Article pubs.acs.org/est
Poly-Use Multi-Level Sampling System for Soil-Gas Transport Analysis in the Vadose Zone Philipp A. Nauer, Eleonora Chiri, and Martin H. Schroth* Institute of Biogeochemistry and Pollutant Dynamics (IBP), ETH Zurich, Zurich, Switzerland S Supporting Information *
ABSTRACT: Soil-gas turnover is important in the global cycling of greenhouse gases. The analysis of soil-gas profiles provides quantitative information on below-ground turnover and fluxes. We developed a poly-use multi-level sampling system (PMLS) for soil-gas sampling, water-content and temperature measurement with high depth resolution and minimal soil disturbance. It is based on perforated access tubes (ATs) permanently installed in the soil. A multi-level sampler allows extraction of soil-gas samples from 20 locations within 1 m depth, while a capacitance probe is used to measure volumetric water contents. During idle times, the ATs are sealed and can be equipped with temperature sensors. Proofof-concept experiments in a field lysimeter showed good agreement of soil-gas samples and water-content measurements compared with conventional techniques, while a successfully performed gas-tracer test demonstrated the feasibility of the PMLS to determine soil-gas diffusion coefficients in situ. A field application of the PMLS to quantify oxidation of atmospheric CH4 in a field lysimeter and in the forefield of a receding glacier yielded activity coefficients and soil-atmosphere fluxes well in agreement with previous studies. With numerous options for customization, the presented tool extends the methodological choices to investigate soil-gas transport in the vadose zone.
■
INTRODUCTION Soils play a crucial role in the turnover of greenhouse gases such as CO2, CH4 and N2O.1,2 For instance, the sole terrestrial sink for atmospheric CH4 is provided by aerobic methaneoxidizing bacteria that consume CH4 diffusing from the atmosphere into the soil vadose zone.3 Among the methods to quantify soil-gas turnover, the soil-gas profile method is one of the few techniques that can provide depth-resolved qualitative and quantitative information on soil-gas fluxes, turnover rates, and isotopic signatures.4−11 The method involves the sampling of soil gas at different depths in a soil profile. In most cases the gas samples are collected with a syringe, transferred to an airtight container and measured ex situ. However, alternative techniques for continuous in situ measurements of gas concentrations have also been developed.12−14 Independent of the measurement approach, equipment needs to be installed in the soil to sample gas at different depths. A variety of techniques have been described in the literature, ranging from relatively simple open capillaries or tubes7,15,16 to wells or silicon coils that equilibrate with soil-gas due to molecular diffusion.15,17 Multi-level samplers (MLS) featuring ports at different depths have initially been developed for groundwater sampling,18 but were later adapted to sample soil gas in the vadose zone.19 Most of these methods involve excavation and permanent burial of equipment, which is tedious in stony soils and might create significant disturbance to natural © 2013 American Chemical Society
soil structure. Moreover, the deployment of multiple sampling probes is a substantial investment in time and material, often leading to compromises in depth-resolution and number of sampled locations. For quantification of gas-turnover rates or momentary unit fluxes Jz at a certain depth z, Fick’s first law can be applied: ⎛ dC ⎞ Jz = −⎜ ⎟ Deff ⎝ dz ⎠ z
(1)
Measurements of the concentration gradient dC/dz from the gas samples need to be complemented by determination of the effective diffusion coefficient in soil, Deff [cm2 h−1]: Deff = θaDaτ
(2)
The diffusion coefficient in air, Da [cm2 h−1], is an inherent property of the gas in question, the air-filled porosity θa (total porosity θt [m3 m−3] minus volumetric water content θw [m3 m−3]) and the tortuosity factor τ are properties of the investigated soil. Total porosity often shows little variation within a soil profile, but θw can vary from very low values to near saturation.20 Hence, to quantitatively analyze soil-gas Received: Revised: Accepted: Published: 11122
May 2, 2013 July 24, 2013 August 20, 2013 August 20, 2013 dx.doi.org/10.1021/es401958u | Environ. Sci. Technol. 2013, 47, 11122−11130
Environmental Science & Technology
Article
Figure 1. Illustration of the poly-use multi-level sampling system (PMLS) and involved equipment. The system is based upon the permanent installation of perforated ATs in the soil (a). Soil-gas samples are taken through the walls of the ATs with a MLS featuring an inflatable packer system (b), while water content is measured with a capacitance profile probe (c). Between sampling events, the ATs are sealed from the inside with inflatable sealing tubes that can be equipped with data-logging temperature sensors (d). The seal prevents exchange of soil gas between different depths and the atmosphere and is removed before a sampling event.
profiles it is necessary to accurately determine θw at the time and location of soil-gas sampling. Many studies that employed the soil-gas profile method used time-domain reflectometry (TDR) to determine θw.7,11,13 TDR measures the travel time of an electric pulse along the length of a conductor installed in the soil. The resulting apparent dielectric constant primarily depends on θw.21 However, the placement of the conductor away from the location of gas sampling and potential disturbance during installation might be a source of bias, caused by soil-specific lateral heterogeneity. Also, multiple sampling depths and locations generally require multiple conductors, as layered soils may exhibit abrupt changes of θw with depth. Recently, a new generation of multi-level electrical-capacitance probes have been introduced that measure θw at multiple depths through previously installed access tubes (ATs).22,23 To our knowledge, the principle of using the AT of such probes for multiple measurement and sampling purposes has not been attempted in soil science before. The tortuosity factor τ accounts for higher diffusion resistance in porous media compared to with free air. Various predictive models exist to derive τ from θt and θw.24,25 The accuracy of such models greatly depends on the quality of measured parameters (they generally incorporate power laws), and the choice of the appropriate model for the respective soil.26 In-situ determination of τ using radon concentration profiles27 or tracer-test methods28 can be a valuable alternative, but equipment needed for this purpose is similar to that for soil-gas sampling. Finally, almost all studies investigating soils require the measurement of soil temperature as an important parameter in physical and biological models. Often the necessary equipment is little more than a side note in soil-science publications.
