    DRAFT VERSION FOR COMMENT ONLY - NOT INTENDED FOR FIELD IMPLEMENTATION
                                       
                      DRAFT PERFORMANCE SPECIFICATION ?? 
PERFORMANCE SPECIFICATION FOR HCl CONTINUOUS EMISSION MONITORING SYSTEMS IN STATIONARY SOURCES

1.0 Scope and Application.

0.1 Analytes. This performance specification is applicable for measuring gaseous concentrations of HCl on a continuous basis in the units of the applicable standard or in units that can be converted to units of the applicable standard(s) regulated under Title III of the 1990 Clean Air Act Amendments. The specification includes requirements for initial acceptance including (1) instrument accuracy and (2) stability assessments.

0.2 Applicability. A source that demonstrates their continuous emission monitoring system (CEMS) meets the criteria of this performance specification may use the system to continuously monitor gaseous hydrogen chloride (HCl). If your HCl CEMS is capable of reporting the HCl concentration in the units of the existing regulation, no additional CEMS components are necessary. If your HCl CEMS does not report concentrations in the units of the existing standard, then other CEMS components (e.g., oxygen, temperature, and pressure) are necessary to convert the units reported by your HCl CEMS to the units of the standard. This specification is not designed to evaluate the installed CEMS performance over an extended period of time. The Administrator may require the operator under Section 114 of the Act, to conduct CEMS performance evaluations at other times besides the initial test to evaluate the CEMS performance. See 40 CFR Part 60, §60.13(c).

1.0 Summary of Performance Specification.

0.1 This specification covers two basic topics: (1) the instrument requirements for an HCl CEMS and (2) the requirements that each HCl CEMS must meet during the performance specification test.
 
0.2 The technology used to measure gaseous HCl must provide a distinct response and address any appropriate interference correction(s). It must accurately measure HCl in a representative sample (path or point sampling) of stack effluent.

0.3 Relative accuracy (RA) may be established against a reference method. Alternatively, accuracy tests require a secondary source of certified calibration standard which is spiked through the sampling probe.

0.4 Calibration drift and calibration error tests are also required.

3.0 Definitions.

 3.1 Batch Sampling occurs when a gas cell is alternately filled and evacuated and a measurement of each cell fill is performed.

 3.2 Calibration Cell is a gas containment cell that is placed in the optical path of a CEM instrument to perform precision and bias checks. The cell can be evacuated and purged to remove calibration gas. When charged it contains a known concentration of HCl calibration gas.

 3.3 Calibration Drift (CD) is the difference in the CEMS output readings between the established reference value and the output after a stated period of operation during which no unscheduled maintenance or adjustment took place. A separate CD determination must be performed for the pollutant and diluent analyzers in concentration and for the CEMS in the units of the emission standard.

 3.4 Calibration Error (CE) is the mean difference between the concentration indicated by the CEMS and the known concentration generated by a calibration source, divided by the span, when the entire CEMS, including the sampling interface is challenged. A CE test is performed to document the accuracy of the CEMS.

 3.5 Centroidal Area means a concentric area that is geometrically similar to the stack or duct cross section and is no greater than 1 percent of the stack or duct cross-sectional area.

 3.6 Certified Gas Standard: is a EPA protocol compressed gas standard with known concentration certified by the supplier to be accurate to  2 percent of the value. Note: If protocol gas standards are not available vendor supplied compressed gas standards certified to be accurate to  2 percent of the value may be used.

 3.7 Constant emissions occurs when the composition of the emissions remains stable within 5% over a measurement cycle so that a single analytical procedure and instrument algorithm or program is suitable for analyzing all of the sample spectra.

 3.8 Continuous Emission Monitoring System (CEMS) means the total equipment required for the determination of a gas concentration or emission rate as a continuous operation. The sample interface, pollutant analyzer, diluent supply/analyzer, and data recorder can be major subsystems of the CEMS.

 3.9 Continuous Operation is the time between periodic maintenance when an instrument and sampling system operates without user intervention, continuously samples flue gas, records measurement data, analyzes the data for HCl, and saves the results to a computer file. User intervention is permitted for initial set-up of sampling system, initial calibrations, and periodic maintenance.

 3.10 Continuous Sampling occurs when sample gas is continuously flowing through a measurement path. Measurements of the flowing sample are made at regular intervals.

 3.11 Data Recorder is the portion of the CEMS that provides a record of analyzer output. The data recorder may record other pertinent data such as effluent flow rates, various instrument temperatures or abnormal CEMS operation.

 3.12 Diluent Analyzer means that portion of the CEMS that senses the diluent gas (e.g., O2) and generates an output proportional to the gas concentration.

 3.13 Dynamic Spiking is the procedure used to document the precision and bias of the CEMS by quantitatively spiking a certified gas into the measurement system while sampling.

 3.14 Gas Cell. For extractive CEMS, a gas containment cell that can be evacuated. It contains the sample enabling the optical beam to pass from the instrument through the sample, and to the detector. The gas cell may have multi-pass optics depending on the required detection limit(s) for the application.

 3.15 High-level drift is the absolute difference between a high-level calibration gas and the CEMS response over a specified period of time divided by the span in units of the applicable standard.

