Missing Data Estimation Procedures

1. Parametric Monitoring Procedure for Missing SO 2 Concentration or NO X Emission Rate Data 1.1 Applicability The owner or operator of any affected unit equipped with post-combustion SO 2 or NO X emission controls and SO 2 pollutant concentration monitors and/or NO X continuous emission monitoring systems at the inlet and outlet of the emission control system may apply to the Administrator for approval and certification of a parametric, empirical, or process simulation method or model for calculating substitute data for missing data periods. Such methods may be used to parametrically estimate the removal efficiency of the SO 2 of postcombustion NO X emission controls which, with the monitored inlet concentration or emission rate data, may be used to estimate the average concentration of SO 2 emissions or average emission rate of NO X discharged to the atmosphere. After approval by the Administrator, such method or model may be used for filling in missing SO 2 concentration or NO X emission rate data when data from the outlet SO 2 pollutant concentration monitor or outlet NO X continuous emission monitoring system have been reported with an annual monitor data availability of 90.0 percent or more. Base the empirical and process simulation methods or models on the fundamental chemistry and engineering principles involved in the treatment of pollutant gas. On a case-by-case basis, the Administrator may pre-certify commercially available process simulation methods and models. 1.2 Petition Requirements Continuously monitor, determine, and record hourly averages of the estimated SO 2 or NO X removal efficiency and of the parameters specified below, at a minimum. The affected facility shall supply additional parametric information where appropriate. Measure the SO 2 concentration or NO X emission rate, removal efficiency of the add-on emission controls, and the parameters for at least 2160 unit operating hours. Provide information for all expected operating conditions and removal efficiencies. At least 4 evenly spaced data points are required for a valid hourly average, except during periods of calibration, maintenance, or quality assurance activities, during which 2 data points per hour are sufficient. The Administrator will review all applications on a case-by-case basis. 1.2.1 Parameters for Wet Flue Gas Desulfurization System 1.2.1.1 Number of scrubber modules in operation. 1.2.1.2 Total slurry rate to each scrubber module (gal per min). 1.2.1.3 In-line absorber pH of each scrubber module. 1.2.1.4 Pressure differential across each scrubber module (inches of water column). 1.2.1.5 Unit load (MWe). 1.2.1.6 Inlet and outlet SO 2 concentration as determined by the monitor or missing data substitution procedures. 1.2.1.7 Percent solids in slurry for each scrubber module. 1.2.1.8 Any other parameters necessary to verify scrubber removal efficiency, if the Administrator determines the parameters above are not sufficient. 1.2.2 Parameters for Dry Flue Gas Desulfurization System 1.2.2.1 Number of scrubber modules in operation. 1.2.2.2 Atomizer slurry flow rate to each scrubber module (gal per min). 1.2.2.3 Inlet and outlet temperature for each scrubber module (°F). 1.2.2.4 Pressure differential across each scrubber module (inches of water column). 1.2.2.5 Unit load (MWe). 1.2.2.6 Inlet and outlet SO 2 concentration as determined by the monitor or missing data substitution procedures. 1.2.2.7 Any other parameters necessary to verify scrubber removal efficiency, if the Administrator determines the parameters above are not sufficient. 1.2.3 Parameters for Other Flue Gas Desulfurization Systems If SO 2 control technologies other than wet or dry lime or limestone scrubbing are selected for flue gas desulfurization, a corresponding empirical correlation or process simulation parametric method using appropriate parameters may be developed by the owner or operator of the affected unit, and then reviewed and approved or modified by the Administrator on a case-by-case basis. 1.2.4 Parameters for Post-Combustion NO X Emission Controls 1.2.4.1 Inlet air flow rate to the unit (boiler) (mcf/hr). 1.2.4.2 Excess oxygen concentration of flue gas at stack outlet (percent). 1.2.4.3 Carbon monoxide concentration of flue gas at stack outlet (ppm). 1.2.4.4 Temperature of flue gas at outlet of the unit (°F). 1.2.4.5 Inlet and outlet NO X emission rate as determined by the NO X continuous emission monitoring system or missing data substitution procedures. 1.2.4.6 Any other parameters specific to the emission reduction process necessary to verify the NO X control removal efficiency, (e.g., reagent feedrate in gal/mi). 1.3 Correlation of Emissions With Parameters Establish a method for correlating hourly averages of the parameters identified above with the percent removal efficiency of the SO 2 or post-combustion NO X emission controls under varying unit operating loads. Equations 1-7 in § 75.15 may be used to estimate the percent removal efficiency of the SO 2 emission controls on an hourly basis. Each parametric data substitution procedure should develop a data correlation procedure to verify the performance of the SO 2 emission controls or post-combustion NO X emission controls, along with the SO 2 pollutant concentration monitor and NO X continuous emission monitoring system values for varying unit load ranges. For NO X emission rate data, and wherever the performance of the emission controls varies with the load, use the load range procedure provided in section 2.2 of this appendix. 