Nonetheless, temperature sensors and appropriate data loggers can contribute substantially to equipment costs. Likewise, burial of temperature sensors can cause additional disturbance to soil structure. Here we present a new approach, the poly-use multi-level sampling system (PMLS), for multi-level soil-gas sampling and concomitant determination of θw and soil temperatures through ATs. Our major motivations for its design were (i) to improve depth-resolution of soil-gas sampling; (ii) to allow determination of θw at the exact location of soil-gas sampling and at multiple depths; (iii) to ease the installation of gas-sampling probes and enable sampling in stony soils, for example, glacier forefields; and hence (iv) to substantially reduce time and material needs for extensive sampling campaigns with multiple sampling locations. For proof of concept we compared the performance of the PMLS in a field lysimeter against conventional techniques: capillary probes for soil-gas sampling (by analyzing soil-CH4 concentrations), and TDR for θw measurements. In addition we used the PMLS to perform a tracer test for the in-situ determination of Deff. Furthermore, by applying the PMLS for sampling and estimation of Deff we calculated rate coefficients and fluxes of atmospheric CH4 oxidation in the field lysimeter and in a young glacier-forefield soil.
■
THE POLY-USE MULTI-LEVEL SAMPLING SYSTEM Access Tubes and Installation Procedure. We employed commercially available ATs (Delta-T Devices Ltd., Cambridge, UK) that were compatible with the same manufacturer’s PR2 capacitance profile probe. The 1.115 m long ATs were made of epoxy fiberglass (26 mm i.d.; 1 mm wall thickness). We modified each AT by drilling a total of 160 1 mm diameter holes for soil-gas sampling into its walls (eight holes evenly 11123
dx.doi.org/10.1021/es401958u | Environ. Sci. Technol. 2013, 47, 11122−11130
Environmental Science & Technology
Article
spaced around the tube’s circumference, in 20 depths at 5 cm increments). Installation of the AT into the soil required a hole of 1.15 m depth and 25−28 mm diameter (Figure 1a). To quickly probe whether the installation depth could be reached we used a cordless drill with a 10 mm drill bit of 1.3 m length. In aggregated, well-structured soils the installation hole should be created using an auger according to the installation manual of the manufacturer (Augering Manual 2.0, Delta-T Devices Ltd., Cambridge, UK). This ensures a snug fit between AT and the soil, minimizing preferential gas flow near the tube and facilitating accurate water-content measurements. In sandy soils with negligible aggregation we used a drill bit with a diameter slightly smaller than the AT (25 mm); the AT was then inserted by hand and additional hammering using a shockabsorbing protection cap (Figure 1a). In loose, stony soil devoid of aggregation (e.g., glacier-forefield soil) installation was best accomplished by hammering an iron pin of 28 mm diameter and 1.3 m length into the probing hole. The pin displaced smaller stones and pebbles that would otherwise have blocked any installation equipment, and the slight compaction prevented the hole from collapsing upon removal of the iron pin. The AT was then inserted by hand and light hammering. After installation, the soil should be allowed to consolidate for some time prior to the first sampling, ideally until the passing of several rain events. Multi-level Sampler. In principle, our MLS prototype consisted of 20 stainless steel capillaries of 1 mm i.d. and different lengths (Figure 2a-A), each connected to a sampling disk (Figure 2a-B) made of polyoxymethylen at a specific depth, and to a valve (Figure 2a-C) attached to an aluminum sampling head (Figure 2a-D) on top of the MLS. The capillaries were aligned in a circle around a central steel rod (Figure 2a-E), guided by the sampling disks. While most of the capillaries passed through the disks, the capillary to sample the specific depth ended and connected with a horizontal hole leading to a circular cavity around the disks’ perimeter. This cavity was aligned with the gas-sampling holes of the AT. On the upper and lower part of the disks, tubular latex-rubber membranes (Michelin Aircomp Latex 22/23−622, ClermondFerrand, France) were tightly attached with several windings of nylon string (Figure 2a-F and d), enclosing all capillaries and the interspace between two disks. Once placed inside an AT, the rubber membranes of the MLS were inflated through the central steel rod using a generic bicycle pump (Figure 1b). Via an opening at the bottom of the central rod the air could propagate freely inside all membranes through the open channels guiding the capillaries. The inflated membranes pressed against the inner walls of the AT and therefore created a packer system between the sampling disks (Figure 2e). Individual soil-gas samples from each of the 20 depths could then be collected from the valves at the sampling head (Figure 1b and c), after discarding the dead volume of the capillary and cavity (
