Guide Specification for Service Life Design of Highway Bridges (2020)

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.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 5 C H A P T E R 2 Literature Review and Survey Literature Review Synthesis Rutgers University led the effort of gathering references relevant to the project objectives. The references gathered included code specifications (including both local and federal domestic as well as international), technical literature, and research publications related to avoidance of deterioration, deemed-to-satisfy rules, and probabilistic based design. Over 100 references were reviewed. The approach taken to review the literature began with developing summaries of each identified source. These summaries, at a minimum, included (a) the purpose and/or objectives of the study/document, (b) a description of the methodology employed to achieve the stated objectives, and (c) a summary of the relevant conclusions. For the development of the literature synthesis, the research team decided to prioritize the information gathered from codes/standards and technical reports from professional societies or government entities, and to conduct a synthesis analysis aimed at summarizing and integrating these various sources to identify gaps and shortcomings. State DOT sponsored research was reviewed, as indicated above, as part of the literature review. Summaries of some of the reviewed reports are included in Appendix A of this revised report. However, given the very large number of such studies, and the focus of this research project on the development of specification provisions for service life design, the RT determined that the best way to ensure adequate coverage was to focus the synthesis of the literature review on research results that have influenced design provisions as implemented by the states, as well as other national and international bodies. Last, efforts to reach out to the LTBP Program to obtain access to specific information collected by LTBP staff during initial discussion with the states were unsuccessful. This specific data may no longer be available in any form. The research team instead relied on a questionnaire aimed at gathering tacit knowledge along with owner-specific documented practices and research. The survey results are summarized in this chapter, and the raw responses are provided in Appendix B of this report. Background Over the last three decades there has been increasing attention to the design of bridges to meet specific service lives. In general terms this has meant an expansion of the traditional serviceability limit states to include a focus on durability in addition to excessive live load deflection, stress limitations, foundation settlement, and vibration issues. Although fatigue limit states would technically fall under the design for service life category, it is not the focus of this literature synthesis. Over the last 50 years there has been considerable attention paid to the fatigue limit state and the provisions are both mature and capable of producing bridges without fatigue issues. As a result, unlike the durability-focused performance limit states associated with service life design, there is no evidence that the fatigue limit states of AASHTO LRFD require refinement.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 6 In North American structural design codes, durability-focused limit states are not explicitly identified and are addressed primarily through “deemed-to-satisfy” requirements included in concrete material specifications, crack control provisions (rebar spacing and size limitations), and concrete cover requirements, along with guidance related to drainage, steel coatings, and the design of specific elements for replacement. Over the last few decades, research has been conducted to identify specific durability limit states and to develop rational and quantitative approaches to consider them during the design phase. In general, these efforts have focused on the depassivation of steel in reinforced concrete, as models exist to predict the time-dependent aspects of these deterioration mechanisms. Although corrosion of reinforcement is one of the most common deterioration mechanisms, it is by no means the only one that influences the service life of bridges. Nevertheless, the work on depassivation of concrete provides a compelling example of how service life design provisions will likely evolve, especially as models capable of simulating other deterioration mechanisms become available. The following sections provide a structured summary of service life provisions within current standards as well as recommended service life design approaches provided by various technical committees. Definitions In general, existing codes and standards define two distinct time periods associated with a structure’s expected service life. The first is used to define the design values for variable and time-dependent loads in order to meet target structural reliability indices. Except for AS 5100.5, none of the definitions mention whether or not major repair or rehabilitation may occur during this period. These definitions utilize the words “design” and “reference” to signify the lifetime of expected structural demands after the initial commissioning of a structure. • Design Life — period of time on which the statistical derivation of transient loads is based: 75 years for these Specifications. – AASHTO LRFD (2017) • Design Life — a period of time, specified by an Owner, during which a structure is intended to remain in service. – CSA Canadian Highway Bridge Design Code (2014) (set at 75 years) • Design Life – The period assumed in design for which a structure or structural element is required to perform its intended purpose without replacement or major repairs. – AS 5100.5 (2004) (set at 100 years) • Reference Period – chosen period of time which is used as a basis for assessing values of variable actions, time-dependent material properties, etc. – ISO 16204 (2012b) • Reference Period – chosen period of time that is used as a basis for assessing statistically variable actions, and possibly for accidental actions – EN 1990 (2002), fib Bulletin 34 (2006) The second time period used by many codes and standards aims to define the expected service life from a durability perspective. In contrast to the definitions above, many of these explicitly mention maintenance and repair. They also note that during this time period no major corrective actions will be required. Eurocode and ISO standards provide “indicative” service lives for various types of structures (see Table 1). The durability related provisions for bridges in these standards assume a design working life of 50 years.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 7 Table 1. Indicative design working life (EN 1990 2002, ISO 2012a). Design working life category Indicative design working life (years) Examples 1 10 Temporary structures(1) 2 10 to 25 Replaceable structural parts, e.g. gantry girders, bearings 3 15 to 30 Agricultural and similar structures 4 50 Building structures and other common structures 5 100 Monumental building structures, bridges, and other civil engineering structures Notes: (1) Structures or parts of structures that can be dismantled with a view to being reused should not be considered as temporary. In contrast to a specific time period, the fib Bulletin 34 states that the design service life should be defined based on “(a) the definition of the relevant limit state, (b) a number of years, and (c) a level of reliability for not passing the limit state during this period.” These terms rely on the words “service” and “working” to create a definition that focuses on what constitutes the operating life span of a structure. • Service Life — The period of time that the bridge is expected to be in operation. – AASHTO LRFD (2017) • Service Life — The actual period of time during which a structure performs its design function without unforeseen costs for maintenance and repair. – CSA (2014) • Service Life – A period over which a structure or structural element is intended to perform its function without major maintenance or repair. – AS 5100.5 (2004) • Design Service Life – Assumed period for which a structure or a part of it is to be used for its intended purpose with anticipated maintenance, but without major repair being necessary – ISO 16204 (2012b) • Design Working Life – The assumed period for which a structure or part of it is to be used for its intended purpose with anticipated maintenance but without major repair being necessary - EN 1990 (2002) • Design Service Life – Assumed period for which a structure or part of it is to be used for its intended purpose. – fib Bulletin 34 (2006) • Service life (of building component or material): the period of time after installation (or in the case of concrete, placement) during which all the properties exceed the minimum acceptable values when routinely maintained. – ACI 365.1R (2000) ACI 365.1R goes onto to adopt the following definitions from Somerville (1986), which sub-divides service life based on three perspectives: technical, functional and economic. • Technical service life: the time in service until a defined unacceptable state is reached, such as spalling of concrete, safety level below acceptable, or failure of elements. • Functional service life: the time in service until the structure no longer fulfills the functional requirements or becomes obsolete due to change in functional requirements, such as the needs for increased clearance, higher axle and wheel loads, or road widening. • Economic service life: the time in service until replacement of the structure (or part of it) is economically more advantageous than keeping it in service. The ISO and Eurocode definitions above assume that specific “anticipated maintenance” activities will take place throughout the service life, but assume service life has ended when a “repair” is required. In

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 8 contrast, the definitions offered by AASHTO and fib Bulletin 34 make no mention of maintenance. The definition offered by AS 5100.5 may be interpreted to assume “non-major” maintenance and repair will occur throughout the service life. Given these assumed relationships between maintenance/repair and service life, the definitions associated with repair, maintenance, and rehabilitation included within the standards reviewed are provided below. • Maintenance – A set of activities performed during the working life of the structure in order to enable it to fulfill the requirements for reliability – EN 1990 (2002) • Maintenance – Set of activities that are planned to take place during the service life of the structure in order to fulfill the requirements for reliability. – fib 34 (2006), ISO 16204 (2012b) • Rehabilitation – A process in which the resistance of the bridge is either restored or increased. – AASHTO LRFD (2017) • Rehabilitation – A modification, alteration, or improvement of the condition of a structure or bridge subsystem that is designed to correct deficiencies in order to achieve a particular design life and live load level – CSA (2014) • Repair – The activities performed to preserve or to restore the function of a structure that fall outside the definition of maintenance – EN 1990 (2002), fib 34 (2006), ISO 16204 (2012) The two definitions for maintenance above appear similar, but have an important distinction. In particular, the definition offered by Eurocode states that maintenance is a set of activities that are “performed” during the service life. In contrast, the definition offered by fib Bulletin 34 and ISO define maintenance as a set of activities that are “planned” to take place. Summary of Relevant Research Undertaken as Part of SHRP 2 As part of the Second Strategic Highway Research Program (SHRP 2), which ran from 2006 to 2015, two projects aimed at enhancing the current ability to perform service life design were undertaken. The objectives and key results of these projects, R19A and R19B, are summarized in the following subsections. Specific findings and conclusions from these projects are included throughout the remainder of this literature review in the relevant sections. SHRP 2 R19A Project R19A, titled “Bridges for Service Life Beyond 100 Years: Innovative Systems, Subsystems, and Components,” (Azizinamini et al. 2014a) produced the “Design Guide for Bridges for Service Life”, which aimed to “define procedures to systematically design for service life and durability for both new and existing bridges.” One of the primary goals of this project was to identify the most pressing issues that currently limit the service life of bridges and then focus the development of design provisions to address them. To prioritize these issues, the research team performed an expert elicitation which identified durability as a general challenge to realizing longer service lives of bridges, and the performance of bridge decks, coatings, joints, and bearings as specific challenges. Although this project originally intended to use quantitative performance data to perform this prioritization, the authors concluded that such data is currently not available and thus the expert elicitation approach was employed. This lack of data also hampered the ability of the R19A team in developing new design provisions based on deterioration models.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 9 In response to the identified primary challenges, a framework for the service life design of bridges was developed and presented. The framework starts with bridge system selection and moves in an iterative manner through the design/selection of various components. At each instance of the design, all components are analyzed to estimate their service life. If the component service life does not meet the required service life of the bridge, the component is either modified to extend its service, removed from the design (e.g. jointless bridges), or a replacement of the component is planned during the service life of the bridge. Once a plausible design has been identified that meets the service life requirements of the bridge system, the framework requires the life cycle cost to be estimated by considering all planned maintenance and replacement activities. In its envisioned implementation, a designer would develop multiple plausible designs and then select the final design based upon the estimated life cycle costs. The key activities within the framework are the estimation of service life for the critical components of the bridge. Critical components are defined as components that are required to maintain the functionality of the bridge. To accomplish this, the guide summarizes the literature related to the performance and service life estimation of materials, bridge decks, reinforcing steel due to corrosion, structural steel due to corrosion, steel due to fatigue and fracture, jointless bridges, expansion devices, and bearings. For each of these elements, a fault tree is provided, which summarizes the influences on service life. SHRP 2 R19B Project R19B was titled “Bridges for Service Life Beyond 100 Years: Service Limit State Design” (Kulicki et al. 2015). The goal of this project was to “develop design and detailing guidance and calibrated service limit states (SLSs) to provide 100-year bridge life, and to develop a framework for further development of calibrated SLSs.” As a first step, a survey to identify both critical bridge performance challenges as well as the sufficiency of current SLSs was conducted. The survey was sent to 31 bridge owners and the research team received 16 responses. The results indicated that the most commonly faced durability problems were associated with expansion joints and deck cracking. Additional durability challenges were identified and included deterioration at the ends of beams, issues associated with painted steel members, issues with bearings, corrosion of reinforcement, and deck overlays. Even with these durability challenges, respondents generally noted that the current SLSs were mainly sufficient but recommended some additional limit states, including: • Foundation settlement • Stress-checks on corrosion-reduced sections • Improved crack control and stress limits in concrete flexural members • Limit states for connections, expansion joints, and bearings. Following the survey all of the current SLSs of AASHTO LRFD were evaluated for suitability for calibration. Some of the SLSs were deemed to be “uncalibrateable” as they were either based on expert judgment or deterministic in nature. Table 2 below provides the limit states that were identified for calibration as part of Project R19B. As apparent from this table, SLS were distinguished based on whether they were reversible or irreversible. Irreversible limit states are those in which consequences remain even after the removal of load while reversible limit states were defined as those for which no consequences remain after the removal of load. This distinction is necessary as it influences the total consequences associated with a limit state and thus influences the target reliability index. Thus, reversible limit states are calibrated using lower target reliability indices than irreversible limit states. In general, target reliability indices between 0 and 1.0 were deemed appropriate for SLSs (compared to 3.5 to 4.0 for ultimate limit states).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 10 Of particular interest to this research project are the results of the calibration of crack control-related limit states (inclusive of the tensile stress limitations for pre-stressed concrete). These results are discussed in detail in relationship to the deemed-to-satisfy provision of this literature review. Table 2. Service limit states calibrated as part of SHRP 2 Project R19B (Kulicki et al. 2015). LRFD Article Reversible No. of Lanes MPF 2.5.2.6.2 Criteria for Deflection Yes Single — 3.4.1 Load Factors and Load Combinations for Fatigue No Single — 5.5.3.1 General—Compressive Stress Limit for Concrete—A Fatigue Criterion No Single No 5.5.3.2 Fatigue of Reinforcing Bars No Single — 5.5.3.4 Fatigue of Welded or Mechanical Splices of Reinforcement No Single — 5.6.3.6 Crack Control Reinforcement—To be revised but not calibrated—Deemed-to-satisfy No — — 5.7.3.4 Control of Cracking by Distribution of Reinforcement—Not calibratable-Deemed-to- satisfy No N/A 5.9.3 Stress Limitations for Pre-stressing Tendons No Multiple Yes 5.9.4.2.2 Tension Stresses Yes Single No 6.10.4.2 Permanent Deformations of Steel Structures No Single No 6.13.2.8 Slip Resistance of Bolts No Single No 10.6.2.4 Settlement Analysis of Shallow Foundations No for footing, possible for superstructure Multiple for sands, none for clays — 10.8.2.2 Settlement (related to drilled shaft groups) No Multiple Yes 10.8.2.4 Horizontal Movement of Shaft and Shaft Groups No — — Note: MPF = multiple presence factor; — = current criteria do not specify whether or not the MPF is applicable; N/A = not applicable Summary of Service Life Design Methodologies As put forth in ISO 16204 (2012) and fib Bulletin 34 (2006), there are two generally accepted strategies to perform service life design. The first (termed Strategy 1) aims to provide a means for the structure to withstand environmental and repeated load effects without reaching objectionable limit states during the target service life. In practice, this strategy is carried out through one or a combination of the following: 1. By selecting/specifying materials that have sufficient durability to withstand deterioration throughout the design service life 2. By providing protective systems (e.g. reinforcement coating, membranes, overlays, steel coatings, joints, and scuppers) 3. By providing dimensions and details to reduce the rate of deterioration (e.g. cover dimension, reinforcement size, and reinforcement spacing)

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 11 4. By selecting a shorter service life for specific elements and planning for their replacement such that their deterioration does not govern the service life of the overall bridge (e.g. joints, bearings, scuppers, and traffic barriers) The second strategy (Strategy 2) aims to remove the vulnerability of a structure to deterioration through the removal of vulnerable details (e.g. joints) or the use of corrosion resistant materials (e.g. some stainless steels). Figure 1 below (from ISO 16204) provides the general procedure for Service Life Design. Within this figure, “full probabilistic,” “partial factor,” and “deemed-to-satisfy,” approaches all fall under Strategy 1, while “avoidance of deterioration” falls under Strategy 2. Source: adapted from ISO (2012) Figure 1. Flowchart for service life design. Brief summaries of each of the approaches that fall under Strategy 1 are provided below. • Full probabilistic – Using this approach, the reliability indices of specific limit states are explicitly computed during the design process using deterioration models. • Partial factor – Using this approach, partial safety factors (e.g. load and resistance factors) are used to allow designers to evaluate limit states given specific target reliability indices during the design process. The partial factors are computed using the full probabilistic approach. This approach is similar to what is currently implemented in AASHTO LRFD for structural design. • Deemed-to-Satisfy – This approach provides designers with a set of prescriptive requirements which, if followed, should produce a bridge with a service life above the minimum specified (for assumed Establishing the serviceability criteria Establishing the general layout, the dimensions and selection of materials Verif ication by the full probabilistic method involving: • Probabilistic models • Resistance • Exposure (loads) • Geometry • Limit States Verif ication by the partial factor method involving: • Design values • Characteristic values • Partial factors • Design Equations • Limit States Verif ication by the deemed-to-satisfy method involving: • Exposure classes • Limit States • Other design provisions Verif ication by the avoidance of deterioration method involving: • Exposure classes • Limit States • Other design provisions Execution specif ication Maintenance plan Condition assessment plan Execution of the structure Inspection of execution Maintenance Condition assessment during operational service life No nc on fo rm ity t o th e pe rf or m an ce c rit er ia r es ul tin g in o bs ol es ce nc e or p ar tia l t o fu ll r ed es ig n

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 12 reliability indices). These prescriptive requirements can be developed and or validated using the full probabilistic approach. By far the most common approach to service life design is through the use of “deemed-to-satisfy” rules. This is a result of both the inability to define precise limit states associated with the end of service life as well as the lack of reliable models to simulate the onset and propagation of the deterioration mechanisms that operate on bridges. Notable exceptions to the lack of simulation models include the ingress of chloride in concrete and carbonation. The following subsections provide a summary of current guidance for each of these methods. Full Probabilistic Methods ISO 16204 (2012) and fib Bulletin 34 (2006), put forth a general framework for service life design considering the degradation of concrete and embedded reinforcing steel using the full probabilistic approach. Specifically, ISO 16204 identifies six limit states while fib Bulletin 34 identifies four limit states: 1. Carbonation-induced depassivation (ISO 16204 and fib Bulletin 34) 2. Chloride-induced depassivation (ISO 16204 and fib Bulletin 34) 3. Corrosion-induced cracking, spalling, and collapse (ISO 16204) 4. Freeze-thaw damage resulting in a local loss of mechanical properties, cracking, scaling and loss of cross-section not in the presence of deicing agents and sea water (ISO 16204 and fib Bulletin 34) 5. Freeze-thaw scaling of concrete in the presence of deicing agents or sea water (ISO 16204 and fib Bulletin 34) 6. Freeze-thaw induced deflection and collapse (ISO 16204) Although broadly accepted models exist for depassivation limit states (items 1 and 2), this is not true for the others. For example, there are no widely accepted models in the industry to estimate the time from depassivation to corrosion-induced cracking, which is discussed later in this section. As a result, while the other limit states (items 3 through 6) may have been explicitly identified, the information sources required to quantitatively evaluate them are not available. Limit State: Depassivation Due to Carbonation (Uncracked Concrete) The probability that depassivation occurs (pdep) is equal to the probability that the depth of depassivation at the design service life (xc(tSL)) is greater than the concrete cover, a (ISO 2012): ( ){ }. 0{} 0dep c SLp p p a x t p= = − < < (1) A “design model” that estimates the depth of depassivation with respect to time is provided by ISO 16204 (2012) as: ( )cx t W k t= × × (2) where: W = a measure based on the local climate of the element being considered k = the resistance to carbonation of the specified concrete mix and execution or placement

