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

net/publication/245297469

Geobag Performance as Scour Countermeasure for Bridge Abutments

Article  in  Journal of Hydraulic Engineering · April 2007

DOI: 10.1061/(ASCE)0733-9429(2007)133:4(431)

CITATIONS READS
27 2,974

5 authors, including:

Emilio Martinez Robert Ettema

EDELCA Colorado State University
8 PUBLICATIONS   34 CITATIONS    229 PUBLICATIONS   2,861 CITATIONS   

SEE PROFILE SEE PROFILE

Brian Barkdoll
Michigan Technological University
129 PUBLICATIONS   1,852 CITATIONS   

SEE PROFILE

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

Spencer Dam Failure View project

Port of Nome, Alaska View project

All content following this page was uploaded by Robert Ettema on 21 June 2017.

The user has requested enhancement of the downloaded file.

 Geobag Performance as Scour Countermeasure
for Bridge Abutments
Recep Korkut1; Emilio J. Martinez2; Reinaldo Morales3; Robert Ettema4; and Brian Barkdoll5

Abstract: This paper presents observations and data from a sequence of laboratory experiments conducted to evaluate geobags as a
countermeasure to protect bridge-abutment foundations from failure attributable to scour of the alluvial-river channel in which they are
placed. Geobags comprise geotextile cloth bags filled with local sediment or concrete. The experiments focused on the performance of
geobags placed as an apron around pile-supported wing-wall abutments retaining erodible embankments, and subject to live-bed and
clear-water flow conditions. Though an apron of geobags is shown to substantially reduce or eliminate scour immediately at the abutment,
the apron must be formed flexibly of linked geobags. Moreover, a performance concern is that the apron may shift scour to a location
flanking or downstream of the apron, and in so doing imperil a nearby pier or riverbank. The experiments indicate the importance of
protecting the embankment region beneath and immediately behind the abutment’s pile cap. Live-bed conditions proved to be the more
critical for abutment protection, owing to the capacity of dunes to destabilize geobags around the edges of the apron. Design guidelines
are given and include using current riprap configurations for sizing and placing geobags.
DOI: 10.1061/共ASCE兲0733-9429共2007兲133:4共431兲
CE Database subject headings: Scour; Measurement; Bridge abutments.

