Document Yjg6824QN0agJeZNN1a44yVbN

Meramec 3 Steam Electric-Generating Unit CARL H. RULFS Chief Mechanical Engineer, Union Electric Company. St. Louis,"Wo. Mem. ASME. JOHN K. BRYAN Mechanical Development Engineer, Union Electric Company, St. Louis"Mo. Mem. ASME. C PER COPY C TO ASME MEMBERS This paper describes the principal design features of a 250,000-kw steam electric-generating unit for the Meramec 3 power plant. Basic economic factors, such as fixed charge rate, fuel cost, load factor, thermal performance, and cost of unavailability are dealt with briefly, and the design is viewed largely from the standpoint of such economic considerations. |The Society shall not be responsible for statements or pinions advanced in papers or in discussion at meet- lings of the Society or of its Divisions or Sections, or printed in its publications. discussion is printed only if the paper is pubished in an ASME journal. Released for general publication upon presentation Contributed by the Power Division for presentation at the Semi-Annual Meeting, St. Louis, Mo., June H-18, 1959, of The American Society of Mechanical Engineers. Manuscript received at ASME Headquarters, March 51, 1959. Written discussion on this paper will be accepted up to July 20, 1959. Copies will be available until April 1, 1960. ' US-000261 1/17/02 NUECES I i Meramec 3 Steam Electric-Generating Unit CARL H. RULFS JOHN K. BRYAN Meramec 3 comprises a 250,000-kw electric- aside from the fuel saving that use of reheat generating addition to the Union Electric system additionally provides. In the St. Louis area, placed in commercial oper ation January, 1959. It is neither nuclear nor ECONOMIC ANALYSIS supercritical, but it typifies In many ways the design philosophy behind much of the generating In the case of Meramec 3, over-all project capacity being added to this nation's utility cost was In the order of $40,000,000. The range system. The present paper will make no attempt at covering the plant design in detail. Rather, in cost between a moderately efficient lowest cost plant 1 and a higher efficiency, higher it will attempt to cover the highlights of the cost plant 2 was in the order of $2,500,000. design from the viewpoint of economic-engineer It Is this latter amount that becomes the subject ing considerations. A project of this type has its Inception of engineering-economic analysis as the design is worked out. In making such an analysis, there In a prediction that a utility, some 4 years must be established a number of factors, repre later, will require additional generating capac senting not eternal and Immutable relationships, ity to meet the load demands of Its customers. but best judgement at the particular time as There Is no economic choice here. The decision applied to the particular utility and the par is inherent In the utility's obligation to its ticular project. And since these factors are customers and to the communities that it serves. But there is economic choice involved in the different depending on economic, geographical and other conditions, It follows that individual type of equipment that should be installed to meet the stated load requirement. And It Is in this area of the engineering that various designs may be utilized to fulfill various economic re plant designs will be different. There follows a discussion of some of these basic relationships for the Meramec 3 addition: Fixed Charge Rate on investment was set at quirements. Basically, engineering must begin with a plant design having minimum cost. In dollars per kilowatt, to provide for reliable service together with safe operation, and with adequate provision for maintenance. Additional design features, in general those design features 15 per cent. This Is to say that a $100 Invest ment In plant requires $15 per year In revenue to cover corporate income tax on the revenue, depreciation, and Insurance on the equipment, and 6 per cent net return on the Investment. However, instead of $15 annual revenue, a $15 that Increase investment cost but reduce opera annual reduction In an operating expense, such ting cost, are then adopted as they are shown as fuel or labor, will support the Investment to be Individually attractive. Just as well. Somewhat paradoxically, it can be shown If, for example, $100 additional Investment that for large units the lowest initial cost in a specific feature can be expected to reduce plant for a given electrical output, that is, operating expense by $15 a year, the Investment the cheapest plant on a dollar per kilowatt basis, requires no additional revenue to support It. is also a remarkably efficient plant. It will-- If reduction In operating expense Is more than utilize five or more stages of extraction feed- $15, the Investment Is