 3.16 Independent Measurement means two independent one minute averages of CEMS data taken during sample gas analysis separated by a complete flush of the measurement system with sample gas. 

            3.16.1 For continuous flow extractive measurement systems the interval between independent measurements must be sufficient to pass at least 7 measurement cell volumes of sample gas through the system. Estimate the residence time empirically by: (1) filling the system to ambient pressure with a known analyte concentration gas standard, (2) measuring the gas standard, (3) purging the system with zero gas at the sampling rate and performing a measurements until the gas standard is less than or equal to twice the system detection limit (LOD). 

            3.16.2 For in-situ optical systems, independent measurements are separated by the time required to obtain a single one-minute average measurement.

 3.17 Interference is a compound or material in the sample matrix whose characteristics bias the instrument signal (positively or negatively). The interference may not prevent the sample measurement, but could increase the analytical uncertainty in the measured concentration.

 3.18 Instrument Measurement Range is the range of concentrations the instrument can reliably measure from the lowest concentration to the span.

 3.19 Intercept is the average instrument response at zero HCl concentration.
 
 3.20 Integrated Path Sampling CEMS (IPS-CEMS) is a CEMS that samples the source effluent along a path greater than 10 percent of the equivalent diameter of the stack or duct cross section.
      
 3.21 Level of Detection (LOD) is defined by the procedure in Section 15 of Method 301. The LOD is specific to the measured test gas matrix. The LOD is a calculated value and must be determined before conducting a test to ascertain whether the technique can meet the sensitivity requirements of the CEMS data.
 
 3.22 Linear Regression is a method of finding a mathematical formula that relates two variables (e.g., concentration and instrument response) that are linearly related.

 3.23 Path Length. For extractive CEMS, the distance in meters of the optical path within the gas measurement cell. For cross stack IPS-CEMS, the distance in meters of the optical path that passes through the sample gas.

 3.24 Point Sampling CEMS is a CEMS that samples the source effluent at a single point.

 3.25 Pollutant Analyzer means that portion of the CEMS that senses quantifies and generates an output proportional to the stack gas HCl concentration.

 3.26 Reference Analyzer is an instrument with a separate sampling and sample conditioning system that can be used for comparison measurements such as stratification tests.

 3.27 Reference Value (R) is the certified concentration of a vendor supplied gas or the known concentration of a generated calibration gas in ppmv. For in-situ integrated path (IP) CEMS, the reference value will be calculated such that, when introduced into the optical path of the CEMS, will provide the equivalent concentration at the stack temperature, pressure, and line strength factor (at the stack temperature). The line strength factor is derived from the HITRAN database.

 3.28 Relative Accuracy is the absolute mean difference determined by the CEMS between a) the gas concentration spiked through the sampling probe and the certified value of the spiked gas divided by the certified spike gas concentration or b) the reference method (RM) plus the 95 percent confidence coefficient divided by the average RM value or the applicable emission standard.
      
 3.29 Relative Response Factor (RRF) the response of an instrument detector to an analyte compared to the known concentration of a standard gas.

 3.30 Response Time (RT) is the time interval between the start of a step change in HCl concentration through the measurement system and when the output of the CEMS reaches 95 percent of the final value.

 3.31 Sample Analysis means interpreting a CEMS signal to obtain a sample concentration. This is usually performed by a software routine supplied by the instrument vendor.

 3.32 Sample Interface is the portion of the CEMS used for one or more of the following: sample acquisition, sample transport, sample conditioning, or protection of the analyzer from the effects of stack gas.

 3.33 Sampling System. For extractive systems, equipment used to extract sample from the test location and transport the gas to the analyzer. Sampling system components include probe, heated line, heated non-reactive pump, gas distribution manifold and valves, flow measurement devices and any sample conditioning systems. For IPS-CEMS, all equipment used to take in-situ gas measurements from the measurement probe to the optical signal analyzer.

 3.34 Sampling Time. In batch sampling - the time required to fill the cell with flue gas. In continuous sampling - the time required to make an independent measurement. 

 3.35 Sampling Rate (Qs) for an extractive CEMS is the rate of stack gas drawn through the sampling system.

 3.36 Slope is the rate of change of Y (the instrument response) relative to the change in X (the calibration standard concentration).
      
 3.37 Span is the upper limit of the intended measurement range (highest calibration point). Multiple instrument ranges with spans may be necessary to measure concentrations encountered during normal process operation.

 3.38 Standard Addition Calibration Cell is a cell that is used to add a known concentration of certified gas into the measurement (optical) path of an instrument. 

 3.39 Stratification means the difference in effluent concentration in a duct, when comparing a reference measurement at the centroid of the duct to traversed measurements.

 3.40 System Volume is the volume of the sampling lines plus the instrument measurement chamber volume.

 3.41 Zero drift means the change in the absolute difference between a zero gas concentration and the CEMS response over a specified time divided by the span in units of the applicable standard.

4.0 Interferences. Several compounds, including water, carbon monoxide, and carbon dioxide, are potential interferences with certain types of HCl monitoring technology. Individual vendor equipment often includes procedures to correct or otherwise deal with interferences.

