Economical Sewer Main Rehabilitation Utilizing ArcGISTM and Dynamic Segmentation

 

 

Dale V. Clancy, P.Eng.,

Infrastructure Services Department

Saskatoon, Canada

 

Jan-Mark Gustafson, P.Eng., Ph.D.,

Infrastructure Services Department

Saskatoon, Canada

 

Lynne Higgins, P.Eng

Infrastructure Services Department

Saskatoon, Canada

 

ABSTRACT

 

Traditionally, sewer main condition is determined by video camera inspection and evaluation.  The high cost of sewer camera inspection has led to a selection process that considers maintenance and sewer cleaning records in conjunction with sewer main characteristics to determine inspection priorities. ArcGIS provides a spatial platform to consider deterioration and quantify failed mains and quickly identify sewer camera inspection priorities. To maximize the benefits from limited rehabilitation funds, the sewer rehabilitation industry’s focus has shifted to partial and spot rehabilitation strategies.  Dynamic segmentation and thresholds are used in evaluating partial rehabilitation alternatives that improve on the economics of rehabilitation decisions.

 

INTRODUCTION

 

Customers’ demands and legislative requirements to ensure the sustainability of sewer utilities have motivated utilities to quickly evaluate and quantify their sewer network.  With limited budgets and growing demand to rehabilitate sewer mains, utility managers need analytical frameworks to evaluate and economically renew deteriorating sewers. The focus of this work is the development of frameworks to determine sewers camera inspection priorities and to choose from among those available that treatment which provides the best balance between costs and risk.

 

The traditional approach to sewer rehabilitation is to video inpsect sewer mains to determine their structural and operating condition.  The video is evaluated and a scoring or similar system is used to quantify the condition of the main between manholes. The cost involved in video inspection and video evaluation represents a significant expenditure.  In this work, information from maintenance activities and sewer attributes is considered in identifying the risk of failure and economic considerations in determining whether or not to video inspect.

 

This framework stresses the need for economical collection and analysis of information on the condition and factors that influence the long-term sewer main performance.  In the development of an analytical framework, the need for a spatial platform quickly became apparent. ArcGIS was selected to provide the spatial analysis tools.  ArcGIS permits analysis of factors from a variety of different information sources.  Features available within ArcGIS allow for efficient analysis of information from attribute, maintenance, repairs, video inspection and rehabilitation sources.

 

 

BACKGROUND

 

Centrally located on the Canadian Prairies, the City of Saskatoon was incorporated in 1906 largely to facilitate the construction of water and sewer systems.  The City currently has a population of 208,000 people served by 854 km of sanitary sewer mains and 58,000 sewer services.

 

In 1982, the City initiated a video inspection program.  As of 2001, 496 km of sewer video has been taken using black and white, and color video. This represents 58.2% of the sewer system.  However, very little of the video has been analyzed using a generally accepted framework.

 

In 1989, the City of Saskatoon initiated an annual program of sewer main replacement, followed by a sewer-lining program starting in 1990.  At the start, there were several candidates of clay tile and concrete mains for which the need for rehabilitation or replacement was obvious.  To date, 13.2 km of sewer mains have been replaced and 14.2 km have been lined.

 

In 1990, the City began the long-term commitment to develop spatial models for water, sewer, road and sidewalk assets. Initially, as-built drawings were entered into a spatial database. In-house databases were developed to capture attribute, operating and repair data. 

 

In 1997, the City of Saskatoon initiated a review of its rehabilitation programs as part of a corporate strategic plan. The goal of this review was to ensure the long-term sustainability of Public Works infrastructure at minimum long-term cost. An Infrastructure Group was formed to develop analytical frameworks to consider roads, sidewalks, and water and sewer assets.  The group looks first at the economics by considering approaches that minimize long-term costs.  Minimum acceptable service levels are defined and are considered as constraints.  After the economic models are developed, options that will improve service are considered.  These options are evaluated based on the improvement in services versus the increased cost of the proposed service.  Ultimately, Council is asked to make choices on levels of service versus cost both now and for the future (Clancy, Gustafson1).

