SS Poland

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/342096889

THE SEEPAGE ANALYSIS OF THE EMBANKMENT DAMS OF A FLOOD


RETENTION BASIN IN POLAND

Conference Paper · July 2018

CITATIONS READS

0 829

2 authors:

Burcu Ersoy Ronald Haselsteiner


Björnsen Beratende Ingenieure GmbH Björnsen Consulting Engineers
10 PUBLICATIONS   15 CITATIONS    53 PUBLICATIONS   80 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Woody vegetation on and at flood protection structures View project

Flood Protection Project of the City Nürtingen View project

All content following this page was uploaded by Burcu Ersoy on 11 June 2020.

The user has requested enhancement of the downloaded file.


Commission
internationale Des Grands
Barrages
-------
VINGT-SIXIÈME CONGRÈS DES
GRANDS BARRAGES
Autriche, juillet 2018
-------

THE SEEPAGE ANALYSIS OF THE EMBANKMENT DAMS OF A FLOOD


RETENTION BASIN IN POLAND

Burcu ERSOY, Ronald HASELSTEINER

Hydraulic Engineering Department, BJÖRNSEN CONSULTING ENGINEERS

GERMANY

1. INTRODUCTION

All kind of embankment fill dams are faced with seepage body during
impoundment and during operation. Therefore, seepage control plays an
important role at the design stage to prevent uplift pressures, instability of the
downstream slope, piping through the embankment and/or foundation and
potential suffusion and erosion processes. In order to guarantee safe seepage
conditions for the anticipated load cases, the seepage control design should
comprise sealings, dam body specifications, special drainage elements, and their
characteristics.

The paper provides the main information and results of the seepage and
stability analysis of a case study and aspects considering potential suffusion and
erosion processes.
2. PROJECT OVERVIEW – CASE STUDY

A dry/green flood retention reservoir is under construction in Poland since


2013. The project includes embankment dams (head, right and left dam). The
lengths of all embankments show approx. 22 km. The reservoir volume is 180
Mio. m³ with a perspective to be enlarged to 300 Mio. m³. The maximum height of
the embankment dams is over 11 m in the area of head dam. The maximum
water level is specified at a level of 195.2 m a.s.l. and the crest elevation is
defined at 197.5 m showing a freeboard of 2.3 m. A selected section of the dam
(head dam) is given in Fig. 1. Within specified sections the sand-gravel material
was selected as main dam fill material. Gravel material is used as filter material
and the seepage control function is provided by a geosynthetic clay liner (GCL).
In addition, gravel columns have been placed within the foundation area where
weak formations are located beneath the future dam body.

Fig. 1
Main design section of embankment dam (head dam) used for the seepage and
stability analysis (sketch)

3. SEEPAGE AND STABILITY ANALYSES

3.1. PREPARATION OF THE FINITE ELEMENT MODEL AND BOUNDARY CONDITIONS

The finite element model (FEM) is defined as a mathematical simulation of


a real physical process. For the FEM a typical dam cross section with a height of
11 meters was considered (Fig. 1). The model size (width) was defined in
consideration of the dam height. As minimum size two times the dam height to
up- and downstream were defined so that the model boundaries are not
influencing the seepage conditions within the dam. This approach results in a
model width of 172 m. The modelled underground depth was selected also with a
minimum of two times the dam height. The maximum water level is applied at
195.20 m a.s.l. which is 8.7 m above the upstream ground surface level. The
geological formation consisting of a clay layer, sand-gravel, and London clay
layers are integrated in the model. The thicknesses are established as 6 m, 7 m,
and 7 m for clay layer, sand-gravel and London clay, respectively. The thickness
and depth of the cut-off wall are applied with 60 cm and up to 13 m respectively.
The thickness and depth of gravel columns are assumed as 80 cm and 9 m
respectively. The distance between the gravel columns is assumed as 2 m
(center to center) so that in total 33 gravel columns are considered in the model
section. For a better possibility of evaluation the reference coordinate system
used the origin (x, y) = (0, 0) at the upstream dam toe. The final view of the
numerical model that used in seepage analysis is given in Fig. 2.
To determine the seepage flow – steady state conditions were considered –
through the embankment body, a FEM requires the definition of boundary
conditions. Three hydraulic boundary conditions are applied for the numerical
model as follows:
 ‘Constant head’  Upstream face  at a height of 195.20 masl (water
level with 8.7 m above the terrain surface)  dark blue in Fig. 2
 ‘Constant head’ Downstream face  the maximum water level that is
at a height of -0.44 m  blue downstream line in Fig. 2
 ‘Seepage exit’, ‘Total flux type’  Potential seepage exit areas along the
downstream slope  the ground water flow as 0 m3/s  red line in Fig. 2

