DEFINITION
Two families of curves: 1) Flow lines: Represent the paths that water would flow. 2) Equipotential lines: Lines of constant total head.
Flow Nets
Groundwater flow system can be represented by a 3-d set of equipotential surface (flow lines if in 2-d) and flow lines that intersect at right angles. The flow net should be constructed so that the flow lines and equipotential lines define a curvilinear square.
Flow Nets
Or, for vertical flow in an unconfined aquifer,
If K is uniform and the water table thickness does not vary significantly
Flow Rate
nf = number of flow channels q = total flow through all channels if b = l (square elements)
Steady- State horizontal flow through a sand aquifer confined by layers of clay (same as previous example), with Ksand = 10-4 m/s and n = 0.3
flow line from entrance to exit Draw equipotential lines which are perpendicular to the flow line and divide the BVP into squares
9
Start by drawing a
What direction is flow along the top and the bottom of the sand layer? Lets look at 1 streamtube
dx = distance between two equipotentials dh = head drop between two equipotentials dm = width of stream tube = area per meter to stream tube Q = volumetric flow rate through sand layer Equation for one streamtube is
10
Q = - K dh
11
Now, calculate Total Flow, QT Specific Discharge, q Average flow velocity, Average residence time, t
QT = mQ = mKdh
12
13
1) low K deposit 2) symmetry (due to geometry of flow system) 3) flow lines (because flow cannot cross a flow line!) b) Constant Head boundary h = constant at all points on boundary so h/x along boundary = 0
14
Example at a dam
15
- at water table h = z and = 0 but q and h/x are unknown - h between equipotentials on WT must be constant h1 = h2 = h3 =
16
17
18
19
b) Think! Where does GW enter? Leave? All flowlines must enter and exit. c) Look for symmetry, which could reduce size of problem. Always choose the smallest possible region for flownet. d) Draw your first flowline(s) and then stop to see if they make sense (see b). Occasionally you might draw equipotentials first. e) Draw rest of flow lines and equipotentials obeying all rules and trying to get curvilinear squares. NOTE: it is much more important to observe all rules than to get nice squares.
if water table boundary, measure elevations usually draw boundaries and WT elevations in ink and do rest in pencil.
Cross-section
20
Htube = head drop between inlet and outlet of stream tube nhead = number of head drops or equipotentials along length of stream tube
Plan View
m = number of streamtubes T = Kb transmissivity b saturated thickness of unconfined aquifer, or b = thickness of confined aquifer
21
Flow Nets
The Laplace Equation is solved graphically using a flow net Example
Flow Nets
23
Flow Nets
Draw flow lines at equipotential boundaries and then a few flow lines through the system
Draw flow lines and equipotentials in between original lines and adjust
24
Flow Nets
25
Flow Nets
26
But W = L, so
Flow Nets
Total flow is
27
Flow Nets
28
Flow Nets
Compounding flow nets
Solutions of Laplaces Equation may be superimposed by adding the head values to evaluate the net response of a more complex system For example, consider the case of a well pumping water out of a uniform confined aquifer
29
Flow Nets
Compounding flow nets
30
Flow Nets
Compounding flow nets
Equipotentials for a well recharging at the same rate as the pumping well is If both well are located close to each other
31
Flow Nets
Compounding flow nets
Equiponetial values are added to get net effect Equiponetial values are added to get net effect
32
Flow Nets
Compounding flow nets
33
Flow Nets
Compounding flow nets
The fact that we can add groundwater flow solutions together to get a net effect allows us to deal with complex boundary effects, as we will see later in the course
34
FLOW NETS
Equipotentials are usually denoted by h1, h2, .... (or 1, 2, ...) and flow lines by 1, 2, ... Thus we can determine the discharge between two flow lines,
where
35
Q1 flow rate between 1 y 2 A area of the flow tube K hydraulic conductivity L1 distance between equipotentials 1, and 2
FLOW NETS
The rates of flow can be inferred from flow nets if the hydraulic conductivity is known. Often the total volumetric rate of flow (Q) can be calculated by:
36
where, Nf number of flow channels Ne number of equipotential drops H = Ne h total head drop Multiplied by the length of the section considered (for example, to determine Q under a dam, wed multiply by the length of the dam)
Rules:
1. Head drops between adjacent equipotentials
37
must be constant (or, in those rare cases where this is not desirable, clearly stated, just as in topographic contour maps)! 2. Equipotentials must match known boundary conditions. 3. Flow lines can never cross. 4. Refraction of flowlines must account for differences in hydraulic conductivity. 5. For isotropic media a. Flow lines must intersect equipotentials at right angles. b. The flow line-equipotential polygons should approach curvilinear squares, as shown
Polygons approach curvilinear squares
Rules:
6. The
quantity of flow between any two adjacent flow lines must be equal. 7. The quantity of flow between any two stream lines is always constant. Procedure (Harr, 1962, p.23): 1. Draw the boundaries of the flow region to scale so that all equipotential lines and lines that are drawn can be terminated on these boundaries.
38
Rules:
2. Sketch lightly three or four streamlines, keeping in mind that they are only a few of the infinite number of curves that must provide a smooth transition between the boundary streamlines. As an aid in spacing of these lines, it should be noted that the distance between adjacent streamlines increases in the direction of the larger radius of curvature. 3. Sketch the equipotential lines, bearing in mind that they must intersect all streamlines, including the boundary streamlines, at right angles and that the enclosed figures must be (curvilinear) squares.
