Managing Selected Transportation Assets: Signals, Lighting, Signs, Pavement Markings, Culverts, and Sidewalks (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

OVERVIEW Pipe culverts and box culverts (referred to collectively as cul- verts in this chapter) allow for drainage under and around highways, streets, and sidewalks, providing stability to the road structure and preventing flooding of surrounding areas. Culverts are also increasingly being used for animal passages under highway embankments. Agencies participating in the synthesis survey ranked the transportation objectives that are served by good drainage in priority order, as given in Table 23. Agencies added prevention and reduction of flooding to the objectives listed in the table. Meeting these objectives calls on agencies to observe stan- dards, technical recommendations, and guidelines from a variety of sources. Figures 70 and 71 present agencies’ judg- ments of those sources of guidance that are the important drivers of engineering and management decisions regarding culverts. These results are shown for two key aspects of asset management: new construction and installation, and mainte- nance and rehabilitation, respectively. The importance of individual agency policies, standards, guidelines, and proce- dures, together with national standards for design and statu- tory requirements, is evident for design of new systems or major expansion of existing systems. National standards and guidelines have been published by the FHWA and AASHTO in several relevant topics, including culvert design (e.g., High- way Drainage Guidelines 1987; Brown et al. 2001; Normann et al. 2001); culvert inspection (Arnoult 1986); other areas of hydraulic performance, highway drainage, and scour, as well as guidelines on the safety of roadside features such as drainage headwalls (Roadside Design Guide 2002). Statutory guidelines encompass, among other topics, state requirements for fish passages, which are served by culverts. Although national guidelines exist for maintenance and rehabilitation (e.g., Ballinger and Drake 1995; Maintenance Manual . . . 1999), guidance issued by individual agencies is the primary technical source, as indicated in Figure 71. MANAGEMENT PRACTICES Agencies in general maintain their own drainage systems, as indicated in Figure 72. Among agencies participating in the survey, the use of contractors or other government agencies is limited to DOTs; cities and counties reported no outside entities to conduct or manage drainage work. New Brunswick noted that forestry companies maintain some 88 culverts and these companies do exercise management responsibility. Ohio voiced a concern that current policies need to better account for the life expectancy of pipe mate- rials, and to consider paying a premium up-front for longer- lasting pipe. Under current procedures, a range of pipe materials and coatings is allowable, and contractors decide which to install based on price. Current procedures need to account better for the actual service life to be expected. The Ohio DOT has developed a workbook to evaluate pipe dura- bility during design. Other aspects of asset management practice are revealed through an agencies’ methods of budgeting for preservation, operation, and maintenance of culverts, and their approaches to preserving and maintaining culverts once in service. Survey results for the budgeting method are shown in Figure 73. Explanations of the abbreviated budgeting process descriptions in this figure are given in chapter two. Because agencies could select multiple choices, the percentages in Figure 73 do not sum to 100%. Addressing their methods of budgeting, responding agencies at all levels of government chose a variety of options to best describe their process. The role of professional judgment continued to be important, as it has been for other assets to this point. Many agencies selected multiple options. Texas described a formula that allocates drainage funding among Texas DOT districts based on the proportion of vehicle-miles traveled within each district as compared with the statewide total, and a rainfall factor based on annual rainfall within the district compared with average annual rainfall statewide. Agencies often described their approaches to preserva- tion and maintenance as well in terms of multiple selections of the items shown in Figure 74. Immediate correction of problems was the most prevalent response, followed by the worst-first, prioritized, and preventive approaches. Many agencies explained the multiple approaches that they em- ploy by differentiating how and when they are used. For ex- ample, “immediate work” would be applied to sudden fail- ures; “corrective work” to the biennial inspections of large culverts (similar to those for bridges) or to work managed by the agency’s maintenance management or maintenance quality assurance approach, and “worst-first” to aged cul- verts beyond their design lives that need to be rehabilitated or replaced. In a survey reported in NCHRP Synthesis 303, CHAPTER SIX DRAINAGE CULVERTS

89 27% of responding state and local agencies reported having a preventive maintenance program, as compared with the 20% reported in Figure 74. MEASURING ASSET PERFORMANCE The information provided by agencies on performance mea- surement of culverts is summarized in Figure 75, based on cat- egories of performance factors similar to those described in chapter two. Many agencies reported measures of physical condition and the corresponding qualitative descriptors, and customer complaints as their main indicators of performance. Individual agencies included other physical measures; for example, blockage, geotechnical/embankment risk, pavement or embankment settlement, standing water, and so forth. One agency mentioned basing performance on service of the culvert as a fish passage for environmental objectives. The frequencies with which state, provincial, and local agencies reported conducting their physical performance measures are shown in Figure 76. Another survey conducted by the Ohio Research Institute for Transportation and the Environment (ORITE) on behalf of the Ohio DOT indicated that 48% of DOT respondents inspected their culverts on a frequency of No Response Other Agency Guidelines Public Policy Natíl. Standards Statutes Percentage of Responses 0 20 40 60 80 100 No Response Other Agency Guidelines Public Policy Nat’l. Standards Statutes Percentage of Responses 0 20 40 60 80 100 FIGURE 70 Technical management guidance for new construction and installation of culverts. FIGURE 71 Technical management guidance for maintenance and rehabilitation of culverts. Rank Factor 1 Public safety; accident and accident risk reduction 2 Preservation of the existing road infrastructure; reduced agency life-cycle costs 3 More efficient travel; maintain intended flow and operating speed; reduce travel time 4 Comfort and convenience of the traveling public (motorists, pedestrians, cyclists) 5 Road aesthetics and appeal TABLE 23 PRIORITY OF TRANSPORTATION OBJECTIVES SERVED BY DRAINAGE CULVERTS

0 20 40 60 80 100 No Response No Specific Approach Other Pct. of Total Budget Judgment, Politics Previous + Adjustments Pct. Inventory Annually Budget Drives Target Target Drives Budget Percentage of Responses FIGURE 73 Annual budgeting method for culvert preservation and maintenance. 90 0 20 40 60 80 100 Own Agency Private (Outsourced) Other Gov’t. Unit Other Entities No Response Pe rc en ta ge o f R es po ns es Mgmt. Resp. No Mgmt. Resp. FIGURE 72 Responsibility for maintaining culverts once in service. 1–2 years, whereas 16% reported a range of frequency of 3–5 years, with some states having dual frequencies (e.g., inspecting culverts with larger than a 10-ft diameter at 1–2 year intervals and smaller culverts at 5-year intervals). The remaining 36% reported other ranges of inspection intervals; for example, 1–4 years (Mitchell et al. 2005). These results collectively show that there is no standardized inspection frequency among transportation agencies. Although the survey in this synthesis study did not go into detail on the specific physical measurements that agencies use in assessing culvert condition, NCHRP Synthesis of Highway Practice 303: Assessment and Rehabilitation of Existing Culverts (Wyant 2002) reported that the pipe as- sessment factors cited most frequently by state, federal, and local agencies included joint failures, corrosion, deflection, and cracking. Several agencies also cited hydraulic capacity, soil conditions, and pipe wall thickness. A few noted silt ac- cumulation, debris, clogging, settlement, and scour. Only 15 of 59 state, federal, and local respondents (25%) to the NCHRP Synthesis 303 survey reported having formal guide- lines to assess pipe condition; among state DOTs alone, 10 of 27 (37%) had such guidelines (Wyant 2002). The survey by ORITE indicated that 60% of responding DOTs have some type of inspection policy for highway culverts, but only 12% have developed their own culvert inspection manual (Mitchell et al. 2005). ORITE identified Arizona, California, Connecticut, Indiana, Kansas, and Ohio as having such man- uals; NCHRP Synthesis 303 included Maine, New York, and Pennsylvania. Other agencies may have their own manuals, but might not have been included in survey results or men- tioned in the cited reports. Also, several agencies apply the FHWA Culvert Inspection Guidelines, as discussed here. Culverts of more than 20 ft in span, or a series of adjacent culverts that add up to a crossing greater than 20 ft in length, are included in the National Bridge Inventory. They are inspected in the United States as bridges and are therefore sub- ject to FHWA’s National Bridge Inspection Standards (NBIS) requirements and an agency’s bridge inspection guidelines, data collection and processing procedures, and related man- agement tools and decision criteria. Below a 20-ft-span width there are no NBIS-required inspection intervals, and agency practices differ on what defines a culvert. Approximately two- thirds of state DOTs responding to the ORITE survey reported that they apply the AASHTO definition (span less than or equal to 20 ft). Others have adopted different limiting span widths (e.g., less than 6, 10, or 15 ft) or other definitions based on a “drainage” concept, or have no definition yet in place (Mitchell et al. 2005). NCHRP Synthesis 303 (Wyant 2002) also indicated a wide range of practice in the rating methodology of culvert

