Pipelines 2015 Recent Advances in Underground Pipe... - (Assessment and Rehabilitation II)

Pipelines 2015 1181
Water Mains Degradation Analysis Using Log-Linear Models
Amin Ganjidoost1; Rizwan Younis2; and Mark Knight3

1
Ph.D. Candidate, Department of Civil and Environmental Engineering, University of
Waterloo, 200 University Ave. West, Waterloo, ON, Canada N2L 3G1. E-mail:
aganjido@uwaterloo.ca
2
Centre for Advancement of Trenchless Technologies, Department of Civil and
Environmental Engineering, University of Waterloo, 200 University Ave. West,
Waterloo, ON, Canada N2L 3G1. E-mail: ryounis@uwaterloo.ca
3
Associate Professor, Department of Civil and Environmental Engineering,
University of Waterloo, 200 University Ave. West, Waterloo, ON, Canada N2L 3G1.
E-mail: maknight@uwaterloo.ca
Abstract
The development of reliable lifecycle intervention plans for water distribution

systems depends on better understanding of water main degradation behavior.
Traditionally, water main failures have been studied as Weibull/Exponential
processes. This paper investigates the application of log-linear model for assessing
metallic water main structural degradation. A comparison of the proposed model with
the existing Weibull/Exponential based model is presented. Water mains inventory,
operational and performance data from a Canadian municipality are used in the
analyses. Conclusions concerning the adequacy of existing models and the
applicability of proposed models are made. Municipalities and water utilities can use
the method provided herein as a tool for desktop condition assessment and risk based
failure analysis.
INTRODUCTION
Municipal water mains installed more than a century ago are still in service in
many municipalities and water utilities across North America (Mirza and Haider,
Copyright © 2015. American Society of Civil Engineers. All rights reserved.
2003). The condition and performance of water mains deteriorate over time because
of a number of factors and complex processes that include, for example, mechanical,
biological, and chemical degradation processes, environmental wear, water chemistry
and operating conditions, accidental or intentional interference, defects during pipe
manufacturing process, poor choice of pipe material and faulty design, poor
installation and natural events. Ageing and deteriorating infrastructure systems along
with the lack of maintenance have accelerated the degradation of these vital assets. A
general assessment of the actual structural condition of these systems as well as
analytical tools to assess their rate of deterioration are needed. The breakage history
of the water main is considered as one of the performance measures by many
municipalities and regulatory authorities. To determine the current condition of these
© ASCE
Pipelines 2015 : Recent Advances in Underground Pipeline Engineering and Construction, edited by V. Firat Sever, and Lynn Osborn, American Society of Civil Engineers, 2015. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/uu/detail.action?docID=4415956.
Created from uu on 2021-04-25 00:35:46.
Pipelines 2015 1182
assets, one must collect and assess data on the breakage histories of water mains.
Unfortunately, many municipalities have only been rigorously recording breakage
histories for a decade, while their pipes have been in the ground for much longer
(Pelletier et al., 2003). To investigate the water pipelines failure, this paper uses the
annual number of water main breaks as an indicator of the structural health of water
distribution network.
To assess the performance of water mains, data on the breakage history of the
water main network is considered as a useful indicator. Despite the fact that these
assets are installed for a long time, many municipalities have a short breakage history
of these assets. This study models the occurrence of breaks in metallic water mains by
estimating: (1) the probability distribution of failure times; and (2) the number of
pipelines failures depending on covariates, such as construction period, break time,
and years in service.
The objective of modeling is to adequately reproduce the average tendency of
the annual number of pipe breaks and to predict breakage rates in the future. This can
help the municipalities and water utilities to ascertain present and future states and
performance of their water distribution networks.
The data for this paper come from a Canadian municipality in Southern
Ontario. The water distribution system consisted of metallic (ductile iron, cast iron,
and steel), plastic (PVC, PE) and Asbestos Cement water mains. About 33% of the
pipelines consist of cast iron. However, 95% of the breaks were recorded in cast iron
water mains. Therefore, we considered only cast iron water mains in this paper.
WATER MAIN DETERIORATION MODELS
Deterioration models predict the future condition of infrastructure components

and help decision makers to prioritize future maintenance, rehabilitation, and
replacement activities. For the purpose of the deterioration modeling of water mains,
information about breakage history is needed. For a thorough analysis of pipe
breakage, information must be known on the physical and that have an impact on pipe
failure (Pelletier et al., 2003).
To distinguish between the different orders of breaks, one must identify the
time to failure between the installation and the first break, between the first and
second break, and so on. This is called data stratification in survival analysis
(Kalbfleisch and Prentice, 1980). Times to failure can be modeled by different

distributions, depending on the breakage behavior associated with that break order.
Eisenbeis (1994) used different distributions for different break orders and developed
a calibration strategy, based on maximizing the likelihood function associated with
the model. Such a calibration strategy extends the use of survival analysis to a much
higher number of municipalities. The modeling and the calibration strategies are
presented in Mailhot et al. (2000) and Pelletier et al. (2003).
Rate-of-failure (ROF) models and transition-state (TS) models are two types
of water mains deterioration modeling developed in the literatures (Osman and
Baindridge, 2011). ROF models interpolate the water mains breakage rate based on
pipe age and environmental factors for a specific pipes segment without
distinguishing the times between successive failures; whereas, TS models
© ASCE
Pipelines 2015 1183
differentiate the time between successive failures of a specific pipe segment (Osman
and Baindridge, 2011). The nonhomogeneous Poisson process model (Kleiner and
Rajani 2010) and the Multivariate Exponential (MVE) models (Shamir and Howard
1979; Kliener and Rajani 2002) are some examples of the ROF models. Shamir and
Howard (1979) introduced the first time-dependent Exponential model for forecasting
water main break. They found that the rate of water main breaks increases
exponentially with pipe age.
Gustafson and Calnsey (1999) developed and implemented (Gustafson et al.,
2008) the transition state-life regression (TS-LR) to describe the time to failure
between the first and second, the second and third, and so forth-breaks. These authors
effectively showed that time to failure between breaks up to the 20th order can be
described by Exponential distributions. Osman and Baindridge (2011) compared and
analysed the ROF and TS models using a single data set for cast- and ductile-iron
pipes in the City of Hamilton, Ontario, Canada. They concluded that the TS models
rely upon a large and accurate historical water mains’ breakage record, and enable
municipalities to forecast future performance on the basis of multiple level-of-service
standards such as breakage rates, number of breaks for any specific pipe, and
probability of pipe failure in a defined time frame (Osman and Baindridge, 2011).
Wang et al. (2009) developed a logarithmic regression function to predict the
annual breakage rates of water main. They found that there is a significant correlation
between breakage rates, pipe diameter and length. Artificial neural networks (ANN)
have been used to investigate deterioration of water pipelines by Achim et al. (2007);
Al-Barqawi and Zayed (2008); Fahmy and Moselhi (2009); and Tabesh et al. (2009).
WATER MAIN DEGRADATION ANALYSIS
General description of case municipality. The water distribution network includes

501 km length of water mains, serving a population of 130,000. The following six
characteristics are collected for all pipe segments: (1) pipe diameter; (2) type of
material; (3) year of installation; (4) type of soil; (5) year of first break; and (6) year
of the second to the fifth break. Table 1 provides information about the municipality
and water distribution network.
Table 1. Characteristics of the case municipality.

Characteristics
Population 130,000
Pipe network length (km) 501
Number of pipe segments in database 1,373
Number of pipe breaks in database 807
Year of installation of first pipes 1,850
Number of years of recorded pipe breaks 25
© ASCE
Pipelines 2015 1184
Weibull/Exponential model. The modeling strategy is used two distributions to

model the different break orders. The Weibull distribution is associated with the first
break order (time to failure from installation to first break), while the Exponential
distribution is used to describe the behavior of subsequent breaks (time to failure
from first to second break, second to third, and so forth). The time step for the time to
failure is determined as one (year). The use of an Exponential distribution to describe
the time to failure between the first and second, the second and third, etc. breaks is in
agreement with the results of Gustafson and Clancy (1999). These authors effectively
showed that time to failure between breaks up to the 20th order can be described by
Exponential distributions. Moreover, they observed an almost constant parameter
value after the fifth order.
The first step is fitting a probability distribution to sample data. Fitting a
distribution to a dataset can be done by various methods such as Probability Paper
Plot (PPP), Maximum Likelihood Estimation (MLE), and Moment of Method
(MOM). The most versatile method used to analyze both complete and censored data
is the MLE method.
At the time of analysis, a water main may have experienced the break starting
the failure time under consideration, but the break ending the failure time has not
occurred. This is known as a censored data, or a censor for short. Because the end of
the failure time is not known, the data are said to be right censored. The value of the
censor is the length of time that has transpired since the break starting the failure
time.
Right-censored data are important as they contain valuable information about
the survival. Omitting these data from statistical analysis could lead to an
underestimation of life expectancy. The sample data are divided into two groups. The
first group is completed lifetime and the second group is modelled as right-censored
lifetime data.
To obtain the sample likelihood function given below, it is required to
calculate the probability density function (PDF) of the complete lifetime data and
reliability of the right-censored data.
=∏ ( )∏ ( ) [1]
To compute PDF and reliability with Weibull and Exponential distributions,

we can assume parameters of the distribution, and solve the problem numerically. The
optimal solution is obtained by varying the value of unknown distribution parameters
until the value of the likelihood function is maximized.
© ASCE
Pipelines 2015 1185
Weibull/Exponential model results and discussions. The Weibull distribution is

associated with the first break order (time to failure from installation to first break),
while the Exponential distribution is used to describe the behavior of subsequent
breaks (time to failure from first to second break, second to third, and so forth). The
Weibull distribution is defined by two parameters, and . The Exponential
distribution is a special case of the Weibull distribution when α=1, with only one
parameter, . The value from various types of probability paper plots is used to
determine the “best” probability distribution for the data. The “best” fit distribution
for the first break order of these data set is the Weibull distribution, and for the
subsequent breaks is the Exponential distribution (R2 is closer to 1, and the higher the
R2 value, the better the fit). Calibration parameters are presented in Table 2.
The Weibull and Exponential distributions PDF, cumulative distribution
functions (CDF), and survival (reliability) functions are represented, respectively, in
Equations 2 and 3 as follows:
( )
( )= ( )
( ) [2]
( )=1−
( )
( )=
where, α is the shape parameter and β is the scale parameter of the Weibull
distribution
( )=
( )=1− [3]
( )=
where, is the scale parameter of the Exponential distribution
Table 2. Calibration Parameters of Weibull/Exponential Model.

Parameters α β
2.075 42.9 0.0381
First break. Probability density function with the Weibull distribution (time to failure
from installation to first break) is shown in Figure 1a. The probability of first break
from installation (time to failure) increases from time zero to its maximum value of
31 years, and then, decreases until reaches zero at the age of 100-year.
Survival function associated with the Weibull distribution is shown in Figure
1b. The value of the survival function gives the proportion of pipes that have not
failed at time t. Therefore, the higher the curve, the longer it takes for the first break
to occur, on average, in that municipality. The mean time to failure (MTTF) is equal
© ASCE
Pipelines 2015 1186
to the area under survival curve and mathematically is shown in Equation 4. By

definition (Stephens, 2012):
= ( )= . ( ) = ( ) [4]
The mean time to failure associated with the first break is estimated to be
38.48 years. Alpha determines the shape of the distribution. Weibull distribution with
α=2.075 exhibits an increasing hazard rate (Figure1c) at a constant rate (i.e. the
probability of failure is increasing with time). Hazard rate is calculated as follow:
( ) [5]
ℎ( ) =
1− ( )
Subsequent break. Probability density function with the Exponential distribution

(time to failure from first to second break is shown in Figure 1a. The probability of
second break decreases from time zero until reaches zero at the age of 100 years.
Survival function associated with the Exponential distribution is shown in
Figure 1b. The mean time to failure for the Exponential distribution is equal to
inverse of the parameter of λ. The mean time to failure associated with the second
break is 26.25 years.
1 [6]
=
The Exponential distribution hazard rate is constant and equal to the scale
parameter, λ (Figure 1c). Therefore, a constant hazard rate of 0.0381 is obtained for
the second break.
Figure 2 shows the MTTF between the first to the fifth breaks for the cast iron
water main in the studied case municipality based on a 25-year breakage history data.
The breakage history data obtained from the studied municipality were limited to the
number of fifth break. The result shows that the MTTF is strongly related to the break
number of a cast iron water main. It also indicates that after the fifth break, the value
of MTTF is declining to zero.
© ASCE
Pipelines 2015 1187
(a) First Break Second Break

0.04
Probability Density Function
0.03
0.02
0.01
(b)
1
Survival Function (Ratio)
0.8
0.6
0.4
0.2
(c)
0.12
Hazard Rate, h(t)
0.08
0.04
0
0 20 40 60 80 100 0 20 40 60 80 100
Time to Failure (years) Time to Failure (years)
Figure 1. Weibull/Exponential model results for the first and second-break.
© ASCE
Pipelines 2015 1188
40
Mean Time To Failure (years)

30
20
10
0
0 1 2 3 4 5
Number of Break
Figure 2: Mean time to failure vs. number of break
Log-Linear model. Log-linear model, also known as Poisson regression or Poisson

log-linear model, is useful for modelling count or rate data (i.e., the number of events
per unit time period). For given explanatory variables, , the log-linear model for
expected rate is given as Agresti (2002):
= + + + ⋯+ [7]
Where is the expected value of the sample rate . For water mains’
break data, is the number of water main breaks in a given time period, . The
explanatory variables, , can include, for example, construction period, break
observation period, pipe material, pipe diameter, soil type, location, and depth.
Equation [7] can be written as:
log = log + + + + ⋯+ [8]
Then, the expected number of breaks is given as:
= exp ( + + + ⋯+ ) [9]
For this study, Equation [9] is expressed as:
© ASCE
Pipelines 2015 1189
= exp + ( )+ ( ) [10]
The maximum likelihood method can be used to estimate the parameters of

the Poisson log-linear model as detailed in Simonoff (2003). The residual deviance,
defined as the difference in the deviance of proposed model and the deviance of a
saturated model, can be used for model checking and to perform goodness-of-fit test
for the overall model as explained in Agresti (2002). The saturated model is a model
that fits the data perfectly.
Table 4 presents the water mains break data in a contingency table where
pipes installed during 1850 to 2010 are grouped along with the corresponding number
of failures during the period 1986 to 2010. The table also shows the number of days
the water mains were in service before the break. Table 5 shows the partial dataset
where each case shows the observed number of breaks and covariates such as water
mains’ construction period, observation period, and days in service. For example,
Case No. 29 shows that there were 68 water main breaks observed during the period
2001 to 2005 for water mains installed from 1951 to 1960, and water mains were in
service for 17896 days. There are 50 such cases in total for the data presented in this
paper.
Log-Linear model results and discussions. Table 5 shows the parameter estimates
and deviance goodness-of-fit statistics. Note the days in service were converted to
months before fitting the model. is interpreted as the log expected count of breaks
for pipes that were installed before 1901 and were observed for breaks during the
period 1986-90. ( ) represents the ratio of expected number of breaks for
pipes constructed in 1951-60 and observed in 1986-90 compared to pipes constructed
before 1901 and observed in 1986-90. Figure 3 and Table 6 show that the predicted
values are very close to the observed values confirming that the model fits the data
well. This is further confirmed using the goodness-of-fit chi-squared test which is not
statistically significant because − = . is greater than the usual
significance level of = . .
For Poisson distribution, variance is equal to its mean. However, when
Poisson regression is used on count data, as presented in this paper, variance can
increase faster than the predicted mean. This phenomenon is known as
overdispersion. To check for overdispersion, Figure 4 shows the standardized
residuals versus the predicted values. The majority of points in Figure 4 are within
two standard deviations (shown as horizontal solid lines at ±2). Therefore, there is no
evidence of overdispersion. The model accounts for 96.34% of deviance as shown in
Table 5.
© ASCE
Pipelines 2015 1190
Table 3. Water Main Breaks Data.

Construction Break Observation Period
Period 1986-1990 1991-1995 1996-2000 2001-2005 2006-2010
<1901 No. of Breaks 8 5 1 3 1
Days in Service 37254 33967 35793 37619 39446
1901-1910 No. of Breaks 4 7 4 7 0
Days in Service 30315 32506 34332 36158 37985
1911-1920 No. of Breaks 5 3 5 3 2
Days in Service 26663 28854 30680 32506 34333
1931-1940 No. of Breaks 2 3 11 4 2
Days in Service 19358 21549 23375 25201 27028
1941-1950 No. of Breaks 6 10 7 14 4
Days in Service 15705 17896 19722 21548 23375
1951-1960 No. of Breaks 48 53 61 68 17
Days in Service 12053 14244 16070 17896 19723
1971-1980 No. of Breaks 13 34 25 40 13
Days in Service 4748 6939 8765 10591 12418
1981-1990 No. of Breaks 0 11 16 11 2
Days in Service 1095 3286 5112 6938 8765
1991-2000 No. of Breaks 0 0 11 16 4
Days in Service 0 0 1460 3286 5113
2001-2010 No. of Breaks 0 0 0 1 0
Days in Service 0 0 0 1094 2921
Table 4. Reorganized Water Main Breaks Data.

Case Construction Period Observation Days in Service No. of
No. Period Breaks
1 <1901 1986-1990 37254 8
2 <1901 1991-1995 33967 5
… … … … …
6 1901-1910 1986-1990 30315 4
7 1901-1910 1991-1995 32506 7
… … … … …
10 1901-1910 2006-2010 37985 0
… … … … …
29 1951-1960 2001-2005 17896 68
30 1951-1960 2006-2010 19723 17
… … … … …
49 2001-2010 2001-2005 1094 1
50 2001-2010 2006-2010 2921 0
© ASCE
Pipelines 2015 1191
Table 5. Parameter Estimates.

Estimate Std. Error -vlue (> | |)
-5.71 0.25 -22.60 0.000
( ) 0.28 0.32 0.87 0.383
( ) 0.19 0.33 0.58 0.565
( ) 0.67 0.32 2.11 0.035
( ) 1.47 0.28 5.20 0.000
( ) 3.48 0.24 14.25 0.000
( ) 3.44 0.25 13.65 0.000
( ) 2.91 0.28 10.22 0.000
( ) 3.80 0.30 12.67 0.000
( ) 1.53 1.03 1.49 0.137
( ) 0.16 0.14 1.13 0.260
( ) 0.06 0.14 0.41 0.685
( ) 0.03 0.13 0.25 0.801
( ) -1.44 0.19 -7.72 0.000
: construction period; : observation period

Null deviance: 1219.989 on 49 degrees of freedom
Residual deviance: 43.537 on 36 degrees of freedom, − =
0.1813 > = 0.05
.
Portion of deviance explained by the model = 1 − = 96.43%
.
© ASCE
Pipelines 2015 1192
Figure 3. Observed (o) vs. predicted (p) breaks for the 50 cases (for case
information, please refer to Table 4).
Table 6. Observed and Predicted Breaks.

Construction Period Break Observation Period
1986-1990 1991-1995 1996-2000 2001-2005 2006-2010
<1901 Observed 8 5 1 3 1
Predicted 4 4 4 4 1
1901-1910 Observed 4 7 4 7 0
Predicted 4 6 5 5 1
1911-1920 Observed 5 3 5 3 2
Predicted 4 5 4 4 1
1931-1940 Observed 2 3 11 4 2
Predicted 4 5 5 6 1
1941-1950 Observed 6 10 7 14 4
Predicted 8 10 10 11 3
1951-1960 Observed 48 53 61 68 17
Predicted 43 60 61 66 17
1971-1980 Observed 13 34 25 40 13
Predicted 17 28 32 38 10
1981-1990 Observed 0 11 16 11 2
1991-2000 Observed 0 0 11 16 4
2001-2010 Observed 0 0 0 1 0
Predicted 0 0 0 1 0
© ASCE
Pipelines 2015 1193
2
residuals(model1)
1
0
-1
-2
0 10 20 30 40 50 60
fitted(model1)
Figure 4. Residual plot.
CONCLUSION
The Weibull/Exponential process and Poisson log-linear model provide a

useful framework in which to analyze the deterioration of cast iron water mains.
Using the methods for analyzing survival or failure time data estimates of the
probability distribution of failure times can be obtained. The analysis shows that in
the case of years to first failure, the probability of failure starts out very low and then
increases to a maximum before declining over many years. The MTTF is strongly
related to the break number for a cast iron water main and decreases with increase of
the break number. The results indicate that the MTTF of first to second break, second
to third, and so forth versus number of break for cast iron water main is in agreement
with the results of Gustafson and Clancy (1999). The authors effectively showed that
time to failure between breaks up to the 20th order can be described by Exponential
distributions. Moreover, they observed an almost constant parameter value after the
fifth order. Poisson log-linear model is used to predict the number of breaks per unit
time for water mains installed over different time periods. The proposed models can
be used as a desktop tool for water main failure analysis.
REFERENCES
Agresti, A. (2002) Categorical data analysis (Second ed.), Wiley Interscience,

Canada.
Al-Barqawi, H., and Zayed, T. (2008). “Infrastructure management: Integrated
AHP/ANN model to evaluate municipal water mains’ performance.” J.
Infrastruct. Syst. 14 (4), 305–318.
Eisenbeis, P. (1994). “Modélisation statistique de la prévision des défaillances sur les
conduits d'eau potable (Statistical modeling of drinking water pipes break
prediction).” PhD thesis, Université Louis Pasteur Strasbourg.
© ASCE
Pipelines 2015 1194
Fahmy, M., and Moselhi, O. (2009). “Forecasting the remaining useful life of cast
iron water mains.” J. Perform. Constr. Facil., 23(4), 269–275.
Gustafson, J. M., and Clancy, D. (1999). “Modeling the occurrence of breaks in cast
iron water mains using methods of survival analysis.” Proc. 1999 Annual
Conference of the AWWA, Chicago.
Gustafson, J. M., Osman, H., Bainbridge, K., Homeniuk, R., and Macey, C. (2008).
“Forecasting long-term funding needs for cast iron water mains from analysis of
failure histories.” Proc., 2008 Annual Conf. of the Western Canada Water Works
Association, Edmonton.
Kalbfleisch, J. D., and Prentice, R. L. (1980) The Statistical Analysis of Failure Time
Data, Wiley, New York.
Kleiner, Y., and Rajani, B. (2002). “Forecasting variations and trends in water main
breaks.” J. Infrastruct. Syst., 8 (4), 122–131.
Kleiner, Y., and Rajani, B. (2010). “I-WARP: Individual water main renewal
planner.” Drinking Water Eng. Sci., 3, 71–77.
Mailhot, A., Pelletier, G., Noel, J. F., and Villeneuve, J. P. (2000). “Modeling the
evolution of the structural state of water pipe networks with brief recorded pipe
break histories: methodology and application.” J. Water Resource, 36(10), pp.
3053–3062.
Mirza, M. S., and Haidar, M. (2003). “The state of infrastructure in canada:
implications for planning and policy.” Infrastructure Canada, Ottawa, Canada,
available online at http://www.regionomics.com/infra/Draft-July03.pdf.
Osman, H., and Bainbridge, K. (2011). “Comparison of statistical deterioration
models for water distribution networks.” J. Perform. Constr. Facil., 25(3), 259–
266.
Pelletier, G., Mailhot, A., and Villeneuve, J. P. (2003). “Modeling water pipe breaks:
three case studies.” J. Water Resource Planning Management. Vol.129, pp.115-
123.
Shamir, U., and Howard, C. D. D. (1979). “Analytical approach to scheduling pipe
replacement.” Journal of AWWA, 71(5), 248-258.
Simonoff, J. S. (2003). Analyzing categorical data. New York: Springer.
Stephens, K. S. (2012) Reliability data analysis with excel and minitab, American
Society for Quality, Quality Press. Milwaukee. USA.
Tabesh, M., Soltani, J., Farmani, R., and Savic, D. (2009). “Assessing pipe failure
rate and mechanical reliability of water distribution networks using data-driven
modeling.” Journal of Hydroinformatics, 11(1), 1-17.

Wang, Y., Zayed, T., and Moselhi, O. (2009). “Prediction models for annual breakage
rates of water mains.” J. Perform. Constr. Facil., 23 (1), 47–54.
© ASCE
Pipelines 2015 1195
Rehabilitation and Replacement of the East Layton Pipeline

Adam Murdock, P.E.1; Judd Hamson2; Matt Rasmussen, P.E.3; and Darren Hess, P.E.4
1
Project Manager, CH2M Hill, 4246 South Riverboat Rd., Suite 210, Taylorsville, UT 84123. E-mail:
adam.murdock@ch2m.com
2Vice President, Whitaker Construction, 44 South 1050 West, Brigham City, UT 84302. E-mail:
juddh@whitcon.com
3Project Manager, Weber Basin Water Conservancy District, 2837 East Highway 193, Layton, UT
84040. E-mail: mrasmussen@weberbasin.com
4Engineer Department Manager, Weber Basin Water Conservancy District, 2837 East Highway 193,
Layton, UT 84040. E-mail: dhess@weberbasin.com
Abstract
The East Layton Pipeline is a critical aging potable waterline of the Weber Basin Water
Conservancy District’s (District or Owner) delivery system that serves as a major source of
drinking water supply for the cities of Layton, Kaysville, Fruit Heights, and Farmington, which
are located along the Wasatch Front just north of Salt Lake City. The original East Layton
Pipeline was constructed in 1955 and includes segments of 30-inch-diameter bar wrapped pipe
(AWWA C303) and nonstandard 27-inch-diameter centrifugally cast, reinforced-concrete
pressure pipe with rubber gasket joints. Starting in 2010, the pipeline experienced an increased
number of leaks requiring emergency repairs, which severely jeopardized the District’s ability to
provide a reliable water supply. The pipeline operates with limited system storage and could
accommodate shutdown periods of no more than 24 hours during low demands. This paper
documents the condition assessment of the existing pipeline, evaluation of alternatives for
rehabilitation and replacement of the pipeline, and design and construction of the East Layton
Pipeline using a Construction Manager/General Contractor (CMGC) arrangement between the
owner, engineer and contractor. The rehabilitation and replacement evaluation resulted in
replacing the existing pipeline with a 36-inch-diameter welded-steel pipeline on a separate
parallel alignment through residential streets and slip lining the existing pipeline with a 24-inch
HDPE pipe to provide system redundancy and ability to meet future demands. Key attributes of
the project include construction of a major pipeline in narrow residential streets, crossing of a
canyon with steep side slopes and potential landslide material, coordinating utility relocations,
surge analysis of existing and future pump stations connected to the pipeline, easement and right-
of-way acquisitions, and a significant public involvement effort prior to and during construction.
This paper provides a detailed discussion of key issues and project challenges associated with the
rehabilitation and replacement of aging infrastructure from the viewpoint of the Owner,
Engineer, and the Contractor and also discusses the use and benefit of a CMGC alternative
project delivery method.
INTRODUCTION
The Weber Basin Water Conservancy District’s (Owner or District) drinking water system is
relied upon by a number of cities and water agencies located in northern Utah. The District’s
East Layton Pipeline is the primary source of drinking water for multiple cities located along the
1
© ASCE
Pipelines 2015 1196
populated Wasatch Front north of Salt Lake City, including the cities of Layton, Kaysville, and
Fruit Heights. It is also a supply source for Farmington City. For some of these Cities this
pipeline provides more than 90 percent of their drinking water and currently there is no alternate
or redundant supply for many of the residences if this pipeline were taken out of service.
The northern 7000 foot long reach of the existing East Layton Pipeline has been experiencing an
increasing number of leaks over the last 5 to 10 years which has increased operations and
maintenance costs. Due to limited system storage, the pipeline cannot be taken out of service for
more than 24 hours during low demand periods, making repairs difficult. Currently the pipeline
is operating at design capacity during the summer months and is undersized for meeting
projected future demands.
PROJECT HISTORY AND BACKGROUND

The original 7,000-foot reach of the East Layton Pipeline was constructed in 1955. At the time of
construction, very few residences existed in the area, and the pipeline was constructed in the
center of a 20‐foot‐wide easement that crosses Hobbs Ravine and then traverses open ground
southeast toward Highway 89. The alignment was selected so that the pipeline would operate by
gravity flow from the Davis North Water Treatment Plant (DNWTP) and to provide a relatively
straight alignment across the open terrain. Figure 1 shows the existing pipeline alignment
overlaid on a 1960 aerial image on the right and the same pipeline alignment overlaid on a 2012
aerial image.
Figure 1. 1960 and 2012 Aerial View of East Layton Pipeline
2
© ASCE
Pipelines 2015 1197
As shown in Figure 1, residential development subsequent to the original construction has

greatly encroached upon the pipeline easement and some areas of the pipeline, including air
vacuum and release vaults that are only accessible by foot through the backyards of homes.
CONDITION ASSESMENT OF THE EXISTING PIPELINE

In the mid 2000’s the East Layton Pipeline started experiencing increased leaks and required
more frequent repairs. In 2012 the District conducted a leakage test of the existing pipeline and
hired a consultant to complete a condition assessment and alignment evaluation study to identify
appropriate actions for repairing or replacing this critical facility.
Description of the Existing Pipeline

The majority of the existing pipeline is 27-inch nonstandard reinforced concrete pipe (RCP). The
pipe is nonstandard in that it does not comply with American Water Works Association
(AWWA) or American Society for Testing and Materials (ASTM) standards. The RCP was
manufactured by centrifugally casting the pipe in a mold spun at high velocity and is often
referred to as “Cenviro” pipe. The steel reinforcing in the pipe wall includes a thin wire mesh
designed for internal pressures of 33 to 54 pounds per square inch (psi). The pipe was
manufactured with standard flared concrete bell and spigot joints utilizing a single rubber-gasket.
There is a section of pipe across Hobbs Ravine that includes approximately 1000 feet of 30-inch
bar-wrapped concrete cylinder pipe (CCP) (in accordance with AWWA C303). The CCP was
selected for this reach to handle the higher pressures (90 psi). The District has experienced only
minimal leakage in the CCP and it appears to be in much better condition than the RCP.
Leaks
In 2012 the District performed a leak detection inspection of the existing line with a Sahara Leak
Detection system which uses acoustics and video data to identify leaks and obstructions or
abnormalities in the pipeline. A total of 15 leaks were identified as part of this survey. Two leaks
were estimated to be between 75 and 128 gpm and the other 13 leaks were estimated at 2 to 75
gpm. All identified leaks were located at joints in the 27-inch RCP pipeline and approximately
half were located near the Oak Lane Pump Station Turnout. The video system also captured
locations of obstructions that included rocks and unknown debris in the pipeline. Some of the
leaks identified were visible at the surface and were being monitored daily by District staff.
Figure 2 shows two of the visible surface leaks located near the Oak Lane Pump Station turnout.
Surge Analysis
A surge analysis of the existing pipeline was performed to determine the impacts of starting and
stopping the Oak Lane Pump Station that is connected to the East Layton Pipeline. The East
Layton Pipeline operates at low pressures (less than 20 psi) and is located on the suction side of
the Oak Lane Pump Station. The results of the surge analysis showed that the pipeline was
experiencing significant down surges when the pump station is shut off and it is expected that
many of the leaks at the pipe joints are a result of cyclical negative surge pressures wearing out
the rubber gaskets. Figure 3 shows a hydraulic profile of the pipeline during a surge event. The
red dots shown on Figure 3 show the location of the discovered leaks. Notice the location of
known pipeline leaks relative to the down surges.
3
© ASCE
Pipelines 2015 1198
Figure 2. Pipeline leaks that surfaced near the Oak Lane Pump Station
East Layton Pipeline Hydraulic Profile

Showing Max and Min Surge HGL
4950
4900
4850
Oak Lane PS
Elevation (ft)
4800
Pipe Elevation
4750
Steady HGL
4700 Min HGL

Max HGL
4650 Minor Leak
Major Leak
4600
0 1000 2000 3000 4000 5000 6000 7000 8000
Station (ft)
Figure 3. Hydraulic profile showing surge HGL and location of known pipe leaks
4
© ASCE
Pipelines 2015 1199
Repairs
The existing pipeline has been uncovered
multiple times to repair leaks. In general
the concrete pipe wall and cylinder is in
good condition but failure of the pipe is
occurring at the rubber gasket joints
because of age, suspected pressure surges
in the system, and tree root growth into
the joints. The leaks have compromised
pipeline function and added risk to the
system and to the District. Historically
temporary repairs have been made to the
joints by tightly packing the exterior of the
joints with grout or concrete encasing the
joints. However this method has not
always been successful and often serves Figure 4. Repair of concrete joint with packed grout
only as a temporary fix. Figure 4 shows a
joint of the 27-inch RCP being packed with grout during a leak that required repair in the spring
of 2013.
After several unsuccessful attempts at repairing a
leaky joint and having other leaks show up in the
general vicinity shortly thereafter in the spring of
2013 the District was required to make an
emergency repair to the pipeline which included
slip lining approximately 1000 feet of the
existing 27-inch pipeline with 24-inch outside
diameter HDPE pipe. The District was able to
install some temporary emergency cross
connections with Layton City’s distribution
system which allowed the pipeline to be taken
out of service for about a 5 day time period. This
down time put a major stress on the rest of the
water system and luckily the repair was made in
the early spring when the pipeline was not
operating at peak demands. The 1000 foot
section eliminated 8 of the 15 leaks found during

Figure 5- Slip-lining existing 27-inch RCP with the leakage test conducted in May of 2012.
24-inch HDPE Figure 5 shows slip lining the existing 27-inch
pipe during an emergency repair.
Hydraulic Capacity and Increasing Demands

The existing pipeline was designed to provide a design flow of 8,400 gpm at the upstream end
and 7,700 gpm at the downstream end. The pipeline operates under gravity flow conditions,
with certain reaches operating at pressures less than 20 psi. These low operating pressures make
it difficult for air valves to seat properly and are below the recommended AWWA minimum
operating pressures of 20 psi for a drinking water pipeline. At these low operating pressures
5
© ASCE
Pipelines 2015 1200
any down surges in the pipeline can cause the air valves to become unseated and then quickly
slam closed when pressures are regained.
The 2012 peak hour summer demands in the pipeline were 8,050 gpm. Increasing flow rates in
the pipeline will continue to reduce seating pressures at air valves and increase operation and
maintenance (O&M) requirements. The District’s forecasted 50-year demands require doubling
the pipe capacity (8,050 to 15,600 gpm). Rehabilitation and replacement alternatives will need to
consider the current and long term demand requirements.
REHABILITATION AND REPLACEMENT ALTERNATIVES

Following the condition assessment of the existing pipeline it was apparent that the pipeline
could not be taken out of service for any extended period of time and rehabilitation/replacement
alternatives investigated would require the installation of a separate pipeline on a parallel
corridor that could at least meet existing and near future demands. The District established the
following primary objectives.
• Meet Long Term Demands: The rehabilitation and replacement options considered
must address the 50-year future (2060) peak hour demand of 15,600 gallons per minute
(gpm). The 2012 peak hour demand was 8,050 gpm which is approximately the design
capacity of the existing pipeline.
• Increase System Reliability: The East Layton Pipeline is a major drinking water source
for Layton, Kaysville, and Fruit Heights and cannot be taken out of service for repairs
during the peak summer months (June through September). During other times of the
year the pipeline can be taken out of service for no more than a 24-hour period. A project
objective is to increase the reliability of the District’s water delivery system by
formulating cost-effective ways to replace the existing pipeline and identify rehabilitation
options for the existing pipeline that can improve system reliability and facilitate
maintenance needs.
• Reduce Operation and System Maintenance Costs: The District desires to implement
rehabilitation/replacement options and system improvements that reduce O&M costs and
down time in delivery of water to the cities. The existing pipeline is near or at the end of
its useful design life. The District has experienced frequent pipe leaks causing high O&M
costs. Additionally, many of the air valves on the pipeline are located in the back yards of
residential homes and are challenging to access. Pump station surges on the line are
contributing to the leaks and are causing increased maintenance on air valves that seat
and unseat when the pump stations shut off.
Because the existing pipeline can only be removed from service for 24 hour time periods, the
District was interested in installing a parallel pipeline with the option to rehabilitate the existing
pipeline in the future. Five separate pipeline alignments that provide a replacement for the
existing pipeline were identified and are shown in Figure 6. Each alignment was evaluated
against the Districts objectives along with present worth costs and other engineering factors.
Table 3 provides the advantages, disadvantages, and present worth cost estimate of each
alternative. A brief description of each alternative is provided below.
6
© ASCE
Pipelines 2015 1201
Figure 6. Alternative alignments for the Replacement of the East Layton Pipeline
Alternatives 1 and 2. Alternatives 1 and 2 were selected because they avoided residential
neighborhoods and follow the alignment of a major state highway and frontage road. Portions of
these alignments are above the hydraulic grade line of the DNWTP and require a pump station to
be built which greatly impacts the capital and long term operation costs of these two alternatives.
Alternative 3. Alternative 3 shares the alignment of Alternative 2, but includes tunneling to

avoid the need for a pump station. Tunneling represents a significant project risk and increased
capital cost, but lower long term operating costs without the need for a pump station.
Alternative 4. Alternative 4 follows the same corridor of the existing pipeline. Because the
existing easement is only 20 feet wide and significant development has occurred over and around
the easement over the years, some segments of Alternative 4 would be tunneled to reduce the
impact to local residents. This alternative includes significant challenges associated with the
installation of a second pipeline parallel to the existing unrestrained and leaking pipeline, while
keeping the existing pipeline in service. This presented significant challenges and risks.
Alternative 5. Alternative 5 was developed to provide an alignment that would eliminate

construction in the backyards of residential neighborhoods and also allow gravity flow. This
alternative requires construction of a large diameter pipeline in residential streets and disruption
to the neighborhood.
7
© ASCE
Pipelines 2015 1202
TABLE 3
Advantages, Disadvantages, and Estimated Total Present Worth Costs of Alternatives 1 Through 5
Alternative/PWC Advantages Disadvantages
Alternative 1: - Limited disturbance in residential streets. - Pump station required with high
Pumped Pipeline - Completely separate route from existing O&M Costs
in Frontage pipeline, limiting risk of compromising - Future widening of Highway 89
Road existing pipeline, and stays away from could require relocation of the
Present Worth residents. pipeline.
Cost :
- Construction congestion on major
$9,300,000 commuter route
Alternative 2: - Limited disturbance in residential streets. - Pump station required with high
Pumped Pipeline - Completely separate route from existing O&M Costs
in Highway 89 pipeline, limiting risk of compromising - Future widening of Highway 89
Present Worth existing pipeline. could require relocation of the
Cost : pipeline.
- New pipeline would be out of the private
$7,600,000 backyards of residents - Construction congestion on major
commuter route
Alternative 3: - Gravity flow pipeline - Tunneling required with significant
Gravity - Limited disturbance in public streets. amount of unknowns related to
Pipe/Tunnel in cost and constructability
Highway 89 - Completely separate route from existing
pipeline, limiting risk of compromising - Future widening of Highway 89
Present Worth could require relocation of the
existing pipeline.
Cost : pipeline in the future.
$8,300,000 - New pipeline would be out of the private
backyards of residents - Increased visibility and congestion
on major commuter route
Alternative 4: - Gravity flow pipeline - Disturbance in public streets and
Gravity - Shorter than Alternatives 1, 2, and 3 backyards of residents.
Pipe/Tunnel in - High risk of compromising
Existing - Within the existing easement
existing pipe through construction
Easement and soil disturbance.
Present Worth
- Tunneling with significant amount
Cost :
of unknowns and risk
$6,900,000
Alternative 5: - Gravity flow pipeline - Disturbance to residential streets.

Gravity Pipe in - Lowest Cost Option - Construction challenges of utility
Westerly Streets crossings in public streets.
- Pipeline located in existing easements or
Present Worth
public street rights of way. - Impact to residential neighborhood
Cost :
and will require significant public
$5,900,000 involvement effort.
Recommended Replacement and Rehabilitation Alternative

Based on a qualitative comparison of the alternatives, as well as a comparison of the present
worth cost estimates, Alternative 5 is the best available alternative for installation of a parallel
pipeline in combination with rehabilitating the existing pipeline. Key factors that led to the
8
© ASCE
Pipelines 2015 1203
selection of Alternative 5 as the preferred alternative include: a reduction of risk during

construction by avoiding construction within Highway 89 which would likely require some of
the pipeline to be relocated when the highway is widened in the future, avoids tunneling, and
provides the ability to operate the system through gravity flow (no pump station required). It was
recommended that a 36-inch diameter pipeline be installed capable of delivering at least 10,000
gpm which will meet water demands for the next 10 to 15 years. This will allow the District to
defer the cost of rehabilitation of the existing pipeline and provide additional time to take the
existing pipeline out of service and further investigate rehabilitation options.
CONSTRUCTION MANAGER/GENERAL CONTRACTOR PROJECT
DELIVERY
Because of the difficulty associated with construction of a large diameter pipeline located in
residential streets, the increased leakage that was occurring on the existing pipeline, and the
desire to accelerate the design and construction schedule, the District and their design engineer
selected an integrated team approach to deliver the project by using a Construction
Manager/General Contractor (CMGC) project delivery method. This allowed the Contractor to
provide input during design, and suggestions related to design, cost, schedule, and
constructability issues.
The project schedule was structured to allow multiple design activities to occur concurrently and
keep the project moving forward at an efficient and accelerated pace. The contractual
arrangement and benefits of the CMGC project delivery are discussed below.
Contractual Arrangement
Once the District was ready to move forward with the design of a new pipeline on a parallel
alignment they hired a design engineer to prepare the final design and contract documents for the
project. They also asked that the design engineer support the District with the procurement of a
contractor to provide CMGC services during design which also included the option to move
forward into construction if the pricing of the project was favorable. Figure 7 below shows the
contractual arrangement between the Owner, Engineer, and Contractor.
Figure 7. Contractual Arrangement of CMGC Project Delivery Method.
Using the CMGC delivery process the District was able to select a contractor during the design
phase based on qualifications and best value. The CMGC was asked to provide design reviews at
the 30, 60, and 90 percent design phases along with cost estimates at the 30 and 60 percent
phase. At the 90 percent phase the CMGC was asked to provide an open book Guaranteed
Maximum Price (GMP) so the District could determine if it was competitively priced. If the price
9
© ASCE
Pipelines 2015 1204
was determined to be competitive then the District would move forward with a contract for
construction. If the price was not competitive then the District reserved the right to terminate the
CMGC contract and competitively bid the project similar to a traditional design-bid-build
project.
Benefits of CMGC Project Delivery

Having a knowledgeable and experienced contractor on board during the design process proved
to be very valuable. For a traditional design-bid-build project the design engineer cannot always
predict the contractor’s construction methods, plans for staging of equipment and material,
proposed backfill material that will be used, plans for dealing with traffic control and surface
restoration (paving), and many other items that cannot be established until the construction
phase. With the CMGC delivery process many of these unknowns can be established during the
design phase and gain endorsement from the Owner and jurisdictional agencies. A description of
many of these items are summarized below.
Early and Consistent Public Involvement Team. The CMGC delivery method allowed the
engagement of a public involvement team that started during the design phase and carried
through construction. As part of the predesign phase the contractor was required to provide a
public involvement firm that assisted the District and Design Engineer with several
communication efforts including newsletters, brochures, public open house meetings, and one-
on-one communication with affected residents. With the construction occurring in residential
streets, the District knew that strong communications between project participants would be
necessary for successful completion of the project. Being able to start the public involvement
effort early prepared the community for the construction impact well in advance of any actual
construction taking place. It also allowed the contractor to better understand project constraints
such as coordination with school bus routes and residents requiring special needs.
During construction the CMGC provided a full time public involvement program that included a
24-hour hotline and a project website where residents could receive important announcements,
view detour routes, and track progress of the construction. Figure 8 shows construction between
two residential homes. Figure 9 shows construction in a residential street.
Utility Investigations and

Relocations. During the design phase
the contractor assisted the design
engineer with utility investigations by
potholing and uncovering existing

utilities. This allowed the contractor to
have first-hand knowledge of the
location of key utility crossings, better
understand soil conditions that could
be expected during trench excavation,
and pavement conditions of the
existing roadways. As part of the
potholing and geotechnical
investigations completed during the
design phase, existing utilities on the
Figure 8. 36-inch Pipe Installation Near Residential Homes
10
© ASCE
Pipelines 2015 1205
pipeline alignment were marked in the field by utility locating crews. This provided the
opportunity for the design engineer and contractor to meet in the field and discuss approaches to
utility relocations and installing the pipeline at heavily congested utility crossings and
intersections. Installing a large diameter pipeline in residential streets often requires utilities to be
relocated and there are multiple pipeline alignments that can be selected depending on the
utilities considered for relocation. Gaining agreement during the design phase between the
contractor and engineer on approaches to dealing with existing utilities proved to be very
valuable.
Backfill Materials. Based on the geotechnical investigations completed it was apparent that
much of the pipe zone backfill material would need to be imported. This can represent a
significant amount of the construction cost. During the design phase the contractor was able to
identify local sources of material and provide submittals on the cost and characteristics of
available materials. This allowed for the review and approval of backfill material, including unit
cost, by all parties prior to construction.
Traffic Control. For pipeline

projects designed in traffic congested
roadways it is not uncommon for the
design engineer to provide
preliminary traffic control drawings,
detour routes, and specifications that
outline the contractors requirements
and constraints for maintaining
traffic during construction. This is
primarily done to establish working
limitations and traffic control
requirements that will be enforced
during construction. For a traditional
design bid build project, the final
details of the traffic control and
detour plans are left for the
contractor to negotiate with the
jurisdictional agencies during
construction. The CMGC delivery
approach allowed the contractor to

Figure 9. Pipeline Construction in Residential Streets prepare traffic control plans and
obtain the Cities approval prior to the
award of the construction contract. This eliminated the pricing risk associated with traffic control
and also provided early information to the public involvement team that could be used to help
prepare residents for construction.
Negotiation and Review of Open Book Construction Costs. During the design phase the
Contractor provided open book cost estimates at the 30 and 60 percent design phase and a GMP
following the 90 percent submittal. The design engineer also provided independent cost
estimating at each submittal phase. The District elected to not share the details of the engineer’s
11
© ASCE
Pipelines 2015 1206
cost estimate with the Contractor but indicated to the contractor if they felt certain items were
priced fairly or not and also indicated if they desired to reduce costs of major components of the
project. The contractors open book costs allowed the District and their design engineer to see in
detail how the contractor was pricing the work and provided opportunities to work with the
contractor to identify potential cost savings as well as adding value to certain aspects of the
project. Some of the cost saving ideas included backfill material selection, insurance and
bonding requirements, and assigning final road restoration requirements to Layton City who had
already established a low cost
Construction Cost Estimates paving contract with another
Engineer Estimates and CMGC Open Book Costs contractor. Added value items
included the use of higher
$6,000,000
quality isolation valves,
$5,500,000
increasing the diameter and
pressure class of the HDPE
$5,000,000 pipe installed at the Hobbs
Ravine Crossing, and setting
$4,500,000 aside contingencies for
unknown utility conflicts and
$4,000,000
30 Percent 60 Percent 90 Percent
relocations. Figure 10 shows
the engineer and contractor’s
Engineer CMGC costs estimates during the
three design phases.
Figure 10. Engineer and CMGC Construction Cost Estimates
SUMMARY AND CONCLUSIONS

Many of the critical drinking water facilities installed along the Wasatch Front have reached or
exceeded their intended design life and are starting to fail. Rehabilitating and replacing the East
Layton Pipeline presented significant challenges due to the fact that the pipeline could not be
taken out of service for extended periods of time. This represents a similar situation for most
major pipelines that serve as the backbone of water infrastructure systems. Over the last 50 to 60
years residential development along pipeline corridors has greatly reduced the ability to access
existing pipelines for maintenance, repairs, and replacement. Replacing the East Layton Pipeline
on a new separate alignment allows the District to meet water demands for the next 10 to 15
years and the opportunity to take the old pipeline out of service for repairs and rehabilitation.
Once the old pipeline is rehabilitated, the parallel pipeline system will provide the District with
more reliability and redundancy to meet water demands for the next 50 years.
Using the CMGC project delivery approach the District was able to engage an experienced
design engineer and contractor that were able to address significant project challenges associated
with constructing a large diameter pipeline through residential streets. Together the project team
was able to identify cost savings and added value ideas throughout the design and construction
phase that allowed a quality project to be completed within budget and without major surprises
or cost increases during construction.
12
© ASCE
Pipelines 2015 1207
Validating “Fully Structural”: Development and Testing of a New Carbon

Composite in situ Pressure Barrier for Trenchless Rehabilitation of
Small-Diameter Pressure Pipelines
N. Meyer1; Scott Arnold1; and G. Bontus2

1
Fyfe Co. LLC, 3940 Ruffin Rd., San Diego, CA 92123. E-mail: nico@fyfeco.com
2
Aegion, 17988 Edison Ave., St. Louis, MO 63005. E-mail: gbontus@aegion.com
Abstract
Trenchless technologies similar to Cured in Place Pipe (CIPP) continue to be

embraced for pipeline repair and rehabilitation since the first installation in the early
1970’s. The varying levels of repair from corrosion mitigation, leak protection, to
semi-structural and fully structural repair systems require alternating levels of
strength, stiffness and durability properties under loading conditions. ASTM F1216 is
widely used in water and wastewater pipelines to determine the required thickness of
composite liners for semi structural (class II and III) and fully structural (class IV)
repair/rehabilitation systems as defined in AWWA M28 Appendix A. It is important
to understand the initial assumptions and limitations of these design guidelines. The
ASTM F1216 was developed for felt-epoxy CIPP systems that demonstrate quasi-
isometric properties. Likely because of this, longitudinal loading is not considered in
this design process. As CIPP products continuously develop to resist increasing
external and internal loading conditions, stronger materials are used in specific
orientations to meet those increasing demands. When unidirectional glass and carbon
reinforced polymers (GFRP and CFRP’s) are used to meet the demands of high
internal and external loading conditions, additional design criteria are required to
cover both hoop direction and longitudinal loading. These additional design criteria
extend beyond ASTM F1216. Thus, theoretical calculations from existing and
developing pipeline standards, and experimental validation is required to demonstrate
the capabilities of new technology utilizing high strength and high stiffness materials
like CFRP. This paper will address the additional design considerations appropriate
for CFRP pull-in-place rehabilitations and the validation of a fully structural CFRP in
situ pressure barrier for small diameter (six to fourteen inch) pressure pipe.
INTRODUCTION
This paper will cover practical considerations for CFRP applications for rehabilitation
of pressure pipeline while looking at a few relevant design codes and guidelines and
the assumptions made within these documents. When applications extend beyond the
© ASCE
Pipelines 2015 1208
scope/initial assumptions of such criteria, the need for coupon, element and full scale
testing becomes imperative to validate good engineering judgment. This paper makes
the claim that assumptions written or implicit need to be well understood when
designing, when installing, and when writing specifications/awarding a bid.
Engineers, manufacturers, installers, inspectors and pipe owners cannot afford to
waste time and money to install a product that does not effectively work. Providing
an appropriate solution requires proper specification writing, design and validation
through testing especially when new technologies are developed and brought to
market. In addition, this paper will also discuss the process of validating the claim of
a fully structural pipeline rehabilitation technology, in this case, a CFRP, in-situ,
pressure barrier for small diameter pipe.
Prior to the development of CFRP trenchless technologies for large diameter pipe
renewal and strengthening, CIPP products have provided efficient semi-structural and
fully structural repairs. CIPP rehabilitation systems gained popularity for repairing
and ‘replacing’ degrading pipe since the 1970’s due to the ability to add to the
capacity or take on the entire capacity of a pipeline without digging up the pipeline.
This product originally consisted of a felt sock, impregnated with epoxy or other
resins such as polyester or vinyl ester which cured after the system was pulled in or
inverted into the existing pipeline which acts as a mold for the rehabilitation system.
This application was originally for gravity lines with limited internal pressure.
Because of its success, manufacturers expanded its capabilities, adding additional
reinforcement in the hoop direction to take on increased internal pressures when
necessary. This advancement has drastically increased the effectiveness of this
technology, however design guidelines do not, nor can they be expected to, evolve as
fast as developing technologies. Because of the rapidly degrading pipelines in
America, a large array of technologies have been developed and implemented to
rehabilitate water mains. Beyond the simple reinforcement of CIPP Liners with
unidirectional CFRP or GFRP oriented in the hoop direction, many
manufacturers/contractors have provided hand applied carbon and fiberglass
composite systems when pipe diameters allowed manned entry and application,
typically above 30” diameters. The small diameter CFRP rehabilitation technology
represents the first pull-in place, CFRP in-situ pressure barrier and currently does not
directly fall within the scope of recognized design guidelines and codes.
STANDARDS, CODES AND GUIDELINES
Engineering organizations like American Water Works Association (AWWA) have

developed documents for design, installation methods and potential rehabilitation of
water mains. AWWA M28: The Rehabilitation of Water Mains provides an overview
of the process used for water main rehabilitation covering technologies with a proven
track record within the water industry (AWWA M28 (Foreword) 2014). This
© ASCE
Pipelines 2015 1209
currently is the only document that defines the structural classification of lining
techniques, in Appendix A, which serves as the industry standard for defining a fully-
structural or structurally independent liner. AWWA M28 has defined four
classifications of liners: Class I for corrosion, Class II and III for partial internal
pressure and spanning gaps and holes, and Class IV for the entire internal pressure
with no strength of the host pipe. Because this paper will be focused on validating a
fully-structural liner and the coinciding design considerations, the reader should
peruse AWWA M28 for further explanation of the 1st three classifications. The fully
structural definition is broken down into two main requirements from the technology:
“1. A long-term (50-year) internal burst strength, when tested independently

from the host pipe, equal to or greater than the MAOP (Maximum Allowable
Operating Pressure) of the Pipe to be rehabilitated
2. The ability to survive any dynamic loading or other short term effects
associated with sudden failure of the host pipe due to internal pressure loads”
(AWWA M28 2014)
The document then appropriately clarifies that although such “linings are sometimes
considered to be equivalent to replacing the pipe, they may not be designed to meet
the same requirements for external buckling or longitudinal/bending strength as the
original pipe” (p112 AWWA M28 2014). This clarification can easily be overlooked
and can lead to a fully structural repair that significantly relies on the strength and
remaining performance life of the host pipe. AWWA M28 does not provide design
guidelines for the Class IV lining but its reference to ‘fully structural’ implicitly
references American Society for Testing and Materials (ASTM) F1216: The Standard
Practice for Rehabilitation of Existing Pipelines and Conduits by the Inversion and
Curing Of a Resin-Impregnated Tube. This is widely used in the water industry as the
design methodology for CIPP and other structural lining systems including fiber
reinforced polymer technologies. However, the use of this design guideline far
outreaches its scope and original intent.
ASTM F1216 was originally designed for resin impregnated flexible tubes for use in
gravity and pressure applications for 4 to 108-in diameters (ASTM F1216 – 09).
Resin impregnated felt is a quasi-isotropic material which has essentially equal
strength in every direction. Likely because of this, the guideline only takes into
account the hoop stresses on the pipe. Longitudinal or axial stresses like thrust,
moment or thermal expansion/contraction and Possion’s effects are not considered in
this design standard. Although appropriate for the original CIPP technologies, the
introduction of unidirectional FRP fibers in the hoop direction for pressure pipeline
applications require additional considerations for design. This requirement for
additional considerations becomes more obvious when unidirectional high strength
© ASCE
Pipelines 2015 1210
FRP technologies are used for fully structural designs without the use of resin
impregnated felt at all. These FRP technologies can be seen in hand applied, CFRP
rehabilitation of steel, PCCP, RCP, and many other degrading pipelines in America
and across the world.
Although we must do so in an economic fashion, it is imperative to understand the

life cycle behavior, environment and performance of a pipeline and its relationship
with the rehabilitation technology. ASTM F1216 does not take into account the life
cycle attributes but the design calculations provide a conservative approach to hoop
direction loading for resin impregnated felt CIPP liners. Because unidirectional glass
or carbon fibers are utilized to enhance the strength to thickness ratio of the
rehabilitation technologies, it is imperative to understand the appropriate strength to
consider. Although some carbon fiber systems are three times stronger than steel, the
long term (50-year) design strain limits the effective strength to roughly 40% of their
ultimate capacity based on physical testing (Xian 2008). This is based on Reiner-
Weissenberg (R-W) criterion, a lifetime prediction analysis for sustained strains.
ASTM F1216 does clarify that design properties must be time corrected for estimated
duration of loading, however the strain compatibility with the host pipe, or strain
behavior at all is not considered in this document. Typical rehabilitation technologies
have a lower stiffness than the host pipe which allows the rehabilitation system to
effectively share the load with the host pipe until the host loses structural integrity
due to fatigue or corrosion. It appears important to recognize AWWA M28’s second
qualification for a fully structural liner, that the repair must survive any dynamic
loading due to the failure of the host pipe. Although design guidelines provide these
vital statements, such statements can proceed unenforced throughout the bidding
process removing the effectiveness of the codes/guidelines. One reason ASTM F1216
has been implemented beyond its scope is due to the lack of a guideline or code that
specifically covers unidirectional high strength, high stiffness fiber reinforced
polymer systems for pipeline renewal in the water industry.
When F1216 was not implemented as the controlling design guideline on projects,
CFRP systems were often designed based on the original code utilized during the
initial design. As an example, for steel pipe rehabilitation, FRP was designed with
AWWA M11: Steel Pipe, A Guide for Design and Installation with appropriate
design considerations for the material properties of FRP. This approach can require
owners and other parties to become subject matter experts and know what additional
considerations should be required in the design and installation of these rehabilitation
technologies which can be come thoroughly time consuming. However, when taking
into account the importance of purchasing a product that effectively provides the
desired renewal or strengthening of a degrading water line, parties involved should
consider the concept of validating the claim of a fully structural composite
© ASCE
Pipelines 2015 1211
rehabilitation technology. Such a validation requires good engineering judgment,

understanding the base assumptions of various design guidelines and adjustment of
the design requirements for the validated behavior of developed rehabilitation without
current codes/guidelines. This is well known to many people in the industry such as
AWWA, which has a committee developing a standard for CFRP Renewal and
Strengthening of PCCP.
This AWWA draft standard utilizes an LFRD (Load Factor Resistance Design)
approach for the CFRP renewal and strengthening of PCCP. CFRP has been well
suited for PCCP due to the ability to spot repair PCCP sections that have identified
wire breaks. This document provides a guideline for a 50 to 100 year design life for
CFRP taking into account the entire life cycle and relationship to the host pipe for
either a composite strengthening design or a stand-alone/structurally independent
design.
It remains vital to understand the base assumptions of design considerations. PCCP is

a semi-rigid pipeline that relies on prestressed steel wires to take on all internal
pressures and significant external pressures in the hoop direction. The external
loading of a semi-rigid pipeline is inherently different than a flexible pipe design (i.e.
steel pipelines and CFRP rehabilitation liners). The effective soil loading on a pipe is
based on the relationship of soil settlement relative to the pipe and the corresponding
ability for the soil to help support the pipeline. Semi-rigid and flexible pipes behave
fundamentally different under soil loading. If the host pipe fails due to external
loading, it will likely fail at the hinge regions and then behave like a flexible pipeline
where a CFRP liner now takes on all external forces of the host pipe. The AWWA
M28 document only considers the internal hoop loading (maximum internal operating
pressure) for a fully structural liner, however the external forces during dewatering of
the pipeline may be the governing loading condition depending on the burial depth,
water table depth, and soil modulus.
Although not currently published, the design considerations and load combinations
included in the draft AWWA standard’s design hold relevance as an example of

better design considerations for anisotropic materials used in pipeline renewal and
strengthening. Table 1 shows a direct comparison on some of the considerations
among water industry, oil and gas industry and nuclear industry pipeline codes and
guidelines.
It is also important to consider the determination of the material properties used for
the design, short term and long term. The properties for all CFRP repair system’s
design properties should be found utilizing the appropriate statistical analysis. One
effective method is a Weibull statistical analysis used and implemented on LFRD
designs. The AWWA draft standard requires design properties to be the characteristic
© ASCE
Pipelines 2015 1212
value based on the 80% lower confidence bound on the 5th percentile value of a
specified value according to ASTM D7290: Standard Practice for Evaluating Material
Property Characteristic Values for Polymeric Composites for Civil Engineering
Structural Applications. This test method utilizes the Weibull distribution and allows
for the material properties to be compliant with the base codes of LFRD design
methodology (ASTM D7290 – 06).
Table 1: Guideline/Code Design Consideration Comparison
VALIDATING A FULLY STRUCTURAL LINER
When developing a new product, appropriate testing becomes vital to effectively

prove the theoretically or empirically developed equations for structural behavior.
Because of the complexities of CFRP laminate design, and the typically simplified
initial theories of pipeline design which assume materials are isotropic, it is crucial to
reevaluate the design considerations with good engineering judgment and validate
them through coupon, element and full scale testing. The remainder of this paper will
highlight the process of validating the design of a fully-structural CFRP in-situ

pressure barrier for small diameter (six to fourteen inch) pipeline.
This specific product is designed to be a fully structural repair system for small
diameter pipelines. It utilizes a unique multi-axial hybrid fabric specifically designed
for pipeline loading. The system utilizes a well-known, well tested epoxy that
saturates a low profile carbon and glass hybrid fabric. The epoxy is certified as a
building material after significant long term (10,000 hour) exposure testing under the
ICC ESR AC125 (International Code Council Evaluation Services Report Acceptance
Criteria 125). The Epoxy also has undergone long term performance testing utilizing
the Arrhenius model for both tensile modulus and tensile strength. Gary Steckel’s
© ASCE
Pipelines 2015 1213
accelerated environmental durability testing of this manufacture’s materials

demonstrated a modulus retention exceeding 95% for any exposure temperatures
tested up to 55˚C (Steckel 2015). The testing also demonstrated greater than 95%
normalized tensile strength for exposure temperatures from 22˚C to 49˚C with low
level degradation that increased with increasing exposure temperature. The projected
strength degradation was less than or equal to 10% after 100 years of continuous
exposure to moisture at temperatures up to 55˚C based on Arrhenius analysis of the
testing data.
The major loading conditions and combinations were assessed and the fabric was
designed to specifically orient the strength in the respective loading directions. The
conceptual design has been proven based on testing of large scale, hand applied
CFRP rehabilitation technologies. The small diameter application was initially proven
viable with an installation run, then subsequent material coupon testing such as
ASTM D3039, which tests for tensile properties, and ASTM D790, which tests for
flexural properties, to meet the requirements of various codes and guidelines for the
water industry pipelines.
One major flaw seen in other structural liners was the presence of significant
wrinkles. Surface wrinkles of the coating layer of structural liners are not a structural
concern. However, if the reinforcing structural material layer has a fin which
produces a discontinuity in the path of loading, significant strength reduction results
and a premature catastrophic failure is likely. The ability to install a fully structural
liner without wrinkles ensures the installed product fulfills the design properties. This
particular systems expands and conforms to the host pipe eliminating the presence of
wrinkles for straight pipeline. The installation method used builds confidence in the
initial coupon testing. Because of the industry’s use of ASTM F1216, the
manufacturer performed significant flexural testing on coupons as well. To meet
industry needs, the system was developed to cure in as little as two hours. Curing
profile testing was developed and validated for the product and the coupons were
manufactured with the same cure profile to further validate the actual performance of
the installed product.
Because the coupon tests performed, ASTM D3039 and ASTM D7290, provide a
method to validate fundamental material properties like tensile modulus, E, and do
not necessarily take into account he triaxial state of stress that will occur in the
rehabilitation technology, element and full scale testing was implemented to test both
internal and external loading, along with service connection testing. To validate the
full scale stand-alone performance of this pipeline rehabilitation technology, the
manufacturer produced nominal 8” diameter samples for short-term burst testing in
accordance with ASTM 1599-99: Standard Test Method for Resistance to Short-Time
Hydraulic Pressure of Plastic Pipe, Tubing, and Fittings. The test is an axially
© ASCE
Pipelines 2015 1214
restrained short term burst that mimics worst-case pipeline loading in the hoop
direction without any strength of the host pipe. Figure 1 below, shows an image of a
test specimen.
Figure 1: 8” diameter CFRP rehabilitation technology test specimen in axially

restrained loading apparatus.
Because of the successful short term average burst strength of 1160+/- 90psi for 7
pipe samples installed utilizing the same installation technique and proven cure
regime, the manufacturer performed two 300 ft trial runs (Sheets 2014). This full
scale installation process was implemented to validate the ability to install the product
at the scale required for market use and to test the actual pipe properties during
installation. One 300 ft run was installed into Sonotube® (a cardboard tube acting as
a mold that could be removed after installation), allowing for testing on the CFRP
rehabilitation technology without the host pipe. The second 300 ft install was in 8”
nominal diameter ductile iron pipe with and without mortar lining simulating the
roughness and variability of a degrading pipeline after cleaning.
The Sonotube® trial run produced 8” diameter samples for ASTM 2412-11: Standard
Test Method for Determination of External Loading Characteristics of Plastic Pipe by
Parallel-Plate Loading and ASTM 2290-09: Standard Test Method for Apparent
Hoop Tensile Strength of Plastic or Reinforced Plastic Pipe. These field samples also
demonstrated the flexibility of the product with a 30% defection without failure and
no plastic deformation as expected from the linear elastic properties (Hindman 2014).
This adds additional confidence in the life cycle performance. Unexpected external
loading when a pipe is depressurized might cause significant deformation, but such
deformation would be reversed by the internal pressure as a pipeline returns to

service. In addition to supporting the overall application of CFRP’s in pipe, these
tests help further validate not only coupon testing but theoretical design calculations.
In addition to the element testing, coupons were taken from the host pipe in the
longitudinal direction to validate the material properties of the installed pipe which
was used to confirm the material design properties.
The ductile iron trial run implemented a testing program of service connection testing
with variable protrusion depths and surface preparation methods. Samples were made
to test the ability of the rehabilitations technologies’ bond strength, under various
surface preparation, to prevent leaks at service connection locations. Sixteen samples
© ASCE
Pipelines 2015 1215
were tested under a modified ASTM D1598-02 (09): Standard Test Method for
Time-to-Failure of Plastic Pipe Under Constant Internal Pressure, demonstrating
initial short term leak protection up to 300 psi, the design pressure capacity of the
host pipe, for robotically surface prepped ductile iron. This process and culmination
of testing provides a significant portfolio of validation for the rehabilitation
technology as a fully structural liner that extends beyond the requirements of AWWA
M28 as a class IV liner. Although theory serves us well, initial assumptions must be
validated with full scale testing. Despite the significant monetary and time
investment, full scale testing is crucial to validate the claims produced by
manufacturers of developing technologies. Currently, no guidelines/codes have been
developed for CFRP renewal and strengthening of small diameter pipelines for the
water industry because the developing technologies are so new. Because these
technologies are not fully considered in the development of current codes, a
significant understanding of long-term material properties validated through physical
testing are required to extend the ability for technologies to extend the current
codes/guidelines reach for structural applications in addition to driving the
development of additional guidelines and codes.
CONCLUSION
As applications in engineering extend beyond the assumptions of codes and

guidelines, the assumptions need to be reevaluated and additional considerations must
implemented to ensure effective applications with good, sound engineering judgment.
There is no replacement for full scale testing. This is well known to everyone on a
committee or code council working to better equip those involved with the tools to
protect people and our infrastructure. However, it is imperative that we review the
base assumptions on which we are building our understanding and adapt accordingly
to effectively utilize developing technologies to rehabilitate infrastructure and remove
blight.
REFERENCES
1. ASTM D7290-06 (2011), Standard Practice for Evaluating Material Property

