Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Comparison of 2D and 3D Seepage Model Results for Excavation Near Levee Toe

Rebecca L. Money, P.E.1


1
Project Engineer, Kleinfelder, Inc., 3077 Fite Circle, Sacramento, California 95827; PH (916) 366-1701; email:
bmoney@kleinfelder.com
Downloaded from ascelibrary.org by University of California, San Diego on 08/17/15. Copyright ASCE. For personal use only; all rights reserved.

Abstract
With various two-dimensional (2D) and now three-dimensional (3D) seepage analysis programs available it is
important to understand how results from 2D and 3D models compare and when a 3D analysis may be warranted.
This paper compares the results of SEEP/W© and Seep3D© software programs developed by Geo-Slope
International. The project analyzed is a proposed excavation near the landside toe of the Sacramento River levee in
Sacramento, California. Based on historical underseepage and through seepage emanating from this area during
periods of high river stage, a seepage analysis was required by the City Building Department. The results of a 2D
and 3D analysis were compared for an infinitely long excavation to calibrate the 3D model and then a limited
excavation was modeled in 3D. This comparison revealed calculated seepage exit gradients 50% larger using the
3D model compared to the 2D model results.

Introduction

This paper discusses the results of an analysis comparing seepage results utilizing two-dimensional (2D) and three-
dimensional (3D) modeling programs. The analysis was performed first on a theoretically infinite long excavation
paralleling the toe of the adjacent Sacramento River levee. The project geometry was entered into the programs
SEEP/W© and Seep3D© and the resulting total head contours and seepage gradients were compared. Secondly, an
excavation with limited horizontal boundaries was analyzed in Seep3D© and compared to the previous two models
in order to determine when a 3D analysis is warranted.
The project is located in a residential subdivision with backyards bordering the levee toe within an area
known for historical seepage problems during periods of high river stage. Numerous residences currently have
swimming pools in their backyard, however a new City requirement mandates that a seepage evaluation be
performed for all new pool construction requests for properties abutting the levee.

Seepage Analysis

Two seepage analyses were performed. The first analysis compared seepage results encountered in an open trench
excavated at the landside levee toe extending infinitely in each direction parallel to the levee. This analysis was
compared using SEEP/W©, a 2D modeling program, and Seep 3D©, a 3D modeling program.

Description of Seepage Analysis Programs


This seepage analysis was performed using the programs SEEP/W©, developed by Geo-Slope International (1998)
and Seep3D©, developed by Geo-Slope International (2001). A steady-state analysis procedure with finite element
modeling was used. The program allows multiple soil types to be used with varying anisotropic hydraulic
conductivity parameters.
The boundary conditions used in this model include fixed-head boundary conditions set to the 100-year
flood level (Elevation 27) along the vertical waterside edge and waterside slope of the levee, fixed head boundary
conditions with a value equal to the ground surface elevation along the vertical landside edge of the model to
replicate a worst case groundwater condition (Elevation 15), no-flow boundary conditions along the bottom of the
model, variable head boundary conditions (elevation head) on the top of the model extending from the high water
line on the waterside to the edge of the model on the landside. The excavation is modeled using both variable head
and no-flow boundary conditions to model when the excavation is open before construction and after the pool is
constructed and lined with impermeable material. In Seep3D© the variable head boundary condition is modeled
using a potential seepage option.

Selection of Geometry
Seepage analyses were performed on one representative cross section of the conditions encountered at the proposed
excavation. The soil profile for this analysis was determined after reviewing the subsurface conditions performed by

