure 8-8. Effects of shaft Rubbing Infomation supporting the occurrence of a shaft rub includes: © A vibration problem at time that efficiencies changed © A severe thermal problem at time that efficiencies changed: cold start-up, water induction, etc. 8-18© Por Pat the first stage shell has decreased for a given ‘throttle flow Example: Solid Particle Brosicn ‘The subject of solid particle erosion focuses on the damage done by oxide material which spalls off hot boiler and pipe surfaces, then, carried by steam, cuts and scratches away turbine material in high velocity zones. The immediate effects, as indicated in Figure 8-9, are to increase flow areas and decrease efficiency. This is in contrast to the closing effects caused by larger, more dense particles such as weld beads, discussed in Internal Damage. land Extraction Design Measured & e 3 |_Ne. TIME TIME Figure 8-9. Typical Results of Solid Particle Erosion Damage e-19‘The above example indicates that it is generally the admission stage (say, the first high pressure or first reheat stage) where erosion effects are most severe. ‘The damage declines in subsequent stages as the particles becane finer while pass fing through the turbine; also as they are dram off at extraction points. Scetimes the damaging capability of the fine oxide is revived at the crossover so that the first LP stage will be strongly attacked even though the last IP stage is unaffected, Material differences between one stage and another (spillstrips, packings and nozzle materials) may also cause an unexpected local change in erosion damage. At lower temperatures the softer materials employed are more vulnerable to erosion damage. Supporting information of a solid particle erosion problem is a past history of Note that both erosion and rubbing tend to increase flow areas. The key elements for differentiating which problem exists are: © Erosion occurs relatively gradually, rubbing is sudden. © Erosion can cause large effects in the IP section PN functions, rubbing cannot. © Leakage can (sanetimes) cause an apparent increase in IP efficiency. Example: Internal Damage Internal turbine damage has the effect of closing flow areas. It is comonly caused by: 8-20© Steam-carried, hard particles fran the boiler. © Turbine canponents which have broken or vibrated loose. © Foreign material left in the turbine or boiler. ‘The hypothetical example in Figure 8-10 shows a probable damage situation. Both efficiency and flow areas have been decreased by a sudden incident. SECTION EFFICIENCY 1 WAPTV TIME TIME Figure 8-10. Effects of Internal Damage e-21In addition to the pictured effects on efficiency and flow area, internal damage can be accanpanied by vibration, a permanent vibration increase may result. A change in thrust may also be noticed. Not infrequently an incident will have secondary perturbations. the seriousness of subsequent reliability loss is very difficult to appraise in the case of internal damage. Evidence of heavy erosion damage is occasionally a precursor to potential internal damage caused by tenon erosion and subsequent loss of bucket covers. Example: Deposits Chemical deposits (oxides, sulfates, carbonates, silica, etc.) can heavily impact Performance and capacity. In recent years, improved control of feedwater chemistry has generally decreased the magnitude of such losses. Still, losses of several percent for groups of stages is occasionally found. This is especially true where large amounts of reheat desuperheating flow are encountered. Extreme cases of copper oxide deposits in high pressure turbines have severely Limited flow capacity and increased thrust to a condition of forcing a shutdom, Deposits usually accumulate in the stean path in areas of large pressure drop where steam conditions have been lowered to the point where the chemical satura tion point is no longer exceeded. For impulse stages the deposition will be primarily in the stationary blades, on the suction surface, in the vicinity of the throat. In reaction stages, the deposit will accumulate in both rotating and sta- tionary blades. 8-22Figure 8-11 shows a hypothetical deposition case where both efficiency and flow areas have been reduced with time. Thrust may change (not if the deposit is oc- curring equally in double flow sections); vibration is usually not affected. SECTION EFFICIENCY TINE TIME Figure 811. Effects of Steam Path Deposits Sane deposits are sensitive to temparature and may spall off during a shutdown or ‘a large load swing. Same are also soluble in water and have been washed fram the steam path by steam that includes water to either dissolve or break loose the chemical layer by impact. That particular step does not normally provide a last- ing value and includes sane risk. 8-23Deposit presence can occasionally be confinned by temporary improvements in ef- ficiency and flow area following a shutdown; also by a sudden rise in hot well conductivity during such transients. DETERMINING THE SERIOUSNESS OF INTERNAL CONDITIONS A general awareness of stage characteristics is necessary for even rough assess- ment of internal conditions. A typical stage of a low reaction design is shown in Figure 8-12, Rough estimates of changes that take place with changing nozzle area are shown in Figure 8-12. Note that a 50% increase in nozzle area only causes a 10% change in W//PW; a 200% change in nozzle area causes only a 148 change in W/PW: Tt is apparent that a massive attack on efficiency has occurred. In the case of the 100% area increase, the efficiency might even be negative. Bor a given stage, if the flow area is increased, the pressure drop, efficiency and used energy will decrease. ‘The result will be a lowered inlet pressure with a corresponding higher pressure drop on preceding stages causing off-design condi- tions there as well. For a decrease in flow area, the stage pressure drop and energy will increase, but unfortunately the efficiency will decrease. ‘the preceding stages, where pressure drop and energy has been decreased will also perform with lowered efficiency. 8-24Nozzle Area Desien] sox | +100x | -50x a 14° 2° 28° ane: x 28° 55° 30° 14° wi, [0.5 | 0.80 [1.08 | 0.36 a Toots [mec |mrasime (tees war | vo | v0 [114 8 Figure 8-12. Agproxinate Effect of Nozzle Area Change on Stage Characteristics For simplicity, the above calculations neglect changes in pressure drop at the rotating blades. This is a significant factor, since the rotating blade area may also be increased or decreased. It is also serious because of off-angle flow ap- proach to the blade entrance. 