SWIFT I SELF-TEACHING CURRICULUM:
ILLUSTRATIVE PROBLEMS FOR TE SANDIA WASTE-ISOLATION FLOW AND
TRANSPORT MODEL FOR FRACTURED MED EA
Hark Reeves*
David S. Ward*
DRFT - INFOWMAL AN PRELIA.NARY AN" .AS SUCh*AY OM RRMS OT YET COr RED. FOR#1SE PRIME STRBUON MD OT FORUORA RELaSE WIPOU CONSEN OF AUJTMOS
March 1984
Sandia National Laboratories
Albuquerque, New Mexico 87185
Operated by
Sandia Corporation
For the
U. S. Departmenr of Energy
Prepared for
Division of Waste Management
Office of Nuclear Material, Safety and Safeguards
U. S. Nuclear Regulatory Commission
Washington, D.C. 20555
Under Memorandum of Understanding DOE 40-550-75
NRC FIN No. A1166
8409120064 840228PDR WRES EXBANLA-1 166 PDR
. .
* GeoTrans, Inc.
DAFT INFOAMAL AND riL;., .. * ,, ; LY CONTAIN ERRo NOT YET CORRWCED, FOR
314t0 W PRIVATE DMBM OND NOT fOR ABSTRACTEX9LMREWWWI USOFAOMSE
!vveral documents have been written describing SWIFT II, the most urrent
version of the SWIFT Model. One, Reeves et al 1984a], describes the theory
and implementation, and another, Reeves et 81 [1984b), describes the data
input. Two others, Ward et al 11984a) and Ward et aL 11984bW deal with code
assessment through verification and field application and through
benchmarking, respectively. This document, however, has an entirely different
function. It, like the work of Finley and Reeves 19811 for an earlier
version, is devoted to assisting the analyst who desires to use the SWIFTCode. This code is quite general in terms of both the pocesses (flow, heat,
brine and radionuclide transport) and the media (single and/or dual porosity,
confined or unconfined) which it considers. Consequently a document such as
this one is necersary. Nine examples are presented to illustrate the use ofSWIFT II in doubly porous and unconfined aquifers. Each problem and its
numerical solution are described. Then several exercises are presented inorder to assist the reader with the input and the output and, occasionally,
with the interpretation of the results. For each case a printed input listin
and a microfiche output listing .e provided.
ii
tit
ACKNOWLEDGEMENTS
The authors would Like to acknowledge the work of 'Ka. Lynette S. Knippa
for her typing skills and for her determination to meet rther stringent
deadline. This document would not have been ossible without her help.
IV
v
TABLE OF CONTENTS
Section Page
1 INTRODUCTION .................... *e, .. q1-1
2 SOLUTE TRANSPORT THROUGH DUAL-POROSITY EDIA...................
2.1 PROBLEM 1. TRANSPORT OF A DECAYING RADIONUCLIDE IN A
FRACTURED POROUS MEDIA [TANG F.T AL, 1981).................
2.2 PROBLEM 2. TRANSPORT OF A DECAYING RADIONUCLIDE IN A
FRACTURED POROUS MEDIA HUYAKORN* 19831...................
2.3 PROBLEM 3. TRANSPORT OF A RADTONUCLIDE CHAIN IN A
FRACTURED POROUS MEDIA INTRACOIN, 19831...............*e
2-1
2-3
2-23
2-3S
3 iLUID FLOW THROUGH DUAL-POROSITY MEDIA ......................... ............, 3-l
3.1 PROBLEM 4. ANALYSIS OF WELL-TEST DATA FOR A DOLOMITE
FORMATION PAHWA AND BAXLEY, 19801....... 3-3
3.2 PROBLEM 5. ANALYSIS OF WELL-TEST DATA FOR THE MUSQUODOBOIT
AREA, NOVA SCOTIA (PINDER AND BREDEHOEFT, 19681........... 3-23
4 FLOW AND TRANSPORT THROUGH AN AQUIFER WITH CONFINING LAYERS.... 4-L
4.1 PROBLEM 6. DRAWDOWN FROM A FULLY PENETRATING WELL IN
A LEAKY AQUIFER KANTUSH, 19601 .................... vote*** 4-3
4.2 PROBLEM 7. HEAT TRANSPORT DURING FLUID INJECTION
[AYDONIN, 19641 ........ .............. 4-17
5 FLOW WITH A FREE-WATER SURFACE ........ ..................... 5-1
5.1 PROBLEM S. THE DUPUIT-FORCHHEIHER STEADY-STATE
PROBLEM [BEAR, 197Z! ...... . .... ... . .*.*. . *........ .. 5-3
5.2 PROBLEM 9. THE OUSSINESQ TANSIENT-STATE PROBLEM
(BEAR, 19721 . .... ........................ 5-15
NOTATION
REFERENCES
APPENDIX
................. .....................................
.................................. ...
CONVERSION OF INPUT DATA FROM SWIFT TO SWIFT II ..............................
6-1
7-1
8-I
VI
vii
LIST OF FIGURES
Figure Page
2.1-1 Problem . Schematic Diagram...... ..................... 2-4
2.1-2 Gridding of the Fracture ................................... . 2-9
2.1-3 Gridding of the Rock Matrix ........ ........ .. ....... . 2-10
2.1-4 Listing of the SWIFT II Input Data ............................. 2-11
2.1-5 Radionuclide Concentrations Within the Fracture for a
Prismatic Characterization of the Rock Matrix.................. 2-18
2.1-6 Radionuclide Concentrations Within the Rock Matrix for a
Prismatic Character.zation of the Rock Matrix .................. 2-19
2.2-1 Problem 2. Listing of SWIFT U Input Data.........*.......... 2-25
2.2-2 Radionuclide Concentrations Within the Fracture for a
Spherical Characterization of the Rock Matrix.... .............. 2-31
2.2-3 Radionuclide Concentrations Within the Rock Matrix for a
Spherical Characterization of the Rock Matrix .................. 2-32
2.3-1 Problem 3. Schematic Characterization of Transport Within the
Fracture and Diffusion Within the Matrix......... ............. 2-36
2.3-2 Listing of SWIFT II Input Data ................................. 2-41
2.3-3 Gridding of the Fracture/Matrix System......................... 2-50
3.1-1 Problem 4. Schematic Diagram of Well-Test HRA Within the
Magenta Dolomite Formation ........................ * ** *** .... *.... 3-4
3.1-2 Conceptual Models for the Secondary Fracture Porosity .......... 3-5
3.1-3 Listing of the SWIFT II Input Data . ..................... 3-11
3.1-4 Flow Rate and Drawdown of the H2A Slug Test Within the
Magenta Dolomite Formation* ......... a.... 69e.444*4. ... . s 3-18
3.2-1 Problem 5. Location Map of the Musquodoboit River Basin ....... 3-24
3.2-2 Geologic Map of usquodoboit arbour Area, Nova Scotia ......... 3-25
3.2-3 Geologic Cross Section, Hurquodoboit Harbour Area,
Nova Scotia . 4.. . 4. .... . .. . .. *. ... ... 3-26
vifl
LIST OF FGUMES
(Continued)
Figure . X
3.2-4 Location of Observation Wells and Characterization of
Hydraulic Properties... .... . ...... .. ....... * ... ...... 3-27
3.2-5 Observed and Simulated Drawdowns for Wells 1, 2 and 3 Using
a Homogeneous Dual-Forosity ModeL .............................. 3-28
3.2-6 Time-Drawdown Curves Obtained From the Simulation of Finder
and redehoeft 119681 ................... ........... *. ......... 3-29
3.2-7 Listing for the SWIFT 11 Input D.. 3-32
3.2-8 Simulation of the Observed rawdowns for Well I Using a
Homogeneous Dual-Porosity Model ....... ........ ... ............. 3-36
4.1-1 Problem 6. Schematic Diagram of a Fully Penetrating Constant-
Discharge Well in a Leaky Aquifer .. , . , . 4-4
4.1-2 Numerical Characterization of the Leaky-Aquifer System ......... 4-8
4.1-3 Listing of SWIFT It Input Data ............... ..... 4-9
4.1-4 Graphical Comparison of the Numerical Results from SWIFT and
the Analytical Solutions of Hantush for a Radial Distance
of 117.4 m......................................... 4-14
4.2-1 Problem 7. Radial Heat Transport Within an Aquifer with
Losses to the Over/Underburden .......... 4-18
4.2-2 Listing of SWIFT 1I Input Data................................ 4-22
4.2-3 Temperature Breakthrough Within the Aquifer at 3.5 from
the Injection Welle........... 4-27
5.1-1 Problem 8. Schematic Diagram for the upuit-Forchheimer
Prob5em ........... . . . . . . . . . . . . . .... . . . . . . . . . . .a . . a . . . 5-45.1-2 Listing of SWIFT I Input Data ......... 6.0....Go........ ...... 5-7
Ix
LIST OF FIGURES
(Continued)
Figure Page
5.1-3 Steady-State Free-Water Surface for the Dupuit-Forchheiser
Problem.,.,.*,.6 ... ......................................... 5-LI
5.1-4 Geometrical Characterization of the Gridding, Boundary
Conditions (Labelled "ALF") and the Recharge. .t............ 5-12
5.2-1 Problem 9. Schematic Diagram of the Boussineaq Problem...... . 5-16
5.2-2 Transient-State Free-Water Surface for the Bouasinesq
Problem........ .. .... * *....... a@*ee.0. * .... .* .......... 5-18
5.2-3 Listing of SWIFT It Input Data ................... .. 5-20
5.2-4 Geometrical Characterization of the Gridding and Boundary
Conditions (Labelled "ALF") in Relation to the Numerical
Solution..... ............. .. *. ................. *........ 5-26
x
.
xl
LIST OF TABLES
Table __
2.1-1 Problem . Input Specifications ............................... 2-(
2.3-1 Problem 3. Input Specifications ............................... 2-3'
2.3-2 Nuctide Inventory (12) nd Matrix Retardation ( 3) ............. 2-4(
2.3-3 Breakthrough Parameters.......... .................... ........ 2-4
2.3-4 Breakthrough Profile ............... ...... *..... .... 2-41
2.3-5 Material-Balance Summaries .......... ........................... 2-5
3.1-1 Problem 4. Input Specifications.............................. 3-1
3.1-2 Observed DecLne in Water-LeveL Height and Bottom-Role
Pressure........ ... *S ... 3-....3
3.2-1 Problem 5. Input Specifications........................ ....... 3-3
4.1-1 Problem 6. Input Specifications .......................... ... 4-
4.2-1 Problem 7. Input Specifications.............................. 4-2
5.1-1 Problem 8. Input Specificationa.............................,. 5-
5.2-1 Problem 9. Input Specifications............. ...... 5-1
xii
Wf - INFORMA AND PRELIMNARY MD AS SOANY o TAIN ERRS NOT YET MECED. FOlm4gou PfiVTE DItRBUO MD Not Vol
INT&ODU mP± rElfMASE W 1liu OSENT OF ALITMO.
The SWIFT Model (Sandia aste-Isolation Flow and Transport Model) has
been developed for the evaluation of repository-site performance. It i a
fully-coupled, transient, three-dimensional model, and it is implemented by a
finite-difference code, which solves the equations for flow, heat, brine and
radionuclide transport in geologic media. Having evolved from the U.S.
Geological Survey Code, SWIP (Survey Waste Injection Program (INTERCOHP,
19761), this code has experienced continuous improvements since 1977 Dillon
et al, 1978, and Reeves and Cranwell, 9811. SWIFT I marks a significant
extension of the development in that the model can now treat three additional
types of media in a cost-effective manner. Two are confined dual-porosity
systems and include a fractured porous material and an aquifer with conductive
confining beds. The other is an unconfined aquifer with a free-water
surface. As such, the model has become a very comprehensive and effective
tool for evaluating the processes envisioned for a repository of high-level
nuclear wastes. Furthermore, its applications extend beyond nuclear-waste
isolation to, for example, aquifer thermal-energy storage, liquid-waste
disposal by deep-well injertion and migration of contaminants from surface
disposal sites.
At the same time, however, the use of SWIFT does make rather heavy
demands on the analyst. He must have a familiarity not onLy with the basic
science but also with the mathematical model*, the numerical model*, the
code*, its input and its output. Current documentation (Reeves et al,
1984a,b) is available covering each of these items, evertheless, probably
the esist way to earn the SWIFT Model, or any other model, is through a
meaningful set of worked problems. Such is the basic premise for both this
report and its precursor, SWIFT Self-Teaching Curriculum Finley and Reeves,
19811. As a matter of fact, that earlier report is appropriate for SWIFT
II. A total of 11 problems are presented there in the input format used by
SWIFT, Release 4.81. However, with only minor changes, as described in the
* See Silling 119831 for a definition of these terms.
1-2
Appendix of this report, the input data sets listed there may be used to
illustrate the application of SWIFT II, Release 12.83, to single-porosity
media.
In this document 9 problems are presented. Fracture/matrix problems are
presented first, followed by aquifer/confining-bed and uncvnfined-aquifer
problems. In each case the discussion proceeds from a description of the
problem to a description of the numerical simulation. For convenience, a
definition of the symbols is reserved for Section 6. Further, the oft-
referenced documents of Reeves et a 11984a.bl are referenced simply as the
Theory and Implementation and the Data-Input Guide, respectively. Several
exercises are placed at the end of the discussion for each problem, which are
directed at two different levels of expertise. The more basic Level I
familiarizes the reader with the input, which is printed, and the output,
which appears in microfiche affixed to the back cover. For these exercises
the only necessary resources, other than this document itself, are the Data-
Input Guide and the Theory and Implementation. The more advanced Level 2
tests the reader's ability to extend or interpret the basic results. Here a
computer and the SWIFT I Code may also be necessary.
2-1
2 SOLUTE TRANSPORT THROUGH DUAL-POPOSITY MEDIA
2-2
2-3
2.1 PROBLEM 1. TRANSPORT OF A DECAYING RADIONUCLIDE IN A FRACTURED POROUS
MEDIA [TANG ET AL, 19811.
2.1.1 Objective
* To ilustrate fracture/matrix transport with a prismatic
characterization of the rock matrix.
2.1.2 Description of the Problem
Problem Statement. A thin rigid fracture is situated within a saturated
porous rock matrix as shown in Figure 2.1-1. Both fracture and matrix are
semi-infinite in their extent. Radionuclides, which derive from a source of
constant strength, are convected and dispersed through the fracture by a
constant velocity field and are diffused into the rock matrix.
Transport Equations. Mathematically the transport within the fractures
is governed by the equations
- a3X (ecu) (a D C) r - KOpC * (KPC) (2.-L)
Transport within the rock matrix is assumed to occur in a direction
perpendicular to the fracture and is governed by the equation
a (PD' -+ * 'p'C' a (K'*pOC') (2.1-2)
Coupling then arises through the flux at the fracture/matrix interface:
r .- *o'DI Tc (s,0) (2.1-3)
2-4
C=1.0SOURCE
I .
z
Iv
I
IXU-
8
ccK-
Figure 2.1-1. Problem 1. Schemacic Dlagram.
2-5
The composite parameters are defined here in terms of the basic parameter
set. They are the dispersion/diffusion of the fracture:
0 a Lu X (2.1-4a)
the diffusivity of the fractures:
D a 40* (2. t-4b)
and the diffusivity of the rock matrix:
D' a Do 'D* (2.L-4c)
The remainder of the parameters are defined in Section 6.
Initial/Boundary Conditions. Within the rock matrix the initial
concentration is zero:
C'(X,,t-0) -20 P s > x > 0 (2. L-5a)
and fracture and matrix concentrations are identical at the interface-
C'(X,8=0,t) C(x't) , t > (2.1-5b)
Within the fracture, the initial concentration is also zero:
C(xtw0) - 0 x > (2.1-6a)
and the boundary concentration is unity:
C(x-0,t) t > (2. 1-6b)
Input/Output Specifications. The input data for this problem, given in
Table 2.1-1, is the same as that prescribed by Tang et al for their lcw-
velocity case (p. 561 f). The desired output consists of two sets of spatial
2-6
Table 2.1-1. Problem 1. Input Specifications.
Parameter Symbol VaLue
Fracture Width 2d 10-4 m
Matrix Porosity 0.01
Matrix Tortuosity 0.1
Fracture Dispersivity CL 0.5 m
Molecular Diffusion in Water D* 1.6x1O-5 2
Kalf Life r 12.35 y
Decay Constant 0.056L y1
Matrix Retardation K' 1.0
Fracture Retardation K 1.0
Fracture Velocity v 0.0. M/d
Fracture Porosity 1.0
2-7
distributions. Curves of concentration versus distance within the fractureare called for at the values of time, t 100, 1000 and 10,000 d. Then asingle curve of concentration versus distance within the rock matrix is calledfor at the time t 10,000'd and the position x - 1.5 m.
2.1.3 Numerical Simulation
Discussion of Code Input. Two items are of interest here. Specifically
they are: (1) representing a semi-infinite domain with a finite domain and
(2) specifing the flow and diffusion within the systeot. For the first ofthese, the steady-state solution of Tang et al is most helpful. This solutionprovides one characteristic length for penetration into the fracture and
another for penetration into the rock matrix:
L 1 {(1l2)xL - ((1/4)aL2 + /D21- (2.1-7)and
a " (DtX)/2 (2.1-8)
where
D y + D* (2.1-9a)
D'- D* (2.1-9b)
* - x~ * *t (D'\))2/ (2.1-9c)
With values taken from Table 2.1-1, these expressions yield
L = 1.4 m (2.1-10)and
a - 0.30 m (2.1-11)
.
2-8
The domains of simulation are taken to be somewhat arger than these
characteristic lengths, i.e.#
L : 10.24 m (2.1-12)
for the fracture (Figure 2.1-2) and
a a 1.2 (2.1-13)
for the rock matrix (Figure 2.1-3a).
Spatial increments in the fracture and matrix domains are then taken to
be significantly smaller than he characteristic lengths. Nevertheless, some
numerical experimentation is necessary for gridding both the spatial and the
temporal domains. There is, however, one consistency check:
As (2D'At)/2 (2.1-14)
for the matrix, which is a useful guide. Figures 2.1-2 and 2.1-3 show
pictorially the gridding adopted for the two spatial regimes. Figure 2.1-4,
then exhibits this information as it i *pecified in the data input. As
shown, the globally connected mesh of SWIFT It is used to simulate the
composite fracture/matrx system with the individual increments specified in
Cards R-l (x), R-18 (Ay) and R1-19 (z). The one-dimensional mesh,
locally connected to only one global grid block, is then used tc simulate the
rock matrix, which is here mbedded within the global mesh. ridding, in this
case, is specified by the three parameter values provided in Cards ROD- n (i)
and RD-2-2 (a and As). Given the number of increments n and the
increment As at the fracture/matrix interface, the code generates the local
mesh automatically.
Continuing on to the second item, the flow within the global system is
maintained using injection and production wells in the end blocks of the
global system, he common rate being
q - 2dyv - 1.157 x 10 1m 3S , R2-6 (2.1-15)
f5 10'< .512 1.02
*12.05 3.07 9.22 10.24
I i / / I I2.4
I I / I I I0 2.5 x1O" 7.5 104' .768 1.54 2.56 9.73
I
Figure 21-2. Cridding of che raccure.
2-10
x
1.0
(a) Prismatic Characteri:ecion.
'0.01
I - O d
He - , - - - I
71.2
I 4Ii-I - I S � * Y
A1
I2d=Ix 10
21.0
.
(b) Spherical Characterization.
Figure 2.1-3. Cridding of the ock Matrix.
2-11 DRAFT
l Z 3 4 5 I a
Iz345&70I tz3%sb4 aI345s67Nqz2L356q 012z3S67I9o LZ34QS 7I312 456749q01.3461S09,*._ _ _ _ _ _. 9 e e S * O._ _ _ _ __e_ _ _ _ ,__- -----
II PROPELEM NO. I TRE-COK.PONENT UCLICE tRANSPORT IN
Z h11193 CPARISN WITM AALYTICAL MOL - RISMATIC
FPRCTUREO E01.
REPRESENTATlON
51-1-I
5-1-2I ...*
0 a 0 0 0#(21* 03
.4
5
6
.7
a
9
10
LI
12
1 3
14
1*411* 0
I 21 1
* .9* i* 0.43).
ZI1ORA0 TRACII 00II 0.0I
I.* 1.0
1 914) *
I 12
1 2 '1
II
I 2
I 0
I
9..44
a z o z
0 0 12.35
w**#. REf* 6** U REf
9* .
1 1 1 1 1.15)99*4
L 0 1 0 0 Ce-3-1
A-3-2
RG-1
510-2-1.Ro-2-L
:£RENCEO t4 IEXT . Ro-2-1.EtfMCEO VN ExE51(!SE
Ro-2-2
ROD-I
100-3
1.0 RI-1
0*50 0.0b.6671E-I4R1-2
RI-25
RI-Z.5
Rl-2.5
Rl-2.S
1000.0 Rl-3
R1-6
Rl-7
R1-I
Rlm9
iS * 0.0I
16 1 0.0I
17 1 0.0
fLe 1 e1 To*+***&
i9 I 0.01
20 * *.0001
2t I 2700.0I
22 1 ? 2I
23 1 20.0
24 1 0.20
25 * 100.0
0.0
0.0
0.0
0.0
0.0
0.0
1.013ECS
2 2
0.001
0.00L
0.00
0.0
0.0
0.0
4116.0
0.0
0.0
20.0 1000.0
20.0
0.50
200.0
0.001
oa o;.
0.001I
*------- .*------------------*- --
Fiere 2.1-4. Liting of the SWIFT I nput Data.
2-12 DRAFT
1 2 3 4 S 0 7 a2Iz4%6s qqot I4soefiqoIZ3 #t6 yo I 2 34SfTgqot214S6llotZ3 4 12 0789O1Z3D7S6T9o901Z467qo
-…- ------- *------*---*----* - ---------
26
27
29
30
I
33
34
3S
3b
37
hi
39
40
41
42
1 100.0a
1 20.0: a
I* 20.0
I 2'5.OL-4
I .E-z
£ 1.0
lS 2** ..
I .
I
1 1.0
I1 0.06-21-I +++
o.o
z0.0
Z00.0 0.001 RI-b4
AidS1
At-ILto.0
RI-tU
I
C
IIIII
I
IIIIII
t. 01 HE GS
1.Cf-I
0.L 8
C.0
2.fCE-
0.256
0.0
4.OE-3
0.512
* SS**l**
J.4052E-? 3.40t2E-7 4.167E-5
l.oc-3
9*1.024
0.0
0.0
I . 2 3.2E-Z
I t
1.0
0.0
t 1
1.0
0.0
I 0
0.0 O.ORI-to
Q1-21
R1-26-I
0.01.9196SE09RI-2-Z
R-Z6-3
4 L-Z6-BLN
QL-Z7
RID-l4* 4*90*9*
43 4 ( 7 * 0.n tI).
44 £ 1 1 1.2 0.0tI *9* **058
I.SE-15
0.0
0.0 l.0
45 +
t46 1I
49 II
So0
1.OR1-2-1
Rio-Z-2
R 10-3-ALN
R 10-4
Rl0-S-4LN
0.0
III9.
IIIIIII
0 0
I L
I 1-1
I I I I 1.0 1-4 1
1-4-&LN 4.
* … ---- . - --- - - - - - - 4 * - - - * ---- - - ------.--….
.4
Figure 2.1-4 Continued.
2-13 bRAse
I 2 1 4 S O 7 312,S61490 tZ145 7090:15
,-- ----- ._______ _ -- ------ . _
ISl 1
IS2 I
53 I1
54 1
55 *
56 1I
57 II
Sd t
I
1
6? 1631
t)I
6 4 1I
65 *
2 21 1 t t I 1
I l I I I I t
P*...
2
If.