Copyright ASCE 2006 GeoCongress 2006


GeoCongress 2006
Kleinfelder, Inc. Topographical data obtained by the US Army Corps of Engineers Comprehensive Study was
utilized to draw the levee cross section. The levee geometry was terminated at the waterside toe of the levee for
simplification purposes. The Sacramento River was modeled at a 100-year water surface elevation (27 feet)
determined by MBK Engineers. The subsurface conditions were modeled based on the information obtained by
Kleinfelder during the site investigations and associated laboratory testing. The soil profile utilized included a 15
foot high silty sand levee overlying foundation/blanket materials consisting of 22 feet of near surface sandy clay
underlain by 28 feet of poorly graded sand. The levee geometry included approximately 3H:1V slopes waterside
and landside with a 20-foot wide crown. A soil/cement/bentonite slurry cutoff wall has been constructed at this
Downloaded from ascelibrary.org by University of California, San Diego on 08/17/15. Copyright ASCE. For personal use only; all rights reserved.

location through the middle of the levee to reduce through seepage. This wall is approximately 2 feet wide and 15
feet deep extending along the centerline of the levee. A 10-foot wide infinite excavation was located between 17
and 27 feet from the landside toe. A 20-foot long swimming pool excavation was modeled in half with analysis
dimensions of 10 feet long by 10 feet wide by 5 feet deep.

Determination of Soil Parameters


Coefficients of hydraulic conductivity (permeability) for the soils in the cross section were chosen based on ranges
of values presented by Terzaghi and Peck (1967), and correlation relationships based on grain size distribution
described in EM-1110-2-1913 (USACE, 2000), and NAVFAC DM-7.01 (NAVFAC, 1986). Conservative values
where chosen based on these means and are shown in Table 1, below. The analyses were performed using a soil
anisotropy ratio (kv/kh) of 0.25 for all native soil layers and the levee fill, and an anisotropy ratio (kv/kh) of 1/1 for
the seepage cutoff wall.

Table 1. Horizontal Hydraulic Conductivity for Various Soil Types


Selected Values of Anisotropy
Range of Coefficients of Horizontal Hydraulic Ratio
Soil Type Horizontal Hydraulic Conductivity
Conductivity (cm/sec)
cm/sec ft/day (kv/kh)
-5 -3 -3
Levee Fill (SM) 10 to 10 10 2.8 0.25
Seepage Cutoff Wall 10-7 to 10-5 10-6 0.0028 1
Sandy Clay (CL) 10-7 to 10-3 10-4 0.28 0.25
-3 -2 -3
Sand (SP) 10 to 10 5x10 14 0.25

Discussion of Seepage Analysis Results


First, comparison of the two programs was performed by modeling existing conditions and infinitely long
excavations. The infinitely long excavations were modeled using two different boundary conditions. Boundary
conditions within the excavation along the cut faces were modeled either by allowing water to seep into the
excavation (Q=Review by Elevation (2D) or Potential Seepage (3D)) or by no flow boundaries (Q=0). These
boundary conditions modeled the open excavation prior to construction and post construction conditions with the
excavation lined with impermeable materials creating a dam effect. Comparing these models would allow the
modeling programs to be calibrated to each other before creating the finite swimming pool model. The resulting
gradients of this calibration comparison varied by about 12%. The second portion compared 3D analysis of a
limited excavation, in this case a swimming pool, at the toe of the levee to the 2D infinitely long excavation results.
Comparison to the 2D model showed differences ranging from 44 to 122% for an open excavation and from 50 to
150% for a lined excavation. Refer to Table 2 below for the analysis results. Refer to Figures 1 through 3 below for
printouts showing the vertical gradient contours for these three cases.

Table 2. Calculated Vertical Exit Gradients


Existing Open Excavation Lined Excavation
Analysis Conditions (Gradient at bottom (Gradient at bottom
(w/o excavation) of excavation) of excavation)
2D 0.40 0.9 0.20
3D – Infinitely Long 0.45 1.3 0.30
3D – Limited Excavation -- 2.0 0.50

Copyright ASCE 2006 GeoCongress 2006


GeoCongress 2006
Description: Silva Ranch Way
Hydraulic Conductivity
Comments: Seepage Analysis
1 - Levee Fill (SM), k=2.8 ft/day, 4h:1v
File Name: Silva Ranch Way 5' infinite, open exc.sep
2 - Clay (CL), k=0.28 ft/day, 4h:1v
Last Saved Date: 7/27/05
3 - Sand (SP), k=14 ft/day, 4h:1v
Last Saved Time: 8:25:56 AM
4 - Open Excavation
Analysis Type: Steady-State
5 - Cutoff Wall, k=0.0028 ft/day, 1h:1v
Analysis View: 2-D
Downloaded from ascelibrary.org by University of California, San Diego on 08/17/15. Copyright ASCE. For personal use only; all rights reserved.

