Impact of Heterogeneity on Vadose Zone Drainage During Pumping: Numerical Simulations of the Borden Aquifer
Unconfined aquifers are in direct contact with the earth's surface; hence, they are an important focus in groundwater recharge and contaminant transport studies. While pumping tests have long been used to quantify aquifer properties, the contribution of drainage from the vadose zone during pumping has been the subject of debate for decades. In 2001, a highly detailed data set was collected during a seven-day pumping test in the unconfined aquifer at CFB Borden, Ontario (Bevan et al., 2005). The frequent observation of moisture content profiles during the test has initiated a closer examination of the vadose zone response to pumping. The moisture profiles collected during the test were obtained using a neutron probe. The neutron data depicts a capillary fringe thickness that increases with both proximity to the pumping well and duration of pumping. This capillary fringe extension results in delayed drainage that persists to the end of the seven-day test with the shape of the transition zone remaining constant (Bevan et al., 2005). Simulations of the pumping test were conducted using Hydrogeosphere (Therrien et al., 2006). Initial simulations were completed based on the conceptual model of a homogeneous and slightly anisotropic aquifer. The simulation results replicated the observed piezometric response, but were unable to produce any change in the thickness of the capillary fringe. It was hypothesized that the discrepancy between observations and simulation results may be the result of assumptions such as the homogeneity of the hydraulic conductivity field. In an effort to replicate this potential mechanism for the observed extension, the conceptual model was updated to better reflect the mildly heterogeneous hydraulic conductivity field of the Borden aquifer. Conductivity fields were generated using the statistical description of the Borden aquifer given by Sudicky (1986) with an adjusted mean log conductivity to better approximate the observed piezometric response. The inclusion of heterogeneity appears to have little effect on the hydraulic head drawdown, or the thickness of the capillary fringe. Heterogeneity does lead to delayed drainage in the drier portion of the vadose zone, where volumetric water content is less than 0.13 m3/m3. This effect is more pronounced with proximity to the pumping well, and is negligible at 15 m from the well. The amount of excess moisture in the vadose zone does not appear to be a function of pumping duration.
The Role of the Capillary Fringe and Unsaturated Zone on the Fate of Organic Compounds Following an Oxygenated Gasoline Spill
Oxygenates are typically added to gasoline aiming to decrease greenhouse gas emissions. MTBE used to be the most common additive in North America, but it is increasingly being replaced by ethanol. In the events of accidental spills, besides presenting an environmental concern itself, oxygenates might change the behaviour of the gasoline compounds in the subsurface. Both MTBE and ethanol are extremely soluble, and therefore are expected to behave differently than most of the gasoline compounds. Previous lab experiments indicated that ethanol stays mainly in the capillary fringe, being transported above the water table along with cosolubilized hydrocarbons. Although this has been demonstrated at lab scale, few field tests have evaluated transport in the capillary fringe, largely due to difficulties associated with sampling. To investigate the fate of MTBE, ethanol and gasoline in the vadose zone, a controlled release of gasoline with 10% ethanol and 4.5% MTBE was conducted at the shallow unconfined sandy aquifer in CFB-Borden, Ontario. Groundwater concentrations above and below the water table were monitored using multilevel wells constructed with porous suction samplers, positioned downgradient the source zone in a row perpendicular to groundwater flow. Soil cores and gas samples were also collected to complement the monitoring. Horizontal transport in the capillary fringe was found to be significant, as indicated by high concentrations in the groundwater above the water table for all compounds tested. Ethanol and MTBE partitioned quickly from the NAPL to the pore water due to their high solubilities. However, ethanol horizontal transport was delayed in comparison to MTBE, assumed to behave as a tracer. Soil cores in the source zone after the spill indicate that ethanol was being retained in the unsaturated zone, above the zone of high hydrocarbons saturation. Water table oscillation between 70 and 20cm bgs resulted in spreading of the gasoline in the unsaturated zone, while ethanol peak concentrations were always found at the first 25cm. The retention of ethanol in the unsaturated zone likely caused delay in its horizontal transport.
