Finding Lost Megawatts at the Harrison Power Station

Finding Lost Megawatts at the Harrison Power Station y William C. Kettenacker Scientech LLC Douglas L. Eakle Allegheny Energy Supply Introduction In today s climate of rising fuel costs, deregulation, and increased environmental awareness, it is essential that each power station manage their costs and resources to be competitive. One element of this is to determine areas of equipment degradation and how this degradation affects the performance of the plant and its cost of doing business. At the Harrison Power Station, plant data is routinely collected and analyzed using PEPSE (Performance Evaluation of Power System Efficiencies), a steady-state energy balance software program. Using spreadsheet access, the data is fed into PEPSE where it is used to calculate current plant and individual component performance. These performance indicators are then compared to design performance or performance from an earlier period to determine if degradation has occurred and, if so, the cost of such degradation. This procedure was applied to the Harrison Power Station, with the methods and the results presented here. Unit Description The Harrison Power Station is made up of 3 identical supercritical pulverized coal-fired units located near Clarksburg, WV. The units came on line in consecutive years in the early 1970 s. The boilers were manufactured by Foster-Wheeler and the turbines by Westinghouse. Main steam pressure is 3615 psi, main steam flow at full load is 4.9x10 6 lb/hr, and main steam and reheat temperatures are 1010 o F. Each of the 3 units is capable of producing 720 MW at full load. Data and Testing Personnel at the Harrison Power Station conduct periodic performance tests on all the units to find degraded equipment and breaches in cycle isolation. Data is collected using calibrated test equipment, plant instrumentation, and manual gauges read by plant personnel. This data is assembled and organized into 1-hour averages. Approximately 100 points throughout the unit, both turbine cycle and boiler, are obtained and used. Of the data collected in the plant, the following items are used for routine analysis: 1

Turbine Cycle - Main Steam Temperature and Pressure - Feedwater Flow - Gross Generation - 1 st Stage Pressure - Cold Reheat Temperature and Pressure - Hot Reheat Temperature and Pressure - Crossover Temperature and Pressure - Extraction Pressures to All Feedwater Heaters - Outlet Temperatures from All Feedwater Heaters - Drain Temperatures from All Feedwater Heaters - Condenser Pressures (two zones) - Discharge Temperatures and Pressures from Condensate Pumps and FP s oiler - Economizer Inlet Temperature - Air Heater Air Outlet Temperature - Air Heater Gas Outlet Temperature - Stack Temperature - Furnace Exit Gas Temperature - Cold Reheat Temperature and Pressure - Hot Reheat Temperature and Pressure - Main Steam Temperature and Pressure - Ambient Air Conditions - Pulverizer Outlet Temperature - Fuel Flow and Ultimate Analysis The data that is collected and used typically represents at or near 100% power. For proprietary reasons, actual data values are not presented here. Performance Issues No specific performance issues were pending at the Harrison Power Station. Current performance issues are day-to-day or periodic issues similar to other stations. Equipment begins to wear, tubes get fouled or leak, boilers begin to slag, instruments drift, all as part of an overall deterioration of an aging plant. Periodic testing, as discussed in the previous section, is used to find these problems and identify larger ones before they become costly. Finding the Problems Harrison Power Station personnel use the PEPSE software to pinpoint degraded equipment and to quantify that degradation in terms of lost generation, increased heat rate, and lost revenue. PEPSE is a steady-state energy balance software program that 2

simulates the plant. This is done by the development of a plant schematic or model that mimics the actual plant components and their connections. The user builds this schematic in a Windows environment by dragging and dropping plant component icons onto the screen from a component library. This library contains all the components found in any power plant. A PEPSE turbine cycle model and boiler model were constructed to simulate the Harrison Power Station units. ecause they are of identical design, the same base models were used for all three units. These models are shown in Figures 1 and 2. The general procedure for developing these models and the specifics for the Harrison Power Station models are given below. The first step is to develop base models using the original vendor designs. For the turbine cycle, this original design is represented by the turbine vendor heat balance diagrams and thermal kit. For the boiler, the original boiler manufacturer specification sheet and various drawings showing the boiler layout are used. These models are useful for checking the validity of the model layout, model data, and operation, and should be used for studies that relate back to the original design as the base point. ecause they use the fixed data from the vendors, these models are often not flexible enough to respond to changing boundary conditions or changing conditions in the plant. The next step is to characterize the individual components using detailed geometric and heat transfer information. For example, for the condenser this would include tube material, tube inside and outside diameter (or WG), tube length, and number of tubes. Other heat transfer surfaces in the boiler and turbine cycle require similar types of information. y using this detailed characterization, the models can respond to changes in load and boundary conditions and are ideal for studies involving operations changes, design changes, component degradation, and other studies where the original equipment performance is no longer valid. Each of the three units operates slightly differently; therefore, when analyzing actual plant data, separate cases were performed for each unit, each starting with the common base model. A technique in PEPSE known as Special Option 6 was used along with the plant data to establish the benchmark performance (step 1), reduce the test data (step 2), correct the test data to standard (or benchmark) conditions (step 3), and finally perform a series of upgrades (steps 4 N) to find the performance benefit of upgrading each current component s performance to it s benchmark performance. This entire procedure is processed automatically within PEPSE, steps 1 N. The benchmark performance may be based on the original design from the turbine and boiler vendors, the acceptance test, a previous performance test, or some other previous plant condition. The acceptance test is the ideal benchmark, but few plants have this. Results from a past performance test also make a good benchmark, and are usually available. For the Harrison Power Station, the benchmark was based on a turbine vendor heat balance that was developed using data from a previous plant condition and the turbine vendor s guarantee of a new HP turbine performance. 3

