A ‘fill and draw’ tracer test at the Landgraaf landfill pilot November 2010 to March 2011

Abstract

Fill and draw is a flushing technique that may be used to help remove contaminants from landfilled waste and, in so doing, accelerate the remediation of landfill sites. A simple fill and draw technique involves saturating in situ waste from the bottom up with water and then, normally following a pause, draining the landfill. In this way, soluble and diffusive contaminants can effectively be recovered from the waste mass. Repeating this process will sequentially reduce the masses of contaminants.

This report describes a large-scale fill and draw tracer test that was performed in a 25 000 tonne test cell at a landfill in Landgraaf (The Netherlands), which contained a ~2.5 m saturated zone at its base. Hydrogeological tracers were mixed with 800 m3 water and introduced into the basal drainage system of the landfill test cell over a period of 17 days. 69 kg of bromide (in the form of potassium bromide) and 0.56 kg of Rhodamine WT (RWT) were evenly mixed with the water and injected at a constant rate of 2 m3/hour through three slotted drains equally spaced across the base of the test cell. Towards the end of the injection phase, 0.45 kg of Li was also added to the tracer mix to help assess the properties of the test cell’s basal sand layer.

After a 3 day pause the injected water and tracer was then pumped from the cell using the same infrastructure that was used to introduce it. The tracer breakthrough data and the analysis of contaminants recovered from the waste during pumping were used to determine a number of important contaminant transport processes.

Abstraction of the injected tracer occurred over two periods of constant pumping separated by a 40 day pause. The initial abstraction of 555 m3 over a period of 26 days occurred at an average rate of 0.9 m3/hr. The second 53 day phase removed a further 543 m3 of leachate at an average rate of 0.5 m3/hr. Drainage was from the three basal drains, with flow rates controlled manually. During abstraction, samples were collected from each drain, and from the combined discharge. An inline fluorometer, installed in the discharge, was also used to monitor the tracer return.

Leachate heads were measured in fully-screened wells and piezometers with discrete response zones. By the end of injection, the leachate table had been raised by ~2.2 m. The wells and piezometers were also used to collect water samples, before and throughout the tracer test. Average leachate levels at the end of the test were returned to starting conditions (i.e. ~2.5 m saturated depth).

Water samples were analysed in the laboratory for a wide range of parameters, including the introduced tracers Br, Li and RWT, and indigenous leachate contaminants including chloride, ammonia, electrical conductivity and dissolved organic carbon. The mass recovery of the bromide was between 61-83%.

The concentration of bromide measured in the piezometers with a deeper response zone (i.e. those nearer the base of the test cell), showed a higher peak during injection period than the shallower piezometers. The deeper piezometers also maintained higher concentrations of tracer at the end of the test than those with a shallower response zone. An inverse response was seen in the analysis of indigenous contaminants (including ammonia and chloride). In general, deep piezometers showed a more significant change in concentration than the shallower piezometers.

Contaminant transport parameters for the waste were recovered from model fits to the introduced tracers Br and Li, and to indigenous contaminants Cl and NH4. The tracer Li was mainly used to characterise the influence of the basal sand layer.

A simple 1D, two-porosity advection diffusion model (DP-Pulse) was used to provide a ‘first approximation’ fit to the data. Both the tracers and the indigenous contaminants fit the model well, suggesting that despite the complexity of the system, a simple conceptualisation is sufficient for estimating bulk contaminant flushing parameters.

The results for Br and Cl, return extremely high values of tcb (the characteristic diffusion time of an immobile block of waste), in excess of 500 days. This may indicate that there are either large sections of the waste which are isolated from the flow, or that there are highly-localised preferential flow paths. The recovered values of tcb, indicate that the characteristic dimension of the immobile blocks is at least 20 cm and possibly greater than 60 cm.

The block diffusion time for NH4, was two orders of magnitude lower than for the other species, which may indicate that the spatial distribution of ammonia was a lot more varied than Cl.

The results of this trial demonstrate the viability of the ‘fill and draw’ concept using the basal leachate drainage system of (hydrogeologically suitability) landfills as a potential accelerated landfill remediation technique. The models that have been developed will form the basis of future design tools. Further work is required to assess the efficiency of this modus operandi versus other landfill flushing techniques.

More information

Identifiers

ORCID for T. Rees-White: orcid.org/0000-0001-9009-8432

ORCID for R. Beaven: orcid.org/0000-0002-1387-8299

ORCID for N. Woodman: orcid.org/0000-0002-5571-0451

Catalogue record

Export record