Investigating Unstable Water Infiltration into Alcohol Contaminated Soils
A new mechanism causing highly focused, unstable flow exists in soils contaminated with alcohols due to their surface-activity. For example, surface-active compounds can significantly decrease the interfacial tension of the air-water interface and change the pressure-head of the soil water; directly affecting water flow and solute transport in the vadose zone. This study evaluated the fundamental effects of surface-active alcohols on water infiltration into contaminated soils under controlled laboratory conditions. A small scale 3-D glass flow cell and a mini disk tension infiltrometer were used to monitor the rates and physical characteristics of water infiltration from a constant head point source into sands of various textures contaminated with a butanol solution. The results confirmed that water infiltration into these soils is fundamentally and substantially different than the current understanding of infiltration patterns, including previously described mechanisms of wetting front instability. In butanol-contaminated soils, the wetting fronts exhibited highly focused flow with smaller wetted soil volumes, deeper penetration and substantially higher infiltration rates. In addition, the extent of fingered focused flow was confirmed to be texturally dependent, decreasing with grain size and dependent on the constant head boundary. This study characterized a new mechanism of focused, unstable flow with significant implications for groundwater management and solute transport in alcohol contaminated soils.
Understanding Dynamic Soil Water Repellency and its Hydrological Implications
The adverse effects of water repellent soils on vadose zone hydrology are being increasingly identified worldwide in both rural and urban landscapes. Among the affected landscapes are agricultural fields, forests, effluent application sites, golf greens, wetlands, and wildfire sites. In spite of cross-discipline research efforts put forth in recent years, understanding of fundamental parameters controlling soil water behaviour in these systems is lacking. This is due, in part, to inherent complexities of water repellent soil systems and logistical shortcomings of methods commonly used by researchers in-situ and in the lab. As a result, modeling flow in these systems has further proven to be a difficult task. The objectives of our study were 1) to systematically measure and quantify water infiltration and distribution in dynamic water repellent systems and 2) to identify fundamental hydraulic behaviours that lead to the expression of changes in soil water repellency. To achieve this, we combined techniques to elucidate soil- water interactions at a post-wildfire site. Field tests and subsequent lab work reveal essential hydrological information on fire-affected water repellent soils at variable scales and under different burn conditions. Through the use of traditional and newer techniques, our work shows unique and previously unreported behaviour of soil water in these systems. We also address limitations of current field methods used to study repellency and associated infiltration behaviours.
Determination of the Thermal Properties of Sands as Affected by Water Content, Drainage/Wetting, and Porosity Conditions for Sands With Different Grain Sizes
It is widely recognized that liquid water, water vapor and temperature movement in the subsurface near the land/atmosphere interface are strongly coupled, influencing many agricultural, biological and engineering applications such as irrigation practices, the assessment of contaminant transport and the detection of buried landmines. In these systems, a clear understanding of how variations in water content, soil drainage/wetting history, porosity conditions and grain size affect the soil's thermal behavior is needed, however, the consideration of all factors is rare as very few experimental data showing the effects of these variations are available. In this study, the effect of soil moisture, drainage/wetting history, and porosity on the thermal conductivity of sandy soils with different grain sizes was investigated. For this experimental investigation, several recent sensor based technologies were compiled into a Tempe cell modified to have a network of sampling ports, continuously monitoring water saturation, capillary pressure, temperature, and soil thermal properties. The water table was established at mid elevation of the cell and then lowered slowly. The initially saturated soil sample was subjected to slow drainage, wetting, and secondary drainage cycles. After liquid water drainage ceased, evaporation was induced at the surface to remove soil moisture from the sample to obtain thermal conductivity data below the residual saturation. For the test soils studied, thermal conductivity increased with increasing moisture content, soil density and grain size while thermal conductivity values were similar for soil drying/wetting behavior. Thermal properties measured in this study were then compared with independent estimates made using empirical models from literature. These soils will be used in a proposed set of experiments in intermediate scale test tanks to obtain data to validate methods and modeling tools used for landmine detection.