24
0.0
I I
0.50o
0 a 0 0 0 a
-L I .5S7f- t
'I. 7E-11I ee**eO6eo
RLA-t
tlA-I-oLN
RZA-2
RZ-1
RZ-iR2-,
R-6-L
K2-64LN
K2-?-I
R2-7-1
RZ-T-
RZ-t-tU
66
67
68
TO
7Z?I
t3
'4
is
IiI
I
II
If
IIIIII
I I
1.0
2 21
1.0
1.0
I I
0 0
1
& I
0.0
I I
0.0
I 1
0.0
1.0
I I
0 0
I I
2000
I I
20.O
0.0
0.0
I 1
t -C
000
0
0 0000
0 1
1 I
0.0-7. t21ZE-5
0 13 1 000 12R2-13
Q2-l
R2-9
RZ-10-1
RZ-10-2
RZ-uO-SL
K2-1.S03
).46E0& 1.456Eo
I -I -1 -1 -1 -L 000 0 0000 0 LO -L 000 IOR2-13
,_______ ~ ~ -- -- -__ __ _ _ -- __ _ _ _ _ _ - - - - - - _- -_-_-
Figure 2.1-4 Continued.
2-14 £WAcr
I 2 3 I S 6 7 U1234S6 740 234Sb?%4G12 1SM9012 2456749012306764012145S6760tt34S6TRIOLZ1456t*90
,_______.______ __,_.__.___.__ --- _ _ _ , _ _ _. _ - ,_ _ _ _
K16K1 a a a 0 0 0 a
77 t *.660 S.eB"EO6
te I I -I -I -t I -1 000
J9 a 0 0 0 a a 0
BO * 1.5Z¶Eo0 6.91cofi
at I t -L -I -I -I _1 Goo
42 t O a C a a a 0
*13t Z.5U;'EC? q.?TTLEO
64 1 * -t -L -t -L -1 000
IS K 0 0 0 0 0 a
el I -- - -I -1 -1 000
I
e I o o 0 o o a a
K
59 S4938fal toqs4fOl
9Qo t -I -I -1 -I -I OoO
S Il c o a a a a o
52 I 3.BA.EO7 2.802EOT
671 -1 -I -z t -i 000
6 1 0 0 0 0 0 0 0
SS*1.25IXO 3.906EO?
go 1 l-SO)BO0 s.s2207
qq I I -1 -i -L -t -t 000c
Lee 0 a a 0 a 0 0
0 a RZ-t
O 0000
o 0
0 Good
a 0
0 oo
o 0
0 000
0 a
0 0000
o o
o oooo
0 0
o oooo
0 0
o oaoo
o o
0 10
a
I oo OR2-13
tz- I
O 10 -L 000 102-13
0 t2-1
0 t0 -1 oo 10t2-13
0 a R2-L
O 10 -L 000 10R2-13
0 2t-t
0 10 -1 000 IOR-13
O RZ-I
o to
a
I cOo 102-t3
RZ-l
O tO -1 000 IOR2-LI
0 Rt-l
0 to -t 000 102-13
O Rt-&
._ ____._ _ ______ ______ _.__ _ ____ _ _~ _ _ __* _ e e~ a e ~ f ** - - - - - .
Figure 2.1-4 Continued.
0
=
2-15 D
I 2 3 4 S & I a
jZ3kSE?qotZ567flt3LZ 34567sq012345670qGt2lt:56789012)ST$901Z3456?0lI23456I q0._________,_ _e _* _ _ _ _ _____ .___e_ _, ___. _ ____. _ _ ._________.
I101 I 2*544CoO o 7.tEOT
102 1 1 -I -I -I -t -1 00
La3 I o a e a a o O
104 j 3.*4E06 4.4CO7
LOS I -1 -I -1 -I -I 000
106 1 0 0 0 0 0 0 0
107 1 .4*312EO 6.Ea4Eo
I
ts I -1 -L -1 -1 -1 000
I
1q I C 0 0 0 0 0 0
1
110O S.170EOe 0.007
111t I t -I -I -I -1 -1 ooa
112 I C 0 O a O a
11 I 6.O0#OE0$ e.6E07
11) 1 1 -1 -1 -l -1 -1 000
llS 0 0 0 0 0 0 0
116 t S.976EO a-.b4ECT
LI1 I I -L -t Mt -1 -1 000
lie I 0 0 0 0 0 0 0I
112 1 7.060E08 S*6tE07
120 I1 -1 -1 -I -1 -1 000
11 a a 0 0 0 0
I
12 K 6.60'.OtI .7307
I1 I I -L -1 - I -z 000
121 a 0 0 0 a 0 0I
1S *. 72(E0 1..2SE06
0 COO
a 0
0 0000
0 0
O 000
o a
0 0000
a a
0 0000
0 0
o 0c0c
0 0
o oooo
a 0
Q 0000
o 0
O 10 -1 000 IOR2-13
O 10 -1 000 IORZ-L3
0 12-1
0 10 -1 000 1ORZ-13
0 t-
0 10 -I 000 loiZ-13
O R2-1
O LO -I 000 lORZ-13
O RZ-1.
0 10 -1 000 tORZ-13
0 az-I
O 10 -1 000 IOR2-13
0 R2-1
0 10 -1 00 lORZ-130 12-1
IIIIIIIII
I
III
II
IIIIIIIII
IIIII£
I
IIIIIIIII
I* …*___e____ - - - - - - - -*- - - --- - - -
Figure 2.1-4 Continued.
2-16
1 2 3 4 a 7 i 124^80Z45t9135? 72**00t357~l35970,2 )1I78qOI23S 7590
I 12Z l I -I -1 -1 I -I 000 a 0000 0 10 1 000 tORZ-13 I
12 7 I O o O 1 t2-t-STOP II .* *__ _-. _ _ * e . *..u_...____.___ __.. _*
Figure 2.1-4 Continued.
2-17
Since both fracture and matrix are included within the global blocks (IFD 0
in Card ROD-3), the global porosity is specified by
* = 4/(ad) a 4.167 x 10 5 , RlO (2.1-16)
This ylds, then, from Equation (2.1-4a), the value
D * 6.6672 x 1014 2I , R1-2 (2.1-17a)
for the diffusion of the global system. The corresponding relation for the
rock matrix is, from Equation (2.1-4b):
D' - 1.6 x 1012 M In (2.L-17b)m
Results. Figures 2.1-5 and 2.L-6 show the results obtained from the
SWIFT Code plotted alongside the analytic results of Tang et a. Both within
the fracture and within the matrix, the two show reasonable agreement.
Discussion of the Code Output. The numerical results exhibited in these
figures are presented in much more detail by the microfiche listing (inside
the back cover). At this point, the r 4 er is encouraged to scan the listing
for a general familarity, but with only one specific purpose, i.e., to locate
the data used in these figures. The tables which are denoted there by the
words "dual porosity" or "rock matrix" refer, in this case, to the rock matria
of a fracture/matrix problem. In Problems 6 and 7, however, these tables
acquire a new meaning in that they there refer to the confining beds of an
aquifer.
in order to examine the concentration distributions, the reader should
note specifically three tables. The first of these tables is entitled X-
Direction Distance to Grid-Block Center". It provides distances along the
fracture to the centers of the grid blocks, As shown, the distance x - 1.5
lies in Global Block 13. The second, "Dual-Porosity Block Numbers," then
identifies, with unique number, the local one-dimensional rock matrix units
which are imbedded in the global blocks. Note that such a unit is not
imbedded in Block 1. This block, as discussed under Problem 2, is used only
for the purpose of establishing a constant-concentracion boundary within the
fracture. Note also that Local Unit 12 is associated with Global Block 13 r
2-18
- ANALYTICAL
0.8 o SWIFT
Z 1 0 0 ~~000 d
< ~100 2.d. s *
z .40
0.2-
1.0 2.0 3.0 4.0 5.0DISTANCE DOWN FRACTURE ()
Figure 2.1-5. Radionuclide Concentrations Within the Fracture for a
Prismatic Characterization of the Rock Matrix.
- ANALYTICAL
0 SWIFT
ZiO A\ ~~~~~~~~~~~~t 10,000 d -:
i- .2 1OO
0 .2 A ,6 1.0 1.2
DISTANCE INTO MATRIX ()
Figure 2.1-6. Radionuclide Concentrations Uithin the Rock Matrix for a
Prismatic Characterization of the Rock Hatrix.
2-20
hence, with the distance x 1.5 n measured within the fracture. The third
table "Specific atrix Discretiration Parameters" gives nodal locations within
the rock-matrix units. Note that Node 12 is located at the fracturelmatrix
interface. Thus, for example, Node 4 is located at a distance of about 04
from that interface.
Tables of fracture concentrations, labelled "Component-1 Concentration,"
and matrix concentrations, labelled Component-I Radionuclide Concentration
Within the Rock Matrix," appear at selected time intervals, with the Last such
tables corresponding to 0,000 d. The calculated fracture concentration C a
0.318 for Global Block 13 x 1.5 ) does agree with Figure 2.15. Also, the
matrix concentration C' 0.068 for Local Unit 12 and Local BLock 4 (a 0.4 m
from interface) is consistent with Figure 2.1-6. Other values may be checked,
as desired.
2.1.4 Exercises
Level 1. Using the Input-Data Guide, the Theory and Implementation and
the microfiche listing of the output, as required, complete the following
exercises:
Exercise 1: From Table 2.1-1 and Equations (2.1-16) and (2.1-17), what data
are required for the blanks labelled "" through "5" in the
input-data set?
The control parameters called for here should specify steady-
state flow, the international SI system of units, transient-
state radionuclide transport and no convection within the rock
matrix (the local system). In this case convection within the
matrix is of negligible importance, and, for computational
efficiency, the proper control parameter should be set to
indicate "n.. convection".
Lxercise 2: From the physical properties specified in able 2.1-2, what are
the data required for the blanks latelled "6" through "8" in the
input-data set?
2-21
Exercise 3 Check the answers for Exercises and 2 by referring to the
output listing, specifically to the echo of the input data which
appears at the beginning.
Level 2. The reader who wants to develop a deeper understanding of the
code should also consider the following exercise*:
Exercise 4: A basic assumption of the foregoing analyses, both analytical
and numerical, is that diffusion in a direction parallel to the
fracture may be neglected. Evaluate this assumption for the
problem of Tang et a.
In most applications where fracture transport (convection and
dispersion) dominates over matrix diffusion, it is safe to say
that this assumption is valid. In some applications, however,
it may be necessary to check this assumption. One approach is
simultaneous discretization of both fracture and matrix, hich
involves the gridding of a two dimensional plane. Such a
simulation has been performed by one of the authors with the
basic assumption found to be valid.
* For the reader with no previous exposure to the SWIFT Code this exercise
will Likely be too difficult at this point and should be postponed.
2-22
2-23
2.2 PROBLEM 2. TRANSPORT OF A DECAYING RADIONUCLIDE IN A FRACTURED POROUS
MEDIA IHUYAKORN, 19831
2.2.1 Objective
* To illustrate fracture/matrix transport with a spherical
characterization of the rock matrix.
2.2.2 Description of the Problem
This problem is identical to Problem I with only two exceptions.
Firstly, it is assumed that the exposed surface area between fracture and
satrix, per unit fracture length, is greater here and may be more
realistically approximated by a spherical surface (Figure 2.1-3b) for the rock
matrix. Thus, Equation (2.1-2) of the previous section is replaced by
I a (sp'D' aC + r - K'$0c' C a (OpICI) (2.2-1)s; s m 3 a
The radius for the spherical units is chosen to be identical to that used in
Problem 1 for the length of the prismatic units:
a - 1.2 m (2.2-2)
Secondly, the matrix diffusion is reduced by a factor of approximately
three relative to that of Equation (2.1-17):
D' a 5.787 x 10 3 m2s (2.2-3)m
The desired output, as in the previous problem, consists of two sets of
spatial distributions. Rowever, the values of time are changed to t 441,
3619 and 90,615 d for the distributions within the fracture and the rock
matrix. The position for the latter is also canged to x - 1.0 m.
2-24
2.2.3 Numerical Simulation
Discussion of the Code Input. To simulate the semi-infinite fracture,
two boundary conditions are specified here, just as they are for Problem 1.
At the maximum extent of the system (10.24 ), a convection-only condition
3T (x-L) 0 0 (2.2-1)
is easily specified in the code input (Figure 2.2-1) by default.
At the origin, however, the specification of unit concentration (Equation
2.1-6b) is considerably more complicated. Since SWIFT is designed to evaluate
repository-site performance, only facilities for radionuclide sources have
been provided. owever, a constant-concentration boundary condition may still
be used by taking the boundary block (Block 1, in this case) to be an
essentially infinite well-stirred reservoir. In fact, this is a basic
physical definition for a constant-concentration condition. Distances, of
course, are measured by taking x - 0 at the interface between Blocks I and 2,
and the dual-porosity region begins with Block 2 (see Card ROD-3). To
implement the concept of a well-stirred reservoir, four separate steps are
taken here.
Firstly, the thickness of lock I is increased to
AZ (1.91985 x 10 + .4) , RI-26 and R-19 (2.2-2a)
using the modification cards. Secondly, the concentration of this block is
specified as
Cl ' 1 1-4 (2.2-2b)
Thirdly, provision is made to replace, with a source, the mass lost from the
reservoir by both decay and convection:
q - C p()4ax 6yhz q) 7.12 x 10 kg/s , R2-LO (2.2-2c)
2-25
1 2 q 5 & 7 6L13�.5agq01Z34',6t6q01Z3456Tqq0IZ3i54?5qOIZ34.567Sq0Li345676q01Z3456 leqoKzli5Etoqo
* -- * *-0.-*-. 0.S* *K
I KPR05LEI� P40. 2. THIEE-COMPOP4EffT #4UCLIOE TRANSPORT KN PRAeUREO RECtA.K
2 KRZ�83. CDI4PARISI3N suits 'FTRANS' - TPHEA1CAL IEPREIVI1TATIQNI
I!-1-L
*5-3-2
31 �I
� K ZR.K
54 QK
6 KIOOIAO SI
7! 0.K
6 1 0.I
9 I 1'I
10 * 1.I
ii. I a
12 1 12� e.g..
13 1 2 1I 4@��*
1* 1K
15 *I
161 0.K
17 1 0.I
11 1 0.1I
191 0.Ip. gee....
20 * E..OOEOI.*...g9s.
ZR I Z?80.1
Zz 1 2K
23 1 20.K
Z� K 0.2I
25 * 100.K
C 0 0
I . 2
2 1. 0
bo1 K
0 0 0 1
1 2 0 2
o 0 12.350
i.0
h.
,.0
,.0
0
* *.e.. REF* � REF
l I I I I 0
0
1. 0 1 0 0 011-3-1
Ii-3-Z
10-1
10-2-1
ERENCED IN TEXTERENCED IN EXERCISE * 10-2-2q... eeg*. .... e.... see
10-2-2
ROD-i
100-2
100-3
t0fl-3-SLt4
1.0 11-1
0.50 0.0 1.00E�50RL*2
1l-Z.5-1
11.2 .5-2
11-2.5-1
A 1-2.5-2S.....
1000.0 11-3S..,.'
11-4
11-7
li-I
11-9
0.0
0.0
0.0
0.0
0.0
0.0
1.013E05
2 2
0.001
0.001
0.001
0.0 4tP6.0
0.0 ".0
0
0
0. 100.
20.0
0.60
200.0
0.001
0.001
0.001
-- *f**n *.flSn*. *e�a*
Figure 2.2-t. Problem 2. tLsting of SUIFT 11 Input Data.
2-26
0RAr
; 1 2 3 * 5 7 T a1i3*SbflQ4lR4Sl7S01Z34iSi7l6f11lQ0i13*51a76,0190137901fl45l76901Z3*sb?690
,_______ ,_______ ._______._a______._n______- - *--- -e----- -
126 1 100.0
Iz' t a o
IZS I 0 O
30 * 20.0
1 I ZO5.fo-4
32 1 6.*E-Z
33 1 L.a
34 1 Z64
35 .3.*OSZE-0?I
36 1I
37 1 1 tI
3e t l.8
39 1 0*0I
40t
4I t 0 0I
42 1 1.001-to
43 t (11 *
0.001
20.0
20.0
I.OL3EOS
I.OE-1
O. 12
ZOO.O 0.001 at-to
t1-t1
tL-11
Rt -I,
Rt-Il0.0
2.0C-1
0.256
000
*.OE-3
0.512
3.4052E-7 3.*CSZE-? 4.*6?-5eqc
t .0
0.o
I1 1
1.0
0.0
Z.OE-3
9*.02
0.0
0.0
0.0
1.6E-Z 3.fe-
0.0 00Rl-i0e~o o~etL-21
Rt-26-
0.O1.9195E09ftl-Zb-2O05*0 @@0
ti-2-3
t1-26-LN
RI-2
tl-I
1.0
0.00.0 0.0lo i.oI
* 1
Is I1'
*6 1'
T 1
t
1*5 1
I*9 1
I
I
1002S 0 * l .4* 040421' le 1)?*0e~~~0#4*00
0.0 100
l.ORDl-Z-1
R1D-Z-Z
RID-3-LN
110-A
RtD-S-Lk
* 1*4 +
a 0 I
I I 1 I I 1 1.0
t1-I I
I1-4 1
1-4-OLN *
IS � S-s-a*----a---�n----4n--aa
Figure 2.2-1 Continued.
a
2-27
O4r1 2 3 * S I a
12)651g90120 6 ?lOqatZ 30567Sq19O34% III1Z3456t6,O L3416UIC Zl3456T7IotZ34s6Ie00¢________, ______+___ ____+____- - ------__ _ _ * _ _ * _ _
ISt I
KS2 1
S3 1K
54 t
55;4I
S7 t
sI
S I5I
60 4
K61t
I62 1
63 1
6 *I
66 1
71 1661
?2t
II
69 £1
73
6I
72 I
12
73 1K
1' 1
75 4
£
2 21 I I I I I
I I I t t I t
a.o
L
0.0
02
* L
I
0.sa
0 0 0 0 0
2
1-I.M5E-ll
41.IS7E-11464694*44
RIlA-
R14-1
RI &-t-BLN
R1A-2
2-1
12-4
2-4
RZ-6-e
12-0-2
RZ-6-BLU
RZ-7-1
RZ- 7-Z
12-7-1
R2-7-2
R2- 7-BLN
I I.
1.0
2 21
1.0
0.0
O 0
0 0
1 .
0.0
L I
0.0
0.0
0 0
0 0
I I
200
I I
20.0
0.0
0.0
0 0 000 0 0000
1 0 0 0 1
0 L3 *1 000
I
I I
c.o
1 1
0.0-?.IZIZE-S0* es.e¢*4
IZRZ-13
RZ-1
R2-9
RZ-ic-I
R2-1C-Z
1Z-108LN
12-11.53
.64( 06
1 -1
56.6EOh
-l -l -l -1 00
II
III0 0000 0 00 -l 111 0042-13
*e4 fb4 * I
Fi6ure 2.2-l Concinued.
2-28 OA
I 2 3 4 5 7 a
LZ3456s90123s*TS9O234Q67iZ356?l6012154t7890Z234567890l234567902345678go* - - -- - _ _.___* e...* e_ _ _*
I16 l 0 0 0 0 a 0 0
?1 1 2.00h687 1.222(01£
is I -L -1 -l -I -1 -I 000
72 I 0 0 0 0 0 0 0I
so 1 .oEdt 1.ZIE307
07 I I -t -I -1 1 -1 0001
69 1 0 0 a 0 0 a 0I
63 1 b5E07 2.443107I
14 1 -1 -1 -1 -i -1 -1 000
9S * 0 0 0a0 0 0 0
66 K 9.716f0? 3.4E,5tE
I
87 I -t -1 -1 I -1 000
e I 0 0 0 0 0 a 0
89 I 1.i7EOI 4.684EOI
9. .-1 -i -L -1 -L -1 000
9i I a 0 0 0 0 0I
92 1 2.150108 6*906E07
l - -l -t -L -1 -1 000
10 I a 0 C 0 a 0 0
'IS* 3.126e06 9.765(0?
qU I 1 -1 -L -1 I -1 000
a1 0 0 0 a a 0
98 I 4.507E06 1.381f86
99 1 -L -L -1 -1 -I -1 000
Ica. 0 0 a 0 0 a 0
0 0 RZ-I
o 0000
a 0
a 0000
a 0
0 oc
o a
0 0000
o o
0 0000
a 0
0 0000
0 0
0000
a 0
0 0000
0 0
0 00 -I III OOf2-13
aZ-i
o 00 1 I *1 004Z-I
IZ-1
0 00 -L III 0022-13
RZ-1
0 00 I III OlZ-13
12-1
0 00 -I III 0012-13
1z-1
0 00 -I III CORZ-13
QZ-1.
0 00 I 11 C002-13
I2-1
0 00 -1 III 00R2-13
AZ-1
* * *… * ... i n *.Sb...S * - e * . *~~~~~~~~~~~~~~~~~~~~~~~
Figure 2.2-1 Cont inued.
2-29
Okr
I 2 3 . S T a123*41 7*qO IZ34S67SC01214 6tq9 1 IS9w I* *O . a234 S676q01ZI45679012 34 S6
¢______~~~ ~ ~ ~ ~ ~ ~ ~~~~~~~~ ,_ _ _ _-__. _ _ ._ __ _ --__ _ _ _ _ -_ _
101 I b.'45qE05 .OQSZEOS
LOZ t -I -I -I -1 -I - 000
I03 1 a a 0 a 0 0 0
104 t .20E60 Z.761F0
105 -I -1 -i -1 -L L 000
LC6 I a 0 0 0 a 0 0
IC? I 1.OLE09 3*904e08
I06 I -I -L -3 -1 -1 -1 000
109 1 0 0 0 0 0 0
110 1.B64EO9 5.520EOI
111 1 -1 -1 -I -I -1 -1 000
112 0 0 0 0 0 0 0
113 1 Z.6'sSEQ9 7.60SEOBI
114 1 -1 -I -I -1 '1 - 000
115s 0 0 0 a 0 0 0
316 I TZ Q E O tS
117 1 I -t -t -t I -1 000I....4 *4***
11i £ a a i a a o1
0 0000
0 0
0 COO
0 0
0 0000
o 0
0 0000
0 a
0 oo
o a
0 0000
0 0
0 00 -I ItI 0OR2-13
R2-I
0 00 -L 11 o0t0-13
R2-I
IIIIIIIII4
IIIIIII1
a 00 -1 11 0012-13
R2-I
0 00 -L L11 0012-13
2-L
aII
I
IIII
II
0 00 -1 111 002-13
t2-L
0 00 1 111 OOR2 13 1I
12-L-ST0? tI
*O *- . _ - 0 e. _a _ _
Figure 2.2-1 Continued.
2-30
Here the various component terms, except for C and Ai, which are given
above, are prescribed in Cards R-6 (p), RO-1 (T Ln(2)/)), R1-20 (),
RI-l7 (AxI), R-18 (AyI) and R2-6 (q).
Fourthly, and lastly, to implement the concept of a !'ell-stirred"
reservoir, the dispersivity
aL 4 10 u , R1-2.5-2 (2.2-2d)
is selected for Block I only. This value, without question, is sufficiently
large that the transmissibility (Theory and Implementation, Section 6.1.3)
cannectir. Blocks I and 2 reduces to a boundary-like transmissibility for
Block 2. In te data input, this value is given to Rock-Type 2. Rock-Type 2
is then ascribed in RIA-1 to Block I only. Rock-Type 1,
with aL = 0.5 m, refers to the transporting region of the fracture, Blocks
2 - 21.
Results. Figures 2.2-2 and 2.2-3 show the results from the finite-
dirference SWIFT plotted alongside those of uyakorn's finite-element code
FTRANS. Both within the fracture and within the matrix, the two show
reasonable agreement. The dissimilarity between Figure 2.2-2 and Figure
2.1-5, however, is somewhat striking. Obviously, transport within the
fracture is significantly more retarded by the spherical-matrix
characterization than by the prismatic-matrix characterization. In spite of
reduction of approximately three in the matrix diffusivity, the spherical
characterization still yields a retardation which is greater, by a factor of
three or more, than that for the prismatic characterization.