 

In 2000, a study of the Sanitary Sewer System was initiated to review the work to date and to develop a framework for analyzing the sewer system leading to a sustainable rehabilitation plan. In 2001, ArcGIS was chosen as the City of Saskatoon’s corporate platform largely to facilitate analysis and utilize water, sewer and roadways spatial models. Today ArcMapTM is used to quickly evaluate spatial and attribute information, expanding the opportunities to better understand assets.

 

 

ANALYTICAL FRAMEWORK

 

For sewer utility customers, a sewer system has failed when sewage backs up into their homes or businesses.  Unfortunately, maintaining a sewer infrastructure with few or no sewer backups is not economical.  Each utility, whether implicitly or explicitly, determines the level of risk that defines service and costs to their customers.  For this reason there are no universally accepted rules that define when and how a sewer should be inspected, evaluated, analyzed and rehabilitated.  Each community and each utility has different values and a different tolerance to risk.

 

Risk is the probability of a failure multiplied by the consequences of the failure.  In a sewer system, the consequences are the cost of damage to property, repair costs and the impacts to the customer. The probability of failure is based on the structural and hydraulic characteristics of the main. The analytical framework is used to implicitly consider the probability and consequence of failure.  This framework has been developed to consider risk in answering two specific questions:

 

“What treatments should be applied under what conditions to sustain the sewer system above acceptable levels of service at minimum long-term cost?”

 

and,

 

“Under what conditions should the utility spend money to video camera sewer mains to obtain better information about the condition of the sewer?”

 

In answering the first question, available treatments, the extent of the treatment, the cost of these treatments and the risk of future failures are considered in estimating the long-term costs.

 

A sewer can either be lined or replaced. Lining is significantly less expensive than replacing the main but there are constraints that may require an excavation to repair a main before lining.  These treatments result in a sewer that is expected to have an economic life in excess of 50 years.  The other alternative is to Do Nothing.

 

The extent of the treatments is simply considered as spot, partial or full replacement.  Applied to both replacement and lining, the definitions are used to differentiate between the rehabilitation costs and the technical issue associated with these treatments.

 

The long-term cost of maintaining the sewer includes the costs now and in the future.   Replacing or lining the entire sewer effectively reduces the probability of failure to near zero for the foreseeable future, but the capital costs are spent in the present.   Do Nothing means that capital cost are deferred to the future.  The utility accepts the risk of future maintenance and damage claims and defers the rehabilitation cost to a future date.  The rehabilitation costs may also be significantly higher due to further deterioration of the main, but could be offset by the economic benefits of deferring the capital cost of replacement.

 

A third option is to undertake partial treatments that eliminate defects with the largest impact on the risk of failure.  This strategy expends funding on partial treatments to reduce but not eliminate the risk.  Future repair costs and rehabilitation are expected and considered in defining long-term costs.

 

If the components of risk can be reasonably estimated, risk and costs can both be expressed in monetary terms.  Then the goal would be to minimize the long-term costs plus risk in sustaining the sewer main system, subject to minimum acceptable levels of service and available budget/resources.  In sewer main systems, the probability of failure, given what is known about the conditions around and of the pipe, is difficult to estimate.  The problem becomes one of balancing risk and costs in accordance with the community’s values.

 

The authors ask the second question, “Under what conditions should the utility spend money to video camera sewer mains to obtain better information about the condition of the sewer?” to improve our knowledge about the main and therefore improve our knowledge about the probability of failure. Selecting a main for video inspection is a technical and economic decision to expend funds to gain information about that sewer.  Deciding to inspect means that any improvements in estimating risk as a result of the additional knowledge exceeds the cost of the video inspection and evaluation.  Deciding not to inspect means the improvements in estimating the risk is not worth the additional cost.

 

Video inspection includes the evaluation of the condition of the main.  The methodology developed by the North America Association of Pipeline Inspectors (NAAPI2,3) is used in this work to quantify the sewer main condition. The NAAPI score at a sub-segment level is used as a proxy for the probability of a failure based on pipe condition.  That is, the higher the NAAPI score, the higher the probability of a failure. Video inspection is required for planning a dig or a lining.