0 20 40 60 80 100 No Response Other No Maintenance Responsibility Deferred Maintenance Worst First Prioritized—Avail. Res. Corrected Immediately Preventive—Schedule Percentage of Responses FIGURE 74 Approach to preserving and maintaining culverts. 91 inspections and assessments. The Utah DOT, Vermont Agency of Transportation (AOT), and many local agencies use the method in FHWA’s Culvert Inspection Manual (Arnoult 1986). These culvert inspection guidelines have been incorporated in the NBIS in items 61 and 62 for channels and culverts, respectively (Recording and Coding Guide . . . 1995). The NBIS standards continue to refer practi- tioners to the 1986 Culvert Inspection Manual for additional details on culvert inspection, including specific rating guide- lines for individual pipe materials and photographs illustrat- ing rating levels. The specified rating scheme in the 1986 and 1995 FHWA manuals is a 0–9 scale, analogous to that used for bridge items in NBIS, where 9 denotes a new condition and 0 signals a totally failed culvert requiring replacement. Because the 1986 FHWA guidelines apply only to concrete pipes and corrugated steel pipes (CSP), however, additional PHYS: Structural Condition PHYS: Corrosion QUAL: Debris Accumulation PHYS: Other Asset Age System Reliability Performance or Health Index QUAL: Structural Condition QUAL: Corrosion QUAL: Debris Accumulation QUAL: Other Asset Value Customer Complaints Customer Surveys Other No Response 0 20 10 40 60 80 30 50 70 90 100 Percentage of Responses FIGURE 75 Measuring performance of culverts. PHYS = physical; QUAL = qualitative.

guidelines would be needed for other materials such as plas- tic pipe (Wyant 2002). Other agencies use different inspection processes that they have developed themselves or adapted from another agency. The California DOT (Caltrans) has instituted a Pilot Culvert Inspection program that is based on a 0–4 scale, with 0 indicating no deficiencies found and 4 signaling a critical con- dition. Guidelines and photographs (the set not yet complete) describe and illustrate these different severity levels for several conditions (e.g., waterway adequacy, pipe alignment and shape, and condition of joints, seams, and pipe wall material) for concrete, steel, aluminum, and plastic pipe barrels. The Caltrans guidelines also address drainage appurtenances such as concrete headwalls, flared-end sections of metal pipe, drainage inlets, scour at pipe ends, embankment and roadway condition, and metal riser pipe (Caltrans Supplement . . . 2003). In 2001, the Montana DOT developed a rating system for culverts that is based on 33 individual culvert attributes encompassing general information (location, site character- istics, installation date, etc.); culvert shape, dimensions, and height of cover; and indicators of existing damage mecha- nisms (e.g., age, corrosion, worn invert, side slope failure, piping, perched outlet, etc.) (Cahoon et al. 2002; Baker et al. 2006a,b). For purposes of field testing and calibration, an overall culvert rating was also requested of field inspectors on a 1–5 scale, with 5 indicating excellent condition and 1 de- noting poor condition. The 33 inspection items were believed to be all relevant and of potential value in a comprehensive culvert database. Of the 16 indicators of existing damage, however, only a subset of 9 of these was found to be statisti- cally significant in indicating overall culvert condition and need for remedial work. These nine culvert attributes included culvert age, scour at the outlet, evidence of major failure, degree of corrosion, worn culvert invert, sedimenta- tion of cross section, physical blockage, joint separation, and presence of physical damage (Cahoon et al. 2002). 92 NCHRP Synthesis 303 provides further examples of rating forms, guidelines, and summary reports from different agen- cies. Current agency experience emphasizes the value of clear guidance, photographs, and inspector training in ensur- ing a consistent and accurate inspection result, regardless of the guidelines and rating schemes used. Nonetheless, although 37 of 57 respondents (65%) to the NCHRP Synthe- sis 303 survey reported having an inspection program, only 27 of these agencies retained these condition records, a step that would otherwise be helpful to a preventive maintenance program (Wyant 2002). The methods used by responding agencies to assess culvert condition and performance are reported in Figure 77. Visual inspections are the most common method used, followed by physical measurements, photologging or vide- ologging, and fielding customer complaints. Ohio recently instituted a formalized culvert inspection program. The Ohio DOT inspects any culvert with a 10 ft or more combined span length every year (the Ohio DOT also inspects all of its bridges annually, as required by state statute). Culverts with a span length of less than 10 ft are inspected every 5 years. Under “Other” methods, Minnesota noted inspections of the deflection of steel plate arch pipes. Texas monitors culverts during floods to identify culvert obstructions, damage, ade- quacy of size, or any failures. Portland conducts dye testing and investigations with closed-circuit television. Survey results regarding the use of video equipment in Figure 77 are consistent with findings of the ORITE survey, which indicated that 30% of the respondents use special equipment for inspection of small culverts. This equipment typically consisted of a video camera, including robotic video systems and tractor-mounted video cameras. Some DOTs responding to the ORITE questionnaire noted that they do not inspect culverts smaller than a certain diameter; for example, 4 ft (Tennessee), 5 ft (New Jersey), or 6 ft (Vermont) (Mitchell et al. 2005). More Than Once A Year Annually Biennially Less Freq Than Biennially FIGURE 76 Frequency of physical condition assessments of culverts.