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 13 W and k may be estimated by examining similar structures with similar mix, execution, and exposure or from the literature. In the case of existing structures, the product of W and k may be directly measured through material sampling and testing. In addition, fib Bulletin 34 (2006) provides an analytical approach to estimating W and k. Limit State: Depassivation Due to Chloride Ingress (Uncracked Concrete, Marine Environment) The probability that depassivation occurs, pdep, is equal to the probability that the chloride concentration at the rebar depth at the design service life C(a,tSL) is greater than the chloride concentration required for depassivation Ccrit (ISO 2012): ( ){ }. 0{} , 0dep SLcritp p p C C a t p= = − < < (3) A modified version of Fick’s second law (which assumes a marine environment) is provided to estimate the diffusion of chlorides within concrete: ( ) ( ) ( ) 2 , S ap S i p C D t xx erf t t C C C      ⋅    = + − × ×   (4) 0 0( ) ( )app app tD t D t t α  =     (5) where: C(x,t) = chloride content at depth x and time t Cs = chloride content at the concrete surface Ci = initial chloride content of the concrete Dapp(t) = apparent diffusion coefficient of chloride in concrete Dapp(t0) = apparent diffusion coefficient of chloride measured at reference time t0, erf = the error function α = aging factor, which decreases with time and likely lies between 0.2 and 0.8 For the design of a new structure, the parameters Cs, Ci, Dapp(t0), and α should be estimated by examining similar structures (with similar mix, execution, and exposure) or from the literature. In the case of existing structures, these parameters may be directly measured from the structure of interest with α requiring observations at two different times separated by a sufficient interval. Limit State: Corrosion Induced Cracking and Spalling The probability that corrosion-induced cracking occurs pcrack is equal to the probability that the expansion of rebar due to corrosion at the design service life ∆r(S)(tSL) is greater than the expansion of rebar that may be accommodated by the concrete without cracking ∆r(R) (ISO 2012): ( ){ } 0( ) ( ){} 0crack SLR Sp p p r r t p= = ∆ − ∆ < < (6) For this limit state there is no widely accepted model capable of simulating the progression from depassivation to rebar corrosion and then from rebar corrosion to corrosion-induced cracking and spalling of concrete.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 14 Limit State: Freeze/Thaw Damage Causing Local Loss of Mechanical Properties, Cracking, Scaling and Loss in Cross-Section - Without Deicing Agents or Sea Water The probability that freeze-thaw damage occurs without the presence of deicing agents and sea water, pfreeze/thaw damage, is taken as the probability that the temperature in the concrete during its service life T(t<tSL) is below 0 °C at the same time that the actual saturation of the concrete SACT is greater than the critical level of saturation of concrete SCR (ISO 2012): ( ){ }/ 0{} ( 0),freeze thawdamage SLSL CR ACTp p p T t t S S t t p= = < − < < < (7) Annex B within fib Bulletin 34 provides guidance related for the quantification of these various parameters. Limit State: Freeze/thaw Induced Surface Scaling - with Deicing Agents or Sea Water The probability that freeze-thaw induced scaling occurs in the presence of deicing agents or sea water, pscaling, is taken as the probability that the temperature within the concrete during its service life T(t<tSL) is less than the critical temperature required to cause surface scaling TR, which is assumed to be a function of relative humidity, number of cycles, time, and chloride content, among others (ISO 2012): ( ){ } 0{} ( , 0) , , ,scaling cyclesSL Rp p p T t t T RH n pt Cl−= = ≤ − < < (8) Annex B within fib Bulletin 34 provides guidance related for the quantification of these various parameters. Limit State: Freeze/thaw Induced Surface Scaling ISO 16204 requires designers to include the effects of localized changes in material properties due to freeze-thaw when evaluating load carrying capacity and deformations. Partial Factor Methods Using the full probabilistic approaches described above, principles for partial factor approaches are proposed and are ready for implementation for three of the limit states identified: 1. Depassivation due to carbonation (uncracked concrete) 2. Depassivation due to chloride ingress (uncracked concrete) 3. Freeze-thaw damage (not in the presence of deicing agents and seawater) Limit State: Depassivation due to Carbonation (Uncracked Concrete) This limit state is violated when the design depassivation due to carbonation at the end of the service life xc,d(tSL) is greater than the design cover of the concrete provided ad (ISO 2012). , ( ) 0d c d SLa x t− ≥ (9) where: ad = the nominal cover thickness minus the permitted deviation xc,d(tSL) = the characteristic (or nominal) value of carbonization depth xc,k(tSL) times the partial safety factor, γf.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 15 Limit State: Depassivation due to Chloride Ingress (Uncracked Concrete, Marine Environment) The partial factor approach to depassivation of concrete due to chloride ingress is the same as the depassivation of concrete due to carbonation (ISO 2012). Limit State: Freeze/thaw Damage Causing Local Loss of Mechanical Properties, Cracking, Scaling and Loss in Cross-Section - without Deicing Agents or Sea Water This limit state is violated when the design actual saturation level SACT,d exceeds the design critical saturation level SCR,d during the service life, as (ISO 2012): , , ( ) 0CR d ACT d SLS S t t   (10) where: SCR,d = the characteristic value of the minimum critical degree of saturation SCR,min minus the safety margin for the critical degree of saturation ΔSCR SACT,d = the characteristic value for the actual degree of saturation SACT,k plus the safety margin for the actual degree of saturation ∆SACT. Deemed-To-Satisfy Guidance In some cases, deemed-to-satisfy guidance is not tied to a specific deterioration mechanism. Rather, these provisions are of a “blanket” type and are intended to address all relevant deterioration mechanisms simultaneously. As a result, in this section, the literature is primarily organized based on the type of guidance provided as opposed to the deterioration mechanism addressed by the guidance. Specifically, the following six types of guidance are discussed in the following subsections. 1. Concrete material specifications – inclusive of acceptable types and classes of constituents, water/cement ratio, cement content, compressive strength, air content, etc. 2. Concrete cover dimensions 3. Crack control approaches – inclusive of maximum rebar sizes and spacing 4. Coatings – inclusive of steel coatings and membranes/overlays for concrete 5. Replaceable elements 6. Element-specific guidance The specific requirements provided for each of these categories may be a function of the design service life, the consequences associated with failing to achieve the design service life, and the exposure of the specific element being designed/evaluated. Notable exceptions to the “blanket” approach to deemed-to-satisfy guidance include ISO 16204 (2012), fib Bulletin 34, and ACI 201.2R (2016), which tie their guidance to specific deterioration mechanisms. These provisions are summarized in subsections below. Definition of Exposure Classes A common approach to structuring “deemed-to-satisfy” provisions is through the definition of exposure classes (or categories). In general, an exposure class indicates the harshness of the environment that a bridge or element will be exposed to throughout its service life. Once defined for a bridge or element, this exposure class then triggers specific “deemed-to-satisfy” provisions, with more stringent provisions keyed to harsher exposure classes.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 16 The following section summarizes the exposure classes that are explicitly defined by the Eurocode, the Australian Bridge Design Code, ACI, the Canadian Standards Association, and the Florida DOT Structures Manual. Several definitions for different environments (e.g., marine) provided by State DOTs are also summarized. In addition to standards that explicitly define exposure classes, many other codes and standards (including the AASHTO LRFD Bridge Design Specifications) address exposure conditions directly within each of their “deemed-to-satisfy” provisions. Since these codes and standards do not directly define exposure classes, they are not discussed in the following subsections. Eurocode The Eurocode standards EN 1992-1-1 (2004) and EN 206 (BSI 2014) define exposure classes to reflect the “physical and chemical conditions” that a structure is exposed to “in addition to mechanical stresses”. These exposures classes are defined for five specific deterioration mechanisms as well as one category of “no risk to corrosion or attack”. The following list provides the different levels defined for each category: • X0 Exposure class for no risk of corrosion or attack • XC1 to XC4 Exposure classes for risk of corrosion induced by carbonation • XD1 to XD3 Exposure classes for risk of corrosion induced by chlorides not from sea water • XS1 to XS3 Exposure classes for risk of corrosion induced by chlorides from sea water • XF1 to XF4 Exposure classes for risk of freeze/thaw attack • XA1 to XA3 Exposure classes for risk of chemical attack For each one of these classes, the Eurocode provides a qualitative description of the environment it is intended to reflect as well as examples of where such an environment may occur. Table 3 presents the description of exposure classes as defined by the Eurocode. Table 4 provides limits used to determine the exposure class for concrete subject to chemical attack.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 17 Table 3. Eurocode exposure classes (EN 2004, EN 2013). Class designation Description of the environment Informative examples where exposure classes may occur 1 No risk of corrosion or attack X0 For concrete without reinforcement or embedded metal: All exposures except where there is freeze/thaw, abrasion or chemical attack. For concrete with reinforcement or embedded metal: Very dry Concrete inside buildings with very low air humidity 2 Corrosion induced by carbonation Where concrete containing reinforcement or other embedded metal is exposed to air and moisture, the exposure shall be classified as follows: XC1 Dry or permanently wet Concrete inside buildings with low air humidity; Concrete permanently submerged in water XC2 Wet, rarely dry Concrete surfaces subject to long-term water contact; Many foundations XC3 Moderately humidity Concrete inside buildings with moderate or high air humidity; External concrete sheltered from rain XC4 Cyclic wet and dry Concrete surfaces subject to water contact, not within exposure class XC2 3 Corrosion induced by chlorides other than from sea water Where concrete containing reinforcement or other embedded metal is subject to contact with water containing chlorides, including deicing salts, from sources other than from sea water, the exposure shall be classified as follows: XD1 Moderate humidity Concrete surfaces exposed to airborne chlorides XD2 Wet, rarely dry Swimming pools; Concrete exposed to industrial waters containing chlorides XD3 Cyclic wet and dry Parts of bridges exposed to spray containing chlorides. Pavements, car park slabs 4 Corrosion induced by chlorides from sea water Where concrete containing reinforcement or other embedded metal is subject to contact with chlorides from sea water or air carrying salt originating from sea water, the exposure shall be classified as follows: XS1 Exposed to airborne salt but not in direct contact with sea water Structures near to or on the coast XS2 Permanently submerged Parts of marine structures XS3 Tidal, splash and spray zones Parts of marine structures

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 18 Table 3. Eurocode exposure classes (EN 2004, EN 2013) (continued). Class designation Description of the environment Informative examples where exposure classes may occur 5 Freeze/thaw attack with or without deicing agents Where concrete is exposed to significant attack by freeze/thaw cycles while wet, the exposure shall be classified as follows: XF1 Moderate water saturation, without deicing agent Vertical concrete surfaces exposed to rain and freezing XF2 Moderate water saturation, with deicing agent Vertical concrete surfaces of road structures exposed to freezing and airborne deicing agents XF3 High water saturation, without deicing agent Horizontal concrete surfaces exposed to rain and freezing XF4 High water saturation, with deicing agent or sea water Road and bridge decks exposed to deicing agents; Concrete surfaces exposed to direct spray containing deicing agents and freezing; Splash zones of marine structures exposed to freezing 6 Chemical attack Where concrete is exposed to chemical attack from natural soils and ground water, the exposure shall be classified as follows: XA1 Slightly aggressive chemical environment Concrete exposed to natural soil and ground water according to Table 4 XA2 Moderately aggressive chemical environment Concrete exposed to natural soil and ground water according to Table 4 XA3 Highly aggressive chemical environment Concrete exposed to natural soil and ground water according to Table 4

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 19 Table 4. Eurocode limiting exposure class values for chemical attack from natural soil and ground water (EN 2013). Chemical characteristic Reference test method XA1 XA2 XA3 Ground water SO42- mg/l EN 196-2 ≥ 200 and ≤ 600 > 600 and ≤ 3000 > 3000 and ≤ 6000 pH ISO 4316 ≤ 6.5 and ≥ 5.5 < 5.5 and ≥ 4.5 < 4.5 and ≥ 4.0 CO2 mg/l aggressive EN 13577 ≥ 15 and ≤ 40 > 40 and ≤ 100 > 100 up to saturation NH4+ mg/l ISO 7150-1 ≥ 15 and ≤ 30 > 30 and ≤ 60 > 60 and ≤ 100 Mg2+ mg/l EN ISO 7980 ≥ 300 and ≤ 1000 > 1000 and ≤ 3000 > 3000 up to saturation Soil SO42- mg/kga total EN 196-2 b ≥ 2000 and ≤ 3000c > 3000c and ≤ 12000 > 12000 and ≤ 24000 Acidity according to Baumann Gully ml/kg prEN 16502 > 200 Not encountered in practice a Clay soils with a permeability below 10-5 m/s may be moved into a lower class. b The test method prescribes the extraction of SO42- by hydrochloric acid; alternatively, water extraction may be used, if experience is available in the place of use of the concrete. c The 3000 mg/kg limit shall be reduced to 2000 mg/kg, where there is a risk of accumulation of sulfate ions in the concrete due to drying and wetting cycles or capillary suction. Australian Bridge Design Code The Australian Bridge Design Code (AS 5100.5 2004) provides four general categories of exposure classes for bridge elements. These include (1) surfaces of members in contact with the ground, (2) surfaces of members in interior environments, (3) surfaces of members in above ground exterior environments, (4) surfaces of members in water, and (5) surfaces of members in other environments. Table 5 provides the classification provided by AS 5100.5 (2004). The classifications provided move from the least aggressive environment (A) to the most aggressive environment (C). The designation U essentially corresponds to “unknown” as the code states that for this classification the “degree of severity of exposure should be appropriately assessed”.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 20 Table 5. AS 5100.5 exposure classes (AS 2004). Surface and exposure environment Exposure Classification 1 Surfaces of members in contact with the ground (see Note 1): (a) Members in nonaggressive soil (see Note 2) B1 (b) Members protected by a damp-proof membrane A (c) Members in aggressive soils (See Note 3) U (d) Members in salt-rich arid areas C 2 Surfaces of members in interior environments: Fully enclosed within a structure except for a brief period of weather exposure during construction A 3 Surfaces of members in above ground exterior environments in the following areas: (a) Inland (greater than 50 km from coastline) environment being— (i) non-industrial and arid climatic zone (see Notes 4 and 5); A (ii) non-industrial and temperate climatic zone; A (iii) non-industrial and tropical climatic zone; or B1 (iv) industrial and any climatic zone B1 (b) Near-coastal (1 km to 50 km from coastline), any climatic zone B1 (c) Coastal (up to 1 km from coastline but excluding tidal and splash zones) (see Note 6), and climatic zone B2 4 Surfaces of members in water (see Note 1): (a) In fresh water B1 (b) In sea water (i) Permanently submerged B2 (ii) In tidal or splash zones C (c) In soft or running water U 5 Surfaces of members in other environments: Any exposure environment not described in Items 1 to 4 above U Notes: 1 Members, such as piles without permanent steel casing, shall be classified as members in water unless it is proved by geotechnical investigation that no part of the member is below the permanent water table level. 2 If testing has been undertaken to ascertain that the soil is in contact with concrete is nonaggressive, then exposure classification A may be used, provided that the soil is not subject to wetting and drying. Typically. members in the top 500 mm of soil would not qualify for this reduction. 3 Permeable soils with a Ph less than 4.0 or with ground water containing more than 1 g per litre of sulfate, would be considered aggressive. 4 The climatic zones referred to are those shown in Figure 4.3, which is a simplified version of Plate 8 of the Bureau of Meteorology publication Climate in Australia, 1982 Edition. 5 Industrial refers to areas that are within 3 km of industries that discharge atmospheric pollutants. 6 For the purpose of this Table, the coastal zone includes locations within 1 km of the shoreline of large expanses of salt water, e.g., Port Phillip, Sydney Harbour east of the Spit Bridge and Harbour Bridge, Swan River west of the Narrows Bridge. Where there are strong prevailing winds or vigorous surf, the distance should be increased beyond 1 km and higher levels of protection should be considered. Proximity to small salt water bays, estuaries and rivers may be disregarded, except for structures immediately over or adjacent to such bodies of water. ACI ACI classifies the environment using exposure categories and classes within ACI 318 (2014) and ACI 201.2R (2016). The following four exposure categories representing common deterioration mechanisms or sources of deterioration for concrete are defined:

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 21 • Freezing and thawing (F) • Sulfate (S) • In contact with water (W) • Corrosion protection of reinforcement (C) Each category is then broken down into several classes based on the severity of the environment. Table 6 gives the definition of each ACI exposure class. Table 6. ACI exposure categories and classes (ACI 2014). Category Class Condition Freezing and thawing (F) F0 Concrete not exposed to freezing and thawing cycles F1 Concrete exposed to freezing and thawing cycles with limited exposure to water F2 Concrete exposed to freezing and thawing cycles with frequent exposure to water F3 Concrete exposed to freezing and thawing cycles with frequent exposure to water and exposure to deicing chemicals Sulfate (S) Water-soluble-sulfate (SO42-) in soil, percent by mass1 Dissolved sulfate (SO42- ) in water, ppm2 S0 SO42- < 0.10 SO42- < 150 S1 0.10 ≤ SO4 2- < 0.20 150 ≤ SO42- < 1500 or seawater S2 0.20 ≤ SO42- ≤ 2.00 1500 ≤ SO42- ≤ 10,000 S3 SO42- > 2.00 SO42- > 10,000 In contact with water (W) W0 Concrete dry in service Concrete in contact with water and low permeability is not required W1 Concrete in contact with water and low permeability is required Corrosion protection of reinforcement (C) C0 Concrete dry or protected from moisture C1 Concrete exposed to moisture but not to an external source of chlorides C2 Concrete exposed to moisture and an external source of chlorides from deicing chemicals, salt, brackish water, seawater, or spray from these sources Notes: 1 Percent sulfate by mass in soil shall be determined by ASTM C1580. 2 Concentration of dissolved sulfates in water, ppm, shall be determined by ASTM D516 or ASTM D4130. Canadian Standards Association Environmental exposures are defined within the concrete quality provisions of the Canadian Highway Bridge Design Code (CSA 2014) for chloride-induced corrosion, freeze-thaw attack, and carbonation- induced corrosion without chloride. The various environmental exposures are summarized in Table 7.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 22 Table 7. CSA environmental exposures by deterioration mechanism (CSA 2014). Deterioration Mechanism Environmental Exposure Chloride-induced corrosion Marine Airborne salts Tidal and splash spray Submerged Other than marine Wet, rarely dry Dry, rarely wet Cyclic, wet/dry Freeze-thaw attack Unsaturated Saturated Carbonation-induced corrosion without chloride Wet, rarely dry Dry, rarely wet Cyclic, wet/dry Within the commentary of CSA (2014), guidance is provided for defining the severity of chemical attack on concrete when applicable, as shown in Table 8. Table 8. CSA types of chemical attack (CSA 2014). Type of attack Weak attack Moderate attack Strong attack Very strong attack Water pH value 6.5-5.5 5.5-4.5 4.5-4.0 < 4.0 Aggressive CO2: me CO2/1 15-30 30-60 60-100 > 100 Ammonium: me NH4+/1 15-30 30-60 60-100 > 100 Magnesium: mg Mg++/1 100-300 300-1500 1500-3000 > 3000 Sulfate: mg SO4--/1 200-600 600-3000 3000-6000 > 6000 Soil sulfate: mg SO4--/kg of air- dry soil 2000-3000 3000-12000 12000-24000 > 24000 More explicit definitions of exposure classes are provided in A23.1: Concrete materials and methods of concrete construction (CSA 2009). The following five general exposure classes are defined: • Class C – chloride exposure • Class F – freezing and thawing exposure without chlorides • Class N – not exposed to chlorides or freezing and thawing • Class A – exposed to manure and/or silage gases or liquids • Class S – sulfate exposure Similar to ACI, the exposure is broken down into specific classes based on the severity of the environment, as shown in Table 9. For sulfate exposure, additional provisions are provided in Table 9 to determine the appropriate exposure class. The CSA A23.1 sulfate exposure classes are identical to those

