Join 40,000+ geospatial professionals and get technology news delivered straight to your inbox every week. Subscribe today.

measuring-methane-landfill-emissions-using-thermal-infrared-hyperspectral-imaging.png

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging

Telops

Methane (CH4) and carbon dioxide (CO2) are landfill gases (LFG) originating from the decomposition of organic material as encountered in landfill areas. Both gases are considered greenhouse gases due to their high propensity for absorbing and emitting infrared radiation. For this reason, landfill installations are often equipped with wells and pumping systems to collect these LFGs.

Depending on the quality of the gas mixture, the LFGs may be collected, flared, or simply dispersed in the atmosphere through passive methane vents. In some situations, gas leaks may occur directly on-site at the landfill originating from components such as the wells, tubing system, or a degraded landfill liner. In the case where lateral migration occurs, the gas leaks may even originate on properties located in the landfills’ neighbourhood.

These situations are difficult to address since they require surveillance over large areas. Several gas measurement techniques are currently available. Gas sampling using pressurized canister vessels or bags provides accurate results, even at low concentrations, but is time-consuming since the samples need to be taken into a lab for analysis. Electrochemical cell sensors, also known as gas sniffers, provide accurate real-time results. Conventional gas measurement techniques typically provide single point results and are not well suited for the localization of an unknown gas leak within a large area. In order to illustrate how thermal infrared hyperspectral imaging performs in landfill environments, results from measurements of diffuse methane emissions are presented. Real-time chemical imaging results, carried out on a passive methane vent, are also presented. The results illustrate how thermal infrared hyperspectral imaging systems provide valuable information to survey methane gas in landfill areas.

Introduction

It is well known that carbon dioxide (CO2) is a strong greenhouse gas, yet methane (CH4) has a mass-based greenhouse warming potential (GWPs) 86 times more effective than CO2 on a 20-year timescale1. All objects spontaneously emit infrared radiation in the longwave infrared (LWIR) spectral range (8-12 μm) under ambient conditions. For infrared-active molecules like methane, the absorption/emission of infrared radiation occurs in a very unique fashion as a function of energy (or wavelength). This absorption/emission spectrum is often referred to as the infrared spectral signature. Figure 1 depicts the methane infrared spectral signature in the LWIR. The most prominent spectral features are located around 1306 cm-1 or 7.65 µm.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging LWIR infrared spectrum of methane
Figure 1: LWIR infrared spectrum of methane.

For the work presented below, we were fortunate to work with the city of Sherbrooke’s department of urban infrastructure to illustrate how thermal infrared hyperspectral imaging performs in landfill environments. The data was collected on a former landfill site in Sherbrooke equipped with a biogas collection system. Figure 2 presents some images taken on the site. The Hyper-Cam was used to collect data all around the site of exposed equipment parts susceptible of leaking methane.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging pictures taken during the data collection
Figure 2: Pictures taken during the data collection campaign on the landfill site.

Experimenting with Methods and Materials

In this work, ground-based standoff thermal infrared hyperspectral imaging is presented as an alternative technique to survey methane emissions in landfill areas. The next sections will present the hyperspectral camera and two data analysis techniques. The first analysis technique is an algorithm capable of providing real-time results for localized emissions while the second is capable of achieving higher sensitivity in the case of diffuse methane emissions.

Hyperspectral Camera

The Telops Hyper-Cam (Figure 3) is a lightweight and compact hyperspectral imaging instrument using Fourier Transform Infrared (FTIR) technology. It provides a unique combination of spatial, spectral and temporal resolution for a complete characterization of the substances being monitored. Its high performance and efficiency as a standoff chemical agent detector have been proven through numerous field campaigns.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Telops Hyper-Cam LW
Figure 3: Hyper-Cam

The Hyper-Cam Methane has a bandpass from 1219 to 1355 cm-1 spectral range centred around the most prominent methane spectral line. It features a cooled Focal Plane Array (FPA) detector containing 320x256 pixels over a basic 6.4°x5.1° field of view. Each pixel has a field of view (FOV) of 350 μrad. However, the FOV can be widened to 1.4 mrad by adding the wide-angle lens to the camera. The spectral resolution of the Hyper-Cam is user-selectable between 0.25 and 150 cm-1.

The HyperCam offers high sensitivity for each pixel of the scene under observation, and its lightweight makes it ideal for field operation. The Hyper-Cam was designed from the ground up so instrument control and data acquisition are specifically optimized. The sensor is capable of generating calibrated data (Fourier transformed and radiometrically calibrated) in real-time at the highest data rate. This real-time processing allows lossless compression by a factor of approximately 10-20, a welcomed factor for such a high-throughput instrument.

An imaging Fourier-transform spectrometer (IFTS) such as the Hyper-Cam is a hyperspectral imager that combines a Michelson interferometer with a staring infrared focal-plane array (FPA). The raw datacube is built by accumulating infrared images at precise mirror intervals. A numerical Fast Fourier Transform (FFT) is applied to the raw datacube at each pixel, producing an intensity spectrum in the frequency domain. This intensity spectrum is further converted into a radiance spectrum through radiometric calibration, using blackbody radiation reference sources positioned in front of the camera.