93 ASSET SERVICE LIFE Factors Affecting Service Life The service life of a culvert pipe is influenced by factors related to the pipe and its placement, the drainage water it carries, and the soil that surrounds it. Studies have shown that no single factor alone is an adequate determinant of ser- vice life. Nonetheless, design professionals and asset man- agers need a practical way to evaluate alternate pipe materi- als at each location to identify technically and economically feasible options; evaluate life-cycle costs (including predic- tion of service life); develop construction specifications; and project maintenance, rehabilitation, and replacement requirements. Agencies must therefore balance needed simplicity and practicality against comprehensive, detailed estimation of behavior. The inherent complexity in dealing with pipe durability is indicated by the number of mecha- nisms that can degrade culvert service life (Precast Con- crete Pipe Durability 1991; Gabriel and Moran 1998; Cahoon et al. 2002; Corrugated Steel Pipe Handbook 2005; ODOT Hydraulics Manual 2005). • Physical damage—Physical damage to the pipe can result from crashes by vehicles leaving the road, im- properly performed maintenance, fire, distortion of the pipe caused by applied loads that exceed the pipe’s structural capacity, and settlement. All of these mishaps can lead to reduced hydraulic efficiency and potential damage to the roadway foundation and surface. Settle- ment, which can be the result of improper backfilling, moisture in the roadbed, and exfiltration from leaking pipe joints, can contribute to roadway and side slope damage. Freeze–thaw cycles acting on moisture in con- crete pipe walls can cause spalling and lead to further damage as a result of chemical attacks, although the likelihood of this mechanism is reduced in pipe that is completely buried. Structural collapse of the pipe barrel is a failure both hydraulically and structurally, and can have serious consequences for the roadway pavement and foundation. • Abrasion—Abrasion of pipe material is caused by sands and aggregates (bed materials or “bed loads”) carried by water through the pipe. It is affected by the volume and velocity of the flow and the amount, size, and abrasiveness of material transported. If the pipe invert is completely abraded and worn away, the pipe can fail structurally. When abrasion exposes bare metal subject to corrosion, as in corrugated metal pipe (CMP), corrosion also often becomes a problem. Abra- sion in concrete pipe can be aggravated by chemical attack (e.g., from acids or sulfates), with the resulting combined damage greater than the sum of the individ- ual effects of these mechanisms. Abrasion of aluminum pipe can be a determinant of service life, because the metal is comparatively soft as compared with the abra- sives. Abrasion-resistant pipe materials and proper manufacturing methods, pipe coatings, and paving of the invert are used to resist the effects of abrasion. • Corrosion or chemical attack—Corrosion or chemical attack can occur from within or outside the pipe, and the literature differentiates between water-side and soil-side analyses of these mechanisms. Materials incorporated within the highway foundation design (e.g., lime-treated base) can affect corrosion, as can the chemical composition of the native soil. With concrete culverts, the possibility of chemical attack is increased when a low pH and soluble salts, particularly sulfates and chlorides, are present in soil or in drainage water, although sulfates and chlorides may be a problem more for cast-in-place concrete structures rather than for buried precast concrete pipe (Precast Concrete Pipe Durability 1991). Corrosion effects on the concrete No Response No Info. Collected Other Customer Complaints Customer Surveys Non-Destructive Testing Physical Measurement Photo, Video Visual Inspection 0 20 40 60 80 100 Percentage of Responses FIGURE 77 Data collection methods for culvert condition and performance.

cement and aggregates, as well as reinforcing steel, need to be considered. Acid, caused by acidic soil or aggressively acidic water runoff, can also degrade con- crete. If the drainage water is abrasive, chemical degra- dation of the concrete will accelerate erosion of the pipe wall surface and lead to a destructive cycle of cor- rosion and abrasion. For steel pipe, “most states have found culvert durability correlates with [soil-side and water-side] pH and resistivity; other states have been unable to confirm this” (Gabriel and Moran 1998, p. 17). Notwithstanding the importance of these variables, however, “predictions of useful service life based solely on pH and resistivity are inconclusive” (Gabriel and Moran 1998, p. 18). The presence of soluble salts, soil moisture content, and oxygen also effect corrosion of CSP in soil. The potential for corrosion may be increased on the soil side by soil moisture, soluble salts, and oxygen, and on the water side by abrasion of the steel or its coating, and the presence of soluble salts and dissolved oxygen or carbon dioxide in the effluent. Aluminum pipe is subject to pitting owing to soluble salts, stress corrosion cracking, and electrochemical corrosion. Although plastic pipe is generally resistant to pH and to chemical and electrochemical corrosion, it can be damaged by serious (albeit unlikely) highway spillage accidents involving concentrated acids and bases or prolonged exposure to high concentrations of certain organic chemicals such as crude oil or its de- rivatives. Corrosion and other chemical problems are inhibited by using nonreactive, corrosion-resistant pipe materials, coatings, and linings; providing cathodic protection; or installing an oversize pipe, anticipating future relining after corrosion has occurred. The Ore- gon DOT noted that water containing salt or chemicals can be very corrosive, and site-specific countermeasures are often required (ODOT Hydraulics Manual 2005, pp. 5–20). • Piping—Piping is water flowing through the fill sur- rounding the culvert barrel. It can result from poorly compacted fill around the pipe or improper or deficient end treatments that allow water infiltration outside the pipe barrel. Because piping can displace the fill that sur- rounds the pipe, it can lead to deformation of the culvert barrel as well as to settlement and damage of the road- way foundation and surface. • Other damage or failure mechanisms—Other failure mechanisms such as buoyancy, overtopping, and ero- sion or failure of side slopes can occur as the result of inadequate culvert design or sizing, inadequate protec- tion or armoring of slopes, and blockage of flow at the inlet or within the culvert barrel. Regarding sun-related effects on plastic pipe: although it can be considered prudent to protect the exposed ends of plastic pipe from sunlight, constituents are often added to the pipe mate- rial during its manufacture that can protect it from harm- ful UV rays. Other mechanisms mentioned in the liter- ature include localized corrosion such as pitting, crevice 94 corrosion, stress corrosion and cracking, and microbio- logic corrosion (Gabriel and Moran 1998). • Sedimentation—Sedimentation and debris collection reduce the culvert cross-sectional area and impede flow. Debris can collect at damaged ends of culverts or be de- posited inside the barrel; vegetation at either end may reduce the flow speed, act as a collector, and contribute to debris accumulation. Sedimentation can result from a culvert being installed too low, resulting in a backwa- ter pool at the downstream end. As the speed and flow of water is reduced, additional sedimentation can occur. Culvert sedimentation and debris are usually addressed as a maintenance item; only if ponded water resulting from continued inattention leads to more severe prob- lems will culvert repair or rehabilitation need to be con- sidered. Therefore, determination of culvert service life has many aspects to consider. The previous descriptions, however, greatly simplify the technical, often site-specific, variables, and potential damage mechanisms that need to be evaluated. As one complicating factor, damage mechanisms can inter- act with one another, as illustrated by several earlier exam- ples. Further complexity is caused by the role of the local am- bient environment. For example, the corrosive effect of chemicals in the soil may depend on the degree of soil arid- ity; the acidity of drainage water reflects not only local soil, rock, and rainfall conditions, but also surrounding activities such as mining; the decomposition of vegetation in steel pipes serving in a warm, wet climate can create organic acids that can lead to corrosion; and the effect of freeze–thaw cy- cles and thermal stresses on concrete pipe depend on the de- gree to which the pipe is buried, which reduces atmospheric exposure. The physical and chemical details of possible reactions need to be recognized with sufficient understand- ing and sophistication, and translated to effective design and analysis procedures. Because the relationships among pipe material, water chemistry, and soil chemistry are compli- cated, agencies often find it prudent to specify allowable ranges of factors in terms of both upper and lower bounds, with annotations of critical situations and interactions. As a final point, although each potential damage mechanism can be mitigated by good practice in culvert design, pipe manu- facturing, pipe material selection and specification, construc- tion, and maintenance, culvert performance and service life also benefit from a policy of regular, thorough inspection to identify and remedy problems in their early stages. Synthesis and Other Survey Findings Information on current practice regarding service life was obtained in the study survey for two major components of drainage networks: pipes and box culverts. Responding agencies were first asked to identify how they would deter- mine service-life values. Responses to this question are shown in Figure 78. Among the 30% of reporting agencies