attractive In that It re water heating. It will employ the reheat cycle, duces total revenue requirements. Fifteen per and It will Incorporate considerably more than the minimum required condenser surface. This is so because the low-efficiency plant entails large Taken as 1800 psi, 1000P/1000F, 5-stage steam and water flows, large quantities of fuel, feedwater heating, 130,000--sq-ft condenser, and high cost of all associated equipment (pumps, tandem turbine-generator. feedwater heaters, piping, boiler, fuel handling and preparation, and so on) . In all but the 2 Taken as 2400 psi, 1050F/1050F, 7-stage smallest units, the lowest investment cost per feedwater heating, 160,000--sq-ft condenser, kilowatt requires the use of reheat, entirely cross-compound large exhaust turbine-generator. 2 RS-000262 1/17/02 NUECES l load factor- Fig.2 Relative value of a given percent improve ment in thermal performance. cent represents an approximate break-even point. the unit throughout its life to an equivalent Fuel Cost 22/ per million Btu, estimated load factor of approximately 75 per cent (curve over the life of the unit. Meramec is close to 4). This, then, is the figure used for calcula a relatively ample and low-cost supply of bitu tions involving fuel consumption of the equipment. minous coal, in which continued mechanization of Incidentally, curve 3 of Flg.l points up mining and availability of both rail and water the important requirement that the unit must be transportation may very well hold fuel cost close capable of both base-load operation in early to today's level over a period of some time. years and cycling or peak-load operation in later Load Factor was considered equivalent to years. 75 per cent of capability over the useful life Heat rate finally realized affects fuel of the unit, for purposes of evaluating fuel sav consumption and in certain cases affects net out ings. This figure may be derived in a number of put of the unit. It is clear from Information ways; Flg.l illustrates one approach. Average already given that average fuel consumption of load factor of the Union Electric system over a Meramec 3 may be considered in the order of period of years has been approximately 64 per $4,000,000 a year. Thus, a 1 per cent fuel sav cent (curve 1). After taking into account system ing represents $40,000 a year, which is the an reserve and generating-equipment availability, nual carrying charge on an additional Investment this means the load factor, or capacity factor, In the plant of slightly over a quarter-of-a- on all generating equipment of the system has milllon dollars. For a 1 per cent reduction in been approximately 57 per cent (curve 2). How heat rate, a quarter-of-a-million dollars initial ever, it is recognized that during early years Investment would be considered "marginal" or of operation, load factor on the Meramec 3 unit "break-even" for Meramec 3. will be high since it will be one of the most Where an improvement in heat rate is non- efficient units available to the system; on the.' uniform over the load range, other factors come other hand, Meramec will be relegated to peaking into play. Fig.2 is aui attempt to indicate the service in later years of its life, and load relative importance of the fuel savings resulting factor will be low (curve 3). The "present from an improvement in performance in various worth" of fuel savings effected during early load rangesj note that a given percentage gain in years of operation is greater than the same sav the 75-100 per cent load range Is worth 13 times ings during later years, and as a consequence as much as the same percentage gain in the 0-25 added weight attaches to the high load factors per cent or the 25-50 per cent load range. If, that are expected during early years of operation. in addition, the Improvement at top load is re The result of taking into account present worth flected in increased output from the unit, an of future fuel savings, therefore, is to increase even greater value must be given it. the load factor arrived at arithmetically for Estimated cctst of unavailability of the RS-000263 1717/02 NUECES 3 EQUIVALENT CAPITALIZED COST, DOLLARS $ 30,000 tem alone. The attraction of the large-size unit, of course, lies in Its reduced cost per 25J000 kilowatt of capacity, reduced fuel consumption per kllowatthour produced, and reduced labor and maintenance costs per kllowatthour of output. 20/300 Additions to the transmission system, the pooled reserves, and associated Interconnections made the 250,000-kw size feasible. 