5.0 Safety. The procedures required under this performance specification may involve hazardous materials, operations, and equipment. This performance specification may not address all of the safety problems associated with these procedures. It is the responsibility of the user to establish appropriate safety and health practices and determine the applicable regulatory limitations prior to performing these procedures. The CEMS users should consult instrument operation manuals, compressed gas safety requirements and other safety guidance such as material safety data sheets for specific precautions to be taken.

6.0 Equipment and Supplies. Equipment and supplies for HCl CEMS will vary depending on the measurement technology and equipment vendors. This section provides a description of the equipment and supplies typically found in one or more types of HCl CEMS.

      6.1 Sample Extraction System: The portion of the CEMS that collects and transports the sample to the sampling module. The extraction system must deliver a representative sample to the measurement instrument. The sample extraction system typically consists of a sample probe and a heated umbilical line.

      6.2 Pressure Regulation and Conditioning Module.
The pressure regulation and conditioning module is designed to remove free particulates from the gas stream prior and provide a constant pressure sample gas stream to the CEMS analysis module.
      
      6.3 Analysis module is the portion of the CEMS that senses quantifies and generates an output proportional to the stack gas HCl concentration.

      6.4 Diluent analyzer is the portion of the CEMS that quantifies stack gas concentrations of oxygen or carbon dioxide (CO2). For systems with a multi-component analyzer, the same analyzer quantifies all measured gases.

      6.5 System Controller is the portion of the CEMS that provides control of the analyzer, sample probe, pressure regulation module and the sample interface.

      6.6 Data recorder is the portion of the CEMS that provides a record of analyzer output. The data recorder may record other pertinent data such as effluent flow rates, various instrument temperatures or abnormal CEMS operation.
 
            7.0 Reagents and Standards.

              7.1 A copy of the supplier's certificate of analysis must be provided for each gas cylinder.

              7.2 Certified gas standards must be used within their certification period.

Note: See Table 1 for reference gas concentration ranges.

  

        Table 1. Performance Specification Test Calibration Gas Ranges



                    HCl Calibration Gas Concentrations [a]
                                     Test
                                     Units
                                     Zero
                                   Mid Level
                                  Level High
Calibration Drift and 
                                   % of Span
                                    0 - 30
                                      NA
                                    50-100
Calibration Error Test 
                                  % of Span 
                                    0 - 30 
                                    30 - 50
                                   50 - 100
Relative Accuracy Spiking 
                                  % of Span 
                                    0 - 50
                                   25 - 75 
                                   50 - 100
a  Calibration gases must be certified to  2% bias.
b  The response time test for multi-component instruments may be based on any measured analyte.
 
            8.0 CEMS Installation and Measurement Location Specifications

              8.1 CEMS Installation. 

  0.6.1 Install the CEMS at an accessible location where HCl measurements are representative. Select a representative measurement point or path for monitoring in locations that avoid gas stratification as described in Section 11.5.

  0.6.2 You may locate the instrumental analyzer portions of your HCl CEMS any distance from the sampling point provided the transmission efficiency from the sampling point to the instrumental analyzer point meets the criteria outlined in Section 11.6. 

              2.1 Measurement Location.
 
  0.6.3 The sampling location of the HCl CEMS should be downstream of all pollution control equipment at a position where the HCl concentration is directly representative of total emissions from the stationary source.
 
  0.6.4 You may use either point or integrated path sampling technology.
 

                              0.6.4.1 For a path integrated sampling CEMS the effective measurement path must be:

                                    0.6.4.1.1 Totally within the inner area bounded by a line 1.0 meter from the stack or duct wall, or 

                                    0.6.4.1.2 Have at least 70 percent of the path within the inner 50 percent of the stack diameter or duct cross section, or be centrally located over any part of the stack or duct centroid area.

            7.0 Quality Control.
 
              7.1  The CEMS operating range (zero to span) must encompass the response of the HCl CEMS for all expected HCl concentrations, including the applicable emission limit, if practicable.

              7.2  All CEMS must operate continuously without repairs, unscheduled maintenance, or non-routine adjustments during the performance specification tests to determine calibration drift, error, and RA.
 
            8.0 Calibration and Standardization

              8.1 The source owner or operator is responsible to calibrate, maintain, and operate the CEMS properly.

              8.2  The calibration error test and seven day drift test must be performed by flooding calibration gas through the sampling probe for this performance specification.

              8.3 Limit of Detection (LOD) Determination. You must determine the instrument specific LOD following the procedure in Section 15 of Method 301. This is accomplished by measuring the standard deviation of a calibration gas standard within 2 times the expected LOD or by measuring the noise limited instrument signal with no additional HCl spike. You may not use a CEMS whose LOD is greater than 20 percent of the regulatory limit or the intended use of the data.

            9.0 Performance Specification Test Procedure.

              9.1 The performance testing is used for initial validation upon installation of a new system. After completing the CEMS installation, setup and calibration you must complete the test procedures in this section.

              9.2 Validation is performed by meeting the RA requirements against a reference method or by dynamically spiking certified HCl gas standards through the sampling probe. Note: Figure 3 provides an overview of a typical dynamic spiking arrangement.