95 that identified at least one method, their selections focused on agency experience, professional judgment, and manuals or guidelines that individual agencies have prepared, which inform the estimation of service life. Within this context, agencies explained their evaluation of the suitability of different culvert materials for new installation design (none addressed the use of service life for rehabilitation and replacement). For example: • Kansas—In road design for nonfreeway, cross-culvert applications on lower volume roads, the service life of CSP is estimated in selected counties based on soil pH and resistivity. The calculation determines the suit- ability of CSP in these locations and applications. In situations unsuitable for CSP, concrete pipe is used. The service life of CSP is thus variable, and the service life of concrete is greater than that of CSP. • Oregon—The Oregon DOT has developed design service-life data for several types of drainage installations (e.g., cross culverts, other locations of culverts, storm drains, subsurface drains, and slotted drains), type of facility (e.g., freeway), and locations (e.g., within travel way, shoulders, and between curbs). Design lives are specified for each realistic combination of these factors; values range from 25 to 75 years. The Oregon DOT also lists the candidate materials that would be suitable for each combination of factors; these are analyzed in the design process (ODOT Hydraulics Manual 2005). Comprehensive service-life data reported by agencies in the study survey are given in Table 24. Examples of the distributions of estimated service lives for pipes and box cul- verts are shown in Figures 79 through 82. The labels on the horizontal axis in these figures give the upper values of each range of service-life data. For example, if these labels are 0, 10, 20, 30 . . . , then the column labeled 10 shows the num- ber of responses for estimated service life of zero to 10 years; the column labeled 20, the number of responses for esti- mated service life of more than 10 to 20 years; the column labeled 30, the number of responses for estimated service life of more than 20 to 30 years; and so forth. It should be noted again that the data in Table 24 and Figures 79 through 82 may be derived in part from the professional judgment of agency personnel. These results may be compared with results of another survey of U.S. and Canadian provincial transportation de- partments that was conducted as part of an investigation of life-cycle cost analysis (LCCA) techniques applicable to culverts. The LCCA survey yielded a range of service-life assumptions across 25 responding agencies for the following pipe materials (Perrin and Jhaveri 2004): • Reinforced and nonreinforced concrete pipe: 50 to more than 100 years. • CMP: 35 to 50 years. • High-density polyethylene (HDPE): 30 to 100 years. • Polyvinyl chloride (PVC): 50 years. • Vitrified clay: no responses. Agency Practices Agency practices regarding culvert service-life data and assumptions are described in more detail in the literature. Data in the Oregon DOT culvert manual regarding values of design life to be used were discussed earlier. NCHRP Syn- thesis 303 provides data from several other agencies, noting that most that use service-life base their estimates on soil and water pH and soil resistivity (a measure of the relative quan- tity of soluble salts, which influences corrosion resistance) No Response Do Not Use Service Life Other Manufacturer’s Data Professional Judgment Literature Agency Experience LCC Analyses Model Develop, MIS 0 20 40 60 80 100 Percentage of Responses FIGURE 78 Sources for determining service lives of culverts. MIS = management information systems; LCC = life-cycle cost.

96 Component and Material No. of Responses Minimum (Years) Maximum (Years) Mean (Years) Median (Years) Mode (Years) Pipes Concrete 13 30 100 60.4 50 50 Corrugated metal 16 10 60 37.3 35 50 50 50 Asphalt coated corrugated metal 5 10 75 43 50 Small diameter plastic 7 10 75 50 50 High-density polyethylene 1 — — 50 — — Box Culverts Reinforced concrete 15 30 100 63.3 50 50 Timber 3 10 50 30 30 — Precast reinforced concrete 1 — — 50 — — Polyvinyl chloride 1 — — 30 — — Aluminum alloy 1 — — 50 — — Notes: —, value is undefined for the particular distribution. When distribution is based on only one data point, its value is shown in the Mean column. TABLE 24 ESTIMATED SERVICE LIVES OF DRAINAGE CULVERTS 8 7 6 N o. o f R es po ns es 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Estimated Service Life, Years FIGURE 79 Estimated service life of concrete drainage pipe.

87 6 N o. o f R es po ns es 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Estimated Service Life, Years FIGURE 82 Estimated service life of reinforced concrete box culverts. 8 7 6 N o. o f R es po ns es 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Estimated Service Life, Years FIGURE 81 Estimated service life of small diameter plastic drainage pipe. 97 8 7 6 N o. o f R es po ns es 5 4 3 2 1 0 0 10 20 30 40 50 60 70 80 90 100 Estimated Service Life, Years FIGURE 80 Estimated service life of corrugated metal drainage pipe.

to determine the recommended types of pipe material, coat- ings, and other installation details. Agencies differ, however, in the manner and degree to which they simplify the results for practical use in design. Following are examples of different ways to incorporate service life in design recom- mendations, as drawn from the literature: • For each combination of type of pipe material and level of corrosion resistance, Wyoming indicates whether the material is suitable for use in that corrosion environ- ment, using a Yes–No convention. Service-life data are not explicitly shown (Wyant 2002, Table 10, p. 23). • For each type of highway facility (Interstate, arterial, collector, etc.) and pipe function (i.e., drainage appli- cation or installation), Arkansas lists the recommended type(s) of pipe material. Service-life data are not explicitly shown (Wyant 2002, Table 11, p. 24). • Mississippi identifies service-life criteria in years for each drainage application. In each case, allowable mate- rials alternates are listed, together with specific technical requirements (MDOT Pipe Culvert . . . 2005). • Louisiana lists design service life for each combination of drainage application and type of pipe material and pipe joint. Values range from 30 to 70 years (Wyant 2002, Table 12, p. 25). • Montana’s service-life guidelines explicitly consider pH value and resistivity. For each combination of pH and resistivity, usage options are listed for four types of culvert materials: steel, aluminized steel, aluminum, or concrete. The options are stated simply: OK, No, or reference to a note specifying additional technical information; for example, whether to use a coating on steel or aluminized steel pipe, and type of cement needed on concrete pipe (Wyant 2002, Table 13, p. 26). • New York State’s approach is to divide the state into two zones based on annual metal loss rates for steel pipe, and to estimate the anticipated service life for steel pipes as a function of pipe wall thickness (gauge), type of steel pipe (e.g., galvanized–metallic coated and gal- vanized with polymer coating), and the two geographic zones (Wyant 2002, Tables 14 and 15, p. 26). • Utah DOT’s material selection for metal and for con- crete pipe employs sets of parametric curves that relate pH, resistivity, and percent soluble salts to expected ser- vice life for different classes of metal and concrete pipe. Metal pipe classes are based on corrugated or structural plate pipe and different metallic and nonmetallic coat- ings. Concrete pipe classes are based on the use of Type II or Type V cement, depending on measured sulfate levels (Wyant 2002, Figures F6 and F7, pp. 73–74). • California provides a set of curves that are functions of pH and resistivity, and that indicate the minimum thick- ness of metal pipe needed for a 50-year, maintenance- free service life (Wyant 2002, Figure F2, p. 70). These curves have been developed based on a testing procedure developed by Caltrans to estimate steel culvert service life. This service life is based on the estimated time to the 98 first perforation of the metal pipe resulting from corro- sion (Method for Estimating . . . 1999). Although some other agencies use the Caltrans method for their own determinations of service life, experience in other states has shown that in some regions of the country the method is too conservative, whereas in others it is too liberal (e.g., because of the prevalence of soft water). These results demonstrate the importance of using local information whenever possible; nonetheless, the Cal- trans method is still judged to be the most reasonable for general use (Corrugated Steel Pipe Handbook 2005). • Many studies of culvert performance and that of other in-ground structures have also been done in Canada, including the following as reported in the Corrugated Steel Pipe Handbook 2005): – A study of CSP in Southwestern Ontario conducted by Golder in 1967, which showed that the California method did correctly predict service life in that local area. – A 1993 study of remaining coating thickness on 21 steel plate and galvanized bin-type retaining walls, all more than 20 years old (with the oldest, 60 years old), con- ducted by the British Columbia Ministry of Transporta- tion and Highways. The investigations also included tests of soil and water pH and resistivity. Results indi- cated an expected service life of more than 100 years on all but two structures, both of which had already exhib- ited abrasion significant enough to reduce expected life. – A study of zinc coating loss on 201 CSP installations by Alberta in 1988, which also investigated soil and water pH, resistivity, and electrical potential between the pipe and the soil. The study concluded that a mini- mum service life of 50 years would be attained in more than 80% of the installations, and that the average service life across these pipes would exceed 80 years. • The service life of a culvert pipe may be extended by increasing the thickness of metal pipe walls, paving the metal pipe invert, applying supplementary coatings to metal or concrete pipe, or specifying a reaction-resistant type of cement and aggregate in concrete pipe and cast- in-place structures. Guidelines on treatment options and the expected “add-on” service life are included, for ex- ample, in NCHRP Synthesis 303 (Wyant 2002, Table 16, p. 5; Figure F3, p. 70), the CSP Durability Guide (2000, p. 5), and the Corrugated Steel Pipe Handbook 2005, Table 8.3, p. 353). Durability Studies A number of agencies have conducted research on the service life of culverts and what factors are critical to culvert performance. Through research and analytic models, they have sought to understand the complicated relationships underlying culvert durability on an objective, field-verified basis. Following are examples drawn from Missouri, Montana, Ohio, and California, among other agencies.