Discussion of the Code Output. The results shown are dependent, at teat
in part, upon the convective velocity and upon the constant-concentration
condition discussed earlier. To check these two, the reader is referred to
the microfiche listing. Insofar as the fluid velocity is concerned, note in
the table "Homogeneous Reservcir" the value of porosity for the composite
global system ( - 4.167 10 ). Note also, in the first recurrent series
output, the first table, which is entitLed "INDQ IWELL .. T". The purpose
the computations referred to here is simply to establish a steady-state flow
with boundary wells, Well rates are given and the printing of the Darcy
2-31
1.0 -
-FTRANS
0.8 0 SWIFT
90.615 d
z° 0.6 3619 d
cc ~~441 dzw
0 \
0.20
0 1.0 2.0 ;.0 4.0 5.0
DISTANCE DOWN FRACTURE (m)
Figure 2.2-2. Radionuclide Concentracions Within the Fracture for a
Spherical Characterization of the Rock Matrix.
2-32
0.4
- FTRANS
0.3 90,615 d o SWIFT
z x = 1.0m0 ~~~3619 d
0.2 41 d.2z
0
0.1
0 0.2 0.4 0.6 0.8 1.0
DISTANCE INTO MATRIX (m)
Figure 2.2-3. Radionuclide Concentracions Within he Rock atrix for a
Spherical Characterization of the Rock Matrix.
2-33
velocity is prescribed by the control parameter IIPRT 13. The table "X-
Direction Darcy Velocity" then provides the desired check on the convective
velocities. As shown in the table, it is constant throughout the entire
length of the ystem, and it has the magnitude
v u (4.8208 x o 12 /4.167 x 10 ) w/s - 0.01 m/d (2.2-3)
which is consistent with the problem specifications (Table '.l-0.
To verify the constant-concentration boundary condition, it is sufficient
to examine the radionuclide concentration tables labelled "Component-L
Concentration." As the reader will note, there are seveiral such tables, which
are displayed at selected time intervals out to the MnximuM time, t 90,615
d. In each case, the C 1.000 condition is maidrAined to at least 4
significant figures.
2.2.4 Exercises
Level 1. Using the Input-Data Guide, Theory and Implementation and the
microfiche listing of the output, as required, complete the following
exercises:
Exercise I
Exercise 2
From the description of the problem, Equation (2.2-3) and Table
2,1-1 what physical parameters are required for the blanks
labelled "1" through 4" in the input-data set (Figure 2.2-1)?
Check the answers by referring to the echo of the input data
printed at the beginning of the output.
Interpret the output-control parameters identified in the last
R2-13 Card of Figure 2.2-1.
Level 2. The reader who wants to develop a deeper understanding of the
code should also consider the following exercise:
2-34
Exercise 3 If the retardation of the rock matrix were increased from ' -
to K' a 3 would the penetration of the radionuclide front in
the fracture b increased or decreased? To answer this
question, the problem may be run anew with the value of he
distribution coefficient on Card RID-4 change.d to kd
7.41 x 10'6 kg/m3 .
2-35
2.3 PROBLEM 3. TRAKSPORT OF A RAD1ONUCLtDE CAIN IN A FRACTURED POROUS
MEDIA UINTRACOIN, 19831
2.3.1 Objective
* To illustrate the coupled effects of fracture/matrix transport and
chain decay.
2.3.2 Description of the Problem
Problem Statement. Radionuclides, it is assumed, are buried in a
fractured medium. As indicated in the schematic drawing of Figure 2.3-1, the
system consists of parallel horizontal fractures with apertures 2d and
spacings 2a. It is infinite in its lateral extent but confined to a vertical
thickness b - At by impermeable beds. The radionuclides each from their
storage leach duration, 2) and contaminate the entire unit. They are then
transported laterally by a one-dimensional flow field. Convection and
dispersion occur within the fractures while diffusion and sorption occur
within the rock matrix. The geometry presented for the diffusion process may
"e characterized by one-dimensional prismatic units.
Transport Equations. Mathematically, transport within the fractures for
radionuclide r i governed by the equation
a 3 * aCG. (pe ) ax (PD ax) r (2.3-1)
+k A K #pC - A K~ -a(K toC)rr-l r-l r-l r-l r rPCr ' at r r
and transport within the rock matrix is governed by the equation
a Sc,S.(p D ) r (2.3-2)
k X K -1~ A K'p*C'1 La Ko~~r,r-l r-1 r-1 p r-1 r Arr'P' at rK Cr
Coupling then arises from the flux at the fracture/matrix interface:
AC'rr %0P'Dt -s (0) (2.3-3)r m ~as
PA -IN W.? *.:
N~~~~~~~at-'. ~ ~ ~ ~ ~ ~ ~ ~ ~ at.~
K.'
Figure 2.3-1. Problem 3. Schematic Characterization of Transport Within the
Fracture and Diffusion Within the Matrix~.
2-37
Initial/Boundary Conditions. Within the rock matrix the concentration is
initially zero:
C'(xG.tuO) 0 0( a a , > (2.3-4a)
a no-flux condition holds at the symmetry boundary:
aceSr- (x,6-at) f 0 , C 0 Q (2.3-4b)
and fracture and matrix concentrations are identical at the fracture/matrix
interface:
x(,s0,t) - Cr(xt) P t (2.3-4c)
Within the fractures the initial concentration is also zero:
Cr(xth0) - 0 a x (2.3-5)
and the infinite boundary is held at this initial concentration.
Source Specification. At x 0 however, the boundary condition for the
fractures must take into account decay/production processes within the
inventory and the leaching of the inventory. To do this, the time-dependent
source concentrations C are defined by the Bateman equations:r
dC
dc rr-l r-l r-l r r(2.3-6a)
with initial conditions
C (t-0) * I r/UT (Z.3-6b)
The B2 boundary condition chosen here is then expressed in terms of these
source concentrations:
UCr r x uCr , x (2.3-7)
0
2-38
Input Specifications. The geometry, the leaching process and the
transport process are characterized by the parameter values given in Table
2.3-1. Radionuclide-dependent processes are then cnaracterized y those given
in Table 2.3-2.
2.3.3 Numerical Simulation
Discussion of Code Input. Within the input-data listing of Figure 2.3-2,
two items are selected for emphasis in this section. They are the fracture
velocity v and the boundary specification B2 . In contrast to Problems I and
2, the velocity here is established by means of a pressure boundary condition:
v a (Kfp(gfg )0I(p-pl )/L - 500 /yr (2.3-8)
In this equation the porosity of the global system
f = d/(ad) - 1.99996 x 10 5 RI-20 (2.3-9a)
is fixed by the input specifications (Table 2.3-1). The system Length
L 2,500 RI-17 (2.3-9b)
(the sum of alL increments in Card RI-17) is taken to be Large compared to the
observation length (500 m) in order to simulate a theoretically semi-infinite
system. Of course, the gravitational constant (g 9.81 m/s 2) is fixed
internally by the code.
Aside from these, all other parameters in Equation (2.3-8) are
arbitrarily specified to yield the desired interstitial velocity:
p - 1000 kg/t 3 R1-3 (2.3-lOa)
K - 1.99996 x 10 5 MIs RI-ZO (2.3-lOb)
Po = 388.44495 Pa RI-28-2 (2.3-1Oc)
and
PI ' RI-28-Z (2.3-10d)
2-39
Table 2.3-1. Problem 3. Input Secifications.- ~~ ~~~~~~~~~~. -
Parameter Symbol Value
Observation Length L1 500 m
Aquifer Thickness b 5.0 e
Fracture Aperture 2d 1.G x 10- 4,
Fracture Spacing 2a 5.O X
Fracture Porosity 4. 2.0 x 10-5
Fracture Velocity v 500 mly
Fracture Dispersity aL 50 m
Matrix Porosity of 5.0 x lo-
Rock Density PR 2700 kg/m3
Matrix Diffusivity D. l o1 2m2 t
Leach Duration 10 y
Table 2.3-2. Nuclide Inventory ( ad Matrixl Retardation (R3i.
Number Isotope Inventory Half-Life2 Retardation Distribution Specific
(Ci) (kg) t(y) K Coeff iciend Activity
k (a.3 11:g) A (Cilkg)
l 2 t, 24cm 0.7 4.074xl0-3 8.500x103 570 0.211 171.8
2 2 3 7Np 1.0 1.418 2.140x106 80 2.96xlO0 0.7055
3 0.OC4 4.l47x10-4 1.592x106 30 1.11X10 2 9.646
I No sorption occurs within the fracture.
2 The corresponding decay constant is given by - n(2)/T.
3 The two sorption parameters are related via K p k
2-41 D/4p
I I 3 * s 7 r Sall54?R0901245&7O9 012 34S7qO3I4I,6790Z34#blqoZ1i 567 qO1234S76901Z4S7&Tq0
._�__�_��__�4�___-- - - -___ _ __ --------- *--------- --- *--
I t II fraciLE4 No 3 INr*&COIN LEVEL CSE S 2.13,G2vPZ.TZEtLL 1-1-1 I
I I2 112153s ONE-0 TRAtSPORT IN FRACTURED EDIA WITH CHAIN ECAY R-1-2 t
3
10
5
6
71
a
9
IZ
la
11
12
13
14
15
l19
20
21I
2Z
23
z4
2S
II '. 0
t 126 It* 1Z2I
* 2
1245CM z45IIZ37N4P 237II LI1233 U 233
* z
I*+*e* ***
1 061 *
I a
II Zs
4 '126I
0 a 0
I 1411.
I Z I
I 0
2 I
1.0
3 1
140
O 0 1 1
1 t -1 0
#00*0*0-0¢
0t 121 4
0C (31 **4 44*9*4..
* I 4** *
0.0
1.0
0.0
1. *4e**" IEFEREN1100 . s**,* F .... N
A-Z
0 t 1 O-3-1
01-3-2
Q0-I-1
R0-1--
R0-t-Z
R0-1-t
ft-I-Z
,.. . . . . . . . . . . . . . . 10 -2 -l1E0 IN TEXT CEO N EXERCISE . R0-2-2,.@.............
Roo-IL
R00-2
100-3
Ro0-3-BLN
0.0 1OE-20R1-2
kl-6
t1-10
RI-LI
I I I I L 0
IIIIIII
I
IIIIIIII
I
0.0
0.0
27000.
2 2
Z0.0
0.20
IOC.0
100.0
0.0
0.0
0.0
0.0
2 2
0.41
0.001
0O001
0.001
20.0
0.0 4186.0 1.0
0.0 o.0 SO.O
20.0# 171 *- (8) *4 4 4 * 4 4 4 -*44 4 4* 9 4 *
20.0
0.C0
200.0
200.0
0.001
0.001
0.00t
0*001
*______ - -_ _--- _- - _ _-_- -_-__- -_-_-,_-_-_-- -_ _ -_-_ -
Figure 2.3-2. Listing of SWIFT 11 Input Data.
2-42 O
I 2 3 4 S a 7 6L21 7s,0sLZ3 so7e zqoZ s 67s0oL 3 sT0 I z3s6sqoIz36s6 lso0234s6e,023seqo
._______. ________._.__ _. ---------_ _-------_ _
26
2i
iiZ'p
30
i1
3z
33
31
3S
16
3?
i8
3,
40
41
4Z
43
44
is
4,
4S
49
$ 0
I oo.0 20.0
I O o
II 200 0.0 0.0 0.0
t.0 Z.O 3.0 9105.0
* 30.0 40.0 50.0 60.0I1 1.0tI 5.0001
I I .999q t -51 .qqqIE-51. qqq6f-S 1.q9tq6{-S 0.0
I
RI-lI
RL-L2
It-16
60.0
1o.0
16*100.0
0.0 0.0
I0
IIIIIIIII
I
IIIIIIII
i 0
I 1 1 I
t.0 3.*4449S
126 126 1 1S 00 9 S 0 0
Z.0 0.0
I 1.
20.0
1
20.0
0.0
0.0
20.0
30S.O
RI-Zo
R 1-26-BLN
Rt-2'
RI-Z-I
RI-ZI-Z
R-ZB-1
RL-2i-2
RI-2I-GLN
RI-33-SLN
RLO-1
t.ORt1-z-1
I 1t-2-2
RIV-3-OLN
t10-4
R11-5-BLE
I-I
SIA-Z
RLA-3
0.0
I.cOF-t2 0.0 5.0OE-03
I 1 2.4 t 0OO-3
0.0
0.0
0.0 C.0
I. 19 *, 10) 0* III) *I**v***o**.4..044*-.0....eoI1I 0 0 0I1 00I* 0I* *- e -4 - a n . . e . ... n. ae
Figure 2.3-2 Cont nued.
2-43 £W/Ar
51
52
S13
1 2 3 S 6 ? a12 34S6?GqOt1Z34S7?gqOIqa127gqto123456a1Z34S6790tZ3)56XqJ01231567G 01Z3567690
. __ _ _. _ * _ ____ . ______ _ * _ _ _ _ _ _ -_ . _ __
I II 0.0 0.0 0.19q96 0.0 tlA-a II I
too In.0 L.o tIA-S II II I I I I I I RlA-& II
54 1
5 OCRI
56 INP
57 1 uI
58 I
S4
237
233
0.0* +**..*,*
* 121 0
* 141 i*00 0,***6
IRIA-1 I
IR11-6 0
IIt A-* II
Rl-S II
a I -9 II
10.0 10.0 10.0
sq
60
6?
63
65
66
67
66
70
71
72
73
7 #
75
II
102 I I
I
I a-1
1.
1 0 0
I
I too 1 1 -
II I - LI* a a
12. 115E 11
I *4I IS I
16.3072 EllII 1 -1I
I*2S.ZZISEl1I
K-L
W-1-SLN
aZ-1
R2-2
I
0 .5
a a a 0 0 0 0
L.0
1 -1
0 0
0.0
15.bSE09
-L -I
0 0
15. T7SEOq
-t -1
a a
3.153&E1O
-1 -t
a 0
-1 I 111 00000 0 1 -1 III
2 0 a 0 0 a
1112-13
tz-1
t-tc0*s
I -t 221 I0000 a
a 6 0 0 0 a
0 I 211. O1Z-L3
12-1
I III aI
II I.I
I
-1
C
E12
-1
0
1 -1 Zll
a 0 a
1. 0000 0 GO I 211 0OR2-13
0 0 t2-1
t -1 221 oa
a a a 0 0
a 00 1 211 R002-t3
tz-I
*e4n *flfl p 0 0S - *-in �
Figure 2.3-2 Continued.
2-44
1 2 3 4 5 4 7 512)0s67sq623% s 9gsolrs67801 sol 6023 42~s~SO 6'z34s67ssoz23,stes78ol2345s7es0._ _ _ _ _ - --- - _- - - ------
I76 I5.67a46EI2 3.L53beI0
I77 I I -t -t -1
?a I a 0 a L
I -L 11
0 0 0
I 0000
0 0
III
a 00 1 211 OOtZ-13 1
42-t-E0 II
*-e *n *
Figure 2.3-2 Continued.
2-45
In order to approximate the B2 boundary condition, Grid Bock is
specified as the source in Card RlA-6. This block, of course, is of finite
width with
Axt a 5 X , Ri-17 (2.3-hla)
However, this increment is only one percent of the observation length and
therefore should be an acceptable approximation to the theoretical boundary
condition of Equation (2.3-7), which is rescribed for an infinitesimal
increment.
Further, the source rate is specified within SIFT by the equation
Rwr ' -m rt )/wta (2.3-1 lb)
(Section 3.4 of Theory and Implementation) where
Rwr a source rate of radionuclide r kgrn3(bulk)/si,
Or a waste density of radionuclide r [kg/m3(waste)I,
V = volumetric density of wastes 1m3(waste)/m3(bulk)], and
ta a T = ach duration 1 1
Quantities ta, m(t-0) and pw must be specified in the input.
Prescribing the leach duration is straightforward. However, the other two
require some explanation. Here it is assumed arbitrarily that the waste
inventory occupies one m3 of volume so that
PW I xyz - 0.199996 , RIA-4 (0.3-12)
where Ax, I m, y 1 m and az * 5.0001 are specified in Cards RI-17,
RI-I8 and R-19, respectively. Thus, the initial densities mr(t-0) have the
same numeri,.al values as the nuclide inventories r given in Table 2.3-2, an.
the code internally solves for the Bateman decay (Equation 23-6a).
2-46
Results. Results are presented in the tabular form prescribed by the
INTRACOIN project, and they exhibit reasonable agreement with other codes used
in that study. Breakthrough curves at the observation point L 500 are
given, in abbreviated form, in Table 2.3-3 and, in expanded form, in Table
2.3-4.
D:scussion of Code Output. The concentrations given in these tables (in
Ci/m3) may be obtained from the computed concentrations (in kg/kg) via the
fluid density (1000 kg/m3) and the specific activities of Table 2.3-2. For
example, consider the concentration for 233U at 180,000 y. The point of
breakthrough is located in Block 102, which way be verified by examining, in
the microfiche output listing, the echo of the block sizes (see also Figure
2.'-3). The reader will also note in the echo of Card -l that Block 102 is
designated as a nuclide monitor block. This facility allows one to store (on
Tape 9) computed concentrations at this particular location. Here such
concentrations were stored so that they might be post-processed and Tables
2.3-3 and 2.3-4 written automatically in the INTRACOIN format. The
concentration computed for 23 3 at 180,000 y is printed in next-to-the last
table of the output listing, the one entitled 'Component-3 Concentration."
The value printed there as a mass fraction may be then converted to the
desired units:
C(Ci/ ) pA C(kg/kg) a 1.95x105 Ci/M3 (2.3-13)
using Tables .3-L and 2.3-2.
To provide a check on the calculation of concentrations such a these,
four material-balance summaries are printed at selected time steps. Two of
them, the one for unleached components and the one for leached, but
undissolved, components pertain to the source block. The other two tables.
one for dissolved and sorbed material in the fracture and one for such
material residing in the rock matrix, pertain to the system as a whole. Each
monitors the conservation of mass within a particular phase or subsystem and
then summarizes the information with balance quotients. Ideally these
quotients should equal unity. The various categories used in these tables are
defined in Table 2.3-5.
2-47
Table 2.3-3. Breakthrough Parameters.
NUCLIE C-SAK T-MX T+ (50t) T- (50%)(CI/CUA.M) (YEAR) (YEAR) (YEAR)
1 CM 245 1.6952E-11 49001. 28802. 79362.
2 NP 237 1.6665E-06 l5Booo. 77616. 464564.
3 U 233 1.9524E-06 18000 . 8245°. 688312.
…---.-------- _-----------------.-----._-_ ---------------
A
2-48
Table 2.3-4. Breakthrough Profile.
-------------- e- f ---- a--------------------------
TIKE(YEARS)
CONCENTRATION (CI/CU.K)____--- ---- _ _---- ------ ---- e_ -------
Cii 245 NP 237 U 233_-_a--a - a-------a------ --------- a eec -
3.50E+037.50E+031.15 E+O4I.55E+04I .95E+042.35E+O42.75c+043. 5E+043.55E+O03.95E+044.35E+O04.75E+0L5.15E+045.55E+045.95E+O46.35E+046.75E+047.15E+047.55E+047.95E+048.70E+O09 .50E+041.03E+051.11-+051. 19E+051 .27E+051 .35E+051.43E4051.51E+051.59E+051 .67E+05I . 5E+051 .86E+052.O4E+052. 36E+052 .68E+053.00EE+053. 32E+053 .64E+053.96E+05
5.1824E-t71.4297E-141.9872E-139.131OE-132.4360E-124.7405E-127.5312E-121.0400E-Il1.2972E- I1.4981E-111.6295E-1II .6698E- Il1.6855E-tl1.6280E- II1.5306E-1I1.4064E-111.267OE- I1.122tE-l1I9.7906E-128.4309E-126.1792E-124.276SE-122.8.703E-121.8777E-121.2019E-127.541 41Eal34.6455E-132.8133E-131.6778E-139.869lE-1I5.7358C-143.2981E-141 .5172E-144.1178E-153.7330E-163.1725E-172.5832E-182.0417E-191 .5793E-201.2023E-21
6.6839E-1 )1. 8597E-098.6g84E-ogt.2451E-084.3552E-087.1697E-081.0627E-07IA.654E-071 .9177E-072.4129E-072.944SE-073.5077E-074.0967E-074.7077E-075.3367E-075.9006E-076.6364E-077.3016E-077.9739E-078.6519E-079.9323E-071. 1302E-061.266BE-061.3952E-061.4974E-o61.5705E-06i.6192E-061.6485E-o61.6631E-061 .6664E-061.6611E-061.6494E-o6I.6256E-o6I.5743E-06I.660E-06
1.3541E-O01.2473E-061.1491E-061.0599t-O69.7947E-07
S. 2623E-II1. 1426E-095.2641E-091.4021E-082.8395E-084.8897E-087.5724E-081 .0886E-071.4819E-071.9347S-072.4444E-O73,0079E-073.6221E-074.2841E-O74.9906E-075.'7386E-076.5251E-077.375E-078.2028E-079 .o888E-071.o824E-061.2767E-061.478OE-O61 .6422E-061.7481E-061.8183E-O61 .8665E-O01 .9oooE-061 .9231E-o61.93B3E-061 .9I75E-o6I .9518E-061.9514E-061.9388E-06I.8923E-o61.8281 E-061.7551E-O61.6785E-06I 6013E-061.5255E-06
2-49
Table 2.3-4 Cont nued.
--- - n- -- --- --- --- -- -- -- -- ft --- - -
TIME(YEARS)
CONCENTRATION (Cf/CU.K)---------------------- ____---------
CA 25 NP 237 U 233-- ___________________________________________…__
4 .28E+054.60E+o54.92E+055. 24E+55.56E+055.88E+056.ZOE+056 .52E+056.84E+057. 6E+057.148E+057 .8OE+5
9.0552E-236.8074E-245.5825E-258.5281E-265 .0058E-263. 4527E-263 .5234E-268.2965E-266.8849E-z65. 1387E-263.9422E-261 .3475E-25
9 .0709E-078.4 9I E-077.8321E-077 .3009E-076.8206E-076. 38I43E-O75 .9872E-075 .6247E-075.2931 E-07X .9889E-07X..7092E-074 .4516E-07
I .4520E-06I . 38 I6E-o61 .3144E-061 .2509E-061. 1907E-061 .1339E-061 .0803E-o6I.0299E-069 .8243E-079. 376eE-078 .9559E-078. 559SE-07
.______________________.__..__,___________________
. 0
500
07
1 2 A
BLOCK
a x(m)
125
3
126
5 59
I
0
F i g u r e .3 -3 k r i d d i n go f C h ePr a c t u r Itac g LX S y t m
2-51
-ago10
I.a
0
v
@3-@3
.1
C.0
la
10
.4
1E
14
10C
-
la>rA-G
oas
.-. 14
toLo
C
K
l~ C
-
0
.D
~ o.
.
C :
v 1
(.S
e x*A - *4
0 .10
Ev G
C *
0a,
Cx
C z
.4' ..4
u
10100 L
c
C
oS*C
oH
La
1
C
10-
C:V .4 n :1
e
u 10101>
CU
o
e4. O1
J:X
U4,
10
43L3
a. .4
0
4.'
kw
0W .
0C L o
C~4
I D
4J0 @3
Z C U
I Z
X C t
43 .4'_3.4'
O- @LaO 0.i
t:k- E016IX
0*4 44
s4
*S Z"*
@3-.4
U 3@343
CIJ
10 4b
..
.4 OE I
041
-
.C
.1
'4c
o@3'-Z 0140n Q *W
SLoC
30 '-'.e
@3o
c.1., 3
U-
I XCo .4'
0 C
e
>' C
*Ca
X
u >
La U -
C *4
C41 1
C CS 0
10 6 0
*0a-
-
.0 . C
eV v
C
D4- U
W0
@CU C
Cr i
~a 4'
C._'
U' a
. @3
87*4 U @
Ca C 10
7 0 c.
.4'
U-.4
4.. - '.C
C 10C 4
S. V
Ca
3-.4 A
43
z C
OS
.C C
oa 4'
CU
O4V4'
V4.
3M
e@
0 W-.
p.'
V C
O CC @4.4S
.C C3
S S0
Cu
.0 Q
10
.C @3u
1C UtC
O v0
C O
C
uC
C @3 .4'
. _
p. .4'10. *-.vn7 4...
I0
.,
C
C
0
.
I
v.0 a
a _
a4.)
>v
0
I10E
&A
u
@3
VC
"Dw
0Iv
03
14
-c
C._4.'
