LESSONS LEARNED FROM 10 YEARS OF

LEAK DETECTION SURVEYS ON

GEOMEMBRANES

B. FORGET, A.L. ROLLIN and T. JACQUELIN

SOLMERS INC., 1471 Lionel-Boulet Boulevard, Suite 22, Varennes J3X 1P7, Quebec,
Canada

SUMMARY

Statistics obtained from geoelectric leak detection surveys performed on more than 89 projects
totalling 2 652 000 m2 are presented. Although many papers have been previously published on this
topic, few authors have gone as far as to identify leak densities on geomembranes (exposed and
covered) with respect to their thicknesses, the application or absence of a Construction Quality
Assurance (CQA) program and water puddle leak detection survey on exposed geomembrane prior
to the placement of a covering material. Results obtained within this investigation show that the
average leak density on exposed geomembranes (many types and thicknesses) that were installed
under a rigorous CQA program is approximately 4 leaks per hectare. Conversely, the statistics show
a sharp climb, to 22 leaks per hectare, in the absence of such a CQA program. The situation was
found to be similar with covered geomembranes: a negligible leak density (0.5 leaks/ha) found on
geomembranes installed under a strict CQA program and a prior water puddle leak detection survey
on the exposed geomembrane, climbing sharply to a density of 16 leaks/ha in the absence of both a
CQA program and the water puddle leak detection survey.

1. INTRODUCTION

Leak detection surveys performed on exposed geomembranes (totalling 2 291 000 m2) and covered
geomembranes (totalling 361 000 m2) during a 10 year period, applied to 89 projects located in
8 different countries have been analyzed in order to complement recently published data and to
stress the importance of the implementation of rigorous Construction Quality Assurance (CQA)
programs during geomembrane installation. The data collected was used to calculate the number of
leaks per hectare (leak density) for each project, and to register the types and size of located leaks.
The surveyed sites were lined with HDPE, PVC, or bituminous geomembranes.
Two different leak detection techniques used to verify the integrity of a geomembrane are
presented: the water puddle technique as used on exposed geomembranes either during or after their
installation and the dipole technique as applied on geomembranes covered with a soil layer.
 Statistics indicating a global decrease in the number of leaks found on sites where a rigorous
construction quality assurance program is implemented are then discussed. In the context of this
article, a rigorous CQA program is defined as a constant surveillance by an expert who possesses
recognized expertise in geomembrane installation during all of the liner installation phases,
including subgrade preparation, liner installation, seaming procedures and covering of the
geomembrane.

2. GEOELECTRIC LEAK DETECTION TECHNIQUES

Geoelectric leak detection techniques used on geomembranes have been described in many
publications, such as Peggs (1989, 1990, 1993), Darilek et al. (1988, 1989), Laine et al. (1989, 1991,
1993) and Rollin et al. (1999, 2002, 2004), and in standards such as ASTM D6747 (Standard Guide
for Selection of Techniques for Electrical Detection of Potential Leak Paths in Geomembranes),
ASTM D7002 (Standard Practice for Leak Location on Exposed Geomembrane Using the Water
Puddle System) and ASTM D7007 (Standard Practices for Electrical Methods for Locating Leaks in
Geomembranes Covered with Water or Earth Materials).
The water puddle method consists in the creation of a potential difference between a soil under an
exposed geomembrane and a puddle of water projected from a diffuser onto the surface. Most
geomembranes are highly resistant electrical insulators and inhibit electrical currents. As soon as
water percolates through a perforation and reaches the supporting soil, a ‘bridge’ is created between
these two potentials which generates an electrical current. A detector signals the presence of an
infiltration to the operator (via acoustical and visual signals). This technique permits the detection of
leaks with dimensions of 1 mm2 or greater (ASTM D7002).
On-site preparation is minimal and generally permits the survey to proceed during the
geomembrane installation. The prospecting rate is approximately 5000 m2 /day/operator, depending
on site conditions. To achieve this survey rate, a continuous water supply of approximately
4 m3/day/operator is necessary. This water supply may be provided from a tanker or a direct
connection to a municipal network. If a water supply proves difficult, the use of a closed circuit with
a low point is also possible. Figure 1 provides a general schema of the water puddle method.

Figure 1. Water puddle technique on exposed geomembranes.

 In the dipole leak detection technique, an electrical potential is applied between the covering
material above the geomembrane and the soil below it. Since most synthetic geomembranes are
effective electrical insulators, the presence of a leak creates a localized passage of current, which
perturbs the potential field in a characteristic way. Leaks are located by recording potential readings
with the dipole at predetermined grid densities.
Under moderate climatic conditions, on-site preparation is minimal. Spraying water on the
covering soil surface might be necessary to insure good contact with the dipole under very dry
conditions. The detection limit is variable but generally allows detection of holes with dimensions of
6 mm2 or greater (ASTM D7007). Figure 2 provides a general schema of the dipole method.