Fig. 2
Finite element model for the seepage analysis

3.2. DEFINITION/DETERMINATION OF THE GEO-HYDRAULIC AND SHEAR STRENGTH


PARAMETERS

The next step in developing a numerical model is to assign materials to the


regions as defined in the numerical model before. The geo-hydraulic and shear
strength parameters are the key parameters for seepage and stability analyses.
To assign reliable inputs to the numerical model, geo-hydraulic parameters are
implemented that are included in different books and papers, e.g., in Haselsteiner
(2007) [8] and the referenced literature sources. In Haselsteiner (2007) [8] the
geo-hydraulic parameters are compiled from 17 different literature sources. The
compiled geo-hydraulic parameters are categorized in accordance with typical
embankment dam zones such as gravel drain, gravel fill used for the dam body
and underground, sand material, surface sealing and alluvial clay/loam.
‘Saturated only’ option was selected for the steady state seepage analysis
and for the material shear behavior, the ‘Mohr-Coulomb’ model was selected.
The geo-hydraulic and shear strength parameters that are implemented to the
numerical model are given in Table 1.
Table 1
Geo-hydraulic and shear strength parameters for the zones of the embankment
dam
geo-hydraulic shear strength
Saturated Unit
Color of Permeability Anisotropy Cohesion Friction
Volumetric Weight
No Type Name Zone in kS A=ky/kx c' Angle j'
Water g
Model [m/s] [-] [kPa] [°]
Content [-] [kN/m³]
-9
1 Cohesive Soil Buttress 10 0.5 0.325 19 4 31.0
Embankment Zone

-3
2 Sand-Gravel 10 0.5 0.225 17 0 37.5
-3
3 Protection Layer (Up) 10 0.5 0.200 17 7 31.0
-7
4 Geomembrane Clay liner 10 0.5 0.400 20 4 25.0
-2
5 Drainage / Gravel 10 1 0.195 18 0 35.0
-9
6 Cut-off Wall 10 1 0.500 25 200 40.0
-4
7 Gravel Column 10 0.5 0.300 22 3 40.0
-9
8 Clay Overlay 10 0.5 0.325 19 4 31.0
Layers
Soil

-3
9 Sand-Gravel 10 0.5 0.225 17 0 37.5
-11
10 London Clay 10 0.5 0.325 20 5 28.0

3.3. DESIGN SITUATIONS AND LOAD CASES

The design, construction and operation of a dam with a reservoir need to


be developed in consideration of its long-term performance that is affected by
boundary conditions such as reservoir water level, traffic loads, performance of
the dam zones, climate impacts, etc. Therefore, corresponding design situations
and load cases need to be defined during the design stage especially for the
seepage and stability analyses.
Load cases consist of a set of conditions that are taken into consideration
in the dam safety analysis. For the analyses, the “Deutsches Institut für Normung
e.V.” [1], [2], [3] and DWA [4] codes are considered which regulate dam
engineering aspects. The load cases and the revision of guidelines are given in
Table 2. Seven load cases are considered within the seepage and stability
analyses. The load cases showing unsteady seepage, e.g., for rapid drawdown
conditions,and operation earthquake conditions are not taken into consideration
for the analyses since they were considered irrelevant.
In addition to the load cases, four design conditions were considered for the
numerical model to evaluate the long-term performance of the designed sealing
(geo-synthetic clay liner + cut-off wall) and soil treatment systems, e g., gravel
columns as given in Table 3.
Table 2
Different load cases (design situations) for seepage and stability analyses
according to DIN (adapted after DWA, 2015)
no sidered
yes Con-

Load Case
Design Situations
according to Resistance
No. according DWA M-522/2015 Loads
DIN 19700-10,-11,- Dam status/situation
adapted according EC 7
12
DS-P Normal Load Cases Temporary/ Water Table
Seepage Description
Permanent Design Situations traffic loads Conditions
1 P.1 Load Case 1.1 p1 = 25 kN/m² ZV = 195.20 masl steady All compounds and elements
Full Flood Water Level ZV Full Flood Water 2D seepage are fully working.
1