10.4
-
* ..4
C .4C oO m
a~ aU 10
_ C0 aC" 0
C Xo _
0 v
C ._
@ C:2N
S1vin - N
9
2-52
2.3.4 Exercises
Level L. Using the Input-Data Cuide, the Theory and mplementat.on and
4he microfiche listing of the output, as required, complete the following
exercises:
Exercise I Fill in the blanks lbelled "I" through "15" in Figure 2.3-1,
making reference as necessary to Tables 2.3-1 and 2.3-2. The
data called for here are either source- or radionuclide-
dependent information and may be verified b reference to the
output listing.
Level . The reader who wants to develop a deeper unierstanding of the
code should also consider the following exercises:
Exercise 2
Exercise 3
What effect does the rock matrix have on the transport?
Contrast the computed breakthrough-time parameters (Table 2.3-3)
with those for the case of convective transport in the fracture
with no matrix effects.
Table 2.3-3 indicates that the peak for 237Np occurs before that
for 233U. Why? Since the latter is in secular equilibrium with
the former, one would expect the peaks to occur at approximatley
the same time providing Chat the retardations were equal. Here,
however, 37Np has the larger retardation (Table 2.3-2). Thus,
should not 237Np peak later than does 233U?
3-1
3 FLUID FLOW THROUGH DUAL-POROSITY MEDIA
3-2
3-3
3.1 PROBLEM 4. ANALYSIS OF WELL-TEST DATA FOR A DOLOMITE FORMATION
[PAHWA AND BAXLEY, 19801
3.1.1 Ob jective
* To illustrate the effects of stress-relief fracturing upon a slug-
injectior. test.
3.1.2 Description of the Problem
Background. A number of slug and recovery tests were conducted by Sandia
Rational Laboratories and by the U.S. Geological Survey (ercer and Orr, 979;
Dennehy and avis, 19811 during the site-charactericacion work for the Waste-
Isolation Pilot Plant (WIPP) [Powers et al, 19781. Two formations of
particular interest were the Culebra and Magenta olmite members of the
Rustler unit, both of which are signifleant water-bearing rocks lying above
the Salado salt layer. These tests were analyzed by Dennehy and Davis using
an analytical model and, since many of the tests exhibited an anomalously high
flow rate at small values of time, they were analyzed also by Pahwa and Baxley
119801 using a dual orosity implementation of their numerical code.
In the SWIFT Self-Teaching Curriculum, Problem 11 Finley and Reeves,
19811, the R2A test (Figure 3.1-1) of the Magenta member was chosen, and the
conceptual model of Pahwa and Baxley was used. This model assumes that, in
addition to the primary porosity, which may itself represent connected
fractures extending throughout the dolomite, there are secondary-porosity
fractures which extend radially outward an average distance of 2000 ft (Figure
3.1-2a). It is then these fractures which affect the small-time response of
the flow rate. As in the work of Pahwa and Baxley, heterogeneity was used in
Problem 11 to characterize the secondary porosity.
Problem Statement. Here, a different conceptual model is invoked. It is
assumed that the secondary porosity is provided by stress-relief fracturing
which extends to a radial distance of about one foot surrounding the wellbore
(see Figure 3.1-2b). The generalized dual-porosity approach of the SWIFT I
Model is then used to simulate the measured data.
3-4
INITIAL WATER LEVEL
- STATIC WATER LEVEL
TUBING DIA. = 1.995 in.
CASING DIA. = 6 o
I
Figure 3.1-1. Problem 4. Schematic Diagram of Well-Test H2A Within the
Magenta Dolomite Formation.
3-4
_
E zwtT i �' 2 1., 45 ,-,. � jWTP�Rf,--:,.' , �,.' 4 ; : :I " " I ii,
- . f INITIAL WATER LEVELj
I
5(
i
8'
2'
k I 1+- STATIC WATER LEVEL!5f
57,'
_4I TUBING DIA. = 1.995 In.
*- CASING DIA. =6
_ _,________________ l, - , - , _ . - - - S x .r - U- I -, , , , , , o
,f , .. ..
*7774 W t - / - / - > J
DOLOMITE g -A, leAf z z - F F --
,f . . - -.1 . ..1- z v-' 'SP ' 4 7 *!~ 7 7
, 4 >- 17 ,, / ^- z / f
5'
1 F1 F , * 0~ Z F0 z .01 z le it_
Figure 3.1-1. Problemu 4. Schematic Diagram of VeLt-Tesc H2A W~ithin the
Magenta Dolomite Formation.
3-5
(a) Partially Penetrating Fracture(s).
(b) Stress-Relief Fractures
Figure 3. 1-2. Conceptual Models for the Secondary Fracture Porosity.
3-6
Flow Equations. athematically, within the primary porosity, the flow is
characterized by the equation
T r ar ) - bw Sat * r r a'7
(3.1-1)
and, within the secondary porosity, by the equation
,as2-S 'sat O < s a , rw r r0 (3.1-2)
where a is the average length of the connected secondary fracturing and ro
denotes the maximum extent of the stress-relief fracturing surrounding the
wellbore (radius, r). The coupling between primary and secondary porosity is
provided by the reLation
r '- AVK' a.g (5-O) (3. 1-3)
Initial/Boundary Conditions. Within the secondary porosity, the initial
drawdown is zero:
s'(r,s,t-0) - 0 O s a rw c r r0 (3.1.4a)
the drawdowns are set equal at the interfaces:
6'(r,s-O,t) s(rt) I r. r r0 (3.1-4b)
snd a symmetry condition is prescribed within the interior:
as, (,sna,t) 0 rw c r r (3.1-4c)
Within the primary porosity, the initial drawdown is also taken as zero
s(r,t=O) 0 r r (3.1-5)
.
3-7
and the water-level above the bottom-hole condition i escribed at the
vellbore radius r:
strwt) - s0(t) (3.1-5b)
Input/Output Specifications, Rather than the draudown formulation, the
SWIFT Model uses a pressure formulation. Thus, in the equations above,
drawdown is replaced by pressure as the dependent variable;
a Po P (3.1-6)
where p is the initial pressure. Also the storativity and transmissivities
are replaced by the equivalent expressions:
S a P(g/g c)(eW+c) (3.1-e)
T - Kb (3.1-7b)
and
Si a p(g/g )W'(c ') (3.l-7c)5 c v
Input data appropriate for both formulations are then given in Table
3.1-1. For the primary porosity, the values given there are the same as those
used by Pahwa and Baxley. For the secondary porosity, however, hich is
assumed herein co arise from stress-reLief fracturing, the values given are
entirely different from those of Pahva and Baxely and were developed
especially for this problem by an optimization procedure. The remainder of
the input data, i.e., the time-dependent height of the water Level within the
wellbore, is provided by Table 3.1-2. Actually, the analytical work of
Denneh. and Davis and the numerical work of Phwa and Baxley treated the
wellhisr sorage implicitly so that the height of the water level, in addition
to the injected flow rates, was calculated. We, too, have a special-purpose
update which permits such a computation. owever, in rder to use the
publically available code on his, an illustrative example, the measured wateL
levels are prescribed as input data, and only the injected flow rates are
cal culated.
3-8
Table 3.1-1. ProbLem 4. Input Specifications.
Parameter Symool Value
Density of Water p 62.4 ib/ft 3
Aquifer Thickness b 25.0 ft
Compressibility of Water c, 3.0 x 1o 6 i-
Primary-Porosity Media:
Porosity O.10
Compressibility of Rock CR 4.0 x 10-6 psi-
Hydraulic Conductivity K 2.0 ;: 104 ft/d
Storativity S 7.58 x 106
Transmissivity T 0.005 t 2 /d
Secondary-Porosity Media:
Porosity t 0.10
Compressibility of Rock cR 4. 10-5 i-Hydraulic Conductivity K' 2; .0 ft/d
Length a *. ft
Specific Storativity ' 1.86 x 10-6 ft-
3-9
TabLe 3.1-2. Observed Decliine in Water-Level Height and
Bottom-Hole Pressure.
Time Interval Water Level Bottom-Hole Pressure
(days) Height (ft) (psi)
IQ-4- lo-2 505 219
10-2 _ 101 496 215
0.1 - 0.2 485 210
0.2 - 0.5 468 203
0.5 - 0.8 450 195
0.8 - 1.0 434 188
1.0- 1.5 415 180
1.5 2.0 392 170
2.0 - 2.5 374 162
2.5 - 3.0 360 156
3-10
3.1.3 Numerical Simulation
Diucussion of the Code Input. In contrast to the problems presented in
the previous chapter, this problem, like several that follow, employs a radial
coordinate system. The inner boundary condition derives from the presence of
a well of radius rw 0.276 t (Card R-22, Figure 3.1-3). This well
comrl..;ates with the aquifer through a weLLskin, which, in general,
represents a disturbed zone of modified hydraulic properties. For the SWIFT
Model it is assumed that the wellskin has no storage and that its hydraulic-
conduction property is characterized by a well index (Theory and
Imple entation, Section 4.1):
WI0 2v K3Az/tnIrIr) (3.1-8)
where quantities rl 1.0 ft, the position of the first node within the mesh.
and A .5 Et appear in Cards R-22 and R-23.
Here, the disturbed zone, e region assumed to have stress-relief
fractures, is also rpresented by a one-dimensional unit attached to the first
grid block (Card ROD-3). This unit does have some storage and a relatively
large hydraulic conductivity. In a sense, then, in this case the well index
(Card R2-7-.I) has been superceded by the one-dimensional unit.
Th? outer boundary is chosen consistently with the aximum simulation
time:
re (Tt IS) 100 ft (31-9a)
with r (Card R-22) being greater than the characteristic length r
r ' 2000 ft (3.1-9b)e
At this boundary, the numerical solution is matched onto an analytic
continuation for an (assumed) infinite external aquifer using the method of
Carter and Tracy [19601. The data appropriate for this method appear on Card
Rl-27 hrough R1-33. They should be examined carefull) since, in the
application of SWIFT, they are used frequently (Theory anJ Implementation,
Section 4.1).
3-11
1 2 3 A S 6 t a1z34s6t1901Z3is676soLz3A¶"s921 AS6rsolZ34AS&730oL2 345&J ol2346769s1t3ZI6 Is O
e---------- *------_-*- ------- - --- -.*
II I1PROALEII kNol 4 SVIFT El EXANPLE. ENGLISH UNITSs 12/63.I
M-t-1
2 IhTERPRETATIOn OF WELL
31 t 0 0 a10*+.+ *0*¢
5 s* O t I I1 50 I
1 0
I
I
S I 1' 1 1 1
91
1.0 *. (Z1 ** 21 0
11 1 1.0 1.0I
tZ I 120 100.
I
13 1 0 1 0 Z
£
Ii 1 t0. 1.0
1S * 100. 1.0I
16 1 0.0 70.0I
17 1 1000... 70.0I
13 I 0 01
TEST ATA FOR DOLOMITE FORMATIOI
a 0 0 a 0
0 1 2 0 1 0
pg--- ...e... . s6-660 .*
. '*o* REFERENCED IN rexr. "... PEPEPENCED IN EXERC
1 * 1 0
0.0
1.0
70.
7o.
6.0
62",
1.0
104
,o NEW MEXICO -1-Z
m.2
0 0 0-3-1
91-3-2
I t00 1ZSE .*00 Ro00-2
100-3
RO-3-LN
Rl-
0.0 .0001tt-2
*t-3
R1-6
Rt-7
R1-9
Rl-11
Rl-11
RI-1.
Rl-16
Rl-ZZ
61-23
Rt-26-LN
Rl-2?
Rl-Z9
tt-1t
II
IIIIIIf
IIIIIIII
IIII II.IIII
IIIII
III
IIIII
I
1.0
19 I 10.0 111.S $11.6 514.6
20 4 0.276 l.OC ZO0. S14.6[46*000*¢z4#9vS**OO4S¢¢¢*S*9***.*++.*.
21 1 25.. t31 *. 4 4') (S t *I *4*4**.44* ***4****.*..**.*
22 1
23 1 3
24 1 1 0 1I................s..0e *..*..*0.****4.
25 * o.cos 2.s 2000. 360.140 4 4* 09 $0 40 4446 46 04 4 00 9044 #4004 44* *e�*e . -*- . -- ��*eee� 0*
Figure 3.1-3. Listing of the SWIFT t Input Data.
3-12 IDRAP
I 2 3 & 5 6
lZ3456 Iao"214"1690113456T690 L234% 7490M,4561.I _.__.______ ___._____
21
2,IZ
10
11
32
31
3s
35
37
35
l1T40
41
4'
%4
As
49
SO
I4
IIIII
1*0*0v*00
O..*... * *
.0003
L I
0.Q . 17) * to1 *
ZS.0 2.0 0.0
0.0 0.0
II
* 01
I aI
I 0*. c
II 2 1I* 1£
I I
I I
III I 1
I
II
I I I
I
I 00001II I II
* a to
I I
0 _1 -1
1 _ _ .1I
* -1- -1
0
I
O. S
0 0 0 0 1 0
.-**,
,9999.0. e4*
Ireqo 12)4 So 7690
£31-3)-SLN I
IRIO-I I
Io.oRIO-2-1 I
I410-2-2 I
Ift1D-)-BLN 4
IR1O-'5-BLN I
II-I I
I22.4-2 1
Kaz-I I
I22-2 4
I22-4 1
Iaz-I I
1-R2-6-SLN I
I12-7-1 I
IRZ-?-Z 4
IRZ-7-3L14 I
I12-11 I
III
IZ-LI I£
az-I. 4I
az-i-i iI
22-7-2 1K
RZ-7-6LW II
0.5 0.00011I
022-13 9K
1 220 s
0 220.0*.4.44* 44
40400
1 -3
70.0 0.0
I
0.0001.
I 1
0 0
1 1*4 4* *4444
219.0
0.0
a 0
1 -3
J0.0
0.0
-I -1
0 0
0.0
0.0
0 0000
3.0
0
0.0
a 1 011-1 -I Oil
Figure 3.1-3 Concinued.
3-13 ORA.P
& z S s el34te6o09023456 Iq90ZZI,4567o901234.670l2 367IQ0Z3416aOt523,67sU3.S&7890
. ___ - - _- -_-- _ . _ . __*
I
52 I 1 I I1
S3 t .10 ;t .I^
5'. 1I -. *-
a
I
z.a*0Moo*
0 0
I -3
70.0
0 0
OC
SS * 0.10 0.0 0.0
56 t I I I I I I OltI *9* t
ST I 0 1 0 0 a 0 aI
so I I I I I I -3
591 *10 209.S TO.O
G.o
0 OOD
0
3.0
0
0.0#06*
*9 4*a .Oil
0.0I
60 b
61 1I
6Z tK
63 1I
64 1I
65 *I
6 6 KI
67 I68 1
leq I
t
371 1
1IIL I
I
I
75 I
0.0
- -1
0 1
1 1
.10
O.SO
-1 A
0 t
I I
.10
0.e0
-1 _
0 L
0.0
-1 -1
0 0
I I
203.9
O.0
_t _I
0 0
1 1
Iq.o
0.0
,I -I
a a
0.0
-1 -L
o 0
1 -3
0.0
-I *L
e a
1 -3
70.0
0.0
-I -I
0 a
01
a
0.0
a co
0
3.0
a
0.0
0 1 011
ItZ-1 t
K
Q2-7-1 K
OR2-13 t
R-7-2 1
17-7-BLN I
0.S 0.0001 1I
OR2-11 t
R2-* I
Q2-7-1 t
12-7-ZI
R2-?-OLN II
0.5 00001 I
QRZ-13 I
RZ-I
I17-7-2 *
R2-T-SLH I
0.S; 0.C001K
OR-X I
R2-l- .
0.0
0l
0
0.0
0 0000
a
3.0
0
0.0
0 1 lt
0.0
OII
0
0.o
0 0000
0
3.0 0.0
0 1 oil
*- *e *�*eeae * �
Figure 3.1-3 Continued.
3-L4 D/
t 2 1 4 5 & 7 aIZ *Sstt &l27f5nt90~l23*S**^§QlZ145b&t7011Z4b79(, LZ38~ 7690123'.507692671 *0
--__ __ . _*.- __ _ _ _ . 0 _____ ____ *... _ _ _
IrtX t L
loII I I 7 I .10 1
laI10 13*0**...*
19 1 1.010.... ** 40s**
50 * 1 1 1
SLI a L 0
*2 I L I 1
63 1 .10
05 * 1.25K
.0 i -1 -i -1
er J 0 1 a
16 I 101 1
qo
LQO.0
1 -3
70.0
e. la 0.o
1 I I
a 0 0
I 1 -3
.0 70.o
all
a
0.0
0 0000
a
3.0
0
0.0
a 1 0i
O.0
Le'. 0.0
00
-1
0
1
.79.5
U
S-
SS
q6
q?
9?
Co
IIIIIIIII
IIIIIIIII
*
1.s0
-1 -I
0 I
i I
.1o
1.75
-1 -1
0 1
1 L
0.0
-1 -1
o a
I I
0.0
-1 -1
a 0
1 1
0.0
-1 -I
o 0
I -3
70.0
0.0
-1 -1
a o
1 -3
I.0
0.0
-I -1
o 0
1 -3
011
0
0.0
0.0
0 CO
0
3.0
a
c.o
0 1 01
R2-7-1 £
RZ-7-1 1
0.5 0.00011
0R2-1 3
I
12-i I
R2-7-2 I
IRZ-7-91. I
oes catcall
I
OR2-131
R2-7-1 t
RZ-7-TLN I
0. 5 0.0001
I
ORZ-13 I
RZ-t-
RZ-7-Z Itt~~~t t~
IZ-7-SL ;
01
0
C .0
0 0000
0
3.0
a
0.0
0 1 all
0.
Oil
a
0.0
o 0OZO
0
3.0
a
c.o
0 1 oil
vie w s v
Ftgure 3.l-3 Concinued.
3-15 0?4p
1 2 3 1 S 6 1 eM456 We aIZS156 rqoz234567aqoLj aIsq a 2Xz3457gqO Z14341 9OaIZ45?aqot345676o9
,_ _ _ _, - ___.__. . . ._ *
I101 I
I
10) II
106 tI
109106 I
I
lto II
1
109 1
110I
iu II
LIZ Ii
12
113 II
12
1116 I
II
LIS? I
I
1
I
12 II
I
I
.10 171.0 70.0 0.0
2.00
-1 -I
0 1
I I
.10
2 .2S
-1 -1
a I
I I
.10
Z.50
-1 -1
a I
I I
.10
2.7S
-t -1
a I
I I
*t0
0.0
-I -I
a 0
I I
167.0
0.0
.. I -1
0 a
t I
164.0
0.0
-1 -1
o 0
1 1
11v.0
.0
1 I
15.0
0.0
-l -1
a 0
1 -3
70.0
0.0
-1 -1
0 0
1 -3
0.0
-1 -1
0 0
1 -3
10.0
0 0
L -3
70.0
0.1
0
0.0
0 0000
0
0.
01L
a
0.0
0 O0
0
3.0
a
3.0
0
3.0
0
1.0
0
c.0
o I il
0.0
a 1 01l
0.0
0 1 01
0.0
0 1 01
0.0
Itz-7*? 1
R2-7-2UE I
IOZZ-7-O t
I
I
tt-1 ;~~
12-7-2 t
ORZ-t-Z R2--OLN t
ORZ-13 1
INZ2-1 I
IK2-t-i I
RZ- T-2 t
I
O.S O.OCO1
ORZ-11 t
RZ-I
RZ-?-Bt 1
O.S 0.0001'
I
OR2-13 I
IAZ-I I
12-I I
tz-1-2
Oil
a
0.0
0 0000
'1
c.o
OIL
C
00
0 Coo
(
0.0
-- q
Figure 3.1-3 Continued.
-
3-16
OA/44j4 r
1 2 3 A S I I e
124S&75Vq12356t7q012 345 PI Z 12356 WgoL306799123 S 7aq801Z34567Iq90ZAS6790
t ' L 1t2-T-BLII X
C6 e ***I
;i 1 3.00 0.0 0.0 0.0 3.0 0.0 O.s 000LI
lz I I I 1 I I t £01). 0oJo 0 0 I011 Q2-13 1
X *4a- I
IZZQ 0 0 0 1 0 0 0 0 0 12--STOP !
II
._____ -_ *- a - - - -
Figure 3.1-3 Concinued.
3-17
Results. Figure 3.1-4 shows the simulated flow rates, which agree
reasonably well with the measured results. As indicated earlier, however, the
simulated values for times less than about 10-2 d were fitted to the measured
values by an optimization procedure for 4:he one-dimensional unit. In so
doing, two relations were found to be useful. They were a ime constant,
* - a SK' (3. i-1a)
and a ('ow-rate dependence,
Aq - /a (3.1-lOb)
both of which were inferred from approximate analytic expressions. They show
that the only two values, t 10 d and Aq 15 ft 3/d, which may be inferred
from the experimental data are functions of the three hydraulic
parameters, a, S, and W'. Thus, the inferred parameter set (Table 3.1-1)
is, unavoidably, onunique.
Discussion of the Code Output. Two tables included in the input echo
(see microfiche) may be of interest to the reader since this is the first use
of radial coordinates in this document. One, entitled "Radial e-id Block
Data," exhibits nodal points and grid-block boundaries. In this case these
radii have been internally generated (RI * in Card R-22) assuming an equal
ratio:
aril /r (3. 1-Ua)
for neighboring nodal points. That ratio is
An/2 r /r or A = 1.16597 (3.1-1lb)
where n X 50 (Card M-3-1), re ' RE 2000 ft (Card R-22) and r - RI -
1.0 ft (Card R-22). That all neighboring radii have this same constant rati
rty be verified by the reader.
The other table is exhibited under the title "Data for Carter-Tracy Wate
Influx Calculations". It shows the so-called terminal-race-influence functic
P I (Van Everdingan and Hurst, 19491 as a uncti:n of time. Implicit in this
101 I I I - -I
10'
a A#AbA 6/_ FLOW RATE
- X --- at *1 x w % s:1 0
. .-.4 6 in a -~~~~~~~~~~~e0
-1600
I.-,1M
o:
BL
10'!60 WATER LEVEL
LUWATER LEVE~L 6
- 500
10' -S. a
-Iw
-J
t1- tobIco0
10"-^ SWIFT* OBS, DATA aa
I 400
10'
I I I
104 10' l0' 10' 10
TIME (days)
Figure 3.1-4. Flov Rate atid Draudown of che H2A SIug Test Within the Magenta
Dolomite Formation.
3-L9
function is the assumption that the interior aquifer under detailed numerical
study for r < re (RAQ) may be analytically continued to infinity by an aquifer
having the prescribed Kb (KH) and Ob (PHIH) parameters. The nfilence
function denotes, in dimensionless form, the pressure drop at t aundary
r re between the two regions which results from a unit fow across this
boundary. Carter and Tracy 19601 used this function to achieve an efficient,
though approximate, relation
eW *W b p (3.1-L2)
It expresses the boundary flux e in terms of the pressure-change p at r re
s .ich occurs during a given time step. Quantities aw (P) and b tPOO
according to this tchnique, are themselves functions fo the terminal rate-
influcent unctior nd, of course, are implicit functions of time. Other
relations for P may be used to correspond to, say, a finite external
aquifer. The function presented in the output table, however, which
corresponds to an infinite aquifer, i the only one which is available
internally within SWIFT. The Theory and Implementation Document (Section 5.2)
provides additional information on this boundary condition.
Perhaps the o:- crucial tables of all, however, are those entitled "Well
Operation Summary," which are printed at selected time steps and which provide
the desired results. For example, from the first such table,
-4 3 Q(t-10 d) a 1.48 x 10 Ib/d - 23.7 ft /d (3.1-13a)
and from the et-and
Q(tlO d) a 165 Ib/d - 2.64 ft /d (3.1-3b)
both of which are consistent with the computed results of Figure 3.1-X. The
detailed results shown in this figure were actually obtained by post-
processing a plot file (Table 12) ox. which the ell-summary data was written
automatically for each time step.
3-20
3.1.4 Exercises
Level 1. Using the Input-Daca Guide, the Theory and Implementation and
the microfiche listing of the output, as required, complete the following
exercises:
Exercise I
Exercise 2
Exercise 3
rxercise 4
Exercise 5
From ables 3.1-1 and 31-2, what data are required for the
blanks labelled "1" through "8" in Figure 3.1-3? The data
called for here are the hydraulic properties of the global and
local units and may be verified by reference to the output
listing.
