LiDAR Based Nox Measurements from Exhaust Stacks

Francois Babin, Jean-Francois Gravel, INO

December 2015

Executive Summary

Following the development of a standoff LiDAR for the measurement of NO emissions from compressor station exhaust stacks in collaboration with TransCanada Pipelines, INO performed an assessment of the existing LiDAR for the measurement of NO2. The proposed activities were:

  • Adapt simulated stack for NO2 measurements
  • Perform laboratory absorption measurements on reference cells
  • Perform concentration measurements on NO2
  • Perform concentration measurements of NO and NO2 in a single laser wavelength scan
  • Analyze data – Determine Limit of detection and interferences
  • Report and publicize results

All of these were done, or at the very least attempted. The stated goal of this project was to demonstrate the use of a UV-DiAL prototype for the stand-off measurement of NOx from exhaust stacks. The measurement campaign was done at INO on an instrumented mock-up stack in realistic Canadian weather conditions. This gave insight into the use of the technique for mass emission rate measurements of NOx from pipeline compressor station exhaust stacks.

It was determined that:

  1. It is possible to measure NO and NO2 from a stand-off distance directly at the output of an exhaust stack by the UV-DiAL technique;
  2. It is possible to measure NO and NO2 in a combined sequence of UV laser wavelength scanning;
  3. Optimizing the measurements implies selecting a subset of wavelengths that enhances the uniqueness of the signature for both NO and NO2;
  4. NO2 has a chemical behavior that is troublesome near typical ambient outdoors temperature, especially under 22°C (thus a good part of the year in Canada), an hindrance for the UV-DiAL technique as used and for the referee measurement of NO2 in the exhaust stack mock-up;
  5. The stability of the NO2 reference concentration is difficult to ascertain when the reference cell and ancillary control equipment is not temperature regulated.
  6. Designing, producing and using a stable optical NO2 reference cell will be a challenge for field use;
  7. A reference cell with a constant flow of gas from a calibrated cylinder was required;
  8. Using a reference measurement on a mock-up stack that uses combustion of natural gas in possible addition to the gas from a calibrated cylinder would be preferable;
  9. Sensitivity was, in part, limited by sub-optimal control of the laser pointing;
  10. Calibrating the LiDAR overlap functions with respect to wavelength before the concentration measurements and adapting the data processing mitigates part of the problems;
  11. Slow mechanical wavelength scanning, as done here, is not the most appropriate form of measurement because of the instability in LIDAR return and in the NOx concentration-length product;
  12. The instabilities in LIDAR return and in concentration along with the slow wavelength scanning will ultimately limit the prototype sensitivity;
  13. Detecting and removing outliers caused by instabilities in the particle concentration along the beam path mitigates some of the instabilities;
  14. Single shot whole spectrum measurements would be the best LIDAR approach;

The points above summarize our conclusions and best practices. Although INO was able to measure NO, the prototype is far from optimal. There are two major drawbacks/hindrances.

  • The first is the speed at which the laser wavelength is scanned. The laser wavelength must be scanned across the entire set of wavelengths within a fraction of a second in order to minimize the effect of instabilities in LIDAR return or in the NOx concentration along the laser beam path. This is not trivial. INO is working on a faster version of the measurement, in order to “freeze” the measurement in as short a measurement time as possible.
  • The second is the reference cell for NO2. In order to have a good reference, the concentration of NO2 in the reference cell must be stable. INO found it was helpful to continuously flow a small amount of calibrated gas in the cell. This will be a problem for a fieldable system. The reference cell and the reference cell control system (cylinder, flowmeters, tubing, pressure gages …) will need to be controlled in temperature. This will make for a much more complex system than anticipated. Another type of wavelength reference, one that would be more stable with temperature and more amenable to field use is required.

It was not possible to obtain a flow of measurable NO2 from the mock-up stack. So no ambient temperature mock-up exhaust stack measurement of the concentration of NO2 could be performed.

Using the measurements, it was determined that for a reasonable 1% absorption of the laser beam in the exhaust plume, the sensitivities would be 0.75 ppm-m and 25 ppm-m for NO and NO2 respectively. This would require measurement times longer than those used here. The limit of detection of NO2 is not sufficient and would require major enhancements in the UV LiDAR platform if we keep using the same approach. A different analytical strategy will need to be used for the sensitivity to reach the adequate level. One way would be to reduce the number of wavelengths used and to improve the laser repetition rate. Finally, as the measured LOD depends on the stability of the emission source, it will be necessary to add a reference instrument to the mock-up system in a future implementation.

Full Report

Best Management Practices

# 15-ARPC-04