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 23 given by ACI, with the only difference being the additional limits for sulfate in recycled aggregates provided by CSA. Table 9. CSA definitions of exposure classes (CSA 2009). C-XL Structurally reinforced concrete exposed to chlorides or other severe environments with or without freezing and thawing conditions, with higher durability performance expectations than the C-1, A-1, or S-1 classes. C-1 Structurally reinforced concrete exposed to chlorides with or without freezing and thawing conditions. Examples: bridge decks, parking decks and ramps, portions of marine structures located within the tidal and splash zones, concrete exposed to seawater spray, and salt water pools. C-2 Non-structurally reinforced (i.e., plain) concrete exposed to chlorides and freezing and thawing. Examples: garage floors, porches, steps, pavements, sidewalks, curbs, and gutters. C-3 Continuously submerged concrete exposed to chlorides, but not to freezing and thawing. Examples: underwater portions of marine structures. C-4 Non-structurally reinforced concrete exposed to chlorides, but not to freezing and thawing. Examples: underground parking slabs on grade. F-1 Concrete exposed to freezing and thawing in a saturated condition, but not to chlorides. Examples: pool decks, patios, tennis courts, freshwater pools, and freshwater control structures. F-2 Concrete in an unsaturated condition exposed to freezing and thawing, but not to chlorides. Examples: exterior walls and columns. N Concrete not exposed to chlorides, nor to freezing and thawing. Examples: footings and interior slabs, walls, and columns. A-1 Structurally reinforced concrete exposed to severe manure and/or silage gases, with or without freeze-thaw exposure. Concrete exposed to the vapor above municipal sewage or industrial effluent, where hydrogen sulfide gas might be generated. Examples: reinforced beams, slabs, and columns over manure pits and silos, canals, and pig slats; and access holes, enclosed chambers, and pipes that are partially filled with effluents. A-2 Structurally reinforced concrete exposed to moderate to severe manure and/or silage gases and liquids, with or without freeze-thaw exposure. Examples: reinforced walls in exterior manure tanks, silos and feed bunkers, and exterior slabs. A-3 Structurally reinforced concrete exposed to moderate to severe manure and/or silage gases and liquids, with or without freeze-thaw exposure in a continuously submerged condition. Concrete continuously submerged in municipal or industrial effluents. Examples: interior gutter walls, beams, slabs, and columns; sewage pipes that are continuously full (e.g., forcemains); and submerged portions of sewage treatment structures. A-4 Non-structurally reinforced concrete exposed to moderate manure and/or silage gases and liquids, without freeze-thaw exposure. Examples: interior slabs on grade. S-1 Concrete subjected to very severe sulfate exposures. S-2 Concrete subjected to severe sulfate exposure. S-3 Concrete subjected to moderate sulfate exposure. Notes: (1) “C” classes pertain to chloride exposure. (2) “F” classes pertain to freezing and thawing exposure without chlorides. (3) “N” class is exposed to neither chlorides not freezing and thawing.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 24 (4) All classes of concrete exposed to sulfates shall comply with the minimum requirements of S class noted in Tables 2 and 3 (of CSA 2009). In particular, Classes A-1 and A-4 in municipal sewage elements could be subjected to sulfate exposure. Table 10. CSA sulfate exposure classes (CSA 2009). Class of exposure Degree of exposure Water-soluble sulfate (SO4)1 in soil sample, % Sulfate (SO4) in groundwater samples, mg/L2 Water-soluble sulfate (SO4) in recycled aggregate sample, % S-1 Very severe > 2.0 > 10000 > 2.0 S-2 Severe 0.20-2.0 1500-10000 0.60-2.0 S-3 Moderate 0.10-0.2 150-1500 0.20-0.60 Notes: 1 In accordance with CSA A23.2-3B. 2 In accordance with CSA A23.2-2B. Florida DOT Structures Manual The Florida Department of Transportation (FDOT) requires the environment of superstructures and substructures to be independently classified as Slightly Aggressive, Moderately Aggressive, or Highly Aggressive (FDOT 2017). Under this procedure, the substructure cannot be given a lower classification (i.e. less aggressive) than the superstructure. This procedure is summarized by Figure 2 (FDOT 2017). Source: FDOT (2017) Figure 2. FDOT procedure for classifying the environment of superstructures and substructures.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 25 Oregon DOT Bridge Design and Drafting Manual The Oregon Department of Transportation (ODOT) does not explicitly define exposure classes, but does provide corrosion protection provisions for marine and non-marine environments. ODOT provides the following definitions for marine environments in its Bridge Design and Drafting Manual (ODOT 2016): • A location in direct contact with ocean water, salt water in a bay, or salt water in a river or stream at high tide (substructure). • A location within 1/2 mile of the ocean or salt water bay where there are no barriers such as hills and forests that prevent storm winds from carrying salt spray generated by breaking waves. • A location crossing salt water in a river or stream where there are no barriers such as hill and forests that prevent storm winds from generating breaking waves. Georgia DOT Bridge and Structures Design Manual The Georgia Department of Transportation (GDOT) does not explicitly define exposure classes in its Bridge and Structures Design Manual (GDOT 2018). Instead, alterative material and/or design specifications are given for bridges that are located north of a specified boundary line (termed the “Fall Line”), shown in Figure 3 below. Source: GDOT (2018) Figure 3. GDOT "Fall Line" for material and design specifications. Additionally, GDOT specifies tensile stress limits to control cracking in PS concrete beams based on exposure severity. “Severe exposure” criteria apply to any bridge over waterways located partially or completely within a coastal county (GDOT 2018). “Normal exposure” criteria are used for all other bridges. Virginia DOT The Virginia Department of Transportation (VDOT) does not specify its own exposure classifications within its VDOT Modifications to the AASHTO LRFD Bridge Design Specifications (2018a). Instead it adopts the classifications of the AASHTO LRFD Bridge Design Specifications for Class 1 and Class 2 Exposure Conditions for crack control (see provisions on concrete crack control within Section 5).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 26 Additionally, VDOT specifies alternative provisions for concrete cover based on three condition classes: normal, corrosive, and marine. Corrosive environment affects cover where concrete surface is in permanent contact with corrosive soil. Marine includes all locations with direct exposure to brackish or salt water (VDOT 2018a). PennDOT Design Manual, Part 4 The Pennsylvania Department of Transportation (PennDOT) supplements the AASHTO LRFD Bridge Design Specifications in its Design Manual, Part 4 (2015) by modifying the exposure classes of concrete flexural members with the following: Class 1 applies to all reinforced concrete members except precast and cast-in-place box culverts, segmental construction and for the specific conditions defined under Class 2. Class 2 exposure also applies to precast and cast-in-place box culverts. (Article 5.7.3.4) PennDOT also notes which conditions may be indicative of corrosive soils and groundwater for driven piles: Conditions which are indicative of potentially corrosive soil and groundwater and require consideration of protective measures: • Resistivity less than 2000 ohm-cm in soil • Resistivity between 2000 and 5000 ohm-cm and combined with: – sulfate concentration greater than 200 ppm, or – chloride concentration greater than 100 ppm • pH less than 5.5 • pH between 5.5 and 8.5 in soils with high organic content • Sulfate concentration greater than 1000 ppm in soil or greater than 150 ppm in groundwater • Landfills and cinder fills • Soils subject to mine or industrial drainage • Mixtures of high resistivity soils and low resistivity high alkaline soils Conditions which are low in corrosion potential and which generally do not require protective measures include: • Undisturbed natural soils with no free draining layers, regardless of conditions, noted above as indications of high corrosion potential • pH greater than 5.5 with no organic content • Soils with resistivities greater than 5000 ohm-cm and uniform in profile • Well-aerated loose soils of uniform composition (i.e., sand) Water shall be considered corrosive if it contains any of the following: • Chloride content greater than 1000 ppm • Sulfate content greater than 150 ppm • Mine or industrial runoff • High organic content • pH less than 5.5 Water with high velocity is generally more damaging than standing water. Piles exposed to air containing sulfur dioxide, chlorine concentrations, or other pollutants require protection against deterioration

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 27 Deemed-to-Satisfy and Avoidance Provisions Based on Specific Deterioration Mechanisms ISO 16204 and fib Bulletin 34 Unlike most of the other codes and standards reviewed, ISO 16204 (2012b) and fib Bulletin 34 (2006) differ in two ways. First, they do not provide detailed deemed-to-satisfy provisions. Second, the general guidance provided for deemed-to-satisfy provisions are based on specific deterioration mechanisms. For example, the following is the entire guidance related to deemed-to-satisfy provisions provided by ISO 16204: • Corrosion-induced cracking – “Within this approach a trading-off of geometrical (concrete cover to reinforcement) material parameters (indirectly linked to diffusion and binding characteristics) and execution aspects (compaction and curing) is applied.” • Freeze-thaw attack – “Within this approach one or more of the following remedies are normally required, based on calibration to long- term experience: limitations to the porosity of the concrete (traditionally expressed by the water/cement ratio); available space for expansion of freezing water (air entrainment); selection of type of cement and aggregates.” • Influence of cracks on reinforcement corrosion – ISO 16204 considers a simplified approach that assumes that “corrosion of the reinforcement is not influenced by crack widths under a certain characteristic value”. Specifically, “depending on the severity of the environment and sensitivity of the structure, this limiting crack width is normally given as a characteristic value (5 % upper fractile) in the range of 0.2 mm to 0.4 mm.” • Acid attack – “Within this approach one or more of the following remedies are normally required, based on calibration to long-term experience: limitations to the porosity of the concrete (traditionally expressed by the water/cement ratio); composite cements or the use of additions like silica fume, fly ash and slag; type of aggregates.” • Sulfate Attack – “Within this approach one or more of the following remedies are normally required based on calibration to long-term experience: maximum amount of C3A in the cement. Cements with C3A less than 3 % to 5 % and blast furnace slag cements with more than 60 % slag are often regarded as “sulfate resistant”; composite cements or the use of additions like silica fume, fly ash and slag; limitations to the porosity of the concrete (traditionally expressed by the water/cement ratio); avoidance of limestone aggregates or fillers in case of the risk of thaumasite formation.” ACI ACI 318 (2014) and 201.2R (2016) provide concrete design recommendations based on the environmental exposures that a structure is anticipated to encounter. ACI does not outline broad categories that include all possible deterioration mechanisms but gives each of the previously defined exposure classes its own design requirements. For example, provisions for freeze-thaw are different for temperate regions or cyclic “extremely severe” freezing for structures kept in a state of near constant saturation. The following bullets summarize the deemed-to-satisfy guidance provided by ACI 201.2R, which has been selectively incorporated into ACI 318. • Freeze-thaw – For this deterioration mechanism, recommendations are included to prevent frost damage: – in new concrete – through proper design – through proper construction practices, and – for existing concrete without adequate air entrainment For new design, ACI gives specific recommendations to combat freeze-thaw effects by reducing the amount of freezable water, using an entrained air-void system, and using good design details. The explicit requirements are provided in Table 11.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 28 Table 11. ACI requirements for freeze-thaw exposure classes (ACI 2016). Exposure Class Minimum f’c,* psi (MPa) Maximum W/CM† Air content Limits on cementitious materials F0 No restriction No restriction No restriction No restriction F1 3500 (25) 0.50 Table 13 (ACI 201.2R Table 4.2.3.2.4) No restriction‡ F2 3500 (25) 0.45 Table 13 (ACI 201.2R Table 4.2.3.2.4) No restriction‡ F3a§ 4500 (32) 0.45** Table 13 (ACI 201.2R Table 4.2.3.2.4) Table 12 (ACI 201.2R Table 4.2.3.1c)‡ F3b# 4500 (32) 0.45** Table 13 (ACI 201.2R Table 4.2.3.2.4) No restriction‡ *The minimum average compressive strength that should be achieved before initial exposure to freezing and thawing. †The maximum W/CM for the in-place concrete to provide adequate restriction of freezable water in the properly- cured concrete. ‡High cementitious material replacement for portland cement frequently results in lower rates of strength gain. Care should be taken to ensure that adequate curing (moisture, temperature, and time) is provided so that the minimum f’c is achieved before initial exposure to freezing and thawing. §Hand-finished surfaces. #Formed and machine-finished surfaces. **A lower W/CM may be needed when corrosion is of concern. – Reducing freezable water – Typically, this is achieved by lowering the water to cement ratio (W/CM) and attaining a minimum compressive strength of approximately 3500 psi through proper curing, as demonstrated by the requirements of Table 11. More stringent measures may be required if deicing salts are present. Limiting the W/CM reduces the amount of freezable water that is available in the cured concrete, while the minimum strength requirement ensures sufficient tensile strength and an adequate amount of mixing water has been reduced through the formation of hydration products. Additional limitations are placed on the percentage of cementitious materials for hand-finished surfaces (Table 12) due modifications to the air-void system and changes to the W/CM that this finishing method causes. Table 12. ACI cementitious materials limitations for Exposure Class F3b (ACI 2016). Cementitious materials Maximum percent of total cementitious materials by mass* Fly ash or other pozzolans conforming to ASTM C618 25 Slag conforming to ASTM C989/C989M 50 Silica fume conforming to ASTM C1240 10 Total of fly ash or other pozzolans, slag, and silica fume 50† Total of fly ash or other pozzolans and silica fume 35† *The total cementitious materials also included ASTM C150/C150M, ASTM C595/C595M, ASTM C845/C845M, and ASTM C1157/C1157M cements. The maximum percentage should include: (a) Fly ash or other pozzolans in Type IP blended cement, ASTM C595-C595M or ASTM C1157/C1157M (b) Slag used in the manufacture of an IS blended cement, ASTM C595/C595M or ASTM C1157/C1157M

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 29 (c) Silica fume, ASTM C1240, present in a blended cement †Fly ash or other pozzolans and silica fume shall constitute no more than 25 and 10 percent, respectively, of total mass of the cementitious materials. – Entrained air-void system – Providing a system with an even distribution of air voids within the paste and with adequate spacing is key to achieving maximum resistance to freezing and thawing, and can typically be done using the recommended entrained air contents given in Table 13. However, parameters in addition to the total air content should be evaluated when specifying an air-void system, including the spacing factor ̅L, the specific surface α, and the Philleo factor F’. Table 13. ACI recommended air contents for frost-resistant concrete (ACI 2016) Nominal maximum aggregate size, in. (mm) Air Content, percent* Exposure Class F1 Exposure Class F2 and F3 3/8 (9.5) 7 7.5 1/2 (12.5) 7 7 3/4 (19) 6.5 7 1 (25) 6.5 6.5 1-1/2 (37.5) 6 6.5 2 (50) 6 6 3 (75) 5 5.5 *Field tolerance on air content is recommended as ±1-1/2 percent. Air content recommendations are based on 18 percent air in the paste portion of the concrete with a Vinsol resin air entraining agent (from an analysis of work by Klieger [1952]). Mixture proportions based on guidance in ACI 211.1 for angular coarse aggregate along with the maximum W/CM values from Table 11 (ACI 201.2R Table 4.2.3.1b) were used to determine the air content recommendations. Mixtures using rounded aggregates will require approximately 1 percent less air due to the lower paste contents associated with rounded aggregates. – Design details – Avoid details that allow for repeated wetting and drying of concrete surfaces, such as surfaces directly below bridge joints. Provide drainage for runoff from flat surfaces, particularly when the runoff contains deicing salts or other aggressive chemicals. • Alkali-Aggregate Reaction (AAR) – Alkali-silica reaction (ASR) “results from a reaction between alkali hydroxides in the pore solution and certain forms of reactive silica present in some types of siliceous or carbonate aggregate” (ACI 2016). The risk of ASR can be minimized by using nonreactive aggregates, limiting the alkali content of the concrete, using supplementary cementitious materials (SCMs), and using chemical admixtures (specifically lithium compounds). – Alkali-carbonate reaction (ACR) results from the “use of certain argillaceous dolomitic limestones” (ACI 2016). The risk of ACR can only be reduced by using nonreactive aggregates. • Sulfate Attack – External sulfate attack – This type of attack can occur due to naturally occurring sulfates found in soil or groundwater adjacent to concrete, industrial, agricultural, and municipal waste discharge, and soil fills that contain industrial waste. Recommendations for protection against external sulfate attack include: decreasing the tricalcium-aluminate (C3A) content via sulfate-resisting cements, lowering the W/CM, using acceptable SCMs, minimizing shrinkage cracking, and proper placement, compaction, finishing, and curing of concrete. Requirements to protect against external sulfate attack based on exposure class are summarized in Table 14.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 30 Table 14. ACI requirements to protect against damage to concrete by external sulfate attack (ACI 2016). Exposure Class W/CM by mass, maximum Prescriptive cementitious material requirements Performance cementitious material requirements Cement types* Maximum expansion when tested using ASTM C1012/C1012M ASTM C150/C150M ASTM C595/C595M ASTM C1157/C1157M At 6 months At 12 months At 18 months S0 No W/CM restriction No type restriction No type restriction No type restriction — — — S1 0.50† Type II‡§ IP (MS), IS (<70) (MS), IT (P<S<70) (MS), or IT (P≥S) (MS) MS 0.10% — — S2 0.45† Type V# IP (HS), IS (<70) (HS), IT (P<S<70) (HS), or IT (P≥S) (HS) HS 0.05% 0.10%|| — S3 0.40† Type V plus pozzolan or slag cement** IP (HS), IS (<70) (HS), IT (P<S<70) (HS), or IT (P≥S) (HS) HS†† — — 0.10% *Alternative combinations of cementitious materials to those listed in Table 13 (ACI 201.2R Table 6.1.4.1b) can be permitted when tested for sulfate resistance and meeting the ASTM C1012/C1012M expansion criteria for the severity of potential exposure. ‡Other available types of cement, such as ASTM C150/C150M Type I or Type III can be permitted in Exposure Class S1 if the C3A content is less than 8 percent. §For seawater exposure, other ASTM C150/C150M cement types with C3A contents up to 10 percent are permitted if W/CM does not exceed 0.40. (Refer to Section 6.3 on seawater exposure). #An ASTM C150/C150M Type III cement with the optional limit of 5 percent can be permitted or ASTM C150/C150M cement of any type having expansion at 14 days no greater than 0.040 percent when testing by ASTM C452/C452M. ||The 12-month expansion limit can be used if the 6-month limit is not met, but is not required if the 6-month limit is met. †Values applicable to normal weight concrete. They are also applicable to structural lightweight concrete except that the maximum W/CM of 0.50, 0.45, and 0.40 should be replaced by specified 28-day compressive strengths of 3750, 4250, and 4750 psi (26, 29, and 33 MPa), respectively. **As stated in ACI 318, the amount of the specific source of the pozzolan or slag cement to be used shall be at least the amount that has been determined by service record to improve sulfate resistance when used in concrete containing Type V cement. Alternatively, the amount of the specific source of the pozzolan or slag cement to be used shall be at least the amount tested in accordance with ASTM C1012 and meeting the criteria shown in the table. ††For Exposure Class S3, ASTM C1157/C1157M HS cement must contain pozzolan cement, slag cement, or both. – Internal sulfate attack – also known as delayed ettringite formation (DEF), internal sulfate attack occurs when concrete is cured at elevated temperatures. This delays the natural formation of ettringite to a later time when moisture is available and the concrete has already hardened. The formation of ettringite is an expansive reaction that can damage the concrete. DEF can also occur in mass concrete when the heat of hydration causes a sufficient increase in the concrete temperature. The risk of DEF

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 31 can be reduced by limiting the maximum internal temperature of the concrete to 158°F at all times. If temperatures above 158°F are unavoidable, the prevention measures in Table 15 are recommended. Table 15. ACI recommended measures to reduce DEF potential in concrete exposed to elevated temperatures at early ages* (ACI 2016) Maximum concrete temperature T Prevention required T ≤ 158°F (70°C) No prevention required 158°F < T ≤ 185°F (70°C < T ≤ 85°C) Use one of the following approaches to minimize the risk of expansion: 1. Portland cement meeting requirements of ASTM C150/C150M moderate or high sulfate-resisting and low-alkali cement with a fineness value less than or equal to 430 m2/kg 2. Portland cement with a 1-day mortar strength (ASTM C109/C109M) less than or equal to 2850 psi (20 MPa) 3. Any ASTM C150/C150M portland cement in combination with the following proportions of pozzolan or slag cement: a) Greater than or equal to 25 percent fly ash meeting the requirements of ASTM C618 for Class F fly ash b) Greater than or equal to 35 percent fly ash meeting the requirements of ASTM C618 for Class C fly ash c) Greater than or equal to 35 percent slag cement meeting the requirements of ASTM C989/C989M for Class C fly ash d) Greater than or equal to 5 percent silica fume (meeting ASTM C1240) in combination with at least 25 percent slag cement e) Greater than or equal to 5 percent silica fume (meeting ASTM C1240) in combination with at least 20 percent Class F fly ash f) Greater tan or equal to 10 percent metakaolin meetings ASTM C618 4. An ASTM C595/C595M or ASTM C1157/C1157M blended hydraulic cement with the same pozzolan or slag cement content as listed in Item 3 T > 185°F (85°C) The internal concrete temperature should not exceed 185°F (85°C) under any circumstances. *Assembled from Ghorab et al. 1980, Ramlochan et al. (2003), Thomas (2001), Thomas et al. (2008b). – Seawater – The reaction of concrete with sulfates present in seawater is similar to that with sulfates in fresh water and soils; however, the deterioration is different in that the concrete “often exhibits erosion, softening, or loss in mass as a result of sulfate attack as opposed to…expansion.” Portions of concrete structures in the tidal and splash zones are typically most susceptible. The risk of sulfate attack via seawater can be minimized by using a low permeability mix. The specification of the S1 exposure class W/CM and cement type provisions from Table 14 are recommended, with several sources recommending more stringent W/CM values of 0.45 to 0.40 to combat chloride exposure as well. Pozzolans or ground blast furnace slag can contribute to concrete having 1/10th to 1/100th the permeability of comparable concrete. Design of concrete to minimize cracks and a 28 day strength of no less than 5 ksi is also recommended. • Chemical Attack – exposure to certain types of chemical solutions will shorten the service life of concrete without proper protective measures. Designing a low permeability concrete and using a less-reactive paste can help to mitigate attack from chemicals. Chemicals and factors that influence chemical attack