Characteristic Values for Polymeric Composites for Civil Engineering
Structural Applications, ASTM International, West Conshohocken, PA,
2011, www.astm.org
2. Hindman, J., (2014). Test Report: 1410013: Determination of external
loading characteristics and apparent hoop tensile strength of a composite
tube sample. Boulder, CO: Microbac Laboratories Inc.
3. Sheets, C., and Alexander, C. (2014). Results of Axially Restrained FRP Pipe
Burst Testing. Rep. Houston, TX: Stress Engineering Service Inc.
© ASCE
Pipelines 2015 1216
4. Steckel, G. L., (2015). Long Term Tensile Property Predictions for Fyfe
Company SCH-41/Tyfo S Carbon/Epoxy Composites from Arrhenius
Analyses of Accelerated Durability Data. Space Materials Laboratory.
5. Xian, G., (2008). Service-life of a FRP Composite as a Function of Sustained
Strain and Stress Level. San Diego, CA: Fyfe Technology Center.
© ASCE
Pipelines 2015 1217
Integrated Technology Applications for Effective Utility Infrastructure Asset

Management
A. D. Applegate, E.I.T.1; and V. L. Robinson, E.I.T.2

1
AECOM, 16000 Dallas Parkway, Suite 350, Dallas, TX 75248. E-mail:
alex.applegate@aecom.com
2
AECOM, 112 East Pecan St., Suite 400, San Antonio, TX 78205. E-mail:
veronica.robinson@aecom.com
Abstract
According to ASCE’s “2013 Report Card for America’s Infrastructure”, the

United States received a D+ for an overall “below average” rating. Drinking water
and wastewater utility infrastructure both received a D rating. Comprehensive utility
asset management is becoming increasingly important as aged and degrading
infrastructure begins to fail. Managers of these utility systems constantly face funding
and scheduling challenges to meet the demands of the country’s growing population.
However, effective utility infrastructure asset management does not always have to
involve daunting or expensive processes. Integration of engineering technologies –
such as geographic information systems (GIS), global positioning system (GPS)-
enabled devices, advanced spreadsheet applications, pole-mounted viewer/recorders
(pole cameras), and web-based user interfaces, yield a multitude of dynamic data and
statistics to end users at the touch of a button during and after infrastructure condition
assessments. This paper details a unique approach to infrastructure asset management
in which managers of small to medium-sized utilities (generally less than 500 miles
of linear assets) can decrease cost and increase productivity through selection and
integration of appropriate engineering technologies.
OVERVIEW
Although a myriad of engineering technologies are available to facilitate

infrastructure condition assessments and asset management applications for small to
medium-sized utilities, the level of detail required is dependent upon the needs and
resources of the client. The integrated approach discussed below has been
successfully deployed at several Department of Defense military bases, specifically
within the United States Air Force (Air Force). Select bases across various
commands have required inspection-based condition assessments to support asset
management programs established within the Air Force. These condition assessments
have also investigated the extent to which aged and degraded utility infrastructure
within the bases contributed to environmental compliance issues.
© ASCE 1
Pipelines 2015 1218
AECOM developed the current infrastructure prioritization model and asset

management process for Air Force clients by first conducting a multi-utility pilot
project at Joint Base Elmendorf-Richardson, Alaska. The approach was then
implemented in subsequent drinking water, wastewater, and stormwater sewer
condition assessments at other Air Force bases across the country. The following
methodology will focus primarily on wastewater applications. Utilization and
integration of four main engineering and data collection technologies – GIS; mobile,
GPS-enabled devices; pole cameras; and web-based user interfaces – is essential to
the successful execution of this approach. Incorporating these technologies as a
comprehensive package ultimately provides a simple, client-sustainable asset
management solution once the initial condition assessment has been performed.
METHODOLOGY
Pre-field Data Evaluation. Conducting utility infrastructure condition assessments

requires spending time in the field, thoroughly inspecting assets on site. Before
deploying field inspection teams, existing system data are combined and connected in
a GIS database (geodatabase) from various sources. Examples of data sources include
construction drawings, utility personnel knowledge obtained through interviews,
computer-aided design (CAD) files, and current information in the geodatabase. Pre-
field data evaluation identifies areas of uncertainty or inconsistency with system
mapping that can be corrected by technicians prior to or after the field assessment. It
can also identify locations requiring detailed, visual inspection during the field
assessment. For example, a utility shop technician could identify a section of pipe in
a wastewater system where blockages have occurred on numerous occasions. In turn,
the pipes in this location would be prioritized for direct inspection with the pole
camera to identify a blockage source and obtain a visual of the pipe condition.
Prior to mobilization, comprehensive system maps are developed in GIS that

identify inspection requirements for field teams and provide a means of hard-copy
documentation for field notes and observations. The field notes function as a
redundancy to assessment data collected in the field with mobile, GPS-enabled
devices.
Field Inspection and Data Collection. Ruggedized, mobile GPS devices with sub-
meter GPS accuracy and user-friendly data collection interfaces and software are vital
to accurate field data collection and continual geodatabase updating. With these
devices, assets are GPS-located, assessed, and edited in the field. Because condition
assessment criteria are different for each asset (i.e. manholes, pipes, and lift stations)
in utility infrastructure systems, customized assessment forms developed in GIS are
incorporated into the data collection software. Pre-loaded drop down menus created
and customized with CartoPac data collection software, are linked with specific asset
attribute fields, ultimately reducing the amount of time spent per asset during
inspections while minimizing typographical errors. Examples of a GPS device and a
mobile data entry form are shown in Figures 1 and 2.
© ASCE 2
Pipelines 2015 1219
Figure 1. Field GPS Instrument

(Leica CS25 GNSS Shown)
Figure 2. Example Device Data Entry Form
During the field investigation phase of the condition assessment, sections of pipeline
are visually inspected using the pole camera according to plans strategically
developed in the pre-field data evaluation phase. The specialty device is a basic pole-
mounted viewer equipped with a camera perched on a telescopic, pivoting rig with
illumination lamps and a remote viewing interface. Sewer manholes are the most
direct points of access to pipes across the system and are thus pre-selected in the
inspection plan. After the condition of the manhole has been inspected and entered
into the GPS device, the pole camera is lowered to pipes so that photographs can be
taken as the condition is assessed by a technician operating the remote viewing
interface. Attributes such as pipe material, diameter, and orientation are also noted
during the inspection to confirm and/or update existing data in the geodatabase.
Assessment data can then be entered into the GPS device as the inspection takes place
© ASCE 3
Pipelines 2015 1220
One of the primary advantages of this approach with respect to pipeline

inspection is that not every foot of pipe is directly evaluated with closed-circuit
television (CCTV) inspections. First, pole camera locations are strategically selected
in areas of uncertainty, where issues have been identified, and/or where pipe material
and age (based upon installation date) are statistically correlated with anticipated
structural condition. For example, a vitrified clay pipe installed in the 1940s is more
likely to be defective and require repair or replacement than a polyvinyl chloride pipe
installed within the past decade. Experience using this method at multiple Air Force
Bases has shown that direct visual inspection of up to 35 percent of total pipe length
within a small to medium size wastewater system is appropriate to gain an overall
condition assessment of similar pipes that are not directly inspected. Essentially, the
length of pipe and locations selected for pole camera inspection need to be sufficient
to extrapolate data when the condition assessment is complete.
Parameters required for the asset management prioritization model are those
by which assets are assessed. For example, structural deficiencies in a pipe – such as
breaks, cracks, corrosion, or deformation – are noted if observed during the field
investigation, along with degree of severity. The pole camera utilized in this
approach has a reported range up to 400 feet, although field use reports a practical
range up to 100 feet in a straight run of pipeline. Ultimately, effective viewing range
of the pole camera can be limited by pipe bends, small diameters, and manhole bench
construction hindering ease of access. Sample photographs of the pole camera in use
during an inspection and a wastewater system pipe taken with the pole camera are
shown in Figures 3 and 4.
Figure 3. A technician operating the pole camera during a pipe inspection
© ASCE 4
Pipelines 2015 1221
Figure 4. Examples of zoomed-in (left) and wide-angle (right) imagery of a

vitrified clay pipe
Data collected in the field is uploaded daily from GPS devices to an online
server throughout the duration of the field investigation. The dynamic integration of
mobile data collection devices and the server where data is stored is facilitated by an
enterprise asset management software solution. This intrinsic, technological
relationship enables field teams to track progress of the assessment in real-time and
ensures that no asset goes overlooked while on site.
Post-field Data Reconciliation. Upon completion of the field inspection, GPS-

updated asset locations and associated condition and attribute data are reviewed for
accuracy and completeness. Remaining data gaps with respect to various assets are
resolved from field observations, often combined with knowledge ascertained from
other sources (i.e. the condition and disposition of surrounding assets). Data gaps can
include unknown pipe materials, buildings without service lines, inconsistent flow
directions, etc. In some instances, developing engineering assumptions with respect to
reconciling data gaps is required. At this point, the process of extrapolating pipe
condition data gathered from pole camera inspections can begin.
Data extrapolation is a function of engineering judgment that applies observed

conditions of assets to those not observed, where appropriate. A cornerstone of this
approach to asset management initially replaces extensive CCTV inspection of up to
thousands of feet of pipeline with pre-selected areas for a quick, direct inspection.
Therefore, conditions must be applied to the remaining percentage of system pipeline
to fulfill requirements of the prioritization model – utilizing pipe installation date and
material, at minimum. Essentially, structural condition of pipes observed in the field
is applied to those not observed as long as the material and installation year are the
same. Where pipe installation dates are not known, assumptions are formulated based
upon the material of the pipe and the median of typical historical use. This method is
only recommended for small to medium-sized utilities.
Upon completion of the data reconciliation process, each asset is assigned a

risk score developed from a risk matrix. As detailed in Table 1, this matrix combines
the probability that an asset will fail with the consequences of failure to stakeholders
(criticality). For example, an asset with a high probability of failure (i.e. a concrete
pipe observed near collapse in the field) and high criticality (i.e. large-capacity
© ASCE 5
Pipelines 2015 1222
wastewater pipe serving several facilities across multiple basins) is an extreme risk to
stakeholders in the event of failure.
Therefore, this pipe yields a numerical score reflecting this extreme risk and is
prioritized for rehabilitation in the form of structural repair or replacement.
Repairing a single segment of pipe in isolation is uncommon after a large-scale
condition assessment has occurred. Therefore, projects comprised of multiple pipes
within the vicinity that also require rehabilitation are recommended for the client.
Table 1. Sample Risk Matrix

Consequence of Failure
Probability
of Failure
Low → Medium → High
Low Negligible Negligible Low Low Moderate
Negligible Low Low Moderate Moderate

→
Medium Low Low Moderate Moderate High
Low Moderate Moderate High Extreme

→
High Moderate Moderate High Extreme Extreme
Limitations. Partial, customized pole camera inspection and post-field condition

data extrapolation of an entire system is an effective strategy for small to medium-
sized utilities with less than 500 miles of pipeline. However, this approach may not
work for larger utilities with hundreds of miles of large diameter pipes (typically
greater than 24 inches). In these systems, man inspection or CCTV surveys are most
appropriate.
While the pole camera can provide a good overall view of the pipe to
determine general condition, it does not provide sufficient information to establish the
most suitable rehabilitation method between repair and replacement. Rehabilitation
methods recommended to the client are based on aforementioned limitations of the
pole camera. This approach to asset management with customized pole camera
inspection has the advantage of expediency and functions as a means to identifying

areas that require more investigation for potentially problematic assets. These
inspections, along with the subsequent extrapolation process, can help direct where
more focused CCTV inspection is implemented as opposed to investing a great deal
of time, manpower, and financial investment in a comprehensive CCTV survey of the
entire system.
For these reasons, thorough engineering analysis should precede each

proposed project in order to verify that the recommended solutions are technically
appropriate. At a minimum, the engineering analysis must include a CCTV survey of
© ASCE 6
Pipelines 2015 1223
identified pipes in order to verify the extent of conditions observed in during the field
inspection.
Client-sustainable Asset Management. Practicality of the prioritization model and

ongoing asset management is dependent on the accuracy and efficacy of underlying
data in the geodatabase. Future facility modifications and sanitary sewer projects
must be simultaneously verified in the field and updated in the geodatabase.
Moreover, the client must be equipped with the user-friendly tools necessary to
update their asset management system.
The client receives an updated geodatabase, advanced spreadsheet

applications that compute risk scores utilized in the prioritization model, and a
dynamic, web-based user interface connected to the geodatabase for generating
individual asset server reports upon completion of the condition assessment. The
spreadsheets, connected to attribute tables within the geodatabase, demonstrate how
attributes are weighted and calculated into the prioritization model for each asset.
Server reports are generated through reporting service software connected to
information also stored within the geodatabase. These server reports then graphically
display attribute and condition data for each asset, along with risk scoring data and
assessment photos from the inspection. End users can readily access server reports on
a webpage without having to open the geodatabase to get information on system
assets. Reports are automatically updated when changes to the geodatabase are made.
There are multiple types of software that can be configured to work together in this
way. AECOM utilized ArcGIS, CartoPac, Microsoft SQL Server Report Builder, and
BIRT (Business Intelligence and Reporting Tools). Figure 5 is an example of a
sanitary sewer pipe server report.
© ASCE 7
Pipelines 2015 1224
Figure 5. A sanitary sewer asset server report detailing pipe condition and
risk data
The prioritization model utilized in this asset management approach is

customizable based on client needs. The primary advantage of server reporting is that
clients can easily track the status of their most critical assets, especially when
developing capital improvement plans and programming and scheduling projects.
This proactive style is preferable to a reactive one, where system managers must
respond to financial and operational consequences after a failure has already
occurred. Figure 6 is a snapshot of a server report generated to track and display
assets with the highest risk.
© ASCE 8
Pipelines 2015 1225
Figure 6. Graphical display of assets with corresponding risk scores
CONCLUSION
Experience has proven that the most nimble and client-sustainable asset
management systems require technology that is relatively affordable and easily

accessible small to medium-sized utilities. Applying integrated technology
applications as essential tools before, during, and after comprehensive infrastructure
condition assessments results in a robust asset management data collection and
organizational methodology. Furthermore, these integrated tools assist clients in
developing effective capital improvement along with operation and maintenance
strategies long after consultants have completed field investigations. Once an
assessment database and risk model are established, they are easily updated by end
users – such as utility managers and field technicians – for ongoing system
management and technical support. The process is simplified as extensive, paper-
based asset reports and costly investigations are eliminated. Integrated asset
© ASCE 9
Pipelines 2015 1226
management technology facilitates a one-stop system, decreases costs, and increases

productivity.
© ASCE 10
Pipelines 2015 1227
Beyond Water Audits into Asset Management: The Process of Non-Revenue

Water Reduction and Revenue Enhancement Activities
Brian Skeens, P.E.1
1
Senior Technologist, CH2M Hill, 6600 Peachtree Dunwoody Rd., Building 400,
Suite 600, Atlanta, GA 30328. E-mail: Brian.Skeens@ch2m.com
Abstract
For many water utilities, the true amount of non-revenue water is a mystery, or at
least a guess. With the growing use of tools like the AWWA Free Water Audit
Software, many utilities are beginning to have a better understanding of how their
water system is operating. But with this realization, comes the next question, which is
how to economically reduce water losses. There are many things that can be done, but
which should be done, and to what level of implementation are the tough questions.
The concept of economic level of losses is important, but most water systems will not
be close to determining this level after first assessing their non-revenue water. This
paper will go describe the approach to identifying the quantity and components of
non-revenue water through an initial audit, bottom-up activities, evaluating, selecting
and prioritizing corrective approaches, and implementing water loss control activities
for effective results. The ability to change course in mid-stream is also important, and
so accurate progress reporting is critical. A comprehensive non-revenue water
reduction program integrates and informs with other utility management functions,
such as asset management. In particular, the level of rehabilitation and replacement of
pipelines is related to the non-revenue water goals and economics. This relationship
between asset management and non-revenue water becomes the cornerstone for
ongoing planning and improvement.
BACKGROUND
For a very long time, water systems used the antiquated terminology of
“Unaccounted-for water” to classify the amount of water lost between production and
sales. However, there was no industry standard definition or calculation for this term,
which was often expressed as a percentage, and translated nothing about the financial
impact of water loss. In order to improve this, the American Water Works
Association (AWWA) through Water Loss Control Committee released a committee
report in August 2003, which outlined a methodology developed in partnership with
an International Water Association (IWA) workgroup. Since that time, the AWWA
and water industry has adopted this IWA/AWWA methodology as the best practice,
(AWWA website). Essentially, this methodology classifies all water that enters the
distribution system into standard categories, so that no water goes “unaccounted-for”.
This allows for a complete water balance to be performed, as shown in Figure 1.
Further explanation of this water balance and methodology is contained in the
AWWA M36 Manual, Third Edition, 2009.
© ASCE
Pipelines 2015 1228
Figure 1. AWWA/IWA Water Balance.

The following are a list of key definitions used throughout this paper:
• Non-revenue water (NRW): the difference between water entering the

distribution system and the water billed to users/customers
• Water loss: the difference between water entering the distribution system and
the water used for authorized purposes; also the sum of real and apparent
losses
• Real loss: water that is lost due to leaks and breaks in pipes, overflows at
tanks, and on service connections up to the point of metering (if applicable)
• Apparent loss: water that is lost to unauthorized use (theft), customer
metering inaccuracies, and systematic data handling errors
In an attempt to increase adoption of this new methodology throughout the water
industry, the AWWA Water Loss Control Committee has developed a Free Water
Audit Software © that can be used to perform a top-down assessment of the water
balance shown in Figure 1. This water audit software is available for free from
AWWA and is used in Microsoft Excel. The software has been tested, revised, and
improved since 2005, and version 5.0 was released in 2014. Some key features of the
software are that the inputs required for the audit are given a “grade” by the auditor so
that the overall “data validity” can be assessed, as well as the calculation of key
operational and financial performance indicators.
OBJECTIVES
Objectives of this methodology are to allow water systems to evaluate their non-
revenue water with increasing confidence and determine activities that can be
implemented to reduce the components of non-revenue water, in an economic
fashion.
© ASCE
Pipelines 2015 1229
NON-REVENUE WATER PROCESS

A process can be outlined to incorporate the water audit in a non-revenue water
reduction program. This program integrates with data from multiple sources
throughout the water system, and connects with operations, asset management, and
billing and customer service in the reduction of non-revenue water. Figure 2 shows
this NRW Process and is described in the paragraphs below.
Figure 2. NRW Process.

INITIAL ASSESSMENT
The use of the IWA/AWWA Water Audit Methodology through the Free Water Audit
Software allows for water systems to perform a top-down assessment. This is called a
top-down exercise because it is primarily meant to be performed using data generally
available. When this is combined with the user-entered data grades, an overall data
validity is calculated, and the water audit generates the top three inputs that with
improvement will improve the data validity.
The top down water audit performed using the Free Water Audit Software calculates
some key performance indicators of water loss. These performance indicators can be
used by the water system as metrics to track their progress over time, as well as
© ASCE
Pipelines 2015 1230
benchmark themselves with other utilities, depending on the indicator. There are
financial indicators, as well as operational efficiency indicators.
DETAILED ASSESSMENT
Activities can be performed to improve the confidence in the audit, by increasing
confidence in the data inputs, as well as the implementation of best practices. These
criteria are outlined in the water audit software, in what is called the grading matrix.
Data grades should be improved to the level at which the overall water audit data
validity provides confidence in the use of the results for non-revenue water reduction
activities. This includes field verification of production meter accuracy, customer
meter testing, evaluation and auditing of billing system practices, and component
analysis of real losses.
One of the most important inputs to the water audit is the water supplied. Therefore,
the first thing a water system should do is consider their confidence in the accuracy of
this number. Activities to verify this number could include using a temporary,
secondary meter to verify the accuracy of the primary flow meter or meters for
finished water or testing of bulk water import meters. An example of this is shown in
Figure 3 below, using a temporary insertion meter to verify the accuracy of the
existing flow meter.
Figure 3. Finished Water Meter Testing using a Temporary Insertion Flow

Meter.
Other important activities during the detailed assessment process include customer
meter testing, assessment of meter reading and billing practices, as well as component
© ASCE
Pipelines 2015 1231
analysis of real losses. The first of these is customer meter testing, which is important
to understand the rate at which metering accuracy may be deteriorating. For instance,
most customer meters in use by water systems are mechanical in nature, and
therefore, wear out over time, resulting in reduction of accuracy, especially at low
flowrates. A program of testing statistically significant number of meters can help a
water system determine their overall customer meter accuracy with a greater degree
of confidence. Figure 4 below shows an example of residential-type meter testing.
Figure 5 shows an example of the results of customer meter testing.
Figure 4. Customer Meter Testing.

© ASCE
Pipelines 2015 1232
Figure 5. Customer Meter Test Results.

Another practice that can be performed to increase the detail and confidence of the
non-revenue water balance is an audit of the meter reading and billing practices, as
well as review of the billing system exceptions reporting. This can be done by simply
creating a flow chart for the process of a meter reading being collected, recorded,
converted into a consumption volume, and a customer bill being created. This can
help create more accurate assessments of apparent losses experienced by a water
system.
For many water systems, real losses are a greater component, on a volume basis, so it
may be worthwhile to understand the components of the real losses. These include
reported leakage, unreported leakage and background leakage. By determining how
much each of these contributes to the overall real losses, appropriate reduction
techniques can be identified. For example, reported leakage and unreported leakage
are essentially the same, except that while the reported leakage usually is visible on
the surface of the ground and is thus reported and repaired, the unreported leakage
does not surface, and may remain leaking for much longer before it surfaces.
Unreported leakage is detectable using current acoustic leak detection technology.
Background leakage, however, is much smaller in individual locations, and thus is not
detectable by these means, but it may still make up a large volume component of the
real losses experienced by a water system. Figure 6 below shows a breakdown of
these types of real losses that occur in a water system. (WaterRF, 2014)
Figure 6. Components of Real Losses.

ECONOMIC LEVEL OF LOSSES
It is important to define the economic level of losses for the purpose of understanding
how water loss affects a water system’s operation and financial health. The economic
level of loss can be applied to apparent losses, or real losses separately.
The economic level of losses can be determined by the cost of various water loss
reduction activities and the value of the water loss that is recovered. Those activities
that create a positive return on investment can be defined as economically justifiable.
© ASCE
Pipelines 2015 1233
This requires the accurate determination of the value of the components of non-
revenue water. These can be determined as shown below:
• Real Loss
o valued at wholesale rate or variable production cost (usually)
o Typically $0.14 - $1.50 per 1,000 gallons
o Could use retail rate in water shortage conditions
• Apparent Loss
o valued at retail rate or customer retail unit cost
o Typically $2.00 - $5.00 per 1,000 gallons
• Unbilled Authorized is based on the type of use
o Valued at variable production cost (in water audit software)
o May be political reasons that prevent recovery of some unbilled
authorized use
It should be noted, that in general, the total volume of the real losses is typically
higher than apparent losses, but the total value of the apparent losses is typically
higher because the unit rates applied are different. This illustrates how the different
components of non-revenue water impact a water system operations in different ways,
as well as the types of water loss control activities each water system may decide to
prioritize.
ECONOMIC AND BUSINESS CASE EVALUATION

Based on the value of the recovered water losses, payback calculations can be
performed to determine projects to be performed that can reduce water loss.
Depending on the components of water loss, various intervention activities can be
performed as shown below:
Real loss reduction:
• Active leak detection to reduce unreported leakage

• Improving the speed and quality of repairs to reduce reported leakage
• Pressure management to reduce background and unreported leakage
• Pipeline rehabilitation/replacement to reduce unreported leakage
Apparent loss reduction:
• Customer meter replacement to improve accuracy

• Automatic meter reading/advanced meter infrastructure to eliminate errors in
the meter reading process
• Business process evaluation including interaction with the following:
o Meter reading system evaluation
o Billing system audit/upgrades
o Customer service interface and access
For each activity, the economic implementation level can be determined. Using the
principles in this paper, water systems can economically justify their investments in
© ASCE
Pipelines 2015 1234
water loss control. In the case of apparent loss reduction, increased revenues can be
experienced, which is why many water systems choose to perform activities related to
apparent water loss control before real losses.
INTEGRATION WITH ASSET MANAGEMENT
Many of the activities listed above are integrated with other components of water
system operation and maintenance activities. In particular, pipeline
rehabilitation/replacement is directly related to asset management practices. If a water
system has a detailed history of water line failures, breaks and bursts, the data can be
analyzed to prioritize portions of a water system to be surveyed for leaks, relined and
rehabilitated, or replaced and renewed. Understanding the benefits to water loss
reduction through these activities can help an asset management program assign
further priority to these projects so they are implemented sooner.
IMPLEMENTATION AND MEASUJREMENT OF PROGRESS
As programs are implemented to reduce the various components of non-revenue
water, the water system should track and report on progress to make modifications, as
needed. This is in agreement with the process shown in Figure 2.
One example of tracking implementation progress can be illustrated by tracking the
miles of pipeline that are surveyed using leak detection equipment. This does not
necessarily guarantee reduction of real losses, but should provide reductions. The
economic goal of miles per year can be used as a metric of progress. If the economic
level of active leak detection was calculated, this can become the target length of
pipes to be surveyed annually.
In addition, it is important to perform the water audit on a regular periodic basis. This
could be annually, using calendar or fiscal or water year 12-month basis. It is not
generally recommended to perform auditing on a 30-day basis, due to the difficulty in
reconciling the customer meters with the source meter data. However, the use of 12-
month rolling totals on a monthly basis has been useful for some utilities to identify
trends and issues. Figure 7 below illustrates the implementation of a 12-month rolling
calculation of water loss.
© ASCE
Pipelines 2015 1235
Figure 7. 12-month Rolling Totals of Non-Revenue Water

Finally, the NRW process shown in Figure 2 should be repeated and progress of
improvements on the performance metrics should be reported. Periods of abnormal or
extreme events may interrupt progress or disrupt calculations, however, and these
should be noted. Examples of this may be large customers reducing their use, or
extreme weather that may impact customer use, or flooding and soil shifting. A
catastrophic failure of a large diameter pipe can also create anomalies.
CONCLUSION
Water systems that implement a comprehensive non-revenue water (NRW)
assessment and control program can benefit from understanding the components, and
the economic impact to their operations. By improving the confidence in their
assessment, they can have greater confidence in the reduction practices they
implement. In fact, the economics of NRW reduction practices can be calculated by

determining the return on the investment and prioritizing those that have a positive
value. It is important to periodically track progress, and report to key stakeholders
and leaders. Finally, the process of NRW reduction is one that takes time, and careful
selection of intervention activities. However, it can and should be integrated into asset
management activities within a water system, to help achieve goals of operational
excellence.
In addition, the AWWA Water Loss Control Committee has collected water audits
from systems that have volunteered to share their results. These results have been
validated, and are published on the AWWA website for informational purposes
(AWWA website).
© ASCE
Pipelines 2015 1236
REFERENCES
Kunkel, G, et. al. (August 2003). “Committee Report: Applying Worldwide BMPs in
Water Loss Control.” Journal American Water Works Association.
American Water Works Association. (2009) M36: Water Audits and Loss Control
Programs, Third Edition.
AWWA. Water Loss Control Resource Community. www.awwa.org/waterlosscontrol
Water Research Foundation. (2014). Real Loss Component Analysis: A Tool for
Economic Water Loss Control. Project 4372a. Denver, CO.
© ASCE
Pipelines 2015 1237
Fully Structural Renewal of 39-inch PCCP Water Transmission

Main with Swagelining™ and HDPE
Todd Grafenauer1; Tom Hayes2; James Vanderwater3; Madhu Kilambi4; and David
Kasper5
1,2
Murphy Pipeline Contractors, Inc., 3507 Southside Blvd., Jacksonville, FL 32216.
E-mail: toddg@murphypipelines.com; tomh@murphypipelines.com
3
Gulf Coast Water Authority, 3630 Highway 1765, Texas City, TX 7759. E-mail:
jvanderwater@gcwater.org
4,5
ARKK Engineers, 7322 Southwest Fwy, Ste 1040, Houston, TX 77074. E-mail:
madhu.kilambi@arkkengineers.com; david.kasper@arkkengineers.com
Abstract
Gulf Coast Water Authority (GCWA) and the City of League City, Texas were faced
with the need to renew a critical 39-inch PCCP water transmission main. The 39-inch
PCCP water main, owned by GCWA and operated by League City, required
replacement of approximately 6,800 feet along Calder Road. This paper will outline
the design phase and selection process as traditional slip lining was considered
including an alternate for a compressive tight fitting HDPE pipe. The design criteria
required a fully structural solution capable of 125 psi operating pressure. The
Swagelining™ process was selected over slip lining due to the additional flow
capacity and the contract was awarded to Murphy Pipeline Contractors. This paper
will also outline the construction phase, including the Swagelining™ process, the
pipe installations performed by Murphy Pipelines, and the challenges associated with
the installation of a 1000 mm (39.4 inch OD) DR 17 HDPE pipe. The 2.32-inch wall
thickness pipe was pulled through a single swage die in four pulls ranging from 1,250
feet to 2,100 feet in length. The Calder Road Project represents the largest diameter,
fully structural pipe installed to date in North America utilizing the Swagelining™
technology. The utilization of this technology with HDPE pipe allowed the owner to
meet all design parameters and increase the flow capacity. Swagelining™ offers a
solution for pressure pipe renewal that is unique in today’s trenchless pressure pipe
market as it meets both internal and external loading requirements.
PROJECT BACKGROUND
GCWA had identified a major potable water supply main in need of replacement. The
39-inch PCCP water transmission main, owned by GCWA and operated by the City
© ASCE 1
Pipelines 2015 1238
of League City, feeds a water plant supplying water to the western quadrant of the
City of League City. The transmission main, originally constructed in 1971 to supply
surface water to the City of Galveston, interconnects between GCWA’s Thomas S.
Mackey Water Treatment Plant and the City of Houston’s Southeast Water
Purification Plant. In addition, this approximately 6,800 foot section of Calder Road
was scheduled for reconstruction and widening.
DESIGN PHASE AND SELECTION PROCESS
GCWA issued a very complex Request for Competitive Sealed Proposals (RFCSP) to
allow GCWA to select the best materials and method for rehabilitation. The proposal
included a base price for slip lining allowing HDPE or PVC. Alternate 1 included slip
lining with higher pressure classes of HDPE or PVC, and alternate 2 included the
Swagelining™ method with HDPE. Although not everything allowed was proposed,
GCWA received proposals from five contractors.
After reviewing the cost and capacity of the pipeline, GCWA along with their partner
the City of League City awarded the highest grade to Swagelining™ with HDPE
based on the evaluation criteria. Final Internal Diameter (ID) with Swagelining™
resulted in 33.86-inches, over 4-inches larger than slip lining with PVC or HDPE
(Figure 1). The City of League City could satisfy their demand with the slip lining
option; however, since the Calder Road potable water transmission line also provides
an interconnect between the City of Houston’s Southeast Water Purification Plant and
the GCWA’s Thomas S. Mackey Water Treatment Plant, GCWA provided funding to
pay the difference in cost between slip lining and Swagelining™.
Figure 1: Final ID comparison between slip lining and Swagelining™.
© ASCE 2
Pipelines 2015 1239
SWAGELINING™ HISTORY AND OVERVIEW
The Swagelining™ technology was developed over 30 years ago by British Gas in
conjunction with United Utilities. With an extensive list of successfully completed
projects across the globe, the technology has been proven in many extreme projects
spanning three decades onshore and subsea. Projects have been completed for water,
sewer force main, mining, hydrocarbons, chemicals, bulk products and gas
distribution. The overall confidence of the technology originates from an extensive
physical testing program conducted by British Gas over several years. The process
was established after extensive analysis of material behavior during and after die
reduction. A major result of the research and development program was the
development of the liner system design software. This software program, which is
utilized for each project, ensures installation stresses do not compromise the integrity
of the HDPE.
The Swagelining™ technology specifies a PE4710 High Density Polyethylene

(HDPE) pipe with an outside diameter larger in size than the inside of the host pipe to
be renewed. After the HDPE is butt fused to correspond to the pull distance, the pipe
is pulled through a single reduction die immediately before entering the host pipe.
This reduces the HDPE pipe temporarily below the ID of the host pipe allowing it to
be inserted (Figure 2). While the towing load keeps the HDPE under tension during
the pull, the pipe remains in its reduced size. The HDPE remains fully elastic
throughout the reduction and installation process. As the liner pipe is not permanently
deformed by Swagelining™, the release of the towing load after insertion is the
catalyst for the liner to revert back towards its original size. As its original size is
larger than that of the host pipe, the HDPE pipe expands until it is halted by the inside
diameter of the host pipe. This produces a residual strain that is locked in the liner
and maintains pressure against the inside of the host pipe, even in the absence of
internal pressure from the product conveyed.
Figure 2: Swagelining™ process as HDPE is pulled through the reduction die.
© ASCE 3
Pipelines 2015 1240
The effectively natural compressive tight fit produced by Swagelining™ provides

value for clients looking to maximize the final ID of the pipeline. Due to the tight fit,
thin walled HDPE liners and semi-structural HDPE pipe can be installed in which
operating pressure is delivered through the host pipe. In circumstances such as the
Calder Road 39-inch PCCP water transmission main which required a fully structural
stand-alone solution, Swagelining™ can install a fully structural HDPE PE4710 pipe
such as DR 17 with a working pressure rating of 125 psi, allowable total pressure
during recurring surge events of 187 psi and allowable total pressure during
occasional surge events of 250 psi (ASTM F714, ASTM D3035 and AWWA C901).
Higher working pressure ratings above 125 psi can also be achieved. In addition to
meeting internal pressure loads, the HDPE installed met all external loading
requirements.
CONSTRUCTION PHASE
Calder Road runs parallel to 1-45 in League City, Texas. The urban area is a
combination of residential and commercial use, with a Big League Dreams Sports
Park in the middle of the project. The project limits encompassed a very tight area as
Calder Road is a single lane two way road in which the allowed work area was no
wider than 25 feet. The 39-inch PCCP water transmission main was located along the
edge of the pavement among a congested utility corridor. While shutting down both
lanes would have eased construction, only a one lane shutdown was allowed.
The rehabilitation of the 6,800 linear feet of the PCCP 39-inch diameter waterline
included the replacement of four 36-inch diameter butterfly valves, additional three
36-inch x 24-inch diameter tees with gate valves, the replacement of two air relief
valves, the addition of three blow-off valves, installation of a 39-inch x 36-inch
flanged reducer, placement of two large thrust restraint blocks, and a bypass (Figure
3, Figure 4 and Figure 5). The placement of valves, blow-offs and tees were
relocated based on the constructability of the project and the needs for the future
expansion of the City of League City Water Plant on Calder Road. The north end of
the project was extended about 300 linear feet to an existing 36-inch diameter
butterfly valve. The south end of the project required the addition of a flanged
reducer, butterfly valve and blow-off. Thrust restraints were required at the north and
south end of the project to protect the existing transmission main from stresses during
construction. Each thrust restraint included an ellison type pipe clamp along with
over 40 cubic yards of concrete. During construction the existing water plant was
connected to the City of Houston water supply or the GCWA water supply through a
12-inch diameter bypass laid mostly above ground. The project layout was designed
to meet the needs of the roadway construction along with the expansion of the Calder
Road Water Plant.
© ASCE 4
Pipelines 2015 1241
Figure 3: Left and middle picture of 24-inch stub out by sidewall fusion with Flange
Adapter connected by downhole butt fusion. Right picture of 24-inch Gate Valve
bolted on to Flange Adapter.
Figure 4: Left picture of downhole butt fusion connecting HDPE in receiving pit.
Middle picture of 36-inch side actuated butterfly valve. Right picture of 39-inch by
36-inch PCCP flange reducer.
© ASCE 5
Pipelines 2015 1242
Figure 5: HDPE connections using Flange Adapters with stainless steel back up ring
and stainless steel bolts.
To minimize the impact of the project to the surrounding community Murphy

Pipelines designed the project layout in which four installations ranging from 1,250
feet to 2,100 feet in length were accomplished. The long pull lengths were beneficial
as they allowed for long fused sections of HDPE to be installed eliminating future
leak potential and aided in the reduction of excavations by 87% of what open trench
would have required.
For each pull, the 50 foot lengths of HDPE were butt fused using a rolling McElroy
1648 machine to correspond to each pull length. After each fusion weld cooled, the
external roll-back bead was removed to allow clearance through the swag die. While
the pipe was fused, crews performed a visual inspection of the interior of the PCCP
waterline. This step is critical as it identifies any major obstructions, location and
degree of bends and condition of host pipe interior which determines if any cleaning
is required. Finally, a proving pig was pulled through. A proving pig is a short section
of HDPE fabricated one to two millimeters larger than the installation OD of the
HDPE during Swagelining™ operations. Its purpose is to eliminate risk by ensuring a

free bore path.
Once a free bore path is confirmed, Swagelining™ operations would begin (Figure
6). To complete each pull, a specific bank shoring plan was implemented to
compensate for the amount of force required to pull the long lengths of HDPE with a
wall thickness of 2.32-inches through a single swage die. Two types of constant
tension pulling equipment were used for the project; Hammerhead 173 ton pulling
machine and a TT Technologies 143 ton pulling machine. Both machines performed
well and without incident. As part of the liner system design process, Murphy
Pipelines utilized their software program to ensure installation stresses on the HDPE
met the ASTM standard for the tensile yield design factor.
© ASCE 6
Pipelines 2015 1243
Figure 6: 1000mm (39.4-inch) OD HDPE pipe enters single swage die. The HDPE
pipe is temporarily reduced below the ID of the 39-inch PCCP host pipe to allow for
insertion.
After the HDPE pipe was completely pulled through the host pipe (with pull lengths
of 1,250 to 2,100 feet), the pulling force was removed. This allowed the HDPE to
naturally revert back towards its original diameter until halted by the inside diameter
of the host pipe forming a compressive tight fit (Figure 7). While dependent on
ambient temperatures, the HDPE is typically allowed to relax overnight to regain full
reversion for most thin walled and semi-structural Swagelining applications. Due to
the thicker wall of this fully structural application, the HDPE needed longer to revert
and regain full reversion.
Figure 7: Tight compressive fit of HDPE after reversion.
© ASCE 7
Pipelines 2015 1244
SUCCESSES AND CHALLENGES
A major challenge with the project was the location of working within a tight utility
corridor with limited room for construction activity. This challenge was addressed
early on through extensive communication with all parties involved with an emphasis
in working with local businesses and homeowners to understand and meet their
demands. The success of this project ultimately required an extensive amount of team
work and coordination. GCWA, League City, ARKK Engineers and a number of
other local agencies showed great resolve in working with Murphy Pipeline crews to
properly plan, adapt and execute the project.
The other major challenge was installing a fully structural HDPE PE4710 DR 17 pipe
with a 2.32-inch wall thickness. The thick walled HDPE pipe required more tonnages
during installation than most thin walled or semi-structural Swagelining™
applications. This placed a higher importance on bank shoring, HDPE butt fusion
operations and pulling equipment.
CONCLUSION
The Calder Road Project represents the largest diameter, fully structural pipe installed
to date in North America utilizing the Swagelining™ technology. The utilization of
this technology with HDPE pipe allowed the owner to meet all design parameters and
increase flow and capacity. The larger final diameter with Swagelining™ vs. slip
lining had significant benefits for the project economics. Gulf Coast Water Authority
will be able to realize a higher value by delivering more water to its customers, both
now and in the future.
As communities across North America face the challenges of aging medium and large
diameter water transmission and sewer force mains, Swagelining™ has been proven
as a technology that can add remarkable value for renewal and replacement. The
method’s advanced engineering agenda through research and development coupled
with its ability to meet various internal pressure requirements from thin walled to
fully structural, including designing for external loading make Swagelining™ a vital
method to be considered.
© ASCE 8
Pipelines 2015 1245
City of Baltimore SW Diversion 78-in. Diameter PCCP:

2,140 LF Continuous Carbon Fiber Pipe Rehabilitation
M. Gabbitas1; K. Eysaman2; A. B. Pridmore3; J. Kiladis4; and J. Hall5

1
Project Manager, City of Baltimore, DPW, Office of Engineering & Construction,
200 N. Holliday St., Room 305, Baltimore, MD 21202. E-mail:
Michelle.Gabbitas@baltimorecity.gov
2
Sr. Manager, Municipal Engineering, RK&K, 81 Mosher St., Baltimore, MD 21217.
E-mail: keysaman@rkk.com
3
Ph.D., M.ASCE, Vice President-Pipeline Solutions, Structural Technologies, LLC,
10150 Old Columbia Rd., Columbia, MD 21046. E-mail:
apridmore@structuraltec.com
4
P.E., Project Manager, RK&K, 81 Mosher St., Baltimore, MD 21217. E-mail:
jkiladis@rkk.com
5
P.E., Project Engineer, RK&K, 81 Mosher St., Baltimore, MD 21217. E-mail:
jhall@rkk.com
Abstract
In 2009 a CCTV/sonar inspection revealed that many areas of Baltimore’s Southwest

Diversion Sewer (SWDS) were in need of repair. Following this inspection, the City
of Baltimore divided the Southwest Diversion sewer repairs into phases for the
purpose of retaining consulting engineers and prioritizing repairs. For the Phase III
repairs, the City and their consulting engineer, Rummel, Klepper & Kahl (RK&K),
worked for several years on project planning and analysis to determine the most
effective options to repair the RCP gravity and PCCP pressure segments of this major
sanitary sewer conveyance system. A Cured-In-Place–Pipe (CIPP) repair system was
selected for the gravity portions and Carbon Fiber Reinforced Polymer (CFRP) for
the repair of the pressure portions because each system could provide the necessary
repairs with minimal loss in the system’s hydraulic capacity. Carbon fiber systems
have been typically used by water and wastewater pipeline owners across the United
States to provide full structural repairs to distressed and damaged pipeline segments.
These specialized materials are designed to meet a variety of installation conditions,
including stand-alone repairs where all internal pressures and external loads are
assumed by the CFRP without relying on the existing structure for strength. The
CFRP scope of work for the SWDS Rehabilitation Project represents the largest
continuous installation of carbon fiber completed to date for an internal repair of a
large diameter pipeline in the United States. The project required extensive planning,
design and an expert installation crew trained and certified for installation of carbon
fiber materials. In addition, a comprehensive inspection and QA/QC process was
© ASCE
Pipelines 2015 1246
used throughout the installation process to insure the long-term success of the repair.
This paper will address the successes and challenges of the SWDS Rehabilitation
Project throughout all stages of the project, including options analysis, design,
bidding, best installation practices, collaborative design approaches, coordinating
large-scale trenchless repair system installations and inspection.
PROJECT BACKGROUND AND OPTIONS ANALYSIS
The City of Baltimore DPW processes wastewater for approximately 1.6 million
residents of the metropolitan area at two (2) wastewater treatment plants (WWTP).
The SWDS transports 25 percent of the wastewater generated within the Baltimore
metropolitan area, including portions of Baltimore County, Howard County, and
Anne Arundel County, to the Patapsco Wastewater Treatment Plant for processing.
The City collects and treats upwards of 250 million gallons of wastewater daily and
maintains over 1,400 miles of sanitary sewer mains. As part of the system, the
SWDS is approximately 7.9 miles in length and is critical to the sanitary collection
operations.
The SWDS is comprised of Class III and IV reinforced concrete pipe (RCP) in the
gravity sections and Class IV pre-stressed concrete cylinder pipeline (PCCP) in the
pressurized sections that range in size from 78- to 102-inches in diameter. In 2009,
deficiencies were identified in both the gravity and pressure portions of the system.
To address these deficiencies, the Southwest Diversion Phase III Rehabilitation
project was implemented and included installation of 2,550 LF of CIPP to rehabilitate
the 78-inch RCP gravity sections and approximately 2,140 LF of CFRP installation to
rehabilitate the 78-inch PCCP pressure portions of the conveyance system. To
complete the rehabilitation, sanitary flow in the 78-inch diameter pipeline had to be
by-passed. The by-pass system was sized to handle a peak 2-year, 24-hour flow of
152 MGD. Altogether, the project involved temporary by-pass and rehabilitation of
approximately 4,690 LF of pipeline.
Prior to selecting the CIPP and CFRP lining systems, the City and RK&K worked for
several years on project planning and options analysis to determine the most effective
options to repair the City’s major conveyance system. Rehabilitation options were
required to address the structural deficiencies in the RCP and PCCP sewers without
reducing the hydraulic capacity of the system. The primary rehabilitation methods
considered included CIPP lining, CFRP lining, slip-lining and replacement. CIPP
and CFRP were selected for use on this project because they minimized the diameter
reduction (primary reason slip-lining was not selected) and also minimized the
disruption to the surrounding residential communities and intersecting streets
(primary reason replacement was not selected). For the gravity portions of the
pipeline, CIPP was the selected rehabilitation approach, whereas CFRP was deemed
the most feasible solution for the pressurized region of the pipeline.
© ASCE
Pipelines 2015 1247
SPECIFICATION DEVELOPMENT AND BIDDING PROCESS
Because of the proximity of repairs and by-pass requirements, the CIPP lining, by-
pass system installation, and the CFRP lining were bid as part of the same contract.
The City of Baltimore and RK&K worked closely together to prepare detailed
specifications for each scope of work. In order to establish minimum competency for
potential bidders, the City of Baltimore required that potential bidders or their
subcontractors be prequalified by the City in the following prequalification
categories: sanitary sewer by-pass installation, carbon fiber lining installation, and
CIPP installation.
In addition to the use of prequalification categories, the CFRP lining specification

included additional experience requirements for the CFRP lining installer, CFRP
design engineer, and the CFRP material manufacturer. To ensure appropriate
material pedigree, extensive durability data and material testing as well as an
International Code Council (ICC) Engineering Service Report number were required
for the epoxy and carbon fiber materials. Because there can be significant variability
in performance between different carbon fiber and epoxy materials, the City utilized
performance based specification requirements for the design section of the CFRP
lining specification. Minimum material properties were listed in the specification as
well as the structural demands and minimum safety factors that were required to be
accounted for in the design. To ensure design criteria were met, stamped drawings
and calculations, developed by the CFRP designer, were required with the bid
submission. The drawings and calculations were verified by the City and their
consulting engineer, to confirm compliance with the specification requirements.
The SWDS rehabilitation project was competitively bid in 2013 and the lowest
responsive bidder was selected. Because of complexities associated with installation
and maintenance of a by-pass system through downtown Baltimore City, the costs
associated with the by-pass system governed the overall project cost. The successful
bid team, Spiniello and STRUCTURAL, coordinated closely to accelerate the
construction schedule, minimizing by-pass and overall project costs.
BYPASS SYSTEM
One of the most challenging aspects of the project was that wastewater flows would
need to be diverted for the duration of the repairs using a by-pass/flow control
system. With peak flows around 152 MGD, the bypass pumping and conveyance
system was extensive and required several carefully planned right-of-entry
agreements for routing the eight (8) 24-inch diameter high-density polyethylene
(HDPE) bypass pipes in order to minimize disruption to the urban areas in which the
project scope was located. The HDPE pipe sections were fused on site to create a
continuous and leak-free by-pass system for the project as shown in Figure 1. Flows
were intercepted upstream of the repairs, pumped through a series of fused HDPE by-
pass pipes, and returned to the SWDS downstream of the repairs.
.
© ASCE
Pipelines 2015 1248
Figure 1. By-pass system installed for the SWDS Project
The by-pass pipes were configured to maintain site, vehicular and pedestrian access
through the project by-pass corridor to minimize disruption to the area. This required
right-of-entry agreements with several private commercial property owners. Because
of the high cost of bypassing the flows, completing the project within the scheduled
pipeline shutdown was critical to the success of the project.
PROJECT COORDINATION EFFORTS
Prior to arriving onsite, extensive collaboration took place between all parties to
ensure that the technical and operational details were best tailored to the project
needs. Open communication in the form of numerous conference calls as well as
several in person meetings was critical to making this process move smoothly and
effectively.
Once the project specific technical package was finalized, the CFRP materials
manufacturer, STRUCTURAL TECHNOLOGIES, hosted a half-day training session for the

City of Baltimore and their engineer. Inspectors, as well as project engineering and
project management team members received training in the project specific QA/QC
program that was to be implemented on the project. This training session allowed all
team members to get technical questions answered as well as make sure all parties
were aligned in their expectations regarding the specific logistics of what QA/QC
information was to be documented and by whom. In addition, confined space
training as well as an overview of the project specific safety program to be
implemented on site was covered during this training session. The safety training
portion of the program was such a success that City of Baltimore conveyed an interest
in incorporating some of the project’s best practices into the City’s overall safety
program.
© ASCE
Pipelines 2015 1249
As a part of the design and construction coordination, it was determined that the CIPP
lining repair would be installed before the CFRP system. To provide redundancy in
the isolation of the pipeline, a “turnback” from the CIPP liner was left in place after
installation and helped to serve as an additional form of isolation redundancy in the
pipeline requiring entry for the installation of the CFRP system. This allowed for
maximum protection for manned entry to the pipe during the CFRP repairs. Once the
CIPP system was completed, a pressurized bladder was also implemented at the end
of the CIPP repair to provide additional isolation to the pressure sewer region of
SWDS, which was to be lined with CFRP.
Implementation of the CIPP and CFRP lining repair systems required extensive
planning to coordinate traffic control, dewatering efforts, bypass system operations,
and the sequence through which the CIPP and CFRP systems would be installed.
Based on the location of the intermediate access manhole within a 4-lane road, work
was able to be isolated so only one lane of traffic in one direction was shut down,
allowing for traffic to be kept open in the other lane throughout the duration of the
project. The City and the construction team worked together to provide the
appropriate traffic control in the affected area. The intermediate access manhole
helped to facilitate improved construction schedules for the CFRP scope of work.
Due to the close proximity of residents, noise levels for equipment as well as working
hours were closely coordinated for this project.
The number of access points and their locations, with respect to the required repairs,
can have a significant impact on CFRP installations. Three (3) manholes were
provided for the 2,140 LF repair section, one at each end of the repair and one
intermediary manhole. These access locations allowed for personnel to minimize the
distances traveled for transporting rolls of saturated CFRP to the work location. The
intermediary manhole helped to facilitate improved construction schedules for the
CFRP scope of work.
Because the CFRP lining process was one of the last steps in work associated with
the project, significant focus was placed on construction timing and ensuring that the
work stayed on schedule. Detailed project status emails were distributed on a weekly
basis, at a minimum, to all key team members to facilitate smooth communication
and make sure that any items requiring action could be addressed in a timely manner.
The close coordination between the entire project team allowed for any necessary
actions to be addressed quickly and contributed to the CFRP lining work finishing
two (2) weeks ahead of schedule.
QUALITY CONTROL PROGRAM FOR THE CFRP LINING SYSTEM
Overall quality of the CFRP lining system is determined by conservatism of design,

durability of materials installed, experience of workers and supervisors involved in
installation, and a thorough QA/QC program. The installation process includes
surface preparation, mixing of epoxy, saturation of CFRP layers, and installation of
© ASCE
Pipelines 2015 1250
CFRP and curing. Inspection throughout these processes – including at completion –

helped insure a quality CFRP installation.
QA/QC forms were created for each of the 114 segments of PCCP repaired. In order
to maintain locations of each pipe segment throughout the repair process, a measuring
wheel was used to measure the distances between manholes to the pipe segment. The
QA/QC forms provided documentation for the entire CFRP lining installation
process. It included pre-construction condition, date of installation for each stage of
the CFRP process, unique field conditions, environmental conditions, calibration of
the mechanical saturation equipment used to saturate the composite materials,
inspection/verification of each layer of material installed, and lot numbers of
materials used.
One critical verification which takes place is confirmation that the material properties
of the CFRP system applied in the field are in-line with the properties utilized in
design. To do this, 12-inch by 12-inch test panels were made utilizing the carbon
fiber fabric, epoxy, and saturation equipment from the production runs for the field-
installed CFRP lining system. These panels were then cured in the pipe to ensure the
same curing environment. After curing, these panels were collected by the Engineer
and sent to a certified laboratory to be tensile tested in accordance with ASTM D3039
to gather material properties.
Figure 2. Witness panels utilized to confirm material design values.
As a part of the quality assurance program, inspection “hold points” for each
construction region were provided throughout the construction process. The
inspection hold points included:
• Verification of surface preparation via bond testing in accordance with