8-25Broded Conditions Based partly on previous discussions, it will be apparent that erosion will: © Not be uniform either radially or circunferentially * increase load on previous stages (and decrease efficiency) © Cause mal-distribution of flow © Cause buffeting of buckets or blades © Have unpredictable magnitude if P/V changes 15% or nore © May not be effectively repairable if beyond certain limits The last iten, on repairability, is reflective of recent dbservations that exten- sive ot repeated erosion repairs do not fully recover original perfomance. This results fran degradations of profiles, surface finish, sidewall location, edge thickness and throat size. the unrecoverable loss increases with the severity of erosion damage. Tt is apparent that a local physical increase in area exceeding 50% will have se rious effects on performance. Reliability threats fram thrown bucket covers due to tenon erosion or loss of stationary blade section pieces due to sidewall cuts will increase with continued operation. As noted above, if the erosion damage is confined to one stage (not the first), a local change of /S/V as small as 10% may indicate a 50% area change. If several, successive stages are eroded, then even a 108 change of /P7V at the first stage of ‘the group would not be as serious a threat to efficiency or reliability. This in- dicates why checks ate made of VEWT (or W//B7V} at every practical location, to help the diagnostician estimate how deeply into the group of stages the erosion damage has penetrated. 8-26Where 2 stage is determined to be eroded in excess of 308 area increase, it may usually be assuned that proper repairs will be difficult and time-consuming. Over repeated erosion/repair cycles, losses in performance may justify consideration of new stationary blade rings or diaphragms. This will permit a careful repair of ‘the damaged parts which then can be reinstalled at the next outage. Alternating use and repair cycles of such parts are especially valuable where duplicate units Excessive Leakag All turbines will suffer sane seal damage as a result of shaft rubs. The causes include thermal distortion, rotor bowing, vibration and misalignment. Of these causes, the most damaging is thermal distortion which is usually caused by temperature gradients during start-up but may also be caused by water induction. Measurement of leakage rates at end packings will provide only a good hint of how serious the leakage may be at mid-span. Té other evidence suggests that leakage is a major source of lost performance, it is smetimes possible to estimate what clearance could cause a loss of the cb- served magnitude. Some manufacturers provide tabulations of the Ri loss per mil of clearance for each packing and spillstrip. If such information is not available, it will be difficult to closely estimate leakage losses and outside help may be required. ‘The calculation requires estimation of stage energy, pressure level, reaction, critical dianeters, regain factor, re-entry loss and packing or spillstrip coefficients. 8-27Same leakage losses can be assessed. ‘These include end packings which are shown on heat balances. On the heat balance diagram of Figure 8-13 for example the im Pact of sane leakages can be calculated. igen Figure 813. Heat Balance 8-28Leakage at the end packing of the HP turbine, leakage A in Figure 8-13, bypasses the reheater and the IP stages. As a result, for a given differential flow less eat will be consumed by the unit. The change in heat is calculated in Bq. 8-3. SHEAT = AW, (1520.2 - 1308.8) = 211.4 aW, BTU/HR (8-3) Also, fewer HW will be generated as calculated in Bq. 8-4. tw, (1520.2 - 1349.4) _ A +05 OW, KW (8-4) uw 3412 A secondary loss will result fran a slightly increased flow to the deaerating neater since the enthalpy of the leakage steam is lover than the crossover steam, ‘his can usually be ignored. ‘The change in heat rate can be determined, Eq. 8-5, fran the heat rate equation: 8-29Net Heat Rate 1,846,913 (1461.2 - 453.8) + 1,619,929 (1520.2 - 1308.8) - 211.4 AW, (299,725 - 7312) - .05 AW, (8-5) Leakage through the number 2 (N2) packing, marked leakage B in Figure 8-13, bypasses all stages of the HP turbine, except for the first stage; it also bypasses the reheater, Lost generation and decreased heat input can be determined as shown in By. 8-6 and 8-7. (1420.7 - 1308.8) = 03 Wy KW (8-6) 3412 AHEAT = QW, (1520.2 - 1308.8) = 211.4 AW, BTU/ER (8-7) In addition, the leakage steam, with an enthalpy of 1420.7 will have a chilling effect on the hot reheat steam which has an enthalpy of 1520.2 when mixed in the vicinity of the reheat bowl. This will lower the bosl temperature, reducing the reheat turbine available energy. Reduction in available energy can be determined by canparison of isentropic expansions to the condenser pressure. with the as— sumption that the reheat turbine efficiency does not change, a ratio of the two available energies will provide the percentage effect on reheat section output for a change in N2 packing flow. 8-30Excess N2 packing leakage will also cause an apparent increase in IP section en- thalpy drop efficiency. As shom in Figure 8-14 this effect is more noticeable at Light load than at high load. Since real IP section efficiency is virtually flat across the load range, it is often possible, by trial and error, to determine a magnitude of packing clearance that would explain the deviations of 1P enthalpy Grop efficiency fran that which is expected. te, as, un ‘eg &p th, oy, i er, Fy, _ Ney y Probable Real Efficiency IP SECTION EFFICIENCY ‘© OUTPUT Figure 8-14, IP Section Apparent Efficiency 8-31Where excess leakage is suspected but cannot be clearly assigned to rubbed packing and spillstrips, a variety of other sources should be held suspect. These include: Piston ring seals Nozzle boxes Ast stage pressure taps Shell joint leakage Inner casing hole covers Steam Line bypasses extraction pipe expansion joints (in the condenser) Main stean inlets Blow dom pipes Reheat inlets (double shell design) High pressure extraction pipes In each case, it is helpful to determine how much area would be required to pro- vide a leakage flow of proper magnitude to explain the observed effects on flow or efficiency. This will often eliminate most of the possible sources and enable a concentrated consideration of the one or two remaining possibilities. Internal Damag While any portion of the steam turbine can suffer fran damage, experience shows that the first stage, second stage and first reheat stage are the most vulnerable. 