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 32 on concrete are summarized in Table 16 and Table 17, respectively. Sub-bullets that follow describe specific types of chemical attack in greater detail. Table 16. ACI identified effects of common chemicals on concrete* (ACI 2016). Rate of attack at ambient temperature Inorganic acids Organic acids Alkaline solutions Salt solutions Miscellaneous Rapid Hydrochloric Nitric Sulfuric Acetic Formic Lactic — Aluminum chloride — Moderate Phosphoric Tannic Sodium † hydroxide greater than 20 percent Ammonium nitrate Ammonium sulfate Sodium sulfate Magnesium sulfate Calcium sulfate Bromine (gas) Sulfite liquor Slow Carbonic — Sodium† hydroxide 10 to 20 percent sodium hypochlorite Ammonium chloride Magnesium chloride Sodium cyanide Chlorine (gas) Seawater Soft water Negligible — ‡ Oxalic Tartaric Sodium hydroxide Less than 10 percent sodium hypochlorite ammonium hydroxide Calcium chloride Sodium chloride Zinc nitrate Sodium chromate Ammonia (liquid) *Refer to Portland Cement Association (2001) for a more complete list of chemicals and their potential effects on concrete. †The effect of potassium hydroxide is similar to that of sodium hydroxide. ‡Not applicable. Table 17. ACI identified factors that influence chemical attack on concrete (ACI 2016). Factors that accelerate or aggravate attack Factors that mitigate or delay attack 1. High permeability due to: a) High water absorption b) High W/CM c) Poor consolidation d) Poor curing e) Cracking and microcracking 1. Low permeability concrete* achieved by: a) Proper mixture proportioning† b) Reduced unit water content c) Increased cementitious material content d) Appropriate use of SCMs e) Air entrainment f) Adequate consolidation g) Effective curing‡ 2. Cracks and separations due to: a) Loading/stress concentrations b) Thermal stress c) Shrinkage 2. Reduced tensile stress in concrete by:# a) Using tensile reinforcement of adequate size, correctly located b) Inclusion of pozzolan to reduce temperature rise c) Provision of adequate contraction joints d) Effective curing 3. Leaching and liquid penetration due to: a) Flowing liquid§ b) Ponding c) Hydraulic pressure 3. Structural design a) Minimize areas of contact and turbulence b) Provision of membranes and protective-barrier system(s)|| c) Provision of adequate drainage and through-flow *Factors that control permeability are discussed in more detail in Chapter 3 (of ACI 201.2R). †The mixture proportions and initial mixing and processing of fresh concrete determine its homogeneity and density.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 33 ‡Poor curing procedures result in flaws and cracks. #Resistance to cracking depends on strength and strain capacity. §Movement of water-carrying deleterious substances increases reactions that depend on both quantity and velocity of flow. ||Concrete that will be frequently exposed to chemicals known to produce rapid deterioration should be protected with a chemically-resistant protective-barrier system. – Seawater – chemical attack from seawater exposure can be complex due to the number of different dissolved salts that are present, the most aggressive being magnesium, sulfate, and chloride. Magnesium ions can react with any of the common hydrates in concrete to form phases that can reduce the binding capacity of the cement paste. Sulfates, as described previously, can cause secondary reaction products such as ettringite, gypsum, and thaumasite that are associated with sulfate attack. Reactions with chlorides can result in phases such as Kuzel’s salt and Freidel’s salt, and is also a concern due to the potential for corrosion of reinforcement. Concrete members most susceptible to attack from seawater include foundations below the saline groundwater level, and surfaces located in the tidal and splash zones. Employing a mix design that has low permeability through proper specification of W/CM and type of cementitious materials, achieving good consolidation, and performing adequate curing are key to preventing attack from seawater. Other mitigation strategies include reducing the concrete tendency to microcracking, minimizing the extent and width of cracks through structural design, and cathodic protection. – Acid Attack – Concrete can be exposed to acids from fossil fuel combustion products, sewage, mine and industrial runoff, oxidation of sulfide-bearing minerals in soil, mineral waters, and organic acids from farm and certain manufacturing facilities. Deterioration typically occurs due to the decomposition of cement hydration products below certain pH values. Mitigation is typically performed by using pozzolanic materials such as fly ash and silica fume. However, time of exposure to acids should be minimized and immersion should be avoided. Protective-barrier systems may be required for exposure to highly acidic solutions (pH ≤ 3). – Fresh water – Fresh water is considered “aqueous solutions with nearly neutral pH, very low ionic strength, and low dissolved solids content”. Similar to acid attack, fresh water can cause decomposition of the cement hydration products. Flowing and falling water (e.g., from drainage) can be more aggressive by adding a physical deterioration component. Mitigation approaches include minimizing permeability, reducing the portlandite content of the cement paste, and providing adequate drainage details in design. – Carbonation – This form of attack most commonly occurs when atmospheric carbon dioxide reacts with hydrated cement compounds to form a surface layer of carbonate compounds, which can reduce the porosity but increase the corrosion rate of reinforcement. Low W/CM, good consolidation, and prolonged moist curing can delay or reduce carbonation. Exposure to carbon dioxide of both fresh and mature concrete should be limited. The environment surrounding the concrete may be uncontrollable, but carbonation occurs most rapidly at a relative humidity between 50 and 75 percent. – Industrial chemicals – Concrete may be exposed to an array of industrial chemicals that include “acids, bases, alkalis, corrosives, oxidizers, combustibles, flammables, explosives, cryogenic, and other process-specific conditions.” Exposure conditions could be direct, indirect, or abrasive exposure. Due to the number of chemicals that concrete could be exposed to, the list of possible reactions is nearly inexhaustible. In general, previously defined measures to creating durable concrete should be followed. These include decreasing the permeability, reducing susceptibility to cracking, proper wet curing practices, using corrosion resistant reinforcement, increasing the cover, employing waterstops, and applying surface sealers. – Deicing and Anti-icing chemicals – These chemicals have been used historically to maintain safe driving conditions during the winter and can be broadly categorized as chloride-based and non- chloride-based. Due to the number of chemicals employed, there are numerous reactions that may cause deterioration in the form of increased porosity, acceleration of ASR, scaling, cracking,

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 34 delamination, disintegration, degradation of the cement matrix, and loss of compressive strength, among others. Mitigation strategies include low permeability, good curing, avoiding deicing exposure in the first year of service, and adequate drainage to minimize exposure time. – Environmental structures – These types of structures carry liquids and gases and include “water treatment plants; domestic and industrial wastewater treatments plants; storage tanks and reservoirs; water and wastewater pump stations; conduits, sewers, manholes, and junction chambers; and hazardous materials containment structures.” The reaction mechanisms that occur in these chemical- carrying concrete structures are the same or similar to those previously identified for chemical attack. ACI 350 classifies the chemical agents carried by environmental structures and also establishes the minimum durability and protection requirements for the design of environmental structures. • Physical Salt Attack – This deterioration mechanism is caused by “crystallization of salts in pores near concrete evaporative surfaces”, resulting in damage ranging from surface scaling to disintegration. Environments that are conducive to this form of attack contain water-soluble salts in seawater, groundwater, soil, or other sources in an ambient environment with diurnal fluctuations in temperature and relative humidity, such as arid regions in the southwest U.S. Specific mitigation recommendations cannot be made due to a lack of complete understanding of the deterioration mechanism. However, sodium sulfate and sodium carbonate salts are the tow main types responsible for physical salt attack and, therefore, should be avoided. Means should be provided to separate concrete from salt solutions when physical salt attack is not acceptable. • Corrosion of Metals and Other Embedded Materials – the corrosion process of embedded steel can be briefly summarized as: – Initiation: the passive film layer on the steel breaks down (i.e., depassivation) and the rate of corrosion increases. – Propagation: the steel actively corrodes. – Damage: corrosion has progressed sufficiently to cause cracking and spalling until the structure is no longer able to perform its intended function. Mitigation approaches to corrosion can be broadly grouped by concrete materials, construction, and structural design. – Concrete materials – In general, concrete should be of good quality, have adequate cover, should not be susceptible to cracking, and should have high resistivity. Quality concrete is achieved by using a low W/CM, utilizing SCMs, and limiting the initial chloride content (see Table 18). Cover over reinforcement should be a minimum of 1.5 to 2 in. for moderate to severe exposure conditions. The extent and widths of cracks should be minimized by adhering to appropriate crack width limitation design equations, using the lowest practical water content, supplying sufficient amounts and providing proper detailing of reinforcement, controlling the heat of hydration, limiting restraint effects, using shrinkage reducing admixtures, and employing internal curing techniques. The electrical resistivity can be increased by using SCMs, particularly silica fume.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 35 Table 18. ACI limits on chloride content in newly constructed concrete (ACI 2016). Acid-soluble (ASTM C1152/C1152M) Water-soluble (ASTM C1218/C1218M) Soxhlet method* Pre-stressed concrete 0.08 0.06 0.06 Reinforced concrete in wet conditions 0.10 0.08 0.08 Reinforced concrete in dry or protected conditions 0.20 0.15 0.15 *Soxhlet method described in ACI 222.1R. Note: All chloride contents expressed as percent Cl- by mass of cement. – Construction – Good workmanship, proper reinforcement detailing, adequate curing of sufficient duration, and high quality tight formwork all promote a concrete that will be resistant to corrosion. – Structural design – The general layout of the structure can have a profound effect on the corrosion performance of a concrete element. Controllable design features include the height of a bridge over water and moving an element away from a severe environment (e.g., a splash or spray zone). In addition, drainage should be designed so that water does not pond on surfaces and drainage systems should be sufficiently sloped and extended to carry water away from concrete surfaces. Embedded items (e.g., bolts) should be given special attention such that (1) the embedded item is not susceptible to corrosion and (2) there is not a path around the embedded item to allow for the intrusion of corrosive solutions. Special protective systems may be specified by the designer in an attempt to mitigate corrosion and may include overlays, coated reinforcement, corrosion resistant reinforcement, waterproof membranes, surface sealers, and cathodic protection. • Abrasion – This deterioration mechanism is the result of rubbing and friction on a concrete surface caused by vehicular traffic, wind, water, and other sources. The abrasion resistance is proportional to the compressive strength at the surface and, to a lesser extent, the paste-to-aggregate bond strength and relative hardness of the aggregate. An abrasion resistant concrete can be achieved by utilizing: – a low W/CM at the surface – well-graded aggregates – low-slump – air content consistent with exposure conditions – special aggregates – proper finishing procedures – special dry shakes and toppings – proper curing procedures Canadian Standards Association The Canadian Highway Bridge Design Code (CSA 2014) defers to A23.1: Concrete materials and methods of concrete construction (CSA 2009) for deemed-to-satisfy provisions for deterioration mechanisms other than chloride-induced corrosion, for which CSA (2014) specifies minimum cover values (discussed in later sections). CSA A23.1 contains provisions for chloride content, freeze-thaw attack, sulfate attack, and AAR, among other deterioration mechanisms. CSA A23.1 places the following limits on the initial (before exposure) water-soluble chloride-ion content by mass of cementitious materials: • for pre-stressed concrete: 0.06% • for reinforced concrete exposed to a moist environment or chlorides, or both: 0.15% • for reinforced concrete exposed to neither a moist environment nor chlorides: 1.0%

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 36 Air content requirements to resist freeze-thaw attack are given in Table 19. Air content categories 1 and 2 correspond to exposure classes F-1 and F-2, respectively, from Table 9. CSA A23.1 includes additional requirements on the air-void spacing factor ̅L for concrete in air content Category 1. Table 19. CSA requirements for air content (CSA 2009). Range in air content* for concretes with indicated nominal maximum sizes of coarse aggregate, % Air content category 10 mm 14-20 mm 28-40 mm 1 6-9 5-8 4-7 2 5-8 4-7 3-6 *At the point of discharge from the delivery equipment, unless otherwise specified. Notes: (1) The above difference in air contents has been established based upon the difference in mortar fraction volume required for specific coarse aggregate sizes. (2) Air contents measured after pumping or slip forming may be significantly lower than those measured at the end of the chute. Concrete subject to sulfate attack must meet the cement and expansion testing requirements of Table 20 for the S exposure classes previously defined in Table 10. These requirements are similar to those specified by ACI in Table 14. For the cementing material types listed in Table 20, HS and HSb are high sulfate- resistant hydraulic cement and blended high sulfate-resistant hydraulic cement, respectively; while MS, MSb, and LH are moderate sulfate-resistant hydraulic cement, blended moderate sulfate-resistant hydraulic cement, and low heat of hydration hydraulic cement, respectively. A complete list of cement types is provided in CSA A23.1. Table 20. CSA requirements for sulfate attack (CSA 2009). Class of exposure Cementing materials to be used†† Performance requirements§ Maximum expansion when tested using CSA A3004-C8, % At 6 months At 12 months†† S-1 HS or HSb 0.05 0.10 S-2 HS or HSb 0.05 0.10 S-3 MS, MSb, LH, HS, or HSb 0.10 — Notes: § Where combinations of SCMs and portland or blended hydraulic cements are to be used instead of the cementing materials listed, the performance requirements shall be used to demonstrate equivalent performance against sulfate exposure (see Clauses 4.1.1.6.2, 4.2.1.1, and 4.2.1.3, and 4.2.1.4 of CSA 2009). Such combinations shall not be designated as blended cements. †† If the expansion is greater than 0.05% at 6 months but less than 0.10% at 1 year, the cementing materials combination under test shall be considered to have passed. — “not applicable” Provisions within CSA A23.1, in relation to AAR, state that aggregates “in concrete shall not react with alkalis contained within the concrete to an extent that results in excessive expansion or cracking, or both, of the concrete. When potentially reactive aggregates are to be considered for use, preventative measures acceptable to the owner shall be applied.” Annex B of CSA A23.1 provides additional guidance on AAR

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 37 mitigation. For ACR, the avoidance of the use of susceptible aggregates is recommended. For ASR, risk of attack can be mitigated through selective quarrying or aggregate, reducing the alkali content via a reduction in the cement content, and utilizing fly ash, slag, silica fume, or lithium compounds. ODOT Bridge Design and Drafting Manual ODOT (2016) specifies a minimum “protection system” for corrosion protection of steel reinforcement in structures designated as being in a marine environment. The following minimum deemed-to-satisfy requirements are given: • Stainless steel for all deck, girder and crossbeam reinforcing steel. • Black steel (no epoxy coating) for pre-stressing strands in precast members (to allow for future cathodic protection if needed). • Minimum 2 inches cover on all cast-in-place members. • High-performance concrete (HPC) (microsilica) for all precast and cast-in-place concrete. Table 21 specifies protective practices for deck and end panel reinforcement of structures in different exposure areas. Table 21. ODOT deck and end panel reinforcement protective practices (ODOT 2016). Coastal Areas (within 1 air mile of the Pacific Ocean) Snow/Ice Areas* Mild Areas** Concrete Type HPC (microsilica) HPC (microsilica) HPC (microsilica) Reinforcement Type Deck – Stainless steel top and bottom bars End Panel – Black (uncoated) top and bottom mats Epoxy coated top and bottom mats in both the deck and end panel Black (uncoated) top and bottom mats in both deck and end panel Reinforcement Cover 2 inch top mat 2 inch bottom mat 2.5 inch top mat 1.5 inch bottom mat 2.5 inch top mat 1.5 inch bottom mat *Snow/Ice areas are defined as all areas of central and eastern Oregon, the Columbia River Gorge, Jackson County, and any other areas above 1500 feet elevation. These areas are intended to include all areas with the potential to receive periodic application of deicing chemicals. **Mild areas are defined as all areas not in a coastal or in a snow/ice area. This includes all of western Oregon below 1500 feet elevation that is not within 1 mile of the Pacific Ocean. Virginia Transportation Research Council Over the last two decades, the state of Virginia has sponsored several research projects related to bridge deck deterioration and service life through the Virginia Transportation Research Council (VTRC). Technical reports associated with these research efforts have concluded that epoxy coated rebar provides little benefit to the service life of a bridge, claiming an added benefit of less than 5 years (Weyers et. al 2003, Weyers et al. 2006). These early reports have led the state of Virginia to transition from using coated steels in bridge deck construction to uncoated, corrosion resistant alloyed steel to reduce costs associated with corrosion in bridge decks (VDOT 2018a, VDOT 2018b). VTRC currently has several active research projects to evaluate the performance of corrosion resistant steels used in bridge construction.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 38 PennDOT Design Manual, Part 4 PennDOT specifies the use of “Single Deck” or “Dual Deck” Protection Systems, depending on whether a bridge is designated a Federal Aid Project. Deck replacements are treated differently than new bridge designs or full bridge replacements. At minimum, a Single Deck Protection System must be used for all decks, which includes any of the following: • Filled and partially-fill metal grid deck. • Epoxy coated reinforcement. • Galvanized reinforcement. Dual Deck Protection Systems must be one of the following combination systems: • Epoxy coated or galvanized reinforcement in combination with a 1-1/4 inch thick latex modified or micro silica modified concrete overlay over an 8 inch minimum thickness Class AAAP or HPC concrete deck. At minimum, 2 inches of clear cover over the top mat of reinforcement is required instead of the typical 2-1/2 inches per PennDOT Standards. • Filled or partially filled galvanized metal grid deck system, overfilled 1 inch during initial placement (overfilled monolithically) in combination with a 1-1/4 inch thick latex modified or micro silica modified concrete overlay. 1 inch of overfill is required instead of the typical 1-1/2 inches per PennDOT Standards. • Stainless steel (solid stainless steel) top mat reinforcement and bottom mat reinforcement with an 8 inch minimum thickness Class AAAP or HPC concrete deck. All ties, chairs and hardware in contact with reinforcement must be stainless steel. PennDOT specifies the use of epoxy coated reinforcing bars for specific members and conditions. For superstructures, epoxy coated bars must be used for: • All bars in reinforced concrete deck slabs, curbs, barriers and backwalls, including bars that protrude into these elements from other portions of the structure, as well as diaphragm bars next to expansion dams and protruding into the deck slab. • All bars a distance of 9 ft. from the ends of pre-stressed concrete beams, regardless of beam type. This requirement applies only to the beam end adjacent to a deck joint regardless of the joint dam/seal type. In regard to substructures, epoxy coated bars and breathable sealant are required above a distance 3 ft. below the finished grade for: • Piers located under expansion joints or exposed to salt spray. • Abutments, wingwalls and retaining walls exposed to salt spray. • Portions of abutments exposed to discharge from troughs located below expansion dams. • Abutment stems located below expansion joints. In addition, epoxy coated bars are required for the main reinforcing bars of abutment, wing, and pier footings that extend into a stem or column. Guidelines for the selection of design and protection details of steel and concrete piles are listed in PennDOT (2015) as follows: • Steel Piles