15,000 General arrangement of major equipment, Meramec plant, is shown in Fig.4. Exclusive of 10,000 Wcv /j general plant service areas, units 1 and 2 occupy 215 sq ft of plant area per 1000 kw of rated out 5/300 yS' ..O . <0^ put, exclusive of general plant service areas; unit 3 occupies only 164 sq ft per 1000 kw; while unit 4 will require still less, or only 134 sq ft. These figures illustrate one of the advan tages of larger units. Meramec 3 follows the basis "kllowatthour Fig.3 AMOUNT OF LOAD REDUCTION, KW Equivalent capitalized cost of one day per year outage. factory" approach of the earlier Meramec units, with a minimum of ornamental and architectural treatment, use of corrugated siding for building walls, pipe railings, and so on. Fig.5 illustrates the essential features V of the steam and water cycles, Meramec 3. Steam unit is shown in Fig.3, representing differential cost of generation by the Meramec 3 equipment and by other less efficient equipment of the system. Equivalent capital cost of a one-day outage per at 2000 psi and 1000 F is supplied to the tur bine, is reheated to 1000 F, and flows to the condenser through a diffusing section in the tur bine exhaust and condenser neck. (More about the year on an unscheduled or random basis is in the order of $20,000. Taking into account the longrange effect on system reserve requirements, un availability may cost more like double the amount shown by Fig.3, if it is a forced or random type, rather than a type that can be scheduled off-peak. So much for the economic backdrop against which the Meramec 3 design was carried out. More could be said, but it is hoped that the foregoing gives a sufficiently clear picture of the general relationships involved. Let us now consider the design itself. DESIGN CONSIDERATIONS turbine exhaust arrangement later.) Feedwater heaters number six, of which one is a deaerator. Boiler feed pumps are motor driven, taking their suction from the deaerator and discharging through regulating valves for control of flow. Steam conditions of 2000 psi and 1000 F/ 1000 F were about as low as the turbine manufac turers wanted to go for a unit of this size and type. On the other hand, still higher steam conditions showed inadequate return; namely, 6 per cent gross return for 2400 psi over 2000 psi, and also only 6 per cent for 1050 F main steam temperature over 1000 F. Recall that 15 per cent gross return on an incremental investment is Size. The company's Cahokia power plant, built during the '20s, has as its largest unit, one of 75000-kw size. Venice, built during, the '40s, has a 100,000-kw unit. The first two'units at Meramec were installed in 1953 and 1954, re spectively; each is rated 125,000 kw. But Meramec 3 with its 250,000 kw capacity, had sud denly Jumped to double the size of the preceding units. Why? The answer lies in the fact that between Meramec units 2 and 3, Union Electric had become part of an operating pool including the Illinois Power Company and the Central Illinois Public Service Company, a pool having a capacity of roughly twice that of the Union Electric Sys- considered the breakeven point. Number of feedwater heaters, that is, num ber of stages of extraction heating, is a design variable; additional stages of feedwater heating increase initial cost and maintenance cost, but likewise increase plant efficiency. For Meramec 3, the total of annual investment charges on equipment and operating costs appeared minimal with six heaters. The additional equipment cost of seven heaters more than outweighed the savings in fuel afforded by the more efficient cycle; similarly, the additional fuel cost of a fiveheater arrangement more than outweighed the concomitant savings in equipment cost. Or to put It another .way, going from a five to a six 4 RS400264 1)17/02 NUECES TABLE 1 BOILER PEED PUMP DATA Type Pump he. Plant Year Availability Split-case. Barrel. . . Barrel. . . . 1750 . 1750 . 1500 Venice Venice Meramec 1943 19 48 19 55 96.0 98.0 99-1 heater arrangement showed an attractive return of over 20 per cent on the additional Investment required, while going from a six to a seven heat er arrangement showed only 13 per cent return on the additional Investment. While of course this Is an oversimplification of the analyses which had also to take Into account disposition of drains, terminal temperature differences, degrees of approach to be used In the feedwater heaters-,and placement of pumps In the cycle, it neverthe less correctly demonstrates the method which pointed to a six-heater arrangement for the par ticular conditions for which this plant was de signed. A deaerator Is Included In recognition of the great cost of oxygen corrosion in modern units, and In an attempt to minimise this cost to whatever extent possible. Not only is a deaerator provided, but, in addition, the con denser Is fitted with a deaeratlng-type hotwell. Positive venting of the deaerator to the atmo sphere at all loads from minimum (20 per cent) to full load Is accomplished by connecting It to an extraction point of the turbine which Is under positive pressure throughout this load range. Boiler feed pump. Reliability and avail ability record was summarized in 1955 for recent feed-pump equipment installed in the Union Elec tric system. The purpose was to develop Infor mation to guide In the selection of the size and number of pumps required for Meramec 3. The findings are given In Table 1. It seemed clear that improvements In design and metallurgy had significantly bettered the reliability of the boiler feed pump.