99 Missouri DOT The Missouri DOT has been studying the durability and per- formance of galvanized CSP and reinforced concrete pipe since the 1930s. Its findings have suggested a service life of 50 years for CSP and almost 100 years for reinforced concrete pipe. However, an attempt to model service life as a function of vari- ables such as pH, abrasion, soil resistivity, chemical properties of runoff, and watershed characteristics was unsuccessful. “No single parameter or combination of parameters accurately predicted service life in all areas of the state” (Cahoon et al. 2002, p. 199). Missouri has also used HDPE since 1983, and the material is still under evaluation. Recently the department installed large-diameter (60-in.) HDPE pipes for evaluation, particularly to monitor pipe wall deflections and joint separa- tion for this flexible material (Blackwell and Yin 2002). Montana DOT The Montana DOT (MDT) specifies a culvert design life of 75 years. Using its service-life estimation procedure, which is based on corrosion, MDT can assess materials options for pro- posed culverts and evaluate current and anticipated perfor- mance of existing culverts. MDT’s method is based on Amer- ican Iron and Steel Institute (AISI) formulas, which calculate service life as a function of resistivity alone for soil with pH of at least 7.3, and of resistivity and pH for soil pH values of less than 7.3. Using these equations with values of pH and re- sistivity determined from soil sample collection and labora- tory testing, it analyzes service life for various pipe materials and recommended coatings and treatments. For example, CSP is considered with a galvanized, aluminized, bituminous, or polymeric coating. If none of these options can meet the 75- year criterion, then other pipe materials are recommended; for example, aluminum or concrete. Although this method has provided a way for MDT to analyze the effects of corrosion, it does not address other damage mechanisms (Hepfner 2001; Cahoon et al. 2002, p. 199; Wyant 2002). In a separate study of corrosion of CSP, a consultant re- viewed MDT’s soil sampling, testing, and analysis methods to recommend improved design practice (Hepfner 2001). The study was motivated by a number of premature CSP failures in Montana. In the overview of existing agency prac- tice, it was noted that there existed a wide variation in mate- rials selection criteria, and a lack of standardized procedures for identifying potentially corrosive environments and eval- uating suitable pipe materials to meet design criteria. More specifically, the complexity in pipe–soil interactions that can lead to corrosion, and the variety of sample preparation and testing methods for resistivity were described, an important determination for corrosion analyses. In applying different testing methods to MDT soil samples, it was demonstrated that the differences are not only in terms of resistivity values themselves, but also in the recommended pipe material that would result from the use of the respective resistivity values. The study recommended that MDT adopt an AASHTO procedure for resistivity testing, and that soil sampling in- clude the materials to be used around the pipe and under the conditions (e.g., moisture, chemical transfer, and bacterial growth) that are expected to pertain at the culvert location in the field. In the longer term, the study recommended devel- opment of a database of corrosion-related data, including soil resistivity, pH, chloride and sulfate concentrations, soil type, location, and sampling depth, and mapping of these data to a statewide soil survey map. The study also recom- mended use of “early warning system” field monitoring to warn of impending corrosion so that remedial measures such as cathodic protection can be installed before the need for more expensive culvert replacement. These detection devices would be installed at locations of questionable soil characteristics or where historical performance of galva- nized steel pipe fails to measure up to analytical predictions of service life. These field data could also be used to refine pipe corrosion models (Hepfner 2001). MDT has also sponsored university research on a condi- tion index for rural culverts. The index is defined on a scale of 1 to 5, where 5 denotes excellent condition, and 1 poor condition. The method employs a spreadsheet to compute a pipe’s condition index based on a number of key inspection variables as discussed earlier, including age, degree of scour at the outlet, evidence of past major hydraulic or structural failure, degree of corrosion, extent of invert wear, physical blockage, sedimentation, joint separation, and physical dam- age. Average daily traffic and detour length for culvert repair or replacement are also used as weights on the condition result to reflect road-user impacts. These variables are rated according to qualitative assessments (e.g., no damage, minor damage, or major damage), ranges of condition (e.g., five ranges of numbers of vehicles crossing the culvert—0–500, 501–2,000, and so forth), percentage values, and other simi- larly general assessment measures (Cahoon et al. 2002; Baker et al. 2006a,b). Ohio DOT Ohio has been analyzing data on culvert performance for more than two decades. Its 1982 report on culvert pipe durability represented a 10-year study of 1,600 culverts from around the state, some of which had been installed before 1940 (Meacham et al. 1982). Data collected under this study contributed to the development of models to pre- dict pipe service life for different pipe materials and char- acteristics, as well as soil and water properties (Hurd 1986a,b, 1988; Precast Concrete Pipe Durability 1991). For example, the models for concrete pipe included pipe age, pipe vertical diameter, invert slope, water pH, the depth of sediment, and flow velocity as independent vari- ables (Gabriel and Moran 1998). The models for metal pipe included pipe age, pipe wall thickness, water pH, and abra- sion as independent variables (Mitchell et al. 2005).