              9.3 Pretest Preparation

  0.6.5 Prior to the start of your initial performance specification tests, you must ensure that the HCl CEMS is installed according to the manufacturer's specifications and the requirements in Section 8.

  0.6.6 After the installation, we recommend you check the zero and high-level drift as described in Section 11.8 and the calibration error as described in Section 11.7 to verify that the instrument is functioning properly.

              6.1  CEMS and Data Recorder Scale Check

  0.6.7 Record and document the measurement range of the HCl CEMS. The CEMS operating range (zero to span) and the range of the data collection device must encompass all expected HCl concentrations and the applicable emission limit, if practicable. The CEMS and data collection device output range must include zero and the span value.

              7.1 Stratification Test 

A stratification test must be conducted when the facility is operating during normal operation. The purpose of this test is to verify that excess stratification of the target pollutant does not render sampling point of the CEMS non-representative.
 
  0.6.8 If an isokinetic reference method (RM) is used for the RA testing required in section 11.8.3 you may determine whether effluent stratification exists using a CEMS while traversing the stack following the requirements in 40 CFR Part 60 Method 1.

                              0.6.8.1 One probe, located at the stack or duct centroid, is used as a stationary reference point to measure the change in the effluent concentration over time (Ci).

                              0.6.8.2 The second probe is used to determine the concentration over the traverse points specified in Method 1 (Cref). The traverse points are sampled for a minimum of five minutes at each point or as required by the reference method. You may test for stratification using either:
 
a)	Velocity tests as described in Method 2, 
b)	HCl concentrations using Method 320 or
c)	an acceptable alternative (i.e., Method 10 for CO or Method 7E for NOx).

                              0.6.8.3 Select traverse points that assure acquisition of representative samples over the stack or duct cross section" (40 CFR 60 Appendix B PS2 Section 8.1.3.2).

                              0.6.8.4 To determine stratification, compare the fixed reference CEMS effluent concentrations (Ci) to the reference method effluent concentration, (Cref), using Eq. 1. You must identify and use a fixed sampling location where the percent stratification differences, St, is within 10% of the reference method. 

  0.6.9 If dynamic spiking is used for the determination required in section 11.8.4 you may determine whether effluent stratification exists using a dual probe system and two identical HCl CEMS.  One probe is used to make HCl measurements at a fixed reference point and the second is used to make HCl measurements of effluent concentration at each traverse point as specified in 40 CFR Part 60 Method 1. Both systems must be prequalified following the requirements in Section 11.5 before testing.

                              0.6.9.1 Install a temporary CEMS equivalent to the permanent installation. The second CEMS must allow sampling while traversing the stack or duct following 40 CFR Part 60 Method 1.
                              0.6.9.2  Use the HCl CEMS and the first probe to make measurements at the CEMS fixed reference point (e.g., at the stack or duct centroid).
 
                              0.6.9.3 Use the temporary equivalent CEMS and the second probe to make measurements of effluent concentration at each traverse point as specified in 40 CFR Part 60 Method 1.

                              0.6.9.4 To determine stratification, compare the fixed probe effluent concentration, (Ci) to the measurements of the traverse points, (Cref) using Eq. 1. You must identify and use a fixed sampling location where the percent stratification differences, St, is within 10% of the reference method measurement. 

  6.6 Response Time Determination

1.0.1 Determine the average upscale and downscale response times from three repetitions of each test. You will report the greater of the average upscale or average downscale response times as the response time for the system.
 
      1.0.1.1 Determine the upscale response time by injecting zero gas into the measurement system at the probe inlet.

      2.0.1.2 When the system output has stabilized (no change greater than 1 percent of full scale for 30 sec), introduce an upscale calibration gas and wait for a stable value.

      3.0.1.3 Record the time (upscale response time) required to reach 95 percent of the final stable value. 

      4.0.1.4 Next, reintroduce the zero gas and record the time required to obtain a stable downscale response time (Section 11.6.1.2).

      5.0.1.5 For CEMS that perform a series of operations, (purge-blow back, sample, analyze, etc.) you must time the injection of calibration gases to produce the longest response time.

      6.0.1.6 Repeat the entire procedure three times and determine the mean upscale and downscale response times. The slower or longer of the two means is the system response time.

2.0.2 The HCl CEMS must complete one cycle of operation (sampling, analyzing, and data recording) at least once every 15-minutes, (i.e., a minimum of four samples per hour).

  2.6 Calibration Error Check 

3.0.3 The percent calibration error is the mean difference between the certified HCl calibration gas value, (R), and the CEMS response at each calibration point, (A), calculated using Eq. 2. Calibration Error must be less than 5 percent of CEMs span.

4.0.4 Calibration curves, (linear or quadratic) must demonstrate an intercept equal to or less than 15 percent of the instrument span.

5.0.5 Extractive CEMS calibration error check is performed by flooding the probe and instrument with known concentrations of HCl gas. 

      1.0.5.1 Vendor certified HCL calibration gas is introduced at a minimum of three levels (low, mid, high). Refer to Table 1 for required calibration gas levels. Verify that the measurement is stable by collecting two consecutive independent measurements, at least 2 minutes apart.
 