Figure 2. Dipole technique on covered geomembranes.

3. LEAK DENSITIES - DIFFERENT TYPES AND THICKNESSES OF GEOMEMBRANES

The results obtained from 57 geoelectric leak detection surveys performed on exposed
geomembranes (HDPE, PVC and bituminous) of different thicknesses are presented in Figure 3. The
leak densities used in this investigation were calculated using data collected from the leak detection
surveys.
The majority (80%) of the projects where the liner installation was performed using a rigorous
CQA program had very low leak densities, ranging from 0 to 7 leaks per hectare, with an average of
4 leaks per hectare. The remaining 20% of projects, which included a CQA program, had leak
densities greater than 7. On the other hand, an average leak density of 22 leaks per hectare was
calculated for projects without a rigorous CQA program.
An analysis of eight ( 8 ) projects performed using a rigorous CQA program that had leak
densities greater than 7 leaks per hectare indicated the following:
In one project, the geomembrane installed in the proximity of a rock wall was not properly
protected from falling rocks puncturing the geomembrane;
In another project, seaming problems were encountered during the installation caused by the
inexperience of the installer;
In a third project, extensive extrusion welding caused half of the leaks detected;
 In the remaining five projects, the lined areas were very small (between 600 and 5310 m2) which
rendered the leak density analyses biased.

Comparison of Projects Completed With and Without

Construction Quality Assurance
Exposed Geomembrane (57 Projects, 2 291 000 m 2 total)

30%

25%
With CQA - 43 projects
Without CQA - 14 projects
20%
% of projects

15%

10%

5%

0%
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50
leaks / hectare

Figure 3. Leak Densities – With and Without a Rigorous CQA Program (Exposed Geomembranes).

For leak detection surveys performed on covered geomembranes, where both a rigorous CQA
program and a water puddle leak detection survey on exposed geomembrane during installation
were implemented (corresponding to approximately 80% of the projects), the calculated leak density
was found to be almost negligible, approximately 0.5 leaks per hectare. Conversely, an average leak
density of 16 leaks per hectare was found for projects that did not implement a strict CQA program.

4. LEAK DENSITIES - HDPE GEOMEMBRANES

The data obtained for different HDPE geomembrane thicknesses was gathered to determine the
influence on the quantity of perforations during installation. The survey results have been grouped
as exposed and covered geomembranes. Columns 1 and 2 of Table 1 present the results of the leak
densities for exposed geomembranes with and without CQA implemented during installation. The
results obtained for covered geomembranes are presented in columns 3 and 4. The surveys on
projects with deficiencies at the design level, and projects with a total surveyed area of less than
10 000 m2, were discarded in order to permit an objective comparison of the influence of CQA
programs and leak detection surveys.
Table 1 - Leak Density Comparison (HDPE Geomembranes).
Exposed HDPE Geomembranes Covered HDPE Geomembranes
(Water Puddle) (Dipole)
Column 1 Column 2 Column 3 Column 4
With CQA and Without CQA and
Geomembrane geoelectrical leak no geoelectrical
Thickness With CQA Without CQA survey (water leak survey
puddle) before (water puddle)
covering before covering
Number of leaks per hectare (prospected area m2)
2.0 mm 3.2 (362 460) N/A (0) 0.2 (170 190) 15.6 (50 600)
1.5 mm 5.1 (66 880) N/A (2 760) N/A (0). 24.7 (10 500)
1.0 mm 20.5 (17 070) 31.5 (313 770) N/A (1110) N/A (0)
N/A: Data nonexistent or insufficient to enable a good representation.

As intuitively expected, leak density was found to decrease as the thicknesses of the HDPE
geomembranes increased. However, it was not found to be inversely proportional to the thickness of
membranes installed where a rigorous CQA program was implemented: leak densities of 20.5, 5.1,
and 3.2 leaks per hectare, respectively for 1 mm, 1.5 mm, and 2 mm thick exposed geomembrane
were found.
As shown in Table 1, the leak densities obtained from projects without a rigorous CQA program
correspond well with the data presented by Rollin et al. (EUROGEO 2004), Nosko et al. (2000), and
Phaneuf et al. (2001). More important is the fact that the calculated leak densities for applications
where a rigorous CQA program was implemented are greatly reduced and correspond better with the
densities forecasted by Giroud (1989).
Another essential observation arises from efforts to determine the number of perforations
resulting from the installation of a covering material on top of a geomembrane. A comparison of the
data for exposed and covered 2 mm thick HDPE geomembranes in projects where a CQA program
was implemented (columns 1 and 3 in Table 1), indicating respective densities of 3.2 and 0.2
leaks/ha, leads to the conclusion that only 6% of the perforations were caused during the covering
material installation. In a 1996 survey reported by Nosko et al., and one in 2001 by Phaneuf et al.,
the results obtained in landfills indicated that 73% of damage occurs when the soil layers are placed
on top of the geomembranes, 24% occurs during geomembrane installation, and 2% occurs during
the post construction phase. They concluded that, contrary to the general perception, most damage
detected in landfills occurs during covering layer installation and is not caused by improper seaming.
This conclusion is probably valid only in cases where no rigorous CQA program has been
implemented.
5. LEAK TYPES - HDPE GEOMEMBRANES