The culvert rating approach used by the Ohio DOT since 1982 was updated in 2003, and followed by the development of a culvert risk assessment methodology (Mitchell et al. 2005). The analyses contributing to this methodology included linear as well as nonlinear regression of field per- formance data that were collected under the new inspection procedure. This new procedure was based on a 0–9 scale to rate pipe conditions (with 9 denoting excellent condition and 0 indicating failure), whereas the earlier procedure before 2003 had employed a 4-point scale (good, fair, poor, critical, with 1 denoting good, and 4 indicating critical). The risk as- sessment was calculated by taking the original average culvert rating, based on inspection results, and adjusting it for factors that can indicate a reduction in remaining expected life: current age, water pH, abrasiveness, and the pipe cover height-to-vertical diameter ratio (reflecting a greater risk to motorists if this ratio is smaller). In addition to these methodological investigations, the study also recommended an expanded inspection protocol (revising the 2003 procedures) that would consider 30 to 33 items rather than the 16 items in the current Ohio DOT inspection man- ual. These recommendations were developed for concrete, metal, and plastic pipe used by the Ohio DOT. The culvert inspections conducted in this study indicated that concrete culverts have a service life of 70–80 years, and metal culverts 60–65 years (Mitchell et al. 2005). California DOT (Caltrans) In California Test Method 643, service life is based on the es- timated time to the first perforation of the metal pipe as a result of corrosion (Method for Estimating . . . 1999; Wyant 2002). This test method incorporates a set of curves that indicate the minimum thickness of metal pipe needed for a 50-year, main- tenance-free service life, and that are functions of pH and resistivity. The testing procedure and curves were developed based on field studies of the performance of 7,000 culverts that began in the 1950s and were subsequently updated to yield today’s standardized procedure (Beaton and Stratfull 1962; Ault and Ellor 2000). Although other agencies use the Caltrans method or variations of this method for their own determina- tions of service life, experience in other states has shown that in some regions of the country the method is too conservative, whereas in others it is too liberal. These results demonstrate the importance of using local information whenever possible; nonetheless, the Caltrans method is still judged to be the most reasonable for general use (Corrugated Steel Pipe Handbook 2005). The California method “is now the most widely ac- cepted method to determine culvert durability” (Ault and Ellor 2000, p. 51). The FHWA study also found that the California method works satisfactorily in several locations nationwide, but can over-predict or under-predict service life in others. This is perhaps not surprising, because the California method was based on statistically average values of variables among the randomly sampled culverts in the field studies and does not predict durability well in extreme conditions. 100 Other Examples The FHWA study identified other predictive models that have been developed by Florida and AISI, both of which are similar to the California method, but introduce specific vari- ations. FDOT has developed models to directly predict the service life of concrete pipe, aluminized Type 2 corrugated steel, and aluminum alloy culverts (Ault and Ellor 2000). FDOT has also conducted supplementary studies on the effects of seawater on the durability of reinforced concrete culvert pipe (Sagüés et al. 2001). The AISI method uses a chart similar to California’s, but applies a different criterion for when a pipe is judged to reach its service life. The AISI predictions are thus double those of the California method. Other organizations [New York, Colorado, and the National Corrugated Steel Pipe Association (NCSPA)] have devel- oped procedures based on a service-life concept; however, these are structured to aid in selecting the most appropriate materials rather than to predict a value of culvert life itself (Ault and Ellor 2000). The North Carolina DOT also attempted to derive service-life prediction models for four types of pipe in three geographical regions of the state, but was unsuccessful. Its recommendations for further work included the development of databases on site-specific infor- mation regarding soil and water chemistry and physical drainage characteristics (Gabriel and Moran 1998). NCHRP Synthesis Reports 254 and 303 and the 1996 FHWA study describe these several approaches to analyzing service life. The FHWA study also describes software that agencies have developed to perform related computations. Durability studies have been conducted for other pipe materials. The use of HDPE by Missouri was mentioned ear- lier. Although this material has been in use in the United States for more than 35 years (Design Service Life . . . 2003) and is applied in transportation facility drainage in more than 40 states (Reddy 1999), agencies are continuing to evaluate its performance under field conditions and for larger pipe diameters. Although the durability of a high-quality plastic material itself can yield potentially long service lives (exceeding 100 years for corrugated HDPE pipe—Design Service Life . . . 2003; Gabriel 2005), there are several concerns regarding performance under field conditions, including the following (Reddy 1999): • Pipe-wall deflection, joint separation, and potential buckling of the pipe owing to improper installation and backfilling, or vehicle live loads. • Stress cracking, as a result of improper installation and backfilling, which can lead to catastrophic failure. • Creep of the plastic material and creep rupture. The study acknowledges the complexity of the problem and the need for additional laboratory, analytic (computer simulation), and field investigations (Reddy 1999). The Wis- consin DOT has also installed large-diameter (48-in.) HDPE pipe, and if performance results continue to be favorable, will

101 expand their application (Wilson 2000). The importance of proper installation, including preparation of the bedding soil, use of appropriate backfill material and procedures, and pro- viding sufficient cover, have been emphasized in an evalua- tion of 45 HDPE highway drainage pipes in South Carolina (Gassman et al. 2002). An investigation of the in-service performance of HDPE pipes in six states has been conducted on behalf of the American Concrete Pipe Association (ACPA), documenting pipe-wall deflection; distresses such as buckling, bulging, and cracking; joint separation; and mis- alignment (Nelson and Krauss 2002). Performance studies have also been conducted for the following materials: • PolyRib, a small-diameter pipe manufactured from polymer-coated galvanized steel, which showed favor- able results (Brockenbrough 2002). • Aluminized steel pipe, in studies conducted by the manufacturer (Morris and Bednar 1982) and by a research consultant for the FHWA (Ault and Ellor 2000). Criteria for Determining Service Life NCHRP Synthesis 254 discusses the concept of culvert service life as applied to each type of major pipe material: concrete pipe, steel pipe (including CSP and spiral-rib steel pipe, metal- lic-coated, nonmetallic coated, lined, and paved), Alclad aluminum pipe (Alclad is an alloy that is bonded to the alu- minum alloy core that provides cathodic protection to the aluminum pipe) and aluminum structural plate, plastic pipe (including HDPE, PVC, and acrylonitrile-butadiene-styrene), ductile iron, and clay. Because different damage mechanisms affect these materials, the selection of particular mechanisms as well as the threshold values that define service-life criteria can vary among transportation agencies. The following are examples regarding reinforced concrete pipe (Gabriel and Moran 1998): • California uses the debonding of reinforcing bars as its measure of service life. • Colorado bases service life on functionality, relying on a committee of professionals to determine whether the pipe still meets its intended purpose. • Missouri defines service life as the time until pipe re- placement. • North Carolina defines service life as the age beyond which 80% of pipes may be expected to experience functional failure. • Mississippi assumes that concrete pipe will last the life of the highway facility. Analogous examples of the diversity among agency service-life criteria are given in NCHRP Synthesis 254 for other pipe materials. Variability in assessing culvert durability is subject as well to natural causes, as illustrated by the following statements: Durability is not defined as clearly as structural and hydraulic standards for drainage pipe systems, because it includes the performance of the components of concrete and reinforced con- crete structures. Durability deals with life expectancy and the endurance characteristics of a material or structure. Among other considerations, the varying nature of climate, weathering, soils and geology, fluid chemistry, product installation tech- niques, in-plant production, material mixes, and raw material quality cloud the development of a way to define durability and predict performance (Why Concrete Pipe? 2006). . . . the corrosivity of an environment is based on multiple, in- dependent (and interdependent) variables and their interaction. No single parameter dominates the corrosion process, and therefore a combination of individual indicators is needed to accurately eval- uate the corrosive potential of a particular environment. The inherent complexity of soil corrosion [of corrugated metal culvert pipes] creates great difficulties in estimating a rea- sonably valid service life... for a given site. No single corrosion contributing factor can be utilized to assess corrosion potential of a metal pipe/soil system. At the present, corrosion assessment is typically based on experience; no singular, standardized methodology is used in highway departments or private consult- ing firms (Hepfner 2001). Assessing Remaining Service Life for an Existing Culvert To apply the service-life concept in asset management, a method is needed to determine where an asset is in its service life—that is, how much life is consumed and how much remains. Agencies were presented with a number of ways to determine the current status of an asset regarding its service life, and asked to rank each method by relevance to their agency. The result is shown in Table 25. Under “Other factors,” agencies reported the following: • Kansas—Culverts of 10–20 ft in size or span are inspected every other year, and smaller pipes are inspected occasionally, to identify potential problems. Although an “estimated service life” is not used to proj- ect impending failures, an effort is made to identify the end of the actual, practical service life through regular inspection. • Pennsylvania—The culvert condition is analyzed using a matrix to determine the treatment and cost. • Texas—Culvert condition is analyzed when the road is expanded or rehabilitated. Although several states have long-standing and continu- ing programs to assess culvert condition and infer remaining service life, most agencies do not apply service life in their routine culvert management. Seventy percent of the respon- dents to the question addressed in Figure 78 either reported that they do not use service life or left this item blank. This result echoes findings of an earlier survey reported in NCHRP