      2.0.5.2 The percent relative difference of the reference gas concentration to the measured concentration may not exceed 10 percent. Table 1 provides guidance on the certified HCl concentrations for use as low-, mid-, and high levels.

6.0.6 In-situ, cross stack CEMS calibration check is performed by addition of known concentrations of HCl gas into a calibration cell of known volume and path length. Calculate the equivalent concentration of the Vendor certified HCl calibration gas at the stack conditions.

  6.6 Calibration Drift (CD) Test

7.0.7 Prior to the start of the initial accuracy tests, you must begin a calibration drift test for a period of seven consecutive days.

8.0.8 The calibration drift test must be conducted during normal facility operations. The purpose of this test is to verify that the instrument operation is stable.

9.0.9 The calibration drift tests must be performed using certified calibration gases or a NIST traceable calibration transfer standard. 

10.0.10 High concentration certified standard may be diluted and used to measure high-level calibration drift. Use of this method requires documenting the quantitative introduction of HCL gas into the system; such as a Method 205 or similar procedure.

11.0.11 During the calibration drift test period you must determine the magnitude of the zero-level calibration drift and the high-level calibration drift at least once each day (at 24-hour intervals)using Eq. 2. 

12.0.12 During the intervals between drift tests, no adjustments or calibrations may be made to the CEMS. If periodic automatic or manual adjustments are made to the CEMS zero-level and calibration settings, conduct the CD test immediately before these adjustments, or conduct it in such a way that the CD can be determined.

13.0.13 For Extractive CEMS, determine the zero-level drift by introducing certified zero gas into the CEMS sampling probe (24 hours after the last calibration or calibration drift test). Record the CEMS response. Determine the high level calibration drift by introduction certified calibration gas into the CEMS sampling probe. (24 hours after the last calibration or calibration drift test). Record the CEMS response.

14.0.14 For IP-CEMS, calibration drift checks, introduce certified zero or calibration gas into a permanently mounted calibration cell located in the optical path of the instrument Verify that the measurement is stable by collecting two consecutive independent measurements, at least 2 minutes apart. Calculate the equivalent reference value, Ri, at the stack conditions. See section 12.1 for nomenclature of Ri.

15.0.15 The zero-level and high-level drift each must be less than 5% of span for each of seven consecutive days. Each zero- and high-level drift check must be recorded daily and reported for the seven day drift check tests. 

  15.6 Accuracy Test

16.0.16  The accuracy of the HCl CEMS may be determined by performing either a relative accuracy comparison to a reference method or by dynamic spiking. 

17.0.17 Sampling Strategy for RM Tests. Conduct the RM tests in such a way that they will yield results representative of the emissions from the source and can be correlated to the CEMS data. Conduct the diluent (if applicable), moisture (if needed), and pollutant measurements simultaneously. However, diluent and moisture measurements that are taken within an hour of the pollutant measurements may be used to calculate dry pollutant concentration and emission rates. In order to correlate the CEMS and RM data properly, record the beginning and end of each RM  run (including the exact time of day) with the permanent record of CEMS output.

18.0.18 Conduct the RA test while the affected facility is operating during normal operation, or as specified in an applicable subpart. 

19.0.19 Relative Accuracy Using a Reference Method

      1.0.19.1 Reference Methods (RM). Unless otherwise specified in an applicable subpart of the regulations, Methods 26A or Method 320, or their approved alternatives, are the reference methods for HCl. Other reference methods for moisture, oxygen, etc. may be necessary.

      2.0.19.2  Number of RM Tests. Conduct a minimum of nine sets of all necessary RM test runs.

NOTE: More than nine sets of RM tests may be performed. If this option is chosen, a maximum of three sets of the test results may be rejected so long as the total number of test results used to determine the RA is greater than or equal to nine. However, all data must be reported, including the rejected data. 

      3.0.19.3 Correlation of RM and CEMS Data. Correlate the CEMS and the RM test data as to the time and duration by:

(a). Determining from the CEMS final output of the integrated average pollutant concentration or emission rate for each pollutant RM test period. Consider system response time, if important, and confirm that the pair of results are on a consistent moisture, temperature, and diluent concentration basis and

(b). Comparing each integrated CEMS value against the corresponding average RM value.

      4.0.19.4 If the RM has an integrated sampling technique (e.g., Method 320), make a direct comparison of the RM results and CEMS integrated average value for each test period.

      5.0.19.5 Calculate the mean difference between the RM and CEMS values in the units of the emission standard, the standard deviation, the confidence coefficient, and the relative accuracy according to the equations in Section 12.
 
20.0.20  Accuracy Determination Using Dynamic Spiking

      1.0.20.1 Dynamic spiking may be used to determine and document the accuracy, precision, and bias of the HCl CEMS. Figure 3 provides an overview of a typical dynamic spiking arrangement.

      2.0.20.2 To meet the dynamic spiking criteria, you must perform validation of your HCl CEMS by measuring spiked and unspiked sample gas. You must introduce the spike gas into the permanent CEMS probe, upstream of the particulate filter and as close to the sampling head as practical. A known quantity of calibration gas is spiked into the sample gas and measured by the CEMS during normal facility operation. 