An analysis was performed to determine the breakdown of the four types of perforations generally
found on geomembranes and identified as faulty welds, tears, cuts, and punctures. Perforations can
be caused during the welding process and include separation due to poor double fusion welds,
perforations during the fusion process, insufficient water tightness or perforations as a result of
faulty extrusion welds. Tears are generally caused by difficulties in handling the geomembranes
during installation or by heavy equipment traffic used during the covering phase. The installer is
most likely to be responsible for damage due to cuts. Finally, punctures arise from contact with
static objects (such as sharp edged stones and gravel) left on or under the geomembranes.
Figure 4 presents a breakdown of the different causes of perforations. The first graph considers
the data obtained from exposed 2 mm HDPE geomembranes installed using a rigorous CQA
program, and the second presents the results from exposed 1 mm HDPE geomembranes in the
absence of a CQA program. It must be noted that this comparison deals with 2 variables due to the
complete lack of data concerning 2 mm geomembranes installed without a rigorous CQA program.

Breakdown of leak types on exposed HDPE Breakdown of leak types on exposed HDPE
geomembrane geomembrane
1 mm thick geomembrane - without CQA 2 mm thick geomembrane - With CQA
31.5 leaks/hectare (313 770 m2) 3.2 leaks/hectare (362 460 m2)

Welds : 28%
Punctures :
0.9 leaks/ha
33%
Welds : 18%
1.1 leaks/ha
5.7 leaks/ha

Tears : 6%
1.9 leaks/ha

Punctures : Cuts : 2% Tear : 10%

74% 0.3 leaks/ha
0.6 leaks/ha
23.6 leaks/ha Cuts : 29%
0.9 leaks/ha

Figure 4. Breakdown of Leak Types (Exposed HDPE Geomembranes).

The analysis of the leak types, as presented in Tables 1 and 2, permits the measurement of the
impact of the thickness and of CQA implementation on the leak densities.
A comparison of Tables 1 and 2 permits the following conclusions:
Confirming what has been previously published on this topic, approximately 30% of leaks are
found at seam edges, and 70% are found on the panels;
There is no correlation between the geomembrane thickness and the implementation of a CQA
in regard to the number of knife cuts;
The number of tears found is six (6) times greater for the 1 mm geomembrane without a CQA
program than with implementation of a CQA program (1.9 vs. 0.3 leaks/ha);
The number of faulty seams is six (6) times greater for the 1 mm geomembrane without a CQA
program than with implementation of a CQA program (5.7 vs. 0.9 leaks/ha);
 The number of punctures is twenty one (21) times greater for the 1 mm geomembrane without a
CQA program than with implementation of a CQA program (23.6 vs. 1.1 leaks/ha).

6. LEAK DIMENSIONS - HDPE GEOMEMBRANES

Even though the leak densities on 2mm thick HDPE covered geomembranes are very low after the
application of a rigorous CQA program (0.2 leaks/ha), it is important to mention that the vast
majority of these perforations are caused by the heavy equipment traffic used during the installation
of the covering materials. Consequently, the sizes of these perforations were found to be relatively
large. This is corroborated by Nosko and Touze Foltz (2000) that characterised the dimensions of
the perforations as a function of their type. Table 2 presents their results for covered and exposed
geomembranes, while not taking into account the thickness or the presence or absence of a rigorous
CQA program. It is interesting to note that the perforations caused by heavy machinery represent
75% of the tears greater than 10 cm2.

Table 2. Perforation Sizes and Types (Nosko and Touze Foltz, 2000).

Diam. Stones Heavy Seams (%) Cuts (%) Installer Total

Size (cm2) (%) equipment directly
(%) (%)
< 0.5 332 11.1 - - 115 43.4 5 8.5 - - 452
0.5 – 2.0 1720 57.6 41 6.3 105 39.6 36 61.0 195 84.4 2097
2.0 – 10 843 28.2 117 17.9 30 11.3 18 30.5 36 15.6 1044
> 10 90 3.0 496 75.8 15 5.7 - - - - 6701
Amount 2985 654 265 59 231 4194
Total 71.17% 15.59% 6.32% 1.41% 5.51%

Figure 5 shows a typical geomembrane perforation originating from heavy equipment traffic on the
covering material located at the cell bottom.
Figure 5. A one meter tear discovered using the dipole method, caused by heavy equipment during
covering material placement.