Explain the meaning of the control parameters identified on
Cards M-2, M-3-1 end M-3-2.
Explain the meaning of the welL-control parameters identified on
on Cards R2-6 and R2-7-1 and the physical parameter identified
on the second Card R2-7-2.
Examine the printer-control parameters identified on several R2-
13 Cards. Then check the output to see that they produce the
specified output tables at the times indicated on the
corresponding R2-12 Cards.
How would you obtain the well-summary table for intermediate
time values between 10-4 d and 10 d?
Level 2. The reader who wants to develop a deeper understanding of the
code should also consider the following exercise:
Exercise 6 Check the validity of Equatiors (3.1-10) for this application.
For example, ncrease the value of K' and run the simulation out
to about 10 I d. Are the flow rates increased and the duration
of the dual-porosity effect decreased in accordance with these
equat ions?
3-21
.
3-22
3-23
3.2 PROBLEH 5. ANALYSIS OF WELL-TEST DATA FOR THE MUSQUODOBQZT AREA,
NOVA SCOTIA (FINDER AND REDEHOEFT, 19681
3.2.1 Objective
e To show the feasibility of characterizing the delayed-yield effect it
pumping-test data from the Husquodoboit River Basin with a dual-
poLosity model and to illustrate the manner in which such a
simulation is carried out.
3.2.2 Descriprion of the Problem
Background. The setting for this problem is the usquodoboic River Basi
in Nova Scotia (Figure 3.2-1), and the aquifer o interest is the zone of
glacio-fluvial deposits shown in Figures 3.2-2 and 3.2-3. As a part of a
water-supply feasibility study for the city of usquodoboit Harbor, a pumping
test was performed. The configuration of the pumping well and the three
observation wells is shown in Figure 3.2-4, and the measured data are
presented i Figure 3.2-5.
Pinder and others (Finder and Bredehoeft, 1968, *f.d Finder and Frind,
19721, it. their numerical studies, used these data for calibration of a site
model, with which they predicted drawdowns over a 20-year time period (Figure
3.2-6). These investigations characterized the site by the two-dimensional
geometry and the transmissivities shown in Figure 3.2-5. A delayed yield,
which they observed in the measured data, was characterized by a temporally
varying storativity:
0.003 , t 0
S a (3x10 4)(10-t) + (6x10 3 )t, 0 ( t 10 min (3.2-1)
0.06 , t 10 min
Problem Statement. That this delayed yield ay aLso be characterized a
flow in a doubly porous media is the focus of this probLem. Here it is
assumed that the observed drawdowns derive from he distribution of fine sar
Oleto, 'A 0 * :: . Msusquodobolt River basin 1
*--- * *;8:sz35p>- Musquodobalt Harbour: *g 9 ~ Dartmouth
VC 0~~~~~~C
@ 50~~~~~~~~~s 0 so,
Scale in miles
Figure 3.2-1. Problem 5. Location Map of the Musquodoboit River Basin
(From Pinder and redehoeft 1968)).
3-25
Figure 3.2-2. Geologic Map of Musquodoboit Harbour Area, Nova Scotia
(Corrected From Pinder and redehoeft, 1966).
3-26
L E E DCENOZO C
PLEISTOCENE AND RECENT
El] RECENT ALLUVIUMW GLACIO-FLUVIAL DEPOSITS
PALEOZOIC AND PRECAMBRIANN
GRANITE AND SLATE 5
ScaleHorizontal: 1 in 948 ft.Vertical: 1 in a 130 ft.
AA
100 > M~~~~Fusquodoboit River a
>I~~~~~~~~~~~~~~~~ C@ _ltI'-5_ ) Vs s t94^ z;>6 >,^P^W~s b4L~ws~s > as ^ -
q 4LA '7it.4A.
;m.~~~~~~~~~~~~~Aj
-j jk y 4 AJ%~
Figure 32-3. Geologic Cross Section, tKsquodoboit Harbour Area, ova Scotia
* 491 Sol
SCtan @ESt
go I
3
Meiroi
S. lete2. I1gOIItl
a te~~IS 1, . t\
t t" e. T U.I fal \ls
I XI,* tel O.MU ft _ IO
v pum1l all ^ 0.913 Woas
WI ertt.* l, I t * It,. In - OeyIlnse wll .r * 300 ft.
ful~ter .11 a . X t*
Figure 3.2-4. Location of .- servation Wells and Characterization of
Hdraulic Propert ies.
10.
1.0
z
0
0.1 0
WE WEL
.01I0.1 1.0 10. 100. 1000. 10,000.
TIME (min)
Figure 3.2-5. Observed and Simulated ravdawas for ella 1, 2 and 3 Using a
Homogeneous Dual-Porosity Mode.
. . . . .
2A- -
DigitzI model rsAts(sefected tne stepsi
0
0
_el DO - - ,. _
LaLI
g
W., 0
OA4
. . .......
O1.101
.0 10.000 100,000
PUMPING PER100 mnutes)
1=1=00 1QOOQOOQ
Figure 3.2-6. Time-Oravdavn Curves Obtained From the Simulation of Pinder
and Bredehoeft 19681. I
3-30
and of coarse sands and gravels (denoted hereafter as gravels), that the value
0.06, determined by Pinder, represent the storativity of and the fine sands
and that the measured ransmissivi~ies represent the connected gravel lenses
within the aquifer. The delayed yield, hen, rather than coming from the
time-dependent storativity of Equation (3.2-1) is controLled by the
conductivity of the fine sands and, to a limited extent, the storativity of
the gravels.
tn order to demonstrate the delayed yield effects as simply as possible,
it is assumed that the drawdown cone is confined to Zone 1 (Figure 3.2-4) and,
correspondingly, that the time period of interest is restricted to times less
than 10,000 min, the time scale of the pumping test. For these limited space
and time regimes, an axisymmetric site model may be used.
Flow Equations. The equations depicting coupled flow in the coarse- and
fine-grained regimes are identical to those for coupled flow in the two
fractured regimes, which were employed in the previous problem (Equations
3.1-1 through 3.1-3). ere, however, the dual-porosity zone is present
throughout the system rather than confined to the immediate vicinity of the.
wellbore.
Input/Output Specifications. Data selected for this demonstration are
summarized in Table 32-1. As indicated in the discussion above, all of these
parameters are taken from the work of Pinder, with two exceptions. The
hydraulic conductivity of the ine sands and the storativity of the gravels
were obtained by calibration against the drawdown curve for Well Number
(Figure 3.2-5) for times t 10,000 min. Comparison of measured and
calculated drawdowns is the desiAd result.
3.2.3 Numerical Simulation
Discussion of the Code Input. Although the two porosities here refer to
two different porous-media components, rather than the two different
fractured-media components of the last problem, the code inputs (Figures 3.1-1
and 3.2-7) are quite similar in some respects. Both problems use radial
coordinates with a Carter-Tracy aquifer-influence function (Cards R-27
through R-33) affixed to the outer boundary. There are differences,
however. The well produces in this cse (positive value of Q Card R2-5) itt
a rate control (IINDWI - in Card R2-7-l) rather than a pressure control.
3-31
T&IMe 3.2-1. ProbLem 5. Input Specifications.
Parameter Symbol Value
Gravel s
Transissivity T 0.274 ft 2 /s
Storstivity S 1.95 x 10-5
Fine Sands
Conductivity K' 4.63 x a-7 ft/s
Storativity so 0.012 ft 1
Thickness a 2.5 ft
Location of Observation WelLs
Well I r1 100 ft
Well 2 r2 212 ft
Well 3 r 3 300 ft
3-32
"/?*44
1 2 i # 5 6 I a1Z345780913456?901234567901234578lO1Z34S69012345678901Z34S67S790123S6?690
, * * . . * -- -- -e-- - * _ __
I
2
3
4
S
6
7
a
9
za
11
2a
13
14
15
1 7
la
19
10
z1
Z2
13
24
Zs
IIFRCLEM 140. 5 SWIFT It EXAPLE ENGLISNIpkliE4RErAll0N Of ELL TST DATA IN USCIt L 0 0 0 0 0 0 0II 40 1 1 3 0 t 1L
* a . 1 .I
I ZS
I 1 40 1 1 1 I 1 a
j C++ .. +. C .. ,.+
4. (1) '4 (23 * 0.0 I.0
I 0.0 0.0 0.0 0.0
I t40.0 t.0 66.0 62.4It 0 1 1 2
1 68.0 1.0 65.0 to0
I 60.0 1.0
II 80. 0 1.0....-I . e*¢s' RE1 0.0 69.0 . U'... RE~I .....1 1.0 68.0II 0 01* |4.0 0.0 0.0 0.0I *004*60**e1 0.37S 0.0 1200.0 0.0I * . . e e .+ ¢¢* 4 C C e C 4 4 * . . * . . . . .
to t31 t41 e* (51 .. 4 4*4
I 0.4$7 0.6640 1.0Z 1.S06
I 7.37 1o.977 16.130 Z4.29s
* 100.0 122.4 144.8 167.2
UNTTS, 12/83,
U0OOIOIT AREA. OWA SC0
0
0 0 O
1.0
0.0
2 *4
0.0
rIA N-t-z
o on-3-1, t
RO'-3-ELN I
Rl-7
1-9
Ql-11
it-it0 L-1
Rt-LZ
RI-L6~~~
I1t-22
RI-Z3
4.1909RI-3S
T6.t6711-1
2Z9061tI-zs
EFERENCE, IN TEXTEFEREFMCE0 IN EXERICISE .
0.0
2,141
36.6145
169V6
3.334
S3.773
Z12.0
9----
Figure 3.2-7. Listing for the SWIFT L Xnput Data.
3-33 ORAvr
1 z 3 4 S b 7 a
1234967sq Z123456789OtZ34567S10123Ss67890 3I e456?0 Q1234567T09tZ345fSqO 23ms78904 .______.______.__ _ .__ __. ____. . - -- ._____.________,
26
Z7
2q
29
30
it
32
33
34%
is
36
I?
3 8
3q
40
41
42
43s
44
46
47
As49
so
1IIIIII
Z47Z
415. 1,0
794.363
26 4.6
456 * 285
670&6?6
Z62.4
502.284
qsko^tl
00.0
So.50
104S.629
I I O
t 2.361E4 15.0 1200.0 360.0
I*
* 0.0 0.0. M61 .. 19**
I 99944** 44440 94
1 1 1 110* * I.OOE-4 0.0
ItII
* a 0 0
* 0.0I
I Is 1 I I 0I 0
3 2S I I Z
1 1 0.SI
1 0.0 0.0 0.0 *.3ZE04
* 1 1S 1 t t I
1 1000.0 1.0 .650 0.0
I
1 2 20 1 I 1 II1 1000.0 1.0 68.0 0.0II 3 25 1 I. 1 II* 1000.0 1.0 o.0 0.0t
324.046
603.36t
1146.2317
0.0
345.0092
fi61.289
0.0
l11.52R1-2s
7z4.7791-25
R1-25
Rt-26-BLN
1t-27
R1-29
RI-31
Rt -33-SLN
110-1
0.ORID-1-1
t 10-2-2
* 10-3-BLh
R 10-5-LN
1-1
R1A-Z
RZ-1
RZ-Z
R-4
RZ-S
Q 2-7-1
Q2-7-2
k2-7-1
RZ-7-Z
R2-7-L
R2-7-2
a
*- -. …--_ -- __-_ __--_ _ _0 -- &.
Figure 3.2-7 Concinued.
3-34 _
I 2 3 4 5 6 7 e124s5A$s901234srg40qt 1s56719023s&70$901Z3476?Q s1234S689012345s7s901234S67890
* * _ _ _ _,_ _ _ _ _._ _ _ _ _._ _ _._ _ _ _e__ _ _ _ _._ _ _ _ _._ _,
451 1 4 I 1 I
5Z I 1000.0 1.0I
53 1I
54 1 6oq.44-4k 1.3189E-4
5S * L I * I
57 1 .9444E-3 1*3699E-)
58 1 1 L 1 1.
591 0 0 0 01
60 6.q444E-Z 0.0
61 1 1 1 I, I
62 1 a 0 0 0I
b4 I I 1 1
65. 0 0 a 0I96 gf,q44k 0 a.o
4? 1 1 1 1 I
66 1 0 0 0 0I
69 I6.i444 1 c.0I
ToI
71 I 0 0 0 I
I I
66.0 0.0
RZ-7-L I
R2-7-OLN IfII
1 1 1L '5 0000 a 00 -1 1I OORZ-13
0 0 va 0 0 t2-LIIIII
I 111 0 CO a 00 -S1 ll acZ-II I
.45 * 2S .2O 9. 69444E-2 wool*
1 I I 11 0 0000 0 00 -1 11 OORZ-13 I
o 0 a a 0 RZ-I I
.9S .25 .20 9. 6944'E-Z .001
1 I1 LI a 000 a 00 -I 11 OORZ-ts I
o 0 0 0 0 R- I
.QS .2S .o 9. 69444 *011
. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I
I I 111 0 0000 0 00 -L III OOR2-13 I
I
0 0 a a o K2-1.
I
.95 *2S .20 9. 6.944 .*0011
I
I I III 0 0000 0 00 -. I11 OROZ-3
0 0 0 0 0 QZ-t-ST I
* - -- f - -ain *, -, .- * - - -.-. -. *-
Figure 3.2-7 Continued.
3-35
Further, the nodal radii are specified here in Cards R-25 (control parameter
RI 0.0 in Card R-22) instead of being internally generated. Finally, local
one-dimensional units are here imbedded in each global grid block, as
indicated by the parameters 11B - 40 and IFD - 0 in Card ROD-3.
Results. Figure 3.2-d shows the end result of the calibration of
simulated to field data. For this purpose th- two curves labelled "Lower
bound" and "upper bound" were quite helpful. Both are single-porosity Theis
solutions which assume that the gravels provide all of the transmissivity of
Lhe aquifer. The lower bound uses the combined storativity of the fine sands
and the gravels, whereas the upper bound uses only the storativity of the
gravels. Thus, the system resonse initially coincides with that of the upper
bound but bridges over to the lower bound at a rate controlled by the
magnitude of the conductivity assumed for the fine sands. Figure 3.2-5 shows
calculated and observed drawdowns for ells 2 and 3, as well as for Well .
It would appear that in the vicinity of Well 2, the gravels are a bit more
dominant than for Welt 1, whereas, in the vicinity of WelL 3, the fine sands
are somewhat more dominant. Both Figure 3.2-5 and Figure 3.2-8 demonstrate
the feasibility of using dual porosity to chattcterize the delayed-yield
phenomenna .
Discussion of the Code Output. The latter also indicates that drawdowns
(pressures) within the fine-sands come to equilibrium with those in the
gravels after about 100 in of pumping. After that the fine- and coarse-
grained materials behave asha single-porosity media. This effect may be seen
in the output tables (see . -ofichte) which, in this case, are inappropriately
labelled as "Pressure Within tb:. Rock Matrix." Take, for example, the fine-
sand surrounding Well I (Btozk 5). At t - I mn (6.9 x 104 d) he pressure
drop across the block isAI.0 psi; at t 10 min (6.9 x LO- 3 d) the pressure
drop is 0.1 psi; and at t 100 min (6.9 x 10 4 d) this drop becomes negligible
at 0.01 psi. Because of the choice of an initial pressure p 0, the
pressures may be converted to drawdown using the relation
s - - p/0.433 psi/ft (3.2-2)
in order to spot-check the results shown in Figures 3.2-5 or 3.2-8.
10.
1.0 I v
z
0
0.1
LWER BOUND *FIELD DATA
-. SWIFT
0.1 1.0 10. 100. 1000. 10;000.TIME (min)
Figure 3.2-8. Simulation of the Observed Drawdowns for Well I Using a
Homogneous Dual-Porosity Model. The upper bound is a
singl*.-porosity simulation using the gravel storativity only.
The lower bound i a sing1-porosity simulation using the
r-torativity of both gravel and sand.
3-37
3.2.4 Exercise*
Level 1. Using the Input-Data Guide, the Theory and Implementation and
the microfiche isting of the output, as required, complete the following
exercises:
Exercise I
Exercise 2
From Table 3.2-1, what data are required for the blanks labelled
"I" through "10" An Figure 3.2-77 The data called for here are
the hydraulic properties of the gravel and fine sand. They may
be verified by reference to the output Listing.
In what blocks are the observation wells located? (int:
examine the R2-5 and R2-7 Cards). These wells correspond to
Wells 1, 2 and 3 and their function is simply that of
convenience. The appropriate grid-block pressures are placed
withii he "ell Operation Summary" so that they may be easily
tabulated and plotted. The reader is referred to the output
listing, where he may examine several of these summaries.
Level 2. The reader who wants to develop a deeper understanding of the
code should also consider the following exercise:
Exercise 3 Figure 3.2-8 shows the behavior of the drawdown to lie
intermediate between the two singLe-porosity curves labelled
"upper bound" and lower bound". f one increases the
conductivity of the fine sand, will it move the simulated
drawdown in the direction of the upper bound? Check your
prediction by rerunning the problem.
3-38
-
4-1
4 FLOW AND TRANSPORT THROUGH AN AQUIFEF WITH CONFINING lAYERS
4-2
4-3
4 1 PROBLEM 6. DRAWDOWN FROM A FULLY PENETRATING WELL IN A LEAKY AQUIFER
(HANTUSH, 19601
4.1.1 Objective
* To illustrate the use of the one-dimensional local units to simulate
the effects of a leaking aquitard.
4.l.2 Description of the Problem
Problem Statement. A well fully penetrates an infinit. aquifer and is
pumped at a constant rate. The aquifer is bounded from below by an
impermeable bed and from above by a confining bed or quitard. The latter
influences the aquifer to a moderate degree since it is weakly conductive and
contains some fluid storage. A schematic drawing is shown as Figure 4.1-1.
Flow Equations. There are two flow equations here, one for the aquifer
and one for the aquitard. The former is given by
T r(r a - brw - S -(4.I-1
and the latter, which assumes perp icular flow in the aquitard, by
K' ' s'i* (4.2-r)3 2 W st
Coupling arises through the flow at the interface between aquifer and
aquitard:
r a K' a -z 0) (4.1-3)
where A- 1/b is the interface area per unit of aquifer volume.
?.4.: N* '.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~4
im **
Figure 4.1-1. Problem 6. Scheatic Diagram of a FuLty Penetrating
Constant-Discharge etL n a Leaky Aquifer.
4-5
Initial/Boundary Conditions. For the aquifer the initial drawdown is
zero:
*(r~t-O) * 0 r O0 (4.1-4s)
and the discharge at the origin Is:
lim (r 1) 2 r*O dr T
t >0 (4.1-4b)
For the aquitard the initial drawdown is also zero:
8'(r~z,t-0) 0
and the interface condit ion is
I r > , -b' z O (4. 1-5e)
slr,zwO,t) - r,t) r > O , t > (4. 1-5b)
where z is -astired positive downward from the auifer-quitard interface. At
the upper boundary of the aquitard, there is assumed to be leakage from an
overlying aquifer so that
s'(r,2t-W.0 -t ' r > O. t > 0 (4. 1-5c)
Input/Output Specifications. able 4.1-1 provides the input data, which
is taken directly from the revised Benchmark Problem 3.2 (Ward, 1984bl.
SWIFT, of course, uses a pressure-based formulation rather than the drawdown-
bdsed formulation assumed by the benchmark-problem specification, with the
transformation between the two given by Equations (3.1-6) and (3.1-7). Table
4.1-1 gives the values appropriate for both formulations. In addition, these
data are supplemented by a reasonable value of the wel radius. The output
specification, also taken from Benchmark Problem 3.2, calls for the drawdown-
versus-time profile at a radius of 117.4 from the withdrawal well for a timi
period extending out to t 106 s.
4-6
Table 4.1-1. Problem 6. Input Specificatior.i.
Parameter Symbol Value
Aquifer Trmnsmissivity T 0.05 n2/sec
Aquifer Hydraulic Conductivity K 5.00 x 0 3 n/sec
Aquifer Thickness b 0 M
Porus ity 0.10203
Water Density 1000 kg/m3
Water Compressibility cw 0
Rock Compressibility CR 5.00 x 10 7 WI
Aquifer Sorativicy S 0.005
Aquitard Specific 3torativity St 0.0016 1S
WeLL Radius rw 0.1143 m
Aquitard Hydraulic Conductivity KI' 10 5 /sec
Aquitard Thickness L' 50 m
Aquitard Porosity 0' 0.3265
Well Pumping Rate Q 6.283 3/sec
4.1.3 Numerical Simulation
Discussion of Code Input. Up to this point, one-dimensional local unit
conceptually have been placed interior to the glo~tl grid blocks in order to
simulate dual-porosity effects. Further, the units were assumed to provide
storage only with a no-flov condition assigned to one of its sides. Here, t
role of these units is expanded in that they are plaree5 external to the glob
aquifer block in order to imulate an aquitard. In addiion, they are given
zero drawdown (Equation 4.1-5c) at the external end i order to simulate
leakage from an overlying aquifer. figure 4.1-2 typifies the numerical
gridding by shoving both the global aquifer blo:k, and the one-dimensional
units of the aquitard, which are themselves subdivided into grid blocks.
Figure 4.1-3 contains the necessary code input for the ocal units. t
control parameter
IFD a -3 , ROD-3 (4.L-6a)
effects the external placement f local units for all 50 grid blocks. Cont
paifameter
KBC * l , RID-3-l (4.1-6b)
chess activates the external constant-pressure boundary condition, which is
specified as
PBD - O.C R1D-3-2 (4.1-6c)
Figure 4.1-3 &lso gives the outer radius of the simulated region:
r X RE - 2647 RI-22
The selection of this radius, just as for the two previous problems was a
matter of judgement. Hovever, the appropriate characteristic radius
r * (Tb'/K1)2 X 500 (4.1-7)
4-9 tR,4r
I 2 a I 0 gzZ3s s 7 z9012344 s190oZ3 st679OIZs34S79023 s 7690123s o 6s901Z25m n023s6o7890
.____ __.__ __ _____ ____ . _ _ *_. _ _ _ _
I1 IPROSLE NO. 4. INIFT I EAPLE. St UNIT$* 12/83I
2 IFULLY *ENETI&TINC ELL IN A LEAKY AOUIFER
3 s 1 0 0 0 1 0 I 0
I I 5C ItzG
I
I I **' RE~EREKCEO
I Z
S 1 * 50 1 1 1 1 1. -3
o ' "-2-I
Pt-t-t
A-2
t a I a 04-3-
I u rexr IN EXERCtSE * ROD-I
100-2
A00-3I
9 II
10 t1
1 1 1I
12 11
1 3 I
14 II
15 *
I 1
1 7 II
I
I1
0.0. I11 *
to0 1.0
Z000.0 S,OE5
2 2 2 2
20.0 0.001
0.20 0001
100.0 0.001
L0O.0 0.001
0.0 20.0
LO.0 20.0
0.0
1.0
20.0
ZO.0
0580
.00.0
200.0
1.0
1.0
1000.O
1.0
1.0
1000.0
1.0
ROO-S-dLN
a11
1.021-2
RI-I
RI-m
aRI-i
I-8
1 1-9
11-10
0.001
0.001
0.001
0.00t
IL-tl
Rl-11
RI-lta a
tO.0 5.ES 0.0 0.0I t*0*0000
22 1 0.1143 01263642 2046.7663 0.0I4.*4,..4....4.4*.*.O++*SW***99¢¢¢*+
23 to 121 *' (1) .. f4) *- £5 *
24 II
254 I 0t
0.0
RtL-L
RI-2Z
l-23
Q1-26-BLN
Rl-27
,______._______._______._____ _ + _ _ _ _ _ _ _,_ _ _ _ _ _ _ _._ _ _ _._ _ -_-__-
Figure 4.1-3. Listing of SWIFT 1 Input Data.