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 39 – If anticipated corrosion loss is less than 1/16 in., deduct 1/16 in. minimum from the exposed surface of the pile when computing section capacity. – Apply a coating with adequate dielectric strength, resistance to abrasion during driving, and proven service in the type of corrosive environment anticipated. – Use a 4-in.-thick minimum concrete encasement around the pile. Coat the steel embedded in the bottom few feet of the concrete jacket with an electrostatically applied epoxy. – Install a cathodic protection system in combination with coating the pile with a coating resistant to cathodic disbondment. Electrically connect piles exposed to unpolluted fresh water to anticipate the future installation of a cathodic protection system if corrosion is discovered during future inspections. • Concrete Piles – Dense, impervious concrete – Minimum concrete cover:  4 in. for cast-in-place reinforced concrete  3 in. for precast reinforced concrete  1-1/2 in. for pre-stressed strands and 1-1/2 in. for secondary reinforcement in pre-stressed concrete – Maximum W/CM ratio of 0.45 (by weight) – Air entrainment – No concrete additives containing chlorides – Sulfate resistant cement as shown in Table 22 (particular cases): Table 22. PennDOT cement requirements for concrete piles exposed to sulfates (PennDOT 2015). Water-Soluble Sulfate in Soil (%) Sulfate in Water (ppm) Cement Type 0.10 - 0.20 150 - 1,500 II 0.20 - 2.00 1,500 - 10,000 V > 2.00 > 10,000 V plus Pozzolan – Epoxy coated reinforcement (particular cases) – Cathodic protection with electrical continuity between all reinforcement (particular cases). Cathodic protection should not be used for pre-stressed piles. Concrete Material Specifications Eurocode Based on the Eurocode exposure classes, previously defined in Table 3, EN 206 (BSI 2014) provides concrete mix design recommendations in the form of permitted types and classes of constituents, maximum W/CM, minimum cement content, minimum concrete compressive strength class, and minimum air content (see Table 23). The assumption implicit in these requirements is that the design working life is 50 years. In addition to these requirements, EN 206 also provides maximum limitations on the chloride content of concrete for plain (unreinforced) concrete (1% by mass of cement), reinforced concrete (0.2-0.4% by mass of cement), and pre-stressed concrete (0.1%-0.2% by mass of cement).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 40 Table 23. Eurocode recommended limiting values for composition and properties of concrete (EN 2013). Exposure classes No risk of corro- sion or attack Carbonation-induced corrosion Chloride-induced corrosion Freeze/thaw attack Aggressive chemical environments Sea water Chloride other than from sea water X0 XC1 XC2 XC3 XC4 XS1 XS2 XS3 XD1 XD2 XD3 XF1 XF2 XF3 XF4 XA1 XA2 XA3 Maximum W/CMc — 0.65 060 0.55 0.50 0.50 0.45 0.45 0.55 0.55 0.45 0.55 0.55 0.50 0.45 0.55 0.50 0.45 Minimum strength class C12/15 C20/25 C25/ 30 C30/ 37 C30/ 37 C30/ 37 C35/ 45 C35/ 45 C30/ 37 C30/ 37 C35/ 45 C30/ 37 C25/ 30 C30/ 37 C30/ 37 C30/ 37 C30/ 37 C35/ 45 Minimum cement contentc (kg/m3) — 260 280 280 300 300 320 340 300 300 320 300 300 320 340 300 320 360 Minimum air content (%) — — — — — — — — — — — — 4.0a 4.0a 4.0a — — — Other require- ments — — — — — — — — — — — Aggregate in accordance with EN 12620 with sufficient freeze/thaw resistance — Sulfate- resisting cementb a Where the concrete is not air entrained, the performance of concrete should be tested according to an appropriate test method in comparison with a concrete for which freeze/thaw resistance for the relevant exposure class is proven. b Where sulfate in the environment leads to exposure classes XA2 and XA3, it is essential to use sulfate-resisting cement conforming to EN 197-1 or complementary national standards. c Where the k-value concept is applied the maximum W/CM ratio and the minimum cement content are modified in accordance with EN 206 5.2.5.2. — not applicable

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 41 Australian Bridge Design Code AS 5100.5 (2004) provides minimum required concrete compression strength based on the defined exposure classes. Table 24 provides the minimum requirements that include both the strength following accelerated curing as well as the characteristic compression strength. In addition to these requirements, the minimum compression strength required for various abrasive environments is provided. These vary from 25 MPa for “footpaths and cycleways” to 40 MPa for “steel-wheeled traffic”. Table 24. AS 5100 strength requirements for exposure classes (AS 2004). Exposure classification Compressive strength of concrete at the completion of accelerated curing MPa Minimum characteristic strength, f’c MPa A 15 25 B1 20 32 B2 25 40 C 32 50 For concrete surfaces exposed to freeze-thaw, AS 5100.5 provides two additional requirements. First, minimum compression strengths of 32 MPa and 40 MPa are required for concrete exposed to less than 25 freeze-thaw cycles per year and more than 25 freeze-thaw cycles per year, respectively. Second, air entrainment of 4-8% and 3-6% are required for average aggregate sizes of 10-20 mm and 40 mm, respectively. Finally, limitations on the chemical composition of concrete are provided. Acid-soluble chloride-ion content (per unit volume on concrete) is limited to values of 3.0 kg/m3 and 0.8 kg/m3 for plain (unreinforced) concrete and reinforced/pre-stressed concrete, respectively. In addition, the mass of acid-soluble SO3 is required to be less than 5% of the mass of cement. Canadian Standards Association The Canadian Highway Bridge Design Code (CSA 2014) provides maximum W/CM (by mass) based on the exposure conditions for specific deterioration mechanisms previously defined in Table 7. Table 25 presents the limitations provided by CSA (2014). In cases not covered by this table, a W/CM of 0.50 is required. Additional requirements for concrete materials and constituents are covered in CSA A23.1: Concrete materials and methods of concrete construction (CSA 2009), and are presented in Table 27.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 42 Table 25. CSA maximum water/cement ratios (CSA 2014) Deterioration Mechanism Environmental Exposure Maximum ratio*†‡ Chloride-induced corrosion Marine Airborne salts 0.45 Tidal and splash spray 0.45 Submerged 0.40 Other than marine Wet, rarely dry 0.40 Dry, rarely wet 0.40 Cyclic, wet/dry 0.40 Freeze-thaw attack§ Unsaturated 0.45 Saturated 0.40 Carbonation-induced corrosion without chloride Wet, rarely dry 0.50 Dry, rarely wet 0.50 Cyclic, wet/dry 0.45 *Unless otherwise approved †Water-to-cementing materials ratio by mass. Cementing materials included Portland cement, silica fume, fly ash, and slag. ‡The ratio shall be independently verified on the submitted concrete mix design and concrete materials. Quality control and quality assurance measures shall be taken to ensure uniformity of concrete production so that water/cement limits are maintained throughout production. Such measures shall include measurements of slump, air content, unit weight, and strength. §Air content shall be in accordance with CSA A23.1. The minimum air content shall be 5.5% for concrete in saturated conditions unless otherwise approved.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 43 Table 26. CSA concrete material requirements (CSA 2009). Curing type (see Table 20 CSA 2009) Class of exposure Maximum water-to- cementing materials ratio† Minimum specified compressive strength (MPa) and age (d) at test† Air content category as per Table 19 Normal Concrete HVSCM1 HVSCM2 Chloride-ion penetrability requirements and age at test‡ C-XL 0.40 50 within 56 d 1 or 2§ 3 3 3 < 1000 coulombs within 56 d C-1 or A- 1 0.40 35 at 28 d 1 or 2§ 2 3 2 < 1500 coulombs within 56 d C-2 or A- 2 0.45 32 at 28 d 1 2 2 2 — C-3 or A- 3 0.50 30 at 28 d 2 1 2 2 — C-4** or A-4 0.55 25 at 28 d 2 1 2 2 — F-1 0.50 30 at 28 d 1 2 3 2 — F-2 0.55 25 at 28 d 2 1 2 2 — N As per the mix design for the strength required For structural design None 1 2 2 — S-1 0.40 35 at 56 d 2 2 3 2 — S-2 0.45 32 at 56 d 2 2 3 2 — S-3 0.50 30 at 56 d 2 1 2 2 — †The minimum specified compressive strength may be adjusted to reflect proven relationships between strength and the water-to-cementing materials ratio. The water-to-cementing materials ratio shall not be exceeded for a given class of exposure. ‡In accordance with ASTM C1202, an age different from that indicated may be specified by the owner. Where calcium nitrite corrosion inhibitor is to be used, the same concrete mixture, without calcium nitrite, shall be prequalified to meet the requirements for the permeability index in this Table. For field testing, the owner shall specify the type of specimen and location from which it is taken. If cores are required, the concrete cores shall be taken in accordance with Clause 6.1.2.3.3 of CSA S413. §Use air content category 1 for concrete exposed to freezing and thawing. Use air content category 2 for concrete not exposed to freezing and thawing. **For class of exposure C-4, the requirements for air entrainment should be waived when a steel troweled finish is required. The addition of supplementary cementing materials may be used to provide reduced permeability in the long term, if required. AASHTO The AASHTO LRFD Bridge Design Specifications (2017a) cite the AASHTO LRFD Bridge Construction Specifications (2017b) for concrete class mix design requirements. Table 27 provides a summary of the concrete classes, which are defined based upon minimum cement content, maximum W/CM, air entrainment, coarse aggregate size, and 28-day compressive strength. The AE terms shown in Table 27 refer to “air entrained” and are used when the concrete is exposed to freeze-thaw cycles. The classes with HPC designations in Table 27 refer to “high performance concrete (HPC)” and are intended for use in high strength pre-stressed concrete members (Class P(HPC)) or for cast-in-place concrete with specified

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 44 performance criteria in addition to compressive strength (Class A(HPC)). The commentary of AASHTO (2017a) provides the following guidance for where each of the concrete classes should be used: • Class A – this class is used for all areas of a structure for which other concrete classes are not specified, and specifically for concrete exposed to saltwater • Class B – is used for footings, pedestals, massive pier shafts, and gravity walls • Class C – is used for thin sections (less than 4 in. thick) and as a filler for steel grids, etc. • Class P – is used when concrete strengths greater than 4.0 ksi is required (e.g., pre-stressed concrete) • Class S – is used for concrete deposited under water to seal out water In addition, AASHTO (2017a) requires that pre-stressed concrete elements and bridge decks have minimum concrete strengths of at least 4.0 ksi. In addition, concrete used in “structural applications” is required to have a minimum compressive strength of 2.4 ksi. Table 27. AASHTO concrete mix characteristics by class (AASHTO 2017b). Class of Concrete Minimum Cement Content Maximum Water/ Cementitious Material Ratio Air Content Range Size of Coarse Aggregate Per AASHTO M 43 (ASTM D448) Size Numbera Specified Compressive Strength lb/yd3 lb per lb % Nominal Size ksi at days A 611 0.49 — 1.0 in. to No. 4 57 4.0 at 28 A(AE) 611 0.45 6 ± 1.5 1.0 in. to No. 4 57 4.0 at 28 B 517 0.58 — 2.0 in. to 1.0 in. and 1.0 in. to No. 4 3 57 2.4 at 28 B(AE) 517 0.55 5 ± 1.5 2.0 in. to 1.0 in. and 1.0 in. to No. 4 3 57 2.4 at 28 C 658 0.49 — 0.5 in. to No. 4 7 4.0 at 28 C(AE) 658 0.45 7 ± 1.5 0.5 in. to No. 4 7 4.0 at 28 P 564 0.49 —b 1.0 in. to No. 4 or 0.75 in. to No. 4 7 67 ≤ 6.0 at b S 658 0.58 — 1.0 in. to No. 4 7 — P(HPC) —c 0.40 —b ≤ 0.75 in. 67 > 6.0 at b A(HPC) —c 0.45 —b —c —c ≤ 6.0 at b Notes: a As noted in AASHTO M 43 (ASTM D448), Table 1-Standard Sizes of Processed Aggregate. b As specified in the contract documents. c Minimum cementitious materials content and coarse aggregate size to be selected to meet other performance criteria specified in the contract. FDOT Structures Manual The FDOT Structures Manual (2017) defines the required class of concrete based upon the type of element within the superstructure or substructure as well as the environmental classification. Table 28 provides the required concrete classifications and Table 29 provides the concrete strengths associated with each concrete class.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 45 Table 28. Required concrete classes (FDOT 2017). Component or Usage Environmental Classification Slightly Aggressive Moderately Aggressive Extremely Aggressive Su pe rs tr uc tu re Cast-in-Place (other than Bridge Decks) Class II Class IV Cast-in-Place Bridge Deck (including Diaphragms) Class II (Bridge Deck) Class IV Approach Slabs Class II (Bridge Deck) Precast or Pre-stressed Class III, IV, V or VI Class IV, V or VI Su bs tr uc tu re Cast-in-Place (except as listed below) Class II Class IV Class IV or V Precast or Pre-stressed (other than piling) Class III, IV, V or VI Class IV, V or VI Cast-in-Place Columns located directly in splash zone Class II Class IV Piling Class V (Special) or VI Drilled Shafts Class IV (Drilled Shafts) Retaining Walls Class II or III Class IV Seals Class III (Seal) See FDOT (2017) Table 1.4.3-2 (Table 29) for minimum 28-day compressive strengths. Corrosion Protection Measures: Calcium nitrite, silica fume, metakaolin or ultrafine fly ash admixtures may be required. Admixture use must conform to the requirements of “Concrete Class and Admixtures for Corrosion Protection.” Table 29. Concrete classes and associated strengths (FDOT 2017). Class of Concrete Minimum 28-Day Compressive Strength (ksi) Class II 3.4 Class II (Bridge Deck) 4.5 Class III 5.0 Class III (Seal) 3.0 Class IV 5.5 Class IV (Drilled Shaft) 4.0 Class V (Special) 6.0 Class V 6.5 Class VI 8.5

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 46 Texas DOT Bridge Design Manual The Texas Department of Transportation (TxDOT) specifies the use of TxDOT Class S (HPC) concrete in their Bridge Design Manual – LFRD (TxDOT 2018) for concrete elements if they are to be regularly subjected to deicing chemicals. GDOT Bridge and Structures Design Manual GDOT (2018) requires the use of Georgia Class D concrete which has a 28-day strength of 4.0 ksi, as specified by Article 5.4.2.1 of the AASHTO LRFD Bridge Design Specifications (2017a) for all non-pre- stressed concrete components. A minimum 28-day strength of 5.0 ksi is required for pre-stressed concrete beams. VTRC A VTRC research project on service life and cost of bridge decks (Williamson et al. 2007) validated a chloride-induced corrosion model accounting for variable input parameters including W/CM ratio and the addition of fly ash or slag. The report concluded that bridge decks built with a W/CM of 0.45 and a concrete cover of 2.75 inches are expected to have a maintenance-free service life of greater than 100 years, regardless of the type of reinforcing steel (Williamson et al. 2007). The report also concluded that W/CM has negligible effect on the diffusion properties of the sampled bridge decks; however the addition of fly ash or slag to the sampled bridge deck concrete mixture was found to dramatically reduce the diffusion rate of chlorides into concrete, providing equivalent long-term corrosion protection effects (Williamson et al. 2007). Concrete Cover Concrete cover protects reinforcing and pre-stressing steel through the delay of depassivation of concrete through carbonation and the ingress of chlorides. As a result, the required concrete cover should be a function of both the environmental exposure class of the concrete as well as the permeability of concrete. Although all of the concrete cover requirements summarized in the following subsections address exposure classes, only the Australian Bridge Design Standard (AS 2004) and the AASHTO LRFD Bridge Design Specifications (2017) give cover requirements as a function of concrete compressive strength or W/CM (which are common surrogates for concrete permeability). Eurocode The Eurocode (EN 2004) defines the minimum cover required cmin as: { }min min, min, , ,max ; ;10mmb dur dur dur addc c c c cγ= + ∆ −∆ (11) where: cmin,b = minimum cover due to bond requirement cmin,dur = minimum cover due to environmental conditions (i.e., durability) Δcdur,γ = additive safety element Δcdur,st = reduction of minimum cover for used of stainless steel reinforcement Δcdur,add = reduction of minimum cover for use of additional protection measures However, since the recommended values for the modifiers are all 0 mm (i.e., Δcdur,γ = Δcdur,st = Δcdur,add = 0 mm) the minimum required cover for durability reduces to cmin,dur. Table 30 and Table 31 below provide the values of cmin,dur for both reinforcing and pre-stressing steel based on the previously defined exposure classes in Table 3.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 47 Table 30. Eurocode minimum cover required for durability of reinforcement steel (EN 2004). Environmental Requirement for cmin,dur (mm) Structural Class Exposure Class according to EN (2004) Table 4.1 (Table 3) X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3 S1 10 10 10 15 20 25 30 S2 10 10 15 20 25 30 35 S3 10 10 20 25 30 35 40 S4 10 15 25 30 35 40 45 S5 15 20 30 35 40 45 50 S6 20 25 35 40 45 50 55 Table 31. Eurocode minimum cover required for durability of pre-stressing steel (EN 2004). Environmental Requirement for cmin,dur (mm) Structural Class Exposure Class according to EN (2004) Table 4.1 (Table 3) X0 XC1 XC2/XC3 XC4 XD1/XS1 XD2/XS2 XD3/XS3 S1 10 15 20 25 30 35 40 S2 10 15 25 30 35 40 45 S3 10 20 30 35 40 45 50 S4 10 25 35 40 45 50 55 S5 15 30 40 45 50 55 60 S6 20 35 45 50 55 60 65 Similar to the concrete material requirements, Eurocode implicitly assumes a design working life (i.e., service life) of 50 years for these cover requirements. For cases where designers want to either take a “credit” for higher quality materials and/or construction quality, or to achieve design service lives greater than 100 years, a procedure is included within EN 1992-1-1 (2004) which involves a modification to the Structural Class of concrete based on the exposure class. These recommended modifications to Structural Class are provided in Table 32.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 48 Table 32. Eurocode recommended structural classification modifications (EN 2004). Structural Class Criterion Exposure Class according to EN (2004) Table 4.1 (Table 3) X0 XC1 XC2/XC3 XC4 XD1 XD2/XS1 XD3/XS2/ XS3 Design Working Life of 100 years increase class by 2 increase class by 2 increase class by 2 increase class by 2 increase class by 2 increase class by 2 increase class by 2 Strength Class1,2 ≥ C30/37 reduce class by 1 ≥ C30/37 reduce class by 1 ≥ C35/45 reduce class by 1 ≥ C40/50 reduce class by 1 ≥ C40/50 reduce class by 1 ≥ C40/50 reduce class by 1 ≥ C45/55 reduce class by 1 Member with slab geometry (position of reinforcement not affected by construction process) reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 Special Quality Control of the concrete production ensured reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 reduce class by 1 Notes: 1. The strength class and w/c-ratio are considered to be related values. A special composition (type of cement, w/c value, fine fillers) with the intent to produce low permeability may be considered. 2. The limit may be reduced by one strength class if air entrainment of more than 4% is applied. Australian Bridge Design Code AS 5100.5 (2004) defines minimum nominal cover dimensions based on (a) the exposure class (see discussion above), (b) the concrete compression strength, and (c) the type of formwork and compaction used. Table 33 provides the minimum nominal cover dimensions for concrete cast with standard formwork and compaction procedures defined by AS 5100.5 (2004). If concrete is cast against excavated material, then the values in Table 33 must be increased by 30 mm. In cases where a waterproof barrier is provided between the excavated material and the concrete, the values in Table 33 must be increased by 10 mm.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 49 Table 33. AS nominal cover requirements for standard formwork and compaction (AS 2004). Exposure classification Nominal cover for concrete of characteristic compressive strength (f’c) not less than mm 25 MPa 32 MPa 40 MPa ≥ 50 MPa A 35 30 25 25 B1 — 45 40 35 B2 — — 55 45 C — — — 70 — not applicable In cases where rigid steel forms are used together with intense compaction, the minimum nominal cover dimensions are provided by Table 34. AS 5100.5 (2004) defines intense compaction as “obtained with vibrating tables or form vibrators.” Table 34. AS nominal cover requirements for rigid formwork and intense compaction (AS 2004). Exposure classification Nominal cover for concrete of characteristic compressive strength (f’c) not less than mm 25 MPa 32 MPa 40 MPa ≥ 50 MPa A 25 25 25 25 B1 — 35 30 25 B2 — — 45 35 C — — — 50 — not applicable Canadian Standards Association CSA (2014) defines both the required nominal concrete cover and the allowable construction tolerance for cover. These requirements are based on different environmental exposures, component type, and whether the cover is provided for reinforcing steel, pre-stressing strands, or post-tensioning ducts. Table 35 provides the cover and tolerance requirements set by CSA (2014).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 50 Table 35. CSA concrete cover and tolerance requirements (CSA 2014). Concrete covers and tolerances Environmental exposure Component Reinforcement/ steel ducts Cast-in-place concrete, mm Precast concrete, mm Deicing chemicals; spray or surface runoff containing deicing chemicals; marine spray (1) Top of bottom slab for rectangular voided deck Reinforcing steel Pretensioning strands Post-tensioning ducts 40 ± 10 — 60* ± 10 40 ± 10 55 ± 5 60* ± 10 (2) Top surface of buried structure with less than 600 mm fill† Top surface of bottom slab of buried structure Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 50 ± 10 65 ± 5 70* ± 10 (3) Top surface of structural component, except (1) and (2)‡ Reinforcing steel Pretensioning strands 70 ± 20 — 55 ± 10 70 ± 5 Post-tensioning ducts Longitudinal Transverse (dd ≤ 60 mm) Transverse (dd > 60 mm) 130* ± 15 90* ± 15 130* ± 15 120* ± 10 80* ± 10 120* ± 10 (4) Soffit of precast deck form Reinforcing steel Pretensioning strands — — 40 ± 10 38 ± 3 (5) Soffit of slab less than 300 mm thick or soffit of top slab of voided deck Reinforcing steel Pretensioning strands Post-tensioning ducts 50 ± 10 — 70* ± 10 45 ± 10 60 ± 5 65* ± 10 (6) Soffit of slab 300 mm thick or thicker or soffit of structural component, except (4) and (5) Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 10 — 80* ± 10 50 ± 10 65 ± 5 70* ± 10 (7) Vertical surface of arch, solid or voided deck, pier cap, T- beam, or interior diaphragm Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 10 — 90* ± 10 60 ± 10 75 ± 5 80* ± 10 (8) Inside vertical surface of buried structure or inside surface of circular buried structure Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 50 ± 10 65 ± 5 70* ± 10 (9) Vertical surface of structural component, except (7) and (8) Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 55 ± 10 70 ± 5 75* ± 10 (10) Precast T-, I-, or box girder Reinforcing steel Pretensioning strands Post-tensioning ducts — — — 35 +10 or -5 50 ± 5 55* ± 10