ASTM D4541 pull tests (Figure 6)
• CFRP post-installation inspection (Figure 3)
• Top coat post-installation inspection (Figure 4)
© ASCE
Pipelines 2015 1251
Figure 3. Inspection of the Installed CFRP Lining, prior to Top Coat Installation
Figure 4. Inspection of top coat
The QA/QC forms for each segment of PCCP repaired provided documentation for
the inspection hold points, installation of the CFRP system, and served as permanent
records of the project.
© ASCE
Pipelines 2015 1252
SURFACE PREPARATION FOR THE CFRP LINING SYSTEM
Prior to the installation of a CFRP structural lining system, the following installation
steps must take place: dewatering of the pipeline, set-up and maintenance of
ventilation and environmental controls, initial cleaning of the pipe substrate, surface
preparation, final cleaning of the pipe substrate, and verification of surface
preparation. The by-pass system was put in-place, redundant safety controls were
installed to allow for safe worker entry into the pipe, and an initial pressure wash of
the pipe with a bleach solution was performed to sanitize the walls of the pipe and
allow for surface preparation to take place.
The project requirements for surface preparation were as follows: a minimum surface
profile of ICRI CSP 3 for the inner core substrate and near white metal blast (SP10)
for the exposed steel pipe cylinder at the end terminations of the CFRP lining. The
typical surface preparation methods used for preparation of the inner core substrate
are sponge blasting, sand blasting and hydro blasting. Because either sponge blasting
or sand blasting are necessary for the preparation of the steel substrates in the joint
regions of the pipe, one of these techniques are more commonly used for surface
preparation on smaller projects so that the same preparation method can be used for
the inner core substrate and the joint regions. However, hydro-blasting using
approximately 30,000psi water pressure becomes a more efficient inner core substrate
surface preparation method for longer production runs and was utilized for this
project because it was able to be run continuously through the length of the pipeline
repair, without having to stop and clean a blast pot. Post cleanup for hydro-blasting
was also simpler; the water ran downstream and was pumped out of the pipe.
Figure 5. Deterioration observed at the invert of the pipe
© ASCE
Pipelines 2015 1253
After surface preparation was completed, the entire surface of pipeline was inspected
by the City’s consulting engineer, the CFRP installer, and the CFRP system
manufacturer to document any anomalies and address any unanticipated conditions.
One region of the pipe, shown in Figure 5, had experienced such severe erosion
damage at the invert that the entire inner core, steel cylinder and most of the outer
core was missing for a region approximately 16 feet along the length of the pipe.
While the CFRP lining system was designed as a stand-alone structural system to
take internal pressure and all external loads without reliance on the host pipe for
structural integrity, this severe level of distress required special detailing to make sure
that water intrusion was stopped during lining installation to allow the CFRP
materials to cure properly.
As part of the QA/QC program for the project the surface preparation and adhesion of
the CFRP to the inner core substrate was verified using ASTM D4541 bond testing.
This is a critical step because the installation of the CFRP liner system for pipeline
rehabilitation is considered a bond critical application. Since the failure mode for the
adhesion test is often a tensile failure within the inner core substrate, this test
confirms both satisfactory adhesion of the CFRP to the inner core as well as the
approximate tensile strength of the existing inner core concrete. CFRP panels were
installed on prepared concrete substrate at approximately 60 LF intervals along the
entire repair section. The adhesion tests were performed by the Engineer using
20mm test dollies installed on the adhesion test mock up areas, as shown in Figure 6,
prior to installation of the first layer of the CFRP lining system as the project
progressed down the length of the pipeline.
Figure 6. Pull tests performed to verify adhesion.
© ASCE
Pipelines 2015 1254
INSTALLATION OF THE CFRP LINING SYSTEM
After completion of surface preparation, fabric layers are mechanically saturated with
a two (2) part, 100% solids epoxy as shown in Figure 7. The mechanical saturator
was calibrated several times per shift to ensure consistency in the epoxy to fabric
weight ratio. The application of the CFRP lining system then involved applying a
layer of thickened epoxy onto the prepared pipe substrate and applying epoxy
saturated sheets of glass fiber and carbon fiber composites to the inside of the
pipeline in the orientation designated per the project drawings. Once cured, the
CFRP lining system is designed to take all of the internal pressure and external loads
acting on the original pipeline without reliance on the host pipe. The design relies on
the carbon fiber fabrics for structural integrity and the glass fiber fabrics to serve as
an electrical isolation layer in any area where steel surfaces are exposed.
Figure 7. Mechanical saturation of the carbon fiber fabric.
Quality checks were performed after each layer of fiber was installed to confirm that
proper development length was achieved, verify fabric alignment and spacing of
layers was in accordance with project requirements, and check for any air bubbles.
Any items requiring remediation were documented and addressed prior to continuing
with the subsequent layer of CFRP.
© ASCE
Pipelines 2015 1255
CFRP laminates properties are dependent on the bond between each lamina. The
type of bond required is based on the time lapse between which the layers are
installed. If a new layer is placed upon a previously installed layer within the epoxy’s
cure cycle, a chemical bond is established; otherwise, a mechanical bond is required.
To establish a chemical bond with the epoxy used on this project, new layers had to
be installed within 72 hours of application. Due to the size of the repair, it was
necessary to divide the CFRP installation into three (3) distinct construction regions,
where each region was treated from an operational standpoint as a separate CFRP
installation. The use of separate installation regions, along with careful planning of
work sequences, allowed for subsequent layers of CFRP to be installed within the
necessary 72 hour window to develop a chemical bond between the layers of CFRP.
Figure 8 shows the typical application process for the CFRP layers.
Figure 8. Application of hoop direction layer of CFRP
Because the SWDS pipeline transports raw sewage, a chemical-resistant top coat
specifically designed for sewage applications was installed over the CFRP structural
lining system. Quality inspections were also performed by the Engineer on the top
coat layer as shown in Figure 9 to check for any defects that required the application
of additional top coats to insure conformity with the project requirements.
© ASCE
Pipelines 2015 1256
Figure 9. Inspection of the CFRP lining top coat
CONCLUSION
While the CFRP lining process has been used for structural upgrade of large diameter
pipelines for over 15 years, this project is unique for multiple reasons. At 2,140 LF,
the project’s CFRP installation is the longest continuous CFRP installation performed
to date in the United States. In addition, the complex by-pass system and other
coordination efforts required to rehabilitate the sewer pipeline located in a heavily
populated area presented additional unique project challenges. Due to the size of the
project and additional project complexities, close coordination was necessary among
the City, the engineer, and contractors for the CIPP and CFRP systems. The high
level of communication was necessary to determine site logistics, access
requirements, installation sequences and inspection of the CFRP work. This project
demonstrated that CFRP is advantageous for structural upgrade of extended runs of
large diameter sanitary sewer pipelines.
REFERENCES
ASTM D3039, Standard test method for tensile properties of polymer matrix
composite materials. American Standard for Testing and Materials (ASTM)
ASTM D4541, Standard test method for pull-off strength of coatings using portable
adhesion: American Standard for Testing and Materials (ASTM)
ICRI Guideline No. 310.2, Selecting and specifying concrete surface preparation for
sealers, coatings and polymer overlays. International Concrete Repair Institute (ICRI)
SSPC-SP No.10 / NACE 2 Near-White Blast Cleaning. Society for Protective

Coatings (SSPC) and the National Association of Corrosion Engineers International
(NACE)
© ASCE
Pipelines 2015 1257
Miami-Dade Implements Hybrid FRP Trenchless Repair System
Luis Aguiar1; Anna Pridmore2; and Mark Geraghty3
1
Assistant Director, Water Operations, Miami-Dade Water & Sewer Department,
3071 SW 38th Ave., Miami, FL 33146.
E-mail: LAgui@miamidade.gov
2
Ph.D., M.ASCE, Vice President, Pipeline Solutions, Structural Technologies, LLC,
10150 Old Columbia Rd., Columbia, MD 21046.
E-mail: apridmore@structuraltec.com
3
Vice President, Strengthening Sales & Marketing, Structural Technologies, LLC,
10150 Old Columbia Rd., Columbia, MD 21046.
E-mail: mgeraghty@structuraltec.com
Abstract
In 2010, Miami-Dade Water & Sewer Department (MDWASD) implemented a

comprehensive asset management program to increase the reliability of their large
diameter pipeline system. The program includes inspection, prioritization and
targeted replacement or rehabilitation of pipeline segments as needed. MDWASD
utilizes several methods to address damaged or distressed pipelines including targeted
fiber-reinforced polymer (FRP) structural repairs. Several critical Miami-Dade
pipelines have had FRP systems installed at numerous segments as fully structural
repairs. FRP repairs are trenchless and involve the use of carbon fiber and glass
fiber-reinforcing fabric saturated in an epoxy matrix, then installed on the interior of
the pipeline. Once cured, the FRP provides a standalone structural upgrade of the
pipe, extending the lifecycle equivalent to replacement. Traditional FRP systems are
typically used for segmental repairs when single or multiple non-contiguous pipe
sections require rehabilitation. To address extended runs of pipe in a more cost-
effective manner, a hybrid system has been designed utilizing continuously wound
high strength steel reinforcement embedded in epoxy along with layers of FRP to
form a structural upgrade system inside the existing pipe. In the summer of 2014,
following an inspection, 13 pipe segments in need of structural upgrade were

identified across a 3.5 mile stretch of Red Road. The 54-inch prestressed concrete
cylinder pipe (PCCP) segments were selected as the most at-risk through
MDWASD’s risk ranking and prioritization process. The repair method elected for
the 13 segments was installation of FRP as a stand-alone structural upgrade. Also, as
part of the project, Miami-Dade set aside 3 of the 13 segments for installation of the
Hybrid FRP System, also implemented as a stand-alone structural upgrade. The
Hybrid FRP System installation process includes surface preparation, application of
FRP and installation of the steel tensile reinforcement. The steel is installed using
specialized equipment which places the reinforcement in the hoop direction onto the
pipe’s interior surface. The general design approach utilizes the steel reinforcement
© ASCE
Pipelines 2015 1258
for resisting hoop direction design requirements and the FRP component of the
system for resisting longitudinal design requirements. The steel is typically placed
between layers of FRP with the steel reinforcement set into an epoxy putty for
system continuity and protection. The Miami-Dade Hybrid FRP installation was
implemented successfully within the allotted schedule. Readers will learn detailed
information about MDWASD’s experience selecting and implementing the Hybrid
FRP repair at three (3) segments of 54-inch PCCP. The benefits and limitations of the
system will be explored including design, materials selection, installation and the
appropriate quality control measures.
BACKGROUND
Miami-Dade Water & Sewer Department (MDWASD) operates one of the largest
public utilities in the United States. MDWASD services approximately 2.3 million
people, the state’s highest population. The customer base consists of outlying areas
of Miami-Dade County, 485,000 service connections and 15 municipal wholesale
customers.
MDWASD’s service area consists of over 7,900 miles of water mains ranging in size
from 2 inches to 120 inches in diameter spread across 400 square miles. MDWASD
uses predominantly Prestressed Concrete Cylinder Pipe (PCCP) for large diameter
transmission mains and has over 100 miles of PCCP that is 48-inches and larger.
Much of the PCCP is located under major roadways in densely populated areas.
After a series of high profile catastrophic failures in 2010 and 2011, similar to the
one shown in Figure 1, MDWASD began to focus on pipeline management and
developed a comprehensive asset management program. The objective was to
implement a program to minimize disruptions to residents and other customers.
MDWASD had to develop a program that balanced upgrading and replacing
pipelines to the region’s increasing base of customers.
© ASCE
Pipelines 2015 1259
Figure 1 – Ruptured 54-inch PCCP water transmission main
MIAMI-DADE’S INRASTRUCTURE ASSESSMENT AND

REHABILITATION PROGRAM
To address these needs, MDWASD established the Infrastructure Assessment and

Rehabilitation Program (IAARP). The program consists of routine inspections of the
pipeline inventory on a rotating basis. As part of IAARP, MDWASD adopted
industry best practices which include precision inspection, replacement and
structural upgrade for its large diameter PCCP inventory.
Through specialized electromagnetic inspection services provided by Pure

Technologies, MDWASD is able to evaluate their entire large diameter PCCP
system to identify distressed and high-risk pipe segments. The inspection technique
utilized is able to identify broken prestressing wires throughout each segment. When
a large enough number of prestressing wires break on a given segment of PCCP
structural integrity of the segment is compromised. Utilizing inspection methods
that pinpoint which pipe segments have broken wires allows MDWASD pipeline
systems to be repaired or replaced before failures cause unscheduled and costly
shutdowns.
After an inspection of a pipeline run is completed, MWASD works with their

consulting engineers to analyze the data and develop a failure risk analysis. This
process includes taking into account all of the factors for each specific pipe including
prestressing wire pitch and spacing, cylinder thickness, concrete core thickness,
along with internal and external loads acting on the pipe. From this analysis the pipes
inspected and identified as distressed can be ranked into groups for near term
replacement or repair, mid-term and long term which require monitoring. Following
this, the decision making process for how to address near term repairs is completed.
Many of MDWASD’s pipelines are located underneath major roadways which make
trenchless rehabilitation techniques advantageous over dig and replace or other repair
methods requiring excavation. In cases where rehabilitation is the correct technical
© ASCE
Pipelines 2015 1260
solution, MDWASD addresses the identified high risk pipeline segments with
targeted structural repairs installed without excavation.
For the past 5 years MDWASD has utilized high-strength fiber-reinforced polymer
(FRP) as a method to provide structural upgrade to specific pipe segments identified
during inspections and failure risk analysis. Many projects have been completed
using FRP for PCCP lines ranging from 48 to 96-inches, and a typical installation is
shown below in Figure 2. MDWASD has established an emergency response team
(ERT) of prequalified specialty contractors who can perform this work through an
on-call process.
Figure 2 – Typical FRP installation
Following an inspection and failure risk analysis in mid-2014, MDWASD moved

forward with repair of 13 segments of 54-inch PCCP. Ten (10) of the segments were
repaired using FRP and MDWASD elected to implement the StrongPIPE hybrid
FRP system for three (3) of the 54-inch segments.
HYBRID FRP SYSTEM OVERVIEW

Figure 3, below, demonstrates the composition of the Hybrid FRP system used by
MDWASD on their recent project. The system consists of high strength steel
reinforcement continuously wound around the inside circumference of the pipe. The
steel reinforcement is embedded in an epoxy matrix and sandwiched between layers
of glass or carbon FRP reinforcement that provide the required strength in the
longitudinal direction.
The “sandwiched” composite structure is engineered to meet all design requirements

acting on the system in both the circumferential and longitudinal directions. The
continuous circumferentially installed steel reinforcement is intended to provide the
main reinforcement in the hoop direction, while the FRP materials are oriented
© ASCE
Pipelines 2015 1261
primarily in the longitudinal direction to resist the longitudinal loads acting on the
system.
Figure 3 – Basic composition for Hybrid FRP system
The purpose of the top and bottom longitudinal layers of glass and/or carbon FRP
reinforcement is twofold: (1) to provide resistance of the Hybrid FRP System
against longitudinal stresses resulting from internal pressure thrust, Poisson’s effect
from internal pressure, and temperature variations; and (2) to provide additional
environmental protection for the high strength steel so that the steel is fully
protected from contact with water. Glass and carbon FRP reinforcement have been
widely used in the strengthening of concrete structures since the early 1980’s. Its
use extended to PCCP rehabilitation in the United States in the mid-1990s, and it
has been steadily increasing in the last 15 years.
The Hybrid FRP System has been designed to provide an economical solution for
short as well as long runs of repair, which have been typically considered to be cost
prohibitive for FRP applications. The system is very appealing for the rehabilitation
of PCCP due to its lower cost compared to other available methods, and because it
involves the use of high strength steel, which is a widely accepted material for
structural upgrade applications.
HYBRID FRP SYSTEM DESIGN FOR MDWASD 54-INCH PCCP
Design of both the FRP System and the Hybrid FRP System were tailored to
operating conditions and took into account several factors including internal pressure,
transient pressure, vacuum pressure and all external loadings. Miami-Dade typically
has all structural upgrades on their large diameter PCCP designed for 150psi
operating pressure with operating plus transient pressures up to 225psi. In addition,
designs incorporate full vacuum as well as soil cover along with ground water up to
the top of the soil cover. For vehicular loads, MDWASD accounts for HS-20
© ASCE
Pipelines 2015 1262
vehicular loading for all pipe segments, since most of their pipelines run under
roadways throughout the city.
The Hybrid FRP System for pipe nos. 456, 457 and 459, 20LF segments of PCCP
was designed as fully structural, with no reliance on the host pipe. It consisted of
multiple layers of Carbon FRP (0.08-inch thickness), Glass FRP (0.04-inch) and 260
ksi steel reinforcement (0.208 inch diameter, 30 steel reinforcements per foot)
oriented to meet design requirements.
INSTALLATION OF THE HYBRID FRP SYSTEM FOR MDWASD 54-INCH

PCCP
The Hybrid FRP System installation for MDWASD 54-inch PCCP took place using
trenchless methods which included manned entry into the pipeline structure. Prior to
entry, proper confined space procedures were implemented to meet OSHA
requirements regarding work taking place within a confined space and appropriate
ventilation was installed. Once jobsite safety was addressed, the system installation
commenced.
The project set-up for the Hybrid FRP System required a minimal topside footprint
which included a truck mounted unit for the steel reinforcement installation. The
truck unit is shown in Figure 4 below.
Figure 4 – Truck mounted unit for Hybrid FRP System steel reinforcement installation
The first step of pipeline operations completed for the Hybrid FRP System
installation at the 54-inch PCCP for MDWASD was surface preparation. The inner
core concrete was abrasively blasted to roughen the surface to a minimum concrete
surface profile (CSP) of 3 as defined by ICRI 310.2 guideline. At the termination of
the Hybrid FRP system, the joints of the PCCP were chipped out to expose the bell
and spigot steel and the steel substrate in joint areas was prepared to near-white metal
condition as defined by SSPC-SP10/NACE No.2. Figure 5 below shows the prepared
concrete substrate.
© ASCE
Pipelines 2015 1263
Figure 5 – Prepared concrete substrate – MDWASD 54-inch PCCP
Following surface preparation the substrate was tested for proper bond using the
ASTM D4541 procedure and upon confirmation installation of the longitudinal layer
of Carbon FRP was installed. The FRP materials were prepared topside using a
mechanical saturation machine and installed over each repair segment in the
longitudinal direction. After the Carbon FRP layer was installed, per the design
requirements a layer of Glass FRP was installed.
The next step in the Hybrid FRP System was the installation of the spirally wound
steel reinforcement. The truck mounted unit, shown above in Figure 4, fed the steel
reinforcement in a spiral fashion into the pipe where it was placed using automated
equipment specially developed for the in-pipe installation. The steel reinforcement
installation equipment, shown in Figure 6 below, was dismantled and reassembled in
the pipe, allowing for the equipment to fit through the 16-inch by 18-inch manholes
typically available for entry into MDWASD’s pipelines.
Figure 6 – Hybrid FRP System – Steel reinforcement installation
© ASCE
Pipelines 2015 1264
Following placement, the steel reinforcement was then covered with a specialized
thickened epoxy with a putty-like consistency. As per the approved design, a layer of
Glass FRP was applied following the steel reinforcement.
Termination details were used at the ends of the system to prevent water migrating
through any cracks within the inner core concrete at the termination of the repair. To
achieve this, the mortar in the joint as well as a small portion of the inner concrete
core at the end of the pipe was carefully removed to expose the steel cylinder and to
create a transition region using a wedge built up with epoxy mortar. The system was
transitioned at the termination and bonded directly onto the bell or spigot steel and
steel cylinder. Stainless steel expansion rings were then installed in the joint
terminations to guarantee full intimate contact. Once the rings were installed, the
joint region was filled in flush with epoxy mortar.
QUALITY CONTROL PROGRAM FOR THE HYBRID FRP SYSTEM
Quality Control during installation of the Hybrid FRP System was a critical
component to project implementation and required close coordination with
MDWASD and their consulting engineer. The process included QA/QC checks
typical of standard FRP installations plus a group of additional steps for the steel
component of the Hybrid system.
QA/QC for FRP component. During the installation and cure time of FRP,
environmental conditions are monitored and tracked. These include temperature and
humidity. As materials are prepared for installation there are steps for verification
which include confirmation of the mixing procedure for epoxy and documentation of
calibration of rollers within the mechanical saturation equipment, shown in Figure 8.
Figure 8 – Mechanical saturation equipment
© ASCE
Pipelines 2015 1265
The QA/QC process at the point of installation within the pipe includes several steps
starting at surface preparation. As mentioned in previous sections the substrate is
checked for bond using the ASTM D4541 process.
As installation of the Hybrid FRP System takes place, the materials have QA/QC
checks which include verifying proper alignment of the FRP material and confirming
that appropriate overlaps within the FRP layers are achieved per the approved project
drawings. Following installation of the fiber layers, inspections are performed to
check for air bubbles and insure that intimate contact is achieved between the pipe
substrate and the FRP layers.
The FRP is also tested for tensile properties using the ASTM D3039 testing protocol.
Sample panels are created using the same material installed in the pipelines. The
samples are sent to a laboratory and tested to confirm the properties meet or exceed
design values utilized.
QA/QC for Steel Reinforcement component. The second major component of the
Hybrid FRP System is steel reinforcement and there are several QA/QC steps
associated with this portion of the installation. As the steel is delivered to the jobsite,
the lot numbers of the steel are verified. A visual check of the steel takes place as it
is placed onto the pipe surface. In addition, the spacing of the steel reinforcement –
confirmation that the correct numbers of steel reinforcement wires are placed within
each lineal foot of pipe takes place. There is also a check to determine that the
minimum and maximum spacing between each steel wire in relation to adjacent steel
meets project requirements. This is shown in Figure 9.
© ASCE
Pipelines 2015 1266
Figure 9 – Verification of steel wire placement

After the steel is installed, the thickness of epoxy used to encapsulate the steel is
verified to insure proper coverage and protection of the steel wire.
CONCLUSION
The Hybrid FRP System installation was completed successfully and MDWASD is
exploring the use of this system for upcoming projects. MDWASD produces
approximately 350 million gallons of water every day through its pipeline system for
its customers. Through the development and implementation IAARP, MDWASD
effectively manages its inventory of pipelines. As a result, MDWASD ensures
pipeline reliability, manages limited resources, and improves the life of residents.
© ASCE
Pipelines 2015 1267
REFERENCES
ASTM D3039, Standard test method for tensile properties of polymer matrix composite
materials. American Standard for Testing and Materials (ASTM)
ICRI Guideline No. 310.2, Selecting and specifying concrete surface preparation for sealers,
coatings, and polymer overlays. International Concrete Repair Institute (ICRI)
SSPC-SP No.10 / NACE 2 Near-White Blast Cleaning. Society for Protective Coatings
(SSPC) and the National Association of Corrosion Engineers International (NACE)
© ASCE
Pipelines 2015 1268
Composite versus Stand-Alone Design Methodologies

for Carbon Fiber Lining Systems
Michael Gipsov1; Rasko Ojdrovic2; and Anna Pridmore3

1
P.E., M.ASCE, Principal Structural Engineer, Washington Suburban Sanitary
Commission, 14501 Sweitzer Lane, Laurel, MD 20707. E-mail:
Michael.Gipsov@wsscwater.com
2
Ph.D., P.E., M.ASCE, Senior Principal, Simpson Gumpertz & Heger, 41 Seyon St.,
Bldg. 1, Suite 500, Waltham, MA 02453. E-mail: rpojdrovic@sgh.com
3
Ph.D., M.ASCE, Vice President - Pipeline Solutions, Structural Technologies, LLC,
10150 Old Columbia Rd., Columbia, MD 21046. E-mail:
apridmore@structuraltec.com
Abstract
Fiber Reinforced Polymer (FRP) composite lining systems are used by major
municipalities throughout the United States to structurally rehabilitate and upgrade
large diameter pipelines. For internal Carbon Fiber Reinforced Polymer (CFRP)
lining systems addressing prestressed concrete cylinder pipes (PCCP), there are two
design approaches utilized relative to interaction with the host pipe structure. These
approaches are referred to as stand-alone and composite. For a stand-alone design, the
carbon fiber takes 100% of the loads acting on the pipeline system with no reliance
on the host pipe for structural integrity. Composite designs rely on the carbon fiber
lining system and inner concrete core of the PCCP to interactively provide a
structural system to resist the loads. A composite design approach relies on the inner
core to resist bending and buckling due to external loads such as soil cover, water
table, vehicular loads and vacuum pressure. When applicable, this type of design can
be more cost-effective because the amount of carbon fiber materials utilized can be
less than stand-alone design. This paper presents design limit states and includes
information from recent research, development, and testing. It discusses factors to be
considered, potential challenges and best practices for determining stand-alone versus
composite designs for carbon fiber lining systems.
© ASCE
Pipelines 2015 1269
BACKGROUND
Over the past more than 15 years, Fiber Reinforced Polymer (FRP) composite
materials have been utilized with increasing frequency for internal structural
rehabilitation and upgrade of pipelines. The overall process involves surface
preparation of the internal pipe substrate followed by manual application of layers of
unidirectional carbon fiber fabrics (Figure 1) which have been saturated with a two
part epoxy directly prior to installation using a calibrated mechanical saturator.
Figure 1. Typical process for Installation of CFRP inside a pipeline
The layers of carbon fiber fabric are oriented in the longitudinal and the
circumferential directions and are designed to resist the structural demands acting on
the pipeline. Depending on the design approach, the CFRP liner can be designed as a
stand-alone system or a composite system which relies on the host pipe for partial
structural strength.
CFRP liners are commonly used to structurally rehabilitate prestressed concrete

cylinder pipeline (PCCP) segments which have been identified as distressed. An
embedded-cylinder type (ECP-type) PCCP, the type of PCCP used for larger
diameters pipelines, is composed of an inner concrete core, a steel cylinder, an outer
core, prestressing wires over the outer core, and a protective mortar coating (Figure 2).
A common failure mode of PCCP is breakage of the prestressing wires within
individual PCCP sections.
© ASCE
Pipelines 2015 1270
Once enough prestressing wires break, the concrete core in the region near the broken
wires is no longer in compression and can crack, exposing the steel cylinder to
ground water and thus, causing corrosion. The condition of the host pipe is critical in
determining what extent of the host pipe, if any can be taken into account in the
CFRP lining design. Since the steel cylinder, outer core, and prestressing wires are
debonded from the inner core, only the inner core can be relied on in composite
CFRP design for addressing distressed PCCP segments.
Figure 2. Components of an ECP-type Prestressed Concrete Cylinder Pipe Section
DISTRESS LEVEL OF THE HOST PIPE
As part of the CFRP lining design process, the overall distress level within the host
pipe is considered. These levels of degradation are defined in a draft AWWA
standard for CFRP rehabilitation and strengthening of PCCP as Non-Degraded Pipe,

Degraded Pipe, and Severely Degraded Pipe.
a) A non-degraded host pipe is taken into account in the design when there is no
known damage to the PCCP segment and the CFRP liner is added due to load
increases acting on the pipeline (live load, earth load, pressure, etc.). Based
on the good condition of the pipe, the CFRP system can be designed as
composite action with the entire pipe wall thickness.
b) A host pipe is defined as a degraded pipe when the PCCP has some broken
wires and the outer concrete core may be also cracked and softened, but any
minor cracking of the inner core can be repaired and the inner core is still
© ASCE
Pipelines 2015 1271
circular. The host pipe is expected to continue to degrade with time after the
CFRP repair is in place. Since additional wire breakage, outer core cracking,
and corrosion of steel cylinder are anticipated over time, the CFRP repair of
degraded pipe can be based on either composite action of the host pipe inner
core reinforced with CFRP laminate or stand-alone CFRP liner.
c) A severely degraded host pipe consists of PCCP with a non-circular inner

concrete core showing multiple wide cracks as well as an uneven internal
surface with ovality or waviness. Pipes with this level of severe distress
require special design consideration and additional attention should be given
to determining applicability of CFRP lining for these applications.
DESIGN PROCESS FOR FRP REHABILITATION OF PCCP SEGMENT
Design process used for FRP Rehabilitation of PCCP at WSSC consists of several
steps depicted in below the diagram (Figure 3).
Figure 3. FRP Design Process
The design process involves collection of the design information, including as-built
drawings, lay schedules and pipe specifications, results of structural pipe inspection
(visual and sounding, electro-magnetic, sonic/ultrasonic, etc.) and assessment. Since
stand-alone design for the large diameter PCCP most likely will utilize more layers of
the FRP material than the composite design method, the design process may start
with validation of the less expensive composite method which relies on the existing
condition of the substrate, i.e. inner core concrete. Determination of the concrete
condition is one of the most critical components needed for the “composite versus
© ASCE
Pipelines 2015 1272
stand-alone” decision since inner core concrete compressive strength is used for the
estimate of the FRP-to-substrate bond. Should adhesion bond, σ, be less than the
minimum allowed bond value (Rmin) per AWWA [5] such that σ< Rmin, a stand-alone
design approach may be used. In order to confirm the adhesion bond and therefore
determine applicability of a composite design approach, on site pull-off testing must
be performed.
Determine Condition of Inner Core
In order to determine whether the inner core is capable of being used in a composite
CFRP design, a condition assessment is performed to evaluate the level of
deterioration that has taken place. Several methods are used at WSSC for
determining condition of the inner concrete core include visual and sounding,
adhesion testing, sonic/ultrasonic, and rebound hammer testing.
Visual and sounding inspection of a pipe involves a trained inspector looking for
signs of distress within the pipe which include cracks within the inner core, damaged
joints, areas with severe pipe ovality, and concrete spalling. One sign of severe
distress in a PCCP section involves longitudinal cracks within the inner core, which
could indicate loss of prestress due to broken wires.
Tests to estimate the compressive strength of concrete include sonic/ultrasonic

inspection which can be performed as a part of the structural assessment [8], and the
rebound hammer test (i.e. Schmidt hammer test) per ASTM C805 [4] which involves
a spring loaded hammer hitting a steel plunger, which is in contact with the concrete
as shown in Figure 4. Once the concrete is impacted by the defined energy, the
hammer’s rebound distance is measured. This rebound hammer can be used to
determine the concrete’s compressive strength using the manufacturer’s conversion
chart [9].
Figure 4. Evaluation of Inner Concrete Core via Rebound Hammer Test
© ASCE
Pipelines 2015 1273
In order to validate design based of the estimated inner core concrete values, adhesion
tests must be performed in accordance with ASTM D4541 [3]. Adhesion tests are a
part of the typical QA/QC process for the CFRP lining process. A common failure
mode observed in the adhesion tests is tensile failure within the inner concrete core
substrate so the results from adhesion testing provide a measure of the tensile strength
of the concrete core (Figure 5). Since the tensile strength of concrete is
approximately 10% of concrete’s compressive strength, the compressive strength of
the concrete can be approximated through use of adhesion tests on the inner core
substrate. The calculated compressive strength for the inner core concrete can be
checked against the minimum required values used in the design (σ< Rmin).
Figure 5. Pull Test per ASTM D4541
DESIGN APPROACH
CFRP systems are designed using a Load and Resistance Factor Design (LFRD)
approach (AWWA draft standard), where factors are applied to applied loads and
material properties to account for uncertainties within the design assumptions.
As part of this design approach, design limit states are analyzed separately and the
CFRP lining design is governed by the limit state that has the lowest demand to
capacity ratio for the particular design scenario. Various limit states are accounted
for in the design depending on whether a composite or stand-alone system is being
considered.
Stand-Alone Design
For stand-alone design, the following limit states must be considered:
© ASCE
Pipelines 2015 1274
o Rupture of CFRP laminate in the circumferential direction due to internal

pressure.
o Rupture of CFRP laminate in the circumferential direction due to bending

of empty pipe.
o Rupture of CFRP laminate in the circumferential direction due to

combined pressure and bending due to gravity loads.
o Buckling of CFRP laminate in the circumferential direction due to

external loads and pressures and internal negative pressure
o Rupture of CFRP laminate in the longitudinal direction due to pressure

induced thrust, Poisson’s effect of internal pressure, and temperature
changes in the pipe.
o Shear bond failure of the CFRP at pipe ends.
o Rupture of CFRP laminate in the longitudinal direction due to radial

expansion of pipe in broken wire zones.
o Compressive failure of CFRP laminate in the longitudinal direction due to

radial expansion of pipe in broken wire zones.
o Buckling of CFRP liner in the longitudinal direction due to temperature

increase.
Composite Design
Composite design can be applied in situations where the host pipe is classified as
non-degraded or degraded.
When a PCCP section is considered degraded and only the inner concrete core is
taken into account in the CFRP design, the following additional limit states are
addressed:
o Debonding of CFRP from the concrete inner core under one of the
following circumstances:
• Shear between the CFRP and the concrete inner core.
• Excessive radial tension.
• Concrete core crushing from gravity loads, in absence of internal

pressure.
© ASCE
Pipelines 2015 1275
In situations where the host pipe is considered non-degraded and the CFRP lining is
utilized to upgrade or strengthen the existing pipe, the entire wall thickness may be
considered in composite action.
In the design process, it is initially assumed that the CFRP lining is acting
compositely with the concrete inner core. Since stand-alone designs typically require
higher layer counts than composite designs, in order to not unreasonably increase an
amount of CFRP layers and consequently the cost of the repair, the design may start
as composite. The bond between the CFRP liner and the inner core is checked and if
any of the limit states are not satisfied, then the system must be designed as a stand-
alone.
Recent Testing Affecting CFRP Lining Designs
Over the past several years, significant research and development efforts have taken
place impacting best practices regarding designs of CFRP linings. One of the major
testing programs was completed in conjunction with the Water Research Foundation
(Zarghamee et al.) [10]. The testing included full scale external load tests and internal
pressure tests.
External load testing, such as that recently completed (as shown in Figure 6), assists
in better understanding of the CFRP and inner core composite action mechanism and
ultimately helped validating the design approach which relies on the inner core for
composite CFRP design.
Figure 6. Water Research Foundation Testing Setup (Zarghamee et al, 2013)
One of the most significant findings in recent testing is that watertightness of the
CFRP lining is critical to long term performance, whether in a stand-alone or
composite design approach. The termination details must be effective in preventing
pressure build-up behind the CFRP liner. It was determined that preparation of the
© ASCE
Pipelines 2015 1276
steel substrate at the pipe ends (for PCCP) is to be completed in a manner which
ensures that material bonding is not compromised.
Along with its importance at the terminations, watertightness of the entire CFRP liner
is a recent point of focus with regard to permeability. Best practice for CFRP liner
materials now includes validation of watertightness for different laminate designs
through testing and inclusion of watertightness provisions within each CFRP design.
CONCLUSION
The composite design process is considered a typical design concept for CFRP lining
of PCCP. In order to establish feasibility of a composite design, the pipe must be
verified through inspection to determine the condition of the inner core substrate
within the host pipe. When composite designs are feasible, they have the potential to
help reduce the overall layer count for the CFRP lining system, thereby helping
pipeline owners further extend rehabilitation dollars. When composite designs are
not feasible, the CFRP lining system can be designed as a stand-alone system to take
all loads without reliance on the host pipe for structural integrity.
REFERENCES
1. ANSI/AWWA C304-07 (2007). Standard for Design of Prestressed

Concrete Cylinder Pipe. American Water Works Association (AWWA).
2. ANSI/AWWA C301-07, (2007). Standard for Design of Prestressed

Concrete Pressure Pipe, Steel-Cylinder Type. American Water Works
Association.
3. ASTM D4541, Standard test method for pull-off strength of coatings

using portable adhesion: American Standard for Testing and Materials
(ASTM)
4. ASTM C805, Standard Test Method for Rebound Number of Hardened

Concrete: American Standard for Testing and Materials (ASTM)
5. AWWA (draft) CFRP Renewal and Strengthening of PCCP: American

Water Works Association Research Foundation
6. Gipsov, M.P. and Pridmore A.B. (2014), Conquering Construction

Challenges during PCCP Rehabilitation, ASCE Pipelines Conference
2014, Portland, OR.
7. Pridmore, A.B. and R.P. Ojdrovic (2015), “Chapter 2: Trenchless Repair

of Concrete Pipelines Using FRP Composites”; Ed. V. Karbhari;
Rehabilitation of pipelines using fiber reinforced polymer (FRP)
composites.
© ASCE
Pipelines 2015 1277
8. Gipsov, M.P. and P.S. Fisk (2014), PCCP Rehabilitation with Fiber
Reinforced Polymer Composite – Confirming Installation Conditions
ASCE Pipelines Conference 2014, Portland, OR.
9. Gipsov, M.P. and A.B. Pridmore (2012), WSSC’s Systematic Approach to

the CFRP Liner Installation Process, ASCE Pipelines Conference 2012,
Miami, FL.
10. Zarghamee, M.S., M. Engindeniz, and N. Wang (2013) CFRP Renewal of

Prestressed Concrete Cylinder Pipe, Report #4352. Water Research
Foundation
© ASCE
Pipelines 2015 1278
Better Data Equals Better Decisions: New Developments in Multi-Sensor

Condition Assessment Technologies
Csaba Ékes
SewerVUE Technology Corp., 4650A Dawson St., Burnaby, BC, Canada V5C 4C3.
E-mail: info@sewervue.com
Abstract
This paper describes the development and successful applications of a CCTV,

LIDAR and sonar based pipe inspection system that is robust to gather quantitative
data for critical underground pipe condition assessment. The system that can be
deployed on a ROV or on a float and produces accurate cross-sectional analysis and
sediment volume. This capacity is increasingly critical in large diameter pipes with
high level of flow. The system employs a time of flight LIDAR that is accurate to
1/16th of an inch. Results from recent projects are discussed in detail. The Huntington
Trunk sewer in Abbotsford, British Columbia, Canada is a critical line in the
municipality’s waste-water system. It is a PVC and HDPE pipe that also carries
sewage from Sumas, WA. Pipe diameters vary between 10” and 27” with highly
variable flow conditions. Hard to access, off street manholes located in a swamp and
on a railway right of way created challenges during deployment. The robust, yet
modular SewerVUE multi-sensor pipe inspection system (MPIS) was repeatedly
reconfigured during the project to accommodate the challenging site conditions. The
sonar results provided accurate sediment volumes and cross sectional restrictions. The
Quai George Gorse combined sewer in Boulogne-Billancourt, a southerly suburb of
Paris, France is a critical interceptor in the SEVESC operated collection system. This
2200 mm wide and 2700 mm high, irregular shaped (“cunette avec banquettes”)
reinforced concrete pipe runs parallel to the Seine river and experiences wet weather
overflows during extreme rainfall events. The primary objective of the survey was to
quantitatively measure sediment volume and distribution within a 1275.8 m long
section. This paper presents the methodology and the results of the inspection.
Advanced pipe condition assessment technologies, such as the CCTV, LIDAR and
sonar system described in this paper are cost-effective, non-destructive methods that
are able to help better refine estimated remaining life of an interceptor, accurately
determine overall severity of pipe degradation, as well as provide a basis for
improved cost allocation and timing of rehabilitation efforts.
© ASCE
Pipelines 2015 1279
INTRODUCTION
Obtaining quantitative data which allows for objective assessment of pipes is

of increasing interest to engineers, contractors, and municipalities. Conventional
closed-circuit television inspection technologies cannot adequately meet this need due
to the subjective and imprecise nature of the assessment process. Laser profiling is an
emerging technology that has been shown to provide precise measurements of pipe
parameters such as ovality, unobstructed cross-sectional area, pipe deformations,
lateral size, offset joints, and flow levels.
Accurate pipe dimensional data is especially critical for CIPP (cured in place
pipe) design engineers, as ovality is one of the main influence factors in the CIPP
liner design equation as specified in ASTM F1216-03 (Dettmer et al., 2005). With
accurate measurements, CIPP liners can be designed more cost effectively by
reducing the required thickness of the liner.
Municipalities and contractors have also shown interest in the verification of

the dimensions of newly installed CIPP liners or pipes (Shelton and Travis, 2012).
Having an accurate pipe-wall profile of both pre- and post-installation would
guarantee that the liner was designed correctly and help determine which party would
be responsible in the event of a liner failure. For example, if the liner was designed
for a known geometry but was installed incorrectly by a contractor, accurate pipe
profile data could verify the fact that the contractor is at fault. Conversely, if the liner
was installed correctly as specified by the design engineer, and the liner fails, then the
fault would lie with CIPP liner designer, not the contractor (Dettmer et al., 2005).
The laser profiling concept as well as its inherent measurement errors are
described by Dettmer (2007) and by Dettmer et al. (2005). There are several
commercially available models on the market. Their reliance on accurate calibration
and unreliable field accuracy was pointed out in a seminal paper by Shelton and
Travis (2012).
The approach outlined in this paper employs LIDAR (LIght and raDAR) an
optical remote sensing technology that measures properties of scattered light to find
range and/or other information of a distant target. The prevalent method to determine
distance to an object or surface is to use laser pulses. Like the similar radar
technology, which uses radio waves, the range to an object is determined by
measuring the time delay between transmission of a pulse and detection of the
reflected signal. The SewerVUE MPIS’s LIDAR data is correlated with an onboard
inertial navigation system (INS) that uses a computer, motion sensors
(accelerometers), and rotation sensors (gyroscopes) to continuously calculate via dead
reckoning the position, orientation, and velocity (direction and speed of movement)
of the inspection platform without the need for external references. This technology is
commonly used on vehicles such as submarines and guided missiles and is specially
adapted for the use of multi-sensor inspections for underground infrastructure surveys
where LIDAR is utilized and location and time measurement data is necessary. The
© ASCE
Pipelines 2015 1280
multi-sensor system can be deployed from an autonomous robot (or ROV) or from a
floating platform. Successful applications for each are described in the following case
studies.
METHODOLOGY
LIDAR Theory
LIDAR (also written Lidar or LiDAR) is a remote sensing technology that

measures distance by illuminating a target with a laser and analyzing the reflected
light. Although erroneously considered to be an acronym of LIght Detection And
Ranging, the term Lidar was actually created as a portmanteau of "light" and "radar".
Lidar uses ultraviolet, visible, or near infrared light to image objects. It can
target a wide range of materials, including non-metallic objects, rocks, rain, chemical
compounds, aerosols, clouds and even single molecules. A narrow laser-beam can
map physical features with very high resolution. Wavelengths vary to suit the target:
from about 10 micrometers to the UV (ca. 250 nm) range.
Understanding how each laser profiler works, its advantages and

disadvantages, is imperative for engineers in charge of pipe specification, installation,
maintenance or testing. Output from laser profiling systems can vary greatly. For
example, the difference in results even from the same ring laser profilers operated by
different contractors can be significant (Shelton and Travis, 2012). Municipalities and
engineers must carefully assess the repeatability, accuracy and calibration of the
employed systems.
Continuous-ring profilers use a planar laser whose light rays emanate radially
outward in a continuous fashion from a fixed focal point. The laser plane is
perpendicularly aligned to the pipe axis. Incident rays on the interior wall readily
illuminate its orthogonal cross section. Using a calibrated high-definition digital
camera, the illuminated ring is imaged along the pipe’s axis and then analyzed.
Because of the camera calibration, the digitized image contains usable spatial
information (known relation between pixels and actual distance). By counting the
number of pixels from the center of the pipe to the incident laser, many radial
distance measurements are obtained simultaneously along the pipe wall. When the
camera-laser is in motion, the camera frame rate assures that the illuminated ring is
imaged at fixed intervals along the pipe (Salik and Conow, 2012).
LIDAR systems use a scanning laser that moves back and forth in a single
plane. Distance measurements are acquired by measuring the time it takes for the
laser to bounce off of a target and return to its origin. Because the light propagation
speed is constant, distance can be determined from the “time of flight.” The scanning
motion results in a plane that projects along the interior pipe wall. Because the laser’s
angular step remains constant, the orthogonal measurements from the pipe’s center to
the wall are taken only two at a time (per sweep), but at many non-uniform distances
© ASCE
Pipelines 2015 1281
from the robot. When placed in rotation, many pairs of distances are acquired so that
a ring of measurements is formed. This measurement ring forms a 2-D cross section,
and with many sections obtained simultaneously, a 3-D pipe profile can be created
(Salik and Conow, 2012).
ABBOTSFORD, BC, CANADA
The Huntingdon Trunk Sewer is a 10 inch to 27 inch (250mm to 675mm)

diameter PVC and HDPE and sewer pipe in the City of Abbotsford’s (City) collection
system. This is a critical line since it also carries sanitary sewage from Sumas, WA.
Typical issues include FOGs and high sedimentation reducing capacity and causing
SSOs. The objective of the inspection was to determine the condition of the inspected
pipes by mapping out the accumulated sediments at the bottom of the pipe. The
inspected sections were located between Farmer St and McConnell Rd (Figure 1).
The total inspected length was 5,589 ft (1863 m).
The inspection took place between March 26 and April 1 2014. SewerVUE
crew was assisted by the City and a local contractor for traffic control and site safety.
This paper presents the methodology and results of the inspection.
Figure 1. Overview map of the inspected sewer pipes, Huntingdon Trunk Sewer,
Abbotsford, BC, Canada.
SURVEY EQUIPMENT
The SewerVUE Multi-sensor Pipe Inspector System (MPIS) is a float based

inspection system that uses visual and quantitative technologies (CCTV, LIDAR, and
Sonar) to inspect the condition of underground pipes. This tethered, modular and
© ASCE
Pipelines 2015 1282
customizable second generation MPIS was attached to a light weight and mobile 500
feet (150 m) long tether cable and reel. This allowed truck independent deployment
and operation which was critical for the project since most of the manholes had no
vehicle access (Figure 2).
The floating platform was modified to fit through 18” (450 mm) pipe sections.
CCTV, LIDAR and sonar data were acquired simultaneously in both in and out
directions. The pipe diameter of the inspected sections varied between 18 and 27
inches (450 mm and 600 mm), pipe material was PVC, HDPE and steel. First a guide
rope was installed then the inspection platform was winched through. A total of 5589
ft (1863 m) was inspected.
Figure 2. Deployment of and data collection with the second generation SewerVUE
Multi-sensor Pipe Inspection System on a railway right-of-way in Abbotsford, BC,
Canada.
The primary objective was to measure the height, volume and distribution of
the sediment for subsequent cleaning and maintenance. By quantifying the sediment
distribution over time the City can better maintain the pipe, locate the primary source
of the sediment and take corrective actions.
In total, 2,562 ft3 of debris was detected along the 5,589 ft of the Huntingdon
Trunk Sewer. The average cross-sectional restriction for the sections of pipe ranged
from 3.9% to 28.5% with an overall average of 12.6%. An example of the pipe cross-
sections are shown in Figure 3. The cross-sectional restriction did not appear to
correlate with pipe diameter or material.
© ASCE
Pipelines 2015 1283
Figure 3. Showing an example of pipe cross-sections from the Huntingdon Trunk

Sewer. The line represents sediment level in the pipe. This section had an average
cross-sectional restriction of 9.0%.
BOULOGNE-BILLANCOURT, FRANCE
The 2200 mm wide and 2700 mm high irregular shaped reinforced concrete
interceptor runs parallel to the Seine River in the municipality of Boulogne-
Billancourt in the outskirt of Paris, France. This combined sewer is a critical line in
the sewer network of the municipality that experiences wet weather overflows
directly to the environmentally sensitive Seine River during extreme rainfall events.
Therefore, monitoring the sediment level and volume is critical for the efficient
operation of the sewer. Previously used methods such as measuring sediment depth
with sticks via manned entry provided only point data and are both inaccurate and
potentially dangerous to operators.
SEVESC, the organization in charge of the maintenance of the pipe was

looking for safer and more efficient ways to monitor the condition of the pipe. They
contracted SewerVUE to deploy its MPIS. The primary objective of the survey was to
quantitatively measure sediment volume and distribution within a 1275.8 m long
section. Manholes spaced at regular intervals provided relatively easy access, while
offset manholes provided some operational challenges. The inspection was completed
in late November 2014. The inspection took place while the pipe was in service
flowing between 50 and 70% of capacity.
The SewerVUE’s long range multi-sensor pipe inspection (MPIS) technology

combines state of the art data collection and analysis with proprietary processing and
reporting software. The float based inspection platform is outfitted with high
definition CCTV, LIDAR and sonar sensors and has a 4000 ft maximum deployment
capability. LIDAR measurements determine the exact size and shape of the pipe and
provide quantitative assessment of deformation and corrosion. Sonar accurately
profiles the pipe below the flow line and calculates the sediment and debris volume in
the pipe. The system is customizable and can be deployed through a 18 inch manhole
and can inspect any pipe size over 18 inches. Bypass pumping is not required.
Inspection reports provide integrated and quantitative corrosion and debris
measurements, 180 degree virtual pan/tilt/zoom function, video, laser and sonar flats.
© ASCE
Pipelines 2015 1284
Figure 4. The SewerVUE MPIS and field crew before deployment in Boulogne-
Billancourt.
A total of 285.7 cubic m of sediment was found in the inspected 1214.7 m

long pipe section. In some sections the sediment was up to the top of the central
trough (“cunette”). Lidar profiling did not detect any significant corrosion. The
reported sediment volume and distribution helped mangers of the pipe to prioritize
targeted cleaning and reinforced the need for regular condition monitoring.
© ASCE
Pipelines 2015 1285
Figure 5. Sediment depth, distribution and cross sections for a 125.8 m section of the
Quai Georges Gorse Interceptor.
© ASCE
Pipelines 2015 1286
SUMMARY AND CONCLUSIONS
With limited available funding and budget constraints becoming more

prevalent, timing of rehabilitation and overall intelligent asset management is more
critical than ever for municipalities and asset owners. Advanced pipe condition
assessment technologies, including the SewerVUE multi-sensor pipe inspection
system (MPIS) have demonstrated to be cost-effective, non-destructive methods that
are able to help better refine structural condition and estimated remaining life of an
interceptor, accurately determine overall severity of pipe degradation, as well as
provide a basis for improved cost allocation and timing of rehabilitation efforts.
ACKNOWLEDGEMENTS
The author would like to acknowledge the support received from SEVESC
personnel during data collection and the permission to use the data for this
publication.
REFERENCES
Dettmer, A., Hall, A., Hegab, H., and Swanbom, M. (2005). “Refining laser profiling
method used for pipeline assessment.” Proceedings of the North American
Society for Trenchless Technology (NASTT) 2005 NoDig Show 2005.
Orlando, FL, April 24-27 2005.
Dettmer, A. (2007). “Position and orientation correction for pipe profiling robots.”
PhD thesis, Louisiana Tech University.
Ékes, C., Neducza, B., and Henrich, G. (2011). “GPR goes Underground: Pipe
Penetrating Radar” Proceedings of the North American Society for Trenchless
Technology (NASTT) NoDig Show 2011, Washington DC, March 27-31
2011. pp B-3-02-1-10.
Ékes, C., Neducza, B., Fleury, M., and Hess, D. (2013). Advanced Condition
Assessments: The Benefits of Using Pipe Penetrating Radar. Proceedings of
the North American Society for Trenchless Technology (NASTT) NoDig

Show 2013. Sacramento, CA. pp. TM1-T5-05-1-10.
Salik, J. and Conow, O. (2012). Pipe Inspections: Robotic laser profiling demystified.
Precast Solutions Fall 2012, pp 22-25.
Shelton, J. and Travis, J. P. (2012). Laser profiling – A case study in field accuracy.
Proceedings of the North American Society for Trenchless Technology
(NASTT) NASTT’s NoDig Show 2012, Nashville, TN, March 11-15 2012. Pp
B-5-03-1-5.
© ASCE
Pipelines 2015 1287
Application and Laboratory Tests of Stainless Steel Liner for Trenchless

Rehabilitation of Water Mains in China
Wei Zhou1 and Baosong Ma2
1
Ph.D. Student, College of Engineering, China University of Geosciences-Wuhan, No.
388 Lumo Rd., Wuhan, Hubei Province, China. E-mail: cugzhouwei@163.com
2
Professor, Ph.D., Director, College of Engineering, China University of
Geosciences-Wuhan, No. 388 Lumo Rd., Wuhan, Hubei Province, China.
E-mail: mabaosong@163.com
Abstract
Stainless steel liner, as an emerging method, is being used with increasing frequency
for trenchless renovation of damaged water mains in China. In this paper, the
application of rehabilitating water mains and laboratory testing of stainless steel liners
are described. Practices prove that this new trenchless technology can be effectively
and low-costly utilized to renovate damaged water mains and meet the utility owner’s
requirements. And considering that there have been little studies on the buckling
performance due to external pressure of this new thin-wall structure, the buckling
strength of stainless steel liner is studied by laboratory tests. The research shows that
the buckling resistance of the DR445 liner is more than 7.25 psi. This conclusion
demonstrates that the stainless steel liner meets the requirement in Chinese national
standard GB 50332-2002.
Keywords: Stainless steel liner; Supply pipes; Trenchless rehabilitation; Buckling
strength.
INTRODUCTION
Using a stainless steel liner for the trenchless rehabilitation of water supply pipes is an
emerging approach providing a low cost and a trenchless rehabilitation of water
supply pipes in China. The stainless steel liner can be welded with the host pipe
formed to become a close-fit liner with small annular gap. Generally, due to the
limitation of construction technology, stainless steel liners are mainly being applied to
renovate supply pipelines which are larger than 31.5 inch diameter at present. The
curved stainless plates are manually in-situ welded to be form cylinders inside the
host pipes to be rehabilitated. The most commonly used stainless steels for water
supply pipe rehabilitation have the designation 06Cr19Ni10 which is a form of Type
304 stainless steel. Since most pipes used to build water distribution systems in China
were built decades ago, and research shows that more than 0.16 million miles water
pipes were built before the year 2000( Ma and Zhou, 2013). Now there is an
increasing concern about the remaining service life of these aging water distribution
systems. Due to long term corrosion and stress-induced deterioration and damage, the
structural integrity of many old steel or concrete water pipes is such that they need to
be rehabilitated immediately. Carrying out such rehabilitation using open-trench
© ASCE
Pipelines 2015 1288
construction and full replacement involves a great deal of work, requires a significant
amount of time, and is often very costly (Jeyapalan, 2003 and Najafi, 2013). Stainless
steel liners have many desirable characteristics for the trenchless rehabilitation of
water supply pipes and they have shown excellent performance and a broad market
prospect.
BACKGROUND OF THE PROJECT
The City of Weifang in Shandong Province, which was named as “the World Kite
Capital” is located 410 kilometers northwest of Beijing. The city has a population of
more than 9 million. The Weifang Water Company owns a 47.2 and 55.1 inch
diameter concrete pipes, which are the water mains of the downtown built in 1990 and
2000 respectively, and buried under the cities’ main road as showed in Figure 1. Both
of the length is nearly 1 mile. Considering the fast development of the city’s
population and industry, the currently designed pressure 14.5 psi is becoming lower
and cannot meet the requirement of the city. The owner plans to improve the internal
pressure capacity and supply capacity.
The water mains are located in the main road of the city, the traffic is very busy. These
factors made traditional methods of water main replacement impossible in terms of
cost and customer service. Due to the high tensile strength to resist internal pressure
of the stainless steel, it was applied to renovate the concrete water mains. To
guarantee continuous water supply while constructing the stainless liners and reduce
the cost for installing temporary by-pass pipes, the owner plans to repair the two
water mains in sequence.
Figure 1. Location of the buried concrete pipes
STAINLESS STEEL LINERS CONSTRCTION
1 Required material properties of stainless steel
© ASCE
Pipelines 2015 1289
The most commonly used stainless steels for water supply pipe rehabilitation have the
designation 06Cr19Ni10 which is a form of Type 304 stainless steel. For higher
requirement of corrosion resistance such as high chloride concentration environment,
the designation 06Cr17Ni12Mo2 and 022Cr17Ni12Mo2 which is a form of Type 316
and 316L stainless steel respectively can be utilized as ruled in table 1. The mechanic
properties should meet the requirement in national standard GB/T 228-2010, Code for
metal material tensile test as showed in table 2. In this case, the Type 304 stainless
steel was chosen because the renovated pipes are water pipes.
Table 1. Choice for different types of stainless steel
Type chloride Application
3
concentration/lb/ft
06Cr19Ni10 ≤0.012 Water and gas pipes.
(Type 304)
06Cr17Ni12Mo2 ≤0.062 Pipes with higher corrosion
(Type 316) resistance than type304.
022Cr17Ni12Mo2 ≤0.062 Sea water or medias with
(Type 316L) high chloride concentration
Table 2. Mechanic properties of stainless steel

Property Minimum value Testing code
Tensile/psi 7.54×104
Elongation/% 35
GB/T 228-2010
Yield strength/psi 4.64×104
Area reduction ratio/% 30
2 Construction procedures of stainless steel liners

The construction procedures of stainless steel liners are quite similar as the other pipe
liners. And the process can be showed as follow in Figure 2.
© ASCE
Pipelines 2015 1290
Figure 2. Construction flow-process diagram of stainless steel liners

2.1 Excavation of construction pits
The stainless steel liners are inserted into the water main through the construction pits
excavated along the water mains. The location of the pit was selected to minimize the
number of pits. To minimize the impact of rehabilitation construction to the
surrounded traffic and environments, the area and distance of each construction pit are
limited to be less than 19.6 × 9.8 ft and more than 0.87 mile respectively.
2.2 Pipe cleaning
It is a critical step to cleaning the pipe in the rehabilitation of a water main using
stainless steel liner. The 15 and 20 years old reinforced concrete pipes are fairly clean.
However the deposits and small amount of corrosion on its inside walls have to be
removed to allow the stainless steel liners to adjoin tightly to the host pipes and
restore its flow capacity. The cleaning tools and procedures are quite the same as done
in cured-in-place pipe (CIPP) practice. Cleaning was achieved with the use of a rotary
chain boring tool. Using water pressure, the chains rotate inside the pipe knocking off
the deposits. After cleaning, the pipe was inspected with a CCTV camera to verify
that the corrosions and deposits are removed.
2.3 Joints and leakage treatment
Considering the fact that the host pipes are concrete pipes, any dislocation of the
joints and leakage spots will result in gaps between the liner and host pipe. They
should be treated by grouting mortar to keep the inner wall of host pipe smooth. The
stainless steel liners can adjoin to the host pipes tightly and get enhanced by the host
pipe.
2.4 Welding process of stainless steel liners
The 7.9 ft long and 0.07 inch thick stainless steel plates are pre-produced in factory,
© ASCE
Pipelines 2015 1291
and each plate is curved as showed in Figure 3 by a rounder. The required separation
distance of welded joint between each stainless plate is no less than 0.79 inch. And the
position of 4 o’clock and 8 o’clock are recommended as welding joints for the
adjacent stainless steel plates.
Figure 3. The curved stainless plates

They are manually welded in-situ to form cylinders inside the host pipes to be
rehabilitated. Generally, stainless steel liners can be fitted to the host pipes evenly and
with a good form-fitting shape using this method. For overlapped and manually
welded liners, there would be an inevitable gap adjacent to the overlap point. The first
segment of the overlapped liner will cause a gap which is equal to the thickness of the
liner, and the second part of liner is welded onto the inner wall of the first liner. And
the gap of the second liner near the overlapped part doubles, i.e. d=2t, as shown in
Figure 4. The welding quality inspection should be conducted after the welding
construction based on Chinese national standard GB50235-97, Code for construction
and acceptance of industrial metal pipeline engineering. Figure 5 shows the stainless
steel liners after insertion into the host pipe in practice by in-situ manual welding
method.
© ASCE
Pipelines 2015 1292
d=2t
welding point of
host pipe the second liner
stainless liner
R1
(a)
The second
segment of liner
2t
Overlapped part Inner wall of the

host pipe
(b)
Figure 4. Gap caused by overlapping stainless steel liners
Figure 5. Stainless steel liners after insertion into the host pipe (from BAODING
JINDI SCIENCE&TECHNOLOGY DEVELOPMENT CO., LTD)
QUALITY INSPECTION AND TESTS
1 Welding quality inspection
© ASCE
Pipelines 2015 1293
The welding quality assessment was conducted by tensile and bending strength tests
of the welded stainless steel samples. And the conclusion can be made that the tensile
strength of the welding joints is higher than that of the stainless steel material. Most of
the fracture spots are located beyond the center of the welding joints. And no flaw was
found in the surface after being curved in 180°. The inspection proved that the
welding quality was excellent.
2 Mechanic property tests
The required material properties including tensile strength, yield strength and
elongation at break were tested based on related national standard GB/T3280-2007,
Cold rolled stainless steel plate, sheet and strip, and the results were compared with
the specified values as showed in table 3. The mechanic property tests showed that the
applied stainless steel meet the requirement of the national code.
Table 3. Comparison between the Test results of stainless steel samples and
specified values in GB/T3280-2007
Property Tested results of specified values in the
stainless steel samples standard
5
Tensile/psi 1.02×10 ≥7.47×104
Yield strength/psi 4.13×104 ≥2.97×104
Elongation/% 63 ≥63
3 Hydrostatic pressure test of the stainless steel liner
As specified a hydrostatic pressure test of independent 47.2 inch diameter stainless
steel liner was conducted to investigate the structural resistance to internal pressure.
The test was successfully carried out at 166.8 psi pressure for 72 hours, which was
more than twice the 72.5 psi operation pressure. And the hydrostatic pressure test of
the renovated pipe also was carried out for the final acceptance of construction. The
test was conducted at 116 psi for more 15 minutes, and the pressure did not drop
which showed a good performance of internal pressure resistance.
BUCKLING TEST OF STAINLESS STEEL LINER
In water pipelines, it is the high tensile strength to resist internal pressure of the
stainless steel that needs to be first considered. However, little attention has been paid
to the buckling strength of this thin stainless steel liner, whose DRs typically are more
than 300 in practices (Ma, 2014). And also the bulges inevitably caused by
construction technology will badly impact the buckling strength (Sawy and Moore,
1998). The motivation for vacuum pressure testing is to better understand the critical
buckling pressure for the types of stainless steel liners described above. Vacuum
buckling testing is a practical way to try to acquire the necessary data to create a guide
for the practical use of such liners. Earlier studies showed a length of 10 times its
inside diameter (L/D =10) is adequate for representing the condition of a relatively
long pipeline in the laboratory, and eliminates the effect of the restrained ends of the
liner on the measured buckling pressure (Bakeer, 1999).The length of each test
section L is set to be 32.8 ft, and the diameter of tested liners D is less than 2.62 ft.
The ratio L/D＞10, which will certainly eliminate the effect of the restrained ends.
© ASCE
Pipelines 2015 1294
For the vacuum testing, each end of the liner is sealed with plates to allow the creation
of the vacuum within the test section. Figure 6 shows a schematic of the experimental
setup.
blindplate
steelpipe
IR Camera
Valve c
welded
Valve b
valve a
stainless liner
32.8ft
Valve d
display vacuum pump
air outlet
pressure
recorder
Figure 6. Schematic for experimental liner testing
After the stainless steel liners were inserted into the host pipe, they were welded
together to be a whole pipe inside the 2.62 ft diameter pipe. And the conclusion can be
made that the quality of the liner insertion for the laboratory tests was equivalent to
that typically seen in the field. For the test itself, a vacuum pump is used to decrease
the internal pressure of the liner at the rate of 0.725 psi/min with the inlet valves
opened. The test lasted for less than 10 minutes. Infrared radiation (IR) cameras were
used to record the liner deformation and the buckling failure process. The pressure
recorder was used to read the values of buckling pressure. The 0.07 inch thick,
diameter ratio (DR) 445 stainless steel liner was tested. Figure 7 shows the initial
overall buckling of a stainless steel liner of the test. It can be seen that the buckling is
overall along the axis formed at the place where there were obvious imperfections, i.e.
the gap due to overlapping of the welding points. And the tested critical buckling
strength of the liner is more than 7.25 psi, which means that the stainless steel liner
meets the requirement in criterion in national standard GB50332-2002, Structural
design code for pipelines of water supply and waste water engineering.
Figure 7. Initial buckling of the liner from IR camera
© ASCE
Pipelines 2015 1295
CONCLUSIONS
Stainless steel liners are an emerging and innovative method used in the trenchless
rehabilitation of water supply pipes in China. They have been shown to provide an
excellent performance with good installation practices in China. The application of
the stainless steel liner showed that the internal pressure resistance of this liner is
excellent due to the high tensile strength. And the constructing technology of stainless
steel liner will inevitably cause a gap which is twice the thickness of the liner. This
paper provides an initial investigation of the critical buckling strength of stainless
steel liners inserted into host pipes by conducting full scale laboratory buckling tests.
And the test showed that the DR445 stainless steel liner have a buckling strength of
more than 7.25 psi after the insertion into a steel pipe. The conclusion shows that this
innovative method for trenchless rehabilitation of supply pipelines meets the
requirement in the relative Chinese national standard. And more detailed research on
the buckling strength and the design theory of stainless steel liners is needed in the
future.
REFERENCES
Bakeer, R. M., Barber, M. E., Pechon, S. E., Taylor, J. E., and Chunduru, S. (1999).
‘‘Buckling of HDPE liners under external uniform pressure.’’ J. Mat. in Civ.
Engrg., ASCE, 11(4), 353–361.
El-Sawy, Khaled, and Ian D. Moore. (1998) “Stability of loosely fitted liners used to
rehabilitate rigid pipes.” Journal of Structural Engineering, 124(11),
1350-1357.
GB/T 228 (2010), Code for Metal material tensile test
GB/T3280 (2007), Code for cold rolled stainless steel plate, sheet and strip.
GB50235 (1997), Code for construction and acceptance of industrial metal pipeline
engineering.
GB50332(2002), Structural design code for pipelines of water supply and waste
water engineering.
Jeyapalan, J.K. (2003) ‘‘Future of America’s Water Is in our Hands,’’ in Proc. ASCE
International Conference on Pipeline Engineering and Construction: New
Pipe-line Technologies, Security and Safety, 1026 (2), 13–16.

