COAL-FIRED POWER PLANT PERFORMANCE MONITORING IN REAL-TIME

Sastry Munukutla, Director

and

Robert Craven, R&D Engineer

Center for Energy Systems Research

(Formerly Center for Electric Power)

SUMMARY

Technology for coal-fired power plant performance monitoring in real-time has been developed at the Center for Electric Power, Tennessee Technological University. Customized software, TTURTHRM TM (Tennessee Technological University Real-Time Heatrate Monitor), has been installed at several coal-fired units in the U.S.A. and one has been installed in India. The input data for the software is typically chosen from the already existing data available on the plant computer. All that is needed for the implementation of this technology in any unit is a PC running a Microsoft Windows 32 bit operating system, which communicates with the plant computer. For a typical 500MW coal-fired unit in the US.A., a 2% improvement in heatrate translates into an annual savings in fuel cost of U.S. $1.31 million under given conditions.

INTRODUCTION:

Power plant performance monitoring in real-time affords the opportunity to optimize the performance of a unit continuously. Boiler efficiency, steam cycle heatrate, and net unit heatrate are the three primary performance indices. However, the net unit heatrate, which is the ratio of the rate of energy input (Btu or Kcal per hour) to the net power generated (KW), is the most important parameter. The net unit heatrate, which is the inverse of the unit efficiency, should be minimized. The current industry practice in many parts of the world is to conduct performance tests periodically. In many instances the performance tests are conducted only annually if at all.

Due to the use of modern control technology plant data from various instruments is available on plant computers on a continual basis. Except for alarm function most of the data is rarely used. However, by using part of the data, performance calculations can be made. Since the data is in real-time, the results from the calculation are also in real-time.

The Center for Electric Power, established in 1985, has embarked on the development of real-time performance monitoring technology since 1986. Several original contributions have been made to this field. For the last several years this technology has been successfully implemented in several coal-fired units in the U.S.A. and one unit in India. Currently negotiations are ongoing for further implementing this technology in China, South Korea, Taiwan, Botswana, India, and the U.S.A.

The technical discussion of the real-time performance monitoring methods is presented in the next section. The units at which this technology has already been implemented are listed in the following section. The cost-benefit analysis for a typical 500MW unit in the U.S.A. is presented next. The qualifications of the project managers are discussed in the subsequent section. Some information related to the cost and time needed for implementing this technology in a given unit is presented in the final section.

TECHNICAL DISCUSSION

Calculations for performance monitoring in real-time are based on the well known output/loss method. The output/loss method of determining heatrate is an extension of the heat loss method for the determination of boiler efficiency as prescribed in ASME PTC 4.1[1]. Several published papers describe the output/loss method application to power plant performance monitoring [2, 3, 4, 5, 6, 7, 8].

The basic principle of the output/loss method is that the input to a system is the sum of the output and losses. If the entire coal-fired power plant were to be considered as a system, the input is the product of the coal flow rate (Mc) (1b or kg/hr) and the gross calorific value (H) of the fuel (Btu/lb or Kcal/Kg). The output is the heat transferred to the steam in the boiler, Qs (Btu/hr or Kcal/hr). The losses of the system include, heat loss due to unburned carbon in the fuel, heat loss due to heat in dry flue gas, heat loss due to moisture in the fuel, heat loss due to moisture from burning hydrogen in the fuel, heat loss due to moisture in the air, heat loss due to formation of carbon monoxide and heat loss due to surface radiation and convection. These losses, l, can be estimated per unit mass of fuel. We can, therefore, write the following equation:

M c*H = Q s + M c*l

In the above equation M c can be calculated by knowing H, Q S, and l. The real-time performance monitoring software known as, TTURTHRM TM, performs the calculations based on plant data. Once the coal flow rate is calculated, the following performance parameters can be calculated.

Boiler efficiency:

Steam cycle heatrate

Kwg — gross power generation

Net unit heatrate

K S — Station service power.

In addition to the above parameters other parameters can also be calculated: for example air preheater effectiveness, air preheater leakage, and flow rate of emissions such as CO 2 and SO 2.

The schematic of the system modeled by the output/loss method is shown in Figure 1. It is to be noted that this is a generic example. Deviations from this Such as trisector air heaters, PA fans, etc. can be easily taken into account. The TTURTHRM will be customized for each unit. Typical data needed for the calculations is given below in Table 1.

It is to be noted that certain data described above may not be available. For example cold reheat steam flow and hot reheat steam flow may not be available on the plant computer. However by performing energy balance on feedwater heaters, and extraction steam, these can often be calculated. Additional data such as CO 2 % in the stack and O2 % at air preheater gas side exit could be available. In such a case these data will be used in the calculation procedure.