8-32As with erosion problens, it is helpful to know whether one stage or a series of stages have been damaged. Such awareness will help determine local area effects. It would simplify this determination if all stages had pressure taps, but that is a very rare luxury. Any adjacent pressures (and temperatures) should be used. Tf the damage appears to have penetrated several stages, the chserved effects of in- creased pressure drop should be analytically distributed. Experience suggests that even where the danage is shared by several stages, the First of the group will be more severely attacked than the second or third. A broken turbine part will often confine its worst damage to one stage, being rolled around and battered in that stage (sanetimes captured) to be much less of a threat to subsequent ones. Weld beads fram the boiler, however, often retain their damaging potential through several stages, although as they are centrifuged toward the outer periphery, the damage level is decreased. Broken bucket covers or bucket sections will be accanpanied by sudden and con ‘tinued vibration changes. The failure of one cover or bucket is often followed by additional failures. Other potential clues must be investigated, such a © Was there a temporary vibration spike at the time of the episaie? © Has there been a significant continuous change in vibration level? © Is there a history of weld particle damage fran the boiler? Have recent repairs made this problem more likely? 8-33Increased clearance or leakage is normally a minor threat in regard to forced outages. Deposits in the steam path can usually be monitored to permit a rela- tively safe operation. A diagnosis of internal damage is the most threatening of the four cawmon turbine Problems, And where the apparent damage is increasing with time or involves wor- sened vibration, it should be treated with even more respect and attention. Deposits In most cases the harmful effects of deposits are limited to efficiency. on rare cccasions, however, deposits may grow to severely limit flow capacity and to in- crease thrust to a point of danger. Deposit situations are often preceded by condenser leaks, misoperation of chemical injection equipment and unusual magnitude of reheat desuperheating flow. Diagnosis of deposition conditions suggests increased attention to thrust bearing temperatures and water chemistry. CHANGING HISTORICAL TRENDS Undesirable problems harmful to turbine efficiency will nomally recur with devas tating frequency unless very positive steps are taken. Such steps include: © Determining what is the basic cause of the problem. ‘* Iéentifying possible alleviating procedures. © Evaluating best alternatives. © Modifying operating or maintenance procedures to reduce or elimi- nate the problem. 8-34Very strong emphasis is placed on the value of the above activity. For example, the investigator may conclude that erosion damage is proportional to outages for boiler tube repair and that such outages are increasing exponentially with time. ‘the canbination of lost performance, increasing repair costs and forced outage costs may show justification for improved boiler materials in troublesane sections. A second example is vibration trouble occurring during cold starts. Such vibra~ tion frequently causes enlarged packing and spilistrip clearance that severely degrades efficiency. Investigation may show that temperature change rates are too severe, that acceleration through critical speeds can be a problem, or that boiler/turbine controls are inadequate to permit a proper controlled start. SUMMARY ‘Techniques for diagnosing turbine performance problems are improving. Improved sustained turbine capacity and efficiency will provide a healthy payback for these efforts while maintenance and forced outage costs should also benefit. 8-35‘The performance monitoring of power plants has not developed as rapidly as ex- pected since the start of use of canputers in power plants. Section 9 ‘CONCLUSTON eral factors among which are: ‘The computer system originally installed in the Morgantom generating station was designed to monitor performance based on the criteria of detecting change in per formance as a function of time. for this project. The goals Cost of fuel was very low up to 1973 therefore the incentive to im Prove performance was low. ‘The early applications were dissapointing due to poor application of instrumentation and camputer systems. Results were not dependable. ‘There was little correlation between improved performance and im proved reliability and availability. Most utilities did not have an active performance monitoring pro- gram for. the purpose of providing power generation control centers with realistic unit input/output data for the implementation of economic dispatch, capabilities and develop and demonstrate a system that: Monitors power plant performance Pinpoints causes of unit deterioration Helps identify solutions to deterioration of plant performance Enhances the econanic dispatch of the generating units o1 This is due to sev- ‘This is one of the reasons the unit was selected of this project are to extend Morgantom'sPhase I of this project has taken the first steps tovard these goals. ‘This volune of the final report has described the specification and/or procurement of instrumentation, data acquisition and conputational equipment that will be needed as the project continues. The following have been accamplished: © A solid state watthour metering system has been specified and purchased. ‘The meter was built by Scientific Columbus for this project. @ A tap sequencing systen was built to measure turbine exhaust pressure. © A system of accurately calculating turbine valve position fran the ratio of first stage pressure to main steam pressure has been de~ signed and implemented, © A procedure for measuring leakage flow in the packing between the HP and IP turbine sections (N2 packing) has been developed. Design and implementation, while reported here, are technically part of the project's next phase. Use of the flow measurement system will also be part of the project's next phase. © The specification of additional canputational capability, data ac— quisition and display capability was part of the project's first phase. Purchase, implementation and use are part of the project's next phase. In addition, diagnostic procedures to evaluate a turbine's condition have been developed and docunented. It is expected that these will be