Ma, B. S., and Zhou, W. "China's Municipal Pipelines: Today and Tomorrow."
Pipelines 2013 (2013): 10-18.
Ma, B. S. (2014) “Trenchless Pipeline Rehabilitation and Renewal Technology.”
China Communications Press, Beijing.
Najafi, M. (2013). “Trenchless Technology Planning, Equipment and Methods,”
McGraw-Hill, New York, NY.
© ASCE
Pipelines 2015 1296
Non-Invasive and Remote Pipeline Rehabilitation Technology

Using Reactive and Magnetic Particles
M. Makihata1; B. Eovino1,2; X. Jiang1,2;

A. Toor1,2; K. L. Dorsey3; and A. P. Pisano1,3
1
Dept. of Mechanical and Aerospace Engineering, University of California at San
Diego, P.O. Box 92093-0411, 9500 Gilman Dr., La Jolla, CA. E-mail:
mmakihata@eng.ucsd.edu
2
Dept. of Mechanical Engineering, University of California at Berkeley, Berkeley,
CA.
3
Dept. of Electrical and Computer Engineering, University of California at San
Diego, La Jolla, CA.
Abstract
We propose a novel pipeline rehabilitation technique that uses particles of a

reactive, multi-layer metallic foil to repair internal cracks of a pipeline with minimum
on-site effort and no downtime. The principle of this repair technique is as follows:
when cracks are detected during routine pipeline maintenance, the particles are
introduced into the fluid flow in the pipe, then manipulated by an external magnetic
field to fill the cracks or pits. Once the particles are in the site of interest, induction
heating is externally applied, causing the reactive metallic foil to undergo an
exothermic diffusion reaction and sintering the particles in the crack. In this paper, we
experimentally confirm the feasibility of such a reaction within a pipe repair system.
We investigate the reaction and bonding strength for various simulated crack sizes,
particle sizes, and particle mixture compositions. We observed stable reaction
propagation in 30 mm long, 100 μm diameter glass capillaries at velocities of 80-100
mm/s. Shear tests were also performed on the reacted particles. A maximum shear
stress of 3 MPa was applied between sintered particles and a simulated crack in
carbon steel, demonstrating the ability of reactive particles to repair small pipe cracks.
INTRODUCTION
Pipelines that transport petroleum product, natural oil, or water are our life
lines. Inadequate maintenance of these pipes poses the risk of leaks, ruptures, or
explosions, leading to property damage and environmental destruction. Modern
technology allows early detection of damage to a pipe through the use of non-
destructive inspection tools. However, it is economically difficult to repair these
damages when they are detected (Castanier, B. et al, 2006), especially for pipelines in
remote locations, such as off-shore pipes. Several repair technologies have been
proposed for remote locations, such as mechanical clamping of off-shore pipes
(Espiner, R. et al (2008)), recoating the inside of the pipe, mechanically sealing the
pipe, or inserting a new pipe inside an existing pipe (Morrison, R. et al, 2013)). All
of these technologies require major on-site efforts (e.g., heavy construction) and also
© ASCE
Pipelines 2015 1297
force the suspension of the pipeline’s operation, causing lost revenue while the pipe is
repaired.
There is strong demand for non-invasive technologies that can repair minor
pipe damage with little on-site effort. We introduce a technique for filling pipe cracks
by using a particle mixture of solder and a reactive, multi-layer metallic foil. The
particle mixture is injected in the fluid flow and guided to the crack via an externally
applied magnetic field. The application of an energy source (e.g., induction heating,
electrical current) starts a large exothermic reaction in the foil, which sinters the
solder and repairs the crack. As the particle mixture is introduced to the pipe during
its operation, no revenue losses result from taking the pipe out of operation.
Additionally, this method does not require precise alignment of repair equipment with
the crack, reducing the on-site effort for maintenance.
In this paper, we determine the optimal composition and size of the particle
mixture for pipeline crack repair. We also experimentally confirm the feasibility of
this technique through reaction propagation and bonding strength tests.
NON-INVASIVE REPAIR USING THE PARTICLE MIXTURE
Figure 1 shows the principles of the repair technique. Once the damaged areas
are identified, particles are introduced and suspended in the fluid stream. Mukherjee,
D. et al. (2014) previously introduced a method to guide particles into position using
a longitudinal magnetic field on the pipe wall. The pipe discontinuity at the crack site
Figure 1. An overview of the three major steps in the non-invasive pipe repair
technique.
© ASCE
Pipelines 2015 1298
causes leakage of magnetic flux and preferentially attracts the particles. This paper
does not explore particle guidance within the pipe, but the method is briefly described
to illustrate a complete repair system.
After particle guidance into the crack, the exothermic reaction of the reactive
foil particles is activated through an external source, such as induction heating. Once
the reaction begins, the surrounding material is heated, sintering the particles in the
crack. Multiple damage sites may be repaired in parallel by applying the activation
energy source to several cracks in the same area.
COMPOSITION AND FABRICATION OF PARTICLES
The particles used in this technique are fabricated by modifying 40 μm thick

NanoFoil (NF40, Indium Corp., Clinton, USA), a sheet of reactive foil composed of
alternating nanoscale layers of Al and Ni (inset, Figure 2). Nanofoil undergoes a self-
sustaining exothermic reaction when it is thermally or physically shocked (Indium
Corporation, (2012)). As the reaction is not combustion-based and does not require
oxygen to proceed, the reaction could take place in a sealed pipe.
The NanoFoil particles are prepared by mechanical grinding using a mortar.
The NanoFoil sheets are covered by water or ethanol during grinding to prevent a
premature reaction due to the physical shock of grinding. The reacted product of the
reactive foil does not take a liquid phase, so a low melting point solder, Sn96Ag4
(ASTM96TS), is introduced as part of the particle mixture to improve the crack
conformality. The Sn alloy solder is ground to size using a file. A scanning electron
micrograph of the solder with energy-dispersive X-ray spectroscopy is shown in
Figure 3.
To increase the mobility of the particles in the externally-applied magnetic
field, each sheet of foil may be electroplated with a 1 μm thick layer of Ni just prior
to grinding as demonstrated in Figure 2(b). Similarly, the solder particles are mixed
with Fe particles in the mortar to enhance their magnetic properties.
(a) (b)
Figure 2. (a) A scanning electron micrograph of several reactive foil particles
(inset: cross-section of the NanoFoil). (b) A photograph demonstrating the
motion of the NanoFoil particles in a magnetic field.
© ASCE
Pipelines 2015 1299
(a) (b)
Figure 3 Scanning electron micrographs of (a) one solder particle shown with
energy-dispersive X-ray spectroscopy false color, and (b) several solder particles
To sort the particles by size, the solvent is evaporated and the particle
mixtures are sieved. Solder-reactive foil particle mixtures are separated into four
average particle diameter ranges: <45 μm, 45–65 μm, 65 μm–90 μm, and 90–200 μm.
In the following sections, the solder-reactive foil concentration of these mixtures is
given as a volume percentage of the solder.
RESULTS AND DISCUSSION
Reaction Speed in Glass Capillaries
Glass capillaries were used as simulated cracks to investigate the reaction

speed, and to determine if the reaction is quenched as the crack size decreases. First,
we investigated the reaction speed under various capillary tube diameters and particle
sizes by photographing the reaction with a high speed camera. A quantity of
NanoFoil particles without the filler metal was loaded into a glass capillary and
activated using a pulse of electric current. Figure 4 shows snapshots of the reaction in
glass capillaries with various inside diameter (ID).

The reaction speeds as measured by the high-speed camera footage (~100
mm/s) were much lower than those of a continuous NanoFoil sheet (2–10 m/s). The
NanoFoil particle reaction is self-propagating up to a distance of a few cm, even with
a 100 μm capillary, where quenching effects should be most prominent. To dissociate
influence of capillary size from the experiment shown in Figure 4, The relation
between reaction speed and particle size was investigated under same packing density
and same diameter size of the capillary. As shown in Figure.5, the reaction speed
increases as the particle size decreases. This finding suggests that the packing factor
is critical to the reaction speed, however crack size is less so.
The solder concentration in the particle mixture also plays a role in reaction
propagation velocity. Figure 5 shows the relation between reaction speed and the
© ASCE
Pipelines 2015 1300
Figure 4. High-speed photographs demonstrating the reaction of NanoFoil

particles in a glass capillary with various capillary diameter and average particle
size.
(a) (b)
Figure 5. Change of combustion speed in capillary due to particle size (a) and
mixing ratio of solder (b).
concentration of the solder observed in a 1 mm diameter capillary. In this reaction,

90~200um NanoFoil and solder particles are used. The maximum reaction speed
occurs at a 20% solder particle concentration. We believe that this peak indicates the
trade-off between increasing the thermal conductivity due to melting of the solder and
decreasing the energy density of the mixture. The upper limit of solder concentration
occurs at 60%, and is also the lowest reaction rate we measured. Above this
concentration of solder particles, the reaction does not reliably self-propagate.
© ASCE
Pipelines 2015 1301
Repair of a Carbon Steel Crack
The repairing performance of this technique is evaluated by reacting the

particles in a simulation crack made of a cylindrical hole (1 mm in diameter) in a
carbon steel block (ASTM A516). Figure 6 shows the experimental setup. The test
blocks are prepared by drilling a carbon steel block with a spiral reamer to get a
smooth inside surface. The test blocks are single-use and are cleaned by
ultrasonication using isopropyl alcohol before the experiment. These Nanofoil sheets
were not plated with Ni before grinding. The particle mixture is loaded into the hole
with a plug at the bottom (Figure 6(2)). After activation with an electrical current, the
force required to push out the sintered particle mixture is measured with an FC22 load
cell (Measurement Specialities, Hampton, USA). The shear bonding strength is
calculated by dividing the maximum force by the surface area of the test hole (9.4
mm2).
Figure 7 is a plot of the shear bonding strength for various solder
concentrations. The bonding strength rises with increasing solder concentration until
around 40% solder, and then falls due to decrease in energy concentration. The peak
of bonding strength is measured as 0.3 MPa at 40% solder, as shown in Figure 7(a).
The effect of solder flux on the bonding strength was also studied. In these
experiments, NOKORODE regular paste flux (Rectorseal, Houston, USA) is used to
increase the bonding strength by increasing the solder wettability to carbon steel. The
improvement of wettability was confirmed by measuring contact angle of melted
solder on carbon steel (148° without flux, 44° with flux). The flux is applied by
inserting a flux-coated rod into the hole. The plate is heated to 100℃ to flow the
solder, the particles are introduced at room temperature. Once the solder flux is
introduced, the maximum bonding strength increases by a factor of 10 in the optimal
particle mixture (Figure 7(b)). The solder flux may enable better thermal transport
and improves contact between the particles and the sidewall.
The reacted particle mass is inspected by scanning electron microscopy. As
shown in Figure 8, an energy-dispersive X-ray spectroscopy image indicates the
NanoFoil and solder particles are sintered. Generally, the NanoFoil particles hold
their shape while the surrounding solder appear to coat the NanoFoil particles.
Furthermore, some carbon steel debris in the form of Fe is present on the surface of
the sintered particle mass, suggesting that bonding likely occurred between the
simulated crack and the and particle mixture.
© ASCE
Pipelines 2015 1302
Figure 6. Procedure of the experiment for shear strength measurement

(a) (b)
Figure 7. Plots of the shear strength vs. mixing ratio (a) without and (b) with a
solder flux coating.
© ASCE
Pipelines 2015 1303
Figure 8. Scanning electron microscopy image with false color energy-dispersive

X-ray spectroscopy (mixing ratio: 20%, no solder flux)
CONCLUSION
In this paper, we presented a non-invasive repairing technology designed to

repair a pipe without suspension of operation and with less on-site effort. We
synthesized an energetic and magnetic particle mixture and measured its reaction
speed in simulated cracks. The reliability of this technology is also experimentally
estimated by measuring the shear strength between sintered particles and carbon steel.
The strongest bond (3 MPa max shear) occurred at a mixture composition of 40%
solder/60% NanoFoil particles, which balances the use of solder as a crack fill
material while leaving enough reactive NanoFoil to propagate the reaction.
As suggested by Figures 5 and 7, the concentration of solder is a major factor
in bonding strength, while the propagation speed is governed by the particle size and
mixture composition. Coating the simulated crack sidewalls with solder flux further
increased bonding strength. In future work, we will add microcapsules of flux into the
particle mixture to measure the effects on bond strength.
In conclusion, a preliminary investigation toward novel pipe repairing

technology is completed. These findings will be helpful for more advanced
development of pipeline repair technologies that use reactive particle mixtures,
enabling improved pipeline maintenance without increasing cost.
© ASCE
Pipelines 2015 1304
REFERENCES
Castanier, B. et al (2006), "Maintenance optimization for subsea oil pipelines",

International Journal of Pressure Vessels and Piping, Volume 83, Issue 4, April 2006,
Pages 236-243
Espiner, R. et al (2008), "Inspection & assessment of damaged subsea
pipelines: A case study", Proceedings of IPC 2008, 7th international pipeline
conference, Calgaru, Alberta, Canada (IPC2008-64480)
Indium Corporation, (2012), “NanoFoil Properties”, Market sheet, Form
No.98744R0, www.indium.com
Morrison, R. et al (2013) , "State of Technology for Rehabilitation of water
distribution systems", EPA/600/R-13/036 www.epa.gov/gateway/science (technical
report)
Mukherjee, D. et al (2014), "Investigation of guided particle transport for
noninvasive healing of damaged piping systems using electro-magneto-mechanical
methods", Society of Petroleum Engineers. doi:10.2118/169639-MS
© ASCE
Pipelines 2015 1305
2015 ASCE Pipelines Conference - Paper 117
Engineering Rehabilitations Based on Non-Destructive Examinations

Dan Ellison1 and Andy Romer2
1
M.ASCE, HDR, 701 E. Santa Clara St., Ventura, CA. E-mail: Dan.Ellison@HDRinc.com; 2M.ASCE, AECOM.
Abstract
While a combination of condition assessment and pipeline rehabilitation is used extensively for
wastewater sewer mains, this is not true for water mains. One reason is the greater difficulty of
assessing water mains than sewer mains (a simple video inspection does not suffice). Another
reason is the lack of guidelines for selecting and designing a rehabilitation system which
considers the condition of the old water main. This paper summarizes recently completed Water
Research Foundation Project 4473, “The Assess-and-Fix Approach: Using NDE to Help Select
Pipe Renewal Methods”. This report makes the case for employing condition assessment as part
of rehabilitation projects, selecting and designing the final lining only after first scanning the
pipe for defects. To develop useful guidelines, the study investigated the essential properties for
structural lining systems, and how various linings might be applied to impaired mains.
Application of the assess-and-fix approach is feasible today, but requires knowledge and
diligence on the part of utilities. Guidance is now available through this recently completed
study. Utilities are encouraged to adopt this approach.
INTRODUCTION: WHY USE NDE FOR SMALL WATER MAIN REHABILITATION

Few water utilities currently employ active condition assessment methods for small diameter
water mains. Renewal decisions for these assets are generally based on break frequency. After a
main has been repaired several times, its condition is deemed to be poor, and a new main is
planned. Although the planning for main replacement can be quite sophisticated taking into
consideration many factors, the actual evaluation of pipe condition is normally based largely on
leak and break repair records. This repair-on-failure approach is acceptable because a break on a
small main is generally not a high-consequence event, and many managers would rather spend
money replacing pipe than assessing it.
Even though break frequency is usually the prime criterion for renewing a main, it is not
uncommon for mains to be replaced merely on the basis of age. Although it is well understood
that age is a poor predictor of pipe condition, utility managers will sometimes elect to replace a
main because the perceived risks associated with the main are judged too high. The decision to
replace an old main without a history of breaks is also often driven by other factors—concurrent
work along the street, or the general idea that infrastructure must be renewed in a timely manner.
Contrast these management strategies with wastewater systems. Although wastewater and water
pipes are frequently managed by the same utilities, decisions about renewal are quite different.
For wastewater, few would argue that a whole pipe is worthless because one or two repairs were
needed. For wastewater, no one would decide to replace a pipe merely because of its age. For
wastewater pipe, condition assessment is routine and is the driver for nearly all condition-based
renewal decisions. Why is this? Because the method (video inspection) is inexpensive, easy to
Page 1
© ASCE
Pipelines 2015 1306
deploy, and results are readily understood. Not only does the condition assessment method
produce a picture (which even a layperson can understand), but the industry has developed
standard inspection protocols and defect codes for documenting the results.
The Water Research Foundation has funded two projects which aim to broaden the acceptance of
in-pipe non-destructive examinations (NDE) for small diameter water mains. Project 4471
proposes to use NDE in a relatively non-disruptive manner to “sample” pipe in a system, then
apply the information to infer the conditions of similar pipes. Project 4473 goes a step further,
proposing to combine the assessment, engineering, and rehabilitation of water mains into a single
product delivery (the “Assess-and-Fix Approach”). This latter project was recently completed
and its final report will be published this year (2015).
This is the third paper presented to ASCE Pipelines regarding the Assess-and-Fix study. In
2013, the need for the project and basic concepts were outlined. In 2014, progress was reported,
including laboratory testing of rehabilitation methods and field tests of NDE inspection methods.
This paper discusses the results of the study, providing guidelines for how a rehabilitation
method can be selected and designed using data from detailed NDE scanning of the pipe.
THE ASSESS-AND-FIX CONCEPT

So why are managers unwilling to assess water mains? The technology has existed for nearly 20
years and is well proven. The arguments against using the technology are:
1. It’s too expensive. Getting a tool inside a water pipe takes too much effort, and the
assessment services cost too much.
2. It’s too risky. The tool could get stuck or contaminate the water.
3. It’s too confusing. The condition assessment may produce hard-to-interpret results.
Similar arguments are heard regarding water main rehabilitation and why similar long-
established technologies are not more broadly used:
1. These pipes are too old. How do I know the final product will last?
2. It’s not cost effective. We tried rehab once, and didn’t save much money. We would
rather invest in new pipe, where the life-expectancy is better understood.
3. It’s unproven. We’ve tried several methods, but have not adopted any. Where are the
standards?
By adopting an assess-and-fix method for water main renewal, and by implementing it on a large
scale, a water utility could achieve results similar to how wastewater mains are managed. Bad
sections of pipe would be rehabilitated, good sections of pipe would be left intact, and the
method of rehab would be appropriate to the defects. Most importantly, rehabilitation rather than
open-trench would be employed, the street would not be torn apart and projects would be
completed more quickly.
The assess-and-fix concept is to perform condition assessment at the same time that a pipe
rehabilitation project is underway. The decision to rehabilitate a pipe may be made using various
Page 2
© ASCE
Pipelines 2015 1307
factors, as is currently practiced, but the method of rehabilitation is not selected until the pipe is
scanned and its condition is determined. Unlined cast iron pipes are classic candidates for an
assess-and-fix approach, but so is any pipe whose condition is believed to be compromised and
where a trenchless method of renewal would be beneficial.
The difficulties and risks associated with scanning a water main are eliminated, if the inspection
is performed as part of a cleaning and lining rehabilitation project. In rehabilitation projects,
temporary bypass water systems are first installed, and holes are excavated to gain access to the
pipe. The pipe is then cleaned using mechanical scrapers pulled through the pipe. The final step
is to line the pipe. If scanning is performed after the cleaning, but before the lining, the added
field effort is minor. The scanning tool can be pulled through the pipe at the same time that a
final video inspection is often performed. The NDE data can then be evaluated and a lining
selected and designed. On a project involving multiple mains, crews could be directed to other
work, while the engineering evaluation is completed. In this way, work progresses without
significant delay to the project and impact to crew inefficiency. An assessment during rehab is
thus very manageable.
Making the appropriate lining adjustments should also be manageable. When a spray-applied
polymer lining is being used, the thickness of the lining is increased or decreased by adjusting
the travel speed of the sprayer. With an appropriate contract mechanism, an owner can go from
a non-structural to a semi-structural lining, by agreeing to pay for a thicker lining. If the
evaluation indicates the need for a fully-structural method, the contractor may need to procure
materials for a cured-in-place pipe (CIPP) lining or a pipe bursting application. This could delay
completion of a main by several days. In the meantime, the access holes would be traffic-plated
while work continues elsewhere. If the project is large enough, a wide range of lining choices
should be feasible without significant overall disruption to the schedule, but good up-front
planning would be necessary. An owner could also facilitate these adjustments and mitigate
delays by paying to keep lining materials on hand. Materials that are not used on one project
will find application on another, particularly if the infrastructure program is large and
continuous.
While committing a pipe to rehabilitation before it is assessed is counterintuitive, it’s not really
that crazy. Miles of unlined (pre-1940) cast-iron pipe are still found in many systems.
Rehabilitation of these mains can be justified by the water quality and hydraulic benefits
achieved by lining, not to mention the life-extension attained by eliminating internal corrosion.
Many utilities, large and small, already do this, and have been for decades. Through long-
running programs of rehabilitation, several large utilities have in fact completely eliminated
unlined cast iron pipes from their systems. The assess-and-fix approach merely advocates
deferring final selection of the rehabilitation method until the pipe has been scanned, and its
condition is known. At utilities that have implemented large-scale lining programs, the cost per
foot of pipe accomplished ranges from 20 to 60 percent of the cost of replacement. The added
cost of NDE scanning should not significantly alter this cost advantage, while promising the
added benefit of a longer-lasting, better-defined product.
Similarly, utilities often commit to replacing mains based on leak history, age, and other factors.
If these utilities were to commit to trenchless renewal as their primary method of main
replacement, an assess-and-fix evaluation could be used to optimize these renewals. A few
utilities already use trenchless methods as their primary means of infrastructure renewal. By
Page 3
© ASCE
Pipelines 2015 1308
adding assess-and-fix evaluations to their procedures, renewals could be custom-tailored to fit

the true conditions of the mains. In many cases, less expensive rehabilitation methods could be
employed. Arguably, there is little that is accomplished through open-trench replacement of
small diameter mains that cannot be accomplished just as well with a low-impact trenchless
method. This is particularly true through an assess-and-fix approach that matches the
rehabilitation to the condition of the main.
APPROACH TO DEVELOPMENT OF ASSESS-AND-FIX GUIDELINES

An engineering approach was employed to develop the assess-and-fix guidelines found in this
the study’s report. These guidelines expand upon existing well-established standards and
manuals of practice, while applying the latest research and basic engineering principles. In some
cases, assumptions were made where knowledge gaps existed. Because the guidelines are
intended for small diameter water mains (12 inches and smaller), a perfect methodology is not
needed. Small water mains are “low-consequence” assets. They are allowed to fail occasionally.
By accepting the possibility of such occasional failures, over-conservatism is avoided, and
greater overall economy is achieved. Also, simplicity is favored. A guideline that is overly
complex will not allow for the timely field decisions needed for assess-and-fix rehabilitation, and
will never be widely adopted.
By necessity, these guidelines attempt to tie together several “loose ends”—issues that are not
fully debated (much less resolved) within the industry. For instance, what are the basic
requirements of “fully structural” or “semi-structural” lining systems? Where should different
lining systems be applied? How should lining systems be designed? And most importantly, how
can the likelihood of a future rupture be determined from NDE data? While partial answers can
be gleaned from various sources, this report synthesizes and expands upon the available answers,
providing guidance for assess-and-fix rehabilitation that can be implemented today.
Fundamental Requirements for Structural Lining Systems

This study clarifies several important criteria for structural linings, including the paradoxical
properties of adhesion and tear resistance. In determining the structural value of a lining system,
a primary consideration is whether a lining has the ability to withstand the cracking of the host
pipe. If a lining does not keep the water inside when the host pipe cracks, it cannot be
considered fully-structural, and its value as a semi-structural lining is also greatly diminished.
While AWWA Manual M28, “Rehabilitation of Water Mains” alludes to this requirement, the
criteria for a structural lining system are far from clear. As a result, lining systems have been
advertised as fully structural, when in fact, they are not. Basic material mechanics indicates that
linings which adhere are likely to tear when a host pipe cracks, even if the crack is small. This
means adhesion of the lining to the host pipe is undesirable if cracking of the pipe is likely.
On the other hand, good lining adhesion can also be a good thing. Tight adhesion to the host
pipe is often needed to connect the lining system to the service laterals, and a good connection
between lining and lateral is necessary if the lining is to have structural value. Without adhesion,
a more difficult mechanical connection between the lining and lateral is required. Adding these
mechanical connections increases cost and often involves digging holes at many service
connections, reducing the benefit of “trenchless” construction. This means adhesion of the lining
to the host pipe is desirable, if digging is to be minimized.
Page 4
© ASCE
Pipelines 2015 1309
So the question becomes, to adhere or not adhere? There are advantages and disadvantages.
Spray-applied linings are the easiest and least expensive, but are likely to tear upon pipe fracture.
More robust, non-adhered linings are more likely to survive pipe fracture, but may require added
effort to connect the lining to the lateral. There is no current system that is both adhering and
non-adhering (like a Post-it™ Note). By employing an assess-and-fix analysis, the condition of
the pipe is used to decide whether an adhered lining is or is not appropriate.
Tests performed for this study confirmed that spray-applied linings should not be assumed to
survive host pipe cracking. Even if the adhesion is not good, a frictional bond is created by the
internal pressure in the pipe. On the other hand, an earlier, manufacturer-sponsored test has
indicated that a CIPP lining may be tear-resistant, but there are questions and issues associated
with this test that merit additional investigation. Utilities adopting a large rehabilitation program
are encouraged to perform their own tear-resistance testing on real samples of their own in-situ
lined pipes. Likewise, utilities are encouraged to verify that linings and laterals are positively
connected at service laterals and other discontinuities. This may involve excavating and
extracting a few of these connections for examination.
Methods for Assessing Future Pipe Condition

When this study was first conceived, a specific NDE tool was envisioned, which uses remote-
field electromagnetic testing (RFT) scanning. This particular tool has been around for nearly 20
years, and has been validated in various independent studies. However, the use of other
technologies including newly developed magnetic flux leakage (MFL) scanning tools may also
be feasible for assess-and-fix evaluations of iron and steel mains. The basic requirement for iron
main assess-and-fix assessment is that the NDE method needs to detect the depth, size and
spacing of corrosion pits, and also measure the general thickness of the pipe wall. Technologies
that provide a general assessment would not be suitable.
Because water main rehabilitation is intended to last many decades, it is important to design for
the future (not current) conditions. It is therefore important to distinguish between the external
corrosion pits, which will continue to grow, and the internal pits, whose growth will be arrested
once the lining is applied. RFT and MFL tools do not indicate which defects are on the outside or
inside of the pipe. To differentiate external from internal corrosion, the NDE scanning should be
coupled with in-pipe video inspection (and possibly laser profilometry).
For forecasting external pit growth, a fuzzy-logic model developed through another WRF study
is useful. This model shows that pit growth follows a logarithmic curve, slowing substantially as
the pipe ages. According to this model, a pit that is 8 mm deep after 75 years should grow by
only 1 mm in the next 50 years. Thus, for a relatively old pipe, the future condition will not be
dramatically different from the current condition, but pit growth should still be taken into
consideration. Applying this model is quite simple; all one needs to know is the current depth of
the pit and the age of the pipe.
Selecting a Rehabilitation System to Match Pipe Condition

To select a lining method, a decision tree is provided in the report involving four simple
questions:
(1) Is hoop strength significantly impaired? If yes, a fully structural method is needed.
Page 5
© ASCE
Pipelines 2015 1310
(2) Is bending and axial strength significantly impaired? If so, a tear-resistant liner is
needed.
(3) Is significant joint leakage expected? Then a semi-structural method is appropriate.
(4) Is a through-wall hole likely? If yes, a semi-structural method is also appropriate.
The default condition (no significant impairment) warrants a non-structural lining method.1
Of these questions, the second question is the most difficult, because no standard exists which
defines beam-bending deficiency. Each situation is different. Most mains are not intended to be
bent, yet we know from experience that failures from beam bending are very common. If a
material is brittle, bending can fail even a main with little deterioration. Circumferential breaks
caused by bending and axially loadings are influenced by the soil, traffic loading, variations in
temperature, topography and other factors.
The fundamental purpose of the NDE assessment is to determine both the probability of host
pipe failure and the modes of failure that are likely to occur. Depending on how a host pipe
might fail, different lining designs are warranted. Three different methods of making this
assessment are proposed: statistical modeling, deterministic modeling and risk assessment
modeling. None of these methods is perfect, but by considering more than one approach and
applying good engineering judgment, reasonable results are attainable in a reasonable time
frame. Again, analytical perfection should not be a requirement for low-consequence water
mains.
Statistical modeling
Desk-top studies of available data are usually the first step in condition assessment. By
examining various pipe characteristics (age, material, diameter, pressure, soil type, etc.) and
historic records of repairs, the probabilities of different types of pipe failures can be estimated.
Statistical analyses are thus important for planning assess-and-fix projects. Mains may be
selected for assess-and-fix renewal, based on studies that show a high probability of impairment
(structural, water quality or hydraulic).
These statistical analyses also provide valuable input for calibrating the results from the other
analyses. For good reasons, engineers are taught to be conservative in their assumptions and
analyses. Conservatism saves lives and protects property. But for renewal decisions involving
miles of low-consequence assets, over-conservatism can waste money. When looking at pipe
condition data, there may be may be a tendency to assume the worst—believing a high likelihood
of pipe rupture exists, when the risk may in fact be tolerable. Statistical analyses of break data
are useful for ascertaining the true likelihood (the mean and standard deviation) of various
occurrences, helping an engineer avoid overly conservative assumptions.
Ideally, a utility will eventually perform enough NDE scanning that statistical relationships
between NDE data and break data could be developed. For instance, a utility might know the
likelihood of a beam break at the point when pits reach a certain size, in pipes of a certain
1
AWWA Manual M28 provides guidance regarding lining methods considered fully structural (Class IV), semi-
structural (Class II or III) and non-structural (Class I).
Page 6
© ASCE
Pipelines 2015 1311
vintage, in expansive clayey soils, in a certain part of town. Just as baseball managers use
statistics to decide on a change of pitchers, a pipeline manager might use statistics to decide on a
change of linings.
Deterministic modeling
Deterministic methods involve applying scientific and engineering principles to predict future
conditions and calculate stresses. This is the natural approach for pipeline engineers, because it
is how they are taught to design new pipes and other things. They consider the various load
cases, the properties of the materials, calculate the stresses, apply safety factors, and are assured
their creations will last for many decades.
However deterministic models are fraught with complications that render them difficult to apply
to old water mains. These include difficulties in knowing whether a pipe is under a bending
load, what exact materials were used in its construction, what defects currently exist (including
casting defects and fatigue weakening), and how much additional deterioration will occur. It is
common to have little knowledge of the actual wall thickness and the actual mechanical strengths
of the pipe, yet this information is necessary for an accurate estimate of pressure and bending
stresses. Worst of all, the stress calculations can be quite complex without necessarily
producing accurate, reliable results. Varying patterns of corrosion pits create complex 3-
dimensional structures that are not easily modeled. False negatives and false positives are both
likely to occur with deterministic modeling.
Risk Assessment Modeling

Because decisions need to be made quickly and without significant analytical cost, simplicity is
favored for assess-and-fix evaluations. Risk assessment models can be fairly simple. In risk
assessment modeling, the relative risks of failure are evaluated based on various factors. Where
the likelihood of a particular failure mode is considered high, a lining should be selected that
accounts for that failure mode. Where the consequences of failure are also high, a bias towards
conservatism (higher factors of safety) is warranted.
The problem with most risk assessments is that they are often subjective and only produce
relative risks. One pipe is judged to be riskier than another. So if a pipe is found to be “high-
priority”, does this mean it is about to fail, or merely the worst one in a healthy population? To
account for this, risk assessment models need to be calibrated. The statistical and deterministic
models can provide these calibrations.
Designing Rehabilitation Systems to Match Host Pipe Conditions

The soon-to-be published project report provides recommendations for selecting and designing
lining systems, based on evaluations of future host pipe integrity. These recommendations
include suggestions for various design parameters such as what long-term material strengths and
factors of safety to use. For the most part, these recommendations are based on the approaches
used in AWWA and ASTM standards. Because they are heavily debated by subject-matter
experts before their adoption, AWWA and ASTM standards carry considerable weight.
However, the opinions of the report’s authors also are part of the study recommendations, to
provide starting points for the needed debates. An example is a testing protocol suggested for
Page 7
© ASCE
Pipelines 2015 1312
determining whether a lining resists tearing. It is hoped that future AWWA standards will take
into consideration the ideas presented in this study. The Pipeline Rehabilitations Standards
Committee of AWWA is currently working on clarifying many of these issues.
APPLYING THE ASSESS-AND-FIX APPROACH

Two examples of applications of the assess-and-fix method are discussed in the report. The first
involved the NDE inspection several years ago of 9 miles of corroding ductile iron pipe. In this
example, a risk assessment approach was coupled with deterministic and statistical analyses to
evaluate the likelihood of various failure modes for each stick of pipe in the 9-mile pipeline.
Good pipe was differentiated from not-so-good pipe. Had the owner desired it, different
rehabilitation methods could have been confidently used for different pipe reaches, with the
expectation that many decades of additional service would have been achieved. Equally
important, more than half the pipe was found to be in good condition and could have been left
alone for another generation.
The second example involved an assess-and-fix demonstration in the City of Phoenix. This
demonstration was limited in scope, involving approximately 500 feet of main, and was
performed solely to demonstrate the method. The demonstration illustrated how easily a NDE
tool can be pulled through a water main, once the main has been prepared for lining. The
demonstration also illustrated the potential benefits of using NDE when performing water main
rehabilitation. While the pipe was not badly corroded, multiple through-wall pits were detected
by the NDE scanning. These through-holes justified a semi-structural lining rather than the non-
structural lining which was specified for this main. Fewer future leak repairs would be expected
had the more robust lining been applied, and the added cost might have been marginal.
FINDINGS AND RECOMMENDATIONS

The assess-and-fix method can be used today. The necessary technologies exist. Assess-and-fix
is already offered in the marketplace, and has been demonstrated through this WRF study. All
that is required are utilities that wish to employ it.
While there are technical issues to be figured out, they involve refinements rather than proofs of
concept. By joining water main rehabilitation and NDE technologies together, both will
advance: a better-defined rehab product is achieved and the NDE is completed cost-effectively.
By targeting the bad portions of pipe, making use of the good portions of pipe, and spurring
broader use of trenchless rehabilitation, the added cost of employing NDE should be recovered
through lower infrastructure renewal costs, once the method becomes routine.
There is one missing ingredient in assess-and-fix implementation: one or more large utilities are
needed that see the value in this method, adopt it, and push its development. By adopting this
approach as part of a substantial capital improvement program, the assess-and-fix system of
project delivery can quickly advance. Any large water utility should be capable of filling in the
technical gaps, including:
• Standards, criteria, and test methods for linings
• Inspection and analysis methods for timely and more useful NDE assessments
Page 8
© ASCE
Pipelines 2015 1313
Time to Think Outside the Trench

For run-of-the-mill water infrastructure renewal, there is arguably little that is accomplished
through open-trench construction that cannot be accomplished with rehabilitation and other
trenchless methods, but in the water industry, the adoption of trenchless has lagged. Uncertainty
about the integrity of the old main and uncertainty about the value of the rehabilitated product
are two of the reasons. An assess-and-fix approach helps remove these uncertainties.
If water engineers followed the example set by the wastewater community by defaulting to low-
dig approaches, greater industry-wide adoption of trenchless would drive innovation, providing
savings of money, time, and community impacts. The assess-and-fix marriage of rehabilitation
and NDE assessment is a model for how this can be accomplished.
SUMMARY OF STUDY CONCLUSIONS
1. Condition assessment and pipe rehabilitation are both routinely used in the wastewater
industry, where these methods are economically performed, well understood, and addressed
by widely-accepted standards. Because water mains are more complex than gravity sewer
mains, use of condition assessment and rehab in the water industry has lagged. By
employing condition assessment as part of rehabilitation, an owner is able to select a lining
method with confidence. Broad use of this method would lead to development of applicable
industry standards and substantial economies of scale.
2. Remote-field testing (RFT) is currently the preferred technology for asset-and-fix
application. High-resolution, accurate results have been validated by several independent
tests, and the technique has been used on water mains for nearly 20 years. The tools and
services currently available reflect this long experience. RFT tools are available for pipes
ranging from 4-inch to 36-inch. Magnetic flux leakage (MFL) also provides meaningful data
needed for assess-and-fix applications, but only recently has this method been applied to
water main assessment.
3. A visual inspection should accompany most NDE assessments, as an aid to interpreting data.
It is generally important to distinguish internal from external corrosion pitting, since internal
corrosion will be largely stopped with the application of the lining, but external corrosion
will continue. Video inspections using closed-circuit cameras are often performed prior to
lining anyway, so this is not necessarily an added step or extra cost.
4. No method is perfect and no inspection is 100 percent, but most water distribution mains are
“low-consequence” assets which don’t require perfect, precise analysis. With a combination
of graphical data display and interpretation by a trained technician, adequate information
should be available for routine assess-and-fix rehabilitation decisions within a reasonable
time frame.
5. Because corrosion of iron pipe is a generally decelerating process, the future condition of a
50-year old pipe can be confidently forecast if its current condition is known. A pit that
reaches 8 mm penetration after 75 years should grow by only 1 mm in the next 50 years. The
fuzzy-logic pit growth model of Rajani, et al., (2011) can be used to predict the depth of
future pits. This prediction model relies on information regarding historical pit growth for
the pipe being assessed. If maximum pit size and pipe age are known, future pits sizes can be
estimated. Information about the corrosivity of the environment is not required.
Page 9
© ASCE
Pipelines 2015 1314
6. Pipe corrosion (both pitting and general) contributes to various types of failures, but is not
the only influence or aging factor. Other factors to be considered in assessing failure risks
include:
• Wall thickness and pipe diameter
• System pressures, pressure cycles, and surges
• Potential for ground movement
• Type of joint material
• Material ductility
7. Three methods are provided for interpreting NDE data and evaluating the risk of main break:
• Statistic analysis is useful for assessing the likelihood of different types of breaks
and their association with various factors. As a data base of NDE data is built up,
NDE data can also be used in these assessments.
• Deterministic analysis can be used to forecast future pit size and calculate stress
levels and residual safety factors. These analyses can be difficult to perform due to
complex patterns of corrosion pitting, uncertainties about material strengths, and
unknown strains created by pipe bending.
• Risk assessment is a practical way of prioritizing and categorizing pipes based on
their assessed condition, while also taking into consideration pressure, soil stability
and other factors that contribute to breaks. While this method is somewhat
subjective, its accuracy can be improved by comparing results to the statistical and
deterministic methods.
8. The selected rehabilitation method should reflect the type of pipe break considered most
problematic:
• Class2 IV (fully structural) methods are needed for pipes with insufficient remaining
hoop strength
• Class III (semi-structural), tear-resistant methods are appropriate if circumferential
(beam) breaks are likely
• Class II and III (semi-structural) methods are useful for stopping rust-hole leaks and
joint leaks
• Class I (non-structural) methods are appropriate if little external corrosion has
occurred and the pipe has sufficient residual strength
9. In many cases, the tear resistance and water-tightness of lining products need to be tested in
order to confirm that they meet the desired performance criteria. Samples for testing should
be taken from mains lined in place, and tests should be performed under a pressure that
reflects expected system conditions.
10. Several existing standards provide guidance for evaluating deteriorated mains and designing
appropriate lining systems:
• ASME B31G provides guidance for how closely-spaced corrosion pits may be
analyzed
2
Class I, II, III, and IV refer to the lining classifications of the AWWA M28 Manual of Practice.
Page 10
© ASCE
Pipelines 2015 1315
• ASTM F1216 provides a formula for determining the maximum size of hole for
which a Class II or III lining is appropriate. This standard also provides a formula for
determining the hole that may be spanned by a lining.
• Various ASTM and AWWA standards provide guidance for determining the long-
term material properties of plastic lining materials
The ASME and ASTM standards should be used somewhat cautiously, as they were not
developed with water main lining in mind. Also, FEA performed by Brown, et al., (2014)
found that ASTM F1216 was not always conservative.
11. AWWA currently has standards for two Class I systems (cement mortar and 1mm epoxy).
Standards are needed for the other lining systems as well as guidance in evaluating the
condition of mains from NDE data. Starting points for these standards are suggested in this
study. Because small-diameter water mains are generally low-consequence assets, modest
safety factors are suggested, particularly where a ductile system is provided.
12. A demonstration in Phoenix showed the practicality of performing NDE in middle of an
lining project. Had the project been a true assess-and-fix project, a Class II lining would
have been recommended rather than the Class I lining that was applied. This could have
been easily accommodated with a spray-applied polyurea lining.
REFERENCES
American Water Works Association. 2014. Manual M28 – Rehabilitation of Water Mains.
AWWA. Denver, CO.
Battelle, TTC, Jason Consultants, Virginia Tech. 2008. White Paper on Rehabilitation of
Wastewater Collection and Water Distribution Systems. USEPA.
Brown, Michael, Ian D. Moore, and Amir Fam. 2014. “Design Recommendations for
Performance Limits for Cured In Place Liner in Cast Iron Water Pipe.” NASTT No-Dig
Conference. Orlando, FL
Deb, Arun K., Frank M. Grablutz, Yakir J. Hasit, and Jerry K. Snyder. 2002. Prioritizing Water
Main Replacement and Rehabilitation. AWWARF Project 459. Denver, CO.
Deb, Arun K., Jerry K. Snyder, John O. Hammell, Jr., Elizabeth Tyler, Linda Gray, and Ian
Warren. 2006. Service Life Analysis of Water Main Epoxy Lining. AWWARF Project
2941. Denver, CO.
Deb, Arun K., Sandra B. McCammon, Jerry Snyder, and Andrea Dietrich. 2010. Impacts of
Lining Material on Water Quality. WRF Project 4036. Denver, CO.
Douglas, Bruce D. and Douglas T. Merrill. 1991. Control of Water Quality Deterioration
Caused by Corrosion Cement-Mortar Pipe Linings. AWWARF Project 415. Denver,
CO.
Ellison, Dan, Andy Romer, Ray Sterling, David Hall, and Michael Grahek. 2006. No-Dig and
Low-Dig Service Connections Following Water Main Rehabilitation. AWWARF Project
2872. Denver, CO.
Ellison, Dan, Firat Sever, Peter Oram, Will Lovins, Andrew Romer, Steven J. Duranceau, and
Graham Bell. 2010. Global Review of Spray-On Structural Lining Technologies. WRF
Project 4095. Denver, CO.
Ellison, Dan, Graham Bell, Steven Rieber and David Spencer. 2014. Answers to Challenging
Infrastructure Management Questions. WRF Project 4367. Denver, CO.
Page 11
© ASCE
Pipelines 2015 1316
Folgherait, Brian, Ryan Rogers and Shawn Kirsch. 2013. Water Main Rehabilitation Using
Polyurea Linings – Same Day Return to Service. Presentation at the North American
Society for Trenchless Technology No-Dig Conference. Sacramento, CA.
Hannaford, M. A., W. J. Melia, P. M. Hoyt, R. Z. Jackson, 2010, An Advanced Method of
Condition Assessment for Large-Diameter Mortar-Lined Steel Pipelines. Presentation at
AWWA Annual Convention and Exhibition, Chicago, IL.
Hartman, William F., Ken Karlson, and Roy Brander. 2002. “Waterline Restoration Based on
Condition Assessment – A Case Study”. AWWA Distribution and Operations
Conference. September 2002.
Jackson, Rodney Z., Charles Pitt, and Ronald Skabo. 1992. Nondestructive Testing of Water
Mains for Physical Integrity. AWWARF Project 507. Denver, CO.
Kleiner, Yehuda and Balvant Rajani. 1992. “Using Limited Data to Assess Future Needs.”
Journal AWWA. 91(7):47-61.
Kleiner, Yehuda and Balvant Rajani. 2010. Dynamic Influences on the Deterioration Rates of
Individual Water Mains (I-WARP). WRF Project 3052. Denver, CO.
Klopfer, Danny J. and Jeff Schramuk. 2005. “Field Report -- A Sacrificial Anode Retrofit
Program for Existing Cast-Iron Distribution Water Mains”. Journal - American Water
Works Association. December 2005, Volume 97, Number 12.
Lillie, Kevin, Christopher Reed, Mark Rodgers, Simon Daniels, and David Smart. 2004.
Workshop on Condition Assessment Inspection Devices for Water Transmission Mains.
AWWARF Project 2871. Denver, CO.
Makar, J., and N. Chagnon. 1999. “Inspecting Systems for Leaks, Pits, and Corrosion.” Journal
AWWA. 91(7):36-46.
Makar, Jon, Ronald Rogge, Shelley McDonald, and Solomon Tesfamariam. 2005. The Effect of
Corrosion Pitting on Circumferential Failures in Grey Cast Iron Pipes. AWWARF
Project 2727. Denver, CO.
Marlow, David R. and David J. Beale. 2012. Condition Assessment of Water Main
Appurtenances. WRF Project 4188. Denver, CO.
Matthews, John, Ryan Wensink, Erez Allouche, Shaurav Alam and Jadranka Simicevic. 2011.
Performance Evaluation of Innovative Water Main Rehabilitation Spray-on Lining
Product in Somerville, NJ. USEPA. Cincinnati, OH.
Mergelas, Brian and Xiangjie Kong. 2002. Electromagnetic Inspection of Prestressed Concrete
Pipe. AWWARF Project 2564. Denver, CO.
Muster, Tim, Paul Davis, Stewart Burn, Januar Gotama, Scott Gould, Dhammika De Silva, and
Nicholas Beale. 2011. Life Expectancy of Cement Mortar Linings in Cast and Ductile
Iron Pipes. WRF Project 3126. Denver, CO.

Nestleroth, Bruce, Stephanie Flamberg, Vivek Lal, Wendy Condit, and John Matthews. 2013.
Field Demonstration of Innovative Condition Assessment Technologies for Water Mains:
Acoustic Pipe Wall Assessment, Internal Inspection, and External Inspection. USEPA,
Cincinnati, Ohio.
O'Day, D. Kelly, R. Weiss, S. Chiavari, D. Blair. 1986. Water Main Evaluation for
Rehabilitation / Replacement. AWWARF Project 54. Denver, CO.
Rajani, Balvant, Yehuda Kleiner, and Dennis Krys. 2011. Long-Term Performance of Ductile
Iron Pipes. WRF Project 3036. Denver, CO.
Rajani, Balvant. 2000. Investigation of Grey Cast Iron Water Mains to Develop a Methodology
for Estimating Service Life. AWWARF Project 280. Denver, CO.
Page 12
© ASCE
Pipelines 2015 1317
Reed, Chris, David Smart, and Alastair Robinson. 2006. Potential Techniques for the
Assessment of Joints in Water Distribution Pipelines. AWWARF Project 2689. Denver
Reed, Chris, David Smart, and Alastair Robinson. 2004. Techniques for Monitoring Structural
Behavior or Pipeline Systems. AWWARF Project 2612. Denver, CO.
Rockaway, Thomas D. and R. Timothy Ball. 2007. Guidelines to Minimize Downtime During
Pipe Lining Operations. AWWARF Project 2956. Denver, CO.
Romer, Andrew E. and Graham E. C. Bell. 2005. External Corrosion and Corrosion Control of
Buried Water Mains. AWWARF Project 2608. Denver, CO.
Seica1, Michael V. and Jeffrey A. Packer. 2004. “Mechanical Properties and Strength of Aged
Cast Iron Water Pipes”. Journal of Materials in Civil Engineering © ASCE /
January/February 2004.
Stone, Steve, Emil J. Dzuray, Deborah Meisegeier, AnnaSara Dahiborg, and Manuela Erickson.
2002. Decision-Support Tools for Predicting the Performance of Water Distribution and
Wastewater Collection. US EPA/600/R-02/029
Thomson, James and Lili Wang. 2009. State of Technology Review Report on Condition
Assessment of Ferrous Water Transmission and Distribution Systems. USEPA.
Urquhart, Anthony and Stewart Burn. 2008. Condition Assessment Strategies and Protocols for
Water and Wastewater Assets. AWWARF Project 3048. Denver, CO.
Page 13
© ASCE
Pipelines 2015 1318
Asset Management: Performance, Sustainability, and Resiliency Model

Development
Richard O. Thomasson1,2; and Sunil K. Sinha2

1
Arcadis, 14900 Sweitzer Lane, Laurel, MD 20707. E-mail:
richard.thomasson@arcadis-us.com
2
The Charles E. Via Department of Civil and Environmental Engineering, Virginia
Polytechnic Institute & State University, 117 Patton Hall, 750 Drillfield Dr.,
Blacksburg, VA 24061. E-mail: ssinha@vt.edu
Abstract
In the past five years, EPA and other agencies have been talking about sustainability
and resiliency. Each of these areas has received attention on an individual basis.
There are parameters and interdependencies which cross all three areas; performance,
sustainability and resiliency. This paper will present the research to date on
development of the governing parameters associated with each of the three areas. The
basic asset management plan shall be viewed under the following distinct categories,
performance management, sustainability management, and resiliency management.
Parameters for performance will be selected which are the major contributors, based
on: Industry Standards, Standard Practice, Research, Wide Use, Data Reliability, and
Data Sustainability. Based on the evaluation matrix a performance index will be
assigned. Each area will receive an indicator index of 1 (one) to 5 (five). The results
of the research will be presented at the conference.
INTRODUCTION
Asset Management has become a major driver in the water industry for determining
infrastructure management. The deterioration of the infrastructure has resulted in

severe budget ramifications and a great impact on level of service for the water,
wastewater and stormwater utilities. There has been advancement in asset
management throughout the world during the past 15 years. In most cases the asset
management models have dealt with condition and financial issues. In the US, there
has been a great deal of focus on condition assessment, deterioration curves and triple
bottom line analysis.
© ASCE
Pipelines 2015 1319
In the past five years, EPA and other agencies have been talking about sustainability
and resiliency. Each of these areas has received attention on an individual basis.
There are parameters and interdependencies which cross all three areas; performance,
sustainability and resiliency.
This paper will present the research on development of the governing parameters
associated with each of the three areas. The basic asset management plan shall be
viewed under the following distinct categories, performance management,
sustainability management, and resiliency management. The initial three areas will be
defined as follows:
Performance Sustainability Resiliency
Structural Condition Social Assessment Plan
Internal Environment Environmental Risk Mitigation
External Environment Economic Recovery
Each area will receive an indicator index of 1 (one) to 5 (five).
Performance – Parameters will be selected which are the major contributors, based
on: Industry Standards, Standard Practice, Research, Wide Use, Data Reliability, and
Data Sustainability. Based on the evaluation matrix a performance index will be
assigned.
Sustainability – Parameters will be selected which are the major contributors, based
on the same measures listed above. Based on the evaluation matrix a sustainability
index will be assigned. The EPA Guidance Document on Sustainability will be a
guide in development of the index so that important regulatory support will be
attained for the sustainability index.
Resiliency – Parameters will be selected which are major contributors, based on the
same measures listed above. Based on the evaluation matrix a resiliency index will be
assigned. Several existing tools will be reviewed and used appropriately in the matrix
evaluation to develop the index.
Development of an overall three dimensional measure of service level index from 1

(one) to 5 (five), will take place which shows the relative impact of performance,
sustainability, and resiliency.
Water utility asset management programs have been developed following the
U.S. EPA and WERF core definition of maintaining a level of service at the lowest
life-cycle cost and at an acceptable risk. Most utilities, however, only incorporate
performance measures into their asset management plans. A holistic approach to asset
© ASCE
Pipelines 2015 1320
management is more beneficial because it takes into account the short and long term
goals of the utility and can provide better service socially, economically, and
environmentally. To address this approach, the past focus on performance must be
separated into performance, sustainability and resiliency. By separating the Asset
Management Plan (AMP) into the three areas an index can be developed for each of
the areas. A common definition must be developed for each of the areas to focus on
the areas, as there are many conceptions of what performance, sustainability and
resiliency mean. This project will build on past work at Virginia Tech and previous
students Masters and PhD work.
In coordination with Virginia Tech’s Sustainable Water Infrastructure
Management (SWIM) laboratory and information from many utilities and subject
matter experts, information and data is gathered which will result in development of
parameters and index’s for performance, sustainability and resiliency. A review of all
existing performance, sustainability and resiliency academic work and industry
models has been studied and evaluated. The goal is to identify all parameters required
to determine an index on a scale of 1 – 5, for a utility’s performance, sustainability
and resiliency. Then the weighting for each index will be determined to indicate an
overall index for the utility by combining all 3 indices.
BACKGROUND
Over the past decade, many utilities, organizations, and regulators in the
United States and worldwide have developed and published resources relevant to
infrastructure asset management for water, wastewater and stormwater. EPA,
AWWA, WEF, ASCE and others in the United States have urged utilities to move
from a reactive role in asset management to a proactive role. Asset management is
sometimes defined differently by these entities and there are only some aspects of
AMP’s which are common among them. Each has addressed performance,
sustainability and resiliency, but in different ways, and based on a specific definition,
which in some instances is not focused on water, wastewater, and stormwater
conveyance systems.
Holistic Asset Management Framework
The asset management framework developed by Dr. Sinha incorporates basic

elements that build on and complement one another to provide sustainable municipal
infrastructure asset management. Unlike other asset management structures, this
framework links standard asset management concepts, information systems, and
sustainable and resilience management practices. Ideally, this framework provides
utilities with a support system that handles short and long term holistic asset
management planning (Sinha & Eslambolchi, 2006). The framework is outlined in
Figure 1.
© ASCE
Pipelines 2015 1321
Figure 1: Holistic asset management framework. (Gay & Sinha, 2013
The holistic asset management framework incorporates various aspects of

asset management such as data collection, condition assessments, decision-making,
repairs and maintenance, and future priorities into its strategic model. Like many
asset management frameworks, most of the components in the framework support
performance management. Alternatively, the holistic asset management framework
incorporates sustainability and resilience management concepts into the asset
management discussion (Gay & Sinha, 2013). Sustainability and resilience
management, explained more in the next section, are needed for an asset management
framework to be truly holistic. Performance focuses on the physical functioning of
assets as they pertain to providing the desired LOS. Sustainability and resilience
management concepts bring goals and standards to a utility that include the well
being of the community and the environment as well as preparation for disaster.
Performance Management
Performance Management is defined as managing the infrastructure to minimize

the total cost of owning and operating the system, while delivering acceptable service
levels. Performance of a utility can be defined in terms of service life and reliability.
Both of these criteria rely on factors involved with the structural condition, internal
environment and external environment. Parameters must be identified and data
collected from utilities for each parameter.
The following list of parameters has been identified for water systems for
performance are shown in Table 1.
© ASCE
Pipelines 2015 1322
Table 1.Performance Parameters Water Systems

Parameters Attributes
Coating yes/no
Cathodic Protection yes/no
Soil Type A-1; A-2; A-3; A-4-A-5; A-6-A-7
Poor Drainage; Fair Drainage; Good
Moisture Drainage
Stray Current Yes/No
External
Environment 0-1 per year; 1-5 per year; 5-20 per year; 20-
Flooding 100 per year; >100 per year
Unpaved; Non-National Highway System;
National Highway System; Interstate;
Dynamic Load Railroad/Airport
ADT 0-10; 10-100; 100-500; 500-2000; >2000
Disturbances Yes/No
Coating yes/no
Water pH (Baylis Curve) 5-6; 6-7;7-8; 8-9; 9-10
Internal 50-100 psi; 100-200 psi; 200-300 psi; 300-
Environment Design Operating Pressure 400 psi; 300-400 psi
0-1 per year; 1-5 per year; 5-20 per year; 20-
Pressure Surge frequency 100 per year; >100 per year
Lining Yes/No
Lining Type Non-structural; Semi-Structural; Structural
Tuberculation yes/no
Dissimilar Materials yes/no
Structural
Surge Control Valves Yes/no
Condition
Pump Control Valves No Valve; Suring Check Valve; Other
Age
Material type
Joint type
A questionnaire was sent to many utilities, both public and private to collect
the data for each of the parameters. The results showed what data is available and
what data is not presently collected by most water utilities. Based on the statistical
representation of the data, a determination was made of the applicable parameters to
be included in the development of a performance index dealing with potable water.
These results will be completed and presented in the presentation at the 2015
Pipelines conference.
The following list of parameters has been identified for wastewater systems
for performance shown in Table 2.
© ASCE
Pipelines 2015 1323
Table 2.Performance Parameters Wastewater Systems

Parameter Unit Range
unpaved road, gravel, grass,
Ground Cover Type
dirt, loose particle material;
Below pipe; close to pipe
Groundwater Table level <2ft; Slightly above 2-5ft;
above pipe>4ft;
light >50ft from road or
External railway; medium 50ft from
Location (Traffic) level
Environmen road or railway; heavy 20ft
t from major road or railway;
Pipe Depth ft >18; <4
Pipe Slope % 0-5
Soil Corrosivity level
gravel, coarse sand, fine sand,
Soil Type Type
silt, clay
Tidal Influences Yes/no Yes; No
D/d (flow depth over
% 0-100
Internal diameter)
Environmen Flow Velocity f/s 0-5
t Pipe Surcharging Level Frequent; Occasional
Wastewater pH pH 0-9
Very Dense (>5 per 100ft),
Density of Connections level
Dense (4-5 per 100ft)
Very Often, cleaning and
Inspecting every 1-3 years;
Maintenance Frequency Level
Regularly, 3-5 years; Rarely,
Structural >5 ; Never
Condition
Pipe Age Year 0-130