Figure 1 Schematic of the System Modeled by the Output/Loss Method

Table 1 Data needed

• Primary air temperature
• Secondary air temperature
• Coal/air temperature
• Economizer exit flue gas temperature
• Economizer exit O 2 percent
• Economizer exit Co ppm
• Flue gas temperature at exit of air preheater
• Temperature, pressure, and flow rate of feed water
• Temperature, pressure, and flow rate of main steam
• Temperature, pressure, and flow rate of cold reheat steam
• Temperature, pressure, and flow rate of hot reheat steam
• Temperature, pressure, and flow rate of any “sprays” used
• Ambient temperature
• Relative humidity
• Unburned carbon in ash
• Coal ultimate analysis
• MW generated
• Station service power

Table 2 Sample Static Data Inputs

Table 3 Sample Dynamic Inputs

Table 4 Sample Outputs

UNITS MODELED

• Morgantown Unit 2 (600 MW)
• Pepco , Maryland
• Plant Scherer, Four 880 MW units
• Georgia Power, Georgia
• J. M. Stuart Station Unit 4, 600 MW
• Dayton Power, Ohio
• O. M. Hutchings Station, 65 MW
• Dayton Power, Ohio
• Plant Miller Unit 4, 680 MW
• Alabama Power, Alabama
• Greenriver Unit 2, 60 MW
• Kentucky utilities, Kentucky
• Kingston Unit 9, 200 MW
• TVA, Tennessee
• JPM Unit 300 MW, Genoa Unit 300 MW, Five Alma Units 300 MW
• Dairyland Power Cooperative, Wisconsin
• Monticello Station, Two 500 MW Units
• Texas Utilities, Texas
• Plant Hammond, 500 MW
• Georgia Power, Georgia
• Brandon Shores Unit 700MW
• Baltimore Gas & Electric , Maryland
• Dadri Unit, 500 MW
• NTPC, India
• Clover Station, Two 600 MW Units
• Dominion Generation, Virginia

Field Results

TTU does not perform field tests of power plants, and as such must rely on what performance metrics various plants are willing to share to support the technology.

Equation 1, discussed above, taught how to compute the coal flow rate; so a good check as to whether the calculation is useful would be to compare the computed coal flow rate with a measured coal flow rate. Figure 2 shows the results for one plant and shows a close correlation under normal load conditions. It is of note that the customer requested that a data filter be added such that when a sensor goes out of a prescribed range, a substitute value will be employed. Under this scenario, if the plant completely shuts down the heatrate program will indicate a “normal” heatrate. This explains why the dips in measured coal are not completely matched by equal sized dips in coal flow prediction.

Figure 2 Coal flow rate calculation versus two measurement techniques.

Another indication of the software’s performance would be to compare it with another vendor’s software. One client did such a test, but noted that the other vendor’s software utilized a single static estimate of the coal composition. The solution for the comparison was to have both programs utilize the CEMS predicted coal calculated via TTURHTRM. Figure 3 shows the close matching of the two heatrates calculated by the two different programs over the course of about 12 days.

Figure 2 Coal flow rate calculation versus two measurement techniques.

Another indication of the software's performance would be to compare it with another vendor's software. One client did such a test, but noted that the other vendor's software utilized a single static estimate of the coal composition. The solution for the comparison was to have both programs utilize the CEMS predicted coal calculated via TTURHTRM. Figure 3 shows the close matching of the two heatrates calculated by the two different programs over the course of about 12 days.

Figure 3 Comparison of TTURTHRM heatrate and another vendor using CEMS calculated coal.

A third metric for measuring the usefulness of TTURTHRM would be actual field tests that measure plant performance. A third vendor supplied that data for a series of tests conducted over a wide range of loads where the TTURTHRM performed well as shown in Figure 4

Figure 4 Performance test comparison over a wide range of loads.

BENEFIT ANALYSIS

Real-time heatrate can either be utilized as a tool for operator performance tweaking or as the central tool in a plant optimizing strategy. Several EPRI studies have pointed to 3-4% potential improvement in plants not currently employing optimization. For tweaking, the operator simply notes the heatrate before and after changing a control setting to determine if the change was for better or worse.

Cost Benefit

Benefit analysis will be presented based on the following information:

A 2% improvement in heatrate (decrease from 10,000 Btu/Kwh to 9,800 Btu/Kwh or from 2,520 Kcal/Kwh to 2,470 Kcal/Kwh) will result in savings in fuel cost of U.S. $1.31 million per year.

After implementing real-time performance monitoring, many units were able to improve heatrate by more than 2%. This results in considerable savings in fuel costs.