Introduction This paper presents the findings of a laboratory flume study

Geobags, also called geo-synthetic containers, are quite com-
monly used as an erosion countermeasure in coastal engineering
applications 共e.g., PIANC 1992; Ray 1977; Pilarczyk 2000;
Heibaum 2002, 2004兲. They are used in lieu of riprap stone or
other armor cover, such as cable-tied blocks, that in certain re-
gions can be difficult and expensive to obtain The potential ad-
vantages of the geobags are that they are readily transported
共when empty兲; they can be filled with local sediments and soils
共sometimes concrete兲; additionally, they can be formed to a range
of sizes—geobags can be literally tailored to fit specific applica-
tion situations 共e.g., Pilarczyk 2000; Heibaum 2002, 2004兲.
Geobags, however, rarely have been used as a scour countermea-
sure to prevent erosion or scour of the abutments of bridges span-
ning rivers and streams.
1
Graduate Student, Dept. of Civil and Environmental Engineering,
IIHR Hydroscience and Engineering, The Univ. of Iowa, Iowa City, IA
52242.
2
Graduate Student, Dept. of Civil and Environmental Engineering,
IIHR Hydroscience and Engineering, The Univ. of Iowa, Iowa City, IA
52242.
3 may have to be repaired after a major flood-flow event. In this
Graduate Student, Dept. of Civil and Environmental Engineering,
IIHR Hydroscience and Engineering, The Univ. of Iowa, Iowa City, IA
52242.
4 reliant on the protection afforded by a scour countermeasure such
Professor, Dept. of Civil and Environmental Engineering, IIHR
Hydroscience and Engineering, The Univ. of Iowa, Iowa City, IA 52242.
5
Associate Professor, Dept. of Civil and Environmental Engineering,
Michigan Tech Univ., Houghton, MI 49931.
Note. Discussion open until September 1, 2007. Separate discussions
must be submitted for individual papers. To extend the closing date by
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and pos-
sible publication on September 9, 2004; approved on July 24, 2006. This
paper is part of the Journal of Hydraulic Engineering, Vol. 133, No. 4,
April 1, 2007. ©ASCE, ISSN 0733-9429/2007/4-431–439/$25.00.
that investigated the use of geobags placed as scour protection
around bridge abutments in alluvial channels. The experiments
were aimed at addressing the following questions:
1. Are large geobags, fitted as an apron around bridge abut-
ments, useful as a short-term means to prevent abutment
scour?
2. How do geobags perform as an alternative to riprap or cable-
tied blocks for preventing abutment scour?
3. Can geobags be used as a form of filter underlay in conjunc-
tion with riprap or cable-tied blocks?
The findings comprise general observations about geobag perfor-
mance, data on scour depths, and lead to a simple design guide
for deploying geobags. The experiments were conducted using a
simulated wing-wall abutment 共Fig. 1兲, but their findings are ap-
plicable directly to other abutment shapes.
At the outset of this paper, it is important to state directly that
no scour countermeasure, including geobags, can be used without
its performance being regularly monitored; in this regard, all
highway bridges in the United States are subject to biennial, or
more frequent, inspections. Indeed, in the United States, it is fully
expected that an armoring countermeasure, such as a riprap apron,
regard, bridge design guides in the United States expressly state
that the structural integrity of a new bridge must not be critically
as a riprap apron.
The writers’ experiments focused especially on scour protec-
tion for wing-wall abutments, of the general type shown in Fig. 1.
Wing-wall abutments commonly are used for small, single-, or
two-span bridges. Given the large number of small bridges, espe-
cially in the U.S. Midwest, the great majority of abutment failures
have occurred for small single- or double-span bridges that com-
monly have wing-wall abutments. Fig. 2 depicts a typical failure
example. In most cases of subsequent bridge repair, and now in

JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007 / 431

 Fig. 1. A wing-wall abutment fitted with a riprap apron
the design of new bridges, it is usual to consider use of a protec-
tive armor layer placed to prevent erosion of the channel bed and
bank around abutments, as shown in Fig. 1.
The evident success experienced with geobags in various ap-
plications of erosion protection prompts consideration of geobag
use as scour protection for bridge abutments, and leads to the
foregoing questions. It is important to note that no prior flume
experiments have investigated scour at a wing-wall abutment
founded on piles and retaining an earth-fill embankment.
Flume Experiments
Flume experiments were conducted using a 27.4-m-long,
0.91-m-wide, and 0.45-m-deep sediment recirculating flume.
Wing-wall abutments were replicated at a scale corresponding to
about 1:40 of the width of abutments typical of two-lane roads in
the United States; road width is about 12 m. Fig. 3 shows the
dimensions of a model abutment used. Though a series of pre-
liminary experiments were done with clear-water scour condi-
tions, with u* / u*c = 0.85, the bulk of the experiments were
conducted under live-bed flow conditions, with u* / u*c = 1.5 共here
u*⫽shear velocity; and u*c⫽critical value of the shear velocity
associated with bed-particle movement兲. The main hydraulic pa-
rameters for the flume flow were: mean velocity, U0 = 0.55 m / s;
and flow depth, Y 0 = 0.10 m. The sediment parameters were: me-
dian particle size, d50 = 0.45 mm; standard deviation of sediment
size, ␴g = 1.4; specific gravity of particles= 2.4; and, the angle of
sediment repose, ␣ = 30°.
Dune-bed conditions pose the most severe test for the stability
of an armor cover, such as riprap or geobags, because the passage
of dunes may dislodge portions of a cover. This certainly was
Fig. 2. Bridge failure associated with abutment scour
432 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007
Fig. 3. Model dimensions of a 45° wing-wall abutment of the
pile-supported form commonly used in the United States
found in the present study, and it is amply shown for efforts at
armoring beds around piers 共e.g., Chiew and Lim 2000兲. It is of
interest to note that existing guidelines for riprap design are based
on laboratory experiments performed exclusively in clear-water
scour and do not account for the dislodging effects of bed forms
passing the riprap. The writers’ experiments are documented in
Martinez 共2003兲 and Korkut 共2004兲.
Geobag Sizing
The simulated, large geobags were sized in accordance with a
design method proposed by Pilarczyk 共2000兲. The method esti-
mates a geobag thickness, DB. The aerial extent of a geobag
should exceed DB, and otherwise can be sized for handling ease
or to fit a site. The general form of Pilarczyk’s relationship for
geobag thickness is
0.035 ␾ KTKh U2
Dn = 共1兲
共SSB − 1兲 ␽C Ksl 2g
where SSB = specific gravity of the geobag; U = depth-average
mean velocity; g = gravity acceleration; ␾ = stability parameter;
␪C = critical value of the Shields parameter for particle 共geobag兲
entrainment; KT = turbulence factor; Kh = depth parameter; and
Ksl⫽slope parameter
Ksl = 冑 冋 册
1−
sin2 ␣
sin2 ␪
共2兲
where ␣ = angle of the boundary on which the geobag is placed
and ␪C = angle of repose of the sediment forming the boundary;
for the experiments, ␣ and ␪ were 26.7° and 30°, respectively.
Pilarczyk 共2000兲, who gives the background to Eq. 共1兲, recom-
mends for geobags that ␾, ␪C, and KT be 0.75, 0.05, and 2.0,
respectively. The depth parameter, Kh is defined as a function of
water depth, y, and equivalent roughness ks. Pilarczyk 共2000兲 sug-
gests using ks = Dn. However, since Dn is unknown initially, the
measured thickness of the geobag sample was used as a trial
value. The required thickness of the geobags, Dn, was calculated
as 22 mm, using a bulk specific gravity of the model geobags
measured to be 1.46. Accordingly, the model geobags used in the
Table 1. Values of Ks and n 共Richardson and Davis 1995兲
F KS n
艋0.8 1.02 2
⬎0.8 0.69 0.1
experiments were selected to be 22 mm thick, and 95 mm by
55 mm in plan area.
Riprap Sizing
The diameter of riprap stone used was estimated using the rela-
tionship proposed by Richardson and Davis 共1995兲
DR KS
= Fn
Y 共SS − 1兲
in which DR = equivalent riprap diameter; Y = flow depth in the
bridge section; KS = shape factor associated with abutment shape
共wing-wall or spill-through兲 F = U / 共gY兲0.5⫽Froude number for
the mean flow in the bridge 共contracted兲 section; and SS-specific
gravity of riprap stone. For wing-wall abutments, values of KS
and n are given in Table 1 共Richardson and Davis 1995兲. The flow
conditions used in the experiments, and using SS = 1.4, determined
for the available stone, Eq. 共3兲 gives DR = 22 mm, which is about
the same as the required thickness of geobag estimated using Eq.
共1兲. The riprap stone selected for the flume tests was sieved so
that DR ⬇ 22 mm, with a shape factor 共major axis/minor axis兲 of
about 2.1.
Geobags of approximately equivalent weight were sized, using
Eq. 共3兲, as 90 mm⫻ 70 mm⫻ 18 mm, densely filled with sand;
other geobag sizes were tested, but the length dimensions of these
bags were considered, based on comments from bridge engineers,
to be useful for handling and placement.
Results
Presented here first are observations and data for the baseline
scour condition 共no countermeasure兲, followed by a description of
the performance of a riprap apron such as depicted in Fig. 1. The
