Investigation of Co-boiling Behaviour During In Situ Thermal Treatment: Experiments and Modelling

Cindy Zhao, Kevin G. Mumford, Bernard H. Kueper

In the proceedings of: GeoMontréal 2013: 66th Canadian Geotechnical Conference; 11th joint with IAH-CNC

Session: Contaminated Sites and Remediation I

ABSTRACT: Dense non-aqueous phase liquids (DNAPLs) such as chlorinated solvents, coal tar and creosote are some of the most problematic groundwater and soil contaminants in industrialized countries throughout the world. In Situ Thermal Treatment (ISTT) is a candidate remediation technology for this class of contaminants. Thermal technologies deliver energy to the subsurface resulting in DNAPL removal by vapourization, steam distillation and boiling. Steam distillation (co-boiling) occurs when immiscible liquid mixtures boil at a lower temperature than the normal boiling point of either liquid. For single component DNAPLs, the soil temperature stays constant during the co-boiling process resulting in a well-defined co-boiling plateau. For example, PCE DNAPL and water have a co-boiling plateau at 88 °C under standard pressure conditions, which is less than the boiling point of either compound. There are four stages of heating soil and DNAPL during ISTT that can be identified from the temperature signal. The first stage corresponds to heating of the system up to the co-boiling temperature. The second stage corresponds to constant-temperature co-boiling of the DNAPL and water. At the completion of co-boiling the DNAPL will, theoretically, be depleted and the remaining pore water will be heated to the water boiling temperature during the third stage. In the fourth stage the water will boil under constant-temperature conditions, following which the system will achieve dry-out. Several studies have been carried out to investigate the processes involved in ISTT (e.g., Heron et al., 1998; Burghardt and Kueper, 2008; Francis and Wolf, 2004). However, the relationships between DNAPL mass removal, temperature, vapour generation rate, vapour concentration, and vapour composition are not fully understood. Temperature data and vapour composition can be indicators of the aforementioned remediation process stages. A screening-level mathematical model was developed to predict the co-boiling temperature and transient composition of the vapour phase during heating of a uniformly distributed DNAPL source. A transient heat conduction model accounting for heat capacity and thermal conductivity of the sand, water and DNAPL mixture was developed to simulate the first and third stages of heating. This was combined with a non-isothermal partitioning model accounting for latent heat and vapourization of water and DNAPL components developed to simulate the second stage of heating (co-boiling). This model is based on Dalton™s Law, which states that the total pressure of the gas mixture is equal to the sum of each individual component™s partial pressure, and Raoult™s Law, which states that the partial pressure of each component is related to the mole fraction of the component in the liquid phase and the pure component vapor pressure. Because the boiling temperature of a mixture is the temperature at which the sum of the vapor pressures equals the ambient pressure, the co-boiling temperature of a multicomponent DNAPL increases as the mole fraction of each component in the DNAPL phase changes due to the earlier removal of the more volatile components. This results in changing mole fractions of each component in the vapour phase, and a co-boiling stage that is not characterized by a constant temperature plateau. A laboratory study was conducted to assess the degree of mass removal, as well as the gas generation rate and the composition of the gas phase as a function of various heating times, initial DNAPL saturations and DNAPL composition. Experiments consisted of a soil, water, and DNAPL mixture packed in a 1 litre (L) jar that was heated in a convection oven. Gas produced during each experiment, which consisted of steam and DNAPL components, was collected and condensed to quantify the gas generation rate. The temperature of the contaminated soil was measured continuously using a thermocouple, and was used to identify periods of heating, co-boiling and water boiling. Samples were collected from the aqueous and DNAPL phases of the condensate, as well as from the source soil, at different heating times, and analyzed by gas chromatography/mass spectrometry (GC-MS). Experiments were conducted with single component and multicomponent DNAPLs, for different initial NAPL saturations (5%, 20%, and 40%) and heating times (approximately 1-9 hours). An additional experiment was also conducted to investigate the sensitivity of the temperature measurement location relative to the configuration of a DNAPL source. In this experiment multiple thermocouples were installed in a 1 L jar that contained DNAPL at 20% saturation in only one side of the sand pack. Experiments conducted with a single component DNAPL (tetrachloroethene; PCE) exhibited a co-boiling plateau at 88°C ± 1°C. Heating of multicomponent DNAPLs composed of PCE, chlorobenzene (CB) and 1,2-dichloroethane (1,2-DCA) showed no well-defined constant-temperature co-boiling plateau, but rather a definable co-boiling stage from 80 to 89 °C. The experimental co-boiling temperature for both PCE and multicomponent DNAPLs were found to be matched well by the mathematical model predictions. However, a comparison of predicted and observed boiling behaviours showed a discrepancy at the end of the co-boiling period, with earlier temperature increases occurring in the experiments.

Please include this code when submitting a data update: GEO2013_318

Access this article:
Canadian Geotechnical Society members can access to this article, along with all other Canadian Geotechnical Conference proceedings, in the Member Area. Conference proceedings are also available in many libraries.

Cite this article: