Effects of capillarity and heterogeneity on flow of organic liquid in soil

Research output: Thesisinternal PhD, WU


Contamination of groundwater by organic liquids, such as gasoline, fuel oils and chlorinated hydrocarbons, forms a serious treat to subsurface water resources. These liquids have a low miscibility with water and move as a discrete liquid phase. A small part of the liquid may dissolve in water and small concentrations can be hazardous for humans. These organic liquids that are also referred to as Non-Aqueous Phase Liquids (NAPLs), may enter the subsurface from a surface spill or a leaking underground pipe or tank. Following a release of sufficient NAPL into the subsurface, NAPL moves through the pores of the soil due to gravitational and capillary forces.

An enhanced understanding of NAPL flow in the subsurface is needed to identify the critical processes and to determine key parameters in the context of remediation strategies at the field scale. In this thesis we study the behavior of NAPL in soil experimentally and numerically. we focus on the physical aspects of this behavior in the unsaturated part of the soil. This implies that we study the movement, interaction and (re-)distribution in a porous medium (soil) of three fluid phases: water, NAPL and air. The porous medium is water-wet, i.e. water is the wetting fluid, NAPL is the intermediate wetting fluid and air is the non-wetting fluid. This leads to typical three-phase flow behavior. We focus on two problems that are both related to NAPL redistribution into unsaturated soil after a spill: (1) non-drainable residual NAPL and (2) the effect of soil layers on the redistribution of NAPL after a spill. In addition to the introduction, the thesis contains 4 chapters which can be read as self-contained papers.

Non-drainable residual NAPL is considered to be the NAPL that remains in the capillaries of a porous medium after (gravity) drainage and displacement by air in the presence of residual water in a water-wet soil. Although for remediation strategies it is important to consider this residual NAPL, until now it has not been incorporated into multi phase flow models. In Chapter 2 we develop a constitutive set of saturation-capillary pressure ( Pc-S ) relationships that accounts for non-drainable residual NAPL and that can be used by multi-phase flow models. These relationships are derived from a conceptual pore scale model of which the soil is represented by packed spheres. According to the model, NAPL becomes immobilized (residual) below a certain critical total liquid saturation, S tc. This critical total liquid saturation is porous medium and fluid dependent. Next, the non-drainable residual NAPL saturation increases, when the spreading coefficient of a NAPL decreases. Furthermore, a higher water saturation results in a lower non-drainable residual saturation. Although the model adequately simulated one individual experimental non-drainable residual NAPL data set, this agreement is a shallow basis for assessing whether the model assumptions are valid.

For this purpose we developed an experiment especially designed to measure three-phase Pc-S relationships and non-drainable residual saturation including independent measurement of the water saturation and pressure. We performed four experiments. We used the fluids water, non-spreading dodecane and air, and sand as the porous medium. The measured non-drainable residual saturation was at most 0.081. From the experimental results no correlation could be observed between the water saturation and the residual NAPL saturation. This might indicate that other mechanisms are involved that are not captured by the constitutive model. In advance of the development of the experimental set up, we searched in the literature for available macroscale experimental data that could provide us with non-drainable residual saturation data that are related to porous medium properties, fluid properties and fluid saturations. After a careful selection we obtained 26 non-drainable residual NAPL data from 9 references. Measured values of non-drainable residual saturation varied between 0.001 and 023. From these data only a relationship between the spreading coefficient Cs , and the residual saturation S or , could be assessed. The value of S or decreases as Cs becomes less negative. It would require a much more comparable description of the different porous media, and a systematic data acquisition for S or as a function of macroscale properties to assess which other factor than the spreading coefficient might be hidden in the data. Changes of porous medium properties such as permeability and porosity across an interface between two soil layers might change the redistribution of NAPL. The mobility may decrease or increase across the interface, which is dependent on the permeability of the layers, the capillarity of the layers and the wetting role that is played by the NAPL with respect to water and air. The effect of these changes have been studied in Chapter 4 and 5.

In Chapter 4 we focused on the effect of an inclined soil layer with respect to the watertable. Two experiments were performed using a sand filled transparent chamber. The first experiment consisted of Light-NAPL infiltration into a fine sand matrix containing a coarse sand layer and the second experiment consisted of LNAPL infiltration into a coarse sand matrix containing a fine sand layer. The sand was partially saturated with water. We modeled the experiments numerically. The observed LNAPL behavior can be understood by considering that LNAPL might encounter two types of barriers. On the one hand, it may accumulate and spread above a coarse sand layer, that has a low water saturation and, that acts as a capillary barrier for the LNAPL before it may infiltrate into the layer. An on the other hand, it may accumulate and spread above a water saturated fine sand layer that has a high entry pressure for the infiltrating LNAPL. Sensitivity analysis performed with the numerical model shows that the qualitative LNAPL redistribution, in case of inclined layers, is mainly subject to the capillarity contrast between sands.

In Chapter 5 we considered Dense-NAPL infiltration into a water unsaturated porous medium that consists of two horizontal layers, of which the top layer has a lower intrinsic permeability than the bottom layer. The layer interface forms a barrier to DNAPL flow, which causes the DNAPL to spread out horizontally just above the interface. We have developed an analytical approximation to estimate the DNAPL pressure and saturation and the horizontal extension of the DNAPL plume just above the layer interface at steady state for low water saturations. The analytical approximation shows that the steady state DNAPL infiltration is determined by five dimensionless numbers: the heterogeneity factorg, the Brooks and Corey capillary pressure parameterl, the gravity number Ng , the ratio of the capillary and gravity numbers Nc/Ng , and the critical DNAPL pressure P oc, which indicates the effect of the water saturation on the flow of DNAPL. We compared its predictions with the results of a numerical three-phase flow simulator for a number of parameter combinations. For most of these combinations the analytical approximation predicts the DNAPL pressure and saturation profiles at the interface adequately, although it slightly underestimates the horizontal spreading of the plume at the interface. Using the analytical approximation, we carried out a sensitivity study with respect to the maximum horizontal extension of the plume. The extension of the plumes appears to be highly sensitive to variation of the dimensionless numbers P oc,landg. The extension increases for increasing values oflandgand for decreasing values of P oc.

Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • Wageningen University
  • van der Zee, Sjoerd, Promotor
Award date16 May 2003
Place of PublicationWageningen
Print ISBNs9789058088468
Publication statusPublished - 2003


  • soil
  • water
  • oils
  • air
  • flow
  • water flow
  • air flow
  • capillary rise
  • fluids
  • liquids
  • models
  • drainage
  • soil pollution


Dive into the research topics of 'Effects of capillarity and heterogeneity on flow of organic liquid in soil'. Together they form a unique fingerprint.

Cite this