results then describe the performance of fitted large geobags, and
the performance of small geobags as riprap alternative, and even-
tually discuss the performance of geobags used as substitute for
filter-cloth underlay to riprap. The results include a summary of
the maximum scour depths associated with the baseline scour
conditions, the use of a riprap apron, and selections from the
geobag arrangements tested. The experiments showed that maxi-
mum depths of scour at an abutment fitted with an apron could
occur at three locations, as indicated in Fig. 4, i.e., in front of the
abutment itself 共A兲; in front of loose apron protection 共B兲; and
downstream of an apron formed of armor elements linked like a
mattress 共C兲. Table 2 lists the selection of scour experiments re-
ported herein.
By virtue of the trial-and-error aspect of the study—
determining whether or not a countermeasure concept worked—
considerably more experiments were conducted than are
presented herein. Martinez 共2003兲 and Korkut 共2004兲 provide full
documentation of the results.
Fig. 4. Location of deepest scour: 共A兲 no protective apron; 共B兲 loose
protective apron; and 共C兲 protective apron as mat tied to abutment
Baseline Scour Condition
共3兲
The initial experiments were conducted to observe scour devel-
opment and measure scour depth at the abutment when it was not
protected with an apron of any form, and to determine a baseline
depth against which geobag performance could be compared. As
no prior study has tested wing-wall abutments, such as shown in
Figs. 1–3, it is useful to include here a brief description of the
scour process.
Fig. 5 shows the consequent scour form that developed at the
abutment with the pile cap at the level of the main-channel bed.
The maximum scour depth, dSmax, occurred at the leading corner
of the abutment, where the flow contraction was greatest and
wake eddies were generated. The maximum scour depth extended
146 mm below the average bed level of the flume 共about 5.2 m at
full scale兲.
The baseline experiments revealed two important mechanisms
whereby the wing-wall abutment could eventually fail. One
mechanism was the mass failure of the embankment that occurred
once the scour hole had deepened to the extent that the embank-
ment’s earthfill lost its geotechnical stability. In comparison, Fig.
1 shows the mass failure of an actual embankment approach. The
second mechanism had not been reported heretofore, largely be-
cause it is difficult to observe. It occurred as follows. As the scour
deepened to below the pile cap and exposed the piles, the em-
bankment’s earthfill was eroded out from beneath the pile cap.
Gradually, a cavity developed within the embankment, undermin-
ing the embankment immediately behind the abutment. This de-
velopment is depicted in Figs. 6共a–c兲. Eventually scour deepening
caused the embankment side slopes to become unstable and slide
into the scour hole, whence sediment was removed by the flow.
As the embankment collapsed, the flow passed around the ex-
posed abutment and resulted in a scour form somewhat similar to
that observed at the failed bridge depicted in Fig. 2.
An observation from the experiments is that the embankment’s
earth-fill beneath and behind the pile cap must be protected. Two
options for doing this are to place armor material immediately
behind the pile cap, or to place the pile cap at a lower elevation.
These options were tested in the experiments. The experiment
with riprap placed immediately behind the pile cap produced a
deeper scour 共165 mm at the leading corner兲, but the embankment
did not fail. The experiment with the pile cap lowered showed
that a lower pile cap resulted in a still larger maximum scour
depth of 182 mm, at the abutment’s upstream corner. Although
this scour depth exceeded the baseline scour depth 共Experiment
3兲, scour could not progress substantially lower than about the
JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007 / 433
Table 2. List of Principal Experiments
Experiment
1. Baseline scour
2. Embankment protected with geobags
3. Geobag protection under the pile cap
4. Testing performance of Riprap I
5. Testing performance of Riprap II
6. Protection of apron and embankment
7. Apron with geobags also under the pile cap
8. Partially tied geobag apron
9. Geobag matress
10. Fully tied apron of geobags
11. Steep embankment slope
12. Pile cap lowered, no geobags
Description
This experiment was conducted to produce reference baseline scour depth that can be used to
determine the scour-reducing influence of a geobag apron.
The side slopes of erodible embankment behind pile-supported abutment were protected with
geobags. No geobag apron.
Geobags were placed under the pile cap in addition to the side slopes to prevent winnowing.
This experiment conducted to test performance of riprap to protect pile supported wing-wall abutment
with erodible embankment.
Experiment 4 repeated with rigid embankment.
Geobags placed in a manner replicating the riprap configuration found to be commonly used for Iowa
DOT bridges.
Experiment 6 repeated with geobags placed under the pile cap as a filter.
Only the geobags at the upper and the lower layers of the apron tied together.
In addition to the two rows of the apron, the geobags at the half downstream part of the upper layer
of the apron toe tied together.
The entire apron of geobags tied together.
Performance of the geobag system used in the previous experiment tested for a steeper embankment
side slope.
The pile cap placed deeper in the bed.