5 Q=Review by Elevation
WSE = 27'
40
0.90
30 4
20 1
Elevation (feet)

H=27
10
0 2 0.5 H=15
-10
-20 3
-30
-40
-50 Q=0
-60
-100 -80 -60 -40 -20 0 20 40 60 80 100 120 140 160 180 200 220 240 260

Horizontal Distance (feet)


TOTAL HEAD CONTOURS
Figure 1. Vertical gradient contours for infinite long, open excavation, using SEEP/W.

1.3

0.3
0.5

Figure 2. Vertical gradient contours for infinite long, open excavation, using Seep 3D.

Copyright ASCE 2006 GeoCongress 2006


GeoCongress 2006
Downloaded from ascelibrary.org by University of California, San Diego on 08/17/15. Copyright ASCE. For personal use only; all rights reserved.

2.0

0.5

Figure 3. Vertical gradient contours for limited size, open excavation, using Seep 3D.

Comparison of 2D and 3D seepage results at another larger project nearby was also performed. This project is
located adjacent to the Sacramento River levee where a sump and pumping station are located adjacent to the levee.
The sump is approximately 90 feet wide parallel to the levee and extends perpendicular from the landside levee toe
over 1,000 feet, but was only modeled to 300 feet. The sump depth ranged from 25 feet below ground surface at the
toe of the levee to 17 feet deep at a distance of 300 feet from the levee toe. Several boils had been observed at the
bottom of the unlined sump intake channel. Initial 2D analysis produced seepage exit gradients that showed
acceptable seepage conditions. However, a 3D analysis produced dramatically different seepage results that were
consistent with gradient conditions which would be expected to produce boils such as previously observed.

Conclusions
©
The results shown above illustrate there is a difference between the two modeling programs and the Seep3D results
appear to more accurately reflect seepage gradient conditions within spaces that have variable geometry over short
distances. The sump project, possibly because of the size of the excavation has shown significant differences in the
results especially ranging from acceptable results using 2D modeling to severe seepage gradient conditions from the
3D modeling that matches site observations.

Acknowledgements
I would like to acknowledge Mr. Ray Costa, Jr., G.E., for his review and assistance during this analysis. I would
like to thank MBK Engineers and the Sacramento Area Flood Control Agency (SAFCA) for sharing project
information.

References
Geo-Slope International (1998), SEEP/W software, Version 4, Calgary, Alberta, Canada.
Geo-Slope International (2001), Seep3D software, Version 1, Calgary, Alberta, Canada.
Kleinfelder (2004), “Groundwater Seepage Evaluation Report, Proposed Residential Swimming Pool, 7724 Silva
Ranch Way, Sacramento, California.” File No. 20506-391, prepared for SAFCA, dated June 16, 2004.
MBK Engineers (2001), “Sacramento River Water Surface Profile.” February 2001.
NAVFAC (1986), “Soil Mechanics,” Design Manual 7.01, September 1986, Department of the Navy, Naval
Facilities Engineering Command, 200 Stovall Street, Alexandria, Virginia.
Terzaghi, K., and Peck, R.B. (1967), Soil Mechanics in Engineering Practice, Wiley, New York.
USACE (2000), “Design and Construction of Levees,” US Army Corps of Engineers Engineering Manual EM 1110-
2-1913, dated April 30, 2000.

Copyright ASCE 2006 GeoCongress 2006


GeoCongress 2006

You might also like