4-10
OPrI 1 3 4 5 6 r e
121496RqIIU456 759O*l145&7I90LZ)45618q01234561690 L23 s5lg9olz31s67NolzL34s790: - ea * - *~. .-- : ** ---
16 i 1 0 at
27 I 5.CE-2 1.0io304 Zb4*.Y&63 360.0I
20 1
to 1. Stl *1**vv******. *****.**.-.*-*'.*
30 * 0.0 0.0. top .. 481 *I ,*4 **v¢+++4**.499*,*4....*............ *4v¢
31 1 2 to 191 * 0.50 .03 *1 1 *.* *44 a
32 1 1 1 0 0 Q
i c-19
t t-3 1
41-3 3-SUE
t10-1
0.01t-2-1
*10-Z-2
I
IIIII
II
0.0 0.0
110-3-1 1
13
34
35
16
31
3,
39
".o
41
42
A4)
A4
4S
46
47
46
49
50
I
I
II
I
IIIIII
0.0944040964 o.0 00
0 0 0
0.0
I I 0 0 0 0 0 0 a 0
I A
*0* (101 *
I I .II ]O15IJ 2 3S11 ISGEIOI* 3 isII L.OEIOI1 4 35II .OEL0III
to * 1.
0.0
I I
0.0
1 1
0.0I I
coo
0.0
L I
0.0
0.0
L I
M0.0
I 1
ZO.0
L I
20 0
I 1
20.0
12.0-3-2
RtLD-3-81.k
*L0-S-OLN
1-I.
RAt-'RZ-1
*2-4
l11-s
11-7-1
UtZ-?-2
12-7-1
Rz-7-z
2.1-7-1
*1-7-1
*2- 7-BL
0.0
0.0
0.0
0.0
1.0
I*� y--s *
Figure 4.1-3 Continued.
4-11 D/ ?Ar
I I 3 * 5 U I a
173'$61t1IOLZ345#6 ?Aq01234s4q it Set5*79V9IZ34567S80113456 719OZ 4167sq0tZ1I5&7490,_ --------- ____,___.____ ,_____ _ -
ISt I
I5Z t
53 II
54 1I
5 9
5 I
I57
I59 tSq I
I60;
161 1
I*2 1
I83 1
I
8q £
"S 9
I80 1
I87 I
I68 I
I49I
I10 a
I71 1
I12 I
I73 I
I
IIs
I
l.OE3 0.0
I L I I
o o 0 0L.0e) 0.0
a 0 0 0
l-OES 0.0
I I I I
0 0 0 0190ES 0.0
I 1 I 1.oDE6 00°
la 0 0 1
2 O SERVAT N
0.0 Z.0cr
0.0 0.0
0.*0E3 4063sES
I.OOE3 .532E5
Z.oOE3 f.230ES
4.0OE3 3.S99ES
10.OE3 3.w4WE5
20.0E3 1. I0ES
-I I
a 0
0.0
- I I
o 0
0.0
-1 1
0 0
O.
-1 1
o a
0.0
-1 I
0 0
WELL (I
i.%.EZ
0.0
UIt 0
o 0
0.0
oil 0
0 a
0.0
01 0
0 0
0.0
oil 0
0 0
0.0
oil 0
a 0
117.4 f
3.0ES
0.0
0000 0 00 -I i11 0012-13 1
0~~~~~~~~~~~~~~~~~~~~~~0 0 t-1 t
0 AZ-I~~~~~~~~~~~~~~~~~
IO.E 0.0 L-OE) 0O SEII
0000 0 00 -1 C11 00&2-I) I
ORO Z-1 +
o o R2-1
I.OEA 0.0 l*OES .1(41
CO0 0 cO -I oil 0012-13 I
0 0 KZ-t 1
1.oEi 0.0 l-OEb 0.2E61
0000 0 cO -1 01I COR2-13 I0 0 * -
o 0 *L-T-EUO I
S I44L L r I NE P OZt
5.OES 0.0 0.0f-s-1
'-3.2
F-4
IIIIIIII
I
'P-4
P-,
P-4 II
V-4 1
F_4 t
to-4-EmO II
F-?I
-100.
G&SCAVATION4 WELL 11 117.4 nI SMALL ANO LAVRIE TIlE
* 0 i - 9 i n i n 9 i n
Figure 4.1-3 Continued.
4-12
rZ 3 S 1 7 e
zZ45aJat0Z23As&7e023sA70Z 3is&7ata~l,5&730o1Zis&7e01l35s7esolZ34s&7eso
I Itl I 00 7.0s *O I.PS 1.ZES o.0 0.O-3-i 1
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I77I 0.0 0.0 0.0 00 P-3-i t
I I
81 3.o*E4 3.IO4E5 P-i II I
1c I 4*,< l.6SES p-i It t
60 * 0.aofi Z. )3IES -
*1 1 t0.OE Z.31iI4 P- k II I
90 I IS.OS Z.096ES P-i II
63 1 ZC.KC . .975E5 P-i I 1
S I lO.OEi t.116E$ r-i SKI
05 * S0.OES 1.b)?E5 P-iI l-
86 I - ICO P-I-ENDl I
eJ 1 i 05SERTIr0U SELL (ft * l17.i NI LAICE ritE P-2 1I
65 I 0.0 I.0b Z.0OE 1.7E5 Z.2ES 0.0 0.OP-3-i tI I
jq 1 0.0 0.0 0.0 0.0 P-3-Z tI I
qo * 1.5015 Z.09&ES P-i I t
qL I i.OOEI 1,91615 *-i I , I
Cl I 9.015 1.86315 P-4 I1 3I -IOG. P-i-END II t
9S * P-EN0 I I~~~~~~~~~~~~~~~~~~~~~~~~~~~~
*- ~ ~ f e ._____i ,_ _______.._.__. _* e___.
Figure 4.1-3 Continued.
4-13
eid provide sone guidance. This quantity, which is derived from dimensional
analysis, properly accounts for the fact that the extent of the drawdown cone
is controlled by leakage from the neighboring aquifer. It, therefore, differs
from the characteristic radius used for the two previous problems (Equation
3.l-9a).
Result. The numerical resuLts are displayed graphically in Figure
4.1-4. As shown, there are two analytic solutions, both of which are taken
from Hantush 19601. On2 is for small values of time t < (b')2S/ICG'
4x104s), and the other is for large values of time (t > 2(b') 2 S'( -
8x105s). The comparison between numerical and analytical results is
apparently quite reasonable.
Discussion of Code Output. Thia output illustrates the generation of
SWIFT printer plots*, a facility which provides a quick, inexpensive way to
display computed results. Since the code only plots data pertaining to wells,
observation wells are defined in Cards R2-4 through R2-7. Further, since only
one plot per well is possible, three such wells are completed in BLock 35 at
the specified adius r 117.4 m. The resulting plots then display the timea
periods
0 t 2xlO s 0 C t 4x10i & and 0 t C 10 (4.1-8)
with different scales of resolution, which permits a detailed comparison with
the Hantush analytic solutions.
Plotting is activated by a positive value of the plotting key KPLP - ,
(Card M-2), whose value appears in the initial table of the output listing
(see microfiche). Note, in the three plots, which appear at the end of the
execution, that, due to the expanded scales of the ordinate, the difference
between simulated () and analytic () is readily appar3nt. Exercise 4 deals
with this relatively small (less than 0.4 percent al 106s) discrepancy.
* The term "plot" should not be confused with :he term map". Maps onsist
of two-eimensional contours of pressure, isopleths of concentratic.. and
isotherms of temperature. In contrast to plots, the mapping facility is,
in no way, coupled to the resence of wells.
ol- 0, LARGE TIME
E-X SOLUTION
z
0
10S SWIFT
0- ANALYTICALSOLUTION
S
10 10'1O 10' 101 10°TIME (SEC)
Figure 4.1-4. Graphical Comparison of the Numerical Results from SWIFT and
the Analytical Solutions.of Hantush for a Radial Distance
of 117.4 no
4-15
4.1.4 Exercises
Level 1. Using he Input-Data Guide, the Theory and Implementation and
the microfiche listing of the output, as required, complete the following
exercises.
Exercise 1
Exercise 2
Exercise 3
From Table 4.1-1, what data are required for the blanks labelled
"I" through "10" in Figure 4.1-3?
Consider Card ROD-3. ow would one simulate two confining beds
of the same thickness? How would one include differing
thicknesses?
Refer in the microfiche to the table f pressures within the
aquitard for radius r - 117.4 m, Block 35, and time c - 04 o
Convert to drawdown via s (p-p )/P(g/lg) ane plot. The
initial pressure, as given by PINIT in the table entitled
"Initial Conditions" is po 5x105 Pa*. Nodal positions are
given in the table "Specific Matrix Discretization Parameters"
located in the input echo. Has the aquitard reached steady
state? Repeat the plot for time t 106 s. Has the aquitard
now reached steady state?
Level 2. The reader who wants to develop a deeper understanding of the
code should consider the following exer'ise.
Exer. se 4 In the aquitard the local mesh was generated using single-mesh
generation (KGRD 1, RID-2), but was that the best option
available? In the local mesh, Node 20 is attached to aquifer
* Since SWIFT plotting routines do not permit negative pressures, the
initial pressure p which her i rbitrary. was set sufficiently
high in Card R-16 (see also -ed R-3) that no negative pressures
would be encountered.
4-16
grid and at Node 1 a constant pressure boundary was
prescribed. From the output table "Specific Matrix
DiscretizatLon Parameters" (see microfiche), near the aquifer
the grid increment is 0.5 (input parni.-eter DSD, Card RD-2)
whereas between Nodes I and 2 the increment is 7.33 . The grid
seems quite lopsided in that so many nodes are situated closest
to the aquifer. The reader shouLd try to improve upon this grid
either by using double-mesh generation or by specifying the
individual increments directly. One would not expect any
difference in the short-term response. The long-term response,
however, may be affected, but to what degree?
4-17
4.2 PROBLEM 7. HEAT TRANSPORT DURING FLUID INJECTION AVDONIN, 9641
4.2.1 Objective
O To illustrate the convective heat transport hich results from
injection into an aquifer with heat losses to the confiing beds.
4.2.2 Description of the Problem
Problem Statement. An incompressible fluid of temperature, T, is
injected into a confined aquifer of temperature, To, t-ough a fully
penetrating well (Figure 4.2-1). Both thermal convection and thermal
conduction occur within the aquifer, and thermal conduction is operative
within the confining layers, denoted herein as the over/underburden.
Transport Equations. For the aquifer, the transport equation is
-vpcp r +, r ar (r -) r pc - (4.2-1)
and, for the over/underburden, the transport equation is
Ks 32L + r P9_ 3't (4.2-2)M aZ2 H P
Here the over/underburden coupling term is given by
r a -2/b)K 3T (4.2-3)
and the fluid velocity by
v a Q2wrbP0 (4.2-4)
Initial/Boundary Conditions. The conditicns imposed on the solutions of
these transport equations are, for the anuifer,
T(rt0O) T , r ) 0 (4.2-Sa)
T(rO,t) TI t > t4.2-Sb)
* 1a~~~r...
* :..
- ~ ~ ~ .~~~ WER COFINING BE8K' PR P PR
Figure 4.2I1. Problem 7. Rdist eat Transeort Within an Aukfer with Losses
to the over/Udx L 1dn
4-19
and, for the over/underburden,
T Mxxt-0) T , z 0 (4.2-6a)
T'tX.-t-,t) * T t i. (4.2-6b)
T'(x,*O,t) T(x) t > 0 (4.2-6c)
and
T'(xz-bt) - . , t 0 (4.2-6d)
where is measured positive downward from the lover aquifer boundary.
Input/Output Specifications. Table 4.2-1 displays input specifications
taken directLy from the enchmarking Problem 5.1 [Ross et al, 19821. It also
contains supplementary data necessary for SWIFT input. Aside from the well
radius, for which a reasonable value is choser, the composite heat capacity i
partitioned as follows:
PmpMPC pc + O(1-)pRcpR/ (4.2-7)
Given the composite values (subscript, ) in the benchmark specifications, ti
values for the rock density and the rock heat capacities were chosen to be
consistent with this relation. The output specifications, also taken from
Benchmark Problem 5.1, calls for the thermal breakthrough at a radius of 37.
m from the injection well for a time period extending out to t 1O s.
4.2.3 Numerical Simulation
Discussion of Code Input. Similar considerations arise both within th
aquifer and within the over/underburden. For computer efficiency, it is
desirable to minimize the outer extent of the simulated system but, yet, at
the same time, to adequately minic the infinite physical system. In additi
and for the same reason, it is desirable use as coarse a mesh as possibl
in the spatial and temporal regimes and stll to adequ..zly characterize th
thermal behavior at a radius r - 37.5 m frow the injection well. Within h
aquifer the controlling mechanism is the flow since i provides the convect
for the heat transport. Setting up the proper fluid velocity is no probleff
4-20
Table 4.2-1. Problem 7. Input Specifications.
Parameter Symbol Value
Injection Rate
Injection Temperature
Initial Temp.rature
Over/Urderburden
Thermal Conductivity
Density
Heat Capacity, Composite
Poros ity
Aqui fer
Thermal Conduct ivity
Density
Heat Capacity, Composite
Thickness
Porosity
Heat Capacity, ater
Heat Capacity, Rock
Well Radius
QTI
T0
10 kg/s
160 C
170 *C
K'
P'
PMC'
20 /(M C)
2500 kg/in3
1000 J/(kg 'C)
0.2
Km
PM
cpM
b
4cp
CR
rw
20 W/(m CN
2500 kg/m3
1000 J/(kg *C)
100
0.2
4185 J/(kg 'C)
2.079 x 106 J/(m3 C)
0.766
4-21
since, according to Equation 4.2-4), the flow is at steady state. This is
done by iposing a constant pressure at the external radius ., Card Rl-26-2 of
Figure 4.2-2 (see boxed value).
This external radius, of cou:se, ust be sufficiently large to contain
the movement of the thermal front which results from this flow field. Here
the retarded velocity is
v vocp/Dcal a v /r (4.2-8a)
-he., from Equations (4.2-4) and Table 4.2-1,
v Qc 2rbu c 1.332 x 104 /s (4.2-8b)o p iuPM
and the convective movement of the thermal front in t l09 s is
r * (v t)/'2 370 (4.2-9a)
To allow for a conductive smearing of this front and to include a slight
safety factor, the external radius is chosen to be
r * 1000 m , RI-22 (4.2-9b)
Gridding criteria are unknown in this case, and consequently it is necessary
to use a trial-and-error procedure for the spatial, and temporal, regimes.
Within the over/underburden the only transport mechanism of significance
is thermal conduction. The characteristic length corresponding to the time t
9 is
- (K-t/p c/. 90 n (4.2-lOa)
and the value
a a 300 m , RID-2-2 (4.2-lob)
is adorte; for the simulated thickness of the over/underburden. Here the
sp3tiol gciiding is chosen to adequately resolve the expected behavior within
4-22
L
3
4
6
1
I
2 Z I 4 S 6 7 el?3I 5S6 180 I23 56?Iq0IZ345A7 8901t* 1q 6 qO 3 S07? 01ZqO sq12)4 6 7 Q *3 ,5671 0123 45t,7 6q0
t I1PkOSLER NO. 7, SWIFT II EXA1PPLE S UNITS. 12t/t 111-3-I II IIA&VONIt RAllL HEAT TRhNSOORT WITH LOSS ro CNFIRINC EOS M-1- II***** 00060 I
I -Z 0 0 a a I 0 1 0 .1-2 1
1 30 1 1 3 0 1 I 0 a I 0 0 a ON-3-i tI ..... I* O I 1 I .1-3-2 4
S II 0 ROD-I tt I
QO-2 1
I 1 30 I I I 1 1*311 ROD-3 II 4*.-.
9 'I . .. . . . .. . . ** 9 ** .4+* .+* 4* * 99 * *.4..
10 * 3.ZOE-IS 0.0 0.0. 121 * t31
5 *******.**4**********¢****4*.,*,**+*-+*.*.*....*.*It Ia, £43 go (SI go (63 O too 0.0
12 1 2875.0 S.OOE 06 1,60.0 t71 .. 38l -
13 1 0 I 1 2I
14 1 160.0 0.001 160.0 0.001I
15 * 170.0 0.001I .... P
16 t 170.0 0.001 REFERENCII *-* .* ... * . a4 *a REFERE4CI
17 1 p.o4, M17C Oo 49. *...........
18 I 100000.0. 3101 *
19 I 0 o1
*0 * 170. 0.0 0.0 0.0
It I .09O223 .76559Z2 1000. 0.0I~ ~ ~ ,+40*0**
22 1 100.0 9.0E-6 9.3E-6 0.20
ZsIZ3 1
24 I 4 0*
ZS 30 30 I I I I 0
. . ....
:E IN ED IN E,... .. .
R00-3-LN
Rl-l
O.O I %OOE-SCR0L-X
RI-?
Rl-7
RI-9
.......... al-tEXERCISE
R t-11
R I-L6
A1-22
R1-23
RI-Z6-BLN
41-27
Rl-Z-I
,_ _ _ _ _._ _ _ _ ._____4__4______ ~ ~ ~ ~ ~ ~ - -- -_ _ -_ _ _ * _ * - -.
Figure 4.2-2. Listing of SWIFT 11 Input Data.
4-23
0;"4,jtr
2b
2?Ia
29
30
31
32
33
34
35
3b
1 ?
la
39
40
4 1
I2
41
4 4
45
4 6
47
A 6
49
SQ
i 3 23 4 s 6 L21i67S601214Sb ?89OL3234'5j790j11 Gq3233467TlqoZ345S76q0L2 3A4S6190134561690
I Z.0 0.0 170.0 0.0 0.0 0.0 *1-26-3 tI I*****e¢*< z
X l~~~~~~~~~~~~~~~~~t-tI-ILM It II R1-33-OLN 1
0.0 R10-1 11*++vv-+v*+ ~*****¢* +**+ I
* 0.0 0.0 0.Z0 9.SE-6 0.00 t113 ** 1121 *RtO-Z-t aI *9e*4440e ,..ove,**¢****** 1 I 1 300.60 0.0 0.0 RI0-2-Z II I*****¢X3 I1 Ila-3-SLN !I
tI 0 0
t 0.0
I*.'..I I I1.....
0 a
I oo***
It I
I-0. 010
I
I 1.0II 2 t II I.C
II
I1 C
I
1 1 1
I
* 0 CI
0
1 0 0 0 0 0 0
O.S
R ID-5-OL N
1-1
RZ-I
RZ-Z42-24
R2-5
a2-7-1
R2-7-2
U2-7-1
R2-7-OLN
0.0
* I I I00 *64v*4,0
0.0# 131 *.** ***e
I L I I
0.0 0.0
0.0
0.0
****b . . .
1.00E4
I I I I I
0 0 0 0
i 0.0 0.0
pI I I I 0
J 0 0 a O
00co
ao
coo
0
0
0.0
0
0
0000
0
0000
0
I0 00 -L 001 OORZ-13 I
II42-1 I
ff*,e#**9*v*****
0.0 to0 GOSOES 0.10fst
0 00 -L 11£ 0042-li £1
*Z-t. 9
42-1 +
.______,__.__.__* ___ __ * . __ __,
Figure 4.2-2 Coninued.
4-24
DRAPP
I 2 3 4 S & r a
LZ3567Gq001Z345 0IZ)45h;:93O2237590123 67Si0IZ34S6t69013t 3454OL90 5tI90*--------- ------ - - -------*
tSt I
Iiz I
53 I
I
56 1
5S I
ISq tS I
S9
60 0
t
Ibt I
61 1
I
61
66 1
6I t
7Z K
I
*q I
10 *
1 KI
?zI
I
I7.,#
I
Io* OOE 6
I I
a o
I 1
o a
l.OOEO
I I
0 0
1. 00E
I I
o 0
0.0
100.0
1.10 07
1.S9e 07
2.34E 07
1.%09E O
5.0511 07
7.48E O?
9.26C 07
tilde go
1*403E 08
1.7SE 06
0.0
I L
o a
0.0
I I
o a
0.0
I I
0 0
0.0
I I
0 a
OSS~ft AT 101
.00 09
170.0
0.0 0.0
1 I 000 0 0000
0 0 0 0
0.0 0.0
L I 000 a00fl0
O 0 0 0 0
0.0 0.0
I I UOO 0O0000
0 a a 0 0
0.0 0.0
1 1 000 0 0000
0 0 0 a 0
WELL (tw37.5 METER)
1.01E 07 0.0
0.0 0.0
169.70
169 .81
166.35
163.64
163.09
laZ.65
16Z*27
6l1.94
0.0
0
c.o
0
0.0
0
0.0*04
0
0.0
It1O 0.50E6 0.1CE61
K00 -L 101 OORZ-13 K
12-1 1
1.0 0.50E7 0.l0EU
I
00 -L III ORZ-1 t
R2-1
1.0 0.5013 0610(61
00 -1 101 OORz-13 1
00 -1 L0t 0012-I)
12-1-1140 K
P-2 t
0.0 0Oop-3-1 t
P-3-Z
P-4 Kt-* ~t
flA t
P-4F-4 I
F-* +~~
P-4
'-4
'-4
IIIIIIIII
I* 9--- * �* �.-4- �4
Figure 4.2-2 Continued.
4-25
£ t 1 4 5 7 0 lZ34i6 79qQL12478912)i56I61345679001345Z 41Z90t13ii74012345674901Z34S67,09
1 t76 I Z.197 od 1(1t.5 P-4 t
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I77 S z.qSIE On 161A9 3-' 1
I 81 I 1SO OR 161.22 P-4 1
I t
?§I I.0 zlta -
1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 7 S.50E oe 0116.93 P.
I a) I 6.0E 05 1607.0 P-4 I
1 t62 1 7.50o Oa 160.66 P-I.
I I53 1 O.SOE 06 160. 73 -4 t
I tI' I i.50E 06 160.66 P-i t
i * t.03E 09 160.66 P-i 4
I Ilb 1 -100.0 P-4-END I
I S67 T 0 a -STOP I
,_ _ _ . _ _ _ _ . _ _ _ _ _I__ _ _
Figure 4.2-2 Continued.
4-26
the over/underburden. This behavior, based on comparable solutions for a one-
dimensional system, should look much like a complementary error function which
is centered at the aquifer interface and which becomes ore and more diffuse
as time increases.
Results. Figure 4.2-3 displays the numerical solution alongside the
analytical solution of Avdonin 19641. Comparison with the case of heat
transport only within the aquifer indicater that the effect of conduction
within the over/underburden is relatively small. Consequently, the slight
discrepancy between numerical and analytical results most likely would be
attributed to the presence of some numerical dispersion, which arises in the
simulation of convection within the aquifer. Exercises 5 and 6 focus on this
point.
Discussion of Code Outpuc. The microfiche provides a comprehensive
listing of the code output. It includes the following: (1) an echo of the
input data, (2) preliminary setup computations for the radial grid within the
aquifer and the linear grid within the over/underburden, (3) time-step and
well sumaries, (4) a steady-state pressure distribution and (5) time-
dependent temperature distributions within both aquifer and
over/underburden. The reader should carefully examine this output, noting
particularly those data which are pertinent to Cuss, a heat-transport problem.
4.2.4 Eercises
Level 1. Using the Input-Data Guide, the Theory and mplementation and
the microfiche listing of the output, a required, complete the following
exercises:
Exercise I From Table 4.2-1, what physical parameters are required for the
blanks labelLed "1" through "3" in Figure 4.2-2? Check the
answers by referring to the echo of the input data printed at
the beginning of the output.
Exercise 2 Explain the function of the control parameters enclosed in boxes
in Cards -2, M-3-2 and R2-2.
Radius 37.5 meter! . ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
, ~. . . . . . . . . . . . ....... .... ...... . .. , . . .
0 5 10 15 20 25 30 35 40
TI ME (year)
Figure 4.2-3. Temperature Breakthrough Within the Aquifer at 37.5 from
the tnjection Welil.
4-28
Exercise 3
Exercise 4
Hov are steady-state velocities achieved as required by the
analytical solution? The solution option NCALL -2 (Card H4-2)
results in a coupled transient pressure and temperature
solution. owever, a steady-state pressure solution is achieved
by assigning a very small value of water compressibility. Could
a tero value have been entered? Before the reader attempts
this, he should remember that a non-zero accumulation is
required by the transient solution option. Would it be more
accurate to first solve a steady-state isothermal velocity field
(NCALL 4) and then switch to a coupled transient simulation
(NCALL -2, Card R2-l1.5)? If the value of water
compressibility is small enough .here should be no significant
difference.