pile-cap base. Use of scour protection immediately behind the
pile cap, or use of a lowered pile cap, therefore enables a wing-
wall abutment and its approach embankment to better withstand
scour.
Scour at Abutment with Riprap Apron
Series of experiments were conducted to determine how scour
develops when an apron of riprap stone is placed around the
abutment in a similar manner to that shown in Fig. 1. The riprap
apron comprised a layer of riprap about two stones thick, with a
toe three to four stones thick. The riprap stones simulated were
scaled down to uniform-sized riprap of d50 of 22 mm. Fig. 7共a兲
illustrates the initial arrangement of the riprap apron used in the
flume experiments. The experiments showed that edge failure of
the riprap apron led to apron failure, and scour progression be-
neath and around the abutment, including the abutment pile cap.
The failure started at the apron’s upstream edge, where acceler-
ated flow and the passage of dunes destabilized the apron’s riprap
Fig. 5. Scour development at the abutment and embankment when
unprotected: A = location of deepest scour
toe, and resulted in large parts of the entire riprap apron sliding
into the scour hole forming around the apron. As the riprap apron
slid, it exposed the pile cap so that embankment sediment was
winnowed from beneath the pile cap. The embankment then failed
due to winnowing of sediment from beneath the pile cap. Addi-
tionally, the deepening scour hole caused the embankment side-
slope to become unstable. Fig. 7共b兲 shows the resultant scour
hole, whose maximum scour depth occurred at the upstream cor-
ner of the abutment, and was 85 mm. The manner of riprap apron
failure essentially was the same as observed by Chiew and Lim
共2000兲 and Parker et al. 共1998兲 for riprap aprons placed around
model bridge piers. It is evident from the experiment that riprap
aprons placed locally around an abutment may not work in dune-
bed channels, unless the apron toe extends sufficiently deep so as
to be below the trough elevation of dunes moving through the
bridge opening.
Scour at Abutment Fitted with Single Large Geobag
To address the question posed for this study, a series of trial-and-
error experiments were conducted to determine whether, and how,
scour would be prevented by one or more large geobags fitted as
an apron around the perimeter of the test wing-wall abutment. All
of these experiments essentially showed that for large geobags to
be effective at preventing scour depth at abutments, the geobags
must be tied to the abutment, otherwise they would slide away
from the abutment, expose the abutment footing, and cause scour
of sediment from beneath the footing. However, even though the
large tied geobag eliminated scour at the abutment, it caused the
location of deepest scour to shift downstream of the abutment;
maximum scour depth dSmax was 110 mm, about the flow depth.
Fig. 8 illustrates a typical result from the experiments. The loca-
tion of maximum scour depth moved from Location A to C
共Fig. 4兲.
It was found that, as additional geobags were placed around
the abutment, the scour shifted further downstream of the abut-
ment. Therefore, to be effective, the geobags have to extend out a
substantial distance from the abutment. As shown by the subse-
quent experiments, this finding indicates that, for single- or
double-span bridges, the geobag 共or riprap兲 protection should ex-

