Improving Methods for Quantifying Fugitive Emissions from Open Cut Coal Mining

Fugitive emissions from open cut coal mines are usually estimated for the purposes of reporting under the National Greenhouse and Energy Reporting legislation using either Method 2 or 3. These methods are based on measuring in situ gas content of strata ahead of mining, which combined with the volume of material extracted is used to estimate total emissions from individual mines. While this approach is a substantial improvement over Method 1, which uses generic state-based emission factors, it does require detailed coring and gas content measurements that add significantly to the cost of the methodology. The purpose of the project described in this report was to investigate alternative methods for measuring fugitive methane emissions using atmospheric methods.

The specific objectives of this study were to investigate the potential and limitations of several atmospheric methods, both measurements and modelling, when applied to quantifying fugitive emissions from open cut coal mining operations. The methods examined in detail were:

· Mobile monitoring traverses using an instrumented vehicle and a plume dispersion model. The advantages of using a tracer gas are also investigated. Consideration is given to uncertainties introduced by factors such as access to plumes, frequency of measurement and atmospheric conditions. Some preliminary field measurements were made to test their applicability for the purposes of periodic monitoring of emissions.

· Atmospheric modelling methods. These were evaluated for their ability to locate sources and estimate their emission rates within local scales (less than 10 km) to regional scales (hundreds of km).

· Use of data from an existing monitoring facility to determine the number and locations of monitoring  sites necessary for robust measurements, and uncertainties likely to be associated with the methodology.

The results of the mobile monitoring using a state-of-the-art methane analyser mounted in a vehicle showed that this method was readily suited to locating methane sources both within and outside the mine. While emission rates can in some instances be derived from mobile ground level concentration data, the method is highly dependent upon prevailing meteorological conditions and suitable vehicle access to the plume and local topography. The method is not well suited to routine monitoring and the level of uncertainty is high. The use of a tracer gas on the other hand, proved to be an effective method for determining methane emission rates under a wide range of conditions from sources up to moderate scales. The tracer gas method was able to yield uncertainties of less than 10 % at these scales and was relatively simple to deploy in the field. At the scale of a large open cut mine, the technique was less successful and requires further development. The inherent accuracy and simplicity of this method suggest that it could be developed into a routine measurement technique, provided the cost of the method was acceptable. Further examination of more sensitive analytical methods or the use of other, natural tracers may overcome some of the challenges identified during the limited field trials conducted during this project.

Several methods were developed and tested to infer the emissions and locations of methane sources using fixed-location measurements of concentrations and meteorology, combined with models of atmospheric transport. These methods are often referred to as "top down". They used either a forward optimisation approach, where model-predicted concentration fields are optimised to simulate the observations by varying the source emissions and locations, or an inverse approach, where plumes are tracked backwards in time and source information is optimised using a Bayesian approach. The inverse methods were based on previous applications at local source-receptor scales (~100 metres), using simplifying flow assumptions and two monitoring stations, that successfully located and quantified the emissions from one source. Two regional atmospheric models were used (viz. CCAM and TAPM) for the more complex case of regional scales (~200 kilometres) and multiple sources. They were applied to the region containing coal fields near Emerald, where greenhouse gas monitoring data were available from the CSIRO-Geoscience Australia baseline monitoring station at Arcturus. Forward model simulations showed that about 40% of the methane concentration variance could be explained by emissions from 21 coal mine sources in the region. The unexplained variance is due in part to other methane sources in the region and the high uncertainties associated with the method 1 approach.

A monitoring network with multiple sites designed specifically for the source types and configuration of interest would be preferable to a single site as was used at Arcturus and would complement the regional scale modelling schemes developed here. There is a clear need to discriminate coal mining methane emissions from methane emitted from other sources when monitoring on regional scales and the use of naturally-occurring isotopes of methane (carbon and hydrogen isotopic composition) is considered. Considerations for the design of a network for monitoring individual coal mines and for coal mining regions, incorporating some of the methods developed and tested in this study, are recommended.