Reducing the test data is the technique where the raw data (pressures, temperatures, flows, etc.) are used to define performance parameters for the individual hardware components in the plant. This raw data is used to define efficiencies, heater TTD s, and other parameters used to measure hardware performance. PEPSE has a technique that can look at each piece of raw data and determine if it is good or bad, based on a set of user-defined criteria. If the data does not meet these criteria, it can be tagged, replaced, or clipped using pre-defined limits. For ease-of-use, a Visual asic routine was imbedded in an Excel spreadsheet during this study to automatically pass the raw data from the spreadsheet into PEPSE. The plant data is retrieved from the plant process computer via a spreadsheet, so this process proved to be the quickest and most error-free method of data transfer. In fact, the entire PEPSE analysis was managed and performed from this routine inside the Excel spreadsheet. ecause tests are rarely performed with the same boundary conditions (main steam pressure and temperature, reheat temperature, condenser pressure, etc.) as the condition to which it is being compared, the benchmark, it is necessary to correct the test performance to benchmark boundary conditions. This technique uses the test component performance parameters (from step 2) and the benchmark boundary conditions (from step 1) to determine the plant performance that would have occurred if the test had been performed using benchmark boundary conditions. The result corrects the test results to these standard boundary conditions. However, an assumption must be made here. It must be assumed that the individual component performance parameters, determined from the test data, do not change with changing boundary conditions. How good is this assumption? It depends on how different the two sets of boundary conditions are. The user s judgment should be used here. At the Harrison Power Station, the differences were small because tests are normally performed at nearly the same conditions each time. This is the step that many inquire as to whether the manufacturer s correction curves have been built into PEPSE. The answer is no. ecause PEPSE is calculating the energy balance for the different boundary conditions, correction curves are not needed. The PEPSE heat balance calculation handles all the corrections by the nature of its calculation. In fact, most manufacturers correction curves are generated using heat balance software calculations like PEPSE s. The upgrade steps are the heart of the test data analysis process because they quantify each component s degradation. This is a series of steps where the component performance parameters from the test are replaced, one-at-a-time, with the benchmark performance parameters. Generation improvement, heat rate improvement, and yearly cost savings are reported for each parameter. Table 1 shows the results of a representative analysis of the upgrade step process. THIS DOES NOT REFLECT THE RESULTS AT THE HARRISON POWER STATION. THE RESULTS AT THE HARRISON POWER STATION CONTAIN PROPRIETARY INFORMATION AND CANNOT E SHARED IN THIS PAPER. Using results like those shown in Table 1, Harrison Power Station personnel can immediately pinpoint the problems and degraded areas, and their impact on the plant. Economic decisions can then be based on these results. 4

Once the model is set up, as described above, data from the spreadsheet for any of the three units may be input and results obtained in seconds. The Future These test data analysis models will continue to be used on a regular basis for all three Harrison Power Station units to find degraded equipment, lost megawatts and to assess their economic impact. In addition, the models will be used in many other capacities and studies, such as: Turbine Cycle - Feedwater Heater Replacement - Tube Plugging Effects - New HP Turbines - Cooling Tower Upgrades - Enthalpy Drop Tests oiler - Reheat Design Changes - Cleanliness/Sootblowing Effects - Fuel Changes Conclusion Software tools used to analyze the thermodynamics of steam power plants have become increasingly popular as economic pressures have increased. Market forces such as deregulation, higher fuel prices, and environmental concerns force utilities to be proactive in their search for better performance and lower costs. The Harrison Power Station is taking an active approach in tackling these issues. Windows and Excel are registered trademarks of Microsoft Corporation. PEPSE is a registered trademark of Scientech LLC. 5