434 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007


Fig. 6. As scour exposes piles 共a兲 and 共b兲, embankment soil may be
sucked under pile cap, forming a cavity behind the pile cap 共c兲
tend across the full width of the bridge waterway. Also, if the
additional geobags were not secured to the abutment, they would
slide into the scour hole. To be kept in mind in sizing bag
thickness using Eqs. 共1兲 and 共2兲 is that the slope angle may in-
crease substantially as scour develops, and that bag thickness
should be based on the anticipated slope associated with scour-
hole formation.
Scour at Abutment with Apron of Small Geobags
Series of experiments were conducted to determine how an apron
of relatively small geobags would perform as an alternative to a
riprap apron. The experiments involved an apron of geobags
placed in the design layout essentially the same as for the riprap
apron described earlier. The apron comprised a layer of geobags,
JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007 / 435
Fig. 7. The failure of a riprap apron due to scour 共Experiment 4兲; 共a兲
before scour; 共b兲 after scour 共the location of deepest scour is
indicated兲
two bags thick, with a toe three to four bags thick, and generally
conformed with the layout of the riprap apron shown in Fig. 4.
Early experiments revealed that, though the geobags loosely
placed reduced scour depth, they might not fully protect the abut-
ment pile cap. Figs. 9共a and b兲 illustrate the results of one typical
test finding in which sand was winnowed from between gaps in a
single layer of geobags, causing the bags to settle in a scour
region around the abutment.
The experiments showed that an apron of suitably positioned
and connected geobags 共acting like cable-tied blocks兲, such as
generally conforming to the apron in Fig. 4, can reduce the scour
depth at an abutment. However, as with a riprap apron, scour may
occur at Position B, the perimeter of the geobag apron, if the
geobags were loosely placed, or at Position C downstream from
the apron, if the geobags were aligned together as a mattress.
Once edge scour occurred, at either Position B or C, the edge
geobags 共as with riprap兲 were dislodged into the scour hole.
The experiments showed that the edge failure is the principal
factor that results in the failure of the geobag apron, as in Fig. 10.
Such failure can be eliminated or substantially reduced by fully
linking the geobags so as to form a flexible apron, by then sloping
the apron into the bed, and by forming a suitably deep toe of
geobags 共as for riprap兲 around the apron’s perimeter. It was found
that geobag size did not affect the performance of a tied apron of
geobags, for the experiment conditions tested. Of greater impor-
tance was that the geobag elements be linked to form a flexible
apron of sufficient coverage around the abutment.

Fig. 8. Scour development around an abutment fitted with a large,
continuous geobag: 共a兲 before scour; 共b兲 scour at downstream side
Scour at the abutment itself was eliminated when geobags or
riprap were placed under the pile cap to prevent the winnowing
erosion of riverbank and embankment soil through the exposed
region beneath the pile cap 共Experiment 7兲. Though the abutment
itself was protected, the bed scoured downstream of the geobag
apron, with scour depths exceeding the maximum scour depth at
the abutment for the baseline case.
Fig. 9. Scour failure of an apron formed of small geobags; 共a兲 before
scour; 共b兲 after scour
aprons extending from each abutment. Further experiments inves-
tigating the extent and layout required for an effective mat led to
the mat layout design guide presented in Fig. 11, which reflects
the following recommendations:
1. The mat should extent upstream and downstream of the abut-
ment by a distance of minimally one bridge width to ensure
that the waterway bed is protected from the accelerating flow

Summary of Scour Data

The scour-depth results associated with the experiments are plot-
ted in Fig. 10, which shows how appropriate geobag use may
reduce scour depth at the abutment 共dsA兲, but with the conse-
quence of shifting scour to Positions B 共dsB兲 and C 共dsC兲 共posi-
tions indicated in Fig. 4兲. In shifting scour location, geobag use
may not eliminate scour at a bridge. As is shown in the ensuing
discussion, it quickly becomes evident that the protection 共geobag
or riprap兲 must extend as a mat across essentially the full opening
of a bridge waterway.

Scour with Waterway Mat

The foregoing findings show that the bridge waterway should be Fig. 10. The deep scour depth led to both edge failure and the slope
fully lined with a protective mat, which essentially links the failure of the apron toe and the downstream riverbank, respectively

