Single-well injection-withdrawal tests as a contaminant transport characterisation tool for landfilled waste.

A new single well injection withdrawal (SWIW) test was trialled at four landfills using the tracers lithium and deuterium, and by injecting clean water and measuring electrical conductivity. The aim of the research was to develop a practical test for measuring lateral contaminant transport to aid in the design of landfill flushing. Borehole dilution tests using dyes were undertaken prior to each SWIW test to determine background flow velocities. SWIW tests were performed at different scales by varying the volume of tracer injected (1 to 5,800 m3) and the test duration (2 to 266 days). Tracers were used individually, simultaneously or sequentially to examine repeatability and scaling. Mobile porosities, estimated from first arrival times in observation wells and from model fitting ranged from 0.02 to 0.14. The low mobile porosities measured rule out a purely advective-dispersive system and support a conceptual model of a highly preferential dual-porosity flow system with localised heterogeneity. A dual-porosity model was used to interpret the results. The model gave a good fit to the test data in 7 out of 11 tests (where R2 ≥ 0.98), and the parameters derived are compatible with previous experiments in MSW. Block diffusion times were estimated to range from 12 to 6,630 h, with a scaling relationship apparent between the size of the test (volume of tracer used and/or the duration) and the observed block diffusion time. This scaling relationship means affordable small-scale tests can inform larger-scale flushing operations.

Measuring methane emissions from a UK landfill using the tracer dispersion method and the influence of operational and environmental factors.

The methane emissions from a landfill in south-east, UK were successfully quantified during a six-day measurement campaign using the tracer dispersion method. The fair weather conditions made it necessary to perform measurements in the late afternoon and in the evening when the lower solar flux resulted in a more stable troposphere with a lower inversion layer. This caused a slower mixing of the gasses, but allowed plume measurements up to 6700 m downwind from the landfill. The average methane emission varied between 217 ±â€¯14 and 410 ±â€¯18 kg h-1 within the individual measurement days, but the measured emission rates were higher on the first three days (333 ±â€¯27, 371 ±â€¯42 and 410 ±â€¯18 kg h-1) compared to the last three days (217 ±â€¯14, 249 ±â€¯20 and 263 ±â€¯22 kg h-1). It was not possible to completely isolate the extent to which these variations were a consequence of measuring artefacts, such as wind/measurement direction and measurement distance, or from an actual change in the fugitive emission. Such emission change is known to occur with changes in the atmospheric pressure. The higher emissions measured during the first three days of the campaign were measured during a period with an overall decrease in atmospheric pressure (from approximately 1014 mbar on day 1 to 987 mbar on day 6). The lower emissions measured during the last three days of the campaign were carried out during a period with an initial pressure increase followed by a period of slowly reducing pressure. The average daily methane recovery flow varied between 633 and 679 kg h-1 at STP (1 atm, 0 °C). The methane emitted to the atmosphere accounted for approximately 31% of the total methane generated, assuming that the methane generated is the sum of the methane recovered and the methane emitted to the atmosphere, thus not including a potential methane oxidation in the landfill cover soil.

Validation and error assessment of the mobile tracer gas dispersion method for measurement of fugitive emissions from area sources.

A controlled release test was carried out to assess the accuracy of the tracer gas dispersion method, which is used to measure whole-site landfill methane (CH4) emissions as well as fugitive emissions from other area sources. Two teams performed measurements using analytical instruments installed in two vehicles, to measure downwind concentrations of target (CH4) and tracer gases at distances of 1.2-3.5 km from the release locations. The controlled target gas release rates were either 5.3 or 10.9 kg CH4 h-1, and target and tracer gases were released at distances between 12 m and 140 m from each other. Five measurement campaigns were performed, where the plume was traversed between 2 and 31 times. The measured target gas emissions agreed well with the controlled releases, with rate differences no greater than 1.1 kg CH4 h-1 for Team A and 1.0 kg CH4 h-1 for Team B when quantifying a controlled release of 10.9 kg CH4 h-1. This corresponds to a maximum error of ±10%. A larger error of up to 18% was seen in the campaign with a lower target gas release rate (5.3 kg CH4 h-1). Using a cross plume integration method to calculate tracer gas to target gas ratios provided the most accurate results (lowest error), whereas larger errors (up to 49%) were observed when using other calculation methods. By establishment of an error budget and comparison with the measured error based on the release test, it could be concluded that following best practice when performing measurements, the overall error of a tracer gas dispersion measurement is very likely to be less than 20%.