1.4 Calculations 1.4.1 Use the following equation to calculate substitute data for filling in missing (outlet) SO 2 pollutant concentration monitor data. M o = I c (1-E) (Eq. C-1) where, M o = Substitute data for outlet SO 2 concentration, ppm. I c = Recorded inlet SO 2 concentration, ppm. E = Removal efficiency of SO 2 emission controls as determined by the correlation procedure described in section 1.3 of this appendix. 1.4.2 Use the following equation to calculate substitute data for filling in missing (outlet) NO X emission rate data. M o = I c (1-E) (Eq. C-2) where, M o = Substitute data for outlet NO X emission rate, lb/mmBtu. I c = Recorded inlet NO X emission rate, lb/mmBtu. E = Removal efficiency of post-combustion NO X emission controls determined by the correlation procedure described in section 1.3 of this appendix. 1.5 Missing Data 1.5.1 If both the inlet and the outlet SO 2 pollutant concentration monitors are unavailable simultaneously, use the maximum inlet SO 2 concentration recorded by the inlet SO 2 pollutant concentration monitor during the previous 720 quality-assured monitor operating hours to substitute for the inlet SO 2 concentration in equation C-1 of this appendix. 1.5.2 If both the inlet and outlet NO X continuous emission monitoring systems are unavailable simultaneously, use the maximum inlet NO X emission rate for the corresponding unit load recorded by the NO X continuous emission monitoring system at the inlet during the previous 2160 quality-assured monitor operating hours to substitute for the inlet NO X emission rate in equation C-2 of this appendix. 1.6 Application Apply to the Administrator for approval and certification of the parametric substitution procedure for filling in missing SO 2 concentration or NO X emission rate data using the established criteria and information identified above. DO not use this procedure until approved by the Administrator. 2. Load-based Procedure for Missing Flow Rate, NO X Concentration, and NO X Emission Rate Data 2.1 Applicability This procedure is applicable for data from all affected units for use in accordance with the provisions of this part to provide substitute data for volumetric flow rate (scfh), NO X emission rate (in lb/mmBtu) from NO X -diluent continuous emission monitoring systems, and NO X concentration data (in ppm) from NOx concentration monitoring systems used to determine NO X mass emissions. 2.2 Procedure 2.2.1 For a single unit, establish ten operating load ranges defined in terms of percent of the maximum hourly average gross load of the unit, in gross megawatts (MWge), as shown in Table C-1. (Do not use integrated hourly gross load in MW-hr.) For units sharing a common stack monitored with a single flow monitor, the load ranges for flow (but not for NO X ) may be broken down into 20 operating load ranges in increments of 5.0 percent of the combined maximum hourly average gross load of all units utilizing the common stack. If this option is selected, the twentieth (uppermost) operating load range shall include all values greater than 95.0 percent of the maximum hourly average gross load. For a cogenerating unit or other unit at which some portion of the heat input is not used to produce electricity or for a unit for which hourly average gross load in MWge is not recorded separately, use the hourly gross steam load of the unit, in pounds of steam per hour at the measured temperature (°F) and pressure (psia) instead of MWge. Indicate a change in the number of load ranges or the units of loads to be used in the precertification section of the monitoring plan. Table C-1—Definition of Operating Load Ranges for Load-based Substitution Data Procedures Operating load range Percent of maximum hourly gross load or maximum hourly gross steam load (percent) 1 0-10 2 >10-20 3 >20-30 4 >30-40 5 >40-50 6 >50-60 7 >60-70 8 >70-80 9 >80-90 10 >90 2.2.2 Beginning with the first hour of unit operation after installation and certification of the flow monitor or the NO X -diluent continuous emission monitoring system (or a NO X concentration monitoring system used to determine NO X mass emissions, as defined in § 75.71(a)(2)), for each hour of unit operation record a number, 1 through 10, (or 1 through 20 for flow at common stacks) that identifies the operating load range corresponding to the integrated hourly gross load of the unit(s) recorded for each unit operating hour. 2.2.3 Beginning with the first hour of unit operation after installation and certification of the flow monitor or the NO X -diluent continuous emission monitoring system (or a NO X concentration monitoring system used to determine NO X mass emissions, as defined in § 75.71(a)(2)) and continuing thereafter, the data acquisition and handling system must be capable of calculating and recording the following information for each unit operating hour of missing flow or NO X data within each identified load range during the shorter of: (a) the previous 2,160 quality-assured monitor operating hours (on a rolling basis), or (b) all previous quality-assured monitor operating hours. 2.2.3.1 Average of the hourly flow rates reported by a flow monitor, in scfh. 2.2.3.2 The 90th percentile value of hourly flow rates, in scfh. 2.2.3.3 The 95th percentile value of hourly flow rates, in scfh. 2.2.3.4 The maximum value of hourly flow rates, in scfh. 2.2.3.5 Average of the hourly NO X emission rate, in lb/mmBtu, reported by a NO X continuous emission monitoring system. 2.2.3.6 The 90th percentile value of hourly NO X emission rates, in lb/mmBtu. 2.2.3.7 The 95th percentile value of hourly NO X emission rates, in lb/mmBtu. 