0-5(also comment the
Pipe Condition Level
supporting criteria)
Pipe Diameter inch 0-60
Pipe Length ft 0; 500 ft
© ASCE
Pipelines 2015 1324

be included in the development of a performance index dealing with wastewater.
Sustainability Management
Sustainability management is maintaining a system that continuously satisfies

need without compromising the ability of future generations to satisfy their own
needs from the Triple Bottom Line (TBL) perspective (Gay & Sinha, 2013). Utilities
in other countries, like the UK and Australia, have begun to adopt sustainable
practices into their AMPs (Marlow, Beale, & Burn, 2010 and Rees, Young, &
Richardson, 2009). It has become globally important to address new world challenges
linked to climate change, population growth, damage to ecosystems, and reduction of
greenhouse gases. Because AMPs are always evolving, adding sustainability
management to AMP goals and objectives can help utilities meet these challenges.
Sustainability management can be implemented in small steps (Marlow, 2010). The
most important overarching goal of sustainability management is that in each
decision-making step a TBL perspective be considered (Kenway, Howe, &
Maheepala, 2007). Though the UK and Australia may be global leaders in sustainable
asset management planning, the US EPA has designed a handbook for utilities to start
to imbed sustainable goals into their planning and management of their water and
wastewater infrastructure. They believe that the core mission of water and wastewater
utilities is to provide clean and safe service that includes not just public health but
environmental health and economic sustainability (EPA, 2012). The handbook helps
utilities create goals and implement practices that incorporate TBL thinking into
organizational practices.
Many researchers agree that setting TBL objectives within an AMP into can create
more sustainable services. However, because asset management helps in long term
planning, planning for future wastewater needs and addressing future problems
becomes part of everyday thinking. Nevertheless, adding TBL principles to asset
management help utilities address emerging issues due to climate change along with
changing populations (Marlow, et al, 2010).
Sustainability Management considers the ability of the Infrastructure system
to operate at pre-defined levels of service for indefinite time. The 3 areas of
consideration in sustainability management are Social, Economic, and

Environmental. Parameters must be identified and data collected from utilities for
each parameter.
The following list of parameters has been identified for water systems for
sustainability shown in Table 3.
© ASCE
Pipelines 2015 1325
Table 3.Sustainability Parameters Water Systems
Parameters Unit Comment

System energy use BTU Total Energy used
if yes, please provide details
Asset Management Plan Yes/No (condition assessment
practices; models, tools used)
Utility Revenue $
Utility Expenditure $
Economical Revenue saved for future
$
renewal
Revenue spend for capital
$
improvement
Includes renewal activities,
Revenue spend for
$ routine maintenance,
operation
equipments, salaries, etc.
Type of Energy source (gas,
Source of Energy Type
oil, solar, etc.)
Environmen Source of Water Type
tal Source Water Capacity Gallons Water drawn for treatment
Water Loss Gallons or % unacounted water loss
Pipe Breakages Number
Served Water Capacity Gallons Water supplied to customer
Expected increase in water
Meeting Demand %
supply
Expected increase in
Social Service provision %
customers in 10 years
Customer complaints
Customer Education
Yes/No Being Practiced? (Y/N)
Outreach
be included in the development of a performance index dealing with potable water.
The following list of parameters has been identified for wastewater systems
for sustainability, shown in Table 4.
© ASCE
Pipelines 2015 1326
Table 4.Sustainability Parameters Wastewater Systems

Parameters Unit Comments
System energy use BTU Total Energy used
if yes, please provide
details (condition
Asset Management Plan Yes/No
assessment practices;
models, tools used)
Utility Revenue $
Utility Expenditure $
Revenue saved for future
$
renewal
Economical Revenue spend for capital
$
improvement
Includes renewal
activities, routine
Revenue spend for operation $ maintenance,
equipments, salaries,
etc.
Sewer Flow gallons
% sewer capacity
I&I gallons
Type of Energy source
Source of Energy
(gas, oil, solar, etc.)
Sewer Overflows #
Environmental
Sewer Backups #
Customer complaints #
Customer Education Outreach Y/N Being Practiced? (Y/N)
Social Service provision Expected increase in

# customers in 10 years
be included in the development of a performance index dealing with wastewater.
© ASCE
Pipelines 2015 1327
Resilience Management
Resilience management is the ability to avoid, reduce, mitigate, and ultimately
recover from the effects of natural, accidental, or malevolent incidents with minimal
impact on end-users (Gay & Sinha, 2013). Resilience management is often the most
difficult management structure to add to any utility AMP. It begins with identifying
what hazards to which the utility could be exposed, and then making specific goals to
address them. A deterioration model in the holistic asset management framework
plays a major role in resilience management (Gay & Sinha, 2013). Deteriorating
assets are more susceptible to disastrous events. Resilience management has 3
components: assessment plan, risk mitigation, and recovery.
The following list of parameters has been identified for resilience, shown in
Table 5. These parameters apply to both water and wastewater systems.
Table 5.Resiliency Parameters Water & Wastewater Systems

Parameters Units
Pipe Location
Proximity to other Assets (including
feet
pipes)
Proximity to Hazards feet
Material
Diameter inches
Age
No. of nodes per demand point
Assessment Plan
Likeliehood of Failure
Quality of other Utility Record
Resource availability material,
Y/N
personnel, equipment
Preventive Measures Plan Y/N
Emergecy Preparedness Training Y/N
Financial Impact to Private Property(in
$
case of disruption)
pipe network redundancy Y/N
Coordination Plan with other agencies Y/N
Acceptable recovery cost to original LOS $

Acceptable recovery time to original
Recovery $
LOS
Access to pipe Y/N
Budget at dispense for emergency per
$
area or in total( in $)
Condition
Hazard Type (earthquake, theft)
Risk Mitigation
Type of nearby property
Post Event Plan Available Y/N
© ASCE
Pipelines 2015 1328
CONCLUSION
Previous work at Virginia Tech by Masters and PhD students has developed
models dealing with performance of water and wastewater assets. Data from those
models has been instrumental in the development of the parameter questionnaires for
performance parameters. Those models were run with data from utilities and
produced performance measures for the data collected. Parameters for the
questionnaire for Performance were selected as major contributors, based on Industry
Standards, Standard Practice, Research, Wide Use, Data Reliability, and Data
Sustainability. With Sustainability, the EPA Guidance Document on Sustainability
was a guide to develop the questionnaire parameters. There are several existing
Resiliency references that were used to develop the parameters used on the
questionnaire for Resiliency. The information collected from the questionnaires will
result in identification of the parameters most relevant to determining an index for
performance, sustainability, and resiliency.
A model will be developed to give an index from 1 – 5 for each of the areas:
Performance, Sustainability, and Resiliency. Several utilities which have well
developed asset plans will be polled to review and give verification of the practical
use of the index system rating for their system. The results of the index model will be
presented at the 2015 Pipelines Conference. The relationship between deterioration
curves and performance, sustainability, and resiliency are shown in Figure 2.
Figure 2.Curves of Deterioration Performance, Sustainability, Resiliency
© ASCE
Pipelines 2015 1329
REFERENCES
Allbee, S. (2005). America’s pathway to sustainable water and wastewater systems.
Water Asset Management International, 1 (1), 9 – 14.
Angkasuwansiri, T. (2013)Development of Wastewater Pipe Performance Index and
Performance Prediction Model, Thesis Department of Civil and Environmental
Engineering, Virginia Tech
Association of Local Government Engineers of New Zealand, National Asset
Management Steering Group, Institute of Public Works Engineering Australia. (2006)
International Infrastructure Management Manual (IIMM).(3rd ed.) Wellington, N.Z.:
National Asset Management Steering (NAMS) Group
Duchesne, S., Guillaume, B., Vileneuve, J., Toumbou, B., & Bouchard, K. (2012). A
Survival Analysis Model for Sewer pipe Structural Deterioration. Computer-Aided
Civil Infrastructure Engineering, 28, 146 – 160.
Environmental Protection Agency (EPA). (2012). Planning for Sustainability: A
Handbook for Water and Wastewater Utilities. (EPA Publication No. 832-R-12-001).
Seattle, WA: Ross & Associates Environmental Consulting, Ltd.
Environmental Protection Agency. (2008). Asset Management: A Best Practices
Guide. (EPA Publication No. 816-F-08-014). Washington, D.C.: U.S. Environmental
Protection Agency.
Environmental Protection Agency. (2003). Asset Management: A Handbook for Small
Water Systems. (EPA Publication No. 816-R-03-016). Washington, D.C.: U.S.
Environmental Protection Agency.
Environmental Protection Agency. (2002). Fact Sheet: Asset Management for Sewer
Collection Systems. (EPA Publication No. 833-F-02-001). Washington, D.C.: U.S.
Environmental Protection Agency.
Federation of Canadian Municipalities and National Research Council (FCM).
(2003). Best Practices for Utility-Based Data. Ottawa, Ontario: Federation of
Canadian Municipalities and National Research Council
Gay, L. F., & Sinha, S. K. (2013). Performance, Sustainability, and Resiliency for
Water Infrastructure Asset Management Primer. Unpublished manuscript, Virginia
Polytechnic Institute and State University, Blacksburg, VA.
Gay, L.F. and Sinha, S.K. (2013) Resilience of civil infrastructure systems: literature
review for improved asset management Thesis Department of Civil and
Environmental Engineering, Virginia Tech
Gay, L.F. and Sinha, S.K. (2013) Stochastic simulation methodology for resilience
assessment of water distribution networks Thesis Department of Civil and
Environmental Engineering, Virginia Tech
Ihde, D. (Eds.). (2013). Proceeding from UIM Asset Management Conference 2013.
Distribution System Desktop Condition, Risk, and Replacement Planning. Arlington,
VA: Benjamin Media, Inc.
Kenway, S., Howe, C., & Maheepala, S. (2007). Triple Bottom Line Reporting of
Sustainable Water Utility Performance. Denver, CO: American Water Works
Association.
Marlow, D. R. (2010). Sustainability- Based Asset Management in the Water Sector.
In J. E. Amadi-Echendu, K. Brown, R. Willett, & J. Mathew (Eds.), Definitions,
© ASCE
Pipelines 2015 1330
Concepts and Scope of Engineering Asset Management (216-275). New York:

Springer.
Marlow, D., Beale, D., & Burn, S. (2010). Linking asset management with
sustainability: Views from the Australian sector. Journal American Water Works
Association, 102 (1), 56 – 67.
Marlow, D., Person, L., Whitten, S., MacDonald, D. H., & Burn, S. (2010). Linking
asset management to sustainability though risk concepts: the role of externalities.
Water Asset Management International, 6 (3), 21 – 28
Mathew, J. Ma, L., Tan, A. , Weijnen, M. Lee, J. (Eds.). (2010) Engineering Asset
Management and Infrastructure Sustainability, Proceedings of the 5th World
Congress on Engineering Asset Management (WCEAM 2010)
New Mexico Environmental Finance Center. (2006). Asset Management: A Guide for
Water and Wastewater Systems. Socorro, NM: New Mexico Environmental Finance
Center.
Pedicini, S. E. Utility Asset Management Programming: Performance, Sustainability
and Resilience (2014) Moving from Academia to Practice Thesis submitted to the
Faculty of Virginia Polytechnic Institute and State
Rees, A., Young, T., & Richardson, I. (2009). Change and asset management in
wastewater utilities – a UK perspective. Water Asset Management International, 5
(4), 3 – 7.
Santora, M., & Wilson, R. (2008). Security and Preparedness – Resilient and
Sustainable Water Infrastructure. Journal – American Water Works Association, 100
(12), 40 – 42.
Sinha, S. K., & Eslambolchi, S. S. (2006). Bridging the Gap: An Educational Primer
on Sustainable Water Infrastructure Asset Management. Unpublished manuscript,
Pennsylvania State University, University Park, PA.
Stolte, M. (Eds.). (2013). Proceeding from UIM Asset Management Conference 2013.
WaterID: Real-Life Benefits of Information Sharing and Reporting. Arlington, VA:
Benjamin Media, Inc.
Syachrani, S., Jeong, H. S., Chung, C. S. (2010). Dynamic Deterioration Models for
Sewer Pipe Network. Journal of Pipeline Systems Engineering and Practice, 2, 123 -
131.
Utility Infrastructure Management (UIM). (2008).Guide to Water and Wastewater
Asset Management. Peninsula, OH: Benjamin Media, Inc.
Williams, W., Jones, M., & Stillman, J. (2013). Asset Management: How US utilities
can leverage international experience. Journal - American Water Works Association,

105 (5), 86 – 91.
Water Environment Research Foundation (WERF). (2009). Sustainable Infrastructure
Management Program Learning Environment Knowledge Base (SIMPLE), 03-CTS-
14
Water Environment Research Foundation (WERF). (2009). Remaining Asset Life: A
State of the Art Review, SAM1R06d
Water Environment Research Foundation (WERF). (2010). Strategic Asset
Management and Communication Decision Analysis/Implementation Guidance,
SAM1R06c
© ASCE
Pipelines 2015 1331
Water Environment Research Foundation (WERF). (2010). GWRC Compendium of

Best Practices in Water Infrastructure Asset Management, SAM3R06
Water Environment Research Foundation (WERF). (2010). Phase 1: Development of
Standard Data Structure to Support Wastewater Pipe Condition and Performance
Prediction Predicting Asset Life of Wastewater Pipes, SAM3R06
Water Environment Research Foundation (WERF). (2010). Assessing Utility
Practices with Strategic Asset Management GAP Analysis Tool (SAMGAP),
SAM2R06COa
Water Environment Research Foundation (WERF). (2011). Condition Assessment
Scoring Tool (SIMPLE), SAM1R06
Water Environment Research Foundation (WERF). (2011). Level of Service Tool
(SIMPLE), SAM1R06
Water Environment Research Foundation (WERF). (2011). Capital Improvement
Program Validation and Prioritization Tool (SIMPLE), SAM1R06
Water Environment Research Foundation (WERF). (2011). Business Risk Exposure
Tool (SIMPLE), SAM1R06
Selection Tool (SIMPLE), 03-CTS-14
Water Environment Research Foundation (WERF). (2011). Leading Practices for
Strategic Asset Management, SAM1R06h
Water Environment Research Foundation (WERF). (2011). Research Digest:
Decision Analysis/Implementation Guidance-Asset Management Tool Development,
SAM1R06e
Water Environment Research Foundation (WERF). (2011). Development of Protocols
and Methods for Predicting the Remaining Economic Life of Wastewater Pipe
Infrastructure Assets- Phase 2: Development of Robust Wastewater Pipe Performance
Index, SAM3R06
Water Environment Research Foundation (WERF). (2012). Integration of Cost of
Failure with Asset Management, INFR1SG096
Water Environment Research Foundation (WERF). (2012). Key Asset Data for
Drinking Water and Wastewater Utilities, SAM5C08
Water Environment Research Foundation (WERF). (2012). Towards an Economic
Decision Methodology for Remaining Asset Life: Research Roadmap, SAM1R06g
and Renewal for Water and Wastewater Pipes-National Database (WATERID),
INFR9SG09
Water Environment Research Foundation (WERF). (2014). Leading Practices and
Key Performance Indicators for Asset Maintenance, SAM1R06k
Water Environment Research Foundation (WERF). (2014). Water Infrastructure
Asset Management Primer, INFR9SG09b
Water Research Foundation (WRF). (2002). Costs of Infrastructure Failure, #2607
Water Research Foundation (WRF). (2002). Distribution System Performance
Evaluation, #804
Water Research Foundation (WRF). (2007). Main Break Prediction, Prevention, and
Control, #461
© ASCE
Pipelines 2015 1332
Water Research Foundation (WRF). (2007). Condition Assessment Strategies and

Protocols for Water and Wastewater Utility Assets, #3048
Water Research Foundation (WRF). (2008). Key Asset Data for Drinking Water and
Wastewater Utilities, #4187
Water Research Foundation (WRF). (2010). Dynamic Influences on Deterioration
Rate of Individual Water Mains (I-WARP), #3052
© ASCE
Pipelines 2015 1333
Finite Element Modeling of Full-Scale Concrete Manholes under Soil Pressure
Elmira Riahi1; Xinbao Yu2; Mohammad Najafi3; and Firat Sever4

1
Graduate Assistant, Department of Civil Engineering, The University of Texas at
Arlington , Arlington, TX 76019. E-mail: elmira.riahi@mavs.uta.edu
2
Assistant Professor, Department of Civil Engineering, The University of Texas at
Arlington, P.O. Box 19308, 429 Nedderman Hall, Arlington, TX 76019
(corresponding author). E-mail: xinbao@uta.edu
3
Professor and Director of Center for Underground Infrastructure Research and
Education (CUIRE), Dept. of Civil Engineering, The University of Texas at
Arlington, P.O. Box 19308, 428 Nedderman Hall, Arlington, TX 76019. E-mail:
najafi@uta.edu
4
Project Manager, Utility Infrastructure Environmental Group, 2550 Corporate
Exchange Dr., Ste. 300, Columbus, OH 43231. E-mail: fsever@structurepoint.com
Abstract
Large amount of manholes in the United States are suffering from serious structural
decay and are in need of immediate rehabilitation. Among various candidate
techniques, spray-on-place lining is one of the promising techniques for manhole
rehabilitation. Based on a new manhole rehabilitation classification system proposed
by the authors, epoxy liner is considered as semi-structural material and relies on
residual strength of existing manhole structure to withstand external loads. In this
study, a two dimensional axisymmetric finite element model for concrete manholes is
developed using ABAQUS and calibrated to an existing full-scale manhole
experiment. The calibrated FEM manhole model is then used to model the structural
behavior of the manhole with epoxy liner under in-situ soil pressure. Results show
that concrete residue strength is needed for deteriorated manhole repaired with epoxy
liners to resist in-situ soil pressure load.
INTRODUCTION
Manholes are access points to underground infrastructures. Large amount manholes

are deteriorated due to several reasons, such as design and construction defaults,
climatic or chemical agent damage, corrosion, accidental overloading or impact, fires,
or earthquakes (Nour et al. 2007). The United States Environmental Protection
Agency (USEPA) estimated that 3.5 million manholes are suffering from serious
structural decay and are in need of immediate rehabilitation or replacement (Sever et
al. 2013). Considering the potentially multi-billion dollar market, it is not surprising
that there are already numerous materials and methods available for manhole
rehabilitation. Concrete patching, using manhole linings, corrosion protection and
replacing the whole or some parts of the manhole are some of the methods used to
solve manhole problems. Among all the available methods, No-dig manhole
© ASCE
Pipelines 2015 1334
rehabilitation attracts great interest due to its low cost and minimum interference on
traffic. In this method instead of replacing the manhole structure, different materials
are used to retrieve manholes’ capability to withstand the applied loads and
infiltration.
Several experimental studies have been conducted to evaluate different
properties of manhole rehabilitation materials (Render et al. 2004; Ahn et al. 2009;
IKT 2012). But to the authors’ knowledge, numerical modeling of manhole structures
is limited. Sabouni conducted her doctoral research on load and deformation of new
precast concrete manholes in 2008 (Sabouni 2008). In her study, three manholes
including one reinforced and two unreinforced concrete manholes were fully
instrumented with strain gages and soil pressure cells and tested in a soil test pit under
different load configurations specified in the standards. The manholes were simulated
using PLAXIS to analyze the soil pressure, bending moment, and strains (Sabouni
and El Naggar 2011). A field demonstration project was performed in Cleveland, OH
in a project funded by EPA to evaluate the performance of cured-in-place pipe lining
rehabilitation method (Matthews 2012). According to the report the project resulted in
successful demonstration of an innovative Class IV (fully structural) water main
rehabilitation technology.
Behavior of epoxy coated concrete was studied by testing coated concrete
specimens in flexural and compression tests and by finite element simulation (Riahi et
al. 2014). The knowledge of previous study and existing experimental data on
manholes is adopted to create a finite element model of concrete manhole with epoxy
liner. The model of the manhole structure and surrounding soil is generated in
ABAQUS and then validated with analytical calculations and the results from an
existing full scale manhole experiment. The calibrated model is then used to study the
behavior of the epoxy liner under lateral pressure from the surrounding soil.
FINITE ELEMENT MODEL
ABAQUS/standard is used to model a full scale manhole. A typical 1.2m diameter

concrete manhole structure tested in a previous laboratory study (Sabouni and El
Naggar 2011) was selected for the manhole simulation. This manhole was fully
instrumented with strain gages and pressure cells and tested under various load
configurations in a large laboratory test pit. The available test data allows for
verification of the developed ABAQUS manhole model. After the new concrete
manhole model was validated, deteriorated manhole with epoxy liner was then
modeled to study the structural capacity of epoxy liner for manhole rehabilitation.
The concrete manhole consists of one 1.2m high monobase, four risers, and
one tapered top as shown in Figure 1a. The wall thickness of the monobase is 139mm
with a 1.219m inner diameter and a 150mm thick nonreinforced base. Height of each
manhole part is illustrated in Figure 1a. The total height of the manhole is 5.89m. The
manhole was tested in a soil pit with rigid walls and floor. The soil test pit had a base
of 4.5m by 4.5m and height of 7.26m. A layer of 0.65m sand was located at the
bottom with a layer of 0.3m gravel on top of that. The manhole base was placed on
the gravel layer and the manhole walls were surrounded by 5.33m concrete sand. The
top layer was 0.53m thick gravel (Sabouni 2008). The manhole was installed
© ASCE
Pipelines 2015 1335
following the standard installation procedure specified by the Ontario provincial

standard specification (OPSS516) (MTO and MEA 2005)
The installation simulation was performed by following all the procedures in
the experiment. A 2D axisymmetric model was generated in ABAQUS. CAX4R
element was used to model both the concrete and the soil. In the initial step the layer
of sand and gravel beneath the manhole were generated and geostatic stress was
obtained for those two layers. Stage construction was started by adding the monobase
and sand layer with 1.2m height at the first stage. Each stage was defined in the
model as one step in which the soil layer and the manhole riser were activated. The
interaction between the soil and concrete was modeled using frictional model with
normal and tangential behavior. The friction formulation used for tangential behavior
was “Penalty” with friction coefficient of 0.5. The normal behavior was modeled with
“Hard” contact pressure over closure approach, without allowing separation after
contact. The tapered part of the manhole was ignored since it did not affect the results
significantly (Sabouni and El Naggar 2011). The geometry and mesh used for the
model are shown in Figure 1.
Rigid Wall
Gravel
Tapered
Top
1.31
0.61 Riser 4
0.61 Riser 3
Sand
7.62 0.91 Riser 2
1.2 Riser 1
1.2 Manhole
Base
0.95
0.65
Pressure cell
0.48 0.35 0.43
a. physical manhole test (unit m) b. ABAQUS model c. FEM mesh

Figure 1: Physical full scale manhole laboratory test and ABAQUS model
The boundary conditions were applied as following. The soil was restricted in
horizontal direction at the right most vertical boundary. Fixed condition was applied
to the bottom soil boundary. Axisymmetric boundary condition was applied to the
concrete and soil along the symmetry axis. The concrete and soil interface was
modeled used the interaction described in the above section. The material properties
used for the simulation of the soil layers are presented in table 1. The data was
obtained from the experimental program (Sabouni and El Naggar 2011).
© ASCE
Pipelines 2015 1336
Table 1: Soil properties

Unit Weight Modulus of Angle of internal
Soil Type
(kN/m3) Elasticity (kN/m2) Friction (°)
Sand 19 5.2x105 39
Gravel 23 6.8x105 42
Concrete damaged plasticity model (ABAQUS analysis user’s manual),

available in the ABAQUS material library, was used to simulate the manhole
structure. According to the ABAQUS theory manual, concrete damaged plasticity
constitutive theory aims to capture the effects of irreversible damage associated with
the failure mechanisms that occur in concrete and other quasi-brittle materials. The
input parameters for concrete, presented in Table 2, were determined from laboratory
tests on bare concrete cylinders and bare concrete beams (Riahi et al. 2014). For
parameters that cannot be directly determined from these tests, values were estimated
from literature. For Epoxy lining a bilinear material model was selected with
elasticity modulus of 5019 MPa, yield stress of 31 MPa and ultimate strength of 58
MPa. Material properties of the epoxy were obtained from manufacturer’s data sheet.
Table 2. Concrete Material Property

Modulus of Elasticity (GPa) 36
Poisson Ratio 0.2
Density (kg/m3) 2400
Dilation Angle (°) 38
Eccentricity 0.1
Plasticity fb0/fc0 1.16
K 0.667
Viscosity Parameter (s) 10-7
MODEL VALIDATION
Results obtained from the 2D axisymmetric simulation of the manhole in ABAQUS

are validated using analytical solutions and the results from full scale manhole
experiment. The comparisons of the results are as follows.
Soil Pressure Distribution Under Manhole Base

Pressure distribution directly under the manhole base is shown in Figure 2. As it is
shown, the pressure under the center from ABAQUS model is 74.5 kPa, which is
approximately equal to the pressure induced by the weight of the manhole, 74.79 kPa.
By moving towards the edges of the manhole, the soil pressure increases. This can be
explained by the interaction between the manhole wall and the surrounding soil. The
friction between the soil and manhole structure causes some of the weight of the soil
been carried by the manhole and increases the pressure at the edges. Comparison of
the results with PLAXIS simulation by Sabouni and El Naggar (2011) shows that the
total trend of pressure distribution under the manhole base is similar with a difference
about 20 kPa at the manhole base center. This difference is caused by the different
© ASCE
Pipelines 2015 1337
structural elements (shell element in Plaxis and solid element in ABAQUS) used in
the simulations.
Pressure Due to Weight PLAXIS ABAQUS
Distance from center(m)

-0.6 -0.4 -0.2 0 0.2 0.4 0.6
30
Vertical Stress(kPa)
50
70
90
110
130
150
Figure 2. Pressure distribution under manhole base
Lateral Soil Pressure

Lateral soil pressure acting on manhole structure is shown in Figure 3 and the result is
compared to active and at rest lateral earth pressure predicted by Rankine earth
pressure theory. The results can also be validated by the results obtained from full
scale manhole experiment. The pressure cell located at depth of 1.44m shows a
pressure equal to 10.2 kPa which is close to the results obtained from ABAQUS
simulation equal to 10.5 kPa.
Lateral Soil Pressure (kPa)

0 10 20 30 40 50
0
ABAQUS-MH
1 At Rest
Active
2 Pressure Cell
Depth (m)
3
6
Figure 3. Lateral soil pressure in depth
Moment in manhole base

The moment in the manhole base can be calculated by using clamped plate theory
(Reddy 1999). According to plate theory the radial and angular moment (Mr and Mθ)
© ASCE
Pipelines 2015 1338
in a circular clamped plate with radius “R” under uniform pressure of “p” at a
distance of “r” is calculated as:
M = 1+υ − 3+ν (1)
M = 1 + υ − 1 + 3ν (2)
The above equations indicate that the maximum moment occurs at the center
of the disk and it is equal to:
M = 1+υ (3)
Figure 4 compares bending moment in the manhole base calculated by
clamped plate theory with the results obtained from ABAQUS. The pressure used for
calculating the moment is the average pressure beneath the manhole base obtained
from simulation. The maximum bending moment which is located at the center of the
manhole base is in good agreement with the calculations from clamped plate theory.
By getting closer to the edges the results from the simulation differ from the results
from calculations which can be due to the fact that the manhole base is not acting like
a clamped plate and some movements may occur at the edges. Comparison of the
maximum moment obtained from PLAXIS simulation (Sabouni 2008) is also shown
in Figure 4. There is a 32% difference between the results which can be due to
difference in magnitude of the pressure under the manhole base from different
simulations.
PLAXIS Clamped Plate Theory ABAQUS
5
4
3
2
Moment (kN-m)
1
0
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
-1
-2
-3
-4
-5
-6
Distance from center (m)
Figure 4. Moment in manhole base
Although perfect match is not obtained from the ABAQUS simulation and
PLAXIS simulation of the existing full scale manhole experiment, regarding
analytical calculations and acceptable match in total trend of the results, the generated
model can be a good comparison base to study the effect of applying epoxy coating
inside of the manhole structure.
© ASCE
Pipelines 2015 1339
FINITE ELEMENT MODELING OF EPOXY LINED MANHOLE
The above verified concrete manhole model was modified to simulate behavior of a
deteriorated manhole with epoxy liner. Two manhole conditions are considered. In
the first one it was assumed that the concrete structure was partially deteriorated and
epoxy lining was applied inside of the manhole structure; in the second condition the
worst case scenario was presented in which the whole manhole structure was
deteriorated without any structural support and the loads are carried only by the
epoxy liner.
The first case was simulated by reducing the Young’s modulus of the concrete
to one tenth of its original value and the second case was simulated by replacing the
concrete material of the manhole with soil. In both cases a layer of 6 mm epoxy liner
was applied to inside of the deteriorated manhole. The interaction between the
coating and concrete was considered as frictional behavior with a friction coefficient
of 0.5 in tangential behavior and hard contact in normal behavior.
Results And Discussion

By replacing the concrete in manhole structure with deteriorated concrete and adding
epoxy liner, lateral deformation of the manhole increases as expected. Figure 5
compares lateral deformation of epoxy liner in two cases. Case I refers to the situation
of the deteriorated concrete with one tenth of concrete stiffness and Case II refers to
the situation of replacing the concrete structure with soil. It should be mentioned that
in these simulations the only load acting on the manhole was lateral pressure of
surrounding soil. The other loads such as water pressure and traffic are not considered
in this study. Radial deformation of the manhole at different depths is also calculated
by using theory of elasticity for pipe under uniform pressure (Saada 1993).
According to the theory of elasticity, radial deformation at any distance of “r” from
center in a hollow cylinder with inside diameter of “a” and outside diameter of “b”
under peripheral pressure of “P” can be obtained using following equation:
u = r− (4)
The pressure used in the above equation is obtained from the ABAQUS
outputs at the chosen depth. The results from analytical calculation are in good
agreement with results obtained from ABAQUS simulation.

The lateral soil pressure profiles of the two cases are shown in Figure 6. It
can be seen that soil pressure decreases by reducing elastic modulus of the concrete in
the case of deteriorated manhole. However, the soil pressure is still above active
pressure state and no plastic strain is observed in the model. The soil pressure profile
of the completely deteriorated manhole moves beyond active state which indicates an
active state soil failure. The soil failure can also be observed in plastic strain
development in the soil which is not shown here.
© ASCE
Pipelines 2015 1340
Radial Deformation(mm)
-0.1 -0.08 -0.06 -0.04 -0.02 0
0
Theory of Elasticity
1
Case I
Case II 2
Depth(m)
3
6
Figure 5. Deformation of the Epoxy coating
Lateral Pressure (kPa)

0 10 20 30 40 50
0
Sound Manhole
1 Case I
Case II
2 At Rest
Active
Depth(m)
7
Figure 6. Lateral Pressure in depth
Deformation of the epoxy liner base in case II is shown in Figure 7. The

maximum deformation is located at a distance of 0.52m from the center of the
manhole. The maximum observed strain is 1.6x10-4 which is smaller than the plastic
strain. Soil Pressure distribution under the epoxy liner base (Case II) for this case is
also shown in Figure 8. The pressure beneath the manhole in this case is decreased
significantly as compared to new concrete manhole. As the self-weight of the epoxy
liner is small, the soil pressure is mainly caused by the load transfer from the soil due
side friction.
© ASCE
Pipelines 2015 1341
Figure 7. Deformation of the epoxy coating at the base of the manhole for Case
II (m)
Distance from center(m)
-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
0
Vertical Stress(kPa)
10
15
20
25
Figure 8. Pressure distribution under manhole base for Case II
CONCLUSION
A finite element model of manhole structure and surrounding soil was generated in
ABAQUS and it was validated by the means of an existing full scale manhole
structure and analytical calculations. The calibrated model was used to study the
behavior of epoxy lining inside of the deteriorated manhole structure. The soil
pressure along the manhole depth is close to at-rest soil pressure. The soil pressure
beneath manhole is close to in-situ vertical stress in the soil. Partially deteriorated
manhole with one tenth of remaining modulus can be repaired with the 6mm epoxy
liner. The manhole with epoxy liner only fails due to active soil pressure failure. The
developed model can be used to perform further analysis on behavior of the epoxy
liner under different loading conditions and with different percentage of deterioration
in manhole structure. More studies are needed for exact soil pressure distribution
under concrete manhole base.
REFERENCES
ABAQUS analysis user's manual, Version 6.12 (2012), Dassault Systèmes.
© ASCE
Pipelines 2015 1342
Ahn, N., Park, D.K., Lee, J., and Lee, M.K., (2009), Structural Test of Precast
Polymer Concrete. Wiley Periodicals, Inc., Journal of Applied Polymer
Science, Vol. 114, 1370 – 1376.
IKT, (2012), Rehabilitation of waste water manholes: Large scale tests and in-situ
studies.
J. Matthews, W. Condit, R. Wensink, G. Lewis. (2012), Performance evaluation of
innovative water main rehabilitation cured-in-place pipe lining product in
Cleveland, OH, U.S. Environmental Protection Agency.
Ministry of Transportation of Ontario (MTO) and Municipal Engineers Association
(MEA). 2005. Ontario Provincial Standard Specifications OPSS 516,
Construction specification for excavating, backfilling, and compacting
maintenance hole, catch basins, ditch inlets and valve chambers. Ministry of
Transportation of Ontario and Municipal Engineers Association, Ontario,
Canada.
Reddy, J. N., (1999). Theory and analysis of elastic plat, USA, Taylor and Francis.
Redner, J.A., Hsi, R.P., Esfandi, E.J., Sydney, R., Jones, R.M., Won, D., and
Andraska, J., (1986), Evaluation of Protective Coatings for Concrete. Final
Report, December 2004, County Sanitation Districts of Los Angeles County,
Whittier, CA. (Updated: 2004).
Reem Sabouni. (2008) Experimental investigation and numerical modeling of
circular precast concrete manholes. Ph.D. Thesis.
Reem Sabouni, M.H. El Naggar. (2011), Circular precast concrete manholes:
experimental investigation. NCR.
Reem Sabouni, M.H. El Naggar. (2011), Circular precast concrete manholes:
numerical modeling. NCR.
Riahi, E., Yu, X., Najafi, M., and Sever, F. (2014), Evaluation of the Structural
Performance of Epoxy Linings for Manhole Rehabilitation Using Laboratory
Testing and FEM Simulations. Pipelines 2014: pp. 1334-1344.
Saada, Adel S. (1993), Elasticity theory and application. 2nd Edition, Malabar, Fla. :
Krieger.
V. Firat Sever, Mohammad Najafi, Abhay Jain. (2013), No-Dig Manhole
Rehabilitation Knowledge Gaps. Pipe lines, ASCE.
Matthews, J., Condit, W., Wensink, R., Lewis, G. (2012). "Performance evaluation of
innovative water main rehabilitation cured-in-place pipe lining product in
Cleveland, OH." U.S. Environmental Protection Agency (USEPA), Edison,
NJ, EPA/600/R-12/012.
Sabouni, R. (2008). "Experimental Investigation And Numerical Modeling Of
Circular Precast Concrete Manholes." Ph.D., The University of Western
Ontario, London, Ontario, Canada.
Sabouni, R., and El Naggar, M. H. (2011). "Circular precast concrete manholes:
experimental investigation." Canadian Journal of Civil Engineering, 38(3),
319-330.
Sabouni, R., and El Naggar, M. H. (2011). "Circular precast concrete manholes:
numerical modeling." Canadian Journal of Civil Engineering, 38(8), 909-920.
© ASCE
Pipelines 2015 1343
Comparative Analysis of Geopolymer Technology for Sewer System

Rehabilitation
J. R. Royer, Ph.D.1; and Dan D. Koo, Ph.D., P.E., M.ASCE2
1
Milliken Infrastructure Solutions, LLC, M-153, 920 Milliken Rd., Spartanburg, SC
29303. E-mail: joe.royer@milliken.com
2
Purdue School of Engineering and Technology, Indian University Purdue University
Indianapolis (IUPUI), Department of Engineering Technology - Construction
Engineering and Management Technology, 799 W Michigan St. ET 314J,
Indianapolis, IN 46202-5160. E-mail: dankoo@iupui.edu
Abstract
Asset owners and engineers throughout the U.S. and the world are in search of
cost-effective and environmentally friendly solutionsthat serve infrastructure issues.
One of the most critical areas of concern is wastewater piping and structures. It is
well known that corrugated metal pipes used in storm-water structures are corroding
and microbial induced corrosion of sanitary sewers of various materials results in
structural concern.Geopolymers have long been known to provide enhanced physical
performance to traditional cementitious binders with the addedadvantages of
significantly reduced greenhouse emissions and superior chemical resistance.
Geopolymers are ceramic polymertechnology that creates a chemical material similar
to natural stone that is superior to traditional Portland cement and shotcretematerials.
However, they have not generally been contractor-friendly. This paper reviews a
geopolymer mortar system that has been used in the U.S. since 2011 and is becoming
a preferred solution fortrenchless rehabilitation. The system is spray cast either by
rotary nozzle or via traditional shotcrete delivery systems inside ofexisting structures
to create whole new structures which do not depend on the existing structure, just
using it as formwork. This paperdiscusses competitive advantages over other
trenchless repair solutions such as spiral wound, slip-lining and CIPP through
specificcase studies including a corrugated metal storm drain rehabilitation in
Hidalgo, Texas along with the repair of a stone sewer system inCincinnati, Ohio.
1.0 INTRODUCTION
As the state of infrastructure around the world decays, more cost effective
solutions to repair large diameter pipe systems are required. Typical dig and replace
technology is often not practical as in most urban areas these degrading pipes are
located directly under other critical infrastructure such as major roadways, buildings,
or other assets. As the diameter of these pipes become larger (>48 inches), the cost of
many of the traditional trenchless technologies becomes exponentially more
expensive and often requires significant excavation around access points that present
additional issues related to community disturbance, traffic control, noise and general
© ASCE
Pipelines 2015 1344
disruption. For example, if a 48-inch diameter sewer pipe were located in the center
of town and a standard 30 or 36 inch manhole was the access point, in order to
perform a CIPP (Cured-In-Place-Pipe) repair it would be necessary to excavate an
access hole of at least the 48 inch diameter. While other techniques such as slip-
lining would require even greater excavation for an access hole to install new liners.
Additionally, with many of the standard so called trenchless repair technologies other
issues related to either the shape (round, arched, elliptical) or the layout (straight,
curved, bends of various radius) can make these repair technologies unpractical
(Buczala,1990) (Osborn, 2010).
Over the last decade additional trenchless technologies have been developed
to help fill the need for larger diameter pipe repairs at effective costs with little or no
excavation requirements and minimal community disruption. One such technological
advance is the use of centrifugally cast geopolymer mortars to create a new pipe
inside the existing old pipe (Henning, 2012). This techniques allows for a
cementitious pipe to be created within the existing structure, using the existing pipe as
a form, and can be designed such that a new fully structural pipe is created. The
flexibility of the technique allows for pipes of all shapes and layouts to be repaired
either using automated mechanical casting or manually controlled material placement.
The equipment necessary can easily fit down standard manholes and all excavation
can be avoided if there are access points at least every 800 linear ft.
The benefit of geopolymer mortars as compared to traditional Portland cement
(OPC) materials is detailed in the following discussion. Additionally, case studies are
included.
2.0 GEOPOLYMERS
Geopolymer is a term originally coined by French researcher Joseph
Davidovits to describe a class of “cement” formed from aluminosilicates. While
traditional Portland cement relies on the hydration of calcium silicates, geopolymers
form by the condensation of aluminosilicates. The kinetics and thermodynamics of
geopolymer networks are driven by covalent bond formation between tetravalent
silicon and trivalent aluminum. The molar ratio of these key components along with
sodium, potassium, and calcium have been shown to affect set-time, compressive
strength, bond strength, shrinkage, and other desired properties. In various parts of the
world, this type of material is also industrially known as “alkali-activated cement” or
“inorganic polymer concrete” (Davidovits, 2011). Geopolymers provide comparable
or better performance to traditional cementitious binders in terms of physical
properties such as compressive or tensile strengths (Bell, 2008) (Buchwald, 2006) but
with the added advantages of significantly reduced greenhouse emissions, increased
fire and chemical resistance, and reduced water utilization (Alonso, 2001). The use
of geopolymers in modern industrial applications is becoming increasingly popular
based on both their intrinsic environmental as well as performance benefits.
Historically, trial applications of geopolymers were first used in some concrete
applications by Glukhovsky and co-workers in the Soviet Union post WWII; the
© ASCE
Pipelines 2015 1345
geopolymer was then known as “soil cements” (Davidovits, 2011). Figure 1 shows a
typical aluminosilicate structure that is common among many geopolymer materials.
Figure 1: Example Aluminosilicate Molecular Geopolymer Structure
The structure of a geopolymer is a cross-linked inorganic polymer network

consisting of covalent bonds between Aluminum, Silicon and Oxygen molecules that
form an alumniosilicate back bone with associated metal ions. While any specific
geopolymer structure, such as the one represented here in Figure 1, will be
significantly more complicated based on the chemical make-up of the starting raw
materials, the generic structure shown provides an excellent representation of how a
geopolymer network is constructed. In contrast, OPC is a hydrated complex of small
molecules that are not covalently bonded but rather associated. This is shown in a
simplified structure in Figure 2. OPC itself is sufficiently complex that the structure
shown in Figure 2 is only a basic representation of the molecules but no long chain
covalently bonded backbone or network structure exists in standard cementitious
materials.
H2 O O Ca(OH)2 O H2 O
O
Ca2+ O O Ca2+
Si Si
H2O
H2O O O H2O
Figure 2: Simplified Example Molecular Structure of Hydrated Portland Cement (OPC)
3.0 GEOSPRAY GEOPOLYMER MORTAR

A specific example of a formulated geopolymer repair mortar is GeoSpray
produced by Milliken Infrastructure Solutions, LLC. It is formulated to meet all the
physical and chemical requirements for rehabilitating sewer and storm water
structures. Water is added to the geopolymer at the job site where it can simply be
© ASCE
Pipelines 2015 1346
centrifugally sprayed inside an existing structure that has been properly prepared.
The exact formulation of most products are considered trade secrets, but generally
speaking, geopolymers contains a mixture of the standard materials that are used in
the production of calcium-aluminosilicates. Other components include, but are not
limited to, blast furnace slag, reactive silicas, metal oxides, mine tailings, coal fly ash,
metakaolin, calcinated shale, natural pozzolans, and natural/processed zeolites.
Additional bio-based admixtures are included in the formulation in order to allow the
composite material to set-up quickly and easily hydrate with a single addition of
water. The “just add water” aspect of this particular geopolymer system has been
specifically developed to avoid the typical alkaline activation mechanisms and order
of addition complexities of traditional geopolymers which have limited significantly
the ability of most contractors and asset owners from using geopolymers
commercially. A summary of the physical properties of GeoSpray as compared to
conventional concrete pipe repair mortars is included in Table 1.
Table 1. Typical properties of GeoSpray compared to Conventional

Cement Based Repair Mortars
With this type of repair mortar the entire system is contained within original powder
formulation, allowing a single step addition. It is common for these materials to be
pumped up to 500ft within a pipe and still be centrifugally cast without clogging or
damaging nozzle performance. To achieve this standard of performance, traditional
cement or geopolymer formulations would require much higher water ratios which
would degrade their ultimate strength and require a much thicker final product during
the installation to meet the flexural strength requirements of the rehabilitation.
© ASCE
Pipelines 2015 1347
4.0 GEOPOLYMER ADVANTAGES
4.1 Cold Joints
On real world construction sites, unexpected and unanticipated circumstances

can result in delays or work stoppages. Additionally, many job sites can be subject
to restricted work hours due to local traffic issues or community related ordinance.
When working with the placement cement, these types of work stoppages or delays
can result in the formation of a cold joint. A cold joint is an undesired discontinuity
between two layers of concrete. A cold joint occurs due to the inability of a freshly
poured wet cement to intermingle and bind with an already hardened cement. A
typical cold joint in a poured structure is shown in Figure 3.
Figure 3: Typical Cement Cold Joint

Cold joints can resulting in multiple problems ranging from minor to
catastrophic. The spectrum of resulting issues include: minor cosmetic visual
differences between layers, possible moisture intrusion into the joint resulting in
degradation from environmental conditions, and areas of significantly compromised
strength within a structure. When water is mixed with Portland cement (OPC) the
cement reacts with the water to form a hydrate allowing the cement to harden around
aggregates and form concrete. The chemistry of the reaction uses a hydration
mechanism to create a hardened solid phase structure. However, once the hydration
is complete and the structure is solid, it will not physically or chemically intermingle
with additional cement.
Geopolymers undergo a completely different set of reactions classified as
condensation. This process creates large polymer molecules that react to form large
chain molecules that create the solid structure. When a hardened geopolymer is
contacted with a freshly poured geopolymer mixture the polymer molecules from the
hardened geopolymer are still active and will chemically bond with the new mixture
preventing a cold joint from forming.
© ASCE
Pipelines 2015 1348
To demonstrate the superior properties of geopolymer mortar as compared to

OPC materials with respect to cold joints, a series of compression test were conducted
using 2 inch by 4 inch cylinders using a commercial geopolymer formulation
On the first day of the experiment, full cylinders of both geopolymer and
commercially available competitive material based on OPC, both designed for use in
structural pipe repair, were poured. In addition to the full cylinders, ½ pours of the
same size were produced with both materials and vibrated on a slant to create an
approximately 45º angle in the lower portion of the cylinder (as shown in Figure 4).
A second pour atop the first pour (of the same material) was then done with intervals
of 1, 7, 14 & 28 days. All samples where then compression tested according to
ASTM C39.
Figure 4: Schematic Illustration of Cold Joint Compression Experiment
For all combinations, the full cylinders poured on day 0 have no joints and
break in a standard compression failure throughout the cylinder. For the geopolymer
samples with the 45º joint, compression failure mode is the same as the full cylinder
even when 28 days have elapsed between pours. The leading OPC competitive
material breaks along the cold joint in all of the test intervals, showing that the cold
joint formed in the OPC between the pours is the weakest part of the structure.
Detailed images of the experiments are shown in Figure 5.
(b)
(a)
Figure 5: (a) Geopolymer samples showing compression failures located away

from the joint (b) OPC samples with compression failure located at the joint.
© ASCE
Pipelines 2015 1349
4.2 Chemical Resistance
In sanitary sewers and other wastewater environments, the general corrosion

mechanism of cementitious based materials is well known and widely documented. It
is often referred to as Microbial Induced Corrosion or (MIC). The process of MIC
involves a 3 step mechanism (shown schematically in Figure 6):
• First, hydrogen sulfide gas (H2S), commonly referred to as sewer gas, is
released by the reduction of sulfates in the sewer effluent from anaerobic
bacteria – generally living in a “slime layer” below the water line.
• Secondly, sulfuric acid (H2SO4) is formed on exposed surfaces through the
oxidation of H2S by aerobic Thiobacillus bacteria.
• Finally, the sulfuric acid reacts most often with calcium hydroxide Ca(OH)2
found in many cements to form gypsum CaSO4·2H2O which is water soluble
and will wash away.
Figure 6: Schematic of the Chemical Processes associated with MIC Corrosion

The chemical make-up of geopolymers makes them inherently more acid resistant
to the MIC mechanism found in many sewer environments. Geopolymers (dependent
on the exact formulation) will contain greatly reduced concentrations of Ca(OH)2
(calcium hydroxide) essentially preventing the acid corrosion mechanism found in
many typical cements. Chemical resistant studies were performed following the
procedures of ASTM-C267. Geopolymer sample cubes were cast and allowed to cure
for 28 days before being soaked in both water and 7% sulfuric acid (pH 0.9). OPC
cubes were also cast and soaked as representative samples for standard reinforced
concrete pipes commonly found in sanitary sewer systems.
Samples were measured for weight and dimensional changes after soaking for
1, 7, 14, 28, 56 and 84 days. 3 samples of the materials were soaked and tested, and
the solution volume relative to the cubes was held constant. The chemical solutions
were refreshed on day 14, 28, and 56. Geopolymer samples showed only slight loss
of mass and signs of surface corrosion through the 84 days exposure to 7% H2SO4
(sulfuric acid), while the Portland cement samples lost more than 50% of their weight
© ASCE
Pipelines 2015 1350
over the same time period. Figure 7 shows samples cubes before and after 84 days of
soaking exposure.
Figure 8 shows the effect of the 7% sulfuric acid on weight of the geopolymer and
OPC cubes over the same time period. The results of weight are normalized to the
percentage of weight change of samples soaked in water to account for the absorption
of water. Through the 84 days exposure the geopolymer corrosion was
approximately 1/5th of the standard OPC material.
When tested under the ASTM C-267 protocol against aggressively corrosive
7% sulfuric acid (pH 0.9), the geopolymer showed only approximately 5-7% (note:
the samples are compared to water soaked materials and the below 0 starting point is
due to gel formation of H2SO4 and not true weight loss)weight loss and slight surface
corrosion compared to the >50% weight loss observed in OPC samples that reflect
concrete sewers in use today. Where concrete pipes and structures exhibit the effects
of microbial induced corrosion, geopolymers should result in significant resistance
improvement over OPC.
Geopolymer
Portland
Figure 8:Comparative weight change over time for soaked cubes
© ASCE
Pipelines 2015 1351
5.0 CASE STUDIES
5.1 McALLEN, Texas (114 inch CMP Storm Drain Rehabilitation)
In the rapidly growing Texas border town of McAllen, dealing with storm
water runoff is a challenge. The weather fluctuates quite rapidly, and ensuring that
storm water infrastructure is capable of dealing with large amounts of water quickly
is of paramount concern. As the population of the community has nearly tripled over
the past two decades some of their storm water infrastructure has presented an
ongoing challenge.
Such a problem was the Rado Storm Drain, located within one mile of the Rio
Grande river. The storm drain consists of 2 side by side 114 inch corrugated metal
pipes (CMP) each over 2200 linear feet in length. The pipes had issues ever since
they were installed and have been repaired in various sections over the past decade
with a non-structural shotcrete and a bitumen coating. These attempted repairs were
done over short segments of the pipe, but large scale separation of the joints along
with water infiltration continued to be major concerns.
The local municipalities had experience with non-structural repairs in the past
that had not been successful on this particular application. They investigated several
repair options including Cured-in-Place-Pipe (CIPP), slip-lining, and geopolymers.
Both the CIPP and slip-lining solutions were significantly higher cost with additional
complications due to the large diameter. The community decided to specify a
cementitious lining as their structural application. In addition, because the county was
a member of HGAC Buy (Houston-Galveston Area Council), a competitively bid
contract organization with members in 48 states, they were able to specify
geopolymers and avoid the costly process of bidding the project themselves. Inland
Pipe Rehab, LLC using their Ecocast™ process installed GeoSpray as the repair
solution for this project in the spring of 2014.
Because there were two side-by-side pipe sections, only one was repaired at a
time and all of the flow was diverted to the other section. Each pipe section was
cleaned and inspected for joint failure, cracking and infiltration. Stopping water
infiltration was the primary challenge of the project and required meticulous
preparation. These issues were addressed with hand repair to ensure that all the
infiltration of water was stopped and a continuous surface for the application of the
geopolymer mortar was created. Once these issues were tackled, 150 to 300 foot
sections were then sprayed with the final engineering designed thickness of 1.5 inches
of geopolymer. The ability to apply a 1.5 inch thick layer in a single spray pass saved
both time and cost for the asset owner. During most days of operation, the contractor
was able to apply between 20,000 and 40,000 lbs of geopolymer in a single run
within the pipe, creating a truly monolithic structure.
The use of geopolymer to create a new pipe within the existing CMP structure
that existed was completed on time and under budget. The new pipe is ready to
handle the unpredictable storms of southern Texas for years to come. Figure 9 shows
© ASCE
Pipelines 2015 1352
a series of images from the job site, this includes upstream the entrance to the pipes, a
view of the joint separation on the shotcrete repaired structure, the geopolymer
application and the finished pipe.
Figure 9: Images from the jobsite for the Rado Storm Drain in McAllen, TX
5.2 Cincinnati, Ohio (36 inch Sanitary Sewer Rehabilitation)

During an assessment of their combined sewer system, the Metropolitan
Sewer District of Greater Cincinnati (MSD) discovered that an 800 ft section of
stacked stone and brick sewer line, situated in an area with active natural springs, was
badly leaking. Records indicated that the old stone pipe was likely installed in phases
between 1870 and 1890.
This specific section of stone pipe presented a host of unique challenges. One
section of the pipe was on a 26° slope that lead to a river outfall. A detailed
inspection revealed that the pipe had both round and arched cross sections with two
different diameters. A first 500 ft section had an inner diameter of approximately 60
inches, and a second 300ft section had a 36 inch diameter section. Access to the pipe
itself was also a challenge. The old sewer ran beneath three sets of road crossings, all
at different elevations. It was also located near the University of Cincinnati campus,
so minimizing surface disruption was a high priority.
© ASCE
Pipelines 2015 1353
In reviewing their options to address the deteriorated pipe, the MSD quickly realized
that replacement was not a viable option because of the pipe’s depth and location.
When considering alternative trenchless methods a number of options were evaluated
including Cured-In-Place –Pipe (CIPP) and Slip-lining. CIPP was not a viable
solution due to the variation of the pipe’s different shapes and sizes, the steep slope
and the rough protruding stone. Slip-lining of the pipe was also a poor option for the
same reasons, would have required digging several large access pits, and resulted in
significantly reduced flow capacity. In the end, the MSD chose to apply a geopolymer
lining that would be both hand and machine sprayed to create a new structure lining,
repair the leaking, and return the pipe to its original shape.
Construction on the site began in March 2012.The first task was to clean the
stone pipe with a high pressure wash and then to use a hand spray application of
geopolymer to stop the leaking and to stabilize the existing stone structure. One
critical advantage of the technique of centrifugally spraying a geoplymer liner is that
the equipment foot print can be limited to the size of approximately two 24 ft box
trucks and spraying can occur more than 400 ft from the actual mixing location . This
allowed the crew to avoid any traffic disruption.
Once the initial hand spray was complete, a mechanical sled system was used
to apply the final coats of the geopolymer and arrive at the engineer’s required
thickness allowing for a new structural pipe to be built within the existing pipe. It is
interesting to note that not only did the pipe itself present many challenges, the
weather also was a key consideration. The temperature above ground during the
installation period ranged from just above freezing to highs in the mid 70s (ºF). These
large temperature swings are no problem for geopolymer systems, helping keep the
project on schedule. From start to finish the full project was completed in under six
weeks, ahead of schedule and on budget. The flexibility of geopolymers makes them
an excellent choice for the toughest sewer repairs. Figure 10 shows a series of images
from the job site.
6.0 CONCLUSIONS
Geopolymer mortar repair systems have been developed to be a cost effective
alternative to other trenchless repair systems for large diameter pipes. Geopolymers
have advantages over traditional OPC systems relating to the chemistry of the
materials and how they are reacted that include (a) lower CO2 footprints, (b) reduced
tendency for cold joints and (c) enhanced chemical resistance. Multiple case studies
have been shown where structural pipe repairs were designed and completed for both
storm drain and sanitary sewer applications.
© ASCE
Pipelines 2015 1354
Figure 10: Images from the jobsite for the MSD in Cincinnati Ohio
REFERENCES
Alonso, S., Palomo A. (2001) “Calorimetric study of alkaline activation of calcium
hydroxide-metakaolin solid mixtures.” Cement and Concrete Research, 21,
25-30.
Buchwals, A. (2006) “What are geopolymer? Current state of research and
technology the opportunities they offer, and their significance for the precast
industry.” Betonwerk + Fertigteil-Technik, 72, 42-49.
Buczala, G. S. Cassady, M. J. (1990). Buried Plastic Pipe Technology: (STP 1093).
ASTM International.
Davidovits, J. (2011) “Geopolymer Chemistry and Applications” 3rd Edition,
Geopolymer Institute.
Henning, S., Vellano, M. (2012) “Understanding the Science Behind Advanced
Geopolymer Mortar Lining Systems.” Trenchless Technology, April, 56.
Osborn, L. (2010). Trenchless Renewal of Culverts and Storm Sewers. American
Society of Civil Engineers (ASCE).
© ASCE
Pipelines 2015 1355
An Evaluation of Trenchless Point Repair Solutions for Pipes of

Varying Inner Diameter and Offset Joints
Rudy Ellgass1; Jey K. Jeyapalan2; Brian Gipson3; Mark Biesalski4;