436 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007

Fig. 11. Maximum scour depths for experiments described in Table
1; depths of scour at Locations A 共closed circles兲, B 共open circles兲,
and C 共closed triangles兲 in Fig. 4. When full mat 共double layer兲
across channel, scour depth was zero.
through the waterway. Mats providing this extent of water-
way coverage were able to prevent scour in the waterway.
2. The mat should be sloped as in Fig. 11. The bottom of the
slope should coincide approximately with the trough eleva-
tion of bed sediment dunes passing through the bridge wa-
terway. At this bottom elevation, dunes do not cause the
upstream or downstream edges of the mat to be undermined
and fall apart. An additional advantage of the sloped mat is
that it enables low flows to concentrate at the center of the
waterway; this is an advantage for fish passage. Additionally,
the center slope minimizes flow blockage through the water-
way.
3. The mat should have a toe and a heel, each of which are
three geobag or riprap stone thicknesses deep below the mat
共Fig. 11兲.
Early series of tests with a single layer of geobags not tied to
each other resulted in local failure of the mat near the abutments,
owing to the winnowing of sand from between the geobags. The
same result occurred with a single layer of riprap. An illustrative
series of geobag-mat failure is given in Figs. 12共a and b兲. If
repeated with a double layer of geobags, or a double layer of
riprap, minimal winnowing of sand occurred, and no scour devel-
oped at the bridge opening. For example, the performance of a
mat formed from a double layer of riprap is shown in Figs. 13共a
and b兲. The mat remained intact and inhibited scour at the either
abutment.
Geobags as Filter Underlayer for Riprap
The flume tests showed that for geobags to serve as an effective
form of filter cloth beneath a single layer of riprap, and for the
riprap to be stable, it was necessary for the geobags to be placed
slightly below the local level of the channel bed. For this arrange-
ment the riprap remains stable. Otherwise, when the geobags
were placed on top of the channel bed, and the riprap was placed
on the geobags, it was found that the riprap was exposed such that
riprap stone was less stable than when the geobags were placed
below level. A larger size of riprap stone would be needed in this
situation 共Fig. 14兲.
Design Guidelines
Guidelines are briefly presented here for the geobag layout, siz-
ing, and postgeobag scour location as an apron around a single
JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007 / 437
Fig. 12. Performance of a geobag mat 共single-bag thick兲 with riprap
toe and heel; 共a兲 setup; 共b兲 after test showing failure of mat owing to
winnowing of sand from between bags
abutment or a mat extending across the full bridge waterway.
Each design should be site specific to some degree, but the fol-
lowing general guidelines should be incorporated as much as
possible.
1. Sizing of the geobags should be such that the thickness of an
individual geobag should be equivalent to, or exceed, that of
a riprap stone sized for the abutment site. Several methods
for sizing riprap are available 共e.g., U.S. Army Corps of En-
gineers 1989; Richardson and Davis 1995; Austroads 1994兲.
After calculating the riprap size, the geobags can be sized
with overall plan dimensions that enable convenient assem-
bly of bags as an apron whose extent is comparable with the
apron shown in Fig. 4, with other apron extents commonly
used for riprap 共e.g., Fig. 1兲, or with an apron extent found
necessary for a particular bridge site.
2. It is necessary to link 共e.g., by tying兲 the geobags placed as
an apron around an abutment. So doing enables the system of
geobags to function as a moderately flexible armor cover that
stays intact when the channel bed scours around the abut-
ment, and impedes undermining.
3. The maximum slope of the geobag apron should be about
2H:1V. The geobag apron should have a toe or skirt at the
bottom end that extends deeper than the mattress by at least
two thicknesses of geobag.
4. The region beneath, and immediately behind, the pile cap of
an abutment should be protected so as to prevent loss of
embankment soil. Protection can be by means of geobags
Fig. 13. Recommended minimum extent of mat formed from
geobags or riprap for single-span bridge

and/or riprap. Because the earthfill region of the embankment

adjoining the abutment typically is poorly compacted and
prone to erosion, it is important to ensure that it is protected.
5. If the aprons are linked so as to form a protective mat, the
geobags should be of double layer thickness, but need not be
tied together. It is necessary though to provide toe and heel
protection of the mat, as indicated in Fig. 11.

Conclusions

Local scour cannot be eliminated completely by an apron of

geobags or, for that matter, one formed of riprap. An apron shifts
the scour region away from an abutment. Fig. 9 shows for the
present experiments that an apron can prevent scour developing at
the abutment itself, but that significant scour can occur readily
near the downstream edge of the apron of geobags. A concern in
using an apron is to ensure that shifting of scour does not imperil
a nearby pier or portion riverbank. Moreover, if the scour is likely
to extend to an adjacent pier, then the abutment and pier counter-
measure apron should be placed so as to protect both elements of
a bridge.
Fig. 14. Performance of riprap mat 共double stone layer兲 for a single-
span bridge. Mat layout as given in Fig. 12.
The writers’ experiments show that it is necessary to protect
the following regions of the river bed and banks near an
abutment:
1. The river bed at the abutment pile cap;
2. the riverbank immediately upstream of the abutment, and a
short distance downstream of the abutment;