 

If a NAAPI score was the true estimate of the probability of failure, then risk and costs could be summed to reflect a total expected cost.  However, because NAAPI scores are a proxy to the probability of failure, the concept of thresholds needs to be introduced to permit the balancing of risk and cost.  A threshold then, is a boundary between the conditions (represented by NAAPI score) above which a treatment would be undertaken and the conditions under which the treatment would not be undertaken.  A threshold implicitly represents a utility’s tolerance to risk.

 

Selecting a threshold also defines the utility’s cost of rehabilitating a sewer main.  By considering different thresholds and the estimates for a rehabilitation cost model, the analytical framework defines a rehabilitation cost for each threshold.  These costs can be aggregated over the sewers that have been evaluated to estimate the total funding requirements.

 

In the following sections, the authors identify various sources of information that help in deciding whether or not to video inspect.  ArcGIS is used to bring that information together in a common format and to display any spatial relationships among the various factors, which may affect the decision.

 

 

VIDEO SELECTION PROCESS

 

In a perfect world all necessary information about the sewer network is available and can be evaluated in making a perfect decision.  In reality no utility can afford to collect all the information that would lead to a perfect decision. In fact, it is very possible that the collection of additional data would cost more than the potential savings resulting from even a better decision.  For most utilities it is the information at hand that should be considered in making video inspection decisions. 

 

In the City of Saskatoon, the municipality has been collecting and electronically storing attribute and operations data on the sewer network since 1990. This information is similar to information on activities that is maintained by all utilities.  In the analysis framework, the information is considered outside of the traditional reasons for capturing the data.  Figure 1 shows the available input from these databases that is considered in quantify, the risk of failure.

 

 

Figure 1: Available Input for the City of Saskatoon

 

Information on soil conditions and ground water represent additional data available in reports but not in a computer database.  This and other types of information may be collected in the future if it is determined that it would change a rehabilitation decision and that the decision would result in a better economic or service delivery outcome.

 

Staff Input

 

Few if any data management and analysis systems match the knowledge, experience and skills of the staff.  For this reason, input from the staff is a very important source of information about the sewer system. Direct and anecdotal information about the system greatly improves the understanding of risk.  Bringing that data into the ArcGIS platform means that the information can be stored, analyzed, and quickly reviewed within databases and a spatial environment.

 

In the City of Saskatoon, the initial meeting began with a print of the sewer network and a box of colored pencils.  Staff was asked to freely identify locations of concern where they felt that an inspection and evaluation should be done to determine the sewer main condition and rehabilitation.  This “paper and crayons” method identified fifty locations for consideration. The information was transferred to ArcGIS and became the initial source of information. Staff recommendations were intuitively based on risk.  Figure 2 shows the staff identified locations in red.


Figure 2: Staff Identified Location and Identified Groups

 


From previous work and discussions with staff, it became obvious that there were certain key characteristics that lead to sewer failures (Sexsmith4). Figure 2 also illustrates Identified Groups of sewers that are more likely to have structural defects and will likely have a higher probability of failure. These include:

·        Clay tile sewers greater than 5.5 m in depth (shown in orange).

·        Non-reinforced concrete pipe (shown in green).

·        Concrete sewers immediately downstream of force mains (shown in blue).

·        Major food processing plants discharges to a concrete sewer (none shown).

 

Input from staff represents an informal analysis of consequences of failure, as staff is well aware of the outcomes of failures. They also consider the probability of failure by identifying clues that suggest sewer deterioration as part of their daily activities.

 

Hydraulic Input

 

Development of a hydraulic model for the sewer network is a key step in determining the consequences of a failure.  At the time of this work, a detailed hydraulic model was under development, but not available. Information on the depth of flow is collected during manhole inspections, sewer-cleaning activities and during video inspections.   This information, along with attribute data, is used to estimate velocities and flows using Manning’s Equation.

 

In this study, the sanitary sewer and storm sewer systems are separated.  However, a high percentage of buildings in the City have weeping tile systems connected to the sanitary sewer system.  Maximum flow occurs during intense rainfall events. The observed height of flow is considered as the dry weather flow.  The normal or dry weather flow is multiplied by a load factor to estimate the expected maximum flows during rainfalls.

 

The information has a number of uses in the selection process. These include:

·        Determining the consequence of sewer collapse at key locations in the sewer network.

