Elsevier

Atmospheric Environment

Volume 35, Issue 16, June 2001, Pages 2837-2844
Atmospheric Environment

Rapid measurements and mapping of tracer gas concentrations in a large indoor space

https://doi.org/10.1016/S1352-2310(01)00081-4Get rights and content

Abstract

Rapid mapping of gas concentrations in air benefits studies of atmospheric phenomena ranging from pollutant dispersion to surface layer meteorology. Here we demonstrate a technique that combines multiple-open-path tunable-diode-laser spectroscopy and computed tomography to map tracer gas concentrations with approximately 0.5 m spatial and 7 s temporal resolution. Releasing CH4 as a tracer gas in a large (7 m×9 m×11 m high) ventilated chamber, we measured path-integrated CH4 concentrations over a planar array of 28 “long” (2–10 m) optical paths, recording a complete sequence of measurements every 7 s during the course of hour-long experiments. Maps of CH4 concentration were reconstructed from the long path data using a computed tomography algorithm that employed simulated annealing to search for a best fit solution. The reconstructed maps were compared with simultaneous measurements from 28 “short” (0.5 m) optical paths located in the same measurement plane. On average, the reconstructed maps capture ∼74% of the variance in the short path measurements. The accuracy of the reconstructed maps is limited, in large part, by the number of optical paths and the time required for the measurement. Straightforward enhancements to the instrumentation will allow rapid mapping of three-dimensional gas concentrations in indoor and outdoor air, with sub-second temporal resolution.

Introduction

Measurements of tracer gas dispersion are useful as a means for quantifying atmospheric transport. Research activities in a broad range of fields benefit from this capability. Examples include flow visualization measurements of gas dispersion in large indoor or outdoor spaces. Such measurements can provide experimental verification of computational models of pollutant dispersion, or footprint models of ecosystem-atmosphere trace gas exchange. For many of these phenomena, the spatial and temporal scales of interest range from 0.1 to 10 000 m and 0.1 to 10 000 s. The smallest scales might correspond to that of a personal breathing zone or turbulent eddies, the mid range scales to ventilation and mixed convection in buildings, and the largest scales to turbulent transport in the atmospheric surface boundary layer.

In outdoor applications, technical considerations have generally restricted the temporal resolution of previous mapping measurements to many minutes or hours. Exceptions are LIDAR measurements of atmospheric water vapor, which can be obtained over distances up to kilometers with approximately 1.5 m spatial and 0.3 s temporal resolution along a single line of sight (Cooper et al., 1994; Eichinger et al., 1999). Sequential measurements over many lines of sight are used to generate maps. LIDAR measurements are particularly well suited for outdoor measurements on large spatial scales because scattering of the laser radiation by aerosols or gas species provides the return beam, eliminating the need for retro-reflectors commonly used with many open-path (OP) systems.

Previous work on measuring the dispersion of pollutants in indoor air has not, to our knowledge, probed time scales shorter than minutes. Techniques used for these measurements have included conventional pump and tube sampling (e.g. Baughman et al., 1994; Drescher et al., 1995), and OP remote sensing, in combination with computed tomography (CT) (Yost et al., 1994; Drescher et al (1995), Drescher et al (1997)). Most of this work covered spatial scales on order of 1–20 m, with approximately 1 m spatial and 100–1000 s temporal resolution. Here the temporal and spatial resolution was limited for pump-and-tube sampling by the time required to switch valves and collect samples from some number of locations, and for OP-CT by the time to sequentially measure many overlapping beam paths with a Fourier Transform Infrared (FTIR) spectrometer. While the temporal resolution of these systems has been modest, they can detect a broad range of compounds and are hence an attractive method for monitoring of unknown pollutants.

Here, we describe a system for mapping concentrations a tracer gas with 7 s temporal and approximately 0.5 m spatial resolution. The system rapidly measures path-integrated gas concentrations over many open paths that are defined by a source of radiation and a detector. The measured data are then used to reconstruct a map of gas concentration using computed tomography.

Section snippets

Materials and methods

This section describes the gas measurement system, the chamber used for the experiments, the initial tests of the system, and the mapping experiments. Additional details concerning this system and the overall experimental program can be found in Gadgil et al. (2000).

Results and discussion

Maps of gas concentration are reconstructed using the 28 long path measurements from each complete measurement sequence (every 7 s). This produces approximately 316 reconstructed maps, from the time period (400–2600 s) when CH4 was being released into the chamber. Fig. 5 shows a sample of six reconstructed maps that illustrate how the pattern of gas concentrations within the measurement plane changed during the experiment shown in Fig. 4. Despite the long duration of the experiment, the CH4

Conclusion

We have demonstrated that tunable diode laser spectroscopy and computed tomography can be combined to accurately map tracer gas concentrations with approximately 0.5 m spatial and 7 s temporal resolution. This represents a significant advance over previous methods for mapping tracer gas concentrations, enabling the examination of transient phenomena on time scales of seconds rather than minutes. At present, CT reconstruction is performed after the experiment is completed. However, the current

Acknowledgments

The authors gratefully acknowledge David Wilson, William Fisk, Woody Delp, Darryl Dickerhoff, Nance Matson, Dennis Dibartolomeo for advice and assistance. The Department faculties at LBNL performed necessary modifications to the chamber. Woody Delp, David Lorenzetti, Michael Sohn, and an anonymous reviewer provided thoughtful reviews of the manuscript. This work was supported by the US Department of Energy under contract DE-AC03-76SF00098.

References (13)

There are more references available in the full text version of this article.

Cited by (26)

  • Chemical Species Tomography

    2022, Industrial Tomography: Systems and Applications, Second Edition
  • Gas distribution mapping and source localization using a 3D grid of metal oxide semiconductor sensors

    2020, Sensors and Actuators, B: Chemical
    Citation Excerpt :

    Two alternative experimental methods for visualizing the gas flow in the experimental scenario without adding tracers are optical remote sensing (ORS) and chemical sensor grids (CSGs). ORS coupled with computed tomography (CT) is an old technique to visualize 2D gas distributions over room-sized areas [22–25]. ORS instruments such as the open-path Fourier transform infrared spectrometer (OP-FTIR) [26] or the tunable diode laser absorption spectroscope (TDLAS) [27] are first used to accurately measure the total number of gas molecules in multiple beams of radiation.

  • Chemical species tomography

    2015, Industrial Tomography: Systems and Applications
View all citing articles on Scopus
View full text