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 51 Table 35. CSA concrete cover and tolerance requirements (CSA 2014) (continued). Concrete covers and tolerances Environmental exposure Component Reinforcement/ steel ducts Cast-in-place concrete, mm Precast concrete, mm No deicing chemicals; no spray or surface runoff containing deicing chemicals; no marine spray (1) Top of bottom slab for rectangular voided deck Reinforcing steel Pretensioning strands Post-tensioning ducts 40 ± 10 — 60* ± 10 40 ± 10 55 ± 5 60* ± 10 (2) Top surface of buried structure with less than 600 mm fill† Top surface of bottom slab of buried structure Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 20 — 80* ± 15 40 ± 10 55 ± 5 60* ± 10 (3) Top surface of structural component, except (1) and (2)‡ Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 20 — 80* ± 15 50 ± 10 70 ± 5 70 ± 10 (4) Soffit of precast deck form Reinforcing steel Pretensioning strands — — 40 ± 10 38 ± 3 (5) Soffit of slab less than 300 mm thick or soffit of top slab of voided deck Reinforcing steel Pretensioning strands Post-tensioning ducts 40 ± 10 — 60* ± 10 40 ± 10 55 ± 5 60* ± 10 (6) Soffit of slab 300 mm thick or thicker or soffit of structural component, except (4) and (5) Reinforcing steel Pretensioning strands Post-tensioning ducts 50 ± 10 — 70* ± 10 40 ± 10 55 ± 5 60* ± 10 (7) Vertical surface of arch, solid or voided deck, pier cap, T- beam, or interior diaphragm Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 10 — 80* ± 10 50 ± 10 65 ± 5 70* ± 10 (8) Inside vertical surface of buried structure or inside surface of circular buried structure Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 20 — 80* ± 15 40 ± 10 55 ± 5 60* ± 10 (9) Vertical surface of structural component, except (7) and (8) Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 20 — 80* ± 15 50 ± 10 70 ± 5 70* ± 10 (10) Precast T-, I-, or box girder Reinforcing steel Pretensioning strands Post-tensioning ducts — — — 30 +10 or -5 45 ± 5 50* ± 10

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 52 Table 35. CSA concrete cover and tolerance requirements (CSA 2014) (continued). Concrete covers and tolerances Environmental exposure Component Reinforcement/ steel ducts Cast-in-place concrete, mm Precast concrete, mm Earth or fresh water (1) Footing, pier, abutment, or retaining wall Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 55 ± 10 75 ± 5 80* ± 10 (2) Concrete pile Reinforcing steel Pretensioning strands Post-tensioning ducts — — — 40 ± 10 55 ± 5 60* ± 10 (3) Caisson with liner Reinforcing steel Pretensioning strands 60 ± 20 80* ± 15 — — (4) Buried structure with more than 600 mm of fill† Reinforcing steel Pretensioning strands Post-tensioning ducts 60 ± 20 — 80* ± 15 40 ± 10 55 ± 5 60* ± 10 Swamp, marsh, salt water, or aggressive backfill (1) Footing, pier, abutment, or retaining wall Reinforcing steel Pretensioning strands Post-tensioning ducts 80 ± 20 — 100* ± 15 65 ± 10 85 ± 10 90* ± 10 (2) Concrete pile Reinforcing steel Pretensioning strands Post-tensioning ducts — — — 50 ± 10 65 ± 5 70* ± 10 (3) Caisson with liner Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 — — — (4) Buried structure with more than 600 mm of fill† Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20 — 90* ± 15 55 ± 10 70 ± 5 80* ± 10 Cast against and permanently exposed to earth (1) Footing Reinforcing steel 100 ± 25 — (2) Caisson Reinforcing steel Post-tensioning ducts 100 ± 25 120 ± 15 — — Various Components other than those covered elsewhere in this Table Reinforcing steel Pretensioning strands Post-tensioning ducts 70 ± 20§ — 90* ± 15§ 55 ± 10§ 70 ± 5§ 80* ± 10§ *Or 0.5dd, whichever is greater. †Buried structures with less than 600 mm of fill shall have a distribution slab. ‡For concrete decks without waterproofing and paving, increase the concrete cover by 10 mm to allow for wearing of the surface concrete. §Or as approved.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 53 AASHTO The AASHTO LRFD Bridge Design Specifications (2017a) provide required concrete cover dimensions for main reinforcing bars for a variety of exposure conditions (see Table 36). The basic cover dimensions provided by Table 36 assume a W/CM of 0.45. To recognize the relationship between W/CM and concrete permeability, multipliers of 0.8 and 1.2 are provided to modify the cover dimensions for W/CM less than or equal to 0.4 and greater than or equal to 0.5, respectively. The required cover for ties and stirrups is 0.5 in. less than the values shown in Table 36, but no less than 1.0 in. Further, the minimum cover for main reinforcing bars is 1.0 in., which includes both unprotected and epoxy coated bars. Table 36. AASHTO cover for unprotected main reinforcing steel (AASHTO 2017a). Situation Cover (in.) Direct exposure to salt water 4.0 Cast against earth 3.0 Coastal 3.0 Exposure to deicing salts 2.5 Deck surfaces subject to tire stud or chain wear 2.5 Exterior other than above 2.0 Interior other than above • Up to No. 11 bar • No. 14 and No. 18 bars 1.5 2.0 Precast soffit form panels 0.8 Precast reinforced piles • Noncorrosive environments • Corrosive environments 2.0 3.0 Precast pre-stressed piles 2.0 Cast-in-place piles • Noncorrosive environments • Corrosive environments – General – Protected • Shells • Auger cast, tremie concrete, or slurry construction 2.0 3.0 3.0 2.0 3.0 Precast concrete box culverts • Top slabs used as a driving surface • Top slabs with less than 2.0 ft of fill not used as a driving surface • All other members 2.5 2.0 1.0 FDOT Structures Manual The required concrete cover for bridge components as by FDOT (2017) is provided by Table 37. In this table, the S, M, and E designations refer to the Slightly Aggressive, Moderately Aggressive, and Extremely Aggressive exposure classes defined by FDOT.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 54 Table 37. FDOT concrete cover requirements (FDOT 2017). Component (Precast and Cast-in- Place) Concrete Cover (inches) S or M1 E1 Superstructure All internal and external surfaces (except riding surfaces) of segmental concrete boxes, and external surfaces of pre-stressed beams (except the top surface) 2 Top surface of beam top flange 3/4 (min.) Top deck surfaces: Short Bridges2 2 Top deck surfaces: Long Bridges2 2-1/23 All components and surfaces not included above (including wall copings and traffic and pedestrian railings which are not allowed to be constructed using the slip forming method) 2 Front and back surfaces of pedestrian railing and traffic railings, other than single-slope traffic railings, which may be constructed using the slip forming method. 3 Front and back surface of single-slope traffic railings which may be constructed using the slip forming method 2-1/2 Noise Wall Posts and Panels 2 Precast Concrete Perimeter Wall Posts and Panels 1-3/4 Substructure External surfaces cast against earth and surfaces in contact with water 4 4-1/2 Exterior formed surfaces, columns, and tops of footings not in contact with water and all components or surfaces not included elsewhere 3 4 Internal surfaces 3 Beam/Girder Pedestals 2 Pre-stressed Piling 3 Spun Cast Cylinder Piling4 2 Drilled Shafts 6 Auger Cast Piles 4 Retaining Walls (Excluding MSE walls5 and external surfaces cast against earth) 2 3 Box and Three-sided Culverts 2 3 Bulkheads 4

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 55 1 S = Slightly Aggressive; M = Moderately Aggressive; E = Extremely Aggressive 2 Short Bridges are less than or equal to 100 feet long. Long Bridges are more than 100 feet in length. 3 Cover dimensions includes a 0.5 inch allowance for planning. 4 Concrete for spun cast cylinder piling to be used in an extremely aggressive environment must have a documented chloride-ion penetration apparent diffusion coefficient with a mean value of 0.005 in2/year or less, otherwise 3 inch concrete cover is required. See FDOT (2017) 3.5.17 for further limits on splicing of these piles. 5 See FDOT (2017) 3.13 for MSE wall cover requirements. ODOT Bridge Design and Drafting Manual ODOT (2016) defines a minimum of 2 inches of cover for any coated or uncoated reinforcing bar. Exceptions to this guidance are given in Table 38. Table 38. ODOT minimum concrete cover (ODOT 2016). Location: Cover (in) Top of deck slab (main reinforcing)* 2.5 Bottom of deck slab* 1.5 Stirrups and ties in T-beams, bottom rebar of slab spans, and curbs and rails* 1.5 Stirrups in box girder stems with non-bundled ducts* 2.5 Stirrup ties in box girder stems with non-bundled ducts 2 Bottom slab steel in box girders 1 All faces in precast members (slabs, box beams, and girders) 1 Pier and column spirals, hoops or tie bars + (increase to 4’’ if exposed to marine environment or concrete is deposited in water) 2.5 Footing mats for dry land foundations (use 6’’ if ground water may be a construction problem) 3 Footing mats for stream crossing foundations 6 TxDOT Bridge Design Manual TxDOT (2015) specifies minimum concrete cover for the following concrete components: • Concrete Deck Slabs: 2-1/2 inches top mat to reinforcing bars, 1-1/4 inches bottom mat to reinforcing bars, 2 inches to bar ends. • Segmental concrete spans: 2 inches to reinforcing steel for entire cross-section (Add 1 in. grinding allowance to top slab), 2-1/2 inches of top slab cover in locations subjected to deicing chemicals. • Precast Girders: 2 inches to reinforcing steel for the entire cross-section, 2-1/2 inches for locations subjected to deicing chemicals. GDOT Bridge and Structures Design Manual GDOT (2018) specifies a minimum slab thickness and concrete cover based on the structures location in the state. Bridges north of the GDOT “Fall Line” (see Figure 3) are required to have thicker decks and additional concrete cover as prescribed in Table 39 below. Slab thickness limits and rebar cover are specified separately for overhang slabs. A minimum of 1 inch bottom cover is required for all deck slabs (GDOT 2018).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 56 Table 39 . GDOT slab thickness and concrete cover (GDOT 2017). Case No. Bridge Location Minimum Deck Thickness (in.) Top Cover (in.) Grinding/Wearing Thickness (in.) 1 Below Fall Line* County Road with ADT<2000 7-1/2 2 0 2 Below Fall Line* All Other Routes 7-3/4 2-1/4 1/4 3 Above Fall Line* County Road with ADT<2000 8 2-1/2 0 4 Above Fall Line* All Other Routes 8-1/4 2-3/4 1/4 VDOT VDOT (2018a) specifies the minimum cover values in Table 40 required for reinforcement in various environments.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 57 Table 40. VDOT minimum concrete cover (inches) (VDOT 2018a). Location Normal Condition Corrosive Environment1 Marine2 Pier caps, bridge seats and backwalls: Principal reinforcement Stirrups and ties 2-3/4 2-1/4 3-3/4 3-1/4 4 3-1/2 Pier caps, bridge seats and backwalls (at open joint locations): Principal reinforcement Stirrups and ties 3-3/4 3-1/4 3-3/4 3-1/4 4 3-1/2 Footings and pier columns: Principal reinforcement Stirrups and ties 3 2-1/2 4 3-1/2 4 3-1/2 Cast-in-place deck slabs: Top reinforcement3 Bottom reinforcement 2-1/2 1-1/4 2-1/2 1-1/4 2-1/2 2 Precast and cast-in-place deck slab spans: Top reinforcement3 Bottom reinforcement 2-1/2 2 2-1/2 2 2-1/2 3 Pre-stressed slabs and box beams: Top steel Stirrups and ties 1-3/4 1-1/8 2-3/4 1-1/8 2-3/4 1-1/8 Reinforcement concrete box culverts and rigid frames with more than 2 ft. fill over top of slab: Top slab – top reinforcement Top slab – bottom reinforcement Inside walls and bottom slab top mat Outside walls and bottom slab bottom mat 1-1/2 1-1/2 1-1/2 1-1/2 2-1/2 2-1/2 2-1/2 2-1/2 3 3 3 3 Reinforcement concrete box culverts and rigid frames with less than 2 ft. fill over top of slab: Top slab – top reinforcement Top slab – bottom reinforcement Inside walls and bottom slab top mat Outside walls and bottom slab bottom mat 2-1/2 2 1-1/2 1-1/2 2-1/2 2-1/2 2-1/2 2-1/2 3 3 3 3 Rails, rail posts, curbs and parapets: Principal reinforcement Stirrups, ties and spirals 1-1/2 1 1-1/2 1 1-1/2 1 Concrete piles cast against and/or permanently exposed to earth (not applicable for pre-stressed concrete): 3 3 3 Drilled shafts: Principal reinforcement Ties and spirals 4 3-1/2 5 4-1/2 5 4-1/2 All other components not indicated above: Principal reinforcement Stirrups and ties 2-1/2 2 3-1/2 3 3-1/2 3 1 Corrosive environment affects cover where concrete surface is in permanent contact with corrosive soil. 2 Marine includes all locations with direct exposure to brackish and salt water. 3 Includes 1/2 inch monolithic (integral) wearing surface.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 58 NCHRP Synthesis 333 Russell (2004) concludes that “the most important structural design practice to reduce corrosion of reinforcement in uncracked concrete bridge decks is to provide a minimum cover to the top layer of reinforcement of 64 mm (2.5 in.).” PennDOT Design Manual, Part 4 Minimum concrete cover specified in PennDOT (2015) are shown in Table 41. Table 41. PennDOT cover requirements for unprotected main reinforcing steel (PennDOT 2015). Situation Cover (in.) Concrete cast against and permanently exposed to earth 4.0 Concrete exposed to earth 3.0* Concrete exposed to weather and pier columns 3.0 Concrete deck slab • Top reinforcement • Bottom reinforcement 2.5 1.0 Concrete not exposed to weather or in contact with ground • Primary reinforcement • Stirrup, tie and spiral 1.5 1.0 Precast concrete pipes See PennDOT (2015) A12.10.4.2.4e Precast concrete • Box beams • I-beams See PennDOT BD-661M See PennDOT BD-661M Reinforced concrete box culverts, cast-in-place • Top slab – Top bars at grade – All others • Bottom slab – Top bars – Bottom bars • Walls 2.5 2.0 2.5 3.0 2.0 Reinforced concrete box culverts, precast • Top slab – Top bars at grade – Bottom bars – All others • Bottom slab – Top bars – Bottom bars • Walls 2.5 1.5 2.0 2.0 1.5 1.5 *Except 2-in. minimum may be used for the stem steel of the safety wings and walls supporting barriers as shown in the bridge standards.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 59 Concrete Crack Control In general the codes and standards investigated provide two types of guidance related to crack control. The first, and most general type, defines the maximum acceptable crack width based on exposure classes and/or component type. As an example of this type of guidance, the most stringent crack width requirement for non-pre-stressed reinforcement ranges from a low of 0.25 mm (CSA 2014) to a high of 0.32 mm (AASHTO 2017a). In addition to the general guidance, many of the codes and standards provide specific cases where special consideration maybe required. These include the side faces of deep beams, the flanges of reinforced concrete T-beams that are in tension, and temperature and shrinkage reinforcement, among others. Eurocode The Eurocode (EN 2004) provides recommended crack width limits for concrete bridges (see Table 42). Designers have two choices of how to meet these requirements. First, they can explicitly compute the crack size for the load combinations provided in Table 42 and limit the bar size and/or spacing until the criterion is met. Alternatively, designers can adopt the maximum bar diameter and maximum reinforcement spacing limitations provided by Table 43 and Table 44. Table 42. Eurocode recommended maximum crack width wmax (mm) by exposure class (EN 2004). Exposure Class Reinforced members and pre-stressed members without bonded tendons Pre-stressed members with bonded tendons Quasi-permanent load combination Frequent load combination X0, XC1 0.41 0.2 XC2, XC3, XC4 0.3 0.22 XD1, XD2, XD3, XS1, XS2, XS3 Decompression 1For X0, XC1 exposure classes, crack width has no influence on durability and this limit is set to give generally acceptable appearance. In the absence of appearance conditions this limit may be relaxed. 2For these exposure classes, in addition, decompression should be checked under the quasi-permanent combination of loads. Table 43. Eurocode maximum bar diameters for crack control (EN 2004). Steel Stress1 [MPa] Maximum bar size [mm] wk = 0.4 mm wk = 0.3 mm wk = 0.2 mm 160 40 32 25 200 32 25 16 240 20 16 12 280 16 12 8 320 12 10 6 360 10 8 5 400 8 6 4 450 6 5 - 1Under the relevant combinations of actions - not applicable