The Principles Behind Real-Time Imaging and Quantification of Fugitive Gas Emissions

The first step to present real-time gas emissions imaging is to detect and identify the substance. This is accomplished using Telops Reveal DI software which allows for chemical imaging of multiple gases simultaneously on an interactive interface (refer to Figure 4). The Reveal DI software uses a powerful and specialized generalized likelihood ratio algorithm (GLRT) to detect and identify the gas signature.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Reveal Detect and Identify
Figure 4: Reveal Detect and Identify (D&I) Software showing methane escaping from an opening made in the sensing entry point of the methane collection pipe system.

The second step is to produce the column path concentration map of the gas plume (in units of ppm·m) from the acquired spectral radiance datacube. This is accomplished using a rapid gas quantification retrieval algorithm. At its core lies a physical radiative transfer model. More precisely, by Beer-Lambert-Bouguer’s law, the total radiation from a pixel in the gas plume measured by the camera can be expressed as:

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq1

Whereas the total radiation from a pixel outside the gas plume is:

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq2

Where:

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq3

By making assumptions such as a constant plume temperature across the column path, a localized gas plume and a gas plume temperature equal to the atmospheric temperature, it is possible to solve for t-atm and thus obtain the gas concentration values.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq3.1

It is worth noting that the algorithm uses the brightness temperature of strong water vapour spectral lines to estimate the atmospheric temperature (Tatm). A principal component analysis (PCA) is performed on the image to help select background pixels and thus estimate LRef. Homogeneous media are characterized by constant spectral absorption coefficients. Consequently, the light transmittance through a layer of such media having a thickness l, a concentration N and an absorption crosssection α is expressed as:

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq4

Equation 4 imposes an important limitation to any retrieval procedure. Since the transmittance of any layer depends, for each gas species, on the product Nl called the column density, this product remains indivisible. For this reason, the results are quoted in path length units or ppm×m 3,4.

Assuming spherical symmetry of the plume, it is possible to estimate local concentrations in units of ppm.

Knowing the nature of the gas, its extent as well as the specifics of the camera (field-of-view, distance to the gas plume), the concentration results can be expressed as a total gas mass in units of grams. In addition, the gas mass flow rate (F) can be determined using both the total gas mass (mgas) per pixel and the mean gas flow velocity (vgas):

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging eq5

The mean gas flow velocity can be approximated as the wind measurement values obtained from nearby meteorological stations or from specialized digital image correlation algorithms analyzing the acquired raw data.

Principles Behind Imaging and Quantification of Diffuse Gas Emissions

When the gas emissions are diffuse in nature, there is no singular point of emission and the gas tends to occupy a large portion of the image. These facts make it difficult for the GLRT algorithm described above to detect its presence. Hence, to analyse those measurements we rely on a more rigorous algorithm. Moreover, by processing hyperspectral data offline, we are not as concerned by processing speed and can therefore forgo the assumptions made in the section above and tackle equation 1 head-on. Equation 1 provides the framework to calculate the spectral radiance entering the camera.

The goal of the offline quantification algorithm (which can be described as an optimization approach) is to minimize the difference between the measured radiance spectra acquired by the camera and the theoretical radiance calculated through the radiometric model of Equation 1, as described in Tremblay et al.2. The gas concentrations and plume temperature are allowed to vary until the system finds the best model fit in an iterative fashion. Once the path length results are obtained, the total gas mass as well as the gas mass flow rate (F) can be determined using the method described in the previous section if the gas leak is localized. If the leak is diffuse, i.e. there is no single gas emission point, it is difficult and sometimes impossible to calculate a flow rate.

Results and Discussion

Results From Fugitive Gas Emissions

In order to demonstrate the capabilities of the technique upon arrival to the site, we focused on an exposed pipe (#57) of the methane collection piping system (refer to Figure 2). For this initial demonstration test, the pumping in that specific portion of the biogas collection system was voluntarily interrupted. In normal operation, methane is not leaking from the pipe.

In this case, the camera was positioned 5m from the pipe. Figure 5 presents a sequence of 3 consecutive infrared measurements overlayed with the detected methane plume coloured in pink identified in real-time by the software. The bottom row of Figure 5 presents the calculated methane path length concentration images for the corresponding measurements. We observe concentrations varying from a few hundreds of ppm·m to upwards of 2000 ppm·m. This approach works well in this scenario even though methane is an atmospheric constituent because the concentration of methane emissions is significantly greater than the atmospheric methane concentration.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Sequence of 3 consecutive measurements
Figure 5: Top: Sequence of 3 consecutive measurements showing methane plume in pink. Bottom: Sequence of the same 3 measurements showing the calculated methane path length concentration map in units of ppm·m.