^ However, during recent years both capacity and pressure had Increased, so that the boiler feed pump and its supporting electrical Installation was be coming a more and more important consideration in the over-all power-plant cost picture. Avail- See "A Qualitative Evaluation of Equipment Reliability," by I. J. Karasslk, presented at ASME Semi-Annual Meeting In St. Louis, June 1517, 1959. 5 R8-000266 1/17/02 NUECES BOOSTER PUMP Fig.5 Meramec 3 flow diagram. ability figures given, in Table 1 for the latest pumps installed at Meramec Indicate that a third (spare) half-size pump could not be Justified. Nevertheless, at the time this portion of the Meramec 3 design had to be crystallized, it was felt that there was insufficient experience in our own plants, or, for that matter, industry wide, upon which to omit the spare boiler feed pump. Three half-size pumps, with 500-hp motor drives, were Installed. As the design of Meramec 4 developed, it appeared that the use of a single full-size tur bine-driven boiler feed pump offered attractive savings. However, there was concern for the risk involved during the break-in period, the "infant mortality" type failure during this period Jeop ardizing the entire output of the 300,000-kw unit. In the meantime, the generally high re liability on the modern boiler feed pump was verified with additional operating experience. The decision was made, therefore, to use the- spare pump during the initial year's operation of Meramec 3, until start-up problems could be ironed out, and then to make the pump available as a back-up for the large turbine-driven pump of the Meramec 4 unit. And somewhat surprisingly it proved considerably less expensive as well as providing a simpler installation, to move the pump physically into the fourth unit area, and disconnect it entirely from the third unit. This was found to be the case despite the value of slightly higher availability of unit 3 if the spare pump were piped into that uni: as well as unit 4. So the Meramec 3 unit is now operating with three half-size feed pumps, although a year or so from now it will have but two such pumps. Feedwater Cycle. It may be of interest to mention the fluid velocities being used in the feedwater and steam-piping system. For the feedwater cycle, water velocities are about 20 fps at full load and In the main steam piping, 275 fps, both figures somewhat higher than generally used. Here again, these velocities represent an attempt to achieve the best economic balance, using the relationships described, between cost of piping on the one hand, and cost of pressure drop on the other. Pressure drop in these sys tems adversely affects pumping power, resulting in an increase in net plant heat rate and a de crease in net plant output. The turbine-generator1s principal feature is its axial-flow exhaust. This type design together with a related condenser design, has been described.1*' In a conventional turbine, steam exhausting from the last-stage buckets is collected and turned downward into the condenser In the axial-flow design, the exhaust section of the turbine and the condenser inlet hood are pro "Design Features and Development of the Cross Compound Single Flow Turbine with Axial Exhaust, by J. E. Fowler and C. Matney, ASME paper 57-PWR-6. Also, "Condenser at Portland - With Axial Exhaust Turbine, Introduces New Horizons in Turbine-Condenser Operation," by G. T. Jones, ASME paper 37-PWR-14. RS-000266 6 1/17/02 NUECES TABLE 2 SILICA DATA Meramec 1. 2 Meramec 3 Silica in steam, ppm ......................... Drum pressure, psl .............................. Max. silica in boiler water, ppm Blowdown, per cent of output . . . 