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 60 Table 44. Eurocode maximum bar spacing for crack control (EN 2004). Steel Stress1 [MPa] Maximum bar spacing [mm] wk = 0.4 mm wk = 0.3 mm wk = 0.2 mm 160 300 300 200 200 300 250 150 240 250 200 100 280 200 150 50 320 150 100 - 360 100 50 - 1Under the relevant combinations of actions Canadian Standards Association Table 45 below provides the maximum crack widths based on the type of embedded steel and the exposure conditions per CSA (2014). Unlike the Eurocode, the Canadian Highway Bridge Design Code provides no tabular limitations for bar size or reinforcement spacing. As a result, designers must directly estimate crack sizes and then modify the reinforcement details until the criteria shown in Table 45 is met. Table 45. CSA maximum crack widths (CSA 2014). Type of structural component Type of exposure Maximum crack width, mm Non-pre- stressed Deicing chemicals; spray or surface runoff containing deicing chemicals; marine spray; swamp; marsh; salt water; aggressive backfill 0.25 Other environmental exposures 0.35 Pre- stressed Deicing chemicals; spray or surface runoff containing deicing chemicals; marine spray; swamp; marsh; salt water; aggressive backfill 0.15 Other environmental exposures 0.20 In addition to limitations provided by Table 45, CSA (2014) also provides prescriptive requirements for minimum reinforcement and reinforcement spacing in the following three cases: • Side Faces of Beams - for beams with depths greater than 750 mm, designers are required to provide longitudinal reinforcement with an area of at least 0.01bwd, where bw is the web width in mm and d is the effective depth in mm (although bw need not be taken greater than 250 mm), spread over 70% of the depth with a spacing less than 200 mm. • Flanges of T-beams – for flanges of T-beams exposed to stresses greater than the cracking strength of concrete, the longitudinal reinforcing bars must be equally spaced across the effective width of the flange or over a width equal to 10% of the span length (whichever is smaller). • Shrinkage and Temperature Reinforcement – In cases where principal reinforcement extends in only one direction, temperature and shrinkage reinforcement perpendicular to the principal reinforcement shall be provided. The area of the reinforcement must be at least 500 mm2/m with a spacing not greater than 300 mm.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 61 AASHTO For flexural reinforced concrete members (excluding decks), the AASHTO LRFD Bridge Design Specifications (2017a) provide maximum reinforcement spacing levels to control cracking. For members which are exposed to stress of at least 80% of the modulus of rupture at the applicable service limit state, the maximum reinforcement spacing (s) of the reinforcement closest to the tension face is given by: 700 2 β e c s ss s d f γ ≤ − (12) where: β 1 0.7( ) c s c d h d = + − (13) βs = ratio of flexural strain at the extreme tension face to the strain at the centroid of the reinforcement layer nearest the tension face γe = exposure factor = 1.00 for Class 1 exposure condition = 0.75 for Class 2 exposure condition dc = thickness of concrete cover measured from extreme tension fiber to center of the flexural reinforcement located closest thereto (in.) fss = calculated tensile stress in mild steel reinforcement at the service limit state not to exceed 0.60fy (ksi) h = overall thickness or depth of the component (in.) The Class 1 exposure condition is intended for members in which cracks may be tolerated due to “reduced concerns of appearance and/or corrosion.” Where there is an “increased concern of appearance and/or corrosion” the Class 2 exposure condition should be used. Note that these definitions of Classes 1 and 2 reflect the specific tolerance of an element to cracking as opposed to the exposure of an element to harsh environmental conditions. This distinction allows for the consideration of appearance in addition to corrosion. As a point of comparison, a γe of 1.0 corresponds to a crack width of 0.017 in. (0.43 mm) and a γe of 0.75 corresponds to a crack width of 0.013 in. (0.32 mm). The commentary mentions that the formulation above was selected to allow the authority having jurisdiction to be able to tailor the requirements by modifying γe. In addition, the AASHTO LRFD Bridge Design Specification (2017) provide additional crack control requirements for the following four cases: • Side Faces of Girders - for both non-pre-stressed and pre-stressed concrete members deeper than 3 ft., designers are required to provide skin reinforcement over half the depth of the beam (from mid-height to the tension face) with an area of at Ask, given by: 0.012( 30)sk lA d≥ − (14) where: dℓ = distance from the extreme compression fiber to the centroid of extreme tension steel element (in.) The total area of skin reinforcement (per face) need not be greater than 1/4th the flexural tension reinforcement. The maximum spacing is set as either dℓ/6 or 12 in., whichever is smaller.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 62 • Flanges of T-Girders and Box Girders – for flanges of the reinforced T-girders and box girders, the tension reinforcement must be equally spaced across the effective width of the flange or over a width equal to 10% of the average of adjacent span lengths (whichever is smaller). In cases where the effective flange width is greater than 10% of the span length, additional reinforcement of at least 0.4 percent of the excess flange area must be provided to the outer portions of the span. • Shrinkage and Temperature Reinforcement – For concrete members exposed to daily temperature changes (or in mass concrete) the area of temperature and shrinkage reinforcing steel in each face and in each direction is given by: 1.30 2( )s y bhA b h f ≥ + (15) except that: 0.11 0.60sA≤ ≤ (16) where: As = area of reinforcement in each direction and each face (in.2/ft.) b = least width of component section (in.) h = least thickness of component section (in.) fy = specified yield strength of reinforcing bars ≤ 75 ksi The maximum spacing of the temperature and shrinkage steel should be 3 times the component thickness or 18 in. (whichever is smaller) and 12 in. for walls and footings greater than 18 in. thick. • Bridge Decks – For decks designed using the empirical design method, the minimum reinforcement in each direction of the bottom and top layers must be 0.27 in.2/ft. and 0.18 in.2/ft., respectively, with a spacing less than 18 in. For decks designed using the traditional design method, reinforcement in the secondary direction as a percentage of the primary reinforcement is given by: – for primary reinforcement parallel to traffic: 100 / 50percentS ≤ (17) – for primary reinforcement perpendicular to traffic: 220 / 67 percentS ≤ (18) where: S = the effective span length (ft) To limit cracking in pre-stressed concrete elements, AASHTO (2017a) provides a series of tensile stress limitations at the service limit state (see Table 46). These limitations depend on whether the element is segmental or cast integrally, the corrosive environment, and the presence of bonded pre-stressing tendons or reinforcement.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 63 Table 46. AASHTO tensile stress limits in pre-stressed concrete at service limit state after losses (AASHTO 2014). Bridge Type Location Stress Limit Other Than Segmentally Constructed Bridges These limits may be used for normal weight concrete with concrete compressive strengths for use in design up to 15.0 ksi and lightweight concrete up to 10.0 ksi. Tension in the Precompressed Tensile Zone, Assuming Uncracked Conditions • For components with bonded pre-stressing tendons or reinforcement that are subjected to not worse than moderate corrosion conditions 0.19λ ' 0.6(ksi)cf ≤ • For components with bonded pre-stressing tendons or reinforcement that are subjected to severe corrosive conditions 0.0948λ ' 0.3(ksi)cf ≤ • For components with unbonded pre- stressing tendons No tension Segmentally Constructed Bridges These limits may be used for normal weight concrete with concrete compressive strengths for use in design up to 15.0 ksi and lightweight concrete up to 10.0 ksi. Longitudinal Stresses through Joints in the Precompressed Tensile Zone • Joints with minimum bonded auxiliary reinforcement through the joints sufficient to carry the calculated longitudinal tensile force at a stress of 0.5fy; internal tendons or external tendons 0.0948λ ' 0.3(ksi)cf ≤ • Joints without the minimum bonded auxiliary reinforcement through joints No tension Transverse Stresses • Tension in the transverse direction in precompressed tensile zone 0.0948λ ' 0.3(ksi)cf ≤ Stresses in Other Areas • For areas without bonded reinforcement No tension • In areas with bonded reinforcement sufficient to resist the tensile force in the concrete computed assuming an uncracked section, where reinforcement is proportioned using a stress of 0.5fy, not to exceed 30.0 ksi. 0.19λ ' (ksi)cf SHRP 2 Project R19B The SHRP 2 R19B Project (Kulicki et al. 2015) examined the control provisions discussed in the previous section associated with reinforced concrete and pre-stressed concrete flexure members. In the case of reinforced concrete, the R19B Research Team recommended target reliability indices of 1.6 for Class 1 exposure conditions and 1.0 for the Class 2 exposure conditions. The calculation of these values was based upon an ADTT of 5,000. Since these provisions provided uniform reliability across the examined girder spacing, no changes to the AASHTO LRFD Bridge Design Specifications were recommended. In the case of the Service III limit state, which employs the tensile stress limitation shown in Table 35, the results of the R19B project indicated that the live load factor associated with the tensile stress limitations should be increased from 0.8 to 1.0. The need for this modification was traced to a change in the pre-stress loss calculation procedure in 2005. If designers use the time-step method for computing pre-stress losses,

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 64 then this modification is not needed. Table 47 shows the average reliability indices associated with different girder types for two tensile stress limitations. Table 47. Average reliability indices for different types of girders (Kulicki et al. 2015). Type of Section Maximum Tensile Stress Used in Design (ksi) ft = 0.0948√f’c ft = 0.19√f’c I- and bulb-T- girders 1.33 1.00 Adjacent box beams 1.85 1.31 Spread box beams 1.45 1.01 ASBI box beams 1.41 1.00 NCHRP Synthesis 333 Russell (2004) notes: The most important construction practices to achieve a low permeability, uncracked bridge deck with adequate freeze-thaw resistance is to initiate wet curing of the concrete immediately after finishing any portion of the concrete surface and maintaining wet curing for a minimum of 7 days. Other practices that are beneficial include moderate concrete temperatures at time of placement, minimum finishing operations consistent with achieving the desired concrete surface, gradual development of performance specifications, and warranties. GDOT Bridge and Structures Design Manual GDOT (2018) limits the concrete tensile stress to control cracking in pre-stressed concrete beams for bridges in normal or severe exposure categories, as shown in Table 48. Table 48. GDOT concrete stress limits for pre-stressed concrete beams for LRFD projects (GDOT 2018). Stress/Limit State LRFD Reference Limit Stress Formula Conditions Initial Compression before Losses 5.9.4.1.1 = 0.60f’ci (ksi) At Release Initial Tension before Losses Table 5.9.4.1.2-1 = 0.0948√f’ci (ksi) ≤ 0.200 ksi At Release Final Compression, Service I Table 5.9.4.2.1-1 = 0.45f’c (ksi) = 0.60f’c (ksi) Without Transient Load With Transient Load Final Tension, Service III Table 5.9.4.2.2-1 = 0.19√f’c (ksi) = 0.0948√f’c (ksi) Normal Exposure Severe Exposure Notes: 1. f’c and f’ci shall be in the units of ksi for the above equations. 2. Severe exposure criteria shall apply to any bridge over waterways located partially or completely within a coastal county. The coastal counties are Chatham, Bryan, Liberty, McIntosh, Glynn, and Camden. Normal exposure criteria shall be used for all other bridges.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 65 VDOT VDOT (2018a) adopts the AASHTO (2017a) provisions for control of cracking by distribution of reinforcement. PennDOT Design Manual, Part 4 Concrete cracking is controlled by PennDOT (2015) using the AASHTO (2017a) reinforcement spacing requirements with modifications. PennDOT specifies that the concrete cover dc used in Equations 12 and 13 shall not include the 0.5 in. wearing surface for deck slabs, the 0.5 in. extra on the top and bottom slab of box culverts, and the 1 in. extra cover on the bottom mat of footing reinforcement and bottom slab of box culverts that accounts for uneven ground leveling. Concrete Overlays, Waterproofing Membranes, and Sealers A number of common concrete surface protection strategies were identified during the literature review. Due to the variety of protection systems, the following general definitions are provided: • Overlay: a protective layer of material typically a ½ inch or greater applied over a horizontal concrete surface (e.g., bridge deck). Common overlay materials include asphalt, latex modified concrete, low- slump dense concrete, silica fume concrete, or polymer concrete (Russell 2004). • Waterproofing Membrane: “a barrier placed on top of the concrete and then protected by another material that functions as the riding surface” (Russell 2004). • Sealer: a protective material typically applied in a thin layer. Sealers are often placed in two classes – penetrants and coatings, with several additional subclasses sometimes used (Cady 1994). ODOT Bridge Design and Drafting Manual ODOT (2016) specifies five overlay types for use on new and existing bridge decks if a deck overlay is warranted. The five overlay types offered are: • Silica Fume Concrete (SFC) (also referred to as Microsilica Concrete (MC)) • Latex Modified Concrete (LMC) • Multi-Layer Polymer Concrete Overlay (MPCO) • Premixed Polymer Concrete (PPC) • Asphalt Concrete Wearing Surface (ACWS) The following guidance for existing structures is given in ODOT (2016) for a designer to determine if a deck overlay system is warranted: • Bridge deck overlays are not recommended for any of the following conditions: – The deck condition is rated as a 7 or greater. The deck is still in good condition. – Less than 1 percent delaminated, patched, or cracked deck area. The deck is still in good condition. – The deck condition is rated as a 4 or less and investigations confirm severe deck deterioration beyond repair. – Greater than 15 percent delaminated, patched, or cracked deck area and investigations confirm severe deck deterioration beyond repair. – Corrosion has caused extreme deck deterioration or the chloride content exceeds 0.075 percent by mass of sample at the surface or 0.020 percent by mass of sample at the depth of rebar. • Bridge deck overlays are recommended for any of the following conditions: – The deck condition is rated as a 5 or 6. – The deck condition is rated as a 4 or less and investigations confirm the deck deterioration is not beyond repair.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 66 – Greater than 15 percent delaminated, patched, or cracked deck area and investigations confirm the deck deterioration is not beyond repair. – Greater than 5 percent but less than 15 percent delaminated, patched, or cracked deck area. – Greater than 1 percent but less than 5 percent delaminated, patched, or cracked deck area and annual average daily traffic (AADT) at least 3000. – Greater than 1 percent but less than 5 percent delaminated, patched, or cracked deck area and the structure carries interstate highway traffic. – Corrosion has not caused extreme deck deterioration or the chloride content is less than 0.075 percent by mass of sample at the surface and 0.020 percent by mass of sample at the depth of rebar. For structures in marine environments, or if deck rebar corrosion is visible, or if the structure may be salted during the winter months, ODOT requires designers to consult the state “Corrosion Engineer” for the proposed overlay system (ODOT 2016). ODOT does not typically use waterproofing membranes on new structures as new structures are normally constructed without an ACWS (ODOT 2016). Waterproofing membranes are specified for “side-by-side precast concrete slab, box beam, and deck bulb-T construction with ACWS” (ODOT 2016). NCHRP Synthesis 333 NCHRP Synthesis 333 (Russell 2004) reviewed and compared past bridge design and construction approaches aimed at improving the performance bridge decks. This report summarizes the findings in the following: Bridge deck protective systems that are designed to prevent the primary concrete and reinforcement from conditions that will cause their deterioration include overlays, membranes, sealers, and cathodic protection. Latex modified concrete overlays and low-slump dense concrete overlays have, in general, performed satisfactorily. Results with membranes appear to be mixed. In states with more experience, the results have been better. However, the life of the membrane system is limited more by the life of the protective cover over the membrane than the membrane itself. Sealing of concrete surfaces can be used to delay the effects of deterioration if deterioration is not already underway. However, the performance of sealers is difficult to assess because of inconsistencies between laboratory tests and field tests and a lack of national standard testing specifications. Nevertheless, sealers do offer a low initial cost approach. NCHRP Synthesis 209 NCHRP Synthesis 209 (Cady 1994) looked at the existing technical literature on the use of sealers to protect highway bridge decks along with relying on surveys of bridge owners and sealer manufacturers. Sealers were initially used are used to “counteract freezing and thawing and deicer scaling damage to concrete.” Later, as the use of air entraining agents became more common in the mix design of concrete bridge decks, the primary purpose for sealers “changed to preventing or retarding the ingress of chlorides that cause serious damage through corrosion of reinforcing steel. Other, albeit less frequently employed, uses of sealers include providing abrasion resistance (wind, water/ice, and traffic), aesthetics (graffiti- proofing), protection against aggressive chemical agents (e.g., sulfate-bearing groundwater and acid rain), and limiting water ingress to retard deterious reactions (e.g., alkali-silica reactivity).” There are competing interpretations of the terms used to classify various sealers. This report uses three classifications: water-repellent, pore blocker, and barrier coat: • water-repellent – materials that penetrate and coat pore walls making them hydrophobic. • pore blockers – sealers with low viscosity, which allows for pore penetration and sealing while leaving little to no coating on the surface.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 67 • barrier coatings – sealers with high viscosity, which prevents pore penetration but allows for the creation of a surface coating that blocks the pores. Sealer types were grouped as water repellents (alkylalkoxy silane, oligomeric alkylalkoxy siloxane), pore blockers (gum resin in solvent, boiled linseed oil in solvent), pore blockers or barrier coatings (epoxy resin, urethane, acrylic, chlorinated rubber), and dual systems (combination water-repellent and pore blocker; alkylalkoxy silane primer/acrylic top coat). The document lists a number of different properties that are important to choosing a sealer. Table 49 below compares the relative importance of each property for different sealer classifications. Table 49. Importance of sealer properties (Cady 1994). Classification of Sealer Sealer Properties Water- Repellent Pore Blocker Barrier Coating Related to Durability of Sealer a) penetration depth C C NA b) UV resistance U U C c) abrasion resistance U U I d) reactivity with HCP substrate C NA NA e) weathering U U I f) alkali resistance I I I g) bond strength to concrete NA NA C h) flexibility NA NA I i) service life I I I Related to Protection of Concrete a) chloride absorption I I I b) water absorption I I I c) water vapor transmission I S S d) crack bridging NA NA I e) deicer scaling resistance I I I Related to Performance of Concrete a) surface slipperiness (skid resistance) NA NA C Related to Sealer Use a) ease of application I I I b) reapplicability I I I Related to Economics or Aesthetics a) cost I I I b) surface appearance NA U I C = Critical; I = Important; S = Somewhat important; U = Unimportant; NA = does Not Apply Most existing criteria for sealers are based on the use of sealers for existing concrete, however “Oklahoma specifies the bridge elements that may be treated depending on the anticipated deicing chemical exposure, which is defined in terms of traffic use.” Table 50 summarizes the Oklahoma DOT sealer policy.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 68 Table 50. Summary of Oklahoma DOT sealer policy (Cady 1994). Chloride Exposure Bridge Member High Moderate Low Deck Slabs Top of slab and underside of cantilever None Parapets All Faces Roadway Face None Approach Slabs Top None None Reinforced Concrete Bridges: • Major Grade Separation Top and Bottom of Slab Top of Slab None • Other Locations Top of Slab Top of Slab None Box Culverts at Grade Driving Surface and Curbs None Precast Beams: • Major Grade Separation All Exposed Faces of All Beams End 5’ of All Beams; Outside Faces and Bottom of Other Beams None • Other Locations End 5’ of All Beams; Outside Faces and Bottom of Other Beams None Post-Tensioned Box Girders • Top Slab Top and Underside of Cantilever • Other All Exterior Surfaces Pier Caps Top, Sides, End None Abutments All Exposed Areas of Bridge Seat and Front Face of Backwall None Columns None None None Wingwalls None None None Retaining Walls None None None Diaphragms None None None High = bridges on urban expressways and interstate highways; Moderate = bridges on federal and state highways; Low = others In addition, “Alberta Transportation and Utilities and the Department of Transport in Great Britain employ dual requirements in much the same manner as Oklahoma DOT.” They have regulation regarding the use of sealers in terms of half cell potential and age (i.e. existing structures). It is noteworthy; however, that Alberta and Great Britain have suggestions for sealing the same elements: • Piers, columns, beams, and abutments within 8 m (26 ft) of the edge of the roadway; • Piers, columns, beams, and abutments with a deck joint above, but with no provision for positive drainage; • Bearing areas, ballast walls, and deck ends with a deck joint above; • Deck beams and soffits directly over the roadway; • Parts of wingwalls within 8 m (26 ft) of the edge of the roadway; • Retaining walls within 8 m (26 ft) of the edge of the roadway; • Curbs and parapets; and • Concrete wearing surfaces. The report details the performance characteristics of various sealer types based on laboratory and field testing. It should be cautioned that there is high variability within each sealer class in terms of lab