Interpret the plotting data shown on Cards P-2 through P-4
(Figure 4.2-2). Also note the value of the plotting control in
Card M-2.
Level 2. The reader who wants to develop a deeper understanding of the
code should consider the following exercise:
Exercise 5 The source of the deviation of the numerical curve from the
analytical curve in Figure 4.2-3 is the issue here. As shown in
the R2-12 Cards of Figure 4.2-2, after an initial time step
of At 104 s, automatic time stepping is used, based on a
temperature increment of AT 1.0 C and a relatively large
minimum time step. Could the coarse time step selected by the
code be the source of the observed discrepancy? The reader is
encouraged to reduce both of these controls, rerun the problem
out to. say, t a 2x108 and then examine the effect of these
changes.
4-29
Exercise 6 Also, the temperature distribution within the over/unierburden
may not be resolved adequately at the time t 2 8 s, where
the discrepancy first begins to appear. From the output tables
labelled "Temperature Within the Rock Matric:", examine this
distribution at r - 37.5 (Block 1) to see. if there is a need
for additional resolution at this partiutar time. Reruo the
problem if necessary.
4-30
5-1
S FLOW WITK A FREE-W.TER SURFACE
5-2
5-3
5.1 PROBLEM 8. THE DUPUIT-FORCHHIEIMER STEADY-STATE PROBLEM EAR, 19721
5.1.1 Ob ective
* To demonstrate the simulation of steady-state flow, in a phreatic
aquifer.
5.1.2 Description of the Problem
Problem Statement. The following idealized problem is considered here:
A phreatic aquifer of length L (Figure 5.1-1) with fixed free-water elevations
h(x-O) ah (5.1-la)
and
h (xuL) * h L (.-
at its boundaries is subjected to surface recharge at a rate q. The problem.
is to determine the elevation h(x) of the free-water surface at positions
interior to the boundaries subject to the data given in Table 5.1-1.
Analytical Solution. To solve this problem analytically, the Dupuit
assumption is invoked (Bear, 19721. This amounts to a neglect of vertical
flow:
u a -K H/az - 0 (5.1-2a)
and yields a strictly one-dimensional solution in the x coordinate. Here is
the total head, H h - . Further, it is assumed that for the horizontal
flow, the ransmissivity is controlled by the saturated thickness:
T aK (5.K-hb)
so that the total discharge through a vertical surface per unit of width in
the direction (Figure 5.1-1), is given by
u -T Ia (5.13)
RECHARGE
%AI
.D.
ho
Figure 5.1-1. Problem 8. Schematic Diagram for the Dupuit-Forchheimer
Problem. I
5-5
Table 5.1-1. Problem 8. Input Specifications.
Parameter Symbol Value
Surface Recharge q 7.505xlO0 /s
Lateral Conductivity K 0.03 m/s
Vertical Conductivity Kz 0.003 /s
Height of Free-Water Surface at x 0 ho 0.75 m
Height of Free-Water Surface at x L hL 0.25 m
Length of System L 20
5-6
The equation of continuity Bear, 1972] then becomes, for steady state,
a (}C to ah) + q 0 (5.1-4)
rhich is the Forchheiter equation. For the boundary conditions prescribed in
Equation (5.1-l), the analytical solution of Equation (5.1-4) is the Dupuit-
Forchheimer parabola:
h2 h (h - h2 )x/L + (q/K )(L-)x (5.1-5)a L a
Flow rates at the two boundaries are obtained from Equations (5.1-2b) and
(5.1-3)
(0) (KI2L)h - h1) + (<))(qL/2) (5.1-6)
Here, the first term (on the right-hand side) represents the flow caused
strictly by the difference in heads on the two ends of the system. This rate,
of course is the same on both ends of the system. The second term arises from
the surface recharge. It shows that half of this recharge exits through the
boundary x 0 d the other )If through the boundary x L.
5.1.3 Numerical Solution
To solve this problem numerically with SWIFT, a two-dimensional vertical
cross section is gridded as listed in Cards R2-17 through R2-19 (Figure
5.1-2). Hydrostatic conditions
p - P - (gtg )h - constant (5.1-7)
are then prescribed for both side boundaries (Cards R-27 and R1-28). Hero
the vertical discretization is necessary for two reasons: Firstly, it
provides geometrical resolution for the free-water surface. Secondly, the
vertical discretization is required since the SWIFT formulation does permit
vertical flow. Thus, there is some difference at this point between the one-
5-7 /
I 2 5 A S O 7 SIZi0 ?90*15& ?190123456 r190ot5J6890t34S67 oW34QS7I90oz).Sia6eo9lZ,Ss9o
-a- *- n a e_._----- e- -.. S..ne - p einO o..
Ia 1PROILE 0. 0. SIFT I EXAMPLE. SI UNITS. 12/33.I
2 IUPUIT-FVOMCHHtINfR STEAOv-ITATE FREEVATEM SURFACE VItH ECHARC
31 i* 0 0 1 0 0 0 1 0
41 20 I *0 t 0 1 0 2 0 to 20 0I+...... .
5* 0 0 0 0l
b 1 0.0 0.0 0.0 1.0 L.0
7 I 0.0 0.0 0.0 0.0 1.0 1.0 II
a I 2130.0 t.OOEOT 20.0 130i.4 100040t
9 I a 2 Z 2
t0o 20.0 t.OOE-03 20.0 IOc-03
IL 1 40.0 1.OOE-03 60.0 IOOE-03
IZ I 40.0 I.001-03 10.0 1.OOE-03I
i3 1 0.0 20.0I 0 . S .0 - .. * . . ... a** .. 4,
14 I ICo.0 20.0 . "4*b REFERENCED IN TEXT aI * . EFERENCE IN EXERCISE
10 I 20.0 0.0 0.0 0.0
LI 20$¢<e0I 1 ZSI.0
ISOO@*0*0*OI" I .0
z s zoo.oso
20 s ll .. £21 *. 131 * 0.50 0,0 0.0
ttI
23 I I t £ oI
24 1 1. 0.000I
25 * 1 1 1 t 7 7
N-1-1
K-L-2
K-2
0 09-3-1
x-3-2
RI-I
L.OE-IORI-2
Rt-3
Rl-S
Rl-i
* 1-9
t-L0
RIL-lt
Ri-1
RI-t
RI-lb
0 ,0 1-20
R 1-26-8LN
RI-Z7
11-28
R I-Z
RL-2U
*----------- - - - - -- - - ^- - - v- - - - -* - - s
Figure 5.1-2. Listing of SWIFT t1 Input Data.
5-8 Up, P
t I 3 O 5 ;. I a12306 W69oz14567aq0I2 3'56a90121,7590123S456s9023S6?6012456769012)4567800
*--------*--_ -- - -- - -- - - * --- - - -
I-.e...... 9 .49C44 .26 I 1. Iko.
I .. I . C** * * I* a11 1 I 1 1 I t e
I
29 1I
30;I
I32 £
I3" I
I34 1
BI
I
39 I S
3Z t
41 tI
t
I
431
44 1
461
t
471
I46 1
I
49 I1
50 9
1. 76.
t l I I
1.
I I
I.
L I
1 t
I I
I .
I I
1.0
I L
14640
I I
19S2.
I L
2440.
I I.
Z924.
I 1
1416.
I i.
390'..
1 1
9 9
10 10
1 It
12 12
3 £1
4 14
A t-28
RI-Zs
Rl-26 1RI-Z6~~~
RI-as
R1-20 t
11-28 11
11-2 6 I
RI-aS
R -2 I
R1-20 I
RI-to~~~R-26
Rl-Zs 11-26 I
RL-26 1
tt~~~t ~ t
AI-Z6~~~
RL-24 I
I
RL-Zc t
Iti-Z s g
Rl-ZG t
I
R1-26 t
Rl-Z0 |~~
lS iS
1. Oh1.
I I I I
I. 0ea.
t I I I
I. 5366.
t I 1 1
1t 17
1 1
I.
I I
5656.
1 1 1q 19
t . 634%.
Figure 5.1-2 Continued.
5-9 4
I Z 3 S 7 r a
1234569 90 1Z *S5J'9012345912306390690124567g90123i5 9t235676S QOIZ14tS689O. _ e_ _ - _ _. __ _ .* - - - - - - - - -n- - - - - - -
S I I I * LS2 1 1. 661?.
S 1 20 20 I 1I
54 I 2. 0.00001
5S * 20 20 1 IS
Sh 1 2. dO.I
57 1 20 ZO 1 t
ii t Z. 974.IaI
59 I 20 20 1 11
60 * 2.I
6 1 20 0 1 1
62 I 2. 1952.
63 1
I
to 20
1i 14
1? I?
to to
19 1q
20 20
11-26 K
11t-25 1
ti-ze I
RI-ZO
R 1-ZR 111-26 1
Rt-Z K
RZ-I {~~
11-26 5
IL-lU~~
t-laS SO
65
66
681
69
To
71
72
73
II
K
I
II
I
a 0
0.0
0 0* &*0, 9
1 20
I 20* *9e
0.0
t I
o 0
a
0 0 0 a
t 1* 9) -***.. ~ .4444.
* *a,*
0 1 0 t.4.4.
a
Q.o
a I
0 1
0*0*4
I I.*00
0 o
Olt
0
a 0000
o o
e 0g..
0 1$0. 4
9 *- *fl--.…*n*�en 0 *aae e4
Fi--ure 5.1-2 Continued.
5-10
dimensional analytical treatment and the two-dimensional numerical
treatment. In general, the effects of the vertical flow may be minimized by
making the vertical conductivity, a free parameter in this case, sufficiently
large. Fot this problem, however, such was not necessary, and it as
sufficient simply to set K K
Insofar as the horizontal transmisivity is concerned, however, the
numerical reatment is very close to the analytical treatment. For a
completely saturated grid block, the full thickness A is used for the
cransmissivity. owever, for a partially saturated grid block, only the
saturated thickness h is used:
T Kah (5.1-B)x x
Results. Figure 5.1-3 shows the numerical solution plotted alongside the
analytical solution, and Figure 5.1-4 shows the same numerical solution
plotted in relation to the grid blocks. An evaluation of the flow yields
-3 3UL (SWIFT) a 1.10 x 10 m Is/n (5.1-9a)
and
u (analytic) I.-' x 10 3 3s/m (5.1-9b)L
a difference of about three percent.
Discussion of the Code Output. The SWIFT results are taken basically
from the tables which appear near the end of the output listing (see
microfiche). Water-table elevations come from the table of grid-block
saturations, and the flow rate L comes from the table of aquifer-influx
rates. The former table is printed through control IPRT, and the latter
through control 106, both of which appear on Card R2-13. To interpret these
tables, however, requires rome additional data from he output. For example,
by examining the former table, we note that, among the unsaturated blocks,
(6,K*5) w 0.8943 (5. 1-10a)
1.0
.8 - ANALYTICAL0 SWIFT
.6
.2~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.
0 2 4 6 8 10 12 14 16 18 20X (m)
Figure 5.1-3. Steady-State Free-Water Surface far the Dupuit-Forchheimer
Probl em.
RECHARGE
IXI I .X I I, I I I
z
F
a
A
Figure 5.1-4. Ceometrical Characterization of the ridding Boundary
Conditions (Labelled "AIF") and the RecharRe. The numerical
solution is also shown in relation o the ridding.
5-13
is the aximum saturation. Grid-block tables, located near the front of the
output, are then used to convert the grid-block saturation to the elevation
h (x 6f5.5 ) 0.795 (5. -lOb)
To interpret the aquifer-influx rates requires the use of another table
of the input echo, namely he one for aquifer-influence block numbers. This
table shows that for the boundary x - 0, the aquifer-infLux blocks are
labelled "1" through "15", whereas, for the boundary x L, the boundary of
interest here, they are labelled "16" through "20". Consequently, to obtain
the flow quoted above in Equation (5.1-9), we form the sum
20 3 3UL Nm 16uK 1.10 k/s/mn a 1.10 x 10- M s/rn (5.1-il)
5.1.4 Exercises
Level . Using the Input-Data Guide, the Theory and Implementation and
the microfiche listing of the output, as required, complete the following
exercises:
Exercise I
Exercise
Exercise 3
From Table 5.1-1, what data are required for the blanks labelled
"1" through "4" in Figure 5.1-2, the input data set? Check the
answers by referring to the echo of the input data printed in
the output.
Interpret the boxed control parameters of Cards M-2, -3-t, 2-
1, R2-2.5 and 2-1l.
Interpret -he control parameters VAB and then analytically
verify the boundary pressures P, as indicated in the R-28
Cards. Use Equation (5.1-7). Also verify that these data were
read correctly by the code by examining the tables entitled
"Constant ressure, Temperature, Concentrations Block Types" nd
"Constant Boundary Pressures".
5-14
Level 2. The reader who wants to develop a deeper understanding of the
code should consider the following exercises:
Exercise 4
Exercise 5
Exercise 6
In the problem specifications the flow is essentially
horizontal. To what extent can the simulation be modified
before the Dupuit assumptions are in error? The problem as
specified has a maximum free-water surface slope of
approximately 10. Bear 1979, p.78] indicates that the error
is small as long as 2 (< I where is the slope of the free-
water surface. The reader should challenge the code by
increasing t slope and examining the error. This could be
accomplished by increasing the recharge rate, decreasing the
horizontal domain, and/or increasing vertical domain. An
increase in the vertical conductivity may be required for some
of these tests.
Can the free-water surface option be used in cylindrical
coordinates? CertainLy, in fact, Bear (1972, p. 311 presents
the analytical solution with Aich to compare. The reader is
encouraged to switch to a cylindrical g d (TG 3, X-3) and
execute this problem. We suggest a no-flow condition at the
perimeter of the aquifer and a pumping well at the origin,
completed in multiple layers, and subject to a specified bottom-
hole pressure of zero.
Can the free-water surface option be used for stratified
aquifers? Yes. The reader is encouraged to design such a
simulation in a Cartesian geometry (HTG 2 in Card -3) for
comparison with the discharge formulas of Bear 1972, p. 370
ff1. The reader should introduce heterogeneity by adding
layering in the RI-2t Cards.
5-15
5.2 PROBLEM 9. THE BOUSSINESQ TRANStENT-STATE PROBLEM [EAR, 19721
5.2.1 Objective
0 To dmonstrate the simulation of transient-state flow in a phreatic
aquifer.
5.2.2 Description of the Problem
Problem Statement. The following idealized problem is considered here: A
phreatic aquifer (Figure 5.2-1), semi-infinite in length, is initially
saturated to its full thickness so chat
h(x,tsO) h x > 0 (5.2-la)
Immediately thereafter, however, water is discharged at a sufficient rate to
reduce the saturated thickness at one end to half its original amount, i.e.-,
h(x=Ot) - h2 , t > (5.2-1b)
There is no recharge through the upper surface. The problem is to determine
the elevation of the free-water surface h(x,t) as a function of position and
time for the parameters given in Table 5.2-1.
Analytical Solution. Here, just as n Section 5.1, the Dupuit assumption
(Equation S.1-2a) is invoked, and the horizontal transmissivity is taken to be
proportional to tne saturated thickness h. The equation of continuity [Bear,
19721 may then be written:
a (K h h) _a (5.2-2)
This is the Boussinesq equation. Polubarinova-Kochina [ear, 1972, p. 384J
has obtained a general solution for this nonlinear equation, which is
presented graphically in Figure 5.2-2.
INITIAL
h1= 1.G
IP"
Figure 5.2-1. Problem 9. Schematic T" .ram of the Boussinesq Problem.
5-17
Table 5.2-1. Problem 9. Input Specifications.
Parameter Symbol Value
Lateral Conductivity Kx 0.01 ma/3
Vertical Conductivity* K 100.0 m/9
Porosity 0.50
Initial Height of Free-Water Surface ho t.0 M
Height of Free-Water Surface at z 0 h 0/2 0.50
* Adjusted to a suitably large value to approximate the Dupuit Assumption.
2.0
hlh(o)
U'a
1.8SO
V 2h h(0)
Figure 5.2-2. Transienc-State Free-Water Surface for he Boussinesg Problem.
5-19
5.2.3 NumericaL Solution
Discussion of Code Input. For the conduction mechanism (left-hnd side
of EquaLion (5.2-2)), the discussion here follows that of the previous
problem, which may be summarized by the following three statements: Firstly,
a two-dimensional grid (Figure 5.2-3 and Card -3-1) is required for the
numerical simulation as opposed to the simple one-dimensional analytical
simulation. Secondly, the Dupuit assumption is not used by the numerical
formalism but may be approximated by choosing a relatively large vertical
conductivity Kz (Cord R-20). Thirdly, the assumption that the transmissivity
is proportional co the saturated thickness is used in the numerical model just
as in the analytical model.
For the accumulation mechanism, the nalytical and numerical treatments
are quite close. For the latter, the grid-block saturation is unity for each
saturated grid block, but, for the partially-saturated block, it is
i a WhIA (5.2-3a) -
where Ah measures the saturated thickness. By defining saturation in this
manner, it can be shown that he:-teneral expression of accumulation, as given
in the Theory and Implementation reduces, under appropriate conditions, to the
right-hand side of Equation (5.2-3a):
Gi (pS) Oa (5.2-3b)
Results. Figure 5.2-2 shows the numerical solution plotted alongside the
analytical solution. For the most part, the agreement is quite reasonable.
However, for the smaller values of time the numerical free-water surface does
tend to drop somewhat more rapidly then does the analytical solution. Most
likely, this deficiency could be corrected by a mare refined spatial mesh
andlor a larger value of the vertical conductivity.
Discussion of Code Output. As ndicated in the discussion for the last
problem, the height of the free-water surface is not a direct output of the
code, but must be clculated from the output. Consequently, it is of interest
5-20 0DRpi-
I 2 3 4 5 6 7 at1Z4S6890Z1 4S07690t134Sb7S9l2l4567Q901Z3*45669O 345674901Z3456?8901230567890
I IFRPOLER hO. 9. SIFr I EMPLE* St UNITS, U2/83, -I-L II
2 ISOu5SINES11"S EON. - ItANSINr FREE-VATEA OESATURAttOk FROM t(IT. SW. K-I-2 It Ii 1 a 0 1 0 0 0 1 0 9-2 E
4 l 20 t 20 I 0 L 0 2 0 10 0 a 0 OM-3-1 l
S * 0 0 0 n-3-z
6 1 4.OOE-0 0.0 0.0 t. 1.0 Ql- II
I I 0.0 0.0 0.0 00.0 [.L t.OOGE-10Rl-2 SI I
6 r 2110.0 I.OOE 0r 20.0 1000.0 LO . R1-3 II t
S I 0 Z 2 2 Rl-6 II
10 * 20.0 1.00E-01 20.0 1-OOE-C3 Ri-TI t
i1 1 40.0 1.OOE-01 80.0 LOOE-03 RL-9 1
12 1 40.0 l.O0f-q3 0.0 l.OOE-03 RL-10 II L
11 1 0.') Z 0.O *......... .... e....e.... . , ...... e Rl11 1
I . *.* P SEFERENLCED IN TEXT . I14 I 1000.0 20.0 R "e.. REFER 'ICEO IN EXERCISE . RI-11 1
1 * 0 0 tI-lZI **.94$.**~e*9+.4*. 1
16 I 1U.00 IL) # 1Z * o.0 RI-Lb I
I7 I 590.01 S0.05 S0.10 240.5C 291.0 2.0 11e*** ***@ *¢********************,****,,**1
ld I 1.0 1
Iq I 20'0.005 I.j+¢¢¢+ 9999999C 99 *9999S **.*+- I
20 *. 43 * 1.0 100.0. I) * 0.0 0.0 O.ORl-20I*"*e^e.... **949S99e9.*.+*t...e t
21 1 -t-27L I II I
22I I l2
231 1 1 1 1 i 1t R-28-1 I
at I 1t 0.00001 R1-28-Z I
25' * I 1 I 12 I? U-Z3-Il C … - * O- *…-- - - - - - - - - * - -- - - - * - - - * - -
Figure 5.2-3. Listing of SWIFT tInput Data.
5-20
"R*Apr
1 2 3 4 S 6 1 aI RI I Z01z345 7I l Sh 8912)412)4516ql23467a901Z34%b6790I23#Sb?$9Q
.--- - ---- -- __ _---*-- ----------I I
I CIPOKE.u NiO. 9, SWIFT It fXAMPLE S Ukftnt , 2161a R-I-1 II
2 IOUSS INES4 0SI
I I I aI
4 I 20 II
S # a O
6 1 4.OOE-LOI
7 I 0.0I
8 I 213000I
q a z
10 * 2060 I
t 1 40.0 1
12 t 40o0 1
13 I 0.
14 I 1000.0
I 4 0 a
EON. - tRANSIENT FEE-IATER OESATURATION FROK INIT. S. -1-2
0
20
1 0
1 0
0 0 1 0
I 0 2 0 g0
a
0 .0
0.0
L.OOE 07
2 2
I .OOE-03
L.OoE-O)
L.OOF-03
200
20.0
0.e
O.G
20.a
2C.c
JO.C
Mo.c
I
I
I
1-0
0.0
1000.0
1.0
t.0
1000.0
0 0 0 0X-3-1
l-3-Z
it I-
1 . 1.OE-10RI-2
R1-3
RI-6KI-7
kI-LO
Rl-ITERXIl' a RL-Q(ERCESE . Rl-tS
IIIIII4
II1ICIIIC
IIII.tIII
I t.OOE-02
I 1.OOE-03
I L.OOE-03
. *0*0 EFERENCEO IN EX
. .e"* REFERENCEC CM fX
Ih
IJ7
I a1'?
20
Z1
22
23
24
25
t Z. t 0.0. CI) .. 123 w 0.0194040¢* 440400444e 44 440040 oo44o.00 99494994494 9009 0 0* 9 9to**$
I S090L 5'0.05 SO.l10 200-50 2*1.0 2.0
C20'0.050I t.O~*499*L 4 0040* 4404044, Iii * 1.0 100.0. 141 * 0.0 0.0
I 4 1
It 1-12
ft -16
0.RI-L6
f.t 1-ZOBLRI-O-BLN
Rl-Z?
RI-Z0-2
RI-li-I
IIIIIII
II
IIIIIII
I
I
I
I
I t I I 11 I
1. o.aooot
I I I I IZ 12
_______ - - -___A
Figure 5.2-3. Listing of SWIFT 1 Input Data.
5-21 I.4JRI4A r
I 2 3 4 S 6 T a1Z34Sb?1901234S% 11s9o1231 7qQ19Z3456?19016?890L23sb?4901Z3s45 eo0
*--- ----- *--'-- * ----- *-* ------- --------- *
1.,**+**. .*..9.... ' t26 I. %t* £Ql-Z0-Z I
K27 I I 1 1 3 1 3 11
IZs I
1
30 0
t31 1
I
I
3 13I
I
37I
I19 II
40 *I
4 1 II
4 2 I
, 976*.
I I I I
I . 1464.
I * t S
I I I I
. MRsz.I t 1 1
1. 2440|.
I I I 1
I. 1416.
I 1 1 1
14 14
15 15
Ih~ 16
11-ZO-1 II
R1.18-2 1I
QI-28-l II
RL-28-2 tI
at-26-2 I
41-26-2 t
IRIe-s
I11-23-2 I
Ift1-ZS-1 I
14K1-z8-2 I
17 17
L 1a
19 19
1.
1 I
3904.
IL I
41-28-2
RI-2zt
I
920 20
to . 39Z. Rl-26-? I
43
44
45
46
4 ?
46
so
II 21-Z-LN
RI-33-BLh
IIII
* 0 0 0 1-1l I
0.0 £18-? 1
c n 0 0 0 C 0 0 0 0 0 RZ-L
I.QO E-07 l.00E-07 g
0 a t I 0 11LI 0 0000 0 012 1
t 0 a 0 0 0 0 0 0 0 0 0 12-1I 1* … 9--.-----t-----*---------- 9… -- t--…_-*-- -- ----
Figure 5.2-3 Continued.
5-22
bRA1t
I ) i S 4 7 6123I)567890IZ as 07 0Z345 91210 S 67901234WO0234567690121456 9t Z3061690
.________ ._______. __*e._ _._-..