expanded and modified as the project continues. In sumary, Phase I of this project has provided much of the foundation for the project's future by identifying the improvements requied to obtain the accuracy and repeatability of measured performance parameters consistent with the goals of the project.Appendix A. WATTHOUR METERING SYSTEM HARDWARE REQUIREMENTS AND SPECIFICATION ‘This docunent descrites the functional hardvare requirenents and specifications for a Digital Watthour Metering (DWiiM) System for Potanac Electric Power Campany's (PERCO) Morgantown No. 2 Generating Unit. this system is to provide a measurement. of the electrical power output of the generator suitable for use in the Pover Plant Performance Instrunentation System being developed under the Electric Power Research institute's (EPRI) Project RP1681-1/2153-1. oasecrivas The DWM System shall be able to measure the gross - electrical energy output of the generating unit to .05¢ at the Lowest normal operation of the unit. The sys- tem must be suitable for providing these measurenents as specified by the ASME Performance Test Codes for Stean Turbines, PIC 6-1976, Section 4.10 through Section 4.18 and the Blectrical Measurements in Power Circuits section of the ASME Power Test codes, PIC 196-1955 Section 5 and Section 6. ‘the system shall be capable of measuring the kilowatt hour output and displaying the total Kilowatt hour pulses for each phase. the unit will also have a non-interruptible pulse initiator per phase. In addition to the three individual phase pulse accumulators, the unit shall provide a fourth non-volatile pulse ac- cumulator to indicate total electrical energy output of the generating unit. wlSYSTEM DESCRIPTION The system shall consist of three single phase unidirectional watt/watthour transducers capable of providing an accuracy of +0.05% of reading. Each transducer shall provide two outputs, an analog output proportional to instan- taneous watts and a digital output directly proportional to watthours. ‘The analog output of each phase will go toa rear mounted terminal board. The output terminals provided for each phase shall consist of a ii" signal, termination, "Lo" ignal termination, and a shield termination. ‘the digital output fran each phase should be utilized to provide the following functions: © Bach phase pulse output should provide an input pulse to a front Panel mounted pulse accumulator. * Bach phase pulse output should drive a pulse initiator. © Bach phase pulse output should provide an input to a pulse ac cunulator that records the total pulse outpats of all three phases. Additionally, each of the single phase front panel pulse accumulators shall be able to be similtaneously reset fran the front panel. the reset action will only be enabled when the unit is placed ina test male by a key selector switch. the unit shall also have a power on indicator and an operate/test indicator. FUNCTIONAL SPECIFICATIONS Function High accuracy panel mount digital watthour meter (Wni-directional) a2Blement's 3 single phase, 500 watt Connection 3 phase, 4 wire, grounded neutral Operating Ranges Potential Inputs 69 volts (nominal) ‘Operating Range 0 to 150 volts Overload Capability 250 volts, continuous Burden/element. 0.01 va Surge Withstand Must meet IEEE C37.90a-1974 Current Inputs 4 amperes (naninal) Operating Range 0 to 10 amperes Overload Capability 15A, continuous 50a, 10 seconds/hour 400A, 1 second/hour Burden/element 0.1 Va Surge Withstand Must meet TEEE C37.90a-1974 Frequency Range 58 to 62 Hz Power Factor © lag to unity to 0 lead Calibration Requirement Calibration Input 500 watts/element Calibration adjustment ‘8 minimum range Zero Adjustment None Required Rated Output (RO) at Cal. watts Analog Output 4 to 20 m Load resistance 5 kilohms (max) Compliance 12 we Response time 0.5 seconds to 90% of final value (max) Ripple 0.05% at RO. (maximum) Calibration Adjustment +18 miniman range Zero Adjustment. 58 of 4 ma Accuracy #(0.05% Reading + 0.0058 RO) Digital Output (Bach Phase Panel Accumulator and Pulse Initiator) Watthour output. 8000 counts/hour at 500 watts/element input (standard) (forward power) Accuracy +(0.058 Reading , BF 0.0058 RO) + 1 LSD Pulse Initiator Contact 15 VA at 1K (maximan) or Ratings 150 V (maxima) with resistive load. Contact protection to be provided for inductive loads for 35% a3Contact Life Expectancy Calibration Adjustment Zero Adjustment Normal Operating Conditions Power Input Potential Input Current Input Power Factor Load, Analog Output Anbient Temperature Relative Humidity Auxiliary Power Frequency Influences Affecting Accuracy ‘Temperature External Magnetic Pield Extended Range Operation Load Analog Output Potential Input. 0 to 60 volts and 135 to 150 volts P.E. at 1 A continuous, 109 operations when ‘operated within specified ratings, +18 minimum range None Required 0 to 500 w/element 60 to 135 Volts 0 to 6.58 amp 0 ko 5 kilohns 2s + 10% los 0 908 85 to 135 volts 58 to 62 Hz #0.005%/2C (max imam) Eran -20% to +70% Less than 0.01% of rated ‘output with a magnetic field of 100 ampere-turns produced by a straight conductor six feet long carrying a current of the same frequency and Phase as the applied voltage and positioned in any direc- tion to 10 inches fran the center of the unit. No additional error within analog output load restric tion when power input is 0 to 1000 W/element and cur- rent is 0 - 10 A. Maximum load resistance decreasing fran 5 kilchms to 2.5 kilohms as power in- put increases from 500 to 1000 watts/element (12 volt campliance) . Additional error shall in- crease linearly to not more than 0,018 RO at zero and 150 voits. aInternal Test Timing Interval ‘Timing Interval 10 min. + 1.0 sec. Calibration ‘The DWHM shall be calibrated to primary transfer standards that are traceable to ‘the National Bureau of Standards. The Subcontractor shall provide a certificate of calibration for the DWHM and shall also show proof of traceability to NBS Standard. HARDWARE, DESCRIPTION ‘he DWIM System shall be configured in a switchboard (semi-flush panel mounting) momnting package. The front panel will consist of three single phase pulse accumulators, one Gross Electrical Output pulse accumulator, one States, flush-mounted switch, power on indicator, operate/test switch and reset button as shom in Figure A-L. The voltage and current input leads shall be bottamrear entry with appropriate cable strain relief. The DWIM System pover shall also be input via bottam-rear entry, havever, separate fron the voltage and current inputs