Wayne Miles5; Steve Leffler6; John Kurdziel7; and Mark Bruce8
1
Rausch USA LLC, 4757 Innovation Way, Chambersburg, PA 17201. E-mail:
rellgass@rauschusa.com
2
Civic Enterprises LLC, 9 Sundance Rd., New Milford, CT 06776-3840. E-mail:
jkjeyapalan@earthlink.net
3
PipeCheck LLC, 325R Emmett Ave., Bowling Green, KY 42101. E-mail:
brian@pipecheck.net
4
Uhrig Kanaltechnik GmbH, Am Roten Kreuz 2, 78187 Geisingen, Germany. E-mail:
biesalski@uhrig-bau.de
5
CDMSmith, 5400 Glenwood Ave., Raleigh, NC 27612. E-mail:
milessw@cdmsmith.com
6
Norditube USA, 975 Nutcracker Dr., Redmond, OR 97756. E-mail:
steveleffler@earthlink.net
7
Advanced Drainage Systems Inc., 4640 Trueman Blvd., Hilliard, OH 43026. E-mail:
john.kurdziel@ads-pipe.com
8
Hydromax USA, 344 Inderrieden Rd., Chandler, IN 47610. E-mail:
mark.bruce@hydromaxusa.com
Abstract
There are numerous point repair solutions in the pipeline industry. These can be
broadly grouped into internal joint seals, mechanical sleeves, CIPP short liners and
other. The pros and cons of available methods are not always clearly understood by
the owners, engineers and installers. Some widely known brand name manufacturers
of joint seals and mechanical sleeves are: WEKO-SEAL®, In-Weg seal, Quick Lock,
Link-Pipe, LMK, HydraTite, Amex 10, and other. These products can be used for
storm, potable water, wastewater and industrial pipes, conduits and drainage culverts.
Materials for these types of repairs are specified based on anticipated exposure
conditions after installation. When owners are in need of an economical solution for
the maintenance of pipelines, the point repair solutions present an attractive
alternative to more extensive repair or replace options. Quick Lock that meets ASTM
F3110 is new in North America although it has been used for over 20 years in other
countries. This paper introduces its features that no other point repair solutions offer.
In addition, given that the suitability of the technology for a particular application
shall always be jointly decided by the owner, the engineer and the installer, this paper
provides an evaluation of available methods during this decision making process.
INTRODUCTION
Based on the trends in the market studies the second author has performed for a
number of clients over the past three decades, more than 3,000,000 mechanical
© ASCE
Pipelines 2015 1356
sleeves and internal joint seals have been installed globally for repairing the following
types of defects in pipe: longitudinal, radial and circumferential cracks,
fragmentation, leaking joints, displacement or joint misalignment, closing or sealing
unused laterals, corrosion, spalling, wear, leaks in the barrel of the pipe, deformation
in the pipe and root penetration. Repairs can be made to vitrified clay, concrete,
reinforced concrete, plastics, glass reinforced plastics, cast iron, ductile iron and steel.
When owners are in need of an economical solution for the maintenance of pipelines,
the point repair solutions present an attractive alternative to more extensive repair or
replace options.
Mechanical trenchless repair sleeves with a locking gear mechanism for pipes of
varying inner diameter and offset joints in the range of 6 to 72 in (150 to 1800 mm)
offer many advantages over relining the entire pipe or using other point repair
technologies. These sleeves can be used for storm, potable water, wastewater and
industrial pipes, conduits and drainage culverts. Given that each owner retains the
right to choose the test protocol to verify the efficacy of these sleeves to provide a
leak free repair, the producers have not provided any test protocol of their own -
except for internal joint seals. The maximum internal pressure this sleeve can carry
depends on the diameter and the wall thickness, ranging from 145 to 217 psi (1.0 to
1.5 MPa); the external pressure cannot exceed 21.7 psi (0.15 MPa) as hydrostatic
groundwater pressure acts in a manner which causes the integrity of the compression
seal to be compromised. It should be noted, however, that it is the pressure
differential that must be determined. The external pressure must exceed the internal
pressure by 21.7 psi to result in leakage; ultimately, the worst case scenario is when
the pipe is empty with high groundwater.
More than 200,000 Quick Lock sleeves have been installed globally for repairing the
following types of defects in pipe: longitudinal, radial and circumferential cracks,
fragmentation, leaking joints, displacement or joint misalignment, closing or sealing
unused laterals, corrosion, spalling, wear, leaks in the barrel of the pipe, deformation
in the pipe and root penetration. There are no limitations on the diameters of the
laterals that can be sealed. The degree of pipe deformation that can be repaired is
dependent on the minimum and maximum diameters for which the sleeve is
applicable, up to 5% deflection is acceptable. Repairs can be made of vitrified clay,
concrete, reinforced concrete, plastics, glass reinforced plastics, cast iron, ductile iron
and steel. The suitability of the technology for a particular application shall be jointly
decided by the owner, the engineer and the installer. For example, all materials in
contact with potable water are certified to meet National Sanitation
Foundation/American National Standards Institute (NSF/ANSI) 61/372. When the
materials for the mechanical sleeves or joint seals are selected to meet the project
demands and installed in accordance with the ASTM standard F3110-14, the
renovation extends over a predetermined length of the host pipe as a continuous, tight
fitting, leak free, and corrosion resistant repair.
There are other mechanical repair sleeves and joint seals and the details are presented
in later sections. Then we have short sectional liners. The extent to which these short
© ASCE
Pipelines 2015 1357
liners could offer structural augmentation for withstanding internal and external
pressures is lower compared to that from the use of mechanical or internal seals,
because of the dramatic difference in the engineering behavior of the materials used
between the two types of methods. These are also cumbersome to use and time
consuming to set up, prepare and install. Health of the workers installing them is also
more at risk.
COMPONENTS FORMING THE SLEEVE
A Stainless steel sleeve with flared end, faces the direction of flow and improves the
hydrodynamics, prevents solids from depositing, and increases jetting resistance
shown as (1) in Figures 1 and 3. Flared ends are not used, however, for potable water
applications; Metal overlap is for expanding to the pipe wall shown as (2) in Figure 1;
The locks that keep the sleeve expanded run along the toothed strip shown as (3) in
Figure 1; The lock is a small set of gears that only moves in one direction, thus
keeping the sleeve expanded shown as (4) in Figures 1 and 3. The locking gears are
also shown in Figure 4. There are three sprockets per gear lock. Two of the sprockets
in the gear ride in the corresponding “teeth” in the sleeve. The third sprocket is the
lock. It allows the other two gears to only move in a forward direction. The gear and
the shield are all of the same material as the rest of the sleeve. Furthermore, the gears
are protected by a cover to prevent snagging of waste, build up of sludge or sediment,
and protect it from a root cutter and cleaning nozzle. These features have a proven
track record of over 20 years. Adhesive tape and plastic pin are put on at the factory
to protect the sleeve during transport and prevent it from unrolling shown as (5 and 6)
in Figure 1; circumferential seals are formed from the rubber being compressed
against the host pipe.
For single installations, the damaged section must always be between the sealing
knobs shown in (7) in Figures 2 and 3.There is a trimming line marked in the rubber
jacket. It shows the installer where to cut off the projecting rubber end shown as (8)
in Figure 2. For serial installations, the projecting rubber end is not cut off and acts as
a seal between the sleeves installed shown as (9) in Figures 2 and 3. There are designs
with fewer sprockets and made of materials other than that of the body of the sleeve
with a record of about a year – somewhat experimental.
Figure 1. Stainless steel sleeve (Uhrig, 2015)
© ASCE
Pipelines 2015 1358
Figure 2. Locking gear mechanism and EPDM rubber seals (Uhrig, 2015)
Figure 3. Cross sectional view of the sleeve and the EPDM seals (Uhrig, 2015)
Figure 4. Locking gear mechanism (Uhrig, 2015)
INSPECTION AND PREPARATION OF DAMAGED PIPE
Before using a point repair solution, the pipe must be inspected to ascertain whether it
can be repaired with the system. Cleaning and inspection shall be as per National
© ASCE
Pipelines 2015 1359
Association for Sewer Service Companies (NASSCO) guidelines for sewage pipes.
For other pipes, cleaning and inspection methods that are acceptable to the authority
shall be chosen by the installer. There must be at least one access via a manhole or an
inspection chamber. The manhole must have a diameter of at least 24 in (600 mm) so
that the camera/packer system can be inserted. The interior of the pipeline shall be
carefully inspected to determine the location of any condition that shall prevent both
deployment and proper installation, such as large joint offsets, roots, and collapsed or
crushed pipe. The pipe to be repaired must always be cleaned with a high-pressure jet,
a drag scraper or an equal approved by the owner, the owner’s representative or the
manufacturer, before using the sleeve.
MATERIALS
The sleeve is made of high grade 316 or 316L stainless steel per ASTM 240 (2015)
and ASTM 666 (2010). There are three types of sleeves available: non-flared, one
end-flared and both ends flared. Non-flared sleeves are mainly used for potable water
applications or serial installations, followed by one end-flared sleeves. The both ends
flared type is used for single installations only. Ethylene Propylene Diene Monomer
(EPDM) seal or silicon rubber seal (for potable water with chlorine or fluoride
treatment or even EPDM washed with peroxide) meets the physical property
requirements for elastomeric materials used in cold water supply, drainage, sewerage
and rainwater systems for Type WC in Table 2 of EN 681-1 Hardness Category of 40.
Nitrile Butadiene Rubber (NBR) and similar elastomers may be appropriate for other
effluents or where contaminated soils are surrounding the exterior of the pipe.
INSTALLATION
In pipes up to 32 inches in diameter (800 mm) the sleeves are installed using an
inflatable packer on wheels. The packer is usually connected to the camera or robot
via a bracket and a hollow link bar. For repairs more than 20 ft (6 m) into the line the
sleeve is installed most efficiently and accurately when the packer is pushed or pulled
and positioned by using a crawler camera, equipped with an accurate distance
counter. The sleeve is usually positioned on the packer while in the manhole. Once
the packer/sleeve is in the correct position over the damaged section, the actual
installation takes only about 1 to 2 minutes. If the installer is using a packer equipped
with a laser, position the sleeve so that the laser beam is reflected both on the edge of
the sleeve and the host pipe. Installation in pipe sizes larger than 32 inches (800 mm)
is done manually. Depending on the quantity of flow present, bypassing in
accordance with the authority’s requirements may be necessary.
OTHER MECHANICAL SLEEVES
a) Link-Pipe asserts that if during routine inspection damaged pipe is discovered,

it takes around 20 minutes to have it repaired using Link-Pipe Grouting Sleeve or
SewerSealer, shown in Figure 5. These sleeves must be installed before the resin is
cured, usually within 20 to 25 minutes. Spot repairs can be installed in pipe diameters
© ASCE
Pipelines 2015 1360
of 6-54 inches. Every Grouting Sleeve, SewerSealer Sleeve and Link-Pipe (PVC
Sleeve) carries a manufacture 10-year Limited Warranty and is evaluated to give a
minimum 100-year service life. Link-Pipe contains a wide variety of diameters and
ranges from man accessible to remote repair products. The sizing of the sleeve that
would work needs to be established with precise measurements of the inner diameter
of the pipe that is being repaired. At times, this lack of the sleeve’s ability to
accommodate even minor variations in the diameter of the damaged pipe presents
challenges in the field.
Figure 5. Sewer seal (Link-Pipe, 2015)
b) WEKO-SEAL®’s advantages over conventional pipe-joint repair methods

are, non -corrodible, bottle-tight seal with minimal reduction of the pipeline’s interior
diameter; operating pressures in excess of 300 psi and 100 feet of external head
pressure, with proper design, accommodates normal pipe movement from ground
shifting, thermal expansion or contraction, and vibration; it is patented technology
with a positive mechanical locking wedge design, test valves standard in all seals,
durable cross-sectional seal thickness. Design and installation options for standard
round pipes as well as lines with unusual shapes including oval, square or those
having compound angles are feasible. Installation with access openings can be in
excess of 2,000 feet apart; entrance through manholes, vaults, fittings or cut-outs,
with fast installation and minimal lead time, the technology lends itself to emergency
situations.
Over 40 years of turnkey installation experience with 300,000 seals installed, Miller
claims that they are the most experienced contractor in the industry. All installations
are warranted. Miller Pipeline’s trained, experienced personnel handle the seal
installation process in full compliance with OSHA’s 29 CFR 1910.146 Permit-

Required Confined-Space regulations with full rescue and retrieval equipment on-site
at all times. Miller’s crews are also trained in CPR, SCBA, First Aid and emergency
rescue procedures.
The material specifications for the WEKO-SEAL® fall into four main application
categories: potable water, wastewater, natural gas and seawater/brackish water. Each
application has materials specifically engineered to provide years of worry-free
maintenance through the proper rubber seal and stainless steel retaining band
selections. Seals designed for use in potable water applications are made from EPDM
rubber seal and type 304 stainless steel retaining bands. All components used for
potable water applications are NSF 61 drinking water approved by Underwriters
© ASCE
Pipelines 2015 1361
Laboratories. Storm water installations use the exact material components as

identified for potable applications. Wastewater applications use the EPDM rubber
seal shown in Figure 6 and type 316/316L stainless steel retaining bands.
Seawater/Brackish Water Installations in salt water environments consist of an EPDM
rubber seal and AL6XN retaining bands. Natural Gas Seals designed for natural gas
applications are constructed of Nitrile butadiene acrylonitrile rubber. In areas where
pipelines are likely to be exposed to petrochemicals or where oil resistance is of
concern, Nitrile or Neoprene rubber gaskets can be considered as an option to EPDM
rubber.
Figure 6. Samples of weko-seal (Miller, 2015)
c) Creamer In-Weg® seals are designed for the internal sealing of leaking pipe
joints or cracks in all types of pipe materials, including cast iron, ductile iron, concrete,
reinforced concrete, steel, vitrified clay and plastic piping systems having pipe
diameters of 16 inches or larger. These seals permanently eliminate leaks at internal
pressures up to 300 psi. Excavations for access can be as much as 5,000 feet apart,
causing very little disruption to traffic and greatly reducing restoration cost. Over
50,000 seals have been successfully fitted worldwide.
d) HydraTite Internal Joint Seal is a mechanical, trenchless remediation for
© ASCE
Pipelines 2015 1362
repair of pipe joints as shown in Figure 7. The HydraTite system consists of a

proprietary rubber seal which spans the joint and is held in place by stainless steel
retaining bands in either side of the joint. These retaining bands are expanded and
locked in place using a wedge design which forms an air tight clamp around the joint
eliminating all infiltration and exfiltration. Each HydraTite seal is designed and
custom made for the application to ensure complete compliance with project
specifications. The HydraTite System is a recognized method of joint repair by
AWWA manual M28 (2001).
Figure 7. Joint seal (HydraTite, 2015)
e) AMEX 10 offers internal mechanical sleeves for pipes under the brand
names MONO, VARIO, LEM and SPEED in the range of man entry sizes up to 240
inches for internal pressures up to 25 bars and external pressures up to 8 bars. The
sleeves have been used in broad applications worldwide for over 30 years. In fact,
Amex 10 and Miller had a working arrangement for some years until both decided to
pursue their own future plans. Uhrig that offers Quick-Lock, Quick-Lock Big and
Liner End Seal and Amex 10 have a working arrangement at the present time. Some
of their efforts are focused on developing jointly a new product line to meet
changing market needs.
f) For Snap Lock repair sleeve shown in Figure 8, no chemicals are needed for
installation. Made of high-grade stainless steel and surrounded by a rubber outer
sleeve, the snap-lock system is highly durable and resistant to most chemicals
including hydrogen sulfide. The installation process locks the Snap Lock© module
permanently in position. The Snap Lock© repair module is comprised of a cylindrical

stainless steel sleeve surrounded by an outer sleeve of rubber incorporating a
specially designed seal. The outer sleeve of buadienstryene rubber is also chemically
resistant. Hydrophilic rubber bonded to the outer sleeve expands up to 300% in the
presence of water, ensuring a completely watertight repair for site conditions where
water is present consistently.
CURED IN PLACE PIPE (CIPP) SHORT SECTIONAL REPAIRS
There are many providers of these repairs: Easy Liner LLC, Flow-Liner,
Formadrain®, Magnaline, Master Liner, Nu Flow Technologies, National Liner,
Perma-Liner, Reline America, Stephen's Technologies, and LMK Technologies. CIPP
© ASCE
Pipelines 2015 1363
lining system for short patch repairs is used to fix cracks, holes, stop leakage,
displaced/failed joints, prevent root intrusion, and help maintain the integrity of the
existing pipe. Installation procedure is composed of these steps for CIPP short liners:
1. Locate/identify problem w/TV inspection camera system.

2. Clean and remove debris.
3. Assemble packer w/fiberglass mat & resin.
4. Install assembled packer into sewer line.
5. Center mat over damaged area in pipe.
6. Inflate the packer to 25 to 30 psi until liner touches and cures.
7. Force the resin to migrate into the damaged area(s)
8. Allow resin to cure for 2 to 3 hours.
9. Deflate packer & remove from pipe.
Figure 8. Snap lock system

Typical applications:
• 3" to 36" pipe diameters

• Municipal, commercial and residential uses
• Repair cracked pipes
• Offset and open joint repairs
• Bridging of missing pipe
Advantages of sectional liner repair systems
• Fast ambient cure

• Structural repair of pipe without excavating
• Non-disruptive to traffic
• Corrosion resistance
• Cost effective
• Consistent & uniform thickness
• No by-pass pumping
• Fast and efficient installation
• Helps eliminate future water infiltration and root intrusion
© ASCE
Pipelines 2015 1364
a) Master Liner’s Sectional Pipe Renewal System offers a solution for spot
repairs in a matter of hours. Features include improved flow characteristics; high
strength, leak proof epoxy; and custom diameters and wall thickness.
b) Logiball offers a complete line of reinforced sleeve installers and carriers for
sectional liners (CIP or mechanical locking sleeves) up to 50 feet long for 4- through
36-inch pipes. In their shorter version, the Sleeve Installers are used as end plugs to
process manhole-to-manhole liners (deformed, reshaped, expanded and CIPP liners).
The Sleeve Installers are made of a two-ply, cross-biased reinforced rubber sleeve
that is resistant to hot water and even steam under pressure. The rubber sleeve is
secured to the end plate through a series of wedging points for a strong and safe
attachment even when the rubber is softened by exposure to high temperatures.
c) LMK Technologies’ Performance Liner Sectional system is a unique one-step

air-inversion point repair process and the only system that is compliant with ASTM
F2599-11 which uses swelling gaskets of hydrophilic rubber. The liner is vacuum
impregnated, meaning it is clean and safe for workers and the environment.
Furthermore 100% resin migration is achieved at the point of repair as described per
ASTM in F2599-11. Another advantage to this water-tight, structural system is that
the liner is inverted through the point of repair versus pushing or pulling the
impregnated liner through the damaged section. These liners are also properly
undersized for the diameter of the host pipe and provide a finished product that
complies with the design thicknesses per ASTM F1216. These properly undersized
liners will invert at 2 to 3 psi with a holding pressure of 4 to 7 psi. This system
renews mainline diameters from 6 to 42 inches. Smaller diameter (6 to 12-inch) sewer
pipes can range in length from 3 to 100 feet. Larger diameters can range from 3 to 30
feet. Typically the resin is cured within two hours at ambient temperatures, or as fast
as 30 minutes using LMK's steam curing system. Table 1 provides an evaluation of
the Performance Liner against others.
d) The Perma‐Liner Point Repair System consists of a fiberglass reinforced liner

and ambient cured resins. The Point Repair Kits are sold from 6 to 54 inches in
diameter and lengths from 2 to 30 feet. The point repair forms a structural, permanent
waterproof repair which seals all types of pipe against infiltration and exfiltration.
This repair has a superior bond to the existing pipe in both wet and dry conditions and
because the resin is 100 percent solids, there is no shrinkage and therefore no annular
space between the pipe and repair that could cause leakage. The Perma‐Liner
Sectional Point Repair is considered a structural repair per ASTM F1216-09.
SUMMARY
The most suitable point repair system needs to be chosen by a careful consideration of
the following factors: structural capability, hydraulic characteristics, design life, ease
of installation, need to deploy human labor or can be done remote using a packer or
robotic means, corrosion resistance, ability to stop leaks and root intrusion, need to
use chemicals or can be mechanical, cost of materials, labor and maintenance, track
© ASCE
Pipelines 2015 1365
record, and whether the technology has been vetted by one or more ASTM and NSF
standards. In the final analysis of pros and cons of internal joint seals, mechanical
sleeves, CIPP short liners and other point repairs, the market will always have a need
for multiple products. Therefore, mechanical sleeves and sectional repairs are here to
stay given the mantra “doing more with less,” and it is only a matter of time that the
total number of sleeves used in USA would become higher than all other countries
around the globe in a given year.
ACKNOWLEDGEMENTS
The authors appreciate the help from Messrs Terry Bell of Miller Pipeline, Zachary
Spera of McMillen Jacobs and Rick Gage of LMK Technologies.
REFERENCES
ASTM F2599 (2011). “Standard Practice for the Sectional Repair of Damaged Pipe
By Means of an Inverted Cured-In-Place Liner, “ASTM International, West
Conshohocken, PA.
ASTM A240 (2015). “Standard Specification for Chromium and Chromium-Nickel
Stainless Steel Plate, Sheet, and Strip for Pressure Vessels and for General
Applications, “ASTM International, West Conshohocken, PA.
ASTM A666 (2010). “Standard Specification for Annealed or Cold-Worked
Austenitic Stainless Steel Sheet, Strip, Plate, and Flat Bar, “ASTM
International, West Conshohocken, PA.
ASTM F1216 (2009). “Standard Practice for Rehabilitation of Existing Pipelines and
Conduits by the Inversion and Curing of a Resin-Impregnated Tube,” ASTM
International, West Conshohocken, PA.
ASTM F3110 (2014). “Standard Practice for the Proper Use of a Mechanical
Trenchless Point Repair Sleeve with a Locking Gear Mechanism for Pipes of
Varying Inner Diameter and Offset Joints,” ASTM International, West
Conshohocken, PA.
AWWA M28 (2001). “Rehabilitation of Water Mains, “Manual of Practice,
American Water Works Association, Denver, CO.
EN 681-1 (1996). “Elastomeric seals - Materials requirement for pipe joint seals used
in water and drainage applications - Part 1: Vulcanized rubber, “European
Committee for Standardization, Brussels, Belgium.

LMK Technologies (2015), “Breaking New Ground Without Breaking Any Ground,”
brochure, Ottawa, IL.
NASSCO (2013). “Guidelines: Specifications for Sewer Collection System
Rehabilitation, “National Association for Sewer Service Companies,
Marriottsville, MD.
NSF/ANSI 61(2015). “Drinking water system components, “National Sanitation
Foundation, Ann Arbor, MI.
NSF/ANSI 372 (2015). “Drinking water system components, “National Sanitation
Foundation, Ann Arbor, MI.
Trenchless Technology Online (2010), “Point Repair for Pipeline Rehabilitation,”
Benjamin Media, Brecksville, OH.
© ASCE
Pipelines 2015 1366
Uhrig (2015), “Quick Lock Pipe Point Repair System,” Geisingen, Germany
Table 1. Performance Liner versus Others (after LMK Technologies, 2015)
Factors to The Performance Liner Packer Wrapper Spot

consider Repairs
Configuration The tubular liner allows the Resin soaked flat sheet liners
worker to pour the resin into the are wrapped on a sewer plug
liner tube. This method is clean and held with rubber bands,
because the resin is contained. Velcro, zip ties, or other.
Liner Vacuum impregnation provides Open resin pouring is a messy
Impregnation a thorough wet-out of the liner process and is difficult to
ensuring the tube is completely verify thorough wet-out. There
filled with resin, not air. is also no evacuation of air.
Resin Migration Extra resin that is added to the Flat sheet liners can only carry
liner will penetrate fractures and the amount of resin that the
open joints in pipe. liner will absorb.
Versatility Diameters up to 42'' and Liners wrapped around a sewer
continuous lengths up to 50 ft plug are difficult to maneuver
thru’ 22'' manhole-liner/bladder and limit the dia and length of
assembly is collapsible. the liner through manholes.
Installation The liner is positioned at the Liners on a sewer plug are
point of repair and then inverted pulled through the damaged
through the damaged section, section with pipe pieces being
never pulled through the snagged leaving a convoluted
damaged section. repair or even a total collapse.
Inflation Liner/bladder is at 1-2 psi and a Sewer plugs are inflated at 28-
Pressures holding pressure 5-6 psi. 35 psi damaging the pipe more.
Documentation Inversion installation allows the Wrapped liners are shorter than
installer to view the liner the plug and the liner cannot be
before, during, and after, so that viewed. Therefore, positioning
the placement is exactly where is solely dependent on
the repair needs to be. measuring.
Length of Repair A continuous length liner Multiple short length liners that
provides uniform wall thickness overlap one another leave
over 3-50 feet. inconsistent wall thickness.
Assurance The resin is protected as the Allows the resin to be
liner is carried to the point of contaminated and wiped off
repair. during the winching-in process.
Safety This system is clean and This system requires handling
environmentally safe because of exposed resin soaked liners
there is no exposed resin. by the workers.
Standards Compliant to ASTM F2599-11. Non-compliant to F2599-11
© ASCE
Pipelines 2015 1367
Effective Repair of Incidental Construction Damage to 54-inch PCCP Line
A. B. Pridmore1; L. Bryant2; and J. Le3
1
Ph.D., M.ASCE, Vice President- Pipeline Solutions, Structural Technologies, LLC,
7455 New Ridge Rd., Hanover, MD 21076. E-mail: apridmore@structuraltec.com
2
P.E., Smith Seckman Reid, Inc., 2995 Sidco Dr., Nashville, TN 37204. E-mail:
lbryant@ssr-inc.com
3
Pipeline Solutions, Structural Technologies, LLC, 7455 New Ridge Rd., Hanover,
MD 21076. E-mail: jle@structuraltec.com
Abstract
A water utility located in the Southern US provides water to over 16,000

customers and wastewater service to over 13,000 customers across two counties. In
order to serve a growing population, the utility is in the process of upgrading its water
treatment plant from 51 MGD to 62 MGD. As a part of the construction process, soil
nails were installed for excavation on the property. During the implementation of
these soil nails, two 6-inch diameter punctures were made in a 54-inch Prestressed
Concrete Cylinder Pipe (PCCP) and 45-degree steel elbow respectively. Given that
this pipeline is the main raw water line for the treatment plant, and that the damage
occurred during peak summer demand, the need to repair the line was urgent. The
proximity of multiple bends and steep slopes adjacent to the distressed regions of the
pipelines posed significant challenges for replacement. After extensive research and
a site visit to a large municipality pipeline project utilizing similar technology, the
utility proposed an internal structural repair using carbon fiber. The owner’s options
analysis process, implementation of the bypass system, as well as the internal repair
will be detailed in this paper.
BACKGROUND AND IDENTIFYING THE DAMAGE
In July 2014, a water municipality in the Southern US was in the process of a

phased water treatment plant expansion project to accommodate the significant
growth in the area it serves. While excavation work was being performed and soil
nails were being installed to stabilize a slope at the job site, the Engineers noticed a
leak started to take place on the slope. Initially the leak was thought to be
groundwater from recent rains in the area. However, it was soon realized that this
leak appeared to be from the raw water main in the adjacent area. To verify the
Page 1 of 9
© ASCE
Pipelines 2015 1368
source of the leak, plant personnel and the Engineers began conducting experiments.
First, runoff water from the slope was directed to one channel for sampling chemicals
and measuring flow rate. Second, chlorine was added to the intake location of the
raw water line to test for residual in the slope runoff water. Also, the raw water
pipeline’s flow was ramped up and down to see if measureable changes in slope
runoff could be documented. This testing determined that the source of the leak was
the 54-inch PCCP raw water main, which transports water from the pre-sedimentation
basin to the conventional filter flash mix tank at the utility’s water treatment plant.
To investigate the damage and identify any sections of distressed pipe, a diver
was sent into the pipeline to observe and take video. Two specific damage locations
were identified through the video inspection. The diver’s inspection noted two holes
in the pipeline: one at the straight section of pipe directly upstream of a 45 degree
bend in the pipeline and one within the 45 degree bend section in the pipeline as
shown in Figure 1 below. Measurements of the hole diameters verified the sections
of the pipeline had been punctured by soil nails and the associated 6-inch diameter
pilot holes that were drilled into the pipeline (See Figure 2 and 3 below).
Damage location
Figure 1. Location of distressed pipe sections within the pipeline

(green highlighted line is pipeline repaired)
Page 2 of 9
© ASCE
Pipelines 2015 1369
Figure 2. Soil nail going through pipe wall

Figure 3. Hole in steel elbow pipe section
Page 3 of 9
© ASCE
Pipelines 2015 1370
OPTIONS ANALYSIS
Based on the damage identified by the video inspection, it was determined the
penetrated sections of pipe could not be relied on safely for long term operation. The
sections required either replacement or structural upgrade.
As part of the options analysis, replacement, slip-lining, and stand-alone

structural rehabilitation with Carbon Fiber-Reinforced Polymer (CFRP) were
considered. The proximity of multiple bends in the pipeline to the distressed pipes
ruled out slip-lining as a repair option. Due to the pipeline being positioned adjacent
to a 20 foot drop-off, excavation for replacement would have been extremely
challenging (see Figure 4). Furthermore, lead time for replacement of the special
order steel bend section would have left the pipeline unrepaired for an extended
amount of time during the remaining peak summer demand. As the need to repair the
line was pressing, the team decided that trenchless repair was the best option. Based
on the notable external load carrying requirements and the construction schedule
restraints, a CFRP lining system was selected as the repair method for its structural
capabilities, trenchless application, and timeliness.
Figure 4. Location of pipeline in relationship to excavation slope
The CFRP lining system was designed as a stand-alone system without

reliance on the host pipe for structural integrity. In addition to taking the internal
pressure loads of the pipeline, the repairs were required to take into account all
external loads including soil pressure and the vehicular load of a Manitowoc Model
#555 crane. With a crane weight of 150,000 lbs and a load carrying capacity of 30
Page 4 of 9
© ASCE
Pipelines 2015 1371
tons (60,000 lbs), this totaled to 210,000 lbs of vehicular load. The load is distributed
onto two tracks, each 3ft by 21ft, which translates to approximately 1,700 psf load
acting on the soil.
In addition to addressing the distressed sections, the adjacent sections also

needed to be evaluated to determine whether they had also been structurally
compromised. The owner considered involving a non-destructive evaluator to assess
nearby pipe sections; if damage was found the adjacent sections would also require
replacement or stand-alone structural rehabilitation. In the end, the owner decided to
address the nearby segments assuming they were structurally compromised, and
allocate funds directly to repair, in lieu of inspection followed by repair. Structural
Group was then called in to conduct the repair.
Given that the owner had no previous experience with the use of CFRP, it was
decided that a direct visit to an active CFRP upgrade project would be beneficial.
The CFRP installation contractor had a project ongoing with Miami-Dade Water and
Sewer Department (MDWASD) at that time repairing 54-inch PCCP, so an exact
comparison would be possible. The CFRP installation contractor coordinated with
Miami-Dade WASD and the owner was able to successfully visit the jobsite in South
Florida, observing the CFRP installation process and helping to finalize the decision
to move forward with CFRP repair.
IMPLEMENTATION OF THE BY-PASS SYSTEM
Given the critical nature of the affected 54-inch pipeline, the owner elected to set
up a temporary by-pass upstream and downstream of the repair area, so the pipeline
could remain operational throughout the repair process. As shown in figure 5, the by-
pass system was set up on the Site prior to the repair process moving forward.
Figure 5. By-Pass System
Page 5 of 9
© ASCE
Pipelines 2015 1372
IMPLEMENTATION OF CFRP INTERNAL REPAIR
The CFRP repair process involves layers of unidirectional carbon fiber fabric
being installed in a pipe longitudinally and circumferentially. In areas with steel
substrate such as the joints, a layer of glass fabric was used as a dielectric barrier
between the steel and the carbon fiber. Both the carbon fiber and glass fabrics were
saturated in a two-part 100% solids epoxy using a mechanical saturator (See Figure
6).
Figure 6. Glass fabric being saturated with a two-part 100% solid

epoxy using a mechanical saturator
Prior to application of the CFRP repair system, the concrete substrate was
prepared to a minimum of ICRI CSP-3 using sponge blasting per project drawings.
In addition, the punctured pipe segments needed to be restored to allow for a uniform
substrate during the repair (see Figure 7A-D). Therefore, the inner core and cement
mortar in the region surrounding the punctures were removed, as well as the damaged
reinforced mortar on the steel pipe section. Any length of soil nails which protruded
into the pipe was also removed. Steel plates were then welded over the holes and
prepared to SSPC SP-10 near white metal finish via sandblasting. Chemical grouting
was used before and after welding of steel plates to address leakage and restore
disrupted soil surrounding the pipe segments as needed. The concrete inner core that
had previously been removed was then restored using non-shrink cementitious repair
mortar, creating a uniform substrate for installation of the CFRP system.
Page 6 of 9
© ASCE
Pipelines 2015 1373
Figure 7A- Puncture in pipeline, as found Figure 7B- Welding over holes in the pipe
Figure 7C- Prepared substrate Figure 7D- Surface restored with epoxy
mortar
Figure 7. A-D: Restoring punctured pipeline segments to allow for a uniform substrate
In addition to the “typical” dewatering, localized dewatering efforts were required to

manage water ingress at the hole locations.
To verify surface preparation, adhesion tests were performed on adjacent substrate

per ASTM D4541 to 300 psi minimum (See Figure 8). Once surface preparation was
completed, the pipe substrate was covered by a prime coat of epoxy and a layer of
thickened epoxy (See Figure 9.)
Page 7 of 9
© ASCE
Pipelines 2015 1374
Figure 8. Pull test per ASTM D4541 being performed to verify surface preparation
Figure 9. Primer epoxy being applied to concrete substrate adjacent to hole repair
Saturated layers of CFRP were then installed onto the interior of the pipe per the
project drawings. The design prescribed that the majority of the reinforcement be in
the hoop direction and minimal longitudinal reinforcement in straight regions.
Additional longitudinal reinforcement was provided near the two (2) 45° bends, see
Figure 10 for completed installation.
Page 8 of 9
© ASCE
Pipelines 2015 1375
Figure 10. Completed CFRP installation at 45° bend
CONCLUSION
Given the set of circumstances presented for this damaged 54-inch PCCP and the
owner requirements to maintain water delivery through peak season, CFRP was
applicable and advantageous for the repair. The options analysis and site visit
provided the engineer and owner the opportunity to weigh out and thoroughly
investigate CFRP repairs of pressure pipelines and the typical construction process.
The use of CFRP allowed for a rapid, fully structural repair placing the pipeline back
in service with minimal disruption.
REFERENCES
ASTM D3039, Standard test method for tensile properties of polymer matrix
composite materials. American Standard for Testing and Materials (ASTM)
ICRI Guideline No. 310.2, Selecting and specifying concrete surface preparation for
sealers, coatings, and polymer overlays. International Concrete Repair Institute
(ICRI)
SSPC-SP No.10 / NACE 2 Near-White Blast Cleaning. Society for Protective

Coatings (SSPC) and the National Association of Corrosion Engineers International
(NACE)
Page 9 of 9
© ASCE
Pipelines 2015 1376
Repairing the World's Largest Prestressed Concrete Pipe:

A Case Study of the Central Arizona Project's
Centennial Wash Siphon
Jim Geisbush, P.E., PMP1

1
M.ASCE, Senior Civil Engineer, Central Arizona Project, P.O. Box 43020, Phoenix,
AZ 85080. E-mail: jgeisbush@cap-az.com
Abstract
The Central Arizona Project (CAP) operates and maintains three (3) 21-foot diameter
pre-stressed concrete non-cylinder pipes (PCP) as part of the aqueduct system that
delivers almost 1,950 million gallons of Colorado River water per day, to Maricopa,
Pinal, and Pima Counties in central and southern Arizona. The pipelines, installed in
the late 1970's, have experienced pre-stressing wire breaks and have been repaired
several times since the early 1990's. The wires have broken primarily due to defects
in the wire. One of the pipes, the Centennial Wash Siphon, conveys water beneath
Interstate-10 and the Centennial Wash, about 80 miles west of Phoenix, AZ. The
siphon is upstream of CAP's first customer turnout, and as such is critical in the
aqueduct's conveyance and delivery system. An internal electromagnetic inspection in
January 2013 discovered several pieces of the pipe comprising the siphon had many
broken prestressing wires, prompting a closer inspection and assessment of the
siphon. The case study described herein examines inspections, assessment,
monitoring, and subsequent repair using post-tension tendons of the Centennial Wash
Siphon. This paper further discusses the ongoing efforts of monitoring, assessing,
repairing, and maintenance practices for the largest prestressed concrete pipes in the
world. A brief history of the siphons includes manufacturing of the 252-inch diameter
prestressed concrete pipes, installation of the pipelines, and early investigations and
repairs. The focus of this paper is on the assessment and monitoring since the last
repairs in 2006, specifically newer technologies that have emerged to assist CAP in
monitoring and making decisions in the repair methods and locations. Relevant
points include a discussion on the excavation of the pipes requiring repair, the repair
work (post-tension tendons), and maintenance practices. A brief discussion will

follow on new technology recently installed in all three of CAP's prestressed concrete
siphons.
Central Arizona Project Background

The Central Arizona Project is a 336 mi (541 km) aqueduct system designed and
funded by the Bureau of Reclamation (Reclamation) and operated and maintained by
the Central Arizona Water Conservation District (CAWCD) to deliver water to
multiple entities throughout central and southern Arizona. In general, the aqueduct
system withdraws water from the Colorado River at Lake Havasu in western Arizona
and through multiple pumping plants and a combination of open channel canals and
closed conduit pipes delivers water to the major urban areas within Maricopa, Pinal,
© ASCE
Pipelines 2015 1377
and Pima counties in Arizona – the Phoenix and Tucson metropolitan areas. Water is
also delivered to agricultural users, recharge facilities, and Native American
reservations along the route. The system is generally comprised of:
• 14 pumping plants
o Capacities ranging from 208 cfs to 3600 cfs
o Lifts ranging from 79 ft to 850 ft
• 39 check structures (for water level control)
• 49 turnouts (for customer delivery)
• 6 groundwater recharge facilities
• Two reservoirs
• One Pumping/Generating Plant
• About 350 miles of canals, pipes, and tunnels
Approximately 5 million people (about 80 percent of the population of the state of

Arizona) are provided water from the CAP. In 2014, the CAP delivered more than
1.57 million acre-feet (Ac-Ft) of water to various users within Arizona. Table 1
indicates approximate water usage by category.
Table 1. Water Usage by Customer Type
Acre-Feet(1) Percent of Total
Municipal & Industrial 440,000 28%
Recharge 251,000 16%
Agricultural 393,000 25%
Native American Reservations 487,000 31%
(1) Rounded to nearest 1,000 1,571,000 Total for 2014

Centennial Wash Siphon History

There are 252 pieces of pipe comprising the Centennial Wash Siphon. Each piece is
about 24.5-feet in outside diameter, 22-feet long, and weighs about 225-tons; the
siphon is approximately 5,800 feet long. Due to the size of the pipe pieces, onsite
manufacturing plants were established near each siphon location and specific
manufacturing and transportation equipment was designed for the manufacturing and
installation of the pipe pieces. Figure 1 shows one of the field manufacturing plants.
Figure 2 shows a piece of pipe being maneuvered into place.
A 1984 failure of a prestressed concrete cylinder pipe on another Reclamation

aqueduct system prompted Reclamation to undertake extensive investigations to
determine the failure and evaluate future pipeline serviceability. Defective
© ASCE
Pipelines 2015 1378
(longitudinally cracked) wire was identified as the cause for the failure of the pipe.
Because of the problems Reclamation discovered, they initiated a program to monitor
other prestressed concrete pipelines. In January 1990 surveys indicated potential
corrosion issues at the CAP's prestressed concrete siphons (in 1990 CAP had six
prestressed concrete siphons in service). Widespread distress was confirmed at
several locations on all six siphons, and of 223 individual pipe pieces excavated to
springline, 28% were found to be distressed and requiring repair, some were so
severely distressed complete replacement of the prestressing wire was required. In
the mid-1990's Reclamation abandoned three of the six (6) prestressed concrete
siphons and installed new siphons paralleling those abandoned. The three (3)
remaining prestressed concrete siphons are still in operation – the Centennial Wash
Siphon is one of those.
Figure 1 – 1978-79 Onsite Manufacturing Facility

The Centennial Wash Siphon was manufactured and installed in 1978-1979 – the
largest prestressed concrete pipe in the world. Results of Reclamation’s 1990
investigations resulting from the 1984 pipeline failure lead to the repair of two pieces
of pipe on the siphon using post-tension tendons in 1991 and lining the first 1,000 feet
of the siphon with an internal steel lining in 1996 – that section of pipeline travels
beneath Interstate-10 in western Arizona. Early investigations and attempts to locate
distressed pipes were largely unsuccessful; it was not until the early 2000's when the
Remote Field Eddy Current Transformer Coupling (RFECTC) Technology was
further refined that results became more accurate and reliable, were distressed pipes
able to be more accurately identified and located.
© ASCE
Pipelines 2015 1379
In November 2002 the first RFECTC inspection was performed on the Centennial
Wash Siphon. The results indicated there were 46 pieces of distressed pipe with the
number of broken wires ranging from 20 to 345; just over 18% of the pieces exhibited
some level of distress.
Figure 2 – Pipe Mobile (Note Volkswagen for scale)
A second RFECTC inspection occurred in October 2004; the results indicating there
were 48 pieces of pipe with the number of broken wires ranging from 10 to 380; just
over 19% of the pieces exhibited some level of distress, two new pipes exhibited
signs of distress since 2002, and some pipes had an increase in the number of wire
breaks. Based on these results it was decided to repair several pieces of pipe.
In November 2006 a third RFETC inspection occurred, as well as repairing several

pieces of pipe using external post-tensioned tendons. Results from the 2006
inspection found only one newly distressed pipe, though ten pipes had an increase in
the number of wire breaks, and the remainder unchanged from the 2004 inspection.
Additionally, an acoustic hydrophone monitoring system was installed to be able to
monitor wire breaks in real time. The monitoring system had marginal success;
communications to the remote site was problematic and at least once the system
broke free of its tether in the siphon inlet due to the high turbulence. By 2010 the
system was inoperable so much of the time CAWCD was not able to track wire
breaks reliably, so the decision was made to dewater the siphon and conduct another
round of Electromagnetic (EM) inspections in January 2013.
© ASCE
Pipelines 2015 1380
In January 2013 a fourth EM inspection occurred and during that outage CAWCD
installed an Acoustic Fiber Optics (AFO) Monitoring System to track wire breaks. A
new and more reliable fiber optics system for communications had recently been
installed along the entire CAP system. The results from the 2013 EM inspection
provided a new baseline of distress and provided an idea of wire distress growth since
the first inspection in 2002.
The results of the January 2013 inspection, along with the real-time results of the
AFO monitoring system for the first few months of 2013 indicated the need to repair
several pieces of pipe. The EM inspection indicated 40 pipe pieces had distress
ranging from 25 wire breaks to 280 wire breaks, but it was the activity monitored by
the AFO system on specific pipes that lead to the decision to conduct repairs. Figure
3 is a graph showing the seven (7) most distressed pieces of pipe based on the January
2013 EM inspection; the graph shows the growth rate of distress since the first
inspection in 2002. Also shown are the AFO wire break events for each piece; those
numbers are in addition to the numbers reported by the EM inspection.
Figure 3 – Graph showing distress growth rate for the 7 most distressed pipe pieces
It takes CAWCD approximately two years to schedule and execute a siphon outage,
thus the approximate two years between early inspections. An outage was already
scheduled for two weeks in January 2014 for the aqueduct system in the vicinity of
the Centennial Wash Siphon, so the decision was made by the Engineering,
Maintenance, and Operations groups to fast-track a repair design to be able to hit the
January 2014 outage. The design required the repairs be executed within a 12-day
period – of the 14-day scheduled outage a day was needed to dewater and another to
rewater.
© ASCE
Pipelines 2015 1381
Events Leading to January 2014 Repair

Superimposing the current results of the AFO monitoring system over the EM
inspection results from January 2013 and the rate of wire break increase since 2002
enabled CAWCD engineers to determine which pipes should be further investigated
to initiate repair. Six pipes were identified and in October 2013 excavated to
springline. CAWCD engineers performed acoustic and visual inspections of the
exposed pipe pieces and Pure Technologies performed a surface electromagnetic scan
to validate the previous inspection data as well as the visual inspections. During the
field inspections of the exposed pipes, three additional pipes were identified as
requiring repair.
The results of the internal EM inspections, the AFO monitoring system, the surface
EM scan, and the visual/acoustic inspections all correlated well, and CAWCD
engineers began designing the repair system for nine pieces of pipe. Table 2 provides
a chronology of events for the repair effort.
The selected method to repair the pipe pieces was with post-tensioned tendons.
Essentially a high strength steel tendon was wrapped around the pipe, tensioned to a
specified stress and locked into an anchor block set on the top of the pipe. This
required fully excavating portions of the siphon and removing the soil support
beneath. Nine (9) pieces of pipe were identified that required repair, but because of
their locations along the pipeline, five (5) separate excavations were required.
Excavation of the pipe was conducted in four (4) phases. The first phase (preliminary
excavation) occurred in October 2013 when the pipe was exposed to its springline.
The second phase occurred in early January 2014 when an additional five feet of
material was removed – the siphon was still conveying water for customer deliveries
at this point. The third phase removed an additional 3½ feet of material – this took
the excavation to the 90° bedding angle of the pipe – at this point water was not being
transported through the siphon. The fourth and final phase removed the supporting
material beneath the pipes in specific sequences in specified locations, but only after
the siphon was isolated from the canal.
The preliminary excavation plan was developed to accommodate quickly and

efficiently removing a minimal amount of material during final excavation. It was
determined based on the soil composition the excavation would be stable with 1½:1
(horizontal to vertical) side slopes and 3:1 slopes at the ends for vehicular and
personnel access. The depth of cover over the pipe varied from just under ten feet to
about twelve feet, so excavations to springline averaged over twenty-two (22) feet. If
the excavations were cut directly to the springline, the excavations would have been
about 80 feet wide at the top; but the intent was to remove as much material as
possible during preliminary excavation so a minimal amount was moved during
repair. The excavations were cut as if the side slopes proceeded to a point one-foot
© ASCE
Pipelines 2015 1382
below the bottom of the pipe which set the width of the excavations at about 130-feet.
The 3:1 end slopes dictated an excavation over 200 feet longer than the length of
exposed pipes. Excavation accounted for about 80% of the construction duration.
Table 2. Chronology of Events Leading to Repair Effort
Time Frame Activities
January 2013 EM Inspection
Feb-Apr Tracked AFO Events
Mar-Apr Evaluated EM Inspection Results
May Established Repair Priority
Prepared Internal Planning Documents and Preliminary Design

May-Jun
for Project Approval
July Received Senior Management Approval for Project
Jul-Sep Developed Preliminary Excavation Plans
October Preliminary Excavation – to Springline
November Surface EM Scan – Validation of Inspection Results
Visual Inspection of Exposed Pipes

Oct-Dec Development of Repair Contract Documents
Selection of Repair Method – Post-Tensioned Tendons
6 Jan 2014 Began Additional Excavation
13 Jan 2014 Began Outage – Contractor Began Repairs
25 Jan 2014 Contractor Substantially Complete

26 Jan 2014 Siphon Put Back in Service
Current Repair (January 2014)

In mid-January the water was "turned off" to the siphon and the siphon isolated from
the upstream and downstream canal sections. Stoplogs were set at the siphon inlet
and outlet and a pump positioned in the siphon to begin dewatering the siphon – the
intent was to remove the water pressure in the pipe. At this point the Contractor was
allowed to remove material from around the pipe to the 90° bedding angle (third
excavation phase).
© ASCE
Pipelines 2015 1383
Excavation beyond the 90° bedding angle (to beneath the pipe) was only allowed in
the locations where the pipe was to be wrapped with post-tensioned tendons. The
sequence of repair limited excavation beneath the pipe to 12-feet at mid-span and 10-
feet across a joint; this required three separate excavations to wrap a single 22-foot
piece of pipe.
During design, a repair sequence was established that optimized the sequence of
repair areas. The repair areas would be excavated, wrapped with tendons, the tendons
stressed, and the pipe backfilled with a controlled-low-strength-material (CLSM) up
to the 90° bedding angle, see Figure 4. The CLSM is essentially a "one-sack slurry"
(flowable fill) to mimic the original bedding of the pipe and provide support to the
pipe haunches that was removed to facilitate the repair. The CLSM also provides a
higher pH buffer around the pipe and new tendons from the soil. Once the CLSM
sufficiently cured, the material adjacent to the CLSM could be removed and repairs
undertaken in the new area. There were 17 separate areas for which the following
repair sequence was performed (at this point the pipe was only "buried" up to a point
of about 3½-feet of soil):
• Excavate to one-foot below the pipe

• Wrap post-tension tendons around the pipe
• Stress the tendons and lock them in the surface couplers (anchors)
• Backfill the repair area with CLSM up to the pipe's 90° bedding angle –
ensuring complete flow beneath the pipe
© ASCE
Pipelines 2015 1384
Figure 4 - Placing CLSM Beneath Pipe
Figure 5 shows two pieces of pipe excavated where the tendons have been installed,
some of which have been tensioned. Note the material removed from the mid-span of
adjacent pieces and the pipe is supported under the joints.
© ASCE
Pipelines 2015 1385
Figure 5 - Sequencing of Repairs
When the post-tensioned tendons were installed on all nine (9) pipes requiring repair
and all areas backfilled with CLSM up to the 90° bedding angle, CAWCD was able
to begin conveying water through the siphon – Substantial Completion was given at
this point – two (2) days prior to the end of the outage. The next task was to bond the
anchor blocks to each other and to sacrificial anodes for eventual connection to a
cathodic protection test station. Each pipe that was repaired had two anodes and a
test station installed. The next sequence was to apply a three-inch thick layer of
shotcrete over all the repair areas to completely encapsulate the tendons and anchor
blocks in a high pH environment to minimize corrosion of the anchor blocks and add
additional protection to the pipe.
Once the shotcrete cured the process of backfilling the excavations began. Material
was replaced in the excavation in 12-inch lifts and compacted up to 5-feet below the
springline of the pipe. From 5-feet below springline up to final grade material was
placed in 18-inch lifts and wheel rolled – no special compaction was required.
Cathodic protection test stations were installed and tested upon completion of
backfilling.
Post-Tensioned Tendon System

The post-tension tendon system selected was from Dywidag-Systems International
(DSI) which is generally composed of a high strength tendon (ASTM A416), an
anchor block that was installed on the surface of the pipe, and a three-part wedge that
© ASCE
Pipelines 2015 1386
locked the tendon into the anchor block. The typical sequence of operation was the
Contractor wrapped several tendons around the pipe, typically at a 6-inch spacing, set
each tendon in an anchor, tensioned the tendon to approximately 47,000 pounds,
(~216,000 psi), and locked the tendon in the anchor. The tendon relaxed to about
41,000 pounds, ~189,000 psi when locked in the anchor. The anchors were then
injected with a rust inhibiting grease and bonded to each other prior to being covered
in a protective layer of shotcrete.
Continuing Assessment and Monitoring

CAWCD is actively monitoring all three prestressed concrete siphons with the
recently installed Acoustic Fiber Optics Monitoring Systems and is constantly
reviewing results with the current and past EM inspection results to look for trends in
individual pipes pieces, as well as gauge the overall health of each pipeline. Failure
curves for each pipe class were recently developed and those curves are being used to
assess the pipes and even trying to predict future events based on trending.
CAWCD is also in the process of developing a "Pipeline Reliability Group" to

actively meet and discuss the condition of all of CAP's pipelines, not just the siphons.
The group, composed of engineering, operations, and maintenance personnel will
make recommendations on further monitoring activities, as well as schedule and
begin planning for additional inspections. CAWCD has experience in monitoring,
assessing, and maintaining pipelines, and continues with that experience in a
proactive manner. If repairs are required, it is this group that will support the
engineering team in developing and executing repairs.
References
Peabody, M.T. (1992). "Repair and Rehabilitation of Prestressed Concrete Pipe at the
Central Arizona Project." AWWA Proceedings.
Randolph, R.D., and Worthington, W. (1992) "CAP Prestressed Concrete Pipe:

Investigations and Repairs." USCID Newsletter January 1992.
© ASCE
Pipelines 2015 1387
Motts Run Dam Outlet Rehabilitation—A Case Study Illustrating Design and
Construction Aspects
Chris Edwards, P.E.1; Owais E. Farooqi, P.E.2; and Ahmad Habibian, P.E.3
1
Spotsylvania County Utilities Dept., 600 Hudgins Rd., Fredericksburg, VA 22408.
E-mail: CEdwards@Spotsylvania.va.us
2
Black & Veatch, 18310 Montgomery Village Ave., #500, Gaithersburg, MD 20879.
E-mail: farooqioe@bv.com
3
Black & Veatch, 18310 Montgomery Village Ave., #500, Gaithersburg, MD 20879.
E-mail: HabibianA@bv.com
Abstract
This paper presents a case study covering design and construction phases of sliplining
the outlet pipe of the Motts Run Dam in Spotsylvania County, Virginia. The
reinforced concrete outlet pipe had been previously lined with an HDPE liner in order
to allow the outlet pipe to be used for pumping water back into the reservoir.
However, the liner did not perform as expected. An investigation of the liner
determined that the existing HDPE liner had experienced major deformation. The
County decided to remove the failed HDPE liner and replace it with a steel liner. The
paper provides an overview of the challenges encountered during design and
construction. The project was successfully completed in 2015.
INTRODUCTION
Background and history. Motts Run Dam is located about eight (8) miles upstream
and west of Fredericksburg in Spotsylvania County, Virginia. The dam is located on

Mine Run tributary, approximately 1,000 feet upstream of its confluence with
Rappahannock River. With a storage volume of over 4,000 acre-feet, the Motts Run
Reservoir is an important source of raw water for the Spotsylvania County’s Motts
Run Water Treatment Plant. The reservoir consists of an earth fill embankment dam,
constructed in 1970’s, with a separate intake tower that also serves as the principal
spillway for the dam. The 700-ft long, 48-inch RCP outlet pipe extends through the
dam embankment and connects the intake tower to the outlet structure. The outlet
pipe is designed to convey overflow and intentional releases from the reservoir under
gravity.
© ASCE
Pipelines 2015 1388
In 2002, during a previous improvement the outlet pipe was lined with a 36-inch
HDPE liner and the annular space was filled with cellular grout, with an intention to
utilize the outlet pipe for filling the reservoir from Rappahannock River Intake
pumping station. These improvements were intended to utilize the reservoir as
pumped storage which required the pipe to operate under pressure. In 2008, the
County needed to raise the water level in the reservoir and routed the pressurized
flow from the Rappahannock River Intake Pumping Station through the outlet. While
the pipe was pressurized, a concentrated seep was observed on the downstream slope
of the dam which threatened the integrity of the dam. The pipe was depressurized and
an investigation was performed to identify the cause of the seepage. A CCTV
inspection of the outlet pipe was performed in September 2008. The inspection
revealed significant lengths of the HDPE liner had undergone minor to moderate
distortion. The inspection also revealed several shorter lengths with severe distortion.
Due to concerns associated with the safety of the embankment, a Ground Penetrating
(GPR) survey was performed from inside the outlet pipe to assess the potential
existence of voids behind the HDPE liner and the RCP pipe. Three areas with large
voids along with several minor voids were identified. Figure 3 shows the summary of
the GPR results. However, due to the signal strength it could not be confirmed if the
voids were in the embankment or in the annular space.
To further assess the condition of the embankment, a Cone Penetrometer Testing

