Contaminant transport in the subsurface - PowerPoint PPT Presentation

1 / 31
About This Presentation
Title:

Contaminant transport in the subsurface

Description:

For a one dimensional flow characterized by a uniform seepage velocity vx, we ... Dispersivity is a characteristic of the system that will need to be determined ... – PowerPoint PPT presentation

Number of Views:114
Avg rating:3.0/5.0
Slides: 32
Provided by: httpserve
Category:

less

Transcript and Presenter's Notes

Title: Contaminant transport in the subsurface


1
(No Transcript)
2
(No Transcript)
3
Contaminant transport in the subsurface
  • Two phase medium VOIDS SOLIDS
  • The voids typically filled with some fluid
    liquid or gas, sometimes both, thus giving a
    three phase medium

4
Darcys law
h L K L/T
5
Constant Head Apparatus (in lab)
6
(No Transcript)
7
Continuity equation, incompressible fluid
Dimensions 1/T, multiplying by density gives
M/L3T
Dimensions 1/L
8
(No Transcript)
9
Velocities in two phase media
10
(No Transcript)
11
  • Fig 6.2 BR

12
Dispersive flux in the subsurface
  • For a one dimensional flow characterized by a
    uniform seepage velocity vx, we might expect the
    dispersion coefficients D to be proportional to
    velocity
  • Dispersivity is a characteristic of the system
    that will need to be determined experimentally
    (recall determination of dispersion coefficients
    in the atmosphere and surface waters)

13
Diffusive and dispersive fluxes in the subsurface
  • Since the diffusive and dispersive fluxes are
    defined similarly, if we let Dx represent the sum
    of diffusivity and dispersion coefficient both
    fluxes will have been included in the
    advection-diffusion equation. Thus
  • We have independent means for measuring
    diffusivity (Dd ) which, is determined by
    temperature, pressure, and molecular properties
    of the diffusing molecule as well as the mixture
    in which it is diffusing
  • (recall the theoretical, empirical correlations
    for gases and liquids)

14
Tortuosity
  • The tortuous path taken in the x direction by
    diffusing molecules means that the diffusive flux
    is less than the flux in the case of an available
    straight path.
  • We can define tortuosity as
  • see (Fig 6.2 BRN)
  • Then our dispersion coefficient expression
    becomes
  • t lt 1, 0.56 0.8 for granular media (Bear,
    1972)
  • Not easy to measure independently

15
Adsorption
  • The adhesion of a component in the fluid to the
    solid surface.
  • Depends on
  • Temperature
  • Solute
  • Solid
  • C mass of solute per volume of fluid
  • S mass of solute per mass of dry solid

16
S
17
(No Transcript)
18
Accumulation in two phases
19
(No Transcript)
20
  • Figure 6.1 BR

21
Advection-dispersion in two phases(no decay)
22
  • Recall the case examined before
  • With initial and boundary conditions
  • S(0,t) So for all tgt0
  • ?S/?x 0 at xL
  • S(x,0) 0 at t0
  • And analytical solution

23
  • In BR notation for subsurface contaminant
    transport, assuming constant Dx
  • Initial and boundary conditions
  • C(0,t) Co for all tgt0
  • ?C/?x 0, at xL
  • C(x,0) 0 at t0
  • Analytical solution (Eqn 6.17 BR) at xL

24
  • The previous formulation describes a tracer (dye)
    test (step input) in a column of porous material.
  • The amount of liquid passed through the column is
    usually measured in number U of pore volumes (Eqn
    6.39 BR)
  • When the solution is expressed in terms of U and
    plotted on appropriate coordinates, the
    dispersion coefficient can be obtained from the
    slope. (See Fig 6.10a BR)

25
  • Figure 6.10a

26
Dispersion in a sand column, Example 6.4
BRN(see Ex6_4BR.xls)
27
PARTICLE SIZE DISTRIBUTION
  • Gaussian, or normal

28
ALTERNATE FORM FOR GAUSSIAN DISTRIBUTION
29
GAUSSIAN DISTRIBUTION, INTEGRATED FORM
  • (Table 8.3 de Nevers)
  • Probability scale a scale linear in z
  • Gaussian distribution gives linear plot on
  • normal (y axis) vs probability (x axis)
    coordinates
  • Slope standard deviation
  • mean value is at z 0

30
Table 8.3 de Nevers
  • Values of the cumulative frequency integral

31
Figure 8.8 de Nevers
  • Graph with (log) probability scale
Write a Comment
User Comments (0)
About PowerShow.com