as shown in Figure A-2. ‘The DWEM System output signals shall be through feed-through type single row ter- minal blocks that axe mounted on the rear panel. Potential Input The phase voltages shall be input to the DWIM System via Magnum Type 4813 or equivalent feed-through terminal blocks. Interposed between the terminal block and watt/watthour transducer shall be either a Type "PK-2" Potential Test Block or a States "RSI" ‘ype P Potential Test Switch, This device shall be used for dis- connecting the potential inputs without removing the input leads and also to fa~ cilitate calibration, this device shall have a protective safety cover to prevent tampering or accidental operation. ASPigure A-l. Front Panel Seen wet GERBERRERE = OB Figure A-2. Input Leads a6Current_Tnput ‘The phase currents shall be input to the DWHM System via a Magnum Type 4813 or equivalent feed-through terminal blocks. Interposed between the terminal block and watt/watthour transducer shall be either a Type "PK-2" Shorting Current Test Block or a States "RST" Type C Current Test Switch. This device shall have a protective safety cover to prevent tampering or accidental operation. Pulse Initiator Outputs he pulse initiator outputs for each phase shall be Form C (SPDT) mercury wetted contacts. Each phase output shall be output via a Magnum Type 4813 or equivalent terminal block and shall provide a terminal block position for terminating a shield wire, Figure A-3 shows a typical termination. Analog Outputs ‘the analog output for each phase shall be via a Magnum ‘Type 4813 or equivalent terminal block and provide a position for terminating a shield wire. Figure A~4 shows a typical termination. Calibration Input A Magnan Type 4813 or equivalent terminal block shall be provided for inputting a potential voltage and current for the purpose of calibrating the DWHM System. ATs owKm & 2] on B- ®& Fh a : (J ac ——_ FT Figure A-3. Pulse Initiator Output Termination Front Panel Individual Phase Accumulators ‘The front panel pulse accumulators shall be solid state electronic counters. Bach display is to consist of 6-digits and the individual digit is to be a 7 segnent, +8 in height, high intensity LHD, The three pulse accumulators are to have the capability of being able to be reset from a single front panel reset switch. In the normal made of operation the accumulators will continuously update totaling the energy output per phase, In the test mode, the accumulators will total the energy output for a 10 minute periai, holding the value until reset or returned to the operating mode. ABDWHM SYSTEM [ \ “ —___ FI Figure A-4. Analog Output Termination Gross Electrical Output Pulse Accumulator ‘the Gross Electrical Output Pulse Accumulator shall record the total accumulated pulses of the three phases. This pulse accumulator shall be a non-volatile (mechanical) counter with an 8-digit display. ‘the digits shall have a minimum height of .250 inches and a maximum height of .375 inches. The unit shall be to- tally sealed and not resettable nor change count on application or removal of sys~ tem power. (One pulse = 100 KWH) AoQperate/Test Switch ‘The function of the Operate/Test Switch is to allow the three individual phase £ront panel pulse accumlators to be used to dbtain an indication of generator en- ergy output. In the operate maie, the three front panel accumulators will con- tinually operate accumulating the energy output of the individual phases. In the test made, the front panel phase accumulators will stop counting. Upon depressing the reset button, they will reset to zero, and begin accumulating counts for a 10 minute period. At the end of the 10 minute pericd, counters will stop counting and hold their value until reset or placed into the operate made. The switch shall be a maintained-contact cylinder lock select switch with the key removable in the operate position only. Reset Switch ‘The Reset Switch should be a manentary action switch that provides the resetting action for the three individual phase front panel pulse accumulators. Depressing the reset switch when the Operate/Test Switch is in the Test mode should cause the front panel phase accumulator to reset to zero, Releasing the switch initiates a 10 minute interval for accumulating the output pulses. At the end of the timing period, the accumulators will be gated off and an indicator turned on indicating campletion of timing pericd. Placing the Operate/Test Switch in the Operate mode disables the reset switch, turns off the timing interval camplete indicator, and places the pulse accumulators in the continuous maie. Ae10Timing Interval Complete Indicator Indicates canpletion of timing interval when the Operate/Test Switch is placed in the Test mode, Depressing the reset switch clears the indicator and begins the timing interval. on canpletion of the timing interval, the indicator will Light ‘and remain on until the DIM is reset or placed in the Operate mode. Power On Indicator Indicator to show that auxiliary power has been applied to the DWHM System. cattrarron The DMM System shall have the capability of being calibrated through the rear panel calibration terminal block. Either the "PK-2" test blocks or the States ast" Test Switch shall be used to accamplish the calibration. Replacing the nor mal covers with a modified calibration cover for the "PK-2" test block or operat ing the "KSI" Test Switch shall result in the following @ Current transformer secondaries are shorted and potential trans~ former secondaries are opened. © Watt/Watthour transducer current input are connected in series and connected to the current input terminals of the Calibration ‘nerminal Block. © Watt/Watthour transducer potential inputs are connected in parallel and connected to the potential input terminals of the Calibration ‘Terminal Block. © Disables the Gross Output Pulse Accumulator. With the DWHM System in the calibrate made and the calibration potential and cur- rent connected to the calibration terminal block, the DWHM System is ready for calibration. The unit can be calibrated against a known standard or against an— other watthour meter. AelL‘TECHNICAL DEVELOPMENT OF DWHM SYSTEM ‘The subcontractor shall indicate any changes that his design may impose upon the Functional specifications or methad of operations outlined in this guide. Compliance to Functional Guidelines The subcontractor shall provide, prior to purchase, the following documents to de- monstrate campliance of the equipment to the Functional Specifications: © Block Diagram, with