(CPT) program was conducted on the downstream slope of the embankment. The
objective of the cone testing was to identify embankment distress resulting from
seepage through the downstream slope. Areas near the concentrated seep locations
were investigated for underlying soft or loose material that would indicate weak
zones in the embankment. The CPT investigations did not indicate presence of any
unusual weak zones that could potentially pose any risk to the dam. Based on the
results of the aforesaid investigation, it was concluded that the HDPE liner
deformation was likely due to uncontrolled grouting operation of the annular space
during the previous modifications. It was concluded that remedial measures were
needed to allow the County to use the outlet under pressurized conditions.
Purpose and scope of current rehabilitation. The purpose of the current

rehabilitation is to make the outlet pipe capable of conveying pressurized flow from
the pump station into the reservoir. The County identified other improvements for
enhancing operational efficiency and reliability of the reservoir. These improvements

included works in the intake tower and at the impact basin. The following is the
description of the work scoped under the current rehabilitation:
Works at the outlet pipe. The works included reaming out of the existing 36-
inch HDPE liner from the outlet pipe, and re-line the outlet pipe with a new 36-inch
steel pipe. Upsizing of the liner was considered but was deemed unnecessary from a
capacity standpoint. Additionally, there was a concern that the original host pipe
could have joint misalignments or bellies which could interfere with the installation
of steel liner. As such, the size of the liner was keep at 36 inches. The annular space
between the new steel pipe and the existing 48-inch RCP was filled with cement
© ASCE
Pipelines 2015 1389
grout. Since the GPR results exhibited potential voids in the embankment outside the
existing 48-inch liner, another GPR survey was included in the scope after removal of
the existing HDPE liner and annular grout. The replacement of the existing liner
entailed demolition of the existing impact basin, the outlet gate valve and ancillary
pedestrian bridge structure. The new design included replacement of the gate valve
and modification in the ancillary structures to enhance flexibility in operation and
maintenance of the outlet valve.
Miscellaneous works at the intake tower. To take full advantage of this

opportunity, other miscellaneous works identified separately were included in the
scope of this rehabilitation. These works included replacement of existing sluice gates
and installation of a new low level outlet. The existing sluice gates exhibited
displacement, causing leaks and difficulty in operations. In order to mitigate future
potential for leakage, the new gates were relocated on the outside face of the intake
tower walls, and were installed using through-bolt connections. In addition, a new
low level outlet valve was installed in the intake tower as a replacement for the
existing 36 inch low level outlet that was not operational.
Challenges during design and construction. The design and construction of these
works were challenged because of limited information from previous modifications,
unknown structural conditions of the 48 inch outlet pipe, and limited ability to assess
conditions of the embankment from within the HDPE liner. Additional planning for
flood mitigation and worker safety was required since working within the outlet pipe
would potentially require closure of the principal spillway. The County worked with
the Engineer to clearly define project objectives, identify elements that are critical for
long term performance, and develop Contract Documents to accomplish the desired
level of quality control during construction. The Contract Specifications included
specific measures such as CCTV inspection of host pipe at several stages of liner
demolition, quality control of welds for the new steel liner, extensive review of liner
placement procedures in the outlet, low shrinkage cementations grout, contact
grouting, detailing of end connections to block potential seepage paths and pressure
testing. Extensive coordination between the Engineer and the Contractor was
enforced to ensure continuous evaluation of Contractor’s means and methods with the
progress of the project and with the changes in condition that were not anticipated
during the design. The following sections present aspects of design and construction
that were critical to the successful completion of the project.
CRITICAL FACTORS - DESIGN AND CONSTRUCTION PHASE
The most critical aspect of the design was to put together Contract Documents that
clearly define the project objectives while outlining constraints and risks for the
Contractor. Since the dam is a critical part of County’s water supply infrastructure,
several opinions from the Engineer’s in-house technical experts and outside
contractors were sought to review constructability, identify anticipated means and
methods, potential risks during construction and feasibility of the design. The process
included in-depth discussions with the County staff to lay out available options,
© ASCE
Pipelines 2015 1390
jointly evaluate potential risk and rewards for these options, and choose the best
course for the project. This section summarizes critical factors encountered during
design and construction, and the mechanisms adopted to address those factors.
Demolition of the existing HDPE liner and annular grout. The demolition of the
existing 36 inch liner and annular grout was identified as the critical task with highest
degree of risk and uncertainty. There were limited as-built records from 2002
modifications that forced the designers to base the current design on previous design
drawings. Elements such as the strength of annular grout used in construction added
uncertainty associated with the hardness and associated difficulty in its removal.
There were concerns that if richer mixes were used to exceed the specified strength
parameters in the previous modification, it may make it difficult to demolish the grout
without using mechanical means. There was a concern that the host pipe could be
damaged during demolition of the liner and annular grout
Similarly, several possibilities were contemplated pertinent to the deformed condition

of the existing 36 inch liner. Based on preliminary assessment, the deformation was
likely due to uncontrolled grouting operation of the annular space during the previous
modifications; however, the indication of potential voids in the GPR survey raised
concerns that the host pipe (i.e. existing 48 inch RCP) might have cracked leading to
loss of material along the concentrated seepage path, and subsequent washout of
embankment material into the outlet.
To mitigate these concerns, specific measures such as CCTV inspection of host pipe
at several stages of liner demolition were included in the Contract Specifications. The
Specification required the Contractor to perform CCTV before and after removal of
the annular grout to preempt any risk resulting due to the damage of the host pipe.
The Contractor was limited to demolishing the liner and grout in 75-feet sections to
reduce the risk of any unforeseen circumstances resulting from the demolition. The
Contractor could advance the demolition only after satisfactory review of the CCTV
data by the Engineer for the previous 75 feet section. Enforcing these provisions of
the specifications required a proactive approach, and defined protocols to efficiently
minimize impacts to the construction schedule. The Engineer worked closely with the
Contractor to coordinate the timing of the CCTV review and provide resources for
expeditious reviews and decisions. The Contractor used a small hand held equipment
to cut the HDPE pipe into small pieces and to demolish the annular grout. It was
found that the bond between the annular grout and the host pipe was very weak and
big chunks of grout would come off when impacted by a hammer and chisel. The
outlet pipe was found to be in good condition. The results of a GPR performed after
the removal of the liner and grout did not indicate the presence of any voids. As a
result, minimal repairs and grouting of the areas outside the outlet pipe was required.
Figure 1 shows the view into the host pipe following a section of HDPE and annular
grout were removed.
© ASCE
Pipelines 2015 1391
Figure 1. View inside 48 inch RCP after removal of HDPE liner and annular grout
Temporary diversion. The rehabilitation of the outlet pipe required accessing the
pipe from the downstream impact structure, which necessarily required closure of the
principal spillway. Although, there is an adjoining emergency spillway which is
sufficient for the safety of the dam embankment, the configuration of the reservoir
and emergency spillway does not permit routing of spills or small discharges through
the emergency spillway. The two options were either to provide storm storage within
the lake by lowering the reservoir, or to design a temporary by-pass mechanism that
could divert inflows from the lake to the Rappahannock River.
Lowering of the lake by a few feet was deemed necessary for the safety of the
workers, considering required reaction time to vacate the outlet pipe in the event of an
unprecedented spill from the lake. However, since the reservoir accounts for a third of
County’s raw water storage, the County was hesitant to lower the lake in view of
potential water supply concerns during the peak summer demands. Since the lake
serves as a key recreational facility for the residents of the City of Fredericksburg,
there were additional concerns that the lowering of lake would negatively impact
recreational activities along the periphery of the lake. During the design phase, these
aspects were discussed with the County to determine a workable approach without
shifting too much risk on the Contractor.
It was decided to adopt a middle of the road approach that included lowering of the
lake level by a few feet and including a temporary diversion to suit Contractor means
© ASCE
Pipelines 2015 1392
and methods. The approach essentially provided flexibility to the Contractor to size
the diversion while offering a maximum limit to the allowable lowering of the lake
levels. The maximum allowable lowering was set at 4 feet that could provide
temporary storm storage for a 5-year storm event. In addition, optional temporary
raising of the weir elevation of the principal spillway by 2 feet (from the existing
normal pool) would allow a combined storm storage equivalent to a 10-year storm.
The Contractor utilized this approach effectively by using a multi-pipe siphon by-pass
system to siphon the desired amount of water from the lake. Once primed by small
pumps, the siphons provided flow under gravity on a continuous basis. The number of
pipes was adjusted during the course of the construction, depending on the amount of
inflow into the lake. During construction several storm events were successfully
contained within the reservoir, and gave enough time to the Contractor to vacate the
outlet, remove equipment and prepare the site for potential spills through the principal
spillway. The specified method of temporary diversion was successful, and proved to
be a cost-effective solution for the project. Figure 2 shows multiple siphon pipes used
as temporary diversion.
Figure 2. Downstream view from the top of the dam showing siphon pipes
Encourage innovation from the contractor. During design, the engineering team
reviewed the anticipated means and methods that the Contractor could potentially use
for the project. Because of the unconventional nature of the work, it was envisioned
that subject to review and approval, the design may accommodate innovative means
and methods from the Contractor to provide flexibility. This philosophy was
© ASCE
Pipelines 2015 1393
engrained in the design drawings and specifications that contained a representation of

a basic work scheme that would ensure constructability using a relatively low risk
approach. The work scheme was deliberately made conservative to invoke
Contractor’s interest in the opportunity and motivate him to innovate for financial
gain.
The approach not only invited fresh thoughts and innovation from the Contractor, but
also protected County’s interest since it guaranteed that the Contractor would deliver
the work product as specified. One drawback of the approach was that it required
more review and coordination during construction; however, given the
unconventional nature of the work and the lack of previous engineering records, the
overall benefits of the approach outweighed the additional engineering time during
construction. An example of this approach was to keep provisions for alternative
methods of installations for the steel liner.
The design for the installation of the steel liner was based on the assumption that the
steel liner would be assembled by butt-welding pipe segments over a temporary
platform on the downstream side of the outlet. Following the assembly, the liner
would be pushed using conventional hydraulic jacks into the host pipe. The design
specified the use of flexible spacers to prevent the liner from floatation during
grouting of the annular space. The cementitious grout specified for filling the annular
space was designed to act as a corrosion shield on the exterior of the steel pipe. It was
envisioned that this option minimizes extensive welding within the host pipe and
hence most workable. However, the Contractor submitted an alternative method that
was based on assembly of the liner inside the host pipe. The method did not require
pushing the assembled liner, rather it utilized adjustable jacking bolts to centralize the
pipe. Individual pipe segments were to be butt-welded from inside by using a steel
backing plate at the joints. Following the grouting of annular space, the jacking bolts
would be unscrewed and capped using steel plugs.
The primary advantage of the method submitted by the Contractor was that it allowed
adjustments in the liner on account of unforeseen bends or kinks in the host pipe. The
method also provided ability to align each segment individually by tightening or
loosening the jacking bolts. However, there were concerns pertinent to structural
integrity of the pipe due to concentration of stresses at jacking locations during
grouting. Since the Contractor had planned to perform annular grouting in a single
stage, the jacking locations at the top of the liner were potentially the most affected
due to high stresses resulting from floatation. Additionally, the method submitted by
the Contractor required annular grouting to be performed in sections using temporary
bulkheads. There were concerns that the shrinkage of the grout mass may allow
infiltration that could reach the surface of the steel pipe through these joints that
could potentially corrode the steel pipe.
During the review of Contractor’s submittal, the engineering team performed a stress
analysis using a 3-D model that concluded that the stress concentration on the top
jacking bolts location was about 25% higher than the allowable limits. It was
© ASCE
Pipelines 2015 1394
recommended to strengthening the pipe at these locations using an added plate. In

addition, secondary grouting was recommended to mitigate concerns due to shrinkage
and prevent any compromise in the corrosion protection of the steel pipe. The
Contractor agreed to modify the design and the installation was successfully
completed. Following the grouting, the liner was successfully inspected and pressure
tested.
Figure 3. Completed outlet structure with new gate valve and removable thrust blocks
CONCLUSIONS
The project was completed successfully with minimal design changes during
construction. The total value of the Change Orders paid to the Contractor on account
of these design changes was less than 5% of the total Contract Value, and well within
the contingency. Barring delays due to the bad weather, the project finished on
schedule and without any delays on account of changed conditions during
construction. This is particularly important since the Contractor had to work with
unknown conditions, and plan for contingencies ahead of the time to be on schedule.
The project was a well-managed team effort where the County actively participated
with the Engineer and the Contractor to pre-empt issues, risks and take timely
decisions to keep the project on track.
© ASCE
Pipelines 2015 1395
Among several factors that contributed to the success of the project, the key factors
were preparation of sound Contract Documents, identification and management of
risk, and extensive documentation during construction. This required extensive
engagement between the County and the Engineer during the design phase to discuss
potential options, risks and outcomes during construction. The engagement was
extended during construction to include the Contractor to be able to discuss the
project issues, and bring fresh thoughts on board to steer the project in the most
beneficial manner. The project execution not only catered to the current project needs
but also secured information for future by documenting critical project
communications, decisions during construction and field observations that will serve
as reliable records for future.
© ASCE
Pipelines 2015 1396
Design and Construction of a Raw River Water Welded Steel Transmission Main for a
New Water Supply System in Northern Virginia
Eric J. LaRocque1
1
Dewberry, 8403 Arlington Boulevard Fairfax, VA 22031-4666. E-mail:
elarocque@dewberry.com
Abstract
Construction is ongoing of a 100-year service life 5.25-mile 42- and 48-inch welded steel
pipeline, conveying raw river water from the Potomac River for a 40 MGD water supply system
in an urbanizing county in northern Virginia. Ductile iron pipe (DIP) and welded steel pipe
(WSP) were specified to create price competition. Transient modeling incorporated material-
specific celerity values and showed that vacuum due to surge was more severe than positive
surge. Air vacuum relief valves were located at each high point and 4 non-high points where
transient surges were predicted. Cathodic protection was designed, consisting of a bonded
coating and galvanic anodes. All bids received were for WSP. Lay schedule and shop drawings
review assessed alignment, joint types and pulls. Constrained easements prompted reduced radii
elbows. The heat shrink sleeve dissipates heat from joint welding after backfill. Daily lay
production of 250 feet is planned. The strength of single lap welds was considered. Each weld
is to be magnetic particle tested.
Evaluation of Six Materials Results in DIP and WSP Specified in Documents
Patterned after pipeline material evaluations often developed for large diameter water
transmission mains in the Western U.S., there's a growing trend for Owners and designers in the
Eastern U.S. to perform a similarly comprehensive evaluation for projects with a significant first-
cost investment, i.e. in the range of 60 to 70 percent of life cycle cost. For the Loudoun County,
Virginia raw water transmission (RWT) project, a robust evaluation outlined performance
requirements and assessed each material's ability to offer long-term reliability—an uninterrupted
100-year service life—at an acceptable, budgeted cost. Such an evaluation may challenge an
Owner's standards by bringing to bear current data and empirical evidence of failure modes,
availability, and constructability, for each pipe material. Design of the RWT project included a
comparative evaluation considering six pipe materials: ductile iron, welded steel, bar-wrapped
concrete, PCCP, HDPE and PVC. Each material was scored under the following criteria, in
priority order: i) total installed cost; ii) availability and demonstrated experience in required
diameters and pressure classes; iii) life cycle cost; and iv) failure mechanisms and history. An
example scoring for life cycle cost considered pump power consumption, which translates to the
pipe’s ability to remain smooth over its service life. Selection criteria scoring of the six
materials proved welded steel pipe (WSP) and ductile iron pipe (DIP) to be finalists for the
project, and were accordingly specified in the documents to take advantage of market
competition. The two variables that most affect the choice between DIP and WSP, while also
being the primary drivers of cost differences, were wall thickness and external coatings. Wall
© ASCE
Pipelines 2015 1397
thickness gives the pipe its principal resistance to structural failure while an external coating is
the primary means of corrosion protection from surrounding soil and externalities such as stray
current. While developing bidding documents, the engineer defined these two variables for each
pipe type to ensure Contractor bids are based on two equally performing materials.
Wall Thickness Design: DIP and WSP are manufactured differently. Ductile iron pipe is
centrifugally cast with the deLavaud process, while WSP is helically welded from coiled sheets.
DIP and WSP have similar ultimate and yield tensile strengths, but different elongation, hinting
at a key difference between the two materials: toughness. Generally, WSP’s higher elongation
translates into greater toughness compared to DIP. In turn, different toughness ranges lead to
material specific wall thickness design methods per AWWA, although methods for both consider
the material’s resistance to internal loads (pressure design) and external loads to arrive at a
required wall thickness. Owing to its long segment lengths, WSP wall thickness design also
requires a handling check to evaluate its beam strength. The design methodology for internal
pressure loading of both materials uses the Barlow Hoop Stress calculation to estimate wall
thickness required to resist the maximum expected hoop stress, which is a circumferential tensile
stress of greatest value along the pipe’s inner diameter. Application of the hoop stress equation
to WSP and DIP, per AWWA methods, introduces built-in conservatism such that computed
wall thicknesses results in a relatively small likelihood that the minimum yield stress will be
developed in the pipe wall due to static or surge pressure. For the RWT project, with surge
suppression devices in place—air release and vacuum relief anti-shock values at each alignment
high point, as well as four specific non-high point locations, and a surge relief tank on the
immediate discharge side of the proposed pump station—the maximum steady state pressures are
comfortably less than 200 psi closest to the pump station, and less than 150 psi for the majority
of the line; hence, pressure class 150 and 200 DIP is the analytical solution for wall thickness to
resist internal loading. Accordingly, the equivalent WSP wall thickness to resist these internal
pressures are 0.188 and 0.208 inches for 42- and 48-inch pipe, respectively, based on the
following AWWA-prescribed Barlow Hoop Stress calculation procedures for test and working
pressures: the allowable tensile stress, i.e. hoop stress, in the pipe wall during working pressure
can be up to half the steel’s minimum yield strength (0.5*42ksi), while that fraction during test
pressure is up to two thirds (0.67*42 ksi). These relatively thin-walled pipe solutions were
overridden in the final specifications by the Owner’s standard thick-walled pipe: DIP thickness
class 52 (0.59 and 0.65 inches for 42” and 48” DIP respectively), which translated to a WSP
thickness of 0.3125 and 0.355 inches for 42 and 48 inch pipe respectively, using the
aforementioned fractions of minimum yield strength for allowable wall hoop stress during
working pressure (0.5*42 ksi) and test pressure (0.67*42 ksi). To underscore a clear difference
between current standards of practice in Western and Eastern U.S. geographies—relatively thin-
walled transmission mains are common in the west as long as an external bonded coating and a
cathodic protection system are specified. However, this standard is not as prevalently accepted
for projects in the Eastern U.S.. The cost implications of these thick walled pipe specifications
may be highlighted by considering payment for pipe as dollar per pound of metal, versus the
conventional dollar per linear foot. Applied to the thin-walled pipe, this payment is dollar per
pound of metal required for performance, while payment for thick-walled pipe defined in the
project’s bid documents is dollar per pound of metal specified.
© ASCE
Pipelines 2015 1398
Two Sets of Vaults Designed: The time and temperature controlled annealing operation of the
deLavaud process causes DIP to be less suited to welded outlets, as welding heat stresses weaken
the metal microstructure. Consequently, structural failures of welded connections and outlets are
acknowledged. If welded connections or outlets are used, the manufacturer shall be consulted to
discuss limiting the heat affected area of the parent pipe, restraining lateral joints, and
minimizing moment arm loads onto the weld. It’s noteworthy that one DIP manufacturer
reinforces their welded appurtenances, similar to the reinforcing collar and wrapper plates used
on WSP welded appurtenances. The represented consulting Engineer has helped a DIP
manufacturer develop testing protocols for welded appurtenances to evaluate resistance to
moment arm failure, as opposed to the previously-held focus on axial thrust loads. DIP
manufacturers often require special thickness class 53 parent pipe when fabricating welded
appurtenances.
Savings Predicted if WSP Selected over DIP: Two independent factors governed the direct cost
of WSP and DIP for this project: wall thickness and exterior coating. Pertaining to 42-inch pipe,
Thickness Class 52 DIP—with 0.59 inch wall—generally is more costly than the equivalent
performing WSP wall of 0.3125 inches. As introduced above, the cost implications of additional
metal comparatively disadvantaged DIP. This disadvantage is compounded by the cost increase
imposed by the specified dielectric bonded coating for DIP versus WSP, as detailed in this
paper’s cathodic protection section. When pairing wall thickness and exterior coating
specifications, WSP was predicted by the Engineer’s estimates to be approximately $4M less
than DIP (in this case, a 15 to 20 percent savings in first cost for the entire project. This
anticipated savings was likely realized as all contractor bids were for a welded steel pipeline.
Moreover, a pipe manufacturer that offers both WSP and DIP was likely advantaged by volume
pricing when coupling the WSP pipeline project with DIP on the program’s water treatment plant
project. The adopted approach to create market competition between materials within the bid
documents seeks to optimize derived value to the Owner, as manufacturers are not only
competing within a material—for example, DIP manufacturer A versus DIP manufacturer B—
but also across materials—DIP manufacturer A vs. DIP manufacturer B vs. WSP manufacturer.
Ultimately, a pipe vendor who manufacturers both DIP and WSP earned the supply contract with
the general contractor.
Alignment Design Based on DIP Deflections, WSP Segment Geometry Customizable
The below figure is the detail shop drawing of a special segment of WSP: a 60 degree bend shop
welded on an otherwise straight length of pipe, which sharply contrasts the equivalent 2-piece
DIP arrangement: a straight piece jointed to a 60 degree bend fitting with mechanical restraint.
The WSP solution eliminates a field joint and mechanical restraint apparatus and hence expedites
laying production while reducing leak potential. Other than 90 degree elbows, which are
individual pieces, horizontal and vertical bends are generally accomplished in this manner with
WSP.
© ASCE
Pipelines 2015 1399
Figure 1: Steel pipe segment for 60 degree bend from American Spiral Weld Pipe
Company.
Steel’s geometric solution for this highly constrained alignment—both horizontally due to a
highly developed landscape and vertically due to crossing utilities, streams, and roadways—
relies on mitered bells and field joint pulls. Acknowledging the bid documents originally
prohibited field pulls at mitered bells, this restriction was lifted by the Engineer during lay
schedule development as requested by the manufacturer. This request stemmed from the
project’s distinctive need for multiple consecutive mitered joints to accommodate horizontal and
vertical curves. Pulling at mitered bells prevented an abundance of unique mitered bell shop
cuts, which would have sharply increased the cost of fabrication. While this provision was lifted,
the Engineer retained the stipulation of a maximum allowable five degrees per pipe joint of total
allowable deflection, which is the sum of miter and field pull angles. The Contractor is
especially appreciative of the flexibility for field fit when the Engineer permits field pull. This
five degree rule is rooted to this pipeline being a relatively low pressure installation, and hence
comparatively low induced thrusts when considering steel’s allowable wall stresses. Joints that
are both mitered bell and field pulled are well served to also be restrained, i.e. welded, as the
wall stress intensification linked to this five degree specification is minimized by the reinforcing
afforded by the weld. Furthermore, it’s notable that mitering of spigot pipe ends was prohibited.
Transient Modeling of Delivery System
Startup and future operation of the new water supply system’s four elements were integrated into
a steady state and transient hydraulic model to predict the pressure envelope faced by
transmission main piping. The following four flow modes were modeled: 1) river water pumped
from river pump station (RPS) to quarry storage; quarry water was then subsequently pumped by
quarry pump station (QPS) to water treatment plant (WTP); 2) river water pumped from RPS to a
flow split at control valve vault to deliver to both the WTP and quarry storage, with the QPS
simultaneously pumping to WTP; 3) river water pumped by RPS strictly to WTP; and 4) river
water pumped by RPS to a flow split at control valve to deliver to both quarry storage and WTP.
The RPS features 3 duty pumps, each rated at 9,236 gpm at 297 feet TDH, while the QPS
features 4 duty pumps, each rated at 6,950 gpm at 430 feet TDH. The target transient pressure
envelop was -7 to 250 psi, with -7 based on minimum pressure at a steel or DIP gasketed joint
and 250 psi based on an economical steel pipe wall thickness and DIP pressure class pipe;
moreover, applied conditions coupled the lowest TDH and highest C factors to yield maximum
velocities. Transient modeling input parameters included: i) pump and motor moments of inertia;
ii) fastest wavespeed (celerity) between WSP and DIP; iii) response time of pressure reducing
© ASCE
Pipelines 2015 1400
valve at the flow split vault and WTP influent valve; and iv) double orifice air-vacuum valves
and a surge tank at the RPS and QPS. Pump trips at the RPS and QPS were modeled separately
and simultaneously to establish the allowable pressure envelope. A total of 25 different transient
scenarios were modeled, resulting from the combinations and permutations of the described flow
modes and transient events (i.e., pump starts, pump trips, valve closures, etc.). Air-vacuum
valves’ location and size were adjusted iteratively until pressures fell within the desired pressure-
vacuum envelope. Refinement model runs were then performed with normal pump start/stop in
conjunction with valve closure at maximum and emergency rates. Thereafter, “hydraulic
capacity” model runs sought the maximum flow the system could deliver pressures while
maintaining the established pressure-vacuum envelope. Once aggregated, model results
informed the selection of pipe wall thickness, restrained joint lengths, and the size and location
of air-vaccum valves and hydropneumatic surge tanks, as presented by bid documents.
Conclusions that significantly affected design include the following: 1) abnormally high
pressures were comfortably addressed, while mitigating full vacuum conditions proved difficult;
2) the optimal size of air-vacuum valves was 8 inches, which is not the largest offered by the
manufacturer (12 inches); 3) slow valve closure rates (in some cases, up to seven minutes for full
closure) significantly mitigate transient responses; 4) once future demand triggers the need for
larger pumps at the RPS and QPS, their larger moment of inertia should improve the system’s
transient response; and 5) a surge tank at the RPS and QPS will be relied upon to minimize
vacuum conditions, i.e., the most threatening transient event. Several instances of cross-contract
coordination were incorporated into this modeling effort, e.g., the RPS surge tank was integrated
into the river intake and pump station contract, not the pipeline contract.
Robust Cathodic Protection System
The first step taken during design to determine the need for cathodic protection (CP) was
obtaining profiles of soil resistivity measurements along the proposed alignment and comparing
the results at the proposed pipe depth with benchmark soil corrosivity values. Widely used
sources of benchmark corrosivity data include the Bureau of Reclamation (dominant in western
U.S.), NACE/Corrosion Consultants (empirical; dominant in eastern U.S.), and the AWWA
Corrosion Control for Buried Water Mains Pocket Field Guide. Disparate resistivity values
across these sources that define corrosion severity categories, e.g., most severe soil environment,
engendered the need for Engineer judgment. For example, a 6,000 ohm-cm resistivity
measurement benchmarked against the Bureau of Reclamation’s guidelines suggests DIP and
WSP do not need cathodic protection—AWWA C105, Polyethylene Wrap and AWWA M11
would also suggest soil corrosion is of low likelihood; however, the NACE and AWWA Pocket
Field Guide classification regimes would indicate a corrosive and moderately corrosive soil
environment respectively, justifying a decision to cathodically protect the pipeline. Given the
100-year service life objective, the set of NACE classifications was adopted for this project, and
when contrasted against field soil resistivity measurements, nearly 83% of the data demonstrated
a corrosive to very corrosive soil environment. When paired with the stray current threat from
adjacent and crossing gas mains (which employ active impressed current systems), the need for
CP became paramount.
As the bulk of the pipeline progresses within an existing buried and overhead utility corridor,
namely natural gas mains and high voltage power lines, the Engineer was alert to stray current
© ASCE
Pipelines 2015 1401
and alternating current induction. While soil corrosion threatens long term viability of the
transmission main, the predominant rapid corrosion failure hazard emanates from stray current
from two nearby and crossing 30-inch high pressure natural gas mains, each configured with an
impressed current cathodic protection system. One main’s operating rectifier is rated at 47 volts
and 5 amps, while the second main features a rectifier on each end of the roughly 6,000 foot
length over which it parallels the RWT main—rated at 36 volts and 4 amps, and 45 volts and 3
amps. These large protection currents act as strong stray current sources, which was a primary
factor for the bonded coating specification. Should gas main protection current stray onto the
RWT main, rapid corrosion failure at anodic points where stray current exits the pipe
continuously would be expected. It's estimated that 20 pounds of metal loss can result each year
per amp of continuously exiting stray current at the anodic point of WSP or DIP mains. The
point at which the RWT main crosses beneath both gas mains invited special CP design: a 125
mil HDPE dielectric membrane above the RWT main and a new test station at each gas main
with a direct current (DC) decoupling device providing DC isolation and grounding up to a 3 volt
threshold. This decoupling device seeks to eliminate DC exiting the gas mains’ CP systems.
When potentially corrosive environments are detected, protection schemes pair exterior pipe
surface protection to an active impressed current CP system or passive galvanic anode CP
system. Generally, exterior protection feature polyethylene wraps and bonded dielectric
coatings. Cost and reliability must be carefully weighed, as prices for bonded coatings sharply
vary between WSP and DIP manufacturers. In contrast to WSP fabrication, which is conducive
to shop application, DIP manufactures typically use a third party supplier and applicator with
pipe changing hands from the pipe manufacturer to the coating vendor—this shift of liability
signals significant warranty restrictions for DIP manufacturers when conforming to a bonded
dielectric coating specification. The approximate incremental cost for DIP with a bonded
coating is estimated at over 10 times that of polyethylene encasement. For a 5.25-mile long
pipeline, this differential scales to a considerable sum. As extensively studied by the National
Academy of Sciences based on miles of ductile iron water, gas, and oil pipelines, when it comes
to reliable corrosion protection of DIP mains, polyethylene wraps enhance longevity when
compared to bare pipe, but bonded coatings are generally viewed as superior in this regard. This
is a hotly contested issue in the transmission main marketplace. Based on convincing large
empirical data sets of transmission main condition assessments and failure investigations, e.g.
20- to 40-mile 54- and 60-inch pipelines, practically no failures were evident when a bonded
coating was paired with CP. In turn, this informed the Engineer’s decision to specify three
allowable bonded coating systems—3 layered tape, polyurethane, and polyolefin—and galvanic
anode CP, which centrally features sets of 20-pound magnesium anodes buried adjacent to the
pipeline at trench bottom. During construction, the Contractor sought to raise anode depth to
pipe springline out of concerns that the anodes would hinder satisfactory backfill at the pipe
haunches, which in turn, may generate long-term structural vulnerability to external loading, i.e.
pipe deformation; however, anode placement at trench bottom was enforced, as justified by the
need for anodes to be in moist soils in order to yield sacrificial protection current to pipeline
cathodic surfaces.
© ASCE
Pipelines 2015 1402
Field Welded Joints

Implementing field welding in full compliance with the Engineer’s specifications was paramount
importance, particularly since this was the Owner’s first WSP transmission main project. In fact,
field welded joints garnered the focused attention of the Owner’s executives. Personnel directly
responsible for field welding were to be qualified in accordance with American Welding Society
(AWS) D1.1 for structural steel or ASME, and included: the Contractor’s field welder, the
Contractor-hired independent AWS Certified Welding Inspector (CWI) who executed magnetic
particle testing (MPT), and the Owner/Engineer’s welding specialist. The project gains
maximum benefit if the field welder performs in full and strict accordance to the specifications;
accordingly, a submittal was required on the Welding Procedure Specification (WPS)—
standards for double- and single-lap welds to which the field welder must demonstrate the ability
to weld. Given that the steel pipe coil was Grade C material and since there’s no American
Welding Society (AWS) pre-qualified WPS for this material, the WPS needed to be expressly
certified for Grade C. The Contractor produced certified WPS by invoking the steel pipe
manufacturer’s welding procedure. Upon Engineer’s approval of the WPS, the field welder
demonstrated the ability to weld to these standards by submitting test results conducted at and
witnessed by an inspection services laboratory for:
1) 1-inch thick “V” groove welding of Grade C material with the electrode’s current,
voltage, and travel speed ranges tabulated; this WPS automatically qualifying him for
field fillet welds as the skill demanded by groove welding surpasses that of fillet
welds; results from tensile break and bend tests were offered as the weld Procedures
Qualification Record (PQR); tested specimens’ ultimate strengths exceeded 60 ksi
and ruptured in the parent base metal, rather than the weld material; therefore, this
PQR was accepted; and
2) 5/16-inch fillet weld of 48-inch Grade C pipe, with the electrode’s current, voltage,
and travel speed ranges tabulated.
Strength of Singe-Lap Joint Welds: While the specifications refrained from prescribing
restrained joints be single or double fillet lap welded, the Contractor choose to exclusively
pursue interior single lap welds, with the exception of double lap welds of carrier pipe at
trenchless crossings. The Engineer pre-evaluated the strength of single lap welds according to
the American Welding Society’s (AWS) Effective Throat Length methodology, which accounts
for of the project’s test pressure, pipe wall thickness, and river water temperature range. This
methodology demonstrated the allowable longitudinal force in the pipe wall exceeded the
maximum anticipated value—the anticipated value is the sum of thermal and Poissons’ stresses,
while the allowable value is predicated on a grade C electrode producing a stick weld of 21,000
psi since weld material must afford yield and ultimate strengths equal to the parent metal. A
complete outlook requires consideration of two alternate methods for assessing strength of single
fillet welded joints: i) ASME’s Pressure Vessel (PV) Joint Efficiency; and ii) Joint Eccentricity.
As long as the fillet weld is the full leg dimension, the AWS methodology is acceptable for
evaluating thrust loads. The project’s specifications are consistent with this approach by
requiring steel coils be of a wall thickness with a zero minus tolerance. The ASME PV approach
© ASCE
Pipelines 2015 1403
scales the steel yield strength down by an efficiency coefficient, independent of the AWWA –
prescribed maximum wall stress of 50% yield. The Joint Eccentricity approach is central in the
current debate of single fillet weld strength since it is focused on the bending moment in the
weld material resulting from the inherent eccentricity of the thrust load path between the pipe
bell and spigot via the fillet weld. This approach imposes the bending rotational stress, creating
a significant load for the fillet weld to withstand; for typical pressures, this moment-induced
stress causes total wall stress to exceed allowable levels. This vulnerability has prompted some
design engineers to strictly specify double lap weld restraint for all steel pipe projects.
Contractor Elects Weld After Backfill for Single Lap Interior Welded Joints: To not hamstring
the Contractor by prescribing means and methods—which helps derive market value in a
competitive bid—the Engineer intentionally did not specify the pipe laying-backfill-joint
welding sequence, i.e. joint welding after or before backfilling the pipe. In the context of the
Contractor’s planned daily pipe laying production rate of 200 to 300 feet, joint welds are
performed a variable length of time after the pipe is laid and backfilled; consequently, welding
occurs after the heat shrink sleeve is installed. Thus, the heat shrink sleeve’s ability to sustain a
100-year service life depends on its ability to dissipate heat borne from interior joint lap welding.
This sleeve is installed on the exterior of each joint to shield the pipe from soil corrosion; it’s
noted that the majority of the 5.25-mile alignment traverses through aggressive soils. During
submittal review, the Engineer ensured the sleeve manufacturer was aware that interior joint
welding would be performed with the sleeve previously installed, since an often used, less costly
sleeve is strictly for pipe that has not yet been backfilled; in particular, the sleeve’s engineering
properties sustain high temperatures—namely, its softening point is 401 degrees F. The
Engineer’s field quality control inspector was alerted to an incorrect grade B electrode used by
the Contractor’s welder prior to the first field joint weld; once this error was recognized and
communicated, the Contractor’s welder demobilized and returned with the correct grade C
electrode. It’s noteworthy that grade B steel features an average of 36 ksi yield strength, while
grade C features an average of 42 ksi.
Magnetic Particle Testing (MPT) of Welded Joints: Quality control and Contractor oversight of
the more than 700 field pipe joint welds was a major point of emphasis for the project. With the
weld material magnetized by the electromagnet instrument, small leakage in the magnetic field
develop across cracks in the weld material. A colored powder of magnetized particles placed by
the inspector along the joint are concentrated at and held by these leakage fields, enabling the
inspector to visually determine the presence of cracks. From the MPT instrument manufacturer
chosen by the inspector, Parker, the D400 model used on this project is a two-pole electromagnet
that applies an AC or DC induced magnetic field in the joint weld material. The physical
connection of the pipe and instrument creates a magnetic flux path, with the weld material
becoming highly magnetized. From within the pipe, the inspector passes the instrument
circumferentially along the joint to detect defects (surface and near-surface cracks) in the joint
weld material. AC magnetization is advantageous for surface breaking cracks, while DC
magnetization is best for near surface cracks. Welds for this project were tested with both AC
and DC magnetization. The instrument is powered by an 120 V AC chord routed through 3-inch
weld lead outlets or manway outlet appurtenances, and either passes 6 amps of AC current onto
© ASCE
Pipelines 2015 1404
the pipe weld material or converts the AC to DC via internal electronics as activated by the
operator. Specifications called for carrier pipe double weld joints at trenchless crossings, i.e.
running within casing pipe, to undergo MPT on the interior and exterior welds; as a leak within
the casing would go undetected, carrier pipe joints at trenchless crossing were to be double lap
welded. The Contractor initially disputed the interior and exterior MPT, claiming testing
exterior welds would impede pipe lay production; the Engineer choose to enforce the double
weld MPT specification. Specifications required a Contractor-hired MPT technician, certified at
the proper American Society for Nondestructive Testing Central Certification Program (ACCP)
non-destructive testing MPT level.
Seismic Loading Increases Axial Stress in Pipe Wall: During design, the incremental increase
in pipe wall stress due to seismic loading was assessed. Within the quarry storage water banking
framework, approximately a third of the pipeline alignment is offset from a planned future
quarry by nearly 50 feet. Consequently, future quarry blasting loads are expected to impart onto
the pipe. Ground accelerations translate into strain in the traverse direction of the pipeline;
through Poissons’ relationship, this imposed strain manifests as an incremental rise of axial stress
in the pipe wall. This axial, or longitudinal, stress was factored into our assessment of the
strength of single lap welded joints.
Construction Schedule: WSP Production Rate Verses Typical DIP Production Rate
The Contractor’s construction schedule is predicated on a daily pipe lay production of 250 feet is
planned, far exceeding the typical DIP rate of 100 feet. This creates value for the project overall
by helping reduce delay risk imposed by unanticipated changes, namely changed field
conditions. For example, initial pipe laying has been hindered by repeated problems with the
onsite “rock crushing” operation through which native heterogeneous excavated material
undergoes physical processing to produce a more compactable backfill. The ensuing delay risk
is mitigated by the 250-foot daily lay rate afforded by comparatively long steel pipe segments—
typically 50 feet.
Special 90 Degree Elbow Fittings
In three locations, constrained easement prevented use of standard 90 degree horizontal steel
pipe elbows, which according to M11 are standardized with a minimum radius of 2.5 times outer
pipe diameter, i.e., 2.5 D. This issue emerged during development of the steel pipe lay schedule
by the steel pipe manufacturer—as DIP 90 degree bends are of a much tighter geometry, with a
radius of about 0.7 times pipe diameter, the need for specially fabricated steel pipe elbows was
not realized during alignment design, since it was based on DIP. The easement afforded the lay
of 90 degree elbows with: i) 1.5 D; ii) 1.25 D; and iii) 0.75 D. Stress intensification calculations
for the elbows’ inner diameter walls were performed for these three geometries under a pressure-
thrust loading condition. Elbow geometry must also consider welding borne stresses as the
segment length between shop welds reduces. Stress calculations proved the specified wall
thickness (0.3125 inches for 42 -inch pipe) satisfactory for the 1.5 and 1.25 D cases, while the
manufacturer chose to roll 0.5 inch plate for the 0.75 D elbow. This highlights an unforeseen
advantage of otherwise unnecessarily thick walled pipe.
© ASCE
Pipelines 2015 1405
Conclusions
Design and construction of this Northern Virginia 100 year service life, non-redundant, raw
water pipeline has required the pinnacle level of evaluation, analysis, and quality control. The
material evaluation applied to this project contrasting WSP versus DIP, predicted WSP to be less
costly based on the wall thickness and bonded coating specifications. This prediction was
affirmed as all received bids were for WSP. During design, a series of material-specific
considerations was captured by the Engineer’s bid documents, including: i) vault systems and
appurtenances unique to WSP and DIP; ii) alignment geometry to accommodate joint deflection
capabilities; iii) transient modeling to size and position anti-surge valves and surge tanks at the
high service pump station, which incorporated separate WSP and DIP celerity values; and iv) a
robust passive sacrificial anode cathodic protection system to counter native soil and stray
current –induced corrosion. Horizontal and vertical geometry design was completed in particular
detail given the high land values associated with the fully developed geography traversed by this
pipeline. The utility corridor character of its alignment introduced significant stray current
corrosion potential from parallel high pressure gas mains. Construction quality control has been,
and continues to be, essential for this high profile project. For WSP pipelines, the Owner and
Engineer must be experienced with magnetic particle testing of interior and exterior field welded
joints. An impressive number of field welds are required for this project as the alignment
features a myriad of horizontal and vertical geometry changes to position the pipe within the
existing utility corridor. Moreover, the strength of single lap field welds should bear on decision
makers when designing and constructing WSP pipelines.
References
American Water Works Association. (2004). "Steel Pipe - A Guide for Design and Installation
(AWWA M11) (4th ed.)” Denver, CO: American Water Works Association.
American Water Works Association. (2008). “Thickness Design of Ductile-Iron Pipe (AWWA
C-150) (8th ed.).” Denver, CO: American Water Works Association.
American Water Works Association. (2009). “Ductile-Iron Pipe and Fittings (AWWA M41) (3rd
ed.).” Denver, CO: American Water Works Association.

Branch, Darren and LaRocque, Eric (2013) “Metal Wars: Welded Steel vs. Ductile Iron for New
Large Diameter Raw and Finished Water Transmission Mains, A Comparative Evaluation and
Field Case Studies.” Virginia Water Joint Annual Meeting Conference Presentation.
Brice, Mary, Langston, Joe, and LaRocque, Eric. (2013). “Metal Wars: Welded Steel vs. Ductile
Iron for New Large Diameter Raw and Finished Water Transmission Mains, A Comparative
Evaluation and Field Case Studies.” American Water Works Association North Carolina
Conference Paper.
© ASCE
Pipelines 2015 1406
Committee on the Review of the Bureau of Reclamation's Corrosion Prevention Standards for
Ductile Iron Pipe. (2009). Review of the Bureau of Reclamation's Corrosion Prevention
Standards for Ductile Iron Pipe. Washington, DC: National Academy of Sciences.
Ductile Iron Pipe Research Association. (n.d.). The Pipeline: Ironclad Facts from DIPRA.
Retrieved from http://www.dipra.org/content/uploads/BondedCoatingsVSPolyEncasement.pdf.
NACE International. (2007). “Control of External Corrosion on Underground or Submerged
Metallic Piping Systems NACE SP0169-2007”. Houston, TX: NACE International.
Parker NDT Equipment. “Portable Contour Probe Instruments and Kits. DA-400 Series Contour
Operating Instructions”. Retrieved from http://www.parkerndt.com/productsprobes.htm.
Raghavender V. Joshi, Cronin, Roger, Pearson, Glenn, and Bundschuh, Carl. (2014) “Hydraulic
Analysis of a New Water Supply System.” Virginia Water Joint Annual Meeting Conference
Presentation.
Romer, A. E., & Bosserman II, B. (2009). “Corrosion Control for Buried Water Mains Pocket
Field Guide”. Denver, CO: American Water Works Association.
Smith, Gordon E. and Aschemeier, Uwe. Yoke MT: Part I—A How to Guide for Inspectors.
American Welding Society. Retrieved from http://www.aws.org/w/a/itrends/2005/04/015.
© ASCE
Pipelines 2015 1407
Lessons Learned in the Design, Manufacture, Shipping, and Installation of the

108-inch Integrated Pipeline (IPL) Section 15-1
Robert J. Card, P.E.1; Ed Weaver2; Randall Payton3; Richard Mielke, P.E.4; and Shah Rahman5
1
Chief Pipe Engineer, LAN, 2925 Briarpark Dr., Suite 400, Houston, TX 77042. E-mail:
rjcard@lan-inc.com
2
Asistant Division Manager, Tarrant Regional Water District, 808 East Northside Dr., Fort Worth,
TX 76102. E-mail: ed.weaver@trwd.com
3
Asistant Director of Water Operations, Dallas Water Utilities, 1500 Marilla St., Dallas TX 75201.
E-mail: Randall.Payton@dallascityhall.com
4
Director of Engineering, Northwest Pipe Company, 5721 SE Columbia Way, Suite 200, Vancouver,
WA 98661. E-mail: rmielke@nwpipe.com
5
Director – Trenchless Technology and Pipeline Rehabilitation, Northwest Pipe Company, 351
Longhorn Rd., Saginaw, TX 76179. E-mail: srahman@nwpipe.com
Abstract
The Tarrant Regional Water District (TRWD) and the City of Dallas Water Utilities (DWU)
are engaged in the planning, design and implementation of a 350 MGD raw water transmission
system. Referred to as the Integrated Pipeline Project (IPL), the system consists of approximately
150 miles of 84- to primarily 108-inch diameter pipeline, a 5 mile 120-inch diameter tunnel, six 100–
350 MGD pump stations, one 450 MG balancing reservoir, and ancillary facilities. When complete,
the “integrated” system will provide a critically important source of water for the rapidly growing
Dallas-Fort Worth Metroplex for the next five decades and beyond. After years of study, followed
by design and construction of a 2 mile, 108-inch steel water pipe demonstration project, the Line J
Project, the first segment of the pipeline, Section 15-1, bid in February 2014 and was awarded in
March 2014. Steel pipe was chosen for the construction. Pipelines of this diameter and length are
not common and come with their own set of challenges from the design, manufacture, shipping and
installation standpoints. This paper will review design of this piping material and the practical
aspects of furnishing and installing a line of this magnitude, with particular emphasis on lessons learned
on both the Line J and the Section 15-1 Projects from the view point of the Owners, Engineer and
Manufacturer.
INTRODUCTION
The Tarrant Regional Water District (TRWD) with the City of Dallas Water Utilities
(DWU), are currently engaged in the construction of the beginning phases and planning, design and
implementation of a 350 MGD raw water transmission system, which will run across north central
Texas from Lake Palestine to Lake Benbrook, with connections to Cedar Creek Reservoir, Richland
Chambers Reservoir, and a Dallas delivery point. Collectively, the system consists of approximately
150 miles of primarily 108-inch pipeline, with some sections of 84-inch diameter pipe, a 5 mile 120-
inch diameter tunnel, six 100–350 MGD pump stations, one 450 MG balancing reservoir, and
1
© ASCE
Pipelines 2015 1408
ancillary facilities. The joint program developed by TRWD and DWU is called the Integrated
Pipeline Project (IPL). The project-location in shown in Figure 1.
Figure 1: Project Location
TRWD and DWU currently provide drinking water to an estimated 4.4 million people.
Based on developments and updates of the City of Dallas and Texas Water Development Board
long range planning studies conducted in 2005-2006, it is predicted that population and water
demands are likely to double in the next 50 years. The IPL project is developed to provide an
additional 350 MGD supply to meet these growing needs. The project is being developed in five
distinct phases with completion of Phase 1 (70 miles of 84-inch to 108-inch pipeline, a 350 MGD
booster pump station, three interconnection facilities, a 450-MG terminal storage balancing
reservoir, and ancillary facilities) in 2018, and Phases 2 thru 5 by year 2035. Design of pipelines and
facilities also took into consideration the potential for future expansions.
SECTION 15-1 OVERVIEW
Section 15-1 of the IPL project consists of a portion of mainline pipeline, a 5,400 sq-ft
interconnect facility, and the lowering of an existing 90-inch water line, all to be constructed as part
of Phase 1. The project spans from the site of future pump station JB2 west to the interface point
with section 15-2 all within Navarro County, Texas, Figure 2.
Figure 2: Location of Section 15-1
2
© ASCE
Pipelines 2015 1409
The pipeline consists of approximately 15.5 miles of 108-inch diameter pipe. It also includes
seven tunnels (90-ft to 450-ft drive lengths) crossing various highways, a Union Pacific Railroad
crossing, existing gas utility and a creek. A 5,400 sq-ft interconnection facility is located
approximately 6 miles west of the future JB2 pump station. The facility includes twelve 42-inch
butterfly valves and 78-inch interconnection piping. This provides pressure reduction and
connection from the new IPL system and the existing TRWD 90-inch diameter PCCP Richland
Chambers pipeline. The overall project included a relocation / lowering of approximately 2,000-ft of
the existing Richland Chambers water pipeline with 90-inch steel pipe which was constructed as part
of a Construction Manager at Risk (CMAR) project in advance of the mainline project.
Project design was conducted from early 2011 until late 2013; design commenced with a
route alignment refinement, continued with coordination among other program consultants
(topographical and land survey, environmental/archaeological permitting, geotechnical investigation
and other detail design teams), advanced with preliminary engineering and ended with final detail
design of improvements required to complete the project.
The project was bid from Dec 2013 to February 2014 using a Competitive Sealed Proposal
process. This process allowed selection of successful proposer considering the best value to project
owner. The project bidding included four options for contractors to select in offering their proposal
– options included choice of pipe material (steel pipe or prestressed concrete cylinder pipe) and
choice of embedment type (compacted granular material or native soil Controlled Low Strength
Material, CLSM). Six proposals were received ranging from $92.0 million to $127 million. A
construction contract was awarded to Garney Companies, Inc. of Kansas City, MO in March 2014.
Northwest Pipe Company of Saginaw, TX provided the steel pipe.
Construction began in June following mobilization and preliminary activities. The

Contractor utilized three pipe headings to complete mainline installation in April 2015. Completion
of appurtenance build out continues with anticipated project completion by June 2015.
DESIGN
While the design of steel pipe might appear to be very straightforward, when utilizing
AWWA M11 as a design guide, the steel pipe design for the IPL pipeline was quite complicated.
The initial IPL program design criteria manual and specifications were compiled via a joint effort of
their initial consultants on the project. These two documents were given to all of the individual
design teams for the eight pipeline contracts that make up the IPL pipeline. The original thought
process was that each of the eight sections would therefore have similar pipe material specifications
to assure equal performance and keep the pipe material competition level. However, early on in the
specification process, it became clear that each consultant desired to utilize their own company
“standard” specification for the pipe material options. Moreover, there were discrepancies between
the design criteria manual requirements and the pipe material specifications that the consultants
were instructed to use. Because of the discrepancies between the two documents and the desire of
each consultant wanting to utilize their own familiar pipe material specifications, the IPL program
management stepped in and demanded the use of their original standard specifications unless they
contained a “fatal flaw” that would prevent any of the individual engineers on each segment from
signing and sealing the specifications for their section.
3
© ASCE
Pipelines 2015 1410
After resolving several rounds of “fatal flaw” issues in the design criteria manual and
specifications, two unified documents were developed for the consultants to use on their individual
sections. As the LAN team was designing Section 15-1, which would be the first actual pipeline
segment to be bid, awarded and built, the unified documents formed the base for the design and
specification of this initial section.
108-inch diameter pipe would be considered “large diameter” by any pipeline designer. As
such, design or manufacturing elements utilized in typical 24-inch to 60-inch diameter pipe might
not scale very well to this large size.
One example of this scaling up process is the placing of the cement mortar lining on the
interior of steel pipe via the shop centrifugal spinning method. While it is routine to shop spin
cement mortar line 24-inch through 96-inch diameter pipe, there have only been very limited
amounts of steel pipe produced greater than 96-inch diameter with shop spun cement mortar lining
that could be documented. After researching the available projects that had been supplied by the
various steel pipe manufacturers with shop spun cement mortar linings on 96-inch and larger
diameters, it was decided to specify shop spun cement mortar lining for diameters up to and
including 108-inch. But, as a change to the AWWA C205 requirements for cement mortar lining,
the IPL program specified the lining to be ¾-inch thick. This additional thickness was added to
address water chemistry and quality issues to meet the 100 year design life. The thicker cement
mortar lining also slightly enlarged the outside diameter of the steel cylinder to account for this
greater lining thickness and still provide the 108-inch inside dimension after lining. Additionally, for
any nominal diameters larger than 108-inch, the cement mortar lining was specified to be field
applied after installation.
Pipeline Section 15-1 contained internal pressure classes up to 225 psi working. When using
the typical grade 42 steel (minimum specified yield strength of 42,000 psi), and the program-
specified minimum thickness for handling (D/t) ratio of 230, a resulting pressure class of
approximately 182 psi is derived. Therefore, for the higher pressure classes contained in this
section, an increased steel cylinder thickness would be required. Recognizing that the cost of the
pipe is the major driving force behind the overall cost of the project, grade 46 steel was suggested.
Increasing the grade to 46,000 psi minimum specified yield strength also increased the working
pressure that the minimum cylinder thickness could be designed to hold up to approximately 200
psi. This slight increase in grade of steel would result in 9.5% cost savings in the steel for the class
200 and 225 psi pipe. Nationally water pipe projects with steel grades over 42,000 psi yield strength
that are shop-applied cement mortar lined are somewhat limited and they are typically in the smaller
diameter range. TRWD did however have considerable successful experience using 46,000 psi
minimum yield strength steel on their multi-mile 96-inch and 84-inch Eagle Mountain projects. After
additional discussions between pipeline design professionals, including those on the AWWA Steel
Pipe Committee, along with the program’s consultants and the pipe manufacturers, it was decided to
allow the use of 46,000 psi yield steel on the project. Safeguards were written into the specifications
to limit cement mortar lining damage during handling, transport and installation. Moreover the
specifications required a field hydrotest pressure in excess of the specified working pressure classes,
another safeguard for the owner to receive a pipeline with the desired 100 year design life.
Smaller diameter steel pipe (with sizes up to and including 72-inch diameter) typically have a
push on rubber gasketed O-Ring joint for the standard joint and a single welded lap joint for the
restrained joint. Since the 108-inch diameter exceeded that of any manufacturer’s standard O-Ring
4
© ASCE
Pipelines 2015 1411
joint, a single lap welded joint was chosen for the desired field joint. Single welded lap joints are
designed to function with surge pressures of 150% of the working pressure or Pipe Class and for full
thrust conditions. Deep welded bell joints were specified at regular intervals to prevent the buildup
of excessive thermal stresses during installation. Another safety factor against this thermal stress
buildup was to allow the Weld-After-Backfill (WAB) procedure, which also results in cost savings to
the owner. Because both of these limit the total thermal stress in the pipeline, a full thickness,
single-welded lap joint was required to satisfy the design. As with any pipeline, select field joints will
be flanged along with electrical isolation joints at the proper connection locations.
Several options on the bedding and backfilling of steel pipe were offered for both the typical
cover conditions (5-ft minimum up to 18-ft of fill over the top of the pipe) in the trench details. In
general existing soils were clay or fatty clay materials along the 15 mile Section 15-1 alignment which
is typical of the soils along the majority of the IPL project. The IPL program had invested in testing
various bedding and backfill scenarios that included large diameter steel pipe embedment in various
materials along with testing the on-site manufacture of CLSM out of the native soils. These
demonstration projects showed that native soil CLSM provides an excellent embedment and
structural support for flexible pipe materials. However, these research and development projects
also demonstrated a much slower production rate in pipe laying, embedment and backfill than what
can be realized when using more standard construction methods. This was proved out by the actual
bid results as Garney, and all but one bidder, used the specified “granular embedment option C” in
their proposal, Figure 3.
Figure 3: Section 15-1, Trench Section Option C
In Trench Section C, Figure 3, the “c” dimension was 0.7D, with the “a” dimension = 6-
inch and the “b” dimension = 24-inch for a total trench width of OD + 4-ft. The granular
5
© ASCE
Pipelines 2015 1412
embedment is classified by the specification as cohesionless material such as crushed stone, pea
gravel, river rock, or gravel embedment (GW, GP) or sand embedment (SP, SW), 100% passing a
½-inch sieve all compacted to 95% maximum density.
STEEL CYLINDER MANUFACTURE
Steel water pipe cylinders were manufactured per AWWA C200 standard and project
specifications. The cylinders were made from coil steel, Figures 4a and b that were pulled through
buttress roles, edges prepared for welding, and then spirally free formed into a cylinder that is
submerged arc welded both on the inside, Figure 5a, and the outside, to produce full penetration
welds.
Figures 4a, b: Coil Steel
The cylinders were then cut to typical 50-ft sections. Ends were precisely prepared for lap
welded field jointing, Figure 5b. Pipe cylinders with prepared ends were then hydrostatically tested,
Figure 5c, to a pressure that engaged 75% of the minimum specified yield strength of the steel. For
the 108-inch cylinders on Section 15-1, the applied hydrostatic pressure was 267 psi. This
hydrostatic pressure is typically equal to the design surge pressure and in effect provides a “proof of
design” test against leaks for each section of pipe.
Figures 5a, b, c: Interior Submerged Arc Welding, Expanding Bell-End of Cylinder,

Hydrostatic Testing of 108-inch Cylinder to 267 psi
Quality Control specified by TRWD for the steel cylinders were stringent and matched
closely with Northwest Pipe Company’s ISO 9001 standards. These requirements included material
selection and verification, welding, dimensional tolerances and all forms of destructive and non
destructive testing. A “lesson learned” is that permanent marking or stenciling of the pipe cylinder,
6
© ASCE
Pipelines 2015 1413
Figures 6a and b, can be used to identify each section of pipe as installed in the trench and tie the
specific pipe back to the pipe manufacturer’s QA records. TRWD would then be able to incorporate
this installed information into their GIS and SCADA systems so as to be able to identify each
section of pipe by location and also to be able to trace all material, dimensional and manufacturing
QA records in the future as needed.
Figures 6a, b: Stenciling on Interior on Spigot-End of Each Cylinder
CEMENT MORTAR LINING APPLICATION
Cement mortar lining was applied per AWWA C205 and project specifications. Details of the
cement mortar specifications were discussed previously in the Design discussions. Specifications
required ¾-inch minimum thickness of cement mortar. A number of lessons learned were identified
during the project, listed below.
1. ¾-inch thickness can be applied in 108-inch and larger pipe with modifications to the
cement mortar lining equipment and amount of spin time. Shrinkage cracks were no greater
than that of AWWA C205 for ½-inch lining, and well within allowances. The extra lining
added considerable weight to the pipe. The extra ¼-inch thickness was necessary due to the
aggressive chemistry of the raw water that the pipeline would be conveying.
2. TRWD previously built the 108-inch Line J demonstration project where it was shown that a
number of experienced steel water pipe suppliers could manufacture, handle and ship 2 miles
of 108-inch cement mortar lined and polyurethane coated pipe over great distances. The
qualified manufacturers provided considerable competition for pipe material pricing on
Section 15-1 as the project that came in approximately 20% under budget estimate. Figures
7a and 7b show the lining of 25-ft and 50-ft sections of cylinder, respectively.
3. Stulling or bracing of the 108-inch cement mortar lined steel pipe received considerable prior
study. The primary purpose of the stulls is to keep the pipe and cement mortar lining from
excessive “flexing” during handling, shipping and jointing the pipe in the trench. From the
Line J project experience, IPL settled on a stulling arrangement for the 108-inch pipe that
was in excess of what is typically used in smaller diameter pipe but appropriate for the fact a
50-ft section of 108-inch pipe weighs in excess of 43,260 pounds. The standardized stulling
configuration worked well during the installation of the Section 15-1 project. Figures 8a and
8b show the finished interior surface of 108-inch diameter cement mortar lined cylinder
prior to stulling, and installed 6-point stulling, respectively.
7
© ASCE
Pipelines 2015 1414
Figures 7a,b: Cement Lining of 25-ft Long Cylinders, 50-ft Long Cylinders
Figures 8a, b: Finished Interior of Lined Pipe, Installed 6-Point Stulling
POLYURETHANE COATING APPLICATION
Polyurethane coating was applied to pipe per AWWA C222 standard, modified to minimum 35 mil
thickness. Lessons learned are listed below.
1. Original Line J project specifications required extended time between applying polyurethane
and ability to holiday test the coating during application. The Holiday test, Figure 9a,
provides a DC current directly to the coating at a voltage of 100 volts per mil of specified
coating thickness, or in this case, 3500 volts. Any voids or pinholes (referred to as holidays) in
the coating would be identified and repaired immediately. It was learned during pipe making
for the Line J project that per manufacturer’s recommendations, the holiday test could take
place while the pipe was still on the “lathe” and within a few minutes after application. This
avoided “over handling” of the pipe for the purpose of holiday testing.
2. To reduce the temperature of the polyurethane coating during the hot Texas summer days
where temperatures can easily increase above 100 deg. F, it was decided to utilize an “off
white” polyurethane coating color in lieu of the standard Northwest Pipe Company dark
blue coloring. Lower temperatures may have long term benefits for the pipe coating’s
desired 100 year design service life.
3. Aromatic polyurethanes used for buried pipe coating can be expected to “chalk” or lose the
top 2-3 mils of coating thickness during extended ultraviolet (UV) exposure to the sun.
Often, applicators apply additional thickness to make sure the specified minimum 35 mils of
8
© ASCE
Pipelines 2015 1415
coating are applied. While the additional thickness “is not by design,” the additional mils that
were applied appeared to more than address any coating loss concerns from chalking. It
should also be noted that the specified minimum coating thickness of 35 mils is considerably
greater than the minimum 25 mil guideline in AWWA C222.
Figure 9a, b: Holiday Testing, Coated Pipe Just Off the Lathe
SHIPPING AND HANDLING
Shipping and handling of 108-inch pipe warrant special precautions due to the dimensions,
weights and safety considerations. Following is a list of the lessons learned in this arena.
1. 50-ft sections of pipe were shipped with 3 sets of curved and padded bunks on the truck bed
that corresponded to the placement of the interior stulls. Ends of the pipe were “capped,”
Figure 10, which helped continue the cement mortar lining curing or hydration process,
which in turn limited the cracking of the cement mortar during shipping and handling to
meet requirements of AWWA C205.
Figure 10: 50-ft Section of Pipe on Truck

w/ End Cap
2. Pipe should be unloaded with nylon straps or equal and with two properly placed pick
points. This is accomplished with spreader bars, Figure 11a, two pieces of equipment lifting
the pipe simultaneously or by other means. Pipe should be placed on earthen berms, Figure
11b, or similar to facilitate picking pipe up to install. End caps should remain on until
installation.
9
© ASCE
Pipelines 2015 1416
Figures 11a, b: Nylon Straps and Spreader Bar to Move Pipe, Earthen
Berms for Pipe Placement
3. Handling the pipe during installation in the trench, Figure 12, again requires nylon straps or
equal and again with properly placed pick points. Stulls should remain in pipe during this
operation to limit “flex” of the pipe cylinder and cement mortar lining and to aid in keeping
pipe as round as possible during joint make up. This facilitates joint fit up and equalization
of the gap for full fillet field welding.
Figure 12: Pipe Handling in Trench
INSTALLATION
There are likely a number of lessons learned from both the TRWD Line J project and Section
15-1 project’s installing contractors could share but this paper will focus on just a few key lessons
learned.
1. The proposed use of CLSM as a cost effective alternative to imported granular materials on
Section 15-1 proved not to be the case. Key element was the time required for excavation,
10
© ASCE
Pipelines 2015 1417
staging, mixing of materials and cure time. This has been the case on the following IPL
projects bid to date.
2. Attention should be paid to the haunching of granular material under any large diameter
pipe, whether flexible or rigid. It is important that granular material be placed in the “pie
shaped” haunch section under the pipe all the way to the invert of the pipe. Voids under the
haunches have potential to be problematic for all pipe materials. Use of granular material,
placed in controlled lifts combined with suitable compaction such as a “compaction wheel”
is producing good results on Section 15-1, Figures 13a and 13b.
Figures 13a, b: Compaction Wheels, Application of Bedding

Material in Haunch Zone in Lifts and Using Compaction Wheels
3. Heat Shrink Sleeves are applied after joint is made up when using the specified Weld-After-
Backfill (WAB) method, Figure 14a. Heat shrink sleeves are industry standard practice whose
installation procedures are well known. Attention should be taken to make sure that during
the backfill process, the haunch zone is completely filled in the area of the bell hole (dug to
facilitate joint make up and installation of the heat shrink sleeve). It was found that due to
the size of the bell hole and the large radius of the pipe, the standard practice for backfill
described above would not always completely fill the haunch zone of the bell, Figure 14b.
Figure 14a, b: Application of Heat Shrink Sleeve, Voids in Bell

Holes after Completion of Backfilling
11
© ASCE
Pipelines 2015 1418
Since the bell hole is in the heat affected zone for the WAB process, the haunch zone should
be completely filled. It was learned on Section15-1 that a certain amount of haunch material
needed to be placed by shovel or by vibration equipment while the bottom crew was still
protected in the trench box. This practice has been proven to work well by excavation of
installed and welded joints to inspect material placement under the haunches in the bell hole
area and to test the heat shrink sleeve itself for any damage to the outer backing. Results
have been good with modified procedure.
4. 108-inch Polyurethane Coated steel pipe in casings or tunnels --- The Line J project used
both commercially available casing spacers and “mortar bands,” Figure 15, applied directly to
the polyurethane coated pipe at the shop. Both the casing spacers and mortar bands are
designed to act as “sleds” to facilitate pushing and or pulling the 108-inch pipe into the
casing or tunnel. They are also designed to keep the casing from contacting (or shorting) to
the 108-inch carrier pipe. Shorting of the bare casing pipe or tunnel liner plates would
require the cathodic protection system to protect the uncoated casing pipe which will drive
up the cathodic protection requirements significantly. Experience on the Line J project and
Section 15-1 has shown that both casing spacers and mortar bands require care in the
installation of pipe within tunnels to avoid damage to these features.
Figure 15: Two Mortar Bands on 25-ft Section of

108-inch Polyurethane Coated Steel Pipe
CONCLUSION
TRWD and DWU engaged the LAN team to design Section 15-1, the first segment of the
IPL Project. Design was based on a 100-year service life. Bids were taken in February 2014 on the
15 mile pipeline and steel pipe was selected. The pipe was supplied by Northwest Pipe Company
and the Garney Construction installation team is scheduled to complete installation by June 2015,
almost a full year ahead of schedule. The lessons learned on this segment, which were discussed in
this paper, have served to enhance or improve the overall quality of the pipe material and
installation. To date, the bid prices for not only Section 15-1 but also for the other IPL projects have
been much lower than the project estimates, and steel pipe with granular bedding has been chosen
for all projects. TRWD and DWU are pleased with the work of their consultants, researchers,
contractors and suppliers and expect the IPL project to continue on scheduled and provide raw
water to the Dallas-Fort Worth Metroplex for many decades to come.
12
© ASCE
Pipelines 2015 1419
Steel Water Transmission Mains in Liquefiable Soils in Hillsboro, Oregon,

Planning Considerations
Nebojsa "Nesh" Mucibabic, P.E., MBA1

Yuxin "Wolfe" Lang, P.E., G.E.2
Tyler Wubbena, P.E.3
1
City of Hillsboro, Water Department, 150 E. Main St., Hillsboro, OR 97123.
E-mail: nesh.mucibabic@hillsboro-oregon.gov
2
McMillen-Jacobs Associates, 101 SW Main St., Suite 360, Portland, OR 97204.
E-mail: lang@mcmjac.com
3
City of Hillsboro, Water Department, 150 E. Main St., Hillsboro, OR 97123.
E-mail: tyler.wubbena@hillsboro-oregon.gov
Abstract
Soil liquefaction has long been recognized as one of the greatest hazards for
the integrity and performance of water transmission systems during and after major
earthquakes. The effects of soil liquefaction include relatively large magnitudes of
permanent ground deformations (PGD) in the form of vertical settlements and
horizontal movements (lateral spreading). The effects can also include loss of soil
strengths and flotation. Pipeline performance during previous major earthquakes
showed that flexible, durable, strong pipes and joints can tolerate some degree of the
liquefaction induced deformations. Welded steel pipe is considered one of the better
seismic performing pipes. In some degrees, steel pipe can withstand plastic yield but
still maintain integrity and service during and after earthquakes. For steel water
transmission system planning and design, overall assessment of liquefaction hazards,
appropriate route selection, pipe thickness and weld selection are the crucial
elements. For deep pump stations and vault structures, foundation failure and
flotation of the liquefiable soils should also be considered. At some locations pipe
supports or ground improvements can provide some advantages in liquefaction

mitigation. Flexible/extendable joints can also provide additional benefits. However,
these additional mitigation features are typically associated with relatively high
costs. In some significant PGD zones (i.e. forefront of the lateral spreading zone)
with prohibitive mitigation costs, considerations can be given for planning emergency
repairs and bypass at controlled, accessible location. This paper/presentation explores
necessary geotechnical and liquefaction hazards assessments, steel pipe and joints
evaluations, and mitigation method selections to optimize construction cost and
seismic resiliency requirements for an approximately 30 mile long water transmission
main project for the City of Hillsboro and Tualatin Valley Water District (TVWD) in
Washington County, Oregon.
© ASCE
Pipelines 2015 1420
INTRODUCTION
TVWD and the City of Hillsboro, Oregon are developing their second major
water supply system from the Willamette River in the City of Wilsonville, through
more than 30-miles of large diameter water transmission pipelines to their service
areas. This project is called the Willamette Water Supply Program (WWSP), and its
location is shown in Figure 1.
Segment 3
Segment 4
Reservoir
Different Pipe Segments
Segment 2
Segment 1
WTP
Figure 1. Map of Willamette Water Supply Program
© ASCE
Pipelines 2015 1421
Currently, WWSP is in the process of a siting study and evaluating different

alignment options. One of the design requirements is that new system components
shall be seismically resilient to withstand likely impacts from a design earthquake
event with a return period of 2,475 years (2 percent probability of exceedance in 50
years). The scenarios for a 2,475 year design earthquake include a magnitude 9.0
Cascadia Subduction Zone earthquake (CSZ, for location see Figure2). The
magnitude 9.0 CSZ was selected by the State of Oregon as the earthquake scenario of
the development of Oregon Resilient Plan (ORP) in 2013.
WILLAMETTE WATER SUPLY PROGRAM (WWSP)