Note: For cases where the emission standard is expressed in units of lb/MM Btu or corrected to a specified O2 or CO2 concentration, an absolute accuracy specification equivalent to 5 ppm must be calculated using an average or typical diluent concentration and applicable conversion factors. The appropriate procedures for use in cases where a percent removal standard is more restrictive than the emission standard are the same as in PS-2.

      3.0.20.3 Dynamic Spiking Procedure for Extractive CEMS

11.9.5.3.1	You must collect at least 6 sets of 1 minute average independent parallel unspiked and spiked sample measurements evenly spaced over no less than an hour for a complete accuracy test. 

11.9.5.3.2	The HCl concentration of the flue gas must be determined during each spiking trial. 

20.3.0.20.3 Install a temporary CEMS equivalent to the permanent installation that can perform a stack traverse.  You may use this system to confirm that stratification of HCL meets the requirements of Method 11.5.

20.4.0.20.4 Begin collecting measurements of 2 independent unspiked samples with the temporary CEMS and two spiked sample measurements with the permanent CEMS. 

20.5.0.20.5 While the permanent HCl CEMS is sampling flue gas, HCl reference gas at the regulatory standard or at five times the LOD is introduced into the CEMS sample interface using a mass flow controller (or equivalent).

20.6.0.20.6 Introduce the spike gas into the back of the probe, upstream of the particulate filter. Collect measurement data from the spiked gas stream until sequential measurements are within 5% of each other. Then collect measurements of 2 independent spiked samples.

20.7.0.20.7 The target ratio of the flue gas sampling rate (Qs) to the HCl reference gas is > 9:1. The HCl concentration is quantified by the HCl CEMS.

20.8.0.20.8 Turn off the spike flow and collect data until sequential unspiked measurements are within 5% of each other.

20.9.0.20.9 Repeat the procedure in 11.9.5.3.1 through 11.9.5.3.8 a minimum of 6 times. Simultaneously collect data from the temporary unspiked monitoring system.

20.10.0.20.10  Calculate the dilution ratio of certified gas to stack gas using either:
 
(1) quantitative introduction of the HCl calibration gas with total flow measurement or
 
(2) a dynamically spiked tracer in the HCl calibration gas.

20.11.0.20.11 Total probe flow measurement requires measurement of HCl spike flow plus total flow through the instrument detector.

20.12.0.20.12 Measure the spike gas flow and the total flow with a calibrated flow monitor capable of NIST traceable 2 percent accuracy. Use of this method requires documenting the quantitative introduction of HCl gas into the system such as a Method 205 or similar procedure.

20.13.0.20.13 You may use a tracer gas to determine the dilution ratio and amount of spike as an alternative and you must measure tracer gas flow to an accuracy of 2 percent. 

3.0.20.3 Dynamic Spiking Procedure for Integrated Path CEMS.

11.9.5.4.1	For in-situ IP-CEMS, you must perform an accuracy test using a calibration cell mounted in the optical path of the instrument. Fill the calibration cell in the optical path with a calibration gas at the regulatory limit or five times the LOD. Sequential measurements are performed with calibration gas and zero gas in the calibration cell.

         0.20.3.5 Analyze the results from the dynamic spiking relative accuracy tests using equations in Sections 12. Calculate the mean difference between the instrument response and the reference value (certified gas) for each injection using Eq. 3 and 4. Calculate the equivalent reference value, Ri at the stack conditions. The RA for the average of three independent measurements may not exceed 15 percent of span.

20.1.0.20.1 If the statistical analysis passes the validation criteria, then the accuracy evaluation is completed.
 
Note: If the results do not pass the criteria, temporal variations in the sample gas may be excessive relative to the interval between measurements. Temporal variation may be reduced by:
          *       Averaging the measurements over long sampling periods and using the averaged results in the statistical analysis,

          *       Reducing the total system response time for extractive CEMS sampling systems, for example, using a smaller volume cell or increasing the sample flow rate.

          *       Using two separate sampling lines (and pumps) for extractive CEMS; one line to carry unspiked flue gas and the other line to carry spiked flue gas to decrease the equilibration time in the lines. Both sampling lines include independent flow measurement and are continuously purged. Even with two sampling lines the variation in unspiked concentration may be fast compared to the interval between consecutive measurements. 

  20.10 Reporting

1.0.1 Record and summarize in tabular form the results of the calibration drift, response time, calibration error, and RA test or alternative spiking procedure, as appropriate. Include all data sheets, calculations, CEMS data records (i.e., charts, records of CEMS responses), and cylinder gas or reference material certifications necessary to confirm that the performance of the CEMS met the performance specifications.

2.0.2 Record and report supporting dilution system data including standard cylinder gas flow, total gas flow, and the results of the test measurements.

3.0.3 You must check with the appropriate regional office, State, or Local agency for additional requirements, if any.

4.0.4 5.0.5 All data and records associated with the HCl CEMS performance acceptance must be retained for 5 years.