The role of wettability in dynamic capillary pressure effects: fundamental considerations
Dynamic effects in capillary pressure are receiving increased attention in the literature. These effects, where capillary pressure is a function of the rate of change in saturation, have the potential to produce non-unique capillary pressure-saturation relationships, and significantly change the transient distribution of non-wetting fluids. Although multiple experimental and numerical studies have been conducted, the underlying mechanisms remain unclear. Several studies have examined the influence of porous media and fluid properties, but few have examined the interaction of the solid and fluids (i.e. the role of wettability). This study investigates the potential for dynamic effects in capillary pressure in a series of multistep outflow experiments conducted in water-wet, intermediate-wet, and organic-wet sands. The results of these experiments show that wettability affects the magnitude of dynamic effects in capillary pressure. A conceptual model based on the strength of liquid-solid interactions and dynamic contact angles suggests that the damping coefficient used in dynamic capillary pressure expressions may be a function of energy dissipation at the moving three-phase contact line. Numerical simulations of the multistep outflow experiments were used to quantify changes in the damping coefficient and investigate the effect of different relative permeability expressions. The results of this study provide additional insight into the fundamental mechanisms that govern dynamic capillary pressure behavior, and help to elucidate the relative importance of liquid-liquid and solid-liquid interactions.
Heat as a Tracer for Estimation of Soil Drainage Following Irrigation Above a Tile Drain System
Salt-affected soil is one of the most common environmental issues facing the petroleum hydrocarbon industry. Large quantities of brines are often co-produced with gas and oil and have been introduced into the environment through, for example, flare pits, drilling operations and pipe line breaks. Salt must be flushed from the soil and tile drain systems can be used to collect salt water which is then be routed for disposal. An accelerated remediation experiment by soil flushing over a 2 m deep tile drain system was monitored by tensiometers, and thermocouples. Water table elevation was monitored with pressure transducers. A 20 m by 20 m experimental plot was irrigated with 10 m3 of water on each of three consecutive days for an approximate total of 75 mm of water. The irrigation event was repeated three times over a period of 4 weeks. Due to a lack of access to the individual tile drains, direct measurement of drainage rates was not possible. One component of evaluating the success of the accelerated remediation experiment is the fraction of applied irrigation water that infiltrated to depth. Drainage rates beneath the irrigated plot were estimated by heat transport modeling using HYDRUS-1D, a numerical code for the solution of Richards unsaturated flow equation and the heat equation. Temperature and soil matric potential time series were recorded beneath the irrigated plot and at a control location at four depths, 0.3, 0.6, 1.0, and 1.5 m, at 15 minute intervals. Data was recorded for the duration of the irrigation period and for 8 weeks following. The temperature time series beneath the irrigated plot shows a broad increase relative to the control and shorter duration increases in direct response to the irrigation events. Heat modelling results are compared to field measurements.
Investigating the Response of Microbial Communities to Cyclodextrin
Recent studies have found applications of hydroxypropyl-β-cyclodextrin (HPβCD) to be highly effective in removing DDT from soils in situ. However, the persistence of HPβCD within the soil and its impact on soil microbial communities is still unclear. It has been suggested that cyclodextrin might provide a substrate for microbial communities resulting in changes in the ongoing effectiveness of remediation and/or soil hydraulic properties. The potential exists that stimulation of the soil microbial community may contribute to removal of DDT, along with the solubilization effects normally associated with cyclodextrin treatment. This study investigated the response of soil microbial communities from a site undergoing remediation of DDT with HPβCD through microcosm and bench scale column studies. Phospholipid fatty acid (PLFA) analysis and their natural abundance 13C signatures can be used to identify in situ microbial metabolism of HPβCD. Heterotrophic organisms have PLFA with 13C signatures 3 to 6‰ depleted from their carbon source. Cyclodextrin was found to have a δ13C of -16‰ resulting from its formation via enzymatic degradation of cornstarch. In contrast, soil organic matter, had a predominantly C3 plant derived signature and a δ13C of -25‰. Incorporation of HPβCD by soil microbial communities would therefore cause a shift to a more enriched isotopic value. While microcosm studies demonstrated no noticeable change in biomass and few changes in PLFA distribution, column studies treated with a 10% solution of HPβCD demonstrated an approximate doubling of microbial biomass after 6 weeks of application based on PLFA concentrations. Concurrent changes in PLFA distribution further indicated a response to cyclodextrin. Changes in PLFA concentration and distribution were concurrent with isotopic enrichment of PLFA in treated columns. This isotopic enrichment provided direct evidence for microbial consumption of cyclodextrin. Incorporation of 13C enriched cyclodextrin carbon was observed in a wide range of PLFA, with the greatest incorporation for PLFA me15:0, and 17:0 suggesting that these PLFA are produced by organisms most directly involved in metabolism and may represent target biomarkers for communities that consume cyclodextrin.