Other Benefits :

There are several other benefits that can be realized from real-time performance monitoring. In many units coal blending is used. Many times the optimum blend ratio is not-known apriori. In such cases real-time performance monitoring would enable the operators to know as to which blend ratio leads to optimum performance. The same is true with taking decision on which coal to use among several available.

A real-time program is a great “what if” tool. If your plant is considering putting in new emissions cleaning systems such as SCRs or scrubbers but want to consider using “cleaned coal” instead. TTURTHRM can help. Simply acquire sufficient cleaned coal to perform a test and burn it while monitoring with TTURTHRM. If the dollar equivalent to the improvement in heatrate is greater than the additional cost of the cleaned coal and if the improved emissions targets are met then the capital outlook for the emissions cleaning systems are not required. Additional savings in less maintenance for the proposed equipment as well as reduced cost of ash disposal are a bonus.

QUALIFICATIONS OF PROJECT MANAGERS

Sastry Munukutla received Ph.D. in Mechanical Engineering from the University of Iowa. He is Professor of Mechanical Engineering and Director of the Center for Electric Power. He is an Associate Fellow of the American Institute of Aeronautics and Astronautics and Fellow of the American Society of Mechanical Engineers. He has been advisor/co-advisor of 9 Ph.D. and 28 Masters’ students to completion. He has authored more than 160technical publications; which include journal articles, conference proceedings articles, and technical reports. His expertise is in the general area of fluid/thermal sciences with particular emphasis on energy conversion processes.

Robert Craven received a Bachelors Degree in Chemical Engineering from West Virginia University in 1984.  He worked for 15 years as a researcher at WVU within the Center for Industrial Research Applications where his research resulted in co-authoring  6 patents.  His three years with Tennessee Technological University as an R&D Engineer, maintaining the Center for Electric Powers heatrate code has taught him the nuances of working with various Powerplant’s computer systems.  He has 33 published technical papers and many reports and white papers to his credit.

COST AND TIME

The cost of the project depends on several factors such as travel distance, from Cookeville, amount of work shared by the sponsoring agency, and the number of units to be modeled. Therefore, the cost of the project is negotiable.

The time needed to complete the project depends on the cooperation of the personnel from the sponsoring agency. Typically it would take between 4-6 weeks to develop the software, which is ready to be installed in a unit, after a snapshot of the relevant data is received from the unit. In most units the software installation is completed within one working day. The total project period is generally kept at one year even though the software can be installed and be running within three months at the most. The remaining time would be utilized for fine-tuning the software, validating the software by field tests, and implementing any changes suggested by the customer.

REFERENCES

• Steam Generating Units, ASME PTC 4.1-1964
• E. K. Levy, S. Munukutla, A. Jibilian, H. G. Crim, J. Cogoli, A. F. Kwasnik and F. Wong, "Analysis of the Effects of Coal Fineness, Excess Air and Exit Gas Temperature on the Heat Rate of a Coal-Fired Power Plant," Paper No. 84-JPGC-PWR-1, presented at the ASME Joint Power Generation Conference, Toronto, Canada (October 1984).
• E. K. Levy, S. Munukutla, O. Badr, S. Williams and J. Fernandes, "Optimization of Unit Heat Rate Through Variations in Fireside Parameters," presented at the EPRI Power Plant Performance Monitoring and System Dispatch Improvement Workshop, Washington, D.C. (September 1986).
• S. Munukutla, G. Tsatsaronis, D. Anderson, S. Wilson and J. Harris, "FLAPP: A Microcomputer Software for Analyzing the Effects of Key Parameters on Plant Scherer Performance," presented at the EPRI Heat-Rate Improvement Conference, Richmond , VA (May 1988).
• M. Gadiraju, S. Munukutla, G. Tsatsaronis and Ora Scott, "Steady State Performance Simulation Model for J. M. Stuart Station Unit 2," Paper No. 89-JPGC-PWR-22, presented at the ASME Joint Power Generation Conference, Dallas, TX (October 1989).
• E. Levy, N. Sarnac, H. G. Crim, R. Leyse and J. Lamont, "Output/Loss: A New Method for Measuring Unit Heat Rate," Paper No. 87-JPGC-PWR-39, presented at the ASME Joint Power Generation Conference, Miami, FL (October 1987).
• S. Munukutla, P. Chodavarapu and D. C. O'Connor, "On-Line Coal Analysis from Measurement of Flue Gas Components," Paper No. 91-JPGC-PWR-17, presented at the Joint Power Generation Conference, San Diego , CA (October 1991).
• S. Munukutla and F. Khodabakhsh, "Enhancement of Boiler Performance Evaluation Methods Using CEMs Data," presented at the 1995 International Joint Power Generation Conference, Vol. 29, pp. 11-16, Minneapolis, MN (October 1995).