ISt I &.Zs(-04 .249E-04
5Z I I I I L
S1 1 0 0 0 0I
5' 7q.lb36E-4l.SISO3E-'I
55 * a I I tI
56 I a 0 0 0
57 I1.7b3I11-3?.86547E-1
se I I I
Sq O I
60 *3.qOJ5E-3
61 I I t.62 t
62f£ 0 0I
61 11.SbZ'f-02
beI I l
*s a o
54 I 0.0%34l2
b7 1 1 I
6' I 0.VOO
I10 * I I
71 I 0 0
7S I 0.OQ?6561
1) I l 1
7S * a . jQOh2
* I
O 0
0.0
I 1
o a
0.0
a I
o 0
0.o
1 I
o o
0.0
I I
0 0
('.0
1 1
o 0
9.6'
1 1 111 0 0000 0
O o 0 0 0 a
1 1 *1 00000 0
O O a a a 8
1 1 1L1 00000 0
0 0 0 0 0 0
0.0 0.0 1000,0
1 I ILL 00000 0
o 0 0 0 0 0
0.0 0.0 1000.0
1 I it 0 0000 0
0 a 0 0 0 .
0.0 c.o 1000.0
I 11 t 0 0000 0
0 0 0 0 0 0
0.0 0.0 1000.0
1 I 111 00000 0
o a 0 0 0 a
0.0 0.0 1000.0
1 I 111 00OO a
0 0 0 0 0 0
B.B R. ? WGGB.
11
0 R2-1]
RZ-I
0
RZ-L)
RZ-I
IIIIII
II
IIII
0.0
C
00
II
IRZ-11 tQt-il I
I
QZ-I I
4.0 62S(-04*I
RZ-13 II
6.0 6.2SE-041
I
QZ-13 I
I
6.0 6.1SE-041I
Az-I] I
Rz-i II
6.0 62SE-0"t
IIt-iL I
0.0
0
0.0
0
0.0
0I
R-I tI
I… - _-. * - _ ---…
Fiure 5.2-3 Continued.
5-23
* 2 1 4 S 6 7 6t2345*?61 aIz I *5b1z1 b7Bq0I2) 67901l3b56?6q01Z34¶6 7890ozstl TaOt1z3)sbT6qO
*--- - ---- ------- -
lb 7E I I I
771 0 0
70 1 0.61035I
$I I t t50
ao * Q
I~~~
06 1 0
89 0 0 0
e I
61 ' IO IO
IF t 0 I
631 t 0 0
60 * 2*1 1
91 I a I
I
86 1 0 0
93 1 3175462
qq I Io r&b
I
L6O I. I
IO * t.414
a I
0 0
0.0
I I
o a
0.0
I I
o a
0.0
I I
o 0
00
I I
o o
0.o
1 1
o 0
o.0
I I
o a
0.0
I I
o 0
00
I I
1 1 1 0 I 0000 0
O o 0 0 0 0
0.0 0.0 1000.0
I I11 a0000 0
O 0 0 a 0 a
0.0 0.0 t000.0
I 1I Lit 00000 0
a 4 0 0 a 0
0.0 0.0 1000.0
1 I 111 0000
0 0 0 a 0 a
0.0 0.0 1000.0
t I I1I 00000 0
0 a a a 0 0
0.0 0.0 1000.0
I I LIt 00000 0
O 0 0 a 0 0
D.0 0.0 1oo 0.O
I 1 111 00000 0
0 0 0 0 0 0
0.0 0.0 1000.0
I I tl1 00000 0
0 0 0 0 0 a
0.0 0.0 1000.0
I I III 0C000 0
I1
0.0
0
0.0
a
0.0
0
0.0
0
2-1) 4
RZ2-1 1
6.0 6.25E-041a
R2-13 1
RZ-1 R2-13 ;
I
4.0 6*25E-041
R2-1 t
4.0 62SE-04t
I
22-13 1
R2-tI t
t2-i
6.0 .2SE-041R2-13
R2-I I
a.o i.2SE-oi l
12-. 1
tZ-z ~I4.0 1.ZSE-0'..
12-1 3 ;
0.0
0
0.0
0
0.0
0
0.0
0
._____ __.___ ____._ __ __ __ - __- -__-_-_-_-_-_-_- -_- - -_-_-_-
Fivire 5.2-3 Continued.
5-24 a
L Z * S 5 6 7 a1Z1450 r01 356 If q iLZ 3 b 40123456 7s90123fstya0L234567eX0
* 0---- - - - - - -- - --a- - - - - - - -- - - - - -__
It01 1 0 0
Itoz I 31.e06
It a) I l I
I1o 0 01
o o
0.0
I I
o 1
o a
00
I I
o a
0 0
0.0
11L 0
0 0
0 0
OO0.0
0000 a
0.0
I12-I I
I2S.0 b.2SE-0 1
IItZ-12 I
RZ-1-SIOP I
11
a
a - - - * _ _ __--------_- a---- - _ *
Figure 5.2-3 Continued.
5-25
to go through the appropriate procedure once again. This time, however,
rather than using the grid-block saturations, as was done for Problem , we
shall use the grid-block pressures directly.
Turn in the output to the input echo, and from Lhe output table entitled
"X-Direction Di tance to Grid-Block Center" (see microfiche), note that the
distance
x 0.025 (5.2-4a)
falls at the center of the Blocks I 3. This is one of the distances used
for Figure 52-2. Advancing to the pressure-at-elevation table for time
t - 9.77 x 10 a (5.2-4b)
for which
C - (a/2K x2 x 0.4 (5.2-4c)
we cbserve that
p(3,6) a -287.6 Pa (5.2-5a)
Thus the unsaturated block thickness is
hh(3,6) - 0.0293 (5.2-5b)
Since the block thicknesses are z 0.05 m throughout, we easily compute
h(x,t) - (15)(0.05) - 0.0293 - 0.721 (5.2-6a)
in agreement with Figure 5.2-4. Converting to
maximum elevation h 2 m, we obtain
a dimensionless scale with
h/h (t 0.4) 1.440
(5.2-6b)
in agreement with Figure 5.2-2.
SURFACE AT = .0976 sec.
Z 1 13 _ J 1 20
T _ _ __ ___
AIF--
o .025 .125 5.3
Figure 5.2-4. Geometrical Characterization of the Cridding and Boundary
Conditions (Labelled AIF") in Relation to the ??umerical
Solution.
5-27
5.2.4 Exercises
Level . Using the nput-Data Guide, the Theory and Implementation and
the microfiche listing of the output, as required, complete the following
exercises:
Exercise I
Exercise 2
Exercise 3
From the description of the problem and Table 5.2-1, what data
are required for the blanks labelled "I" through 4" in Figure
5.2-3, the input data set. Check the answers by referring to
the echo of the input data in the output.
Explain the reason for the relatively large value of KZ (Card
RI-20) and verify the boxed numbers in Card R-28. Hydrostati,
equilibrium (Equation 5.1-7) is assumed here just as for Problt
8.
What is the total flow rate passing through the boundary block
Level 2. The reader who want. :o develop a deeper understanding of the
code shoulI also-onsider the following exercise:
Exercise 4 Convert the Problem 9 data set (Figure 5.2-3) to cylindrical
coordinates using a multiply-completed well with a bottom-hoLe
pressure of zero to establish the boundary condition. Rerun t
problem and plot the results using the same variables (hs'h and
& - (/2Kt) 2 r) and scaling a in Figure 5.2-2. Should the
radial results ie above or allov the Cartesian results? Why!
5-28
6-1
NOTATION*
ROMAN SYMBOLS
a thickness of prism or radius of sphere for the local subsystem
aW explicit portion of e, the boundary flow rate of the simulated aquifer
A radial grid block-ratio
AV global/local interface area specific to global volume
Am specific activity
b aquifer thickness
bo aquitard thickness
bW implicit portion of eV, the boundary flow rate of the simulated aquifer
cPR specific heat of the rock
CR compressibility of the pores
cw compressibility of the fluid
C concentratioa of radioactive (trace) components
* The terms "global" and "local" are used in this section. Typically the
global model is used to represent the fractures or the aquifer. It is
reginally connected and may be three-dimensional. Usually the local
submodel is used to represent either the porous rock matrix or the
confining beds. t is one-dimensional only.
6-2
d fracture half-thickness
D dispersion/diffusion
DM molecular diffusion in porous medium
D* molecular diffusion in water
elf boundary flow rate into the simulated aquifer arising from the aquifer
which surrounds it
g acceleration of gravity
9c units conversion factor equal to g for the English system and equal to
unity for the SI system
h head
H total head
I radionuclide inventory
kd radionuclide disCribution coefficient
IV dimensionless distribution coefficient, Kv Rkd
K retardation factor or hydraulic conductivity
L length
m density of radioactive waste, i.e., mass of radionuclide per volume of
w3ste
ns number of nodes within a local unit
p pressure where subscripts 0 nd I refer to boundary locations
6-3
Pt terminal-rate influence function. used for odel external aquifer
q mass source rate or fluid recharge rate
qw radion- :lide source due to waste Leaching
Q rate of fluid withdrawal from well
r radial coordinate
re external radius of aqifer model
r, extent of stress-relief fracturing
rw radius of wellbore
rl radius of skin or radius to center of first grid block
source term for release of nuclides from the waste matrix
s one-dimensional coordinate for the local units. In formal equations
s - 0 represents local/global interface. In code, however, a 0
represents the external boundary of the ocal subsystem.
drawdown
S storativity
Ss specific sracivity
S saturation
t time
C3 total leach time
6-4
tD dimensionless time
T temperature, transmissivity or leach duration time
u Darcy flux
U mass fow rate through radionuclide inventory
v interstitial velocity
WI well index
x,y,z Cartesian coordinates
GREEK SYMBOLS
aL longitudinal dispersivity
r global-to-local radionuclide transfer rate
rH global-to-locaL heat transfer race
rW globaL-to-local flow rate
6p incremental change in pressure over a time step At
Ah head increment
As spatial increment in s for the ' ' subsystem
at time increment
Ax spatial increment in xi where xi x x, x2 y and x3 z
A decay constant
6-5
dimensionless Boltzmann variable
fluid density
PR formation density
p volumecrit waste density, i.e., volume of waste per bulk volume
T radionuclide half-life
t t' ottuosity
0 porosity
SUBSCREPTS
m fluid-plus-rock composite material
r radioactive component. Subscript is suppressed whenever no confusion
should arise. If given as an integer, r is the component numbeL in a
chain of species.
w radionuclide source (repository)
x,y,z directional indicator
SUPERSCRIPTS
denotes local subsystem
indicates characteristic length
7-1
REFERENCES
Avdonin, N. A., 1964. Some Formulas for Calculating the Temperature Field of
a Stratum Subject to Thermal Injection, Neft' i Ga:, Vol. , No. 3, pp. 37-41.
Bear, J., 1972. Dynamics of Fluids in Porous Media, American Esevier
Publishing Co., New York.
Bear, J., 1979. Hydraulics of Groundwater, McGraw-Hill.
Carter, R. D., and Tracy, C. W., 196C. An Improved Method for Caiculating
Water nflux, Trans. SPE of AIME, 219, pp. 415-417.
Dennehy, K. F. and Davis, P. A., 1981. Hydrologic Testing of Tight Zones in
Southeastern New Mexico, Groundwater, Vol. 19, No. 5, pp. 482-489.
Dillon, R. T., Lantz, R. B., and Pahwa, S. B., 1978. Risk Methodolog for
Geologic Disposal of Radioactive Waste: The Sandia Waste-Isolation Flow and
Transport (SWIFT) Model, NUREG/CR-0424 and SAND78-1267, Sandia Narional
Laboratories, Albuquerque, New exico.
Finley, N. C., and Reeves, H., 1981. SWIFT Self-Teaching Curriculum:
Illustrative Problems to Supplement the User's Manual for the Sandia Waste-
Isolation Flow and Transport Model (WIFT), NUREG/CR-1968 and SD81-04lO,
Sandia National Laboratories, Albuquerque, New Mexico.
Hantush, . S., 960. Modification of the Theory of Leaky Aquifers, J.
Ceophvs. Res., Vol. 65, pp. 3713-3725.
Huyakorn, P. S., 1983. FTRANS, A Two-Dimensional Code for Simulating Fluid
FLaw and Transport of Radioactive Nuclides in Fractured Rock far Repository
Performance Assessment, NWI-426, Battelle Memorial Institute, Office of
Nuclear Wste Isolatv., Columbus, Ohio.
7-2
INTERCOMP Resource Development and Engineering, Inc., 1976. evelopment of
Model for Calculating Disposal in Deep Saline Aquifers, Parts I and I,
USGSIWR-76-6L, PB-256°03, Natiunal Technical Information Service, Wash agton,
D. C.
INTRACO1N, l983. ;nternational ucLide Transpoct Code ntercomparison Study,
Sedis;'4 Nuclear ower Inspectorate, Stockholm, Sweden.
Mercer, J. W., and Orr, B. R., 1979. nterim Data Report on the Geohydroiofy
at the Proposed Waste Isolation Pilot Plant Site in Southeast New Mexico,
U. S. Geo. Survey, Water Resour. Inv. 79-98.
Pahva, S. B., and Baxley, P. T., 1980. Detection of Fractures from Well
Testing in Procetdings:_ Workshop on Numerical Modeling of Thermohydrological
Flow in Fractured Rock Masses, LBL-11566 and ONWI-240, Battelle Memorial
Institute, Office of Nuclear Waste Isolation, Columbus, Ohio.
Pinder, G. F., and Bredehoeft, J. D., 1968. Appication of the Digital
Computer for Aquifer Evaluation, Water Resour Res., Vol. 4 No. 5, pp. 069-
1093.
Pinder, G. F., and Frind, E. O., 1972. Application of Galerkin's Procedure to
Aquifer Analysis, Water Resour Res., VoL. 8, No. 1, pp. 108-120.
Powers, D. W., Lambert, S. J., Shaffer, S., Kill, L. R., and Weart, W. .,
1978. Geological Characterization Report, Waste Isolation Pilot Plant (WIPP)
Site, Southeast New Mexico, Vols. I and II, SAND78-1596, Sandia National
Laboratories, Albuquerque, New Mexico.
Reeves, M., and Cranwell, R. M., 1981. User's Mlanual for the Sandia Waste-
Isolation Flow and Transport odel (SWIFT) Release 4.81, UREG/CR-2324 and
SANDS1-2516, Sandia National Laboritories, Albuquerque, New Mexico.
7-3
Reeves, M., Johns, N. D., and Cranwell, R. ., 1984 a. Theory and
Implementation for SWIFT 11, The Sandia Waste-Isolation Flow and Transport
Model for Fractured Media. To be published as a SAND-NUREG report, Sandia
National Laboratories, Albuquerque, New Mexico.
Reeves, ., Johns, N. D., and Cranwell, R. ., 1984b. Input Data Guide for
SWIFT 11, The Sandia Waste-Isoltion Flow and Transport Model for Fractured
Media, to be published as a SAND-NUREC report. Sandia National Laboratories,
Albuquerque, New Mexico.
Ross, B., Mercer, J. W., Thomas, S. D., and Lester, B. H., 1982. Benchmark
Problems for Repository Siting Models, NUREG/CR-3097, U.S. Nuclear Regulatory
Commission, 138 pp.
SilLing, S. A., 1983. Final Technical Position on Documentation of Computer
Codes for High-Level Wste Magement, NREG/CR-0856, U.S. Nuclear Regulatory
Corm-ssion, 11 pp.
Tang, D. ., Frind, E. 0., and Sudicky, E. A., 1981. Contaminant Transport in
Fractured Pcous Media: . Analytical Solution for a Single Fracture, Water
Resour. Res., Vol. 17, No. 3, pp. 555-564.
Van Everdingen, A. F., and Hurst, W., 1949. Application of the Laplace
Transform-j:inn to Flow Problems in Reservoirs, Trans. SPE of AIME, Vol. 186,
pp. 305-324.
Ward, D. S., Reeves, ., and Duda, L. E., 1984a. Verification and Field
Comparison of the Sandia Waste-Isolation Flow and Transport Model (SWIFT),
NUREG/CR-3316 and SAND83-1154, Sandia National Laboratories, Albuquerque,
New Mexico.
Ward, D. S., Reeves, M., Huyakorn, P. H., Lester, B., Ross, B., and Vogt, .,
1984b. Benchmarking of Flow and Transport Codes for Licensing Assistance,
GeoTrans, Inc. and Teknekron, Inc., to be published as a UREG report.
7-4
8-1
APPENDIX
CONWERSION OF INPUT DATA FROM SWIFT O SWIFT I
-
8-2
&M.
IJAIA�,"Pr
1 z 1I 6 S r a
. - -__ __ _ _ _ _ _, __ _ _ _ - _ _ - - - _ _ _ _ _ _ _ _ _ _ _ _ __j~~~~f34S6?#tq a, Z 114jj; jftq~~~~~~~~~~~~~~~~~~~~~~~~~~~~~jjv~ ~ ~
t
Z I
I
S *htREPLKSVD NTOnUTO
717 9
I 0
I ,______
£0 'ICP4VD
11 PS
12 1 uli 116 J1A JItK
13 1 IMP
5 a'
Iii
N-I 1
NPLP N.T NFLCIUNIT LO, _-_
INCHIK' TIE ONT I EK4C AI
I"-t~~~~
Nl-I II
rmrl-. I
20-1 1Ito- I-I I
I10-1-2 I
It0-Z-I I
I
RO-Z-2 II
100-L~~~
e*O 1M0tC&TfS TAT THESE VAIALESARE SITUATED IN REVISCO FORMATF IELS.
R00-2 I
XlA ig 51 I90,- -t - 1 tR If -100-3 1
1
_jq to tI
N-I II
11-1 I
itl-Z 9
11-1.5 UII
N-" INOICATES 4 f Vl5%Lf FOR SwIFT.ll
R 1-
I
I
Rl -5
Z2 I RlJ I
RL-q Z3 t ~~~~~~~~~~~~~~~~~~41-8
I CtRTEC Of TO SPACE LCetRTONS CE!AI4 VARI&4LE NAMES HAVE SEEN 11UHC&TEDI 1…__ _ _ , _ _ , _ ._ _ _ _ - _ _ _ - - - _ _*_ __ _ _ , _ _ _ _ _ _ _ _ _
8-3
l 2 3 4 5 6 7 4~~~~~~~~~~~~~~~~~~~~~Xr56t~qll}*sstSnlz 49st? n1z% s49D oU3456M90tMASo 7040lt14 St.regltz4s,67ss0
* -- *--- - * - * - - - - -~~~~~~~~----- ---
t I
I Z I RIl-1 I
I t
? t I
I 1
2S I Rt-l I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I
30 * Rl-tS I I
11 I Rt-2,6 1
I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I-I
I t13 I Rt-S I
ii I Rt-tS r
1S) * RI-aGI .
36 I RL-I-I 1
34~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ t- ~1-2~ 1I .I
3d I Rl-22 tI S
'.0 * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~R -Z S I ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I
t I
z. t ~FPIt H4DO rHADORl-z6-Z II ,___ ____,**** **.,.:9*,,*** I
'1 I fTUE Flu? FTUZ Rl-26-3 I
'.4 I * 1-27 ...... 1
Rl-ZS-I ...... *I N....
'b I Rl-20-2 .....1
'.t I 11-29 ..... I ......
'.6~~~~~~~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1-30-i I...... I .....
50 * ftI-31 .....I I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
83-4
t 84 5 &I e
I z 3 1 S tzl 7sNq IMU'P. V04o0 31450?iq'z)*s674oI 9 2s6 2345 6 Oz ?A OLZ#shSqoI zr s607aqo
,_________,_____.___,_________._________._________,_________._________,__________.
i I tI51 I ~I-3t I
2 I S Rl-33-1 If I
S) I IRKCKSAYBK ID-3-1 II -I-# -0* -0*- - -
59 1 CR0 IO-I 1
S * -- ---- *--- - -_ -----.. ________..___
21 -I
5b t KCD sCRD SAD O5o OS0o RlD-Z-2 1
5? K OS 1Q-2 1
SE I IR KIIC KPS 'IPS KSIse 4lO-3)- 1 1
Ss I P80 rpeo sSuO 110-3-2 1
60* OIS9 ftL0-4
et I
625 1ZA-I o * I K I
i I
70 RIA-b
I t
1 RIA-02 I
ObI KIt 6? I ftl-3 I
61 I .hA-9 I
eq i COetH RLA-5 II ._______* .K
70 * ISA-b
I I7S I CLA-C K
I JJ2 K I HL R-OS t
I.--.~~~~~~~~~~~~~~~~~~~~~~Rl- It I
75 . 1K5c JOsc COIc iF-Il.___ ___ ___. l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
._ _ __ _ ._ __ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ . _ __ _ _
8-5 Pr
l 2 3 1 5 h a
12 )-sfi7490tl4sbA9otz )*s7490z34sr6 ?6otzj4AhreQ~tz ag0~t}s~sl24hto#---------*--------- -______._______ .- - - - - _ - - - - -
71
7
la
III
I
t II t JI JZ £CHOI ._ _. _ .___. _______.1
t
INC" RI-I II
R -Z
60 '
81I
e8 I
Ia I
6%I
66 1I
U? £
qlI
ed II
69 II
qO
Iq IIQ2I
IRZ-Z.5 t
IAt-3 t
I
RZ-7-3 ;
I
I
I
R2- -L IRZ-1L II
RZ-7-1 KI
RZ-8 II
lRz-9 IK
Pi-to I
I
R2-LI.S II
kZ-L2 ~~I
ITWYPTKqTtXIYlCtYTlN IO TRUD.___ .__ +,_ _.. __. __--- _ ____.---
q
I
Sfp *NJRT~nRtt
b6 I XYXL XYTL XlXL %ZIL0 . _ _ _ _ _ *_ __ _ - - - ______ -- - _ _ _ _ .--------
q i II
Q I II
q I
(050 I080LItR~Z-93IOSO ~dDI ttRZ-1 3
If
K
It-t*ol II
11-IS II
P-.2 II
P-1 II I* -434fI N4U : t IEa OROAI XCG tKtR1. Z-11; 3 .E. C 'Cf 1 TO *IIWI I* _________*------- … .__-_____ . _______ .___ ___ - - _______ .__ __ __ - - - - - -- - -
8-6
MINIMAL CHANGES FOUIEO T CVERT a nTA SET FROM SWIFT I .4t ro
SWIfI I IZ. 3
CARD C14hC$s
P'-z JCD IF.fE 0 AND SHIfT VARIABLES PLC-IUNIt ro T tIcHtA0D LJloO
M-) SPLIT 0aJTo To CRDS, noOP A&&E, D0 NRCm-XuO. KmIEArO,KSLYOsO, NtDO. OItJoO
RI-it-Z ADO FP"IsO., Sfr HA0O &NO 1tuODO AIGCt
Rl-2h-1 ADD A LLNK FtU~aO. FTUVuD, FTUO1
ta-s nOU CnNVHsO.
R2-1 Aon 1RCH USE OfFAtLT8o
12-IL 50n IXVTP lZt, tXIC. tXYZC* 6UO, rUo (ALL SET a 0£
12-I) LCO 1050, 1050. I IIRTO (ALL St * Of
R-14 REVSED VARIABLES FOR EHANCED R&PPINC S NECESSARY
RZ-l§.S RkVISED VAMIABLES FOR EHRNCED APPING S ECESSARY
___________________________________.__________________________- - -
PROPOSED TAB TITLES
INIRODUCTtON
TRANSPrAT WITH DUAL
POROITY
PROB. 1: One Nuclide,
Frac./Prism. Matrix
PROB. 2: One uclide,
Frac./Spher. Hairix
PROS. : Nuclide Chain,
Frac./Prism. Matrix
FLOW WITH UAL POROSITY
PROB. 4: Well-Test in
Dolomui ce
PROB. : Well-Test in
Glacial Aquifer
FL1OW AD TRANS. WITH
AQUFERICONFINNG; BEDS
PROB. 6: FLow tIhru
Aquicards
PROB. 7: Heat Trans. from
Confining Beds
FLOW WITH FREE-
WATER SURFACE
PROB. 8: Steady State
PROD. 9: Transient State
NOTAT tON
REFERENCES
CONVERSION FROM
SWIFT TO SWIFT It