PHYSIC-CHEMICAL PROCESSES FOR TRANSPORT OF

CONTAMINANTS.
Yetzabbel Flores1
1
Institute of Environmental Management

INTRODUCTION
The study of contaminant transport is an important part of hydrogeology science- It allows to describes,
prevent, forecast and manage migration of hazardous agents on the groundwater flow. It combines
geochemistry science with hydrogeology basis to model the contaminant migration trough the aquifer
systems based on the particular behavior of every class of contaminant.
It involves the interface between solution and solids, where the properties of the surface as particle size
play a main role with three significant consequences (Langmuir, 1997):
 Increase in the solubilities of small particles
 Stability on suspension and transport availability.
 Unsatisfied surface charge making them a potential sorbent
At first, it is important understand the mechanisms of movement of ‘x’ pollutant plume as well as the
factors than enhance it migration. For example, the flowrate velocity has determinative roll in the
movement of the pollution (Stumm, 1992). The fundamental physic-chemical processes controlling the
contaminant transport will depend of the general contaminant class and the geological medium where
the spread event is taking place. The mechanism for the three main contaminants classes are (Hiscock,
2005):
- For non-reactive contaminants: Advection, and dispersion.
- For reactive contaminants: precipitation, sorption, microbiological reactions, and attenuation.
- For non-aqueous phase liquids (NAPLs): immiscibility and dissolution.

TRANSPORT OF NON-REACTIVE DISSOLVED

CONTAMINANTS
The non-reactive contaminants are affected by advection and hydrodynamic dispersion processes
(Hiscock, 2005). In the advection, the contaminant is transported by water flow due to a hydraulic
gradient (Favaretti and Cossu, 2018) where the velocity is calculated by Darcy’s Law. In the dispersion,
the concentration of the contaminant decrease with distance from the source by effect of mechanical
mixing and molecular diffusion (National Research Council, 1984; Hiscock, 2005).
To shallow groundwater, advection and mechanical dispersion is dominant. While, diffusion is dominant
in geological environments of low gradient and low hydraulic conductivity (Hiscock, 2005; Favaretti and
Cossu, 2018). It is important point that a continuous pollution source will produce an elongate plume,
whereas a single pollution point source will produce a slug that grows with time while becoming less
concentrated as a result of dispersion as the plume moves in the direction of groundwater flow.
The net movement of solute under a concentration gradient is known as diffusion. In one dimension, the
second law of Fickian theory describe time-varying change in concentration for a change in solute flux as:
2
∂C ¿∂ C
=−D 2
∂t ∂x
¿
Where: D _ molecular diffusion coefficient for a solute in a porous media.

For an instantaneous step-change of concentration in an infinite aquifer space exist the following
analytic solution:

C
C0
=erfc
( x
2 √D t
¿ )
Where: erfc _ complementary error function (Appendix 8; Hiscock,2005)
C ¿ initial concentration at x=0 and t=0
C ¿ concentration at x position at time t
The mechanical dispersion of a dissolved solute is represented by:
Mechanical dispersion=∝ v
Where: ∝_ dispersivity of the porous media
v ¿ advective velocity (in average the linear velocity) of groundwater flow
Dispersivity is a natural and scale dependent physical characteristic of porous media, it determines the
degree of contaminant spreading. Also, the dispersivity is affected by the anisotropy of the porous
media, usually having greatest longitudinal direction than perpendicular or transversal direction of the
groundwater flow (National Research Council, 1984; Hiscock, 2005)..
Combining the mathematical description of mechanical dispersion with the molecular diffusion
coefficient gives an expression for the hydrodynamic dispersion coefficient, D, in longitudinal ( l) and
transversal (t) directions, as follows:
¿
Dl=∝l v l + D
¿
Dt =∝t v t + D

By laboratory and field experiments was found that at low velocity, molecular diffusion is the important
contributor to hydrodynamic dispersion, although with little effect in spreading the tracer front. At high
velocity, mechanical dispersion dominates and the breakthrough curve adopts a characteristic S-shape
with some of the tracer moving ahead of the advancing front and some lagging behind, as controlled by
the tortuosity of the flowpaths (Hiscock, 2005).
From the above explanations about solute transport process, the one-dimensional form of the solute
transport equation describing the time-varying change in concentration of non-reactive dissolved
contaminants in saturated, homogeneous, isotropic material under steady-state, uniform flow conditions
and undergoing advection and hydrodynamic dispersion in the longitudinal direction, is given as:
 2
∂C ∂ C ∂C
Dl 2
−v l =
∂l ∂ l ∂t
The analytical solution to the advection-dispersion for a step function concentration input, provided by
Ogata and Banks (1961), is written as:

C
C0 [ ( )
= erfc
l−v l t
2 √ Dl t
vl
( ) (
+exp l erfc
Dl
l+v l t
2 √ Dl t )]
Under the following boundary conditions: C ( l , 0 )=0l ≥ 0

C ( 0 , t )=0 t ≥ 0
C ( ∞ , t )=0 t ≥ 0
When the medium has a large dispersivity value, longitudinal distance, or time; the second term of the
equations is negligible. Besides, if molecular diffusion is assumed smaller than mechanical dispersion,
then √ D l t can be write as √ ∝t v t t .

TRANSPORT OF REACTIVE DISSOLVED CONTAMINANTS

The reactive substances change in concentration not only similar than the conservative solute species,
but also by chemical reactions with the aqueous or/and solid phase (National Research Council, 1984;
Hiscock, 2005).
The dominant process affecting the transport of this type of contaminant is sorption, such as adsorption
and the partitioning of contaminants.
Adsorption is the attachment of an ion in solution to a pre/existing solid surface (White, 2013). It affects
the surface of suspended particles and colloids which influence their aggregation and transport, and for
its full understanding is required the physical and chemical property characterization of the solute, the
sorbent and the solvent (Stumm, 1992; Langmuir, 1997).
 Surface complex formation: Coordinate bond between metals and ligands at the surface.
 Electrostatic interactions: The electrical charge of the surface, called electrostatic force, affects the
surface complex formation and loosely binds other ions.
 Hydrophobic adsorption: Non polar substances are insoluble in water, and they tend to be adsorbed
on the surface because they are repelled by water.
When inspecting contaminant plumes from field data, it is also useful to choose a non-reactive tracer
such as chloride to represent the advective transport of groundwater for comparison with the contami-
nant behaviour.

The adsorption is described in terms of adsorption isotherms, where initially the measurements were
done at constant temperature (Merkel and Planer-Friedrich, 2002).

The partition coefficient, K ad , describe the process of contaminant sorption from aqueous phase to the
solid phase. The values are calculated in laboratory and derivate in the sorption isotherms (National
Research Council, 1984; Hiscock, 2005).
Linear regression isotherm:
The simplest form of adsorption isotherm with advantage of simplicity, but with a serious disadvantage
of no upper limit because the relation is linear.

θ M =K ad [ M ]

Where: θ M _ fraction of surface sites occupied by species M

[ M ]_ dissolved concentration of M
K ad _ adsorption equilibrium constant
Using this relation, a transport equation (Merkel et al., 2002; Stumm, 1992) can be writing as:
Bd
∗θM
q Bd
Rf =1+ =1+ ∗K ad
[M] q
Where: Rf_ retardation factor
Bd _ bulk density
q _ water content
Langmuir isotherms:
It reflects saturation of the surface. It is derivate from kinetic fundamentals or from thermodynamic
principles, under the assumptions (1) the free energy of adsorption is independent of the number of sites
available, (2) the law of mass action applies, and (3) only a monolayer of adsorbate can form.

K ad [ M ]
θM=
1+ K ad [ M ]
Using this relation, a transport equation (Merkel et al., 2002) can be writing as:

Rf =1+
Bd
q
∗
[ K ad
( 1+ K ad [ M ] )
2
]
Freundlich isotherms:
It is a merely empirical exponential relation between sorbet and dissolved molecules, and it doesn't
show saturation (White, 2013). It is based on a model of a multi-lamellar coating of the solid surface
assuming a priori that all sites with the largest binding energy are occupied (steep section of the curve)
and with increasing grade, sites with lower binding energy (flattening of the curve) are occupied. (Merkel
et al., 2002).
n
θ M =K ad [ M ]
Where: n _ an empirical constant
Transformation on a retardation factor for transport equation (Merkel et al., 2002) can be writing as:
Bd n−1
Rf =1+ ∗n∗K ad∗[ M ]
q
TRANSPORT OF NON-AQUEOUS PHASE LIQUIDS
The transport of non-aqueous phase liquids (NAPLs) concerns the contamination of groundwater by
organic compounds and includes dense non-aqueous phase liquids (DNAPLs) with a density greater than
water. They have relatively low solubility in water and partition preferentially towards organic material
contained solid phases, showing hydrophobic sorption behavior (Hiscock, 2005; Fetter, 2001).
In a condition of multiphase flow, each phase is competing for the available pore space, and appears the
wettability definition as tendency for a given fluid to be attracted to a surface more than another fluid
(Hiscock, 2005). For example, water can always be considered the wetting fluid with respect to NAPLs or
air, because if a porous material is water-wet and a compound such as oil is introduced, the water will
continue to occupy the capillary space.
Chlorinated solvents are DNAPLs, they are volatile and low viscosity, consequently they are more mobile
than water in a porous material. On infiltrating through the unsaturated zone, solvents leave behind a
residual contamination which partitions into a vapour phase that subsequently migrates upwards and
laterally by diffusion. The remaining contaminant mass migrates downwards under its own weight and
through the water table until halted by the base of the aquifer, or by some other intermediate
impermeable barrier (Hiscock, 2005; Fetter, 2001).
In fractured material, if the weight of DNAPL exceeds the displacement pressure of water contained in a
fracture, the DNAPL can potentially migrate to significant depths.
LNAPLs behave in a similar manner than DNAPL, except that by density they will float on the water table.
In this case, when a spilled occurs, the oil will migrate vertical into the vadose zone until it reaches the
top of the capillary fringe. Some LNAPL will be left trapped in the vadose zone, but the other portion will
reach the capillary fringe and accumulate until the ‘oil table’ develop. As the weight of LNAPL increases,
the capillary fringe will become thinner until mobile or ‘free’ product accumulates. The mobile LNAPL can
migrate in the vadose zone, following the slope of the water table, while the dissolved components can
disperse with the advecting groundwater (Hiscock, 2005; Fetter, 2001).
In fractured rocks, LNAPL will typically resist migration below the water table but if there is sufficient
weight, LNAPL can penetrate below the water table to a limited extent when the pressure exerted by the
LNAPL exceeds the displacement pressure of the water in the fractures (Hiscock, 2005, Fetter, 2001).
Non-polar organic molecules (VOCs), tend to partition preferentially into non-polar environments. At
low concentrations, the sorption of non-polar compounds on to organic material can very often be
modelled with a linear isotherm. For a rapid assessment of sorption behaviour of VOCs, find empirical
correlations between KOC, the organic carbon-water partition coefficient, and the properties of known
substances like the octanol-water partition coefficient, KOW (Hiscock, 2005; Fetter, 2001).

The partition coefficient, K ad , for application in the retardation equation can now be normalized to the
weight fraction organic carbon content of the sediment, f OC, assuming that adsorption of hydrophobic
substances occurs preferentially on to organic matter, as follows:
K ad
K OC =
f OC
BIBLIOGRAPHY
Favaretti, M. and Cossu, R. (2018) ‘Mineral Liners’, in Solid Waste Landfilling. Elsevier Inc., pp. 289–312.
doi: 10.1016/b978-0-12-407721-8.00016-4.
Fetter, C. W. (2001) Applied Hydrogeology. Fourth. Edited by L. Patrick. Prentice Hall.
Hiscock, K. M. (2005) Hydrogeology. Principles and Practica. Available at: www.blackwellpublishing.com.
Langmuir, D. (1997) AQUEOUS ENVIRONMENTAL GEOCHEMISTRY. Edited by R. McConnin. Upper Saddle
River: Prentice Hall.
Merkel, B. J. and Planer-Friedrich, B. (2002) Groundwater Chemistry. A practical guide to modeling of
natural and contaminated aquatic systems. Edited by D. K. Nordstrom. Springer.
National Research Council (1984) Groundwater Contamination, Groundwater Contamination.
Washington D.C.: The National Academies Press. doi: 10.17226/1770.
Stumm, W. (1992) Chemistry of the solid-water interface: processes at the mineral-water and particle-
water interface in natural systems. John Wiley & Sons Inc.
White, W. M. (2013) ‘Mineral surfaces and their nteraction with solutions’, in. Wiley & Sons, pp. 253–
265.