The WWSP consists of a new raw water intake on the Willamette River, a 110 MGD
water treatment plant and pump station, a 72-inch 16-mile long welded steel
transmission main, a 30 MG terminal reservoir, and two approximately 10-mile long
54-inch and 60-inch welded steel gravity pipelines connecting the new terminal
reservoir to the existing transmission mains in the service areas. The approximate
locations of these project components and pipeline corridor segments are shown in
Figure 1. The WWSP pipeline segments are anticipated to be constructed mainly
using open-cut method with a minimum soil cover of 7 feet and average trench depth
between 11 to 13 feet below ground or roadway surfaces. Exceptions are for some
deep crossings below river/creek, railroad, highway, and utilities, where trenchless
methods are considered for pipe installation.
The majority of the selected pipe corridor segments (main stem and gravity branches)
will be laid within the existing or future roads of two counties and four cities. The
corridor segments are also challenged by several major creek and river crossings,
state highway and railways crossings, vicinity of protected wildlife refuge areas,
wetlands, hazardous waste sites, and major sewer, water, gas, electric and oil
transmission mains within the existing public rights-of-way.
Each pipe corridor segment consists of three different pipe sub-routes that will need
to be further explored to more precisely elaborate additional site specific
characteristics including more detailed location of existing utilities, soil
characteristics, ground water levels, corrosivity of soils, potential for soil liquefaction,
need for soil improvements and other sub-route specific seismic characteristics and
concerns that need to be considered to meet pipe seismic resiliency requirements.
OREGON RESILIENCY PLAN

The 2013 Oregon Resiliency Plan (ORP) for Water and Wastewater Systems
identifies that "Re-establishing water and wastewater service will be a crucial element
in the overall recovery of communities after a Cascadia subduction zone earthquake.
Water for fire suppression, first aid, emergency response, and community use, as well
as water for normal health and hygiene, will be required soon after the event".
In addition, the 2013 ORP recommends that water-related industry associations and
manufacturers evaluate the need for seismic design standards for pipelines, and
encouraging The Oregon Health Authority (OHA) to include a seismic design
© ASCE
Pipelines 2015 1422
requirement as part of routine design review of water system improvements with the
goal to ensure that seismic considerations are incorporated into designs for critical
facilities.
The following Table 1 describes expected availability of potable water at different
locations and operational system requirements for different water system components
following the major earthquake.
Water availability at Availability of potable water in days

different locations and 1-3 3-7 7-14 14-30 30-90 90-180
through different components Operational requirements in percentage (%)
At water sources 20-30 50-60 80-90
Through transmission mains 20-30 50-60 80-90
At critical facilities 20-30 50-60 80-90
At key fire points 20-30 50-60 80-90
At fire hydrants 20-30 50-60 80-90
Through distribution pipes 20-30 50-60 80-90
Table 1. Expected availability of potable water after major earthquake
PROJECT SEISMICITY
The seismicity of WWSP project area is subject to two major earthquake
sources: (1) local, shallow crustal earthquake with relatively low magnitude (typically
less than 7.0), and (2) CSZ earthquakes with large magnitudes (typically above 8.0
and with potential up to 9.0 or even 9.2). For a 2,475 year design scenario, CSZ
earthquakes are considered to cause more damage than local crustal earthquakes.
CSZ earthquakes originate along at the interface of the Juan de Fuca and North
American Plates (Figure 2), which is located approximately 20 to 30 miles beneath
the coastline from north California to British Columbia. Recent seismological and
geological researches (Atwater 1995 and Goldfinger et al. 2012) disclosed compelling
coastline and ocean sediment evidences that CSZ earthquakes represent the most
eminent seismic hazard in our region.
Figure 2. Cascadia subdaction zone location (DOGAMI 2010)
© ASCE
Pipelines 2015 1423
PROJECT GEOLOGIC CONDITIONS

The WWSP pipeline is located within the southeastern portion of an
approximately 20-mile-wide valley of the Tualatin Basin. The bedrock at the base of
the valley is the Columbia River Basalt which is also outcropped on the valley sides.
The basalt is overlain by a few hundred feet thick Hillsboro Formation of ancient
fine-grained clayey and silty sediments from more than 1 million years ago. Above
the Hillsboro Formation, the valley ground surface is typically covered by
"Willamette Silt", of approximately 100 feet thick sandy silt, silt, clayey silt and silty
sand deposited 15,000 to 10,000 years ago by glacial flooding events. Also, within
the past 10,000 years, Tualatin River and its many tributaries deposited 10 to 20 feet
of fine-grained Recent Alluviums of soft clay, silt, fine sand, and organic soils above
Willamette Silt along the low lying, narrow flood plains.
The average seasonal groundwater level is typically between 10 to 20 feet deep

within the project areas with localized shallow groundwater (less than 5-feet in depth)
near the floodplains of the Tualatin River and its tributaries. The Willamette Silt and
recent river/creek alluviums are expected to have relatively high liquefaction
potentials and will likely generate large PGDs. The soil liquefaction profile is
typically extended from groundwater table to approximately 50 to 70 feet deep.
Considering the pipeline depths and the groundwater conditions, the WWSP pipeline
will likely be located on the upper portion of the liquefaction zone or within the non-
liquefiable soil crust. The general soil deposits and relative location of the WWSP
pipeline in the Tualatin Basin are shown schematically in Figure 3.
Figure 3. Generalized subsurface zone location and schematic section
POTENTIAL FOR LIQUEFACTION OF SOILS WITHIN PROJECT AREA
Based on published regional liquefaction maps from the State of Oregon

(DOGAMI 1995, 1997 and 2013), a soil liquefaction potential map and general
seismic hazards evaluation were developed for the WWSS pipeline corridors. A
portion of this liquefaction map for Hillsboro and Beaverton areas (where the east and
west gravity branches will be located) is shown in Figure 4.
© ASCE
Pipelines 2015 1424
Probability of Liquefaction
None
Low
Medium
High
Figure 4. PGD Map Hazard Area - Segments 3 and 4
In general, soil liquefaction risk is generally low in the south and north sections of the
main stem, but relatively high (medium to high risks) for the mid-portion of the main
stem and the entire east and west gravity branches. Also the entire WWSP pipeline
will very likely be exposed to different seismic hazard phenomena caused by transient
loading, a shaking hazard caused by seismic wave propagation and the amplifications
due to surface soil conditions and topography, and PGD resulting from surface fault
rupture, landslides and soil liquefaction related phenomena including vertical
settlement, lateral spreading, differential settlement and buoyancy movement
For WWSP project, majority of the pipeline is planned to be steel pipe with welded
joins. In general, strong shaking generated transient stress and strain are not
significant issues for welded steel pipe due to its high strength and flexibility.
However, PGD hazards are considered the main concern for steel pipe performance
during and after the earthquake.
For WWSP, risk for fault rupture is considered low and seismic landslide hazard is
limited to a few steep localized areas along the corridors. Soil liquefaction is caused
by the drastic increase of pore-water pressure (excited by the rapid cyclic earthquake
ground shaking) and dramatic decrease of effective contact stress between soil
particles which leads to the substantial loss of shear strength of the soil matrix.
EXPECTED TYPE AND RANGE OF SOIL LIQUEFACTION PHENOMENA
PGDs manifested by soil liquefaction mainly include post-liquefaction

settlement, lateral spreading, flow failure and flotation. Post-liquefaction settlement is
© ASCE
Pipelines 2015 1425
caused by the reconsolidation of liquefied soils during dissipation of the elevated pore
water pressure, typically after the earthquake shaking stops. The reconsolidation
typically leads to vertical settlement in the range of a few inches to more than 1 foot.
Lateral spreading is a form of ground failure, that is typically shown as blocks of

mostly intact surficial soil crust on top of the liquefied layer moving down slope on
very gentle inclinations (in some cases less than 2 degrees), or towards a free-face,
such as a river channel or bluff. It is caused by the dramatic reduction of shear
strength when soil liquefies, which leads to ground instability under the combination
of seismic cyclic loading and static gravity loading. In past large earthquakes, lateral
spreading typically occurred within 1,000 feet from the river banks and has generated
large horizontal PGDs ranging from a couple of feet to more than 10 feet.
Flow failure is also caused by the substantial loss of shear strength of the liquefied
soils. However, flow failure has typically occurred at steeper slope areas (Youd 1978)
near rivers or at the steeper river banks where even the static stability of the soil mass
cannot be maintained by the residual shear strength of the liquefied soil (Kramer
1996). In other words, the flow failure is driven by the static gravity stress, and the
seismic cyclic stress only “triggers” the soil liquefaction and the associated strength
loss which bring the soil to an unstable state. The flow failure is typically
characterized with sudden failure, rapid flow movement, and large movement
distance (sometimes exceeding tens of meters).
Soil liquefaction creates a buoyancy effect upon buried structures or pipelines.

Numerous flotation cases of underground tanks, structures, manholes and pipelines
have been reported in previous earthquakes, with upward movements ranging from a
few inches to a few feet.
For the WWSP, post-liquefaction settlement and lateral spreading movement are
considered as the main liquefaction manifested PGDs. Various degrees of
liquefaction settlements will likely occur in the liquefaction hazard areas identified in
Figure 4, and lateral spreading will likely affect the areas a few hundred feet on both
sides of Tualatin River and its major tributaries. Flow failures may also be possible at
some locations, but will likely be concentrated in some small areas along steep river
and creek banks. Additionally, risk for floatation of the pipeline is considered to be
low because the general pipeline depths are with or near the non-liquefiable soil crust
(discussed above).
HISTORIC PIPE FAILURE LOCATIONS WITHIN LIQUEFACTION ZONE
Pipelines within lateral spreading and flow failure zones suffered severe
damages during the past earthquakes (Eckel 1967, O’Rourke and Tawfik 1983, and
O’Rourke et al. 1989). The authors have noticed that remarkably high damage ratios
of pipelines were located at liquefied ground near the boundaries between the
liquefied and non-liquefied area and concluded that the damages were due to sharp
change of ground characteristics at soil boundaries.
© ASCE
Pipelines 2015 1426
For continuous pipelines depending on the pipe alignment relative to the ground
failure direction, large axial tension/compressive stresses, and bending/shear stresses
will likely developed at the boundaries of the lateral spreading/flow failure and
between major ground moving blocks (O’Rourke and Lie, 2012). Additionally,
flotation also caused pipeline failure within liquefaction zones. But this failure mode
was observed more concentrated to sewer pipes (ALA 2005).
EXPECTED PIPE FAILURE LOCATIONS IN WWSP
Considering the relatively high liquefaction hazard within WWSP areas, the
potential pipeline failures will likely occur at large liquefaction PGD areas including:
• Boundary zones between liquefiable soils and non-liquefiable soils,
• Within lateral spreading and flow failure zones,
TYPICAL PIPE FAILING MECHANISMS

Strong ground shaking and ground deformations may act in different
directions related to the pipeline alignment and may be affected by axial and
compressive forces or large bending moments caused by soil and pipe interaction that
may locally damage steel water lines, or cause their failure. The four, most expected
failure modes for continuous welded pipelines include:
• Pipe fracture due to extensive tensile strain related to pipe wall structure
and tensile capacity of lap joints which are usually the weakest points.
• Pipe wall local buckling or wrinkling due to extensive compressive strain
influenced by D/t ratio, the presence of internal and external pressure, and on
the yield stress of steel material. If the ductility of the steel is not exceeded,
buckled pipeline will be able to fulfill the basic function of carrying the flow.
• Beam buckling due to extensive compressive, axial loading, insufficient
cover depth and lateral resistance of surrounding soils. In shallow trenches
and loose backfill this mode of failure may occur. In deep trenches with and
dense backfill, the pipe will develop local buckling before beam buckling.
• Pipeline welded slip joints failure (fracture or crushing).
MITIGATION MEASURES AGAINST SEISMIC ACTIONS AND ABILITY
TO RESIST LARGE PGD TO MEET RESILIENCY REQUIREMNTS
Following the 9.0 Cascadia earthquake, it is expected that many local roads
and bridges may not survive. Availability of spare parts (pipes, butt straps, valves,
etc.) due to large pipe diameter, and their transport from different states to the pipe
damaged areas may significantly exceed the time to meet the transmission facilities
operational schedule suggested by the 2013 Oregon Resiliency code. Silty and sandy
soils may liquefy, and access to pipeline alignments outside major roads may not be
possible. Other utilities located within the same corridors may also break and require
simultaneous repair action. Availability of heavy construction equipment could be
questionable and its transport to potential pipe damaged locations may take
significant time. Availability of shoring for deep excavation to access the pipe and
availability of specialized labor forces to do repair work could also be problematic.
© ASCE
Pipelines 2015 1427
Designing the pipeline that may not break is almost impossible and/or cost
prohibitive, as one cannot predict exact earthquake magnitude and duration, ground
shaking intensity and direction, site specific PGD and differential
settlement/movements. For example, several oil and gas pipelines with relatively
thick wall (low diameter "D" to wall thickness "t" ratio, (D/t < 80) compared to water
lines (D/t >100) with butt welded joints were damaged during major earthquakes in
the past. (Rourke, Palmer, 1996)
On the other hand, City of Hillsboro and TVWD shall further explore geotechnical
conditions along selected pipeline segments to be able to evaluate potential mitigation
options and approaches to improve design and resiliency of transmission facilities,
minimize construction and post-earthquake repair cost, and explore means and
methods to repair and bring transmission facilities to expected operational conditions
to meet 2013 Oregon Resiliency Requirements. In selecting potential mitigation
options, the following important issues and options should be considered:
A) Soil Conditions along Pipe Alignments

• Provide more detailed geotechnical investigation along the pre-selected pipe
corridors to identify major liquefaction concerns and hazard related areas.
• Identify locations and depths of liquefiable soil layers and estimate potential,
amount and directions for liquefaction induced lateral spreading, and potential for
flotation or settlement.
• Check if the pipe would be located within, above or below liquefiable soil layers.
• Identify locations of expected changes from liquefiable to non-liquefiable soils
along pipe alignments, measure distances and widths of those layers to estimate
seismic loads and stresses within liquefiable and non-liquefiable soil layers.
• Consider selecting routes outside liquefiable zones, or re-route pipeline alignment
to avoid majority of liquefaction areas, if possible.
B) Pipe - Soil Interaction

• The smaller the pipe frictional forces on the pipe (pipe-soil interaction), the
greater the capacity of the pipe for surviving soil movement,
• Provide adequate pipe trench cover to avoid pipe flotation, and beam buckling,
• Consider placing pipes in a trench with shallow sloping sides to be able to

accommodate itself to the transverse as well to longitudinal components of the
soil movement in part by moving slightly out of trench.
• Improve ground conditions or use deep foundation piles under the pipe to reduce
amount of ground settlement, in areas expecting high subsidence.
• Consider using trenchless applications to install the pipeline below and beyond
the liquefaction zones.
• In soil transition areas, stiff soil conditions introduce higher stress and strains in
the pipeline. Use of soft backfill soils would result in reduced stresses but it also
may reduce its resistance in global buckling.
© ASCE
Pipelines 2015 1428
• For high expected pipe compression/buckling or bending areas, plan for access to
those areas following the major earthquake, due to high potential for damage.
• Consider providing in line valves in non-liquefiable soils, at adequate distance
upstream and downstream of major hazard zones and sufficient size pipe/valve
outlets to allow for installation of temporary bypasses, if needed
C) Pipe Characteristics
Steel pipelines strength to withstand different external forces are based on pipe
diameter, pipe wall thickness, and d/t ratio, knowing that:
• The lower the grade of steel, the more yielding could take place, and the greater
is the capability to resist liquefaction phenomena. Material ductility and
deformation capacity may be more important than strength for pipe to survive
operational capabilities after the earthquake
• Pipe movement capacity increases slightly with increasing pipe wall thickness.
• Increase in wall thickness will increase pipeline strength against seismic actions.
D) Pipe Joints
Pipes are connected using different pipe joints including full penetration butt weld
and single and double lap welded joints, and flanges.
Full Penetration Butt Welded Joints

• Provide joint efficiency between 95% and 100 % of the pipe.
• Mostly used in high seismic areas and for thicker oil and gas pipelines, to
withstand higher pressure and provide higher confidence and safety levels.
• Have higher installation cost.
Single Lap Welded (bell and spigot) Joints

• Have joint efficiency of approximately 40% to 45%.
• The strength of single welded joints are influenced by pipe wall thickness, pipe
bell geometry, alignment of pipe bell and spigot, uniformity of eccentricity
between spigot outside and pipe bell inside diameters.
• Are used in water mains due to lower internal pressure and lower construction
cost. The aligning of pipes (fit-up) is easier and take less time to weld
compared to butt welds.
• Could be welded internally for pipes larger than 30-inches or externally.

Efficiency of inner weld slip joints are often larger than outer welds as the
eccentricity of the weld in respect to pipe radius of inner welds are smaller
than for outer weld. However, some tests showed opposite results.
• Lap welded joints are geometrically eccentric that introduces additional stress
within the joint. As a result of this eccentricity, several case studies have
documented the poor performance of single lap welded joint in earthquakes,
which raises concerns about their use in regions of high seismicity, as
described by O’Rourke and Liu (1999).
• Bell radius should be large enough so that an excessive amount of steel
stability is not used in forming the bell.
© ASCE
Pipelines 2015 1429
Double Lap Welded (bell and spigot) Joints.

• No significant changes in joint efficiency between single and double welded
joints because of failure mechanism (joint efficiency of 45 % for single and
55% for double welded joints)
• New testing results for double welded pipe joints, expected in the near future
E) Emergency Repair Planning and Readiness
As mentioned previously, designing the pipeline that may not break is almost
impossible. Therefore emergency response planning shall be developed to address the
necessary repairs and/or emergency bypass. This involves emergency response
procedures, post-disaster access planning, emergency condition and damage
assessment, establishing repair priority and strategy, and personnel and logistic
readiness. It also involves storage at appropriate places of sufficient lengths of
bypass piping, valving, apparatus for welding or joining pipe sections, fuel, and
education of staff to be familiar with the equipment and emergency response
planning.
SUMMARY
Seismic hazards and their potential effects on life lines have been recognized and
addressed in the 2013 ORP by the State of Oregon. As a local critical water
transmission pipeline, TVWD and the City of Hillsboro have established the design
goal of seismic resiliency for their WWSS project. The main seismic event and
hazard effect are considered to be the large magnitude CSZ earthquake and the
associated soil liquefaction PGDs. Based on historical pipeline failures within
liquefaction zone and failure mechanisms, the design focuses of the planning and
routing levels of the WWSS are:
• Sub-route selection to minimize pipeline exposures to liquefaction PGDs,

• Investigation of pipe-soil interaction and soil improvements,
• Selection of appropriate grade and section for the steel pipe,
• Selection of appropriate welded joints,
• Emergency repair planning and readiness.
REFERENCES
Andrus Ronald D. Chung Riley M: Ground Improvement Techniques for

Liquefaction Remediation Near Existing Lifelines, Building and Fire
Research Laboratory National Institute of Standards and Technology,
Gaithersburg, MD 20899, October 1995
Atwater, B.F., Nelson, A.R., Clague, J.J., and 13 others, 1995, Summary of coastal
geologic evidence for past great earthquakes at the Cascadia subduction
zone: Earthquake Spectra, v. 11, p. 1–18.
Dama E, S.A. Karamanos, A.M. Gresnigt; Failure of Locally Buckled Pipelines,
ASME, 2007; 272 / Vol 129. May 2007
© ASCE
Pipelines 2015 1430
Eckel, E. B., 1967, Effect of the Earthquake of March 27, 1964 on Air and Water
Transport, Communications, and the Utilities Systems in South-Central
Alaska, Geologic Survey Professional Paper 545-B, US Government Printing
Office, Washington DC.
Goldfinger, C., Nelson C.H., Morey, A.E., Johnson, J.E., Patton, J.R., Karabanov, E.,
Gutierrez- Pastor, J., Eriksson, A.T., Gracia, E., Dunhill, G., Enkin, R.J.,
Dallimore, A., Vallier, T., 2012; Turbidite Event History-Methods and
Implications of Holocene Paleoseismicity of the Cascadia Subduction Zone,
U.S. Geological Survey Professional Paper 1661-F, 170 p.
Jianwen Liang, Shaoping Sun, Site Effects on Seismic Behavior of Pipeline:
Review; Journal of pressure Vesel Technology, November 2000, Vol.
122/469
Mason James, Thomas D. O Rurke; Scott Jones Lleker Tutuncu Compression
Performance of Steel Pipelines with Welded Slip Joints, Journal of Pipeline
Systems Engineering and Practices
O’Rourke, T. D. and M. S. Tawfik, Effects of Lateral Spreading on Buried Pipelines
During the 1971 San Fernanda Earthquake, Earthquake Behavior and Safety
of Oil and Gas Storage Facilities, Buried Pipelines and Equipment, ASME,
PVP-Vol. 77, 1983, pp. 124-132.
O’Rourke, T. D. and P. A. Lane, 1989, Liquefaction Hazards and Their Effects on
Buried Pipelines, Technical Report NCEER-89-0007. O’Rourke, X. Liu,
Response of Buried Pipelines Subject to Earthquake Effects, Multidisciplinary
Center for Earthquake Engineering Research, 1999. pp. 59- 105
O’Rourke, MJ., and Liu, X. (1999) “Response of Buried Pipelines Subject to
Earthquake Effects”, MCEER Monograph No.3 MCEER, Buffalo, New York.
O’Rourke (Jack) X. Liu Seismic Design of Buried and Offshore Pipelines.
Monograph MCEER-12-MN04, 2012
T.D. O’Rourke, MJ and M.C. Palmer: “Earthquake Performance of Gas Transmission
Pipelines, Earthquake Spectra: August 1996, Vol. 12, No. 3, pp. 493-527.
Van Greunen, Johannes Seismic Design of Bell-and Spigot Joints for large Diameter
Pipes, Pipelines 2008, ASCE
American Lifeline Alliances, 2005, Seismic
Guidelines for Water Pipelines. Oregon Department of Geology and Mineral
Industry (DOGAMI), 1995, GMS-90, Relative Earthquake Hazard Map of
Beaverton Quadrangle, Washington County, Oregon.
Oregon Department of Geology and Mineral Industry (DOGAMI), 1997, IMS-1,

Relative Earthquake Hazard Map of Portland Metro Region, Clackamas,
Multnomah, and Washington Counties, Oregon.
Oregon Department of Geology and Mineral Industry (DOGAMI), 2010, Cascadia.
Oregon Seismic Safety Policy Advisory Commission, 2013, The Oregon
Resilience Plan – Reducing Risk and Improving Recovery for the Next
Cascadia Earthquake and Tsunami.
Oregon Department of Geology and Mineral Industry (DOGAMI), 2013, Open-File
Report O-13-06, Ground Motion, Ground Deformation, Tsunami Inundation,
Seismic Subsidence, and Damage Potential Maps for the 2012 Oregon
Resilience Plan for Cascadia Subduction Zone Earthquake.
© ASCE
Pipelines 2015 1431
Addressing Rehabilitation Challenges for the Underwood Creek Force Main
Bryon Livingston, P.E.1; Jeremy Clemmons, P.E.2; and Keith Kalinger, P.E.3
1
Black & Veatch, Senior Project Engineer, Buried Infrastructure Practice, 8400 Ward
Parkway, Kansas City, MO 64114. E-mail: livingstonb@bv.com
2
Black & Veatch, Engineering Manager, Water Division, 300 Rancheros Dr.,
Suite 250, San Marcos, CA 92069. E-mail: clemmonsJ@bv.com
3
Milwaukee Metropolitan Sewer District, Senior Project Manager, 260 W. Seeboth
St., Milwaukee, WI 53204. E-mail: kkalinger@mmsd.com
Abstract
This paper discusses the challenges associated with the rehabilitation design and
construction of the 36-inch ductile iron pipe (DIP) Underwood Creek force main.
Based on the condition assessment and risk analysis, the Milwaukee Metropolitan
Sewerage District (MMSD) minimized the risk of continued pipe failures by
rehabilitating the force main. The rehabilitation design was based on a review of the
currently available technologies and methods to identify those that appeared to be
feasible. The four methods selected for further evaluation were cured-in-place pipe
(CIPP) (for pressure), swagelining, tight liner (rolldown), and fold and form
rehabilitation technologies. The final design documents included the use of two
methods, CIPP and high density polyethylene (HDPE) liner, allowing the contractor
to select the method and provide MMSD with the most cost-effective solution. The
challenges that were addressed included minimizing the number of pits for
construction of the liner because of location and constraints, the need to minimize
impacts on the Underwood Creek Parkway and crossing a railroad and highway. This
paper also discusses the challenges associated with implementing the design during
construction and the coordination with many stakeholders, including the full
reconstruction of a U.S. Highway 45 interchange by the Wisconsin Department of
Transportation (WisDOT), and the construction of overhead high voltage trans-
mission lines by the American Transmission Company (ATC).
PROJECT BACKGROUND
The Milwaukee Metropolitan Sewerage District’s (MMSD’s) Underwood Creek

pump station and force main were designed in 1981, and construction was completed
in 1983 to provide an additional 34 million gallons per day (mgd) of relief flow
capacity to the Underwood Creek Metropolitan Interceptor Sewer System during peak
flows. The 36-inch ductile iron pipe (DIP) force main is approximately 10,000 feet
long and runs from the Underwood Creek pump station to a point on Watertown
© ASCE
Pipelines 2015 1432
Plank Road where it transitions to gravity flow as shown on Figure 1. The force main
alignment runs along Underwood Creek for approximately 5,800 feet and then runs in
the center of Watertown Plank Road for the remaining 4,200 feet. The pipe is CL50
ductile iron with a cement mortar liner and an asphaltic coating. The force main was
installed with an 8 mil polywrap and sand bedding. The force main was constructed
with bonded joints to be electrically continuous, and 11 test stations were installed so
that the electrical potential between the force main and the ground could be
monitored.
Figure 1. Underwood Creek force main alignment. (Source: Black & Veatch)
The force main experienced two leaks along Underwood Creek, one in October 2007
and one in June 2009. Following each incident, the MMSD initiated investigations
into the causes for the failures. These investigations concluded that the likely cause
of failure was localized external corrosion of the pipeline, because of a damaged or

improperly installed polywrap, exposing the force main to corrosive soils.
In 2009, following the second leak, a consultant helped the MMSD develop an
approach for inspection of the force main to determine the existing condition of the
pipe. The approach included specific inspection technologies in distinct phases using
indirect condition assessment methods to identify areas that were the most corrosive,
followed by direct condition assessment (pipe inspection and wall thickness measure-
ments) with ultrasonic methods in the areas most susceptible to corrosion.
© ASCE
Pipelines 2015 1433
The condition assessment results were reviewed with MMSD and several alternatives
were presented. The alternatives ranged from regular leak detection to rehabilitation
of the pipeline. MMSD evaluated the alternatives and the associated risks and elected
to install a new liner in the pipeline to minimize the risk of exposure from another
failure of the pipeline. The next step was to complete a preliminary design to evalu-
ate rehabilitation technologies to determine which methods should be considered.
PRELIMINARY PROJECT DESIGN PROCESS
The preliminary design was initiated following the condition assessment with
evaluation of several alternatives. There are currently many technologies available
for rehabilitation, and they were reviewed in the preliminary design. The spray-
applied semi-structural liners were not considered because the previous failures
appeared to be from external corrosion, and it was decided to look at structural
rehabilitation, assuming there were large holes in the host pipe. The following
methods were considered for evaluation:
• Sliplining.
• Pipe bursting.
• Cured-in-place pipe (CIPP).
• Swagelining and tight liner.
• Fold and form pipe.
Sliplining. Sliplining would significantly reduce the potential of leaks from further
corrosion of the DIP since it creates a new pipe within the existing pipe. The use of
sliplining was not considered because it significantly reduces the cross-sectional area
of the flow by inserting a smaller diameter pipe into the existing pipe that would
impact the overall system operation.
Pipe bursting. The use of pipe bursting could be performed for the required project
diameter and would allow the use of high density polyethylene (HDPE) or fusible
polyvinyl chloride (PVC) pipe. However, pipe bursting for this diameter has the
potential for heaving in Watertown Plank Road and may be problematic in the
relatively sandy soils that are present along Underwood Creek.
Because of the project-specific risks associated with pipe bursting, this technology
was not considered for rehabilitating the Underwood Creek force main.
Cured-in-place pipe. CIPP for pressure applications is a newer technology gaining

popularity in the rehabilitation of pressure pipe. There are several manufacturers of
this product with similar application requirements.
© ASCE
Pipelines 2015 1434
The system can be designed as a fully structural, Class IV system in accordance with
American Society for Testing and Materials (ASTM) F1216 guidelines. This system
does not rely on the host pipe for any structural support. The system can be designed
as a Class III semi-structural system, which relies on the host pipe for some structural
support. A diagram of a CIPP retrofit is presented on Figure 2.
Figure 2. Cured-in-place pipe components. (Source: Norditube)
The installation of a cured-in-place liner would reduce the potential of leaks from
further corrosion of the DIP. The hydraulics of the cured-in-place liner has an
improved flow characteristic and an increase in flow capacity resulting from the
improved friction factor. The final design would have to evaluate the impact of the
liner on the operation of the pumps.
Design of CIPP is determined by the normal working pressures in the force main, and
when serving as a structural replacement, the external loading conditions are
included. The normal working pressures and loadings were confirmed during the
design process. The use of CIPP was considered for the rehabilitation of the force
main.
Swagelining. Swagelining is a trenchless rehabilitation process that involves running

the inserted HDPE pipe through a die to slightly reduce its diameter and allow it to be
pulled through the host pipe, as shown on Figure 3. Once through, the tension is
relieved and the new lining will elastically return to its original dimensions. The
process can utilize either thin or thick polyethylene material. The thickness of the
liner is determined during the design process.
© ASCE
Pipelines 2015 1435
Figure 3. Swagelining process. (Source: Murphy Pipelines)
The hydraulics for polyethylene liners has an improved flow characteristic and an
increase in flow capacity resulting from the improved friction factor. The final design
would have to evaluate the impact of the liner on the operation of the pumps. The use
of swagelining with HDPE was considered for the rehabilitation of the force main.
Fold and form. This installation process is similar to cured-in-place but uses a close-
fitting polyethylene pipe that is custom designed to match the existing conditions of
the pipe to be rehabilitated. The pipe is fused together in length for the project and
then held together by bands that are broken after installation into the host pipe, as
shown on Figure 4. The polyethylene liner will improve the flow characteristics in
the force main and any flow loss from the reduced diameter is typically recovered.
The fold and form process using HDPE was considered for the rehabilitation of the
force main.
Figure 4. Fold and form process. (Source: Insituform)
Results of the preliminary design determined that the CIPP, swagelining, tight liner
(rolldown), and fold and form lining should be further evaluated in the design of the
rehabilitation.
© ASCE
Pipelines 2015 1436
Table 1 provides the design criteria that were established for the rehabilitation:
Table 1. Rehabilitation Design Criteria.

Design Parameter Value
Host Pipe 36-inch DIP
Flow Capacity 34.0 mgd
Internal Working Pressure 52 psi
Occasional Surge Pressure (emergency shutdown) 120 psi
Soil Depth (above crown) 15 feet
Soil Density 110 lb/ft3
Ground Water Depth (above crown) 3 feet
Live Load, HS20 0.0 psi (buried greater
than 8 feet)
Modulus of Soil Reaction 1,000 psi
Modulus of Elasticity HDPE PE 4710 (50 year design) 28,500 lb/in.2
Modulus of Elasticity CIPP 350,000 lb/in.2
(Norditube CIPP resin)
Soil Support Factor 1.0
With the potential rehabilitation methods identified, the next step was to compare
each alternative to identify advantages and disadvantages for each. The evaluation
included a review of the use of PE 4710 that, at this time, was not included in the
standard. Based on the brief review for this application, it was determined that it
provided the required strength to meet this design. Table 2 summarizes the design
criteria for those methods considered for further evaluation.
© ASCE
Pipelines 2015 1437
Table 2. Design Criteria.

Pipe Surge Flow Distance
Pressure Pressure External Liner Liner Wall Capacity Technology Needed Time to
Technology Rating Rating(1) Load Material Thickness (2) Constraints Between Pits Construct
CIPP 80 psi 160 psi ASTM Felt w/glass Approx. 34.5 mgd Limited to less than Approximately 15-20 weeks
(Design F1216 fiber 0.75 inch 45 degree angle; 500 to 700 feet
Safety Factor Design reinforcement (18 mils) temperature impacts
[SF] of 2) Epoxy Resin cure time
CIPP 83 psi 166 psi ASTM Felt w/glass Approx. 34.5 mgd Limited to less than Approximately 15-20 weeks
(Design SF of F1216 fiber 0.75 inch 45 degree angle; 500 to 700 feet
2); short-term Design reinforcement (20 mils) temperature impacts
burst pressure Epoxy Resin cure time
of 325 psi
Swagelining 101 psi 202 psi 9.66 SF HDPE – PE 1.8 inches 33.0 mgd Thicker wall limits Longer along 6-8 weeks
(Thick against 4710 (3) bends, increases pull Parkway 2 – after pits are
Pipe) buckling DR-21 times; temperature 3,000; Water- ready
impacts roll down town Plank will
meet conditions
Swagelining 81 psi 162 psi 6.91 SF HDPE – PE 1.4 inches 33.0 mgd Thinner wall Longer along 6-8 weeks
against 4710 (3) negotiate bends; Parkway 2 installation
buckling DR 26 temperature impacts 3,000; Water- after pits
roll down town Plank will ready
meet conditions
Swagelining 64 psi 128 psi 4.89 SF HDPE – PE 1.1 inches 33.5 mgd Thinner wall Varies with 6-8 weeks
(Thin Pipe) against 4710 (3) negotiate bends; existing installation
buckling DR 32.5 temperature impacts conditions after pits
roll down ready
Tight Liner 64 psi 128 psi 4.89 SF HDPE – PE 1.1 inches 33.5 mgd Negotiate bends < 45 Estimated 6 weeks for
(Rolldown) against 4710 (3) degree 1,500 feet, varies installation
buckling DR 32.5 with conditions after pits
ready
Subline 64 psi 128 psi 4.89 SF HDPE – PE 1.1 inches 33.5 mgd Negotiate bends < 45 Lengths 6 weeks for
(Fold/Form) against 4710 (3) degree depending on installation
buckling DR 32.5 conditions after pits
ready
Notes:
(1) Surge pressure allowances in AWWA C906 are applied above the pressure class, and for occasional surges, are equal to the pressure class.
(2) Flow capacity with four of the five pumps in operation. Existing DIP estimated capacity is 31.0 mgd based on C value of 110.
(3) PE 4710 is the newer polyethylene that is being used; the design factor allows for a higher pressure rating.
© ASCE
Pipelines 2015 1438
Other factors. The preliminary design also had to consider how the project would be
impacted by other projects planned for the area.
There are two proposed projects that would impact the rehabilitation of the
Underwood Creek force main. The first is a proposed plan by the Wisconsin
Department of Transportation (WisDOT) to revise the intersection of Highway 45
and Watertown Plank Road as part of the Interstate 94 Zoo Interchange Project. A
preliminary site layout in the vicinity of Watertown Plank Road is presented on
Figure 5. As shown, Watertown Plank Road is scheduled to be reconstructed along
the entire length of the Underwood Creek force main. The modifications will
increase or decrease the depth of cover one to two feet in the area from Manhole
No. 34704 (near Watertown Plank Road and the railroad bridge) to Manhole
No. 34909 (near the entrance to the County Parks Building). The changes in the final
depth of cover, as well as the planned construction of new bridge piers by the force
main, will require that WisDOT structurally assess the changed loading conditions on
the pipe. However, the planned construction does not afford any appreciable cost-
saving opportunity to either add bonding to the pipe or replace the pipe because the
depth of cover over the pipe will not be significantly reduced.
Figure 5. DOT planned changes to Watertown Plank Road. (Source: WisDOT)
The second project is the proposed Western Milwaukee County Electric Reliability
Project to construct power transmission lines to a new substation north of Watertown
Plank Road. There are several routes being considered. Route B would construct
overhead power along the Underwood Creek Parkway near the force main and then
© ASCE
Pipelines 2015 1439
construct underground power along Watertown Plank Road. At this point, the final
route alignment is being determined by the Public Service Commission. If Route B is
selected, American Transmission Company (ATC) would need to complete a more
detailed design of the underground power line along Watertown Plank Road. After a
review of the ATC design, MMSD could decide if the ATC project affords a better
opportunity to re-bond or replace the force main.
Because of the potential for stray currents from high voltage power lines to induce
corrosion on buried metallic pipe, ATC was required to demonstrate there would be
no adverse impact on the force main from any potential stray currents from this power
line.
The impact from these two projects was included in the final design.
FINAL PROJECT DESIGN PROCESS
The final design process began by refining the methods that could be constructed with
the limitations imposed by the existing alignment and surface conditions, the layout
of the project site and avoidance of other work in the area. Primary constraints
considered for the final rehabilitation method selection included the requirements for
accessing the pipeline via excavation pits (including material layout lengths), pipeline
geometry and constructability challenges to meet operating conditions and
parameters.
The existing alignment of the pipeline included both horizontal and vertical changes
in direction, including some bends greater than 11-1/4 degrees. The alignment also
was located in environmentally sensitive regions, an active park and bike path,
beneath a major roadway and under a freeway crossing – all conditions affecting
potential excavation pit locations and pipeline layout lengths. These existing
conditions required the selected methods to be achievable while meeting the
requirements and limitations of the various stakeholders, including a large portion of
the public.
The critical nature of the force main during a rainfall event also resulted in the
necessity of its placement back into service within a three-day period, should the
long-term forecast predict significant rainfall. Any bypassing of the force main was
limited because of this constraint, thus the rehabilitation methods selected had to meet
these requirements so that full reliability would be ensured within a short duration.
These constraints had a significant impact on the selection of the final rehabilitation
method. Early in the final design, CIPP and swagelining with HDPE were selected as
the final rehabilitation methods. The next step in final design was to determine the
locations in which either or both technologies were feasible. The alignment was
separated into four segments based on the project constraints outlined above.
Segments A through D are presented on Figure 6.
© ASCE
Pipelines 2015 1440
Figure 6. Project segments. (Source: Black & Veatch)
Segment A: Both swagelining and CIPP rehabilitation techniques are applicable

along Segment A. Swagelining the first several hundred feet of this section would
require several pits at existing horizontal bends. Alternatively, the swagelining
contractor could elect to use CIPP in this first section to avoid digging multiple access
pits. The remainder of the segment is well suited for swagelining because of ease of
access and long straight reaches. Given the large amounts of staging areas available
in Segment A, MMSD may benefit from competitively bidding the two technologies
against each other, even if the swagelining contractor needs to use a CIPP liner for the
first several hundred feet.
Segment B: Work associated with rehabilitation of Segment B will be performed by

coordinating with the reconstruction of a portion of Watertown Plank Road in the
vicinity of Mayfair Road. Use of CIPP rehabilitation technology for this segment is
recommended.
Segments C and D: The rehabilitation of the force main in these segments will be
designed as CIPP because of the limitations imposed by construction from the
WisDOT project and active roadway conditions.
A challenge during final design was the selection of the excavation pit sites and
access to the sites. Because of the environmentally sensitive locations along
Underwood Creek and the necessity to maintain traffic flow on Watertown Plank
Road, there were limited site locations. The design team overcame these constraints
© ASCE
Pipelines 2015 1441
by working with stakeholders and utilizing the correct rehabilitation method to limit
impacts on the environment and community.
A key step during final design was development of specifications to address the
planning of the project installation and the confirmation of a quality product
installation. These included development of detailed sections for pre- and post-
installation closed circuit televising of the pipeline, cleaning of the existing force
main prior to installation, and bypassing requirements and limitations. The careful
development of these specifications was implemented during construction when an
unknown obstruction was discovered in the pipeline that would have resulted in
detrimental impacts that would have compromised the final pipeline liner.
BIDDING AND CONSTRUCTION
The bid allowed the alternative for both rehabilitation methods. Based on the bids
received, the preferred rehabilitation method was CIPP liner for the entire alignment,
primarily because of the project conditions and field constraints, including access pit
locations.
Construction of the pipeline began in late summer 2014 and continued through the
winter. The number of access pits required several connections between the CIPP
liner sections. The design required the use of PVC pipe to connect the CIPP liner at
the access pit as shown on Figure 7.
Figure 7. PVC used to combine sections of the pipe. (Source: MMSD)
© ASCE
Pipelines 2015 1442
The challenge of construction was increased with the cold weather and required a
large boiler system to create the steam to cure the resin. The CIPP process was
successful, and the pipeline was rehabilitated.
CONCLUSION
The rehabilitation of the force main required the evaluation of several technologies to
address the many challenges and limitations from the location of the alignment. The
process from preliminary design through final design and construction included
addressing these constraints to provide the liner that would extend the service life of
the pipeline and minimize the risk of failure. The CIPP process was successful (as
shown on Figure 8), and the pipeline was rehabilitated.
Figure 8. CIPP lining in pipe. (Source: MMSD)
© ASCE
Pipelines 2015 1443
Decision-Making Guidance for Culvert Rehabilitation and Replacement Using

Trenchless Techniques
He Jin, M.S.1; Kalyan R. Piratla, Ph.D., A.M.ASCE2; and John C. Matthews, Ph.D.,
A.M.ASCE3
1
Graduate Research Assistant, Glenn Department of Civil Engineering, Clemson
University, 109 Lowry Hall, Clemson, SC 29634. E-mail: hjin@clemson.edu
2
Assistant Professor, Glenn Department of Civil Engineering, Clemson University,
109 Lowry Hall, Clemson, SC 29634. E-mail: kpiratl@clemson.edu
3
Principal Research Scientist, Energy and Environment, Battelle Memorial Institute,
7231 Palmetto Dr., Baton Rouge, LA 70808. E-mail: matthewsj@battelle.org
Abstract
Culverts which are typically located under roadways and embankments for the
passage of water are designed to support the super-imposed earth and live loads from
passenger vehicle and trucks as well as the internal hydraulic loading from the stream
flow. Many of the existing culverts in the U.S. are in a deteriorated state having
reached the end of their useful design life, making them vulnerable to failures with
potentially catastrophic consequences. Traditionally, deteriorated culverts have been
replaced by the conventional open-cut construction method. Due to higher costs,
adverse environmental and societal impacts associated with open-cut method,
particularly in high population and busy roadways, transportation agencies are
increasingly looking to adopt trenchless techniques for addressing their culvert
problems. This paper reviews several trenchless rehabilitation and replacement
techniques investigating their suitability to address different defects, and their
compatibility with various host pipe materials and diameters. With focus on
reinforced concrete pipe (RCP), corrugated metal pipe (CMP) and high density
polyethylene (HDPE) culvert materials, easy-to-use decision-making flowcharts are

presented in this paper. State transportation agencies, U.S. Forestry Service and other
local government agencies that manage culvert infrastructure will benefit from this
paper.
1. INTRODUCTION
The U.S. culvert infrastructures have served the American society for over 100 years
(Selvakumar et al., 2014). Due to their invisibility from the surface, they often get
ignored until a problem such as road settlement or flooding arises. Many culvert
structures are currently in a deteriorated condition reaching the end of their design life
(Yang et al., 2009). Their state of disrepair is mainly attributed to: 1) a general lack of
1
© ASCE
Pipelines 2015 1444
uniformity and improvement in design, construction, and operation practices; 2)

insufficient quality control during pipe installation; 3) little or no inspection and
maintenance; and 4) reduced funding (Ge et al., 2014). The deteriorated culvert
infrastructure could potentially lead to surface depression, extensive cracking, and in
extreme cases a collapse of the road surface.
Traditionally, deteriorated culverts have been replaced using the conventional open-
cut construction method. Due to higher costs, adverse environmental and societal
impacts associated with the open-cut method, particularly in areas of high population
and busy roadways, transportation agencies are increasingly looking to adopt
trenchless techniques for addressing their culvert problems. Therefore, several
techniques for rehabilitating and replacing culverts have been developed thus far;
however, some still need further validation for wider adoption in practice.
Previous studies have proposed decision-making frameworks for choosing culvert

rehabilitation and replacement techniques. Thornton et al. (2005) presented a Multi-
Criteria Decision Analysis (MCDA) tool in Microsoft Excel for selecting one
appropriate culvert rehabilitation method from slip lining, close-fit lining, spirally
wound lining, cured-in-place pipe lining (CIPP), and spray-on lining. Their decision-
making is based on scores given to each rehabilitation technique considering several
criteria such as design life, capacity reduction, installation time, etc. Matthews et al.
(2012) developed a set of decision-making flowcharts for appropriately selecting a
rehabilitation or replacement method for corrugated metal pipes (CMP). Their
decision-making approach was based on specific defects normally observed in CMP
culverts; namely, inadequate hydraulic capacity, inadequate structural capacity, and
inadequate bedding support. Hollingshead et al. (2009) briefly summarized the
description, installation procedures, and highlighted the advantages and disadvantages
of segmental lining, spiral lining, CIPP, fold and form lining, deformed-reformed
HDPE lining, and cement mortar spray lining in order to appropriately select a
suitable rehabilitation method.
Although many researchers evaluated culvert rehabilitation and replacement methods,

few have presented a decision-making framework that is classified by defects for
individual culvert materials for better utility. This paper briefly reviews several
trenchless rehabilitation and replacement techniques, highlighting their suitability to
different defects and materials. With focus on reinforced concrete pipe (RCP), CMP
and high density polyethylene (HDPE) materials, easy-to-use decision-making
flowcharts for selecting appropriate culvert rehabilitation and replacement techniques
for different defects are developed and presented in this paper. Some techniques were
excluded in the analysis presented in this paper because of their limited use and lack
of sufficient performance data.
2. TRENCHLESS TECHNIQUES FOR CULVERT REHABILITATION AND

REPLACEMENT
This paper evaluates slip lining, cured in place pipe (CIPP), cement mortar lining,
fold and form lining, spiral-wound lining, and pipe bursting techniques. Based on the
2
© ASCE
Pipelines 2015 1445
evaluation of these techniques, easy-to-use decision making guidance is presented.

This section briefly describes the aforementioned techniques with their specific
advantages and limitations highlighted in Table-1 (Caltrans, 2013; Mitchell et al.,
2005; Meegoda et al., 2009; Hunt et al., 2010; Hollingshead et al., 2009; Syachrani et
al., 2010; and Yazdekhasti et al., 2014).
2.1 Slip lining (SL)

Slip lining entails inserting a smaller diameter pipe directly into a deteriorated or
failed host culvert. It is accomplished by either pulling or pushing the liner pipe into
the host culvert using jacks or other equivalent equipment as shown in Figure-1a. The
space between the host pipe and liner is grouted forming a composite pipe that is
stronger and smoother. Flexible pipes such as PE, HDPE or PVC, with mechanical
(segmental) or fused (continuous) joints are typically used as liners.
2.2 Cured in place pipe (CIPP)

CIPP entails inserting a polymer fiber tube or hose impregnated (or coated) with a
thermosetting resin (such as unsaturated polyester, epoxy vinyl ester, or epoxy with
catalysts) into the host culvert through an inversion process, as shown in Figure-1b.
The liner is then expanded to closely fit the host pipe after which it is cured using hot
water, steam or UV light. CIPP can be applied to all shapes of host pipes, and the
liners in this method are known for their flexibility and suitability for even 90 degree
bends, making them ideal for cases where access to culvert is limited.
2.3 Fold and form lining (FFL)

Fold and form lining entails inserting a folded liner pipe into the host culvert after
which the liner is heated and expanded to tightly fit into the host culvert. The liner is
then cooled to maintain its shape. Figure-1c illustrates the fold and form lining
method.
2.4 Spiral-wound liner (SWL)

Spiral-wound lining entails feeding coiled inter-locking plastic strips through a
winding machine that moves along the host culvert forming a smooth plastic pipe, as
illustrated in Figure-1d. The space between the host culvert and the plastic liner is
grouted to form a robust composite pipe. This technique is more suitable in cases with
non-circular host culverts and strict access restrictions.
2.5 Cement mortar Lining (CML)

Cement mortar lining method entails spraying cement mortar on the interior of the
host culvert using rotating head of air-powered machine. The rotating or conical
trowels then provide smooth surface on the inside, as can be observed from Figure-1e.
This technique is typically employed in the case of steel or iron culverts.
2.6 Pipe bursting (PB)

Pipe bursting entails forcing a larger size expansion head into the host culvert which
fractures or splits it while pulling along a new pipe, as shown in Figure-1f. Pipe
3
© ASCE
Pipelines 2015 1446
bursting is one of the popular trenchless replacement techniques which can also be
used for upsizing the culvert subject to favorable soil conditions.
Based on the advantages, limitations, and descriptions presented in Table-1
and the preceding paragraphs, decision-making guidance is developed. The guidance
is based upon specific defects often observed with RCP, CMP, and HDPE culverts.
Figure 1. Illustration of rehabilitation and replacement methods: (a) Slip lining, (b)
CIPP, (c) Fold and form lining, (d) Spiral-wound lining, (e) Cement mortar lining, (f)
Pipe bursting
4
© ASCE
Pipelines 2015 1447
Table 1. Advantages and limitations of culvert rehabilitation and replacement methods
Technique Advantages Limitations

• Simplest and cheapest technique • Reduced culvert size
SL • Can be used with live flow in host culvert • Needs larger pits for liner insertion
• Offers structural capacity • Not easy to reconnect laterals
• Requires no access pits • Need a lot of water or steam
CIPP • Can negotiate bends • Toxic resins could infiltrate ground water
• Applicable for different culvert shapes and tight curves • Cannot be used with live flow
• Increased liner size compared to SL • Applicable to limited host culvert sizes and shapes
FFL • Can negotiate bends • Toxic resins could infiltrate ground water
• Doesn’t need grouting • Requires additional resources for folding the pipe
• Can be used with live flow in host culvert • Larger installations require man entry
SWL • Applicable for different culvert shapes and tight curves • Need specialized equipment
• Requires no access pits • Reduced culvert size
• Effectively fights corrosion in unlined metal culverts • Suitable for bends up to 45 degree
CML • Long-term protection at cheaper cost • Requires significant curing time
• Easy mixing and application of lining material • Used mostly for corrosion protection only
• Provide structure support • Could pose threat to surrounding sub-structures
PB • Capable of installing larger than host culvert size • Not suitable for all soil conditions
• Faster and cheaper than open-cut method usually • Cannot fix line and grade problems of host culverts
© ASCE
Pipelines 2015 1448
3. DECISION-MAKING GUIDANCE FOR TRENCHLESS CULVERT

REHABILITATION AND REPLACEMENT TECHNIQUES
The three common types of deficiencies that would constitute rehabilitation or
replacement are (1) structural, (2) hydraulic capacity, and (3) bedding. Structural
issues can include: collapse, corrosion/abrasion, invert deterioration, local damage in
the wall or joint such as crack or spalls, shape distortion, or defective/misaligned
joints. Hydraulic capacity issues can include roadway overtopping/flooding, scour at
the inlet or outlet, sediment/debris buildup, and embankment damage. Most hydraulic
issues are addressed with methods and practices that are outside of the focus of this
paper such as culvert realignment, endwalls and wingwalls, energy dissipaters, and
bank stabilization. However, in cases where additional capacity is needed (i.e.,
flooding), methods such as pipe bursting can be used to increase capacity. Bedding
deficiencies can lead to surface depressions, voids around the culvert, and
undermining defects. These issues can be addressed with various forms of grouting,
which are outside of the focus of this paper. Rehabilitation or replacement techniques
that are specifically suitable for commonly observed defects in RCP, CMP, and
HDPE culverts are categorically summarized in the form of decision-making flow
charts which are presented in Figures 2 and 3. These flow charts are developed based
on the capabilities, advantages and limitations of the trenchless alternatives as
discussed in SECTION-2, along with the support of the previously cited literature.
6
© ASCE
Pipelines 2015 1449
Collapse Pipe Bursting (if possible)
CIPP
Joint Misalignment Fold and form lining
Spiral-wound lining
Defective Joint/Seam Sectional CIPP
Pipe Bursting (Major)

RCP or CMP
CIPP
Corrosion/Abrasion (CMP)
or Sliplining
Shape Distortion
or Spiral-Wound lining
Cracks/Spalls (RCP)
Fold and Form Lining
Cement Mortar Lining
Invert Deteriortion Segmental Sliplining
Figure 2. Trenchless rehabilitation and replacement techniques for CMP and RCP
culverts
7
© ASCE
Pipelines 2015 1450
Collapse Pipe Bursting (if possible)
Slip Lining
CIPP
Joint Infiltration Fold and Form Lining
Spiral-Wound Lining
Pipe Bursting
HDPE
Cement Mortar Lining
Joint Gasket Failure CIPP
Sliplining
Wall Deflection/Cracking
to Fracture
Spiral-Wound Lining
CIPP
Cracking/Not Fractured
Figure 3. Trenchless rehabilitation and replacement techniques for HDPE culverts
4. CONCLUSION
This paper provides an overview of trenchless rehabilitation and replacement

techniques that are applicable to commonly encountered structural issues in RCP,
CMP, and HDPE culverts. The two flowcharts presented provide an easy to follow
decision making process for when certain techniques are applicable. Additional
considerations such as cost, contractor availability, and local preferences must always
be taken into account before selecting a final technology, but these flowcharts provide
a general process for selecting a group of technically applicable methods.
ACKNOWLEDGMENT: This research was partly supported by the South Carolina
Department of Transportation (SCDOT) under Grant SPR718. The views and
conclusions contained in this document are those of the writers alone and should not
be interpreted as necessarily representing the official policies, either expressed or
implied, of SCDOT. The support of SCDOT is greatly appreciated.
8
© ASCE
Pipelines 2015 1451
5. REFERENCES
California Department of Transportation (Caltrans). (2013) “Design Information
Bulletin No. 83 - 03Caltrans Supplement to FHWA Culvert Repair Practices
Manual.” http://www.dot.ca.gov/hq/oppd/dib/dib83-03.pdf
Ge, S., and Sinha, S. (2014). “Failure analysis, Condition Assessment Technologies,
and Performance Prediction of Prestressed-Concrete Cylinder Pipe: State-of-the-
Art Literature Review.” Journal of performance of Constructed Facilities, 28(3),
618-628.
Hollingshead, T., and Tullis, B. P. (2009). “In-Situ Culvert Rehabilitation: Synthesis
Study and Field Evalutation.” Utah Department of Transportation, Salt Lake,
Utah.
Hunt, J. H., Zerges, S. M., Roberts, B. C., and Bergendahl, B. (2010). “Culvert
Assessment and Decision-Making Procedures Manual for Federal Lands
Highway.” Federal Highway Administration Central Federal Lands Highway
Division, Lakewood, CO.
Matthews, J. C., Simicevic, J., Kestler, M. A., and Piehl, R. (2012). “Decision
Analysis Guide for Corrugated Metal Culvert Rehabilitation and Replacement
Using Trenchless Technology.” United States Department of Agriculture Forest
Service.
Meegoda, J. N., Juliano, T. M., and Tang, C. (2009). “Culvert Information
Management System – Demonstration Project.” New Jersey Department of
Transportation, Trenton, NJ.
Mitchell, G. F., Masada, T., Sargand, S. M., and Jobes Henderson & Associates, Inc.
(2005). “Risk Assessment and Update of Inspection Procedures for Culverts.”
Ohio Department of Transportation, Columbus OH.
Selvakumar, A., Tuccillo, M. E., Martel, K. D., Matthews, J. C., and Feeney, C.
(2014). “Demonstration and Evaluation of State-of-the-Art Wastewater Collection
Systems Condition Assessment Technologies.” Journal of Pipeline Systems
Engineering and Practice, 5(2), 1949-1190.
Syachrani, S., Jeong, H., Rai, V., Chae, M., and Iseley, T. (2010). “A risk
management approach to safety assessment of trenchless technologies for culvert
rehabilitation.” Tunnelling and Underground Space Technology, 25, 681–688.

Thornton, C.I., Robeson, M.D., Girard, L.G., and Smith, B. A. (2005). “Culvert Pipe
Liner Guide and Specifications.” Federal Highway Administration Central
Federal Lands Highway Division, Lakewood, CO
Yang, C, and Allouche, E. (2009). "Evaluation of Non-destructive Methods for
Condition Assessment of Culverts and Their Embedment." ICPTT 2009,361, 28-
38
Yazdekhasti, S., Piratla, K. R., Khan, A., and Atamturktur, S. (2014). “Analysis of
Factors Influencing the Selection of Water Main Rehabilitation Methods.” North
American Society for Trenchless Technology (NASTT)
9
© ASCE

Pipelines 2015 Recent Advances in Underground Pipe... - (Assessment and Rehabilitation II)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pipelines 2015 Recent Advances in Underground Pipe... - (Assessment and Rehabilitation II)

Uploaded by

Copyright:

Available Formats

Pipelines 2015 1181

Water Mains Degradation Analysis Using Log-Linear Models

Amin Ganjidoost1; Rizwan Younis2; and Mark Knight3

The development of reliable lifecycle intervention plans for water distribution

WATER MAIN DETERIORATION MODELS

Deterioration models predict the future condition of infrastructure components

(Kalbfleisch and Prentice, 1980). Times to failure can be modeled by different

WATER MAIN DEGRADATION ANALYSIS

General description of case municipality. The water distribution network includes

Table 1. Characteristics of the case municipality.

Weibull/Exponential model. The modeling strategy is used two distributions to

To compute PDF and reliability with Weibull and Exponential distributions,

Weibull/Exponential model results and discussions. The Weibull distribution is

where, is the scale parameter of the Exponential distribution

Table 2. Calibration Parameters of Weibull/Exponential Model.

to the area under survival curve and mathematically is shown in Equation 4. By

Subsequent break. Probability density function with the Exponential distribution

(a) First Break Second Break

Mean Time To Failure (years)

Log-Linear model. Log-linear model, also known as Poisson regression or Poisson

Equation [7] can be written as:

log = log + + + + ⋯+ [8]

Then, the expected number of breaks is given as:

For this study, Equation [9] is expressed as:

The maximum likelihood method can be used to estimate the parameters of

Table 3. Water Main Breaks Data.

Table 4. Reorganized Water Main Breaks Data.

Table 5. Parameter Estimates.

: construction period; : observation period

Table 6. Observed and Predicted Breaks.

Figure 4. Residual plot.

The Weibull/Exponential process and Poisson log-linear model provide a

Agresti, A. (2002) Categorical data analysis (Second ed.), Wiley Interscience,

modeling.” Journal of Hydroinformatics, 11(1), 1-17.

Rehabilitation and Replacement of the East Layton Pipeline

PROJECT HISTORY AND BACKGROUND

Figure 1. 1960 and 2012 Aerial View of East Layton Pipeline

As shown in Figure 1, residential development subsequent to the original construction has

CONDITION ASSESMENT OF THE EXISTING PIPELINE

Description of the Existing Pipeline

East Layton Pipeline Hydraulic Profile

4700 Min HGL

section eliminated 8 of the 15 leaks found during

Hydraulic Capacity and Increasing Demands

REHABILITATION AND REPLACEMENT ALTERNATIVES

Alternative 3. Alternative 3 shares the alignment of Alternative 2, but includes tunneling to

Alternative 5. Alternative 5 was developed to provide an alignment that would eliminate

Alternative 5: - Gravity flow pipeline - Disturbance to residential streets.

Recommended Replacement and Rehabilitation Alternative

selection of Alternative 5 as the preferred alternative include: a reduction of risk during

Figure 7. Contractual Arrangement of CMGC Project Delivery Method.

Benefits of CMGC Project Delivery

Utility Investigations and

potholing and uncovering existing

Traffic Control. For pipeline

approach allowed the contractor to

SUMMARY AND CONCLUSIONS

Validating “Fully Structural”: Development and Testing of a New Carbon

N. Meyer1; Scott Arnold1; and G. Bontus2

Trenchless technologies similar to Cured in Place Pipe (CIPP) continue to be

STANDARDS, CODES AND GUIDELINES

Engineering organizations like American Water Works Association (AWWA) have

“1. A long-term (50-year) internal burst strength, when tested independently