0.01 1450 1.3 0.25 0.01 2300 0.24 1.3 portioned so as to convert a substantial portion of the steam's leaving velocity into static pres sure. This results in the paradox of steam flow ing from the low pressure (but high velocity) area in the last stage exhaust to the high pres sure area in the condenser inlet. For a given exhaust pressure at the last stage, then, temper ature of condensate returned to the feedwater cycle is higher than for a conventional turbine, improving cycle efficiency. Or stated conversely, it Is indicated that if silica in steam can be kept to no more than 0.01 ppm, significant de posits do not form. Further, the ratio between silica in steam to silica in the water from which steam is formed has been established by several investigators, and is known to increase markedly with increase in boiler drum pressure. An analysis of Meramec 1 and 2 experience and extrapolated requirements for Meramec 3 boiler are given in Table 2. for a given pressure in the condenser, steam is For Meramec 3, the total average annual expanded in the turbine to a somewhat lower pres cost of blowdown, in terms of heat loss, opera sure and develops more work than in a convention ting and maintenance charges on demlnerallzer, al turbine. and fixed charges on demineralizer and heat-re In all other applications of the axial-flow covery equipment, came to a total of $20,000 as turbine to date, the condenser has been mounted estimated. This was the expected, and accepted, at about the same level as the turbine Itself. cost of preventing or minimizing turbine blade However, in the present instance, circulating- deposits. The dual-circulation feature of the water requirements established the elevation of Meramec 3 boiler is expected to reduce this the condenser, and the desire to have No. 3 tur boiler blow-down by some 80 per cent. bine-generator at the same operating level as The dual-circulation design Itself is de No. 1 and 2 established its elevation. As a scribed in a companion paper. Briefly, the de result, the condenser is located at the basement sign separates the circuit from which steam is level, 35 ft below the operating floor. supplied to the turbine and the circuit from The better performance of the axial-flow which blowdown is taken. A higher concentration turbine tended to favor it over the single-shaft of silica can be tolerated in the secondary tandem, although the latter required less ex (blowdown) circuit while still holding very low penditure for auxiliaries, foundation, and in concentrations in the primary (steam-producing) stallation. The axial-flow arrangement also circuit. The reduction in blowdown thus achieved Involved use of two fairly conventional genera made the increased cost of the dual-circulation tors in place of the tandem's single large gen feature attractive for this boiler. erator for which there was no operating experi Coal handling is entirely from river barges, ence. In a comparison with a conventional dou with the capacity of the coal-handling system ble-exhaust cross-compound unit with signifi such that two-shift operation is required to cantly greater exhaust area, the higher cost of supply the three units. Natural gas is available the latter could not be Justified. However, for -on a "dump" basis to carry load on units 1, 2 the larger 300-mw size Meramec 4, the double and a portion of unit 3 load. rather than single l800-rpm exhaust is required. Instrumentation and control is an important Silica deposition on turbine blades has area of design In which the "economic tools" de been, industry-wide, one of the more important operating problems, a recent estimate^ placed scribed at the outset have little application, and yet Is is of such Importance that It must be cost to the electric utility industry of result ing loss in efficiency at $20,000,000 a year. In the recent experience of the authors' company, at least mentioned. We face a perplexing dicho tomy: (1) We must keep the central operator's Job from becoming too complicated, while at the 5 "Observed Effects of Deposits on Steam Turbine Efficiency," by J. Angelo, ASME paper 57-A-llo. same time (2) we must give the central operator enough information and control to handle unusual and emergency conditions. At least, most emer- RS-000267 1)17(02 NUECES 7 gency conditions. Under some conditions we ac cept the proposition that the operator does not have sufficient Information or control, or cannot act rapidly enough. In such cases the equipment is designed to shut Itself down In a safe manner. In developing the control board designs, It has been found of value to build full-scale mock-ups of the boards as they will appear In the central control room of the plant. This was done for Meramec 3, and a number of Improvements and consolidations were effected In the dry-run operation of these mock-up boards. Operating, design and manufacturers' engineers all partic ipated In this development. Much of the Meramec 3 design remains un touched by this paper. However, It Is hoped that features of particular Interest has been covered, and in a way that has served to illustrate some of the basic economic considerations Involved in the design of a modern steam-electric power plant. c c RS-000268 1/17/02 NUECES 8 I1 COMPTCW PRESS INC. i j