2.2.3.8 The maximum value of hourly NO X emission rates, in lb/mmBtu. 2.2.3.9 Average of the hourly NO X pollutant concentrations, in ppm, reported by a NO X concentration monitoring system used to determine NO X mass emissions, as defined in § 75.71(a)(2). 2.2.3.10 The 90th percentile value of hourly NO X pollutant concentration, in ppm. 2.2.3.11 The 95th percentile value of hourly NO X pollutant concentration, in ppm. 2.2.3.12 The maximum value of hourly NO X pollutant concentration, in ppm. 2.2.4 Calculate all monitor or continuous emission monitoring system data averages, maximum values, and percentile values determined by this procedure using bias adjusted values in the load ranges. 2.2.5 When a bias adjustment is necessary for the flow monitor and/or the NO X -diluent continuous emission monitoring system (and/or the NO X concentration monitoring system used to determine NO X mass emissions, as defined in § 75.71(a)(2)), apply the adjustment factor to all monitor or continuous emission monitoring system data values placed in the load ranges. 2.2.6 Use the calculated monitor or monitoring system data averages, maximum values, and percentile values to substitute for missing flow rate and NO X emission rate data (and where applicable, NO X concentration data) according to the procedures in subpart D of this part. 3. Non-load-based Procedure for Missing Flow Rate, NO X Concentration, and NO X Emission Rate Data (Optional) 3.1 Applicability For affected units that do not produce electrical output in megawatts or thermal output in klb/hr of steam, this procedure may be used in accordance with the provisions of this part to provide substitute data for volumetric flow rate (scfh), NO X emission rate (in lb/mmBtu) from NO X -diluent continuous emission monitoring systems, and NO X concentration data (in ppm) from NO X concentration monitoring systems used to determine NO X mass emissions. 3.2 Procedure 3.2.1 For each monitored parameter (flow rate, NO X emission rate, or NO X concentration), establish at least two, but no more than ten operational bins, corresponding to various operating conditions and parameters (or combinations of these) that affect volumetric flow rate or NO X emissions. Include a complete description of each operational bin in the hardcopy portion of the monitoring plan required under § 75.53(e)(2), identifying the unique combination of parameters and operating conditions associated with the bin and explaining the relationship between these parameters and conditions and the magnitude of the stack gas flow rate or NO X emissions. Assign a unique number, 1 through 10, to each operational bin. Examples of conditions and parameters that may be used to define operational bins include unit heat input, type of fuel combusted, specific stages of an industrial process, or (for common stacks), the particular combination of units that are in operation. 3.2.2 In the electronic quarterly report required under § 75.64, indicate for each hour of unit operation the operational bin associated with the NO X or flow rate data, by recording the number assigned to the bin under section 3.2.1 of this appendix. 3.2.3 The data acquisition and handling system must be capable of properly identifying and recording the operational bin number for each unit operating hour. The DAHS must also be capable of calculating and recording the following information (as applicable) for each unit operating hour of missing flow or NO X data within each identified operational bin during the shorter of: (a) The previous 2,160 quality-assured monitor operating hours (on a rolling basis), or (b) All previous quality-assured monitor operating hours in the previous 3 years: 3.2.3.1 Average of the hourly flow rates reported by a flow monitor (scfh). 3.2.3.2 The 90th percentile value of hourly flow rates (scfh). 3.2.3.3 The 95th percentile value of hourly flow rates (scfh). 3.2.3.4 The maximum value of hourly flow rates (scfh). 3.2.3.5 Average of the hourly NO X emission rates, in lb/mmBtu, reported by a NO X -diluent continuous emission monitoring system. 3.2.3.6 The 90th percentile value of hourly NO X emission rates (lb/mmBtu). 3.2.3.7 The 95th percentile value of hourly NO X emission rates (lb/mmBtu). 3.2.3.8 The maximum value of hourly NO X emission rates, in (lb/mmBtu). 3.2.3.9 Average of the hourly NO X pollutant concentrations (ppm), reported by a NO X concentration monitoring system used to determine NO X mass emissions, as defined in § 75.71(a)(2). 3.2.3.10 The 90th percentile value of hourly NO X pollutant concentration (ppm). 3.2.3.11 The 95th percentile value of hourly NO X pollutant concentration (ppm). 3.2.3.12 The maximum value of hourly NO X pollutant concentration (ppm). 3.2.4 When a bias adjustment is necessary for the flow monitor and/or the NO X -diluent continuous emission monitoring system (and/or the NO X concentration monitoring system), apply the bias adjustment factor to all data values placed in the operational bins. 3.2.5 Calculate all CEMS data averages, maximum values, and percentile values determined by this procedure using bias-adjusted values. 3.2.6 Use the calculated monitor or monitoring system data averages, maximum values, and percentile values to substitute for missing flow rate and NO X emission rate data (and where applicable, NO X concentration data) according to the procedures in subpart D of this part. [58 FR 3701, Jan. 11, 1993, as amended at 60 FR 26547, 26548, May 17, 1995; 63 FR 57313, Oct. 27, 1998; 64 FR 28652, May 26, 1999; 67 FR 40459, June 12, 2002]