signal flow, of proposed DWM. © Preliminary circuit schematic of system. © Preliminary assenbly drawings including: case assenbly drawings, pe board assenbly drawings, and connection diagrams. © DWIM System Specifications as interpreted by subcontractor and a preliminary Operations Manual. ‘The documents that are supplied shall be used as a basis for the purchase of this equipment. RELIABILITY ‘The equipment must have reliability in operating conditions that is equal to and, if possible, greater than the reliability objective. ‘The Supplier shall express his om reliability objective as an overall MIBF (mean time between failures), in hours, applicable to the proposed system, ot else, if appropriate, as an MIBE applicable to the different subsystens that make up the system. In the overall MMBE, the Supplier shall also take into account the con tribution of subsystens supplied by other manufacturers. This MIBE value will serve as a basis for canparison between equipment fron different suppliers. Ael2Stress Factor With respect to equipment of his om design, the Supplier shall see that the fac- tor is kept sufficiently low for it to camply with the MIBF he set as his objective. the Supplier shall give the stress factor value for each of the dif- ferent canponent classes of the equipment proposed, both for “off-the-shelf” equipment and for equipment supplied by other manufacturers. Printed-circuit Boards Single-sided printed-circuit boards and two-sided boards shall have plated holes. ‘The whole board shall be conformally coated except at the canponents' solder points as specified by MIL-1~46058. Al1 boards shall canply with the principles stated in ANSI C83.96-1975. The boards shall be soldered using the "wave" solder- ing method. The acceptability of the quality of the printed circuit is to be Judged according to Standard ANSI C83.107. Plug-in printed-circuit boards shall have position cutouts to avoid faulty connections. Printed-circuit boards shall be made with type @, GF or Gi materials in accord~ ance with Standard MIL-P-13949D. [AL] camponents shall be mounted as suggested in IPC-CM-770 (Guidelines for Printed Circuit Board Component Mounting) al3Integrated Circuits All integrated circuits used in electronic equipment must undergo a 1008 inspection. The Supplier shall supply proof of this check to PEPCO on request. Integrated circuits must be soldered directly onto the board. Sensitivity to Interference In order to reduce the equipment's sensitivity to electramagnetic and radicelectric interference and to ensure its reliability, the Supplier shall fol- Low the recamendations of IEEE Std. 518 and, finally, take the following precautions: a) use of metal covers or enclosures for effective HP shielding. b) Limitation of the equipment's frequency response to the only fre- quencies used. ©) effective short-circuiting to ground of HE signals superimposed on Power supply conductors. 4) appropriate shielding of exterior conductors. e) ensurance that all holes, e.g., those required for instruments, are covered or shielded. £) shielding of printed-circuit boards, especially the back sides. 9) good grounding of circuits, cases, frames and cabinets. hh) appropriate choice of path of cables and conductors. MAINTAINABILITY ‘The Supplier shall express his om maintainability objective as an MITR (mean time to repair) in hours. This must be an intrinsic value for the equipment since it is assumed that a repairman is already on the spot with spare parts and all neces- sary tools required to restore the equipment. Ald‘This MTTR will serve as a basis for camparison between equipment fron different suppliers. ‘The Supplier shall support his arguments when he claims that his equipment can be returned to service very quickly, i.e., within the stipulated time. As regards maintainability, the technical specification must refer to the section labeled Maintenance of this document. Components or Parts All mechanical or electronechanical parts as well as small equipment shall be of the standard North American type and available at all times. ‘The Supplier shall use one kind of parts and canponents so as to facilitate main- ‘tenance and interchangeability and to reduce inventory. Devices with contacts shall be protected fran dust by covers if they are not al- ready installed in hermetically sealed enclosures. ‘The covers shall be made of a nonconbustible material. Inventory Level of Essential Parts With his knowledge of the failure rate of the components, boards, modules or systens, the Supplier shall suggest, for each category, the necessary supply of mechanical, electranechanical or electronic spare parts that should be stocked to maintain availability of the equipment. als‘Spare Parts Provisioning The Supplier shall assure the supply of spare parts; he shall also indicate in the maintenance instructions manual for each spare part at least two (2) addi- tional separate supply sources. Spare Parts Lists These lists must contain divisions for the different types of equipment and sub- divisions for the levels of spare parts, i.e., printed-circuit boards, electronic caponents and mechanical parts subject to wear and tear. The following informa ‘tion must be supplied in the maintenance instructions manual: Printed- Electronic Circuit Mechanical Components Boards Parts Part No. or Serial No. x x x \JEDBC No. or Commercial x Equivalent Function of the Part x x x Name of Manufacturer x x x Quantity Used x x x wplicity of Design ‘The equipment must be of malular design and include plug-in printed-circuit boards and subsystens that are easily accessible and therefore quickly replaceable. However, this requirenent does not relieve the Supplier of his obligation to meet his reliability objective. Al6‘The Supplier shall design circuits that require as few adjustments as possible: he shall leave enough roan to facilitate handling of connectors and to facilitate the removal or insertion of adjacent integrated circuits. The Supplier shall use attachments that cannot be lost and handles for pluggable devices. Soldered or wire-wrapped connections are considered acceptable. No "t™ connections will be accepted. Conductors crossing a joint must be of extra-flexible wire and protected by a sliding sleeve. Identification ‘The Supplier shall see that the different checkpoints are easily accessible and correctly identified. ‘The Supplier shall clearly identify functions at the subsystem and board levels; he shall also identify the components that are mounted on the printed-circuit