AERMOD as a Gaussian dispersion model for planning tracer gas dispersion tests for landfill methane emission quantification.

The measurement of methane emissions from landfills is important to the understanding of landfills' contribution to greenhouse gas emissions. The Tracer Dispersion Method (TDM) is becoming widely accepted as a technique, which allows landfill emissions to be quantified accurately provided that measurements are taken where the plumes of a released tracer-gas and landfill-gas are well-mixed. However, the distance at which full mixing of the gases occurs is generally unknown prior to any experimental campaign. To overcome this problem the present paper demonstrates that, for any specific TDM application, a simple Gaussian dispersion model (AERMOD) can be run beforehand to help determine the distance from the source at which full mixing conditions occur, and the likely associated measurement errors. An AERMOD model was created to simulate a series of TDM trials carried out at a UK landfill, and was benchmarked against the experimental data obtained. The model was used to investigate the impact of different factors (e.g. tracer cylinder placements, wind directions, atmospheric stability parameters) on TDM results to identify appropriate experimental set ups for different conditions. The contribution of incomplete vertical mixing of tracer and landfill gas on TDM measurement error was explored using the model. It was observed that full mixing conditions at ground level do not imply full mixing over the entire plume height. However, when full mixing conditions were satisfied at ground level, then the error introduced by variations in mixing higher up were always less than 10%.

Doublet tracer tests to determine the contaminant flushing properties of a municipal solid waste landfill.

This paper describes a programme of research investigating horizontal fluid flow and solute transport through saturated municipal solid waste (MSW) landfill. The purpose is to inform engineering strategies for future contaminant flushing. Solute transport between injection/abstraction well pairs (doublets) is investigated using three tracers over five separate tests at well separations between 5m and 20m. Two inorganic tracers (lithium and bromide) were used, plus the fluorescent dye tracer, rhodamine-WT. There was no evidence for persistent preferential horizons or pathways at the inter-well scale. The time for tracer movement to the abstraction wells varied with well spacing as predicted for a homogeneous isotropic continuum. The time for tracer movement to remote observation wells was also as expected. Mobile porosity was estimated as ~0.02 (~4% of total porosity). Good fits to the tracer breakthrough data were achieved using a dual-porosity model, with immobile regions characterised by block diffusion timescales in the range of about one to ten years. This implies that diffusional exchanges are likely to be very significant for engineering of whole-site contaminant flushing and possibly rate-limiting.

Multiple-tracer tests for contaminant transport process identification in saturated municipal solid waste.

Two column tests were performed in conditions emulating vertical flow beneath the leachate table in a biologically active landfill to determine dominant transport mechanisms occurring in landfills. An improved understanding of contaminant transport process in wastes is required for developing better predictions about potential length of the long term aftercare of landfills, currently measured in timescales of centuries. Three tracers (lithium, bromide and deuterium) were used. Lithium did not behave conservatively. Given that lithium has been used extensively for tracing in landfill wastes, the tracer itself and the findings of previous tests which assume that it has behaved conservatively may need revisiting. The smaller column test could not be fitted with continuum models, probably because the volume of waste was below a representative elemental volume. Modelling compared advection-dispersion (AD), dual porosity (DP) and hybrid AD-DP models. Of these models, the DP model was found to be the most suitable. Although there is good evidence to suggest that diffusion is an important transport mechanism, the breakthrough curves of the different tracers did not differ from each other as would be predicted based on the free-water diffusion coefficients. This suggested that solute diffusion in wastes requires further study.

Investigating the effect of compression on solute transport through degrading municipal solid waste.

The effect of applied compression on the nature of liquid flow and hence the movement of contaminants within municipal solid waste was examined by means of thirteen tracer tests conducted on five separate waste samples. The conservative nature of bromide, lithium and deuterium tracers was evaluated and linked to the presence of degradation in the sample. Lithium and deuterium tracers were non-conservative in the presence of degradation, whereas the bromide remained effectively conservative under all conditions. Solute diffusion times into and out of less mobile blocks of waste were compared for each test under the assumption of dominantly dual-porosity flow. Despite the fact that hydraulic conductivity changed strongly with applied stress, the block diffusion times were found to be much less sensitive to compression. A simple conceptual model, whereby flow is dominated by sub-parallel low permeability obstructions which define predominantly horizontally aligned less mobile zones, is able to explain this result. Compression tends to narrow the gap between the obstructions, but not significantly alter the horizontal length scale. Irrespective of knowledge of the true flow pattern, these results show that simple models of solute flushing from landfill which do not include depth dependent changes in solute transport parameters are justified.