Level conditions
2 P.4 Not considered within p1 = 25 kN/m² Drawdown starting from unsteady All compounds and elements
Operational drawdown from ZV the requirements of ZV. conditions are fully working.
DIN 19700. analytic approach
(not considered)
DS-T Temporary Rare Load Cases Temporary/ Water Table Seepage Description
Design Situations traffic loads Conditions
3 T.1 Load Case 2.1 p1 = 25 kN/m² ZH1 = 196.20 masl steady All compounds and elements
Flood water level at ZH1 Flood level at ZH1 2D seepage are fully working.
2

conditions
4 T.2 Load Case 2.2 p1 = 25 kN/m² Rapid drawdown with unsteady All compounds and elements
Rapid drawdown from water level maximum velocity conditions are fully working.
ZV starting from ZV. analytic approach
(not considered)
5 T.3 Load Case 2.3 p2 = 33,3 kN/m² Partial impoundment steady All compounds and elements
Extraordinary operation situation Half of the dam height 2D seepage are fully working.
3

Zpartial = 192.00 masl conditions


6 DS-E Load Case 2.4 "For dry basins an analysis considering the load "operational earthquake" can be
A) neglected." (DIN 19700-12) (not considered)
Operational earthquake
DS-A Extreme Load Cases Temporary/ Water Table Seepage Description
Accidental Design Situations traffic loads Conditions
7 A.1 Load Case 3.1 p1 = 25 kN/m² ZH2 = 196.70 masl steady All compounds and elements
Flood water level at ZH2 Flood level at ZH2 2D seepage are fully working.
4

conditions
8 A.2a Load Case 3.2 p1 = 25 kN/m² ZV = 195.20 masl steady Restricted functionality to
Restricted functionality of 2D seepage complete malfunction of the
5

sealing/male function conditions sealing system.


9 A.2b Load Case 3.3 p1 = 25 kN/m² ZV = 195.20 masl steady Restricted functionality to
Restricted functionality of to drain 2D seepage complete malfunction of the
6

conditions toe drain.


10 A.3 Load Case 3.4 p1 = 25 kN/m² ZCrest = 197.50 masl steady All compounds and elements
B) 2D seepage are fully working.
7

Crest Water Level


conditions
DS-E Earthquake load Temporary/ Water Table Seepage Description
Design Earthquake case traffic loads Conditions
Load Case 3.3 not considered not considered not considered not considered
11 Design Earthquake

Notes:
In DWA M-522/2015 the operational earthquake is not considered within the design situations.
The design situation considering the crest water level is usually not considered within the mentioned load situation but is frequently analysed hand in hand with the

Table 3
Considered design conditions together with considered load cases/design
situations (DS)
Load Cases/Design
Design Conditions
Situations No.
Name Abbreviation 1 2 3 4 5 6 7
Full Sealing with
Load Case A.2a (LCA2a)
Load Case A.2b (LCA2b)

FS+GC
Load Case P.1 (LCP1)
Load Case T.1 (LCT1)
Load Case T.3 (LCT3)
Load Case A.1 (LCA1)

Load Case A.3 (LCA3)

Gravel Column
Full Sealing without
FS-GC
Gravel Column
Open Sealing with
OS+GC
Gravel Column
Open Sealing without
OS-GC
Gravel Column
3.4. RESULTS AND INTERPRETATION OF SEEPAGE ANALYSIS

The 2D steady state seepage analysis is performed by using SEEP/W


software program that is a part of GeoStudio 2016 software package. As a result
of 28 seepage analyses considering the different design situations and load
cases, following results are obtained.
Generally, the differences obtained for the different dam/design conditions
(FS, OS, with/without GC) regarding total head and pore water pressure are not
substantial. Therefore, the application of sand-gravel columns does not worsen
the seepage situation critically. But also does not show a benefit in consideration
of the performance of the embankment dam.
The comparison of the results for the different design/dam conditions and
for selected load cases are given in Fig. 3. The load cases LCT1, LCA1, LCA2a
and LCA2b are selected in order to compare the design/dam conditions. In
general, the results of the seepage line and pressure head reflect the same
trends as theoretically expected.
For the load cases LCA2a “Malfunction of the sealing” and LCA2b
“Malfunction of drain”, the drain capacity is not enough to drain the dam body. In
addition, an unfiltered exit is the result that may lead to backward erosion.