7. CONCLUSION

The water puddle and dipole techniques used in geoelectric leak detection surveys are standardized
methods (ASTM) that enable a control of the global integrity of the geomembranes during the
installation and covering phases. The vast majority of leaks discovered after the geomembrane
installation (exposed geomembranes) are found in the panels (up to 70 %), not at the seams. It is
therefore mandatory to check the total lined area to ensure the geomembrane integrity and not only
to implement destructive and non-destructive seams testing.
The relationship between the leak density and the presence or absence of a rigorous CQA
program and geomembrane thickness has been found to be crucial. It was found that most
perforations are caused during geomembrane installation and not during its covering phase
whenever a rigorous CQA program is implemented. However, larger tears and holes are usually
encountered during the geomembrane covering phase.
For landfills, a rigorous CQA program combined with leak detection surveys on exposed and
covered geomembranes is recommended. Even though designers use high factors of security to
reduce the possibility of perforations in many projects, leak detection surveys are an alternative that
significantly reduce these factors, resulting in a reduction of the thickness of the covering material
layers, the protective geotextiles and the geomembranes themselves. In certain cases substantial
economic savings would result. A leak detection survey is a useful tool in assuring the integrity of a
containment project using geomembranes.

REFERENCES

ASTM D6747 Standard Guide for Selection of Techniques for Electrical Detection of Potential Leak
Paths in Geomembranes.
ASTM D7002 Standard Practice for Leak Location on Exposed Geomembrane Using the Water
Puddle System.
ASTM D7007 Standard Practices for Electrical Methods for Locating Leaks in Geomembranes
Covered with Water or Earth Materials.
Darilek G.T. & Parra J.O. (1988). The Electrical Leak Location Method for Geomembrane Liners,
USEPA/600/S2-88/035, Sept.
Darilek G.T., Laine D. & Parra J.O. (1989). The Electrical Leak Location Method for Geomembrane
Liners, Proceeding of Geosynthetics’ 89, IFAI, pp 456-462, San Diego.
Giroud J.P. & Bonaparte R. (1989). Leakage through Liners Constructed with Geomembranes, II:
Composite Liners, Geotextiles and Geomembranes, vol 8, no 2, pp 71-112.
Laine D.L. & Miklas M.P. (1989). Detection and Location of Leaks in Geomembrane Liners Using
an Electrical Method, Proceed. 10th Nat. Conf., Superfund’89, Washington, Nov.
Laine D.L. (1991), Analysis of Pinhole Seam Leaks Located in Geomembrane Liners Using the
Electrical Leak Location Method, Proceed. Geosynthetics’91, Atlanta, pp 239-253.
Laine D.L. & Darilek G.T. (1993). Locating Leaks in Geomembrane Liners of Landfills Covered
with a Protective Soil, Proceed. Geosynthetics’93, vol 3, pp 1403-1412, April.
Nosko V., Andrezal T., Gregor T. & Garnier P. (1996). Sensor Damage Detection System – The
Unique Geomembrane Testing Method, Proceed. EuroGeo, Netherlands pp 743-748.
Phaneuf R. & Peggs I.D. (2001). Landfill Construction Quality: Lessons Learned from Electrical
Resistivity Testing of Geomembrane Liners, Geotechnical Fabrics Report, vol 19, no 3, April pp
28-35.
Peggs I.D. & Pearson D.L. (1989). Leak Detection and Location in Geomembrane Lining Systems,
ASCE Annual meeting, Fort Lauderdale, September.
Peggs I. (1990). Detection and Investigation of Leaks in Geomembrane Liners, Geosynthetics World,
vol 1, issue 2., winter, pp 7-14.
Peggs I.D. (1993)a. Practical Geoelectric Leak Surveys with Hand-Held, Remote and Water Lance
Probes, Proceedings of Geosynthetics’ 93, IFAI, pp 1523-1532, Vancouver.
Peggs I.D. (1993)b. Advances and New Thinking in Landfill Liner Construction Quality Assurance
Practices, Proceedings of Sardinia’ 93, CISA’ Cagliari, pp 121-128.
Rollin A.L., Marcotte M., Jacquelin T. and Chaput L., (1999). Leak Location in Exposed
Geomembrane Liners Using an Electrical Leak Detection Technique, Proceedings of
Geosynthetics’99, Boston, pp 615-626.
Rollin A.L., Marcotte M., Chaput L., Caquel F. (2002). Lessons Learned from Geo-electrical Leaks
Surveys, Proceedings Geosynthetics 2002, Nice pp 527-530.
Rollin A.L., Jacquelin T., B. Forget, Saunier P. (2004). A Guide to Detect Leaks on Installed
Geomembranes, Proceedings EuroGEo, Munich, pp 235-240.