      11.0 Calculations and Data Analysis

 5.1 Nomenclature

A 	= Actual analyzer response to calibration gas(ppmv),
b0	= the intercept of the regression line,
b1 	= the slope of the regression line,
Cave 	= average concentration or velocity at all sampling points,
Cc 	= corrected CEMS HCl concentration,
CC 	= confidence coefficient,
CD 	= calibration drift (percent),
CE 	= calibration error (percent),
Cflowavg	= Average calibration HCl gas flow rate into system (Lpm),
Ci 	= measured concentration or velocity at sampling point I,
Cref 	= the reference method effluent concentration,
davg 	= Mean difference between CEMS response and the reference gas (ppmv),
di 	= difference of one minute average CEM response and the reference gas concentration (ppmv),
HClref 	= Calculated reference HCl value for a test run (ppm),
n 	= Number of one minute average values in this data set,
Ri 	= Reference gas concentration of zero or calibration gas concentration introduced into the CEMS (ppmv),
RSD 	= Percent Relative Standard Deviation of the data set,
S 	= Span of the instrument (ppmv).
Sd 	= the standard deviation (of the differences),
St 	= percent stratification,
t0.975 	= the one sided t-value obtained from Table 2 for n-1 measurements,
Tflowavg	= Average Total flow through the system (Lpm),
RMave 	= The mean value measured by the reference method or the mean dynamic spike concentration.
xi 	= Value of the calibration gas concentration (in ppm),
X 	= the one hour measurement average of HClref concentrations, xi
yi 	= One minute average value of the CEM response (in ppmv),
Y 	= the one hour average of HClmeas concentrations, yi
LSM	= Line strength factor, which is derived from the HITRAN database. Dependent on stack temperature
Tstack	= Temperature of the stack at the monitoring location for an IP-CEMS
Treference	= Temperature of the calibration cell
Ri,eff	= Equivalent concentration of the reference value, Ri, at the specified stack temperature.

 5.2 Stratification.  The difference between the stationary measurement point and each traverse point (taken simultaneously), is compared to the average value for all traverse sampling points to determine the percent stratification using the following equation:

			St=Ci-Cref-CaveCave *100		Eq. 1


 5.3 Calibration Drift and Calibration Error. Determine the CD and CE using Eq 2. 

			CD=CE=Ri-A S*100		Eq. 2

Additionally, see equation 3 when using in-situ IP-CEMS.

0.4 Quantitative HCl Calibration Gas Introduction and Total Flow to Calculate Reference HCl.  If you use the quantitative introduction of calibration gas and total system flow to calculate the reference HCL concentration for a given dynamic spiking test run, use Equation 3 to calculate the reference HCL concentration for a given dynamic spiking test run.

		HClref=CflowavgTflowavg*Ri+BaselineHCl		Eq. 3

When utilizing an in-situ IP-CEMS and using an unheated calibration cell, the equivalent Ri is calculated per equation 3a.

			R - i,eff=RixCalibration Cell PathlengthStack PathlengthxTstackTreferencexLSM

BaselineHCl will be calculated by averaging a period of time before and after the dynamic spike.

0.5 Relative Accuracy using Dynamic Spiking Test Data. If quantitative HCl calibration gas introduction and total flow was used to calculate reference HCl challenge concentration, calculate the accuracy and precision of the HCl CEMS measurements, the correlation between the HCl challenge gas concentration (HClref) and the average measured HCl concentration (HClmeas) is determined as follows:

1.0.1 Calculate the arithmetic mean difference of the one minute RA measurements using equation 4.

				davg= 1ndi  			Eq. 4
	
where:
				di= Ri-yi 			Eq. 5

For in-situ IP-CEMS, use equation 5a.

				di=HCl - ref-yi			Eq. 5a

2.0.2 Calculate the standard deviation of the differences (Sd) of measured and reference method results calculated using equation 6.

			Sd=1ndi2-1ndin2n-11/2			Eq. 6 

3.0.3 Calculate the confidence coefficient, (CC) for the relative accuracy tests using equation 7.

			CC=t0.975*Sdn1/2 			Eq. 7

4.0.4 Calculate the relative accuracy (RA) for the CEMS using equation 8.

	RA=davg+CC/RMavg*100		Eq. 8


0.4 Relative Response (RRF). Calculated RRF using equation 9.

		RRF= (yi)(Ri) 					Eq. 9


      12.0 Method Performance

0.1 Calibration drift for the HCl CEMS must not drift or deviate from the reference value of the gas cylinder by more than 5 percent of the span value for 7 consecutive days (Eq. 2).

0.2 Calibration Error Check, (linear or quadratic) 

      1.0.2.1 Calibration intercept must be equal to or less than 15% of the instrument's span. 

      2.0.2.2 The relative response factor (RRF) of the instrument calibration must be greater than 0.85 for all calibration check points (Eq. 9).
 
      3.0.2.3 The mean percent difference between the known calibration gas value and the CEMS measured concentration at each of the three points (Eq.5) must be less than 5 percent of span.

0.3  Relative Accuracy Check  -  Reference Method 

      1.0.3.1 The  RA of the CEMS compared to a reference method in the units of the emission standard, must be less than or equal to 20 percent (Eq. 8) of the reference method when RMavgis used in the denominator of Eq. 8 or

      2.0.3.2 In cases where the average emissions for the test are less than 50 percent of the applicable standard, substitute the emission standard value in the denominator of Eq. 8 in place of RMavg and the  RA must be less than or equal to 10 percent. 