Fig. 3
The comparison of different load cases and design situations in form of on line of
seepage and pressure head
The sealing of the actual design does not reach to the crest level.
Therefore, a considerable seepage flow may occur within the unsaturated zone,
bypassing the sealing in the crest area. Especially, for high water levels this
shows an adverse effect.
The gravel columns show a minor effect on the underground seepage
conditions. The gravel columns do not show a practical effect when the sealing
system is not working, anyway, or the “open sealing situation” is established in
the model. The gravel columns increase the seepage flow amount, pressure
head and seepage line when the sealing system is working in full sealing (FS)
conditions.

3.5. RESULTS AND INTERPRETATION OF THE STABILITY ANALYSIS

The slope stability analysis is performed by applying the limit equilibrium


analysis (LEA) method in order to determine the factor of safety (FoS) only for the
downstream embankment slope (using SLOPE/W software, GeoStudio 2016). The
FoS values are obtained only for the sliding slope stability considering the
downstream slope. Results are presented in Table 4.
The required FoS ([1], [2], [3], [5]) are relatively below the calculated
values. In addition, thanks to the conservative dam design and flat slopes in
respect to the limited height of the embankment the static stability is not a matter
of concern. The FoS range show up to 30 % safety margin.
Table 4
FoS Evaluation of the Downstream Embankment Slope Slip Stability
Load Cases
LCA2a

LCA2b
LCA1

LCA3
LCP1

LCT1

LCT3

Source Required Factor of Safety


BS I BS II BS III
DIN 19700/2004
1.3 1.2 1.1
Fell et al., 2005 1.5 1.3 - - - - -
USACE 1.5 1.4 1.3 - - - -
USBR 1.5 1.5 - - - - -
FERC 1.5 1.4 - - - - -
Design condition Calculated Factor of Safety
FS + GC 1.8 1.8 1.7 1.7 1.7 1.7 1.6
FS - GC 1.8 1.7 1.7 1.7 1.5 1.6 1.5
OS + GC 1.8 1.7 1.7 1.7 1.7 1.7 1.6
OS - GC 1.7 1.6 1.7 1.5 1.5 1.6 1.4

4. HYDRODYNAMIC SOIL DEFORMATION PROCESSES

Fell & Fry [6], [7] developed a process dependent approach for assessing
erosion and suffusion processes. The consideration of more assessment steps in
order to obtain a full risk assessment approach also in consideration of important
aspects such as system identification, detection, measures, etc. [9] supports of a
better understanding and more detailed risk assessment.
Beneath the assessment steps (initiation, development, continuation,
breaching), which are directly addressing the particle transport, the pre- and post-
phases need to be investigated and assessed as well. For the assessment it is
also important to localize the transport processes. Within this localization step
erosion paths shall be defined which are strongly depending on the dam design
and the underground situation. In total three erosion paths are identified as dam
body erosion path (DB EP), underground erosion path 1 (UB EP 1) and
underground erosion path 2 (UB EP 2). Along the erosion paths different sorts of
processes (A to E) are evident which are also checked within the main four
phases of the erosion and suffusion assessment (see Fig. 4).

Fig. 4
Defined erosion paths within the embankment dam section

The continuation of erosion shall be assessed by mean hydraulic gradients


derived from the seepage analysis. The definitions of the hydraulic head
difference and the flow path length are shown in Fig. 5.
The determined hydraulic gradients are given in Table 5. The critical
gradient for the present sand-gravels should be approx. ic = 0.20 to 0.40. For
finer sands this can decrease to ic = 0.15 and for silts to ic = 0.10 or even less.
For the dam body erosion path the material shows the firstly mentioned
gradients. However, the gradients within the dam body (Table 5) are generally
small i < 0.20, except for load case LCA3 where it reaches a value of i = 0.30.
Thus, for reaching the crest water level also backward erosion may continue
since hydraulic forces are strong enough. For the other load cases the erosion
through the dam body should not be a problem due to low gradients.

Fig. 5
Determination of the Hydraulic Head difference and Flow Path length using the
Total Head results of the Seepage Analysis (here: LCP1, OS - GC)
For the crest water level (LCA3) the hydraulic gradients are strong enough
to continue the erosion process through the dam body. This is a result of the
seepage control design which does not meet the requirements to European
codes and guidelines, especially German design philosophy. Processes within
the underground are also considered to be not critical thanks to the protective
clay layer and the conservative dam design.
Table 5
Hydraulic and critical gradients for the analyzed load cases and design situations
Load Cases