438 / JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007

3. The sides lopes of embankment immediately behind the
abutment 共standard stub for a wing-wall abutment or spill-
through abutment兲; and
4. The area beneath and immediately behind the pile cap.
With regard to the specific use of geobags for wing-wall abut-
ments, the following conclusions can be drawn:
1. Geobags are a promising alternative to riprap for use as a
bridge abutment-scour countermeasure.
2. It is necessary to connect the geobags placed as an apron
around an abutment. The initiation of the failure of geobag
apron shown in Fig. 11 was due to the failure of an indi-
vidual geobag placed in front of the abutment.
3. It is necessary to place geobags 共or riprap兲 immediately
under the pile cap in order to prevent the winnowing of em-
bankment sediment from beneath the pile cap.
4. Geobags may serve as a useful alternative to a geotextile
filter cloth placed beneath a riprap apron, because geobags
are more readily placed than is an underlay cloth for block-
ing the winnowing of sediment from between bed-armor el-
ements like riprap stone. The geobags, though, should be
placed somewhat below bed level so as not increase riprap
exposure to flow.
Acknowledgments
This study was conducted under NCHRP 24-18, Scour Counter-
measures for Bridge Abutments.
References
Austroads. 共1994兲. Waterway design—A guide to the hydraulic design of
bridges, culverts, and floodways, Sydney, Australia.
Chiew, Y.-M., and Lim, F-H. 共2000兲. “Failure behavior of riprap layer at
JOURNAL OF HYDRAULIC ENGINEERING © ASCE / APRIL 2007 / 439
View publication stats
bridge piers under live-bed conditions.” J. Hydraul. Eng., 126共1兲,
43–55.
Heibaum, M. H. 共2002兲. “Scour protection and repair by filtering geosyn-
thetic containers.” Proc., 1st Int. Conf. on Scour of Foundations
(ICSF-1), College Station, Tex., Vol. 1, 1–13.
Heibaum, M. H. 共2004兲. “Geotechnical filters, the important link in
scour protection.” Proc., 2nd Int. Conf. on Scour of Foundations 共CD-
ROM兲, Singapore.
Korkut, R. 共2004兲. “Geobags as abutment-scour countermeasure.” MS
thesis, Dept. of Civil and Environmental Engineering, The Univ. of
Iowa, Iowa City, Iowa.
Martinez, E. J. 共2003兲. “An assessment of two countermeasures to reduce
abutment scour.” MS thesis, Dept. of Civil and Environmental Engi-
neering, The Univ. of Iowa, Iowa City, Iowa.
Parker, G., Toro-Escobar, C., and Voigt, R. L., Jr. 共1998兲. “Countermea-
sures to protect bridge piers from scour.” Final Rep., Project No.
NCHRP 24-7, Prepared for National Cooperative Highway Research
Program, Univ. of Minnesota, Minneapolis.
Permanent International Association of Navigation Congress 共PIANC兲.
共1992兲. “Guidelines for the design and construction of flexible revet-
ments incorporating geotextiles in marine environment.” Supplement
to Bulletin Nos. 78/79, Brussels, 122–124.
Pilarczyk, K. W. 共2000兲. “Geomettresses in erosion control—An over-
view of design criteria.” Filters and drainage in geotechnical and
environmental engineering, Balkema, Rotterdam, The Netherlands,
331–338.
Ray, R. 共1977兲. “A laboratory study of the stability of sand-filled nylon
bag breakwater structures.” U.S. Army, Coastal Engineering Research
Center, Vicksburg, Miss.
Richardson, E. V., and Davis, S. R. 共1995兲. “Evaluating scour at bridges.”
Hydraulic Engineering Circular No. 18 (HEC-18), 3rd Ed., Rep. No.
FHwA-IP-90-017, Office of Technology Applications, HTA-22, Fed-
eral Highway Administration, U.S. Dept. of Transportation, Washing-
ton, D.C.
U.S. Army Corps of Engineers 共USACE兲. 共1989兲. “Riprap protection.”
U.S. Army engineer manual, Engineering Research Center, Davis,
Calif.