0.4 Accuracy Check - Dynamic Spiking

      1.0.4.1  The accuracy of the CEMS compared to a reference method in the units of the emission standard or to a dynamic spike, must be less than or equal to 20 percent when TMavg is used in the denominator of Eq. 8 or

1.0.1 In cases where the average emissions for the test are less than 50 percent of the applicable standard, you may substitute the emission standard value in the denominator of Eq. 8 in place of TMavg.  The accuracy must be less than or equal to 10 percent when the applicable emission standard is used in the denominator of Eq. 8.

16.0 References.

1. Method 318, 40 CFR, Part 63, Appendix A (Draft),
"Measurement of Gaseous Formaldehyde, Phenol and Methanol
Emissions by FTIR Spectroscopy," EPA Contract No. 68D20163,
Work Assignment 2-18, February, 1995.

2. "EPA Protocol for the Use of Extractive Fourier
Transform Infrared (FTIR) Spectrometry in Analyses of
Gaseous Emissions from Stationary Industrial Sources,"
February, 1995.

3. "Measurement of Gaseous Organic and Inorganic
Emissions by Extractive FTIR Spectroscopy," EPA Contract No.
68-D2-0165, Work Assignment 3-08.

4. "Method 301 - Field Validation of Pollutant
Measurement Methods from Various Waste Media," 40 CFR 63,
App A.


17.0 Tables, Diagrams, Flowcharts, and Validation Data.
                                       
              TABLE 1. ARRANGEMENT OF VALIDATION MEASUREMENTS FOR
                             STATISTICAL ANALYSIS.

                                  Measurement
                                 (or average)
                                     Time
                                    Spiked
                                     (ppm)
                                   di spiked
                                   Unspiked
                                     (ppm)
                                  di unspiked
                                       
                                       1
                                       
                                      S1
                                    S2 - S1
                                      U1
                                    U2 - U1
                                       
                                       2
                                       
                                      S2
                                       
                                      U2
                                       
                                       3
                                       
                                      S3
                                    S4 - S3
                                      U3
                                    U4 - U3
                                       4
                                       
                                      S4
                                       
                                      U4
                                       
                                       5
                                       
                                      S5
                                    S6 - S5
                                      U5
                                    U6 - U5
                                       6
                                       
                                      S6
                                       
                                      U6
                                       
                                       7
                                       
                                      S7
                                    S8 - S7
                                      U7
                                    U8 - U7
                                       8
                                       
                                      S8
                                       
                                       
                                      U8
                                       
                                       9
                                       
                                      S9
                                   S10 - S9
                                      U9
                                    U10- U9
                                      10
                                       
                                      S10
                                       
                                       
                                      U10
                                       
                                      11
                                       
                                      S11
                                   S12 - S11
                                      U11
                                    U12-U11
                                      12
                                       
                                      S12
                                       
                                      U12
                                       
                                    Average
                                       
                                      Sm
                                       
                                      Mm
                                       


                               TABLE 2. t-VALUES

                                    n-1[a]
                                    t-value
                                     n-1 a
                                    t-value
                                     n-1 a
                                    t-value
                                     n-1 a
                                    t-value
                                       5
                                     2.571
                                      11
                                     2.201
                                      17
                                     2.110
                                      23
                                     2.069
                                       6
                                     2.447
                                      12
                                     2.179
                                      18
                                     2.101
                                      24
                                     2.064
                                       7
                                     2.365
                                      13
                                     2.160
                                      19
                                     2.093
                                      25
                                     2.060
                                       8
                                     2.306
                                      14
                                     2.145
                                      20
                                     2.086
                                      26
                                     2.056
                                       9
                                     2.262
                                      15
                                     2.131
                                      21
                                     2.080
                                      27
                                     2.052
                                      10
                                     2.228
                                      16
                                     2.120
                                      22
                                     2.074
                                      28
                                     2.048

 a The value n is the number of independent pairs of measurements (a pair consists of one spiked and its corresponding unspiked measurement). Either discreet (independent) measurements in a single run, or run averages can be used.

SOURCE:
DATE:
 CEMS:
LOCATION:
SERIAL NUMBER:
SPAN:


DAY
DATE
TIME
CALIBRATION
VALUE
 CEMS
RESPONSE
DIFFERENCE
PERCENT
OF SPAN
 ZERO/LOW LEVEL
1







2







3







4







5







6







7






 HIGH LEVEL
1







2







3







4







5







6







7






   1. Acceptance Criteria: <15% of span for seven days.

                   Figure 1 Calibration Drift Determination
                                       
                                       
                                       

SOURCE:
DATE:
 CEMS:
LOCATION:
SERIAL NUMBER:
SPAN:

                                      RUN
                                    NUMBER
                               CALIBRATION VALUE
                                      CEMS
                                   RESPONSE
                                  DIFFERENCE



                                   Zero/Low
                                      Mid
                                     High
1





2





3





4





5





6





7





8





9






                                                              Mean Difference =




                                                            Calibration Error =
                                                                              %
                                                                              %
                                                                              %


                   Figure 2: Calibration Error Determination



Figure 3. Extractive Measurement Dynamic Spiking Apparatus