boards or elsewhere. ‘The Supplier shall use a consistent color code and an adequate labelling methai to facilitate identification. al?Extension Boards In order to facilitate field checks, the Supplier shall supply self-supporting ex- tension boards equipped, if necessary, with extension cables, for all the dif- ferent plug-in beards, even those fron other suppliers, mentioned in his bid. ‘the extension boards must have test points that are easily accessible and correspond to each of the connection terminals. Maintenance Taking into account the deterioration of reliability with time and in order to maintain the equipment's level of reliability throughout its useful Life, the Supplier shall specify in the maintenance instructions manual the nature and fre~ quency of preventive-maintenance operations as well as the precautions that should be taken with respect to corrective maintenance, the various kinds of handling, ete. Any redundant module or circuit must be located on a board that is mechanically separate fram the first one so as to allow maintenance on it without any interrup- tion in service. The removal or insertion of this board must not cause any deterioration in the perfomance of the other cizcuits. the Supplier shall supply a list of all specialized apparatus needed for the maintenance and trouble-shooting of all equipment proposed, including input-output interfaces for lagnostics, when applicable. alsMaintenance Instructions Manual ~ Normal. Content Maintenance instructions manual relating to all equipment supplied must contain the following information at the very least: a) Functional diagrams of the equipment, general specifications and a technical description of the operation of the equipment and circuits. b) Detailed drawings of circuits including overall dimension drawings of canponents on the printed-circuit boards with identification of each canponent . ©) Drawings of connections, junctions and terminals. 4) Diagrams of Layout and overall mechanical dimensions. ©) A List of all parts and camponents that make up the circuits ac- cording to the section Spare Parts List of this document. ‘Tis list must contain the minimm infomation for rapid replacement. Parts and/or “black boxes" exclusive to a supplier must necessarily be covered by a supply guarantee. In order to avoid any ambiguity, the name "black box" will apply to any circuit whose canponents are mechanically inaccessible and whose function is defined in terms of its inputs and outputs (e.g., integrated circuits). ‘The Supplier shall include detailed diagrams of any "black boxes" aswell as the data sheets and any infomation for understanding ‘the operation of the internal circuits. £) Trouble-shooting procedures (including voltage and _ resistance tables, test points, standard oscillograms, etc.) along with a list of special reading instruments. g) The supplier shall provide test points on all circuit boards as per ANSI C82.77-1972. The test points shall be referred to in the trouble-shooting procedures (point £ above) to aid in diagnosis of failure. hh) Recommendations for maintenance as well as checking methods recam mended by the manufacturer. algMaintenance Instruction Manual - Special Contents For equipment using canputers or microprocessors ual must comply with the following requirenente: a) The description, nanenclature and sets of instruments in the source language used must be supplied. b) Flowcharts must be detailed and documented; an explanatory text. must accampany each flowchart, if necessary. ©) Programs must include cannentary. The Supplier shall use structured programing and a structured language. Al of the source programs fran the original equipment manufacturer (0.£.M.) and other basic and application software must be supplied on magnetic tape or disks. ALL utility programs required for the maintenance and trouble-shooting of the system must be supplied and documented. d) The operation and use of all of the 0.B.M.'s manipulators and others must be described in detail. ALI cannon and individual variables of the 0.E.M.'s programs must be defined. Block diagrams of the overall design of the system must be supplied 90 days after the contract is awarded. ) The Supplier shall supply the CPU percentage used to fulfill the function requested on delivery and ultimately. Compliance Input and output terminals for connection with PEPCO installations must be iden tified in compliance with the identification indicated on the wiring plans. Symbols used on drawings and diagrams must comply with the ANSI ¥32.14, ¥32.16 and ANSI ¥32.20 standards. 20 the maintenance instructions man-Human Beror The design must minimize sources of human error in assembly, maintenance and operation of the equipment. The Supplier shall make a systematic effort to make it easy to use the equipment safely and appropriately and to make it very dif ficult to use the equipment incorrectly or dangerously. CLIMATIC CONDITIONS AND ENVIRONMENTAL CONTROL REQUIREMENTS The equipment normally operates in an atmosphere in which the temperature may range fran +5° to +40%, and the relative humidity fram 5 to 95%. Ventilation For equipment installed in enclosures (cabinet or case type, etc.) the Supplier shall see that there is adequate ventilation so that the canponents operate within the allovable temperature Limits and the reliability level of the equipment is maintained, In order to ensure this, there shall be a maximum temperature gradi- ent of 15% between the temperature taken at any point inside and the ambient roan temperature. Natural ventilation techniques (convection) shall be used as much as possible be- fore the use of forced ventilation techniques (fan, circulation of fluid, etc.) may be contemplated. In the event fans are used, the design of the system shall take all existing fac tors into account. the MIBF of the fan shall necessarily form part of the cal~ culation of the overall MIBF of the equipment. a-2L‘he fan or fans shall be self-lubricating and must use the same power source as the electronic equipment. There shall be a manual switch to control the start and stop in addition to the autamatic stop that occurs when the door to the enclosure is opened. A fan failure alarm shall be provided. Air filters shall be placed at air intake points. These filters shall be change- able without disassenbling the equipment. Air intake shall be at ground level and air discharge at the top of the enclosure. Forced ventilation may not be used for telecamunications equigment. Minimum Rules of Thermal Design In view of the fact that temperature has a detrimental effect on the reliability and accuracy of the camponents, the Supplier shall comply with a certain number of minimum rules: a) See that ohmic losses (resistors, semiconductors, incandescent lamps, etc.) are kept down by ‘the use of lov-dissipation can ponents (MOS, etc.) b) Use heavy duty power units of the ferroresonant type when the regulation required is not excessive, or of the ultrasonic type when the cost and reliability are canparable to those of the linear type. ©) Locate high-dissipation camponents together and at a distance fram camponents that are inert but heat-receptive. 