·        Identifying sewer flows that exceed available emergency sewer pumping capacity.

·        Identifying low velocities in concrete sewers, which leads todeterioration due to hydrogen sulfide attack.

·        Defining the threshold for sewers that can be lined without flow testing or a detailed design review.

 

Operationally Critical Sewers

 

The Water Research Council (WRc5) defines specific sewers as “Hydraulically Critical,” based on the amount of sewerage that the main carries.  The authors take a slightly different approach to the WRc definition by considering sewers for which the utility currently does not have the ability or resources to quickly divert sewage around a collapsed main. One option to costly sewer rehabilitation is to develop operational procedures that utilize response plans, sewer backup alarms, bypass pump equipment or alternative methods of diverting flow around a sewer collapse. In a comprehensive Asset Management Strategy for the sewer network, the ability to react to main failures can greatly reduce the consequences of failures and may allow the utility to accept a higher probability of failure. 

 

Termed Operationally Critical Sewers, these include major sewers with flows exceeding available emergency pumping, force mains and around lift stations. It also includes sewers that cross major roads and locations where a sewer backup would be unacceptable including hospitals, level-care homes, and could include schools and other public buildings. Figure 3 shows Operationally Critical Sewers based on operations considerations.  Segments indicated in red exceed the current pump capacity of 3,000 gallons (US) per minute using conventional pumping procedures.  Sewers that cross major roads are shown in brown, arterial roads in yellow, river crossings in orange and force mains in green.  Sewers servicing hospitals are shown in blue.

 


Figure 3: Operationally Critical Sewers

 


Low Lying Areas

 

The outcomes of a sewer collapse can be very different depending on the number of customers in low-lying areas of a sewer catchment. Every utility has mains where customers’ basements are near the crown of the sewer main.  On a main serving a large number of upstream consumers, little time is available to react to a backup situation.  For these reasons, mains downstream of these locations pose a greater consequence of failure. Figure 4 illustrates low-lying areas at risk for sewer backups.


Figure 4: Critical Locations-Low lying Areas

 


Low Velocities in Concrete Sewers

 

Concrete sewer mains in which the flow velocities are low are susceptible to hydrogen sulfide attack. Figure 5 shows a portion of the Saskatoon sewer network.  Concrete sewers with low velocities are identified in red. Concrete sewers with low velocities are considered to have a higher risk of structural failure.

 

                                                                                                                                                          


 


Figure 5: Estimated Velocities for Concrete Pipe.

 

 

Operational Input

 

Operations input differs from staff-identified locations in that information is collected during inspection, cleaning and repair activities.  Relying on field observations rather than the memories of individuals, locations are identified and documented within the spatial model.

 

Manhole Inspection and Sewer Cleaning

 

Sewer cleaning is a major activity in the City of Saskatoon’s sewer operations strategy. Sewers are cleaned at frequencies varying from 6 months to 5 years.  During cleaning, staff observes and documents the materials removed from the sewer.  Pieces of clay tile pipe, concrete debris, and bricks from manholes are direct evidence of ongoing deterioration.  It signals a change in the sewer condition and an increased risk of failure in the sewer being cleaned or in an upstream main.

 

At the time of this paper approximately 30 percent of the system had been evaluated during cleaning and the information documented in ArcGIS. Figure 6 shows the results to date.  Red segments indicate that clay pipe pieces were found in the segment during cleaning.  Blue segments indicate that gravel was removed during the process indicating concrete pipe deterioration.


Figure 6: Sewer Cleaning Triage

 


Information on tree roots, grease and debris is also collected during the process using definitions developed by staff.

 

Sewer Backups

 

Although not illustrated in this work, locations where sewage backs up into homes and businesses represent a significant cost to the utility in damage claims or to the community as high insurance premiums.  A similar analysis should be carried out where the utility has documented this information in a spatial model.

 

Historical Target Input

 

With the flexibility of the ArcGIS platform, it is possible to look at a number of different factors. Repairs and rehabilitation locations are considered given the characteristics of the sewer main including age, depth, location, and pipe material. By considering pipe characteristics, the factors that led to this work can be evaluated and applied to other sewers in the network.

