|
Exhaust emissions from a compressor engine are tested to determine the quantity of NOx and CO being vented to the atmosphere. Three 30-minute tests are performed on the engine to determine emissions of these criteria pollutants. This section of the report describes the analytical methods and procedures used during these tests.
The sampling and analyses procedures used during these tests conformed with those outlined in the Code of Federal Regulations, 40 CFR 60, Appendix A, Methods 3a, 7e, 10 and 19.
Stack gas enters the system through a stainless steel probe with a glass wool filter. The sample is transported via 3/8-inch heat-traced TeflonR tubing to the “wet” side of the sample manifold via a stainless steel/TeflonR diaphragm pump. The sample is then delivered to a specially designed stainless steel minimum-contact condenser, which drys the sample without removing NOx or other compounds of interest. The sample is then passed to the “dry” side of the manifold where it is partitioned to the NOx, CO, O2, analyzers through glass and stainless steel rotameters for flow control of the sample.
The sample system is also equipped with a separate path through which a calibration gas can be delivered to the probe and back through the entire sampling system. This allows for convenient performance of system bias checks as required by the testing methods.
A conversion factor to convert PPM TO LB/SCF is 1.194X10-7 for NOx 7.27X10-8 for CO. A fuel rate is recorded from a calibrated fuel flow meter on site and the gross calorific value is received from a fuel analysis done prior to testing.
The stack gas analyses for O2 concentrations are performed in accordance with procedures set forth in EPA Method 3a. Instrumental analyses are utilized by BACT in lieu of an Orsat of a Fyrite procedure due to the greater accuracy and precision provided by the instruments. The O2 analyzer utilizes a paramagnetic cell to quantify the oxygen content of the exhaust gases.
EPA Method 7e is used to determine concentrations of NOx. A chemiluminescence analyzer is used for this purpose. NOx mass emission rates are calculated as if all of the NOx were in the form of NO2. This approach tends to overestimate the amount of NOx emissions but corresponds to the EPA’s convention.
CO emission concentrations are quantified during the tests in accordance with procedures set forth in EPA Method 10. A continuous nondispersive infrared (NDIR) analyzer is used for this purpose. This analyzer is equipped with a gas correlation filter, which also eliminates any interference from moisture, CO2 and other combustion products.
All instruments are housed in an air-conditioned, mobile laboratory. Gaseous calibration standard are provided in aluminum cylinders with the concentrations certified by the vendor. EPA Protocol Number 1 gases are used to determine the cylinder concentrations where applicable (i.e. NOx calibration gases).
All data is recorded on one data acquisition recorder (Monico-Computer). This data acquisition collects data every one second for each instrument and data is recorded on a color chart.
A number of quality assurance activities are undertaken before, during and after this testing project. The remaining section of the report, combined with the supporting documentation, describes each of these activities.
Each instrument’s response is checked and adjusted in the field prior to the collection of data via multi-point calibration. The instrument’s linearity is checked by first adjusting its zero and span responses to zero (nitrogen) and an upscale calibration gas in the range of the expected concentrations. The instrument response is then challenged with other calibration gases of known concentration.
The instrument’s response is accepted as being linear if the response of the other calibration gases agreed within 2 percent of range of the predicted values. (The response of the infrared absorption type CO analyzer is electronically, linearized.)
As a minimum, before and after a set of test runs, the analyzers are checked for zero and span drift. This allows each set of test runs to be bracketed by calibrations, and documents the precision of the data just collected. The criterion for acceptable data is that the instrument drift is no more than 2 percent of the full-scale response.
The quality assurance worksheets summarize all multi-point calibration checks and zero to span checks performed during the tests. These worksheets, as prepared from the chart records, show that no analyzer calibration drifted in excess of any method’s specifications during the tests. These worksheets are contained in Section 1V.1 of the report.
Interference response test on the instruments are conducted by the instrument vendors and BACT Resources, Inc. on the NOx, CO and O2 analyzers. The sum of the interference responses for H2O, CO, CO2 and O2 (as appropriate for each analyzer) is less than 2 percent of the applicable full-scale span value. The instruments used for the tests meet the performance specifications for EPA Methods 3a, 7e and 10. The interference response documentation is contained in Section 1V.1 of the report.
The sampling systems are leak checked by demonstrating that a vacuum greater than 10” Hg could be held for at least 1 minute with a decline of less than 1” Hg.
A leak test is conducted after each sample system is set up and before the system is dismantled. These tests are conducted to ensure that ambient air does not dilute the sample. If no leaks are detected, documentation of this procedure is included with the quality assurance in Section 1V.1.
The absence of leaks in the sampling system is also verified by system bias checks. Comparing the responses of the analyzers to a calibration gas, introduced via two paths, test the sample system integrity. The first path is into the analyzer via the zero/span calibration manifold. The second path is to introduce zero and calibration gases into the sample system at the sample probe valve.
Any difference in the instrument responses by these two methods is attributed to sampling system bias or leakage. These bias checks are conducted before the testing. The same sample system is used throughout the tests.
Examination of the chart excerpts and Quality Assurance Data worksheet, show that the analyzer responses, via both sample paths, agree within 2 percent in all cases.
The efficiency of the NO2 to NO converter in the NOx analyzer is checked by introducing a tedlar-bag of 50% NO and 50% ambient air balance (EMC ALT 013). The criterion for acceptability is a change in NOx of less than 2% over a 30 minute time, period. The above mentioned quality assurance worksheet also summarizes the results of the converter efficiency test.
The control gases used to calibrate the instruments are analyzed and certified by the compressed gas vendors to + 1% accuracy for the remaining gases. EPA Protocol Number 1 gases are used, where applicable (i.e. NOx gases), to assign the concentration values traceable to the National Institute of Standards and technology, Standard Reference Materials (SRM’S). The gas calibration sheets, as prepared by the vendor, are contained in Section 1V.E of this test report.
BACT collects and reports the enclosed test data in accordance with the procedures and quality assurance activities described in this test report. BACT makes no warranty as to the suitability of the test methods. Assume no liability relating to the interpretation and use of the test data.
N/A
Operational data is taken from available instrumentation control panels or gauges by BACT personnel once per test run. Operational parameters obtained during the testing, include engine rpm, fuel consumption, percent torque and compressor operational data.
|