4) Improve heat transfers by setting up paths with high thermal con- ductance between sources and radiators. Heat transfer methods such as conduction, convention and radiation should be taken into consideration. ) Ventilate enclosures adequately. a2EQUIPMENT POWER SUPPLY For each piece of equipment, the power source provided by PEFCO is established in the technical specification; the Supplier shall first refer to this specification. AC Power Supply Electronic equipnent may be supplied by an uninterruptible power supply system with the following specifications: Nominal voltage: 120 Vac (#58) Frequency 60 Hz (#1 Hz) Interruption (static): 4 ms Harmonic distortion: 58 Surge withstand: Must meet IBBE C37.90a-1974 Required Information The Supplier shall include the relevant information for the calculation of the permanent and temporary consumption of the equipment. ‘This information shall include the naninal power supply voltages as well as the variations allowed, the load in watts or in amperes in normal and extreme operat- ing conditions, the power factor and the heat dissipation. B23(GROUNDING A campression-type cable lug for 16 to 14 gauge grounding conductor must be mechanically attached to the inside of case. The mechanical and electrical grounds of the equipment must be separated fram one another with the possibility of connecting both grounds by means of a connecting strip provided for this purpose. EQUIPMENT SAFETY ‘The DwHM shall comply with ANSI C39.5-1974 - Safety Requirements for Electrical ‘and Electronic Measuring and Controlling Instruments. The intent of this specification is to ensure reasonable personal protection and protection against damage to the surrounding area in event of equipnent failure. REFERENCES ‘The following publications are quoted in the text of this standard. American National Standards Institute (ANST €37.90-1978 Relays and Relay Systems Associated with Electric Power Apparatus ©37.90a-1974 Guide for Surge Withstand Capability ©83.98-1975 Printed Wiring Board Dimensions and Tolerances - Single and ‘Wo Sided Rigid Boards ©83.107-1975 Guidelines for Acceptability of Printed Wiring Boards ©82.77-1972 Test Point Locations for Printed Wiring Assemblies Ae24¥32.14-1973 Graphic Symbols for Logic Diagrams ¥32.16-1973 Reference Designations for Electrical and Electronic Parts and Bguipment ¥32.20-1975 Graphic Symbols for Electrical and Electronic Diagrams ©39.5-1974 Safety Requiranents for Electrical and Blectronic Measuring and Controlling Instruments Department of Defense Military Specifications Insulating Campound, Electrical (for Coating Printed Circuit Assemblies) 139490-1967 Plastic sheet, Laminated, Copper Clad (for printed wiring) Institute of Printed Circuits (IPC) IPC-CM-770 Guidelines for Printed Circuit Board Camponent Mounting Institute of Electrical and Electronic Engineers (IEEE) STD-518-1977 Guide for the Installation of Electrical Equipment to Minimize Electrical Noise Inputs to Controllers fran External Sources The American Society of Mechanical Rgineers (ASME) PIC6-1976 Steam Turbines PIC19.6-1955 Electrical Measurements in Power Circuits A258Export Control Restrictions ‘Access to and use of EPRI Intellectual Property is ‘granted with the specific understanding and Tequirement that responsibilty for ensuring full ‘compliance with all applicable U.S. and foreign export laws and regulations is being undertaken by you and your company. This includes an obligation to ensure that any individual receiving access hereunder who is not a US. ciizen or permanent U.S. resident is Permitted access under applicable U.S. and foreign ‘export laws and regulations. In the event you are Uncertain whether you or your company may lavifully ‘obtain access to this EPRI Intellectual Property, you ‘acknowledge that itis your obligation to consult with your company’s legal counsel to determine whether this access is lawful. Although EPRI may make available on a case-by-case basis an_ informal assessment of the applicable U.S. export classification for specific EPRI Intellectual Property, you and your ‘company acknowledge that this assessment is solely for informational purposes and not for reliance purposes. You and your company acknowledge that it is stil the obligation of you and your company to make your own assessment of the applicable U.S. export Classification and ensure compliance accordingly. You ‘and your company understand and acknowledge your ‘biigations to make a prompt report to EPRI and the ‘appropriate authorities regarding any access to or use ‘of EPA Intellectual Property hereunder that may be in violation of applicable U.S. or foreign export laws or regulations, (©2008 Elec Power Rasaareh Insite (EPR), ne. A Hgts ‘eaered. Ecc Power Research nts and EPRI ae registred ‘sore mars fhe Elec Power Resor iru Ine. @ Pred on eoeepaprn te Unies Stes of neca ‘The Electric Power Research Institute (EPRI) ‘The Electric Power Research Institute (EPRI), with major locations in Palo Alto, California, and Charlotte, North Carolina, was established in 1973 as an independent, nonprofit center for public interest energy ‘and environmental research. EPRI brings together members, participants, the Institute's scientists and ‘engineers, and other leading experts to work collaboratively on solutions to the challenges of electric power. These solutions span nearly every area of electricity generation, delivery, and use, including health, safety, and environment. EPRI's members represent over 90% of the electricity generated in the United States. International participation represents nearly 15% of EPRI's total research, development, and ‘demonstration program, Together...Shaping the Future of Electricity, ELECTRIC POWER RESEARCH INSTITUTE {FER Hew Avan, Pala AS, Calor S04 1955 «PO Box TOU, Palo ANS, CONST BETOETS = USA 800.318.3714 = 650.858.2121» aakepa@ op.com swe op.com