 

Sewer Repairs and Rehabilitated Sewers

 

Sewer repairs are completed for a number of technical reasons, but from a customer’s point of view it is to minimize the risk of sewer backup. Repairs from 1995 to 2001 are illustrated in Figure 7 as red lines. Also illustrated are the rehabilitation projects from 1982 to2001. Sewer lining projects are shown in green and sewer replacements in blue. 


Figure 7: Sewer Repairs and Rehabilitation Projects

 


Information on those segments either repaired or rehabilitated represents locations with a higher probability of failure. Those locations are considered in relationship to other locations with the same characteristics.  The detail in Figure 7 illustrates the use of dynamic segmentation to indicate not only the location of the repair but also the extent of the repair.   In this example, segment 6386 and 6653 are shown in red as locations initially identified by staff.

 

Video Inspection

 

In this case study, the City has 80 km of main that were evaluated in the early 1990’s. Although the evaluation method was different than the current methodology, the data can be utilized to consider the timing of future inspections.  Figure 8 illustrates locations in the 1990-92 studies. Because the information is 10 to 12 years old, it is divided into groups and considered under broad definitions. Locations that were identified as having few or no defects were considered as being in good condition and are shown in blue.  Sewers in fair condition are shown in yellow and sewers that were observed in poor condition are shown in red. 


Figure 8: 1990-1993 Sewer Video and Analysis

 


These locations are considered in conjunction with mains that have been evaluated under the NAAPI process in defining the risk of failure.  Unless repairs have been undertaken, it is reasonable to expect that sewers that were in poor condition are deteriorating and have a higher probability of failure than when inspected in the early 1990’s.

 

Water Main Failures

 

Sewer mains are influenced by water main breaks.   This relationship is unique in that the metallic water main fittings fail due to corrosion indicating a clay soil in the pipe zone.   Clay soils tend to extend the life of clay tile sewers by holding the fractured pipe in place.  This is in contrast to the effects that escaping water from a main break has on the sewer. Infiltration under pressure and the washing of bedding material into the sewer pipe accelerates deterioration.

 

Figure 9 illustrates the location of water mains repairs 1990-2001in one neighbourhood. Because of the different information sources breaks are illustrated in a variety of ways. Since 1995, failures have been linked to the GIS numbers for each water main segment. Failures from 1995-2001 are shown as blue lines coded with the number of failures. Failures before 1995 were recorded by civic address. These are shown as blue dots against each site. In this example, service clamps were installed on nonmetallic pipe.  Clamps were either replaced under emergency repairs and are shown in purple, or as part of a replacement program shown in green.


 


Figure 9: Water main failures 1990-2001 and Sewers influenced by Water Main Failures

 

Using the Selection by Location in ArcMap, the sewer mains influenced by the water main failures are also illustrated in Figure 9.

 

 

Other Utilities Input

 

Other utilities in the right-of–way will influence rehabilitation decisions.  Where a number of utilities will conflict with the excavation of a sewer main for the purposes of either repair or replacement, the utility should consider lining all or part of the main before the sewer deteriorates to the point where lining is not an option.  Where other utilities approach their threshold for rehabilitation, a joint or combined replacement is considered depending on the location of underground utilities within the right-of-way (Higgins, Gustafson, Clancy6).

 

 

SELECTING VIDEO INSPECTION CANDIDATES

 

Based on the analysis of available information, the utility will decide to undertake video inspection based on their understanding of the risk.  In reviewing the available input, the following general statements were used in defining video inspect for the Saskatoon case study.

 

1.      Staff input is the first consideration in video inspection decisions as staff is keenly aware of the consequences of failure and can effectively consider the probability of failure.

2.      Direct information from manhole inspections, sewer cleaning, past videos, sewer backups, repairs and rehabilitation provides evidence of sewer main condition and the probability of failure.

3.      The ability of the utility to react to failure at any point in the network is a key consideration in the consequence of failure in the system.  The flow in the sewer, the utility’s response capabilities, the proximity of basement elevations to the sewer and the number of upstream users are considered in the consequence of failure.