Synthesis 303. When asked whether they predicted service life as part of their decision process for selecting culvert re- medial treatments, only 13 of 55 agencies (24%) responded affirmatively. Most of these reported using service-life data provided by manufacturers. When it is used, service life gen- erally informs decisions on relining or replacing metal cul- verts, although some agencies (e.g., Utah) may also apply service life to concrete pipe. NCHRP Synthesis 303 reported that service life is to date rarely used with plastic pipe. The two local agencies that responded affirmatively in the NCHRP Synthesis 303 survey to use of service life noted that they apply their respective state DOT’s data and procedures (Wyant 2002). NCHRP Synthesis 303 reported that only 5 of 56 respon- dents (9%) had a standard set of guidelines to select the most appropriate culvert repair method, and 4 of 56 respondents (7%) standard guidelines to select the most ap- propriate rehabilitation method. However, a somewhat larger number, 15 agencies (27%), noted that they do consider several factors in making decisions on pipe reha- bilitation, including hydraulic capacity, traffic volume, height of fill, service life (12 of these 15 agencies), and risk assessment (Wyant 2002). The ORITE survey addressed factors affecting state DOT decisions on culvert replacement, with the following rates of affirmative response (numbers do not sum to 100% because agencies could select more than one response): • Degree of culvert material degradation: 80%. • Roadway surface conditions over the culvert: 50%. • Deflections in the culvert: 38%. • The sum of numerical rating scores: 23%. • Culvert age: 8%. • Other factors including joint conditions, fish passage is- sues, roadway expansion/rehabilitation/replacement, failure or imminent failure of the culvert, inadequate flow capacity, replacement criteria used for bridge class structures, and video inspection results: 33%. 102 Only in a relatively small number of instances were any of these criteria identified as the sole basis for decisions on cul- vert replacement. In most cases, agencies reported multiple criteria driving culvert replacement (Mitchell et al. 2005). On the related issue of identifying the extension in service life owing to maintenance, agencies provided some examples of current practice in their responses to the synthesis survey: • Ohio field paves CMP inverts before complete failure and assumes 75 years of additional life for the structure. • New Brunswick expects 50 or more years of additional life after rehabilitating a culvert using a concrete invert or aluminum alloy and grout slip-lining. • Saskatchewan has used cathodic protection to prolong the life of corrugated steel pipes. Impacts of Culvert Performance Although the value of proper culvert performance to the public appears to be well understood in concept (Table 23), there is relatively little guidance on how to demonstrate these benefits analytically for highway assets, let alone to mount compelling arguments publicly for the benefits of stronger culvert management. Viewed in another way, there are rela- tively few examples in the literature of methods to analyze and communicate the consequences of culvert failure, even though it is well understood that potential impacts to the road surface, highway embankment, and resulting mobility of road users can be severe. Forensic studies of pipe failures at spe- cific locations appear in the literature (e.g., Freeman 2003, ad- dressing the problem of backfill in the culvert trench). Several cases of culvert failure were reviewed in a study of LCCA to culverts, identifying the significant component of total costs that is attributable to road-user delays (Perrin and Jhaveri 2004). The concept of incremental road-user cost owing to closures resulting from culvert failures is recognized implic- itly by factors such as traffic volume and detour length that are included in the rating systems discussed earlier. Caltrans’ inspection rating guide includes a photograph of pavement Rank Factor 1 Assets are repaired or replaced as soon as they fail without regard to service life 2 The agency does not use/does not monitor service life for this type of asset 3 Monitor condition of the asset occasionally 4 Monitor condition of the asset on a periodic schedule 5 Other factors 6 Compare current age of asset with the maximum age that defines service life TABLE 25 RANKING OF METHODS TO DETERMINE WHERE CULVERTS ARE IN THEIR SERVICE LIVES.

103 and foundation failure resulting from a failed culvert (Caltrans Supplement . . . 2003). From an environmental perspective, the benefits of an effective culvert and drainage management system for hydrologic analysis and better storm water man- agement have also been recognized (Venner 2005). A general approach to deal with impacts of failure has been recommended based on LCCA, including culvert replacement costs and road-user delay costs. However, this proposal is acknowledged to be but a beginning, and the state of practice generally leaves much room for advancement. A survey conducted as part of the LCCA study indicated that only 4 of 25 respondents reported that they employ some type of least-cost procedure for culvert material selection. Fifteen of the 25 agencies in the LCCA survey reported that they doc- ument culvert failures; however, the level of detail varies greatly from one agency to another. The assumed service life of different pipe materials, as well as the unit costs of these materials used in the LCCA procedures, likewise varied among agencies (Perrin and Jhaveri 2004). The American Concrete Pipe Association describes LCCA methods devel- oped by the U.S. Army Corps of Engineers and ASTM to an- alyze the most cost-effective, long-term investment options for alternate pipe materials, structures, and systems (Design Data 25 . . . 2002). Although these methods consider several engineering and economic factors for agency first cost, remedial costs while in service, and residual costs, there is no consideration in either method of the benefits or other impacts of culvert performance to road users and the public. INFORMATION TECHNOLOGY SUPPORT Agencies participating in the study survey identified their key IT capabilities for culverts as shown in Figure 83. Physical measures, age, inspection data, GIS-generated maps, and 0 10 20 30 40 50 60 70 80 90 100 No Response None of the Above Other Historical Database PMs, Dashboards, Accountability GIS Maps, Reports GIS Interface Est. Asset Impacts on Public Track Public Comments Cost Models for Treatments Other Optimization Procedures Benefit-Cost, LCC Decision Rules or Trees Inspector Recommendations Established Mntce. Schedule Deterioration Models Anticipated Service Life Dates of Inspections, Assess. Asset Age Usage, Traffic Volume Photograph Current Condition, Performance GPS Coordinates Location (e.g., Rte-Milepost) Number/Quantity of Asset Percentage of Responses FIGURE 83 IT capabilities to help manage culverts. GPS = global positioning system; LCC = life-cycle cost; GIS = geographic information system; PMs = performance measures.

performance/accountability reports such as dashboards were the most prevalent items selected. Agencies characterized their IT systems for culverts as shown in Figure 84. The great- est number of responses pertained to broad-based manage- ment systems (such as maintenance management systems). The agencies that reported using a culvert management system or a maintenance management or transportation infrastructure asset management system that includes culverts are listed here. • Culvert Management System – New York State DOT – Ohio DOT – Oregon DOT – New Brunswick DOT. • Maintenance or Asset Management System That In- cludes Culverts – Arkansas DOT – Florida DOT – Iowa DOT – Maryland SHA – Minnesota DOT – North Carolina DOT – Pennsylvania DOT – Texas DOT – Utah DOT – Vermont AOT – Virginia DOT – Colorado DOT Region 4 – New Brunswick DOT – Saskatchewan Highways and Transportation – Dakota County, Nebraska – City of Jacksonville, Florida – City of Portland, Oregon. “Other” options mentioned by agencies included the following: • Ohio is in the initial start-up phase of establishing a cul- vert inventory system. 104 • Texas accesses inventory information from its inven- tory database and maintenance expenditures from its maintenance management system. The earlier survey summarized in NCHRP Synthesis 303 reported that 11 of 57 state, federal, and local respondents, or 19%, had a management system that made use of pipe assessment data gathered by the agency (Wyant 2002). Sev- eral states are reported in the literature as having established culvert pipe management systems, including California, Con- necticut, Maine, Minnesota, New York, North Carolina, and Pennsylvania (Beaver and McGrath 2005), and the Maryland SHA (Venner 2005). Initial work to develop such systems is underway in Utah (Beaver and McGrath 2005) and New Jersey. The work for the New Jersey DOT is intended to produce a Culvert Information Management System, a com- ponent of the New Jersey DOT’s Transportation Asset Man- agement System. The project has therefore developed an ini- tial set of survival probability curves for CSP in urban areas based on data from an ASTM study, although the authors em- phasize that a more definitive set of curves will need to be es- timated from actual historical data or accelerated test results in the future. Preliminary technical and cost data and decision rules have also been proposed, defining a framework for fur- ther Culvert Information Management System development (Meegoda et al. 2005). The ORITE survey, which asked about computerized databases rather than management systems, indicated that 23 of 40 DOTs (58% of those responding) had such capabilities. These databases include culvert-specific storehouses (e.g., developed in Microsoft Access® in California and Vermont), and the Pontis® database for culverts larger than 10-ft span, with smaller culverts addressed in a separate database (Minnesota). Washington State reported that it is beginning to use GPS to track the location of culverts (Mitchell et al. 2005). MDOT is also proposing to include culverts in a GPS-based database of roadside assets and features (see chapter eight). Examples of software used to implement particular design procedures or service-life cal- culations have been compiled in an FHWA study (Ault and Simple Program(s) for this Asset Broad-Based MMS, TIAMS, etc. Culvert Management System Percentage of Responses Workbook, Spreadsheet Other Products or Procedures 0 20 40 60 80 100 FIGURE 84 Types of analytic tools to support culvert management. MMS = maintenance management system; TIAMS = transportation infrastructure asset management system.