Table 1 Test Data Analysis Sample Results Using PEPSE (NOT Representative of Harrison Power Station Results) Upgrade Upgrade Lost MW Lost Heat Rate Cost Case Item (MW) (tu/kw-hr) ($/yr) 1 HP Turbine 0.80 42 $63,066 2 IP Turbine 4.14 188 $282,295 3 LP A Turbine 1.04 78 $117,122 4 LP Turbine 0.54 41 $61,564 5 Heater 5 0.34 5 $7,508 6 Heater 4 0.02 6 $9,009 7 Heater 3 0.04 1 $1,502 8 Heater 2 0.74 56 $84,088 9 Heater 1 0.04 2 $3,003 10 Gland Steam Condenser 0.00 0 $0 11 Air Ejector 0.00 0 $0 12 oiler Feed Pump 0.10 2 $3,003 13 Condensate Feed Pump 0.00 0 $0 14 Air Heater 0.60 40 $60,063 15 Primary Superheater 1.01 62 $93,097 16 Secondary Superheater 3.78 105 $157,665 17 Primary Reheater 4.00 106 $159,166 18 Secondary Reheater 0.60 10 $15,016 19 Economizer 6.03 202 $303,317 TOTAL 23.82 946 $1,420,482 6

Main Steam Gov. Stage HP 3 300 7 htr. bfpt IP IP 5 15 25 60 80 90 100 110 120 150 960 160 170 180 190 210 220 20 50 75 85 95 105 145 115 135 155 950 185 195 205 215 10 112 165 175 200 700 500 505 510 515 520 955 525 530 535 Reheater L N 710 705 715 615 630 610 620 720 625 725 653 652 L 527 810 532 6 htr. 795 797 O C 600 605 654 M LP L 655 E Steam Coils 230 225 540 730 735 545 235 240 D 255 245 250 270 265 550 275 260 555 980 982 G 280 steam coil steam LP 285 295 305 308 290 302 307 560 565 309 312 310 314 LP 313 311 315 M 318 316 320 317 321 319 656 F 785 780 334 330 331 340 332 775 Make-up 765 760 770 G H 990 977 809 975 987 985 802 807 Glycol Htr. 801 799 800 798 650 645 640 635 804 803 350 L H 345 O C 745 815 820 965 970 750 J 755 410 880 872 870 400 395 4 Htr. 995 992 390 385 3 Htr. 380 375 585 580 575 900 910 2 Htr. 370 365 1 Htr. 360 355 905 Gland Stm. Cond 915 740 570 737 920 925 657 E F D N 465 470 450 455 SH Spray 445 7 Htr. 6 Htr. 435 440 430 590 425 865 860 850 840 845 851 RH Spray 825 830 835 FP "Hga Figure 1 Harrison Turbine Model 460 J 855 7

MAIN STEAM HOT REHEAT STACK 230 HR1 310 HR2 400 210 210 220 220 102 100 100 92 90 90 82 80 80 IC 122 120 121 112 110 RF 120 FINISH SH SLOPE TUES AP WALLS CONV ROOF FINAL RH AP SCREEN HRA SCREEN IC 582 125 580 130 130 132 SH 240 240 250 250 SH SPRAY RADIANT ROOF TUES FURNACE 70 70 SH PLATENS 141 140 140 150 150 152 PSH C SH 50 50 60 60 HR1 180 HRA 180 HR2 155 HRA 155 UPPER REAR WW 182 157 40 40 185 VERTICAL RH 185 PSH 160 160 187 162 ECN WING/DIV WW LOWER WW UPPER FRONT/SIDE WW 46 10 20 30 45 10 20 45 30 261 260 260 271 270 270 C 272 HRA to Screen 188 187 189 PSH A 165 RF 165 167 5 6 OTTOM ASH Figure 2 Harrison oiler Model 58 510 AIR 510 COLD REHEAT IN 280 280 290 570 570 FUEL 500 560 500 550 560 550 501 561 PULV (6) 420 425 420 425 430 PRIMARY AIR 426 PA FAN GLYCOL HEATER SECONDARY AIR 460 440 460 470 440 442 430 300 300 RH SPRAY 531 530 290 530 450 450 480 480 FD FANS HRA HRA 190 190 192 170 170 172 HORIZ RH UPPER ECON 195 175 ECN 195 175 197 177 200 200 FEEDWATER IN LOWER ECON 320 520 520 320 322 AIR HEATER PRECIP 332 340 330 350 340 AIR SCR 351 330 491 490 490 540 540 360 410 SCRUER OOSTER FANS 350 370 370 ID FANS 390 390 391 380 380 8