4.      Joint projects are a good economic decision when the water and sewer mains are at the same elevation and one or both are close to meeting the threshold for rehabilitation.   Where the sewer is substantially deeper than the water main, the economics of trenchless rehabilitation of the sewer and replacement of the water main could be a good decision.  The thresholds for inspection include consideration of the proximity of the sewer to other utilities and the condition of the utilities and road treatments.

5.      The probability of failure for clay tile pipe considers staff identified locations and operations identified locations.  Previous repairs and rehabilitation, past video inspection, surcharge in the sewer and water main failures in proximity to the sewer are also considered in determining the threshold.

6.      In considering the probability of failure of concrete pipe, staff identified locations and operations identified locations, previous repairs and rehabilitation, and past video inspection are also considered but in this case concrete sewer with low velocities is a contributing factor.

 

Within this analytical framework, a significant amount of information can be accessed within ArcGIS in the decision process to choose video inspection candidates.  Because the probability of failure is difficult to estimate in a sewer system, the benefits of understanding the risk of failure become an important factor in balancing risk and costs in keeping with community’s values.

 

SEWER CONDITION EVALUATION

 

With video inspection completed on the identified segments, the next step is to evaluate the video and determine the need for rehabilitation using the NAAPI evaluation process. This process was developed to allow utility owners to obtain all information about a sewer main assuming that it has never been inspected and in some cases cleaned.  It allows for a numeric representation of the structural and operational sewer condition.  Information is collected within accepted and documented definitions and stored as either spot defects or as continuous defects.

 

Defects are entered into an Access database and can be displayed in tables or graphically. The NAAPI sub-segment score is illustrated in Figure 10.  The Access program allows for the accumulation of defects scores in 1-metre increments along the segment. Some of the coding within the NAAPI process is not related to the structure of the pipe segment (Jennings7). These are shown as negatives in Figure 10.

 


Figure 10: Access Graph of Defects Identified Using the NAAPI Process.

 


This information can be shown in ArcGIS using dynamic segmentation and route event tables generated from Access. Examples of route events are entered as points as illustrated in Table 3, or as lines in Table 4.

 

Table 2: Route Event Table for Continuous Defects

GIS Number

From  Station

To Station

NAAPI Defect Code

NAAPI Defect Weighting

82

81

78

CL

10

82

88

85

CL

10

82

92

90

CL

10

82

9

7

CL

10

Table 3: Route Event Table For Point Defects

GIS Number

From  Station

 

NAAPI Defect Code

NAAPI Defect Weighting

82

3

 

CC

10

82

5

 

CC

10

82

6

 

CL

10

82

7

 

CL

10

CONTINUED

 

Using the Route Editor in ArcMap, a route is identified by combining one or more segments and defining a start point and end point.  In development of this model the upstream manhole is the starting point and the down stream is the endpoint with the route being equal to the length of the main between manholes.  This convention was chosen based on the standard procedures for video inspection and allows for spatial representation of defects.

 

Using the data from the above tables and the Tools Add Route Events procedure in ArcMap, the defects are entered as elements on the sewer main and illustrated in Figure 11.

 


Figure 11:  Major Defects, Minor Point Defects and Continuous Defects in ArcMap.

 


Major defects are evaluated as possible spot excavated repairs and shown separately as major defects.  Minor defects are shown as either point or line information and are shown separately in Figure 11.

 

With the defects entered in the spatial model and quantified using the NAAPI scoring process, an analysis is carried out to determine alternative treatment strategies on a sub-segment level.

 

Thresholds for Rehabilitation

 

In the conventional NAAPI evaluation, soil conditions and surcharge are considered for each segment and the segment is classified with a grade 1 through 5 based on a need to consider rehabilitation.  The results are forwarded to the client and a determination of treatment is left to the utility owner or an engineering consultant. In this work, thresholds are defined in determining when a utility will initiate a treatment. The following general statements reflect selection of treatments for the City of Saskatoon:

 

1.      Emergency repairs are undertaken to maintain service or ensure service where the risk and/or consequences of failure are deemed to exceed the cost of repair or represent an unacceptable risk to the public.

2.      In a non-emergency situation, lining will be considered unless the outcome of lining could result in future backups or the severity of the defect will not permit conventional lining.  Where concrete pipe is cracking or has exposed reinforcing steel, the main will be considered for immediate lining.