105 Ellor 2000). Louisiana has developed an expert system to analyze the costs of installing cathodic protection on its metal culverts (Garber and Smith 1999). GIS are just beginning to be used to organize, display, and analyze hydraulic data for culvert design. In a survey of state DOTs conducted as part of a GIS drainage-application development for the Texas DOT, only 10 of 32 respondents (31%) reported using GIS, mainly for mapping and data man- agement. Of these, only Maryland was applying its GIS for hydraulic analysis (Olivera and Maidment 1998). KNOWLEDGE GAPS AND RESEARCH NEEDS Over the long term, industry sees a strong role for technology to improve culvert products and installation procedures. The TRB Millennium paper addressing culverts and drainage structures provided a vision of potential advances that can be anticipated in the future (Hill 2000): • Improved materials for concrete pipe, including im- proved concrete mixes, greater use of polymeric and epoxy coatings, and synthetic fiber reinforcement of the concrete and polymeric coatings, all of which will increase durability and strength. • Substantial changes in metal pipes, including greater structural economy through the use of higher-strength, lighter-gauge materials; improved metallic, nonmetal- lic, and organic coatings; increased abrasion resistance; greater use of automation and tighter manufacturing tol- erances to aid jointing and installation, including onsite manufacturing and assembly in some installations; and prevalence of hydraulically smooth pipe profiles. • Improved culvert installation, involving methods such as trenchless technology, more automated control of backfilling, tighter joints, or jointless pipe. • As a result of the above advances, improved durability such that a 100-year design life will become a mini- mum requirement. • Greater use of computer-assisted design of culvert sys- tems, including use of satellite imaging and GPS data for pipe location and sizing, to better relate drainage de- sign to water resource management and potential envi- ronmental impacts. Within the shorter term, knowledge gaps and research needs tend to focus more on management needs than tech- nology. Agencies at all governmental levels that responded to the synthesis survey provided a number of comments in this regard for culverts. • Basic management data and tools—There is a need for better condition and performance data and analytic tools to support culvert and pipe management. Many agencies cited data needs that include a complete inventory in terms of number, size, and date of installation (therefore age); condition of culverts; records of maintenance and resources used; evaluations of both the structural and the hydraulic performance of culverts; models of deteriora- tion or service life and life-cycle cost; and institution of a periodic inspection program. Several agencies mentioned an important need for periodic inspection and a compre- hensive and consistent assessment of this infrastructure, but recognized funding and personnel limitations to do this. Analytic tools are also needed to interpret the data, help prioritize repair or replacement, and assist in man- aging the asset. • Stormwater infrastructure performance—The initial selection of culvert material can be complicated by local factors, and evaluation of performance needs to recognize these and other complexities of in-service in- frastructure. The life of steel culverts is affected by both soil type and the content of runoff materials. Knowledge of these effects is specific to certain sites and is acquired only after the fact. Research on life ex- pectancy needs to account for these various conditions. This point is reinforced by New Brunswick, noting that durability must be related to water chemistry and pipe exposure thereto. Ohio notes that simple equations (or procedures) of service-life expectancy that require very little field data are most desirable. Determination of both the pH and abrasiveness of runoff carried in the culvert are difficult to determine for most designers, and soil resistivities are even more difficult. Texas noted that it has had few failures of pipe culverts or box culverts, and most of these are the result of corroded galvanized metal pipe. New Mexico likewise does not have many failing culverts; they are generally replaced in road reconstruction before they wear out. Again, the lack of a periodic culvert inspection program is a major gap in knowledge. • Applying performance to management—Because culvert performance varies across the country, agencies saw an issue in how to apply performance information to man- agement and communication. Oregon voiced a need to develop standardized methods for determining the remaining culvert service life of different materials. New York indicated a need to define a state of good repair both for individual culverts and for the inventory as a whole. This would enable the agency to define performance goals and targets, and to analyze the impacts of different maintenance strategies on system performance in terms of to what degree these targets have been met. Ohio noted that managers may not be fully aware of how a missing invert or poor pipe joints can collapse a road. Tampa ob- served that the lack of a reliable funding source inhibits attainment of the level of service that would accomplish a standard measure of performance. • Broader culvert performance management—Several agencies commented on the broader implications of culvert management to the hydrology of a region. Michigan noted that there is no evaluation of the impact of changing land use on the drainage area to the culvert.

Furthermore, there is inadequate or no evaluation of the impacts of extending existing culverts during road re- construction, in terms of changes to the structural and hydraulic characteristics of a culvert. Dakota County, Nebraska, commented that there should be more in- volvement in culvert management with the Natural Re- sources Conservation Service (a unit of the U.S. De- partment of Agriculture). • Agency evaluation of materials and performance— Several agencies identified the need for research in how agencies evaluate culvert materials and perfor- mance. Maryland noted that the determination of ser- vice life for different culvert materials should be based on field performance. Most research today, by con- trast, is conducted in controlled environments. Perfor- mance based on actual site conditions, accounting for the variability in water and sediment among locations, is not widely known. Vermont identified a need for re- search on new methods and materials to repair and re- habilitate culverts in place to extend service life. Saskatchewan proposed research involving field as- sessments to validate lifespan predictions by the man- ufacturing industry. This more immediate focus on management needs is echoed in a recent research problem statement for culverts. This research topic calls for the development of recom- mended rehabilitation techniques for concrete, steel, and plastic pipe. The research would include the establishment of critical design criteria for each of several possible failure mechanisms in each pipe material, development of test meth- ods to determine the performance of remedial treatments as well as environmental effects on pipe materials that affect durability, and conduct of accelerated testing of the remedial 106 treatments in the laboratory and the field (Committee Research Problem Statements 2005). The literature has identified the need for materials-related research as well as improved installation procedures for cul- verts, and a better understanding of the role of drainage cul- verts in the water resource environment. The TRB Millennium Paper on culverts and drainage structures based its projections on a future culvert service-life requirement of at least 100 years and a need for hydraulically smooth walls. New coatings can be expected to lengthen service life and be more abrasion re- sistant. Joints in pipe will be more tightly gasketed, welded, or eliminated in jointless pipe; onsite manufacture or “in-the- ground manufacturing” of a water conduit are also foreseen. New installation procedures will improve the quality and econ- omy of installation. These innovative procedures will include gains in trenchless technology (directional boring, tunneling, and jacking), as well as in open-trench installation, with greater use of automation and improved control of backfill. Environmental considerations will recognize water as a pre- cious resource, and seek to minimize effects on stream flow (including fish passage), control flow rates, and focus more on recharging groundwater resources where technically possible rather than releasing water into nonusable bodies (e.g., salt water or polluted water) (Hill 2000). Ongoing research is also identifying better techniques for rehabilitating pipes. A research study investigated materials for cost-effective, non-flammable pipe liners to rehabilitate corroded metal pipes. After considering a number of differ- ent coatings and pipe liners, the researchers concluded that the best solution was to reline the metal pipe with HDPE pipe and use concrete end caps to resist grass fires (“Cost Effective . . .” 2005).