LCA2b
LCA2a
LCA1

LCA3
LCP1

LCT1

LCT3
Erosion Mean actual hydraulic gradients Design/dam
path i [-] condition
Critical hydraulic gradient for sand-gravels
ic = 0.20 to 0.40 (Brandl & Hofmann, 2006)
Dam
0.08 0.12 0.02 0.13 0.16 0.09 0.30 OS without GC
body
Underground
0.08 0.09 0.05 0.10 0.08 0.08 0.10 OS without GC
1
Underground
0.09 0.10 0.04 0.12 0.09 0.09 0.12 OS with GC
2

Suffusion plays a minor role, but as a result of checking selected suffusion


criteria some of the applied sand-gravels in the dam fill as well as in the
underground might be susceptible to suffusion.
The applied monitoring concept for a 22 km long dam is not able to
contribute to an effective detection of potential erosion processes. Intervention
measures to stop the erosion through the dam body, once it occurs during
accidental load cases, are difficult to realize. Due to the cohesionless fill material
in the dam body breaching should proceed relatively fast.
Following recommendations were given. The sealing should be extended to
the crest elevation. The drain should be redesigned and extended to be
connected to the drainage ditch. The gravel columns do not show a critical
negative effect on the design and the dam behavior in respect to erosion and
suffusion risks but, therefore, can be spared.

5. CONCLUSION

The dam shows an unsatisfactory seepage behavior resulting in an


unfiltered seepage exit which is the first phase to backward erosion. This risk is
mainly present for the dam body, which shows a design which does not comply
with modern dam design criteria and modern dam safety philosophy. Therefore,
the application of an additional sealing or drainage element and redesign of the
downstream toe drain are considered to be required. The static stability of the
downstream slope is not of concern thanks to the applied flat slope. Erosion risks
are mainly related to the embankment dam body. Within the underground erosion
risks are tolerable.
REFERENCES

[1] DIN 19700-10 Stauanlagen, Teil 10: Gemeinsame Festlegungen.


Deutsches Institut für Normung e.V. (DIN), Berlin, Juli 2004.
[2] DIN 19700-11 Stauanlagen. Teil 11: Talsperren. Deutsches Institut für
Normung e.V. (DIN), Berlin, Juli 2004.
[3] DIN 19700-12 Stauanlagen. Teil 12: Hochwasserrückhaltebecken.
Deutsches Institut für Normung e.V. (DIN), Berlin, Juli 2004.
[4] DWA M 522 Kleine Talsperren und kleine Hochwasserrückhaltebecken.
Deutsche Vereinigung für Wasserwirtschaft, Abwasser und Abfall e.V.,
2015.
[5] FELL, R. ET AL. Geotechnical engineering of dams. A. A. Balkema
Publishers, Leiden, London New York Philadelphia Singapore, 2005.
[6] FELL, R. ET AL. A Framework for Assessing the Likelihood of Internal
Erosion and Piping of Embankment Dams and Their Foundations. Internal
Erosion of Dams and their Foundations, Fell & Fry (editors), Taylor &
Francis Group, London, pp. 65-70, 2007.
[7] FELL, R., FRY. J. J. The State of the Art of Assessing the Likelihood of
Internal Erosion of Embankment Dams. Water Retaining Structures and
their Foundations, Taylor and Francis, London, pp. 1 - 24, 2007.
[8] HASELSTEINER, R. Hochwasserschutzdeiche an Fließgewässern und
ihre Durchsickerung. Berichte des Lehrstuhls und der Versuchsanstalt für
Wasserbau und Wasserwirtschaft, Technische Universität München, Nr.
111, 2007.
[9] HASELSTEINER, R. Die Beurteilung von hydrodynamischen
Bodendeformationsvorgängen in Dämmen und Deichen - Ein integraler
Ansatz. 3. Symposium Sicherung von Deichen, Dämmen und
Stauanlagen, Universität Siegen, S. 289-338, 2009.

SUMMARY

The seepage control concept that comprises sealing, dam body and
drainage elements in order to guarantee safe seepage conditions for the
anticipated load cases should be developed where remarkable seepage
conditions are expected through the dam body. As long as the seepage
conditions are not controlled and hydraulic gradients occur, erosion may occur in
form of backward erosion in the underground/subsoil and along all contact
borders/interfaces of different materials in the dam body.
The paper presents the seepage and stability analyses of a case study
to emphases the importance of seepage control design and the assessment of
hydrodynamic soil deformation process. Based on the analyses’ results,
recommendations and conclusions are presented to overcome and to improve
the specific drawbacks and inherent risks of the present dam design.

View publication stats

You might also like