3.      Where lining is reasonable, consideration will be given to spot lining, blind lining and then full lining based on the cost-benefits.  Under normal circumstances, spot lining will be used to rehabilitate clay tile sewers.  Because of the cause and uniformity of deterioration due to hydrogen sulfide attack, full lining will be utilized in rehabilitating concrete pipe.

4.      Where the costs of repair are significant and other utilities are under consideration for rehabilitation, spot repairs (Band-Aid) will be considered as a short-term solution.

5.      Where a location cannot be lined, excavated repairs will only be undertaken on that portion required to permit lining.

6.      Where necessary, a combination of lining and spot repairs will be considered to minimize the cost.

7.      Only when full replacement is the most economical solution should the main be replaced.  This will invariably occur where the water main is failing and can be economically replaced.  It may also be necessary where the sewer main needs to be upsized to meet existing or future demands or where the deformity of the sewer will not permit lining.

8.      Prioritization of rehabilitation candidates will consider the possible outcomes in deferring rehabilitation particularly where other utilities or surface assets, soil conditions, bypass pumping requirements, detouring and other factors that could greatly increase the costs of emergency repairs.

 

Within the scope of this work, spot lining is not only considered as a long-term solution to a deteriorated pipe, it is also considered as a short-term or band-aid solution.  The authors suggest that, within the goal of minimizing long-term economic cost, it may be more cost-effective to spot line severely deteriorated segments of a main to minimize the possibility of collapse and defer the capital cost of lining or replacing the entire sewer main. This has particular advantages where other utilities or surface assets are close to rehabilitation and the opportunity exists to defer treatment in order to optimize rehabilitation across assets within the right-of-way.

 

The thresholds can also be applied independently to each segment.  This allows for consideration of sewer characteristics, flow, soil condition, ground water, the proximity to other utilities and street classification in determining thresholds. 

 

By considering ranges of thresholds, a utility considers not only what would be rehabilitated on a segment but also what would be left after rehabilitation.  In effect, by setting lower or higher thresholds for different segments, the utility is estimating the probability of future failures. Using the NAAPI defect descriptions, the decision can be considered in terms of what defects would be left if the main were not fully rehabilitated in a given set of circumstances. 

 

Where a utility has a limited rehabilitation budget, setting a high threshold defines locations with the highest risk of failure for considering spot treatments.  In Figure 12, a threshold of 100 points is applied to segment 82.  The threshold would result in treatments being undertaken at four locations along the pipe.

 


Figure 12 Threshold for Rehabilitation

 


Extent of Rehabilitation

 

The other consideration in sewer rehabilitation is the extent of a treatment. In modeling the alternatives for treatment, the authors utilize cost models that consider a fixed cost plus variable cost for each treatment based on past work. In estimating the cost-benefit for each treatment, the cost is evaluated based on the number of NAAPI points that would be eliminated per $1000 of rehabilitation spending.  In the case of spot lining, the extent of the repair varies from 1 m. to 10 m. For blind lining, the repairs are considered from either end of the sewer segment for the full length of the pipe.  Similarly, excavated repairs are from 3 m to the full length of the pipe.  Table 4 illustrates the method for determining the optimum extent of a repair. In this example, a spot repair from station 11 to 13 would result in the maximum NAAPI points per $1000 being eliminated.

 

Table 4:  Optimization of Treatment Based on Points per $1,000

Length

Points

Point/$1000

From

To

1

140

46.7

12

12

2

270

72.7

11

12

280

75.3

12

13

3

270

60.3

10

12

410

91.6

11

13

280

62.6

12

14
  •  

10

280

34.7

12

21

 

 

In each case all possibilities are considered and the selected repair strategy or combination of strategies is determined that will optimize the number of points that are eliminated per $1,000. Figure 13 illustrates the remaining sub-segments scores that would remain if the threshold were set at 100 for segment 82.