From the known specifics of the camera (field-of-view, distance to the gas plume) and the column path concentration results, the total gas mass contained within a specified area of the image is calculated. Assuming that the gas velocity is equal to the wind speed, we can calculate a gas flux as a function of time, refer to Figure 6. The solid blue line assumes a wind value of 2.2 m/s while the shaded areas are bounded using minimum and maximum wind gust values corresponding to that time period.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Methane flux
Figure 6: Methane flux as a function of time since beginning of the test on pipe #57.

A survey of several points of interest on the site revealed methane emissions near a concrete cylinder collection point. Figure 7 presents an infrared measurement overlayed with the detected methane plume in blue and the calculated methane path length concentration image. In this case, the rapidly changing wind conditions made it difficult to obtain gas flux values as a function of time.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Infrared measurement
Figure 7: Right: Infrared measurement overlayed with the detected methane plume in blue. Left: Methane path length concentration image.

Results From Diffuse Gas Emissions

Diffuse methane emissions were measured in the landfill area. An efficient survey of large areas, on the order of hundreds of square meters, can be carried out from the same location. In this case, the surface of the pipe provides an excellent background against which we can determine the diffuse methane concentration between the camera and the pipe. As shown in Figure 8, the average methane concentration in the landfill area was found to be more than 3-4 times higher than the global mean ambient methane concentration (1.9 ppm) normally measured at ground level 5. As indicated in Figure 8 D, the low residuals between measurement and model give us confidence in the retrieved methane concentration values. The results of a similar analysis performed on a different part of the site are presented in Figure 9. Methane concentrations are found to be slightly lower than those of the previous site but still higher than normal atmospheric values. The residuals are not as good as those of the previous site. This may be due to the presence of vegetation just in front of the pipe which affects the signal measured by the camera and thus may negatively impact the signal to noise ratio.

Measuring Methane Landfill Emissions Using Thermal Infrared Hyperspectral Imaging Visible image captured
Figure 8: A: Visible image captured from the position of the Hyper-Cam showing the field and an exposed portion of the piping system 52m away. B: Closeup image of the pipe. C: IR image of the pipe overlayed with the methane concentration map. Each pixel is being processed independently of the others. D: Typical infrared spectrum of a pixel associated with methane diffuse emissions as measured with the Hyper-Cam-LW sensor (red curve) and the spectrum of the best fit model (blue curve). The rms residuals are of the order of 0.3 K. Black arrow points to the location of the most prominent spectral feature of methane.

Conclusion

The versatility of the Telops Hyper-Cam for landfill characterization was demonstrated for both mappings of diffuse methane emissions and imaging of methane gas clouds. This approach allows the survey of large areas without prior installation of any particular equipment as it relies on the strong infrared self-emission of methane in the LWIR. The temporal resolution combined with the selectivity provided by hyperspectral imaging allows efficient characterization of gas leak situations which can be useful to locate fugitive emissions in landfills. Standoff thermal infrared hyperspectral imaging represents an interesting alternative to the conventional tools used for landfill characterization.

Acknowledgements

We would like to thank the city of Sherbrooke and Bruno Marin for their collaboration on this project, it is greatly appreciated.

References

[1] IPCC, 2013. The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. In: Stocker, T.F., others (Eds.), Climate Change. Cambridge Univ. Press.

[2] P. Tremblay, et. al.,”Standoff gas identification and quantification from turbulent stack plumes with an imaging Fourier transform spectrometer,” Proceedings of SPIE - The International Society for Optical Engineering. Vol. 7673, no. 10.1117/12.850127, 2010.

[3] Savary, S., Tremblay, P., Villemaire, A., “System and method to remotely measure the directional velocity map of gaseous emissions”, U.S. Patent pending No. 61/332,260 (2010).

[4] Savary, S., Gagnon, J.-P., Gross, K., Tremblay, P., Chamberland, M., Farley, F., "Standoff identification and quantification of flare emissions using infrared hyperspectral imaging," Proc. SPIE 8024, Advanced Environmental, Chemical, and Biological Sensing Technologies VIII, 80240T (26 May 2011);

[5] NOAA Global Monitoring Laboratory Website: https://www.esrl.noaa.gov/gmd/ccgg/trends_ch4/

​​​​​​​​​​​​​​

Do you have questions about this case study?

Get in touch with Telops, and they would be happy to answer any questions you have about pricing, suitability, availability, specs, etc.

Ask a question

Telops team is happy to assist.

Telops

We're available on all weekdays and usually reply within 1-2 working days.

Please provide detailed information. Your input helps Telops assist you effectively.

Let us know a little about you and your contact information.

Do you represent a company or organization?

Before we proceed, would you be interested in subscribing to our newsletter?

By completing this form, you acknowledge and agree to abide by our Terms and Conditions, Privacy Policy and Cookie Policy.

Thank you for your message! Our team tipically to respond to inquiries within 1-2 business days.