39
Procedure
4. Adjust the locations of the streamlines and the quipotential lines to satisfy the requirements of step 3. This is a trail-and-error process with the amount of correction being dependent upon the position of the initial streamlines. The speed with which a successful flow net can be drawn is highly contingent on the experience and judgement of the individual. A beginner will find the suggestions in Casagrande (1940) to be of assistance. 5. As a final check on the accuracy of the flow net, draw the diagonals of the squares. These should also form smooth curves which intersect each other at right 40 angles.
Example 2
Flow net showing the topographic control of groundwater flow (Hubbert, 1940)
41
Example 3
42
Examples of flow nets: a) flow in the x-z plane of a pervious stratum underlying a dam b) flow toward a discharging well in the xy plane as influenced by a line source (constant head boundary). (Domenico & Schwartz)
Example 4
Some common errors include (a) equipotentials entering or exiting 43 a no-flow boundary, and (b) disappearing flow lines.
Example 5
Flow conditions in the vicinity of a lake demonstrating the effect of a high permeability layer at depth 44
Anisotropic Soils
45
Computing Pressures
46
Mathematics
Steady state: conservation of mass, continuity equation, stream function (), streamlines. Energy balance: Potential function (), total head lines, irrotational flow (flow around a curve by distortion rather than rotation).
47
Streamlines
Seepage velocity vector is tangent to streamline. Streamlines cannot cross. Commence on inflow boundary, finish at outflow boundary. Streamlines have constant stream function value.
48
Streamlines
stream function value dQ = 2 - 1
Stream tube
dQ
2
Seepage velocity
49
50
51
52
Flownet definition
Selection of streamlines and total head lines. Constant difference in stream function values between all streamlines. Constant difference in potential function values between all total head lines. Same difference value for both stream function and potential function.
53
Flownet definition2
Streamlines = 24 = 27 = 30
= 33
dq = 3
54
dq = 3
= 11 =5 =8
Flownet features
Orthogonal net. Inscribed circles. Diagonals to curvilinear squares form another orthogonal net. Ratio of number of streamtubes to number of potential drops is constant for a given flow boundary.
55
Flownet Examples
Basis:
56
For a given boundary geometry, only ONE true flownet exists. Find the true flow net by trial and correct.
Strategy
Identify the two streamlines and two total total head lines that define the problem. For 4 or 5 STREAM TUBES draw the 3 or 4 internal streamlines. Draw TOTAL HEAD lines to attempt to form curvilinear squares (a true flownet). Adjust the trial lines to get better curvilinear squares.
57
58
Cross-section to scale: 10 mm = 1 m.
W1
E F
A
59
Datum RL 60m
Flownet Interpretation
Discharge (m3/sec/m) Pore pressure Hydraulic gradient.
60
k - coefficient of permeability. HL - drop in Total Head across flow. N - number of stream tubes. N - number of potential drops. Note N might not be an integer.
N = 3 N = 8
0 0 1 1
8 7
K = 7 x 10-5 m/s
2 3 4 3 5
2
62
63
u = (H - z) * w
B
H=91.5m
65
dH = 0.5m
RL95m
H=91.5m H=94.5m
dL = 4.5m
C
H=92m H=94m H=93m 66 Total head datum = RL 0m H=93.5m H=92.5m RL72m
Liquifaction of soil may occur if the exit hydraulic gradient exceeds a critical value such that effective stress equals zero.
Icrit = (Gs - 1)/(e +1) Gs - Specific gravity of soil particles e - Void Ratio Icrit ~ 1.0
67
Numerical Solutions
Finite element grid for numerical solutions of seepage below a dam (from SEEPW) and typical output from a finite element seepage analysis (from SEEPW)
68
70
Seepw solution
60
50
40
A
5.8120e-003
0 10 20 30 40 50 60 70 80
30
20
10
Other typical output from a finite element seepage analysis (from SEEPW) 69
Feet
70
71
based on numerical approximation methods (usually finite element or finite difference) to solve the appropriate groundwater flow equation requires that flow system be represented as a boundary value problem a variety of models available well-suited for heterogeneous or anisotropic systems (that are difficult to solve with flow nets) and for irregularlyshaped systems some models allow for transient flow plan view or cross-section models available usually 2D, but some 3D models available the accuracy of predictions based on a numerical model is usually limited by the amount and reliability of the hydrogeological data, particularly K, and by the boundary conditions
72
Define hydraulic boundary conditions Specified head conditions (Dirichlet) constant head boundary water table Specified flux conditions (Neumann) q = 0 at impermeable or symmetry boundary q = f(x,y,z) known flux, can represent recharge at water table, well, river, etc. Assign material properties (K,T,S,Sy) for different layers or regions. Can be anisotropic Specify initial head conditions for transient flow FD models are relatively simple and widely used
73
MODFLOW available from US Geological Survey versatile (2D, 3D, transient, etc.) inexpensive (free) difficult to use, poor graphics Visual MODFLOW proprietary version of MODFLOW easier to use, expanded graphics cost $1000 /copy FLOWPATH 2D plan view, steady-state easy to use, but less versatile cost $700 FLONET/FLOTRAN 2D cross-section, steady-state easy to use, but less versatile cost $700 Numerous others
74
75