 


Figure 13: Remaining defects based on Threshold of 100

 

 


Treatment Strategy Selection

 

In considering a segment, spot excavations are undertaken at locations where the main has deformed or collapsed or where the conditions would require an excavation and repair before lining due to holes, broken pipe or offset joints.  In Figure 11 Major Defects, three holes are identified at 11, 43 and 83 m from the upstream manhole. If these defects would not permit lining, an excavation would be undertaken to make the necessary repairs. The extent of the excavation would depend on a cost benefit review based on the number of NAAPI points per $1,000 with a minimum repair length of 3 m.  Where these repairs are undertaken, the analysis of the lining strategy is considered based on the elimination of the defects at these locations. 

 

Different thresholds and the treatment options are shown in ArcGIS using dynamic segmentation. Figure 14 illustrates the different possibilities for segment 82 for various thresholds that the utility may wish to consider.  In this example the major defects were linable.


Figure 14:  Options for Treatment In ArcGIS

 


The development of this process means that an Asset Preservation Plan can be defined for each segment that has been evaluated. Where multi-block segments have been identified as meeting their rehabilitation threshold that analysis can be repeated for the entire length and alternative asset action plans can be considered based on undertaking a larger rehabilitation project.

 

Annual Program and Backlog

 

For those segments now evaluated, it is possible to determine an estimated cost to repair all defects for a given threshold. Figure 15 shows the estimated cost of rehabilitation of 27 segments in the initial database. In this example all thresholds are illustrated to provide a means of comparing the cost to the selected threshold.   It is also possible to assign independent thresholds for each segment and determine the annual funding.

 


Figure 15:  Budget Demand Versus Threshold

 

The funding demand is evaluated against the available funding and a backlog is identified where demand exceeds the budget.  Where backlog is identified, projects are considered in light of the economics and consequences of failure identified during video inspection and the NAAPI Scores to evaluate the probability of failure. This prioritization is used to define the annual program and those candidates deferred to next or subsequent annual program.

 

Staff Review and Process Improvement

 

A review of thresholds is carried out as new information is identified and added to the sewer network knowledge. With the first cycle complete and an annual program generated, a review is undertaken to consider the findings and fine-tune the process.  Staff reviews thresholds annually at the end of the construction season.  With this review complete, the process is repeated in determining sewers for video inspection and rehabilitation. 

 

The initial program was based largely on staff input regarding locations that need to be considered for video inspection and evaluation. Staff identified a number of locations where the results of the NAAPI score was between 500 and 19,000.  By comparison the NAAPI process recommends serious consideration of locations that score above 165, suggesting that staff have a good understanding of sewers in poor condition.

 

Ideally, the analytical framework would generate a rehabilitation strategy that would minimize the long-term costs.  The difficulty in defining the probability of failure means that utilities must balance the risk and costs in providing sewer service.  Improving understanding of risk of failure and the community’s values will allow the utilities to refine thresholds that reflect customer’s acceptance of the risk of sewer backup and their willingness to pay for their sewer service.

 

 

CONCLUSION

 

This analysis framework is highly dependent on data from a variety of sources.  ArcGIS facilitates the integration of available data into a spatial environment and provides the tools required to analyze the risk of sewer failures in determining camera inspection priorities and rehabilitation strategies.

 

 

REFERENCES:

 

 

Clancy, D.V., Gustafson, J-M; Forecasting Outcomes Versus Costs In The Repair And Replacement Of Cast Iron Water Mains Proc. 2000 AWWA Infrastructure Conference, Baltimore (2000)

 

Canderbank, Ron; North American Association of Pipeline Inspectors, CCTV Certification and Reviews Course, Surrey, B.C, Canada,

 

UK Water Industry Engineering and Operation Committee; Manual of Sewer Condition Classification, Third Edition, 1993 (Amended February 1996)

 

Sexmith, W.C.; Setting Rehabilitation Priorities For Sanitary Sewers In The City Of Saskatoon, Master’s Degree Project, University of Saskatchewan, 1997.

 

WRc; Sewerage Rehabilitation Manual, Second Edition, Water Research Centre, Swindon England, 1986

 

Higgins, L, Gustafson, J-M. & Clancy, D.V.; Integrated Infrastructure Management Decision-Making

Using ArcGIS, Proc. 2002 Esri Conference San Diego CA. (2002).

 

Jennings, Roger; Special Edition Using Microsoft Access 2000, Que Corporation, Indianapolis, IN, 1999.