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 69 performance as well as almost no documented correlation with lab performance and performance in the field. Table 51. Ranking of concrete sealers by various types of laboratory tests (Cady 1994). Generic Type Absorption Water Vapor Transmission Chloride Penetration F-T/Deicer Scaling Rebar Corrosion Accelerated Weathering Carbonation Weighted Overall Rating Rank Rating Rank Rating Rank Rating Rank Rating Rank Rating Rank Rating Rank Rating Rank Silane 49 5 42 8 56 5 28 9 55 2 52 5 57 2 48 4 Siloxane 64 3 57 6 59 1 26 11 5 9 82 3 28 5 42 8 Epoxy 40 7 27 10 57 4 73 3 39 6 40 7 - - 44 6 Gum Resin 45 6 57 6 46 7 33 8 72 1 9 10 71 3 50 2 Linseed Oil 17 11 23 11 40 8 77 2 25 7 91 1 - - 33 11 Stearate 38 8 60 3 30 10 26 11 - - 27 8 - - 40 9 Acrylic 28 9 55 7 20 11 56 5 40 4 2 11 29 4 36 10 Silicate 5 12 58 4 9 12 33 8 0 10 82 3 - - 22 12 Urethane 70 2 18 12 49 6 67 4 54 3 16 9 14 6 49 3 Chlorinated Rubber 54 4 31 9 57 4 54 6 39 6 - - - - 47 5 Silicone 28 9 86 1 33 9 7 12 - - 64 4 0 7 42 8 Dual Systems 76 1 73 2 58 2 92 1 11 8 46 6 86 1 62 1 No. of Data Sets Used 15 11 11 5 9 2 1 54 The chloride penetration test results and overall rankings from Table 51 were compared with field tests performed by both Minnesota and Indiana DOT, as shown in Table 52. Rankings and correlation coefficients were compared between the tests. Only the Indiana DOT and the lab overall rankings showed significant correlation at the 95 percent confidence level; however, a comparison of the rankings shows similarities.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 70 Table 52. Comparison of sealer generic class rankings: field tests versus lab tests (Cady 1994). Laboratory Test Rankings (from Table 51) Rank Chloride Penetration Test Overall Ranking Minnesota DOT Field Test 1 Epoxy Dual System* Dual System* 2 Acrylic Epoxy Silane 3 Dual System* Silane Epoxy 4 Silane Acrylic Acrylic 5 Silicate Silicate Silicate Indiana DOT Field Test 1 Dual System** Siloxane Dual System** 2 Gum Resin Dual System** Gum Resin 3 Epoxy Epoxy Silane 4 Siloxane Silane Epoxy 5 Silane Gum Resin Siloxane 6 Stearate Stearate Stearate Spearman’s Rank Correlation Coefficients Lab Tests Field Tests Chloride Penetration Overall Minnesota DOT +0.500 +0.200 Indiana DOT +0.429 +0.829*** *Rankings by generic class. **Penetrating sealer (commonly silane) primer plus topcoat (commonly methylmethacrylate). ***Correlation coefficient significant at 95% level. Additional commentary on the performance data is also provided: • Silane and Siloxane: In the class of water-repellent sealers. Exhibited the widest variation in field testing performance; some of the tested products were the best of all classes as well as the worst of all classes. These sealers work best when applied to relatively high quality, low permeability concrete. • Barrier Coating/Pore-Blocking Sealers (General): These generic classes of sealers can function as either pore blockers or barrier coatings depending on the amount of solids that are contained within the dilutant. This class includes “epoxies, acrylics, urethanes, and chlorinated rubbers.” Water absorption is inversely related for solids content in epoxies and urethanes, while vapor transmission is positively correlated to solids content for epoxies. Acrylics show no such dependence on solids content for either water absorption or vapor transmission. • Linseed oil: A pore-blocking sealer. It has shown effectiveness in combating deicer scaling in concrete, however it has a relatively short period of effectiveness – about 2 years – before reapplication is needed. Linseed oil treated with UV radiation in laboratory tests or applied in the field displays rather high performance, while linseed oil applied in laboratory tests without UV exposure performs poorly. • Silicates: A pore-blocking sealer. Silicates have not performed well in the field, however they have performed well in laboratory tests where accelerated weathering is utilized. • Dual systems: These systems consist of penetrating primer (either water-repellent or pore blocker) with a second coating that consists of either a pore blocker (when the primer is water-repellent) or a barrier coating. These systems appear to perform well due to the “synergistic” effects of combing different sealer types.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 71 Other parameters that must be considered when selecting a sealer include service and recoating. • Service: Severe service refers to surfaces subject to traffic wear as well as direct application of deicing chemicals. "Moderately severe" service pertains to members subject to splashing or spraying of brine by traffic, or drainage of deicer melt water through open deck joints. "Moderate" service relates to areas generally protected from direct deicer application or generally out of the range of direct traffic splash, but occasionally subject to runoff and/or aerosols containing deicing salt. • Recoating: Important for planning future coats. Some sealers may not be reapplied over other types of sealers. Applying “epoxy-, urethane-, acrylic-, or chlorinated rubber-coated surface” prohibits later use of silanes/siloxanes or linseed oil/gum resin. Linseed oil/gum resin prohibits later use of silanes/siloxanes. Epoxies, urethanes, and acrylics “generally” recoat themselves poorly. Chlorinated rubber “generally” re-coats itself well. Sealer service life is primarily determined by the sealer’s ability to slow the influx of chloride ions to the top surface of a concrete bridge deck. “Service life of a concrete sealer relative to chloride ingress is a function of three categories of factors: (1) sealer material properties, (2) service conditions related to sealer durability, and (3) chloride diffusion related factors.” With a sealer in place, diffusion of chlorides into concrete is result of those that ions that have leaked through the sealer. “All sealers are permeable to chloride ions to some degree.” Cady (1994) estimated sealer service life from existing literature, as shown in.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 72 Table 53. Summary of sealer service life data from the technical literature (Cady 1994). Generic Type Conditions Service Life (years) Silanes/Siloxanes Bridge Decks High traffic volumes 4 – 8 General (better sealers) > 10 Piers, Pier Caps, Beams Subject to abrasion < 10 Light-moderate exp 6 – 8.5 Vertical surfaces Indefinitely General 5 – 7 Epoxies Penetrants (<50% solids) General < 2 – 10 Wave or ice action 1 – 3 No wave or ice action < 1 Coatings Moderate exposure conditions 7 – 8 Light exposure conditions < 10 Sea spray and splash 6 – 10 Deicer runoff 10 – 14 Abrasive wear conditions < 1 General 10 – 15 Urethanes Coatings Sea spray and splash 10 – 14 Deicer runoff 14 – 18 General 10 – 15 Acrylics Coatings Moderate exposure conditions 5 – 7 Light exposure conditions 7 Sea spray and splash 9 – 13 Deicer runoff 13 – 17 General 10 to > 15 Boiled Linseed Oil Not exposed to wave or ice action 1 – 3 Exposed to wave or ice action < 1 General 1 – 5 (avg. 2) Dual Systems General > 15 Sealers – General Bridge Deck Sealers 7 (avg.) Median* 4 – 5 Mode* 5 Range* 1 – 25 Interquartile range* 2 – 10 Mean* 16.5 Range* 10 – 25 Mean from lit.* 5 Non-deck – Cl runoff or spray Median* 5 – 10 Mode* 10 Range* 2 – 25+** Interquartile range* 3 – 10 *Questionnaire Data **Chloride spray: 2 – 35 See Cady (1994) for References Steps to calculate sealer service life are also provided and require laboratory testing data. The procedure is as follows:

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 73 1. Estimate field environmental effective chloride diffusion constant, Dc, using field data or reference values (Table 54). Table 54. Range of equilibrium chloride content and mean diffusion constant for 10 states (Cady 1994). State Range of Equil. Chloride Content Co (kg/m3) Mean Diffusion Constant Dc (cm2/yr) California 0.0 – 2.3 1.61 Delaware 4.7 – 5.8 0.32 Florida 2.4 – 4.6 2.13 Indiana 4.7 – 5.8 0.58 Iowa 4.7 – 5.8 0.32 Kansas 0.0 – 2.3 0.77 Minnesota 2.4 – 4.6 0.32 New York 5.9 – 8.8 0.84 West Virginia 4.7 – 5.8 0.45 Wisconsin 5.9 – 8.8 0.71 Calculate estimated average equilibrium chloride concentration level, C′0, resulting from chloride leakage that is just sufficient to produce the corrosion threshold chloride content at the surface of the reinforcing steel C(x,t) by the end of the expected useful life of the structure, t. ( , ) 0' 1 erf 2 x t c C C x D t =   −     (19) where: erf = the “error function”. x = depth of cover of the shallowest 2.5 percent of the reinforcing steel. 2. Calculate the estimated equivalent field time teq which produces a chloride-ion concentration C′(x,t) in 30 weeks of ponding untreated (control) specimens using the previously estimated field environmental effective diffusion constant Dc and the equilibrium chloride content C0-1c from the 30- week ponding test, i.e. solve following equation for teq ( , ) 0 1 '' 1 2 x t cc eq Cxerf CD t −       = −       (20) 3. Assuming linear increase in equilibrium concentration versus time, the total allowable equilibrium concentration at time t (the service life of the structure), C0-total, equals two times the average equilibrium chloride concentration C′0, i.e., 0 02 'totalC C− = (21) 4. Calculate the average allowable equilibrium concentration, C0-eq

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 74 0 0 eq eq total t C C t− −   =     (22) 5. Determine the laboratory leakage factor, LR, from the ratio of the equilibrium chloride concentration (at 1.27 cm depth) in the sealed and unsealed 30-week ponding tests 0 1 0 1 100 (%)s c CLR C − −   =     (23) where: C0-1s = 30-week ponding test equilibrium chloride concentration at 1.27 cm depth for the sealed specimens C0-1c = 30-week ponding test equilibrium chloride concentration at 1.27 cm depth for the control (unsealed) specimens 6. Determine the allowable leakage factor, LRallowed, as follows 0 1 0 1 100 (%)sallowed c CLR C − −   =     (24) 7. Determine the estimated life of the sealer (i.e., reapplication period), tsl, from the relationship: allowed sl eq LRt t LR  =     (25) NCHRP 209 provides tables of values of the error functions used in Equations 19 and 20. NCHRP Synthesis 425 NCHRP Synthesis 425 (Russell 2012) provides information on waterproofing membranes for concrete bridge decks, including material performance, specification requirements, and costs. Waterproofing membranes generally consist of either 1) a constructed-in-place system or 2) a preformed system. Constructed-in-place systems may consist of either a bituminous membrane or sprayed resinous liquid. Preformed membranes may be made out of asphalt-impregnated fabric, or polymer-, elastomer-, or asphalt- laminated boards. These waterproofing systems also may have an additional adhesive layer or layer to improve the riding surface. European agencies have consistently reported success in using waterproofing membranes to combat deck deterioration, however use in the United States remains limited. Previous NCHRP projects have noted that while the frequency of use over the past 40 years has been variable in the United States, owners have reported success. Both AASHTO and states have their own specifications for the use of waterproofing membranes. Specifications generally follow the following format: 1) Pre-installation, 2) Surface preparation, 3) Installation, 4) QA/QC. AASHTO has specifications for both constructed-in-place asphalt membranes a well as preformed membrane systems. States have various specifications for waterproofing membranes, some with more details, and other with fewer. The number and type of membranes varies among states, as well. Massachusetts, for example, allows three types of membrane systems, while Virginia DOT specifications permit five types. A comparison of state and AASHTO requirements is shown in Table 55.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 75 Table 55. Summary of state specification membrane requirements (Russell 2012). Property AASHTO States Minimum thickness for rubberized asphalt, mil. 65 50 and 60 Minimum thickness for modified bitumen, mil. 70 50 and 60 Minimum deck or air temperature, °F 35 40, 45, and 50 Puncture resistance, lb — 40 and 200 Maximum permeance, perms — 0.10 Minimum longitudinal overlap, in. 2.0 2.0, 2.5, 3.0, 4.0, and 6.0 — = Not specified. Both AASHTO and the states generally refer to ASTM standards for materials specifications and test methods. Table 56. ASTM Standards related to waterproofing membranes (Russell 2012). ASTM Designation Title D5 Standard Test Method for Penetration of Bituminous Materials D36/D36M Standard Test Method for Softening Point of Bitumen (Ring-and-Ball Apparatus) D41/D41M Standard Specification for Asphalt Primer Used in Roofing, Dampproofing, and Waterproofing D146 Standard Test Methods for Sampling and Testing Bitumen-Saturated Felts and Woven Fabrics for Roofing and Waterproofing D173 Standard Specification for Bitumen-Saturated Cotton Fabrics Used in Roofing and Waterproofing D449 Standard Specification for Asphalt Used in Dampproofing and Waterproofing D517 Standard Specification for Asphalt Plank D882 Standard Test Method for Tensile Properties of Thin Plastic Sheeting D1228 Withdrawn Standard Methods of Testing Asphalt Insulating Siding Surfaced with Mineral Granules (Withdrawn 1982) D1668 Standard Specification for Glass Fabrics (Woven and Treated) for Roofing and Waterproofing D1777 Standard Test Method for Thickness of Textile Materials D3236 Standard Test Method for Apparent Viscosity of Hot Melt Adhesives and Coating Materials D3515 Historical Standard Standard Specification for Hot-Mixed, Hot-Laid Bituminous Paving Materials D4071 Standard Practice for Use of Portland Cement Concrete Bridge Deck Water Barrier Membrane System D4541 Standard Test Method for Pull-Off Strength of Coatings Using Portable Adhesion Testers D4632 Standard Test Method for Grab Breaking Load and Elongation of Geotextiles D4787 Standard Practice for Continuity Verification of Liquid or Sheet Linings Applied to Concrete Substrates D6153 Standard Specification for Materials for Bridge Deck Waterproofing Membrane Systems D6690 Standard Specification for Joint and Crack Sealants, Hot Applied, for Concrete and Asphalt Pavements E96/E96M Standard Test Methods for Water Vapor Transmission of Materials E154 Standard Test Method for Water Vapor Retarders Used in Contact with Earth Under Concrete Slabs, on Walls, or as Ground Cover

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 76 Six out of nine Canadian provinces have specifications for the use of waterproofing membranes, with the Ontario provincial specification OPSS 914, Construction Specifications for Waterproofing Bridge Decks with Hot Applied Asphalt Membranes (OPSS 2009) providing the most details. The major differences between U.S. and Canadian standards are: • Canadian specifications generally require the use of hot applied rubberized asphalt; U.S. specifications permit other types of membranes. • Some Canadian specifications require the installation of rubber membranes or reinforcing fabric over cracks and joints before asphalt membrane application. • Most Canadian specifications require the use of protection board on top of the waterproofing membrane. The use of waterproofing membranes in the United Kingdom is mandatory for integral abutment and continuous bridges. The design, materials, and workmanship for waterproofing and surfacing concrete decks for highway bridges follow BD47/99, Waterproofing and Surfacing of Concrete Bridge Decks (UKDOT 1999). All highway bridges are required to have adequate protective measures to prevent water from coming into direct contact with the bridge deck. Besides adequate drainage, waterproofing the upper surface of the deck is required. Waterproofing systems “are required to have a British Board of Agreement Roads and Bridges Agreement Certification” or European equivalent, which requires both laboratory tests and site trials. Within the synthesis, the following performance criteria are identified for ideal waterproof membrane systems (Sohanghpurwala 2006): • Impermeability to water, • Good adhesion to the deck and riding surface, • Tolerant to deck surface roughness, • Resistant to traffic prior to riding surface application, • Ability to bridge cracks in the deck or joints between adjacent precast members, • Safe to apply and with low volatile emissions, • Ability to withstand high and low temperatures and large temperature ranges, • Extended service life of 50 to 100 years. Conversely, the limitations include: • The membrane service life may be limited by the wearing surface service life. • The system is not suitable for grades greater than 4% due to limited bond capacity of certain systems, creating the possibility for shoving and debonding. Performance data on membranes in the U.S. is limited to state surveys and a few small-scale tests: (1) A study in Colorado reviewed 16 bridges built from 1958 to 1985 (Xi et al. 2004). Performance was varied and the report is limited as the age when membrane system was installed is unknown; (2) A study in Kansas reviewed two bridges that were restored with membranes after approximately 15 years of service (Distlehorst 2009); (3) Another report reviewed six bridges that had membranes installed after 20-25 years of service in the late 60’s and early 70’s (Wojakowski and Hossain 1995). The lack of studies shows that there is little current empirical evidence in the United States on membrane performance, especially on the use of waterproofing membranes on new bridge decks. In terms of cost, Kepler et al. (2000) found hot rubberized asphalt membrane to be between the second- lowest and sixth-lowest cost strategy out of 33 evaluated protection systems, depending on the discount rate. The analysis used a service life of 75 years and assumed replacement of the top 1.6 inches of asphalt overlay at 20 and 60 years, and replacement of both the membrane and asphalt overlay at 40 years.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 77 Hearn and Xi (2007) compared the service life costs of four types of reinforcement protection using the history of 82 bridges built between 1969 and 1991 using both annualized cost and present value measures: • Uncoated reinforcing bars with rigid overlay • Epoxy coated reinforcing bars and a concrete surface sealer • Uncoated reinforcing bars protected with a waterproofing membrane and bituminous overlay • Epoxy coated reinforcing bars protected with a waterproofing membrane and bituminous overlay They concluded that decks with waterproofing membranes were the least expensive. Distlehorst (2009) compared the cost of retrofit epoxy coated reinforcement, Iowa system overlays, Kansas system overlays, and membrane overlays, and concluded that the membrane overlays were the most cost-effective rehabilitation technique, with an average cost of $0.12/ft2 /year of service life (based on 1979 dollars). A survey performed as part of this synthesis of U.S. and Canadian agencies returned extremely variable costs of membranes including labor, equipment, materials. Costs were reported at $0.56 to $42.80 per ft2 for the U.S. and at $1.69 to $8.55 per ft2 for Canada (in Canadian dollars). PennDOT Design Manual, Part 4 PennDOT (2015) does not recommend overlays for new bridge decks. Protective sealants are specified for decks that are poured between Fall and Winter months (September through February). Concrete sealants for substructures are also specified. For buried structures, specifically precast box culverts, a waterproof sealer is required between the joints of adjacent sections. Structural Steel Coatings Coating systems for structural steel can be defined by specifying three items: surface preparation, coating material/application method, and dried final thickness (DFT) of the coating (Helsel and Lanterman 2018). The following bullets summarize conventional coating systems by providing common examples of each of these three items: • Surface preparation – hand/power tool cleaning or abrasive blast cleaning • Coating material – paint, hot dipped galvanizing, metalizing, or weathering steel – Paint – multi-layer systems that generally consist of a primer, intermediate coat, and top coat, which serves to both shed water and meet aesthetic requirements (SHRP R19A). – Hot dipped galvanizing – consists of layers of zinc-iron alloy that are metallurgically bonded to the steel substrate through the submersion in a molten zinc bath. This system provides galvanic protection to the steel substrate. – Metalizing – application of layers of zinc and/or aluminum applied to the steel substrate by “thermal spray” and typically utilizes a sealer coat since the thermal spray leaves a porous surface (Helsel and Lanterman 2018). This system provides galvanic protection to the steel substrate. – Weathering steel – the addition of small amounts of copper, phosphorus, chromium, nickel, and silicon to the structural steel, which causes it to form a “patina” of tight corrosion products that, in the correct environment, can greatly reduce the rate of corrosion. • DFT – typically ranges from a low of 4 mils for two-stage alkyd systems up to 20 mils for certain polyurethane systems (Helsel and Lanterman 2018).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 78 The guidance related to coatings, in general, is more qualitative in nature than the guidance discussed previously for concrete. Notable exceptions that provide more quantitative estimates of service life do exist (e.g. Helsel and Lanterman (2018)) but current codes and standards have yet to adopt this quantitative guidance. Rather, current code-based guidance generally focuses on (a) prescribing specific coatings for various exposure classes, (b) the estimation of section loss in areas that either cannot be accessed or cannot be reliably protected from corrosion, (c) accessibility to provide access to inspect, repair, etc., and (d) geometry requirements to avoid “corrosion traps” and provide for drainage. The following subsection summarizes the guidance provided by various codes/standards and technical reports related to these items. NACE Helsel and Lanterman (2018) provide service life and cost estimates for several coating systems and a methodology to estimate the life cycle cost to support the selection of an appropriate coating system. This information (and paper) is updated periodically and republished in the NACE conference proceedings. To structure the discussion, the authors note that there are four general stages within the lifecycle of a coating system: • Original application of paint • Spot repair and touch-up • Spot priming and application of a full recoat • Total coating removal and replacement To support life cycle cost analysis, one would ideally want to estimate the time associated with each of these stages, but this is difficult since it depends greatly on the spatial distribution of the coating breakdown. As a result, Helsel and Lanterman (2018) take a different approach and define the end of “service life for practice maintenance” as the point when “5-10% coating breakdown occurs (SSPC – Vis 2 Rust Grade 4) and active rusting of the substrate is present.” The authors point out that this definition does not necessarily correspond to the stages summarized above, since it does not address the spatial distribution of the coating breakdown. For example, in cases were the 5-10% breakdown occurs in a local region, maintenance painting as opposed to total coating replacement would be appropriate. On the other hand, if the coating breakdown is distributed uniformly across an entire structure, it is unlikely that anything but the complete coating removal and replacement would be feasible. Notwithstanding this shortcoming, the definition is useful as it offers a means to define service life and thus quantify and compare various coatings. In order to effectively estimate the service lives of coatings based on this definition, Helsel and Lanterman (2018) adopt some of the exposure classes defined by ISO 12944-2. In particular, for “atmospheric” exposure, four of the six exposure classes identified in ISO 12944-2 (2017) are adopted: • Low - Atmospheres with low levels of pollution; mostly rural areas (corresponds to Environmental Class C2 from ISO 12944-2) • Medium - Urban and industrial atmospheres, moderate sulfur dioxide pollution; coastal areas with low salinity (corresponds to Environmental Class C3 from ISO 12944-2) • Very High, Industry - Industrial areas with high humidity and aggressive atmosphere (corresponds to Environmental Class C5-I from ISO 12944-2) • Very High, Marine - Coastal and offshore areas with high salinity (corresponds to Environmental Class C5-M from ISO 12944-2)

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 79 In addition, for elements that are immersed in water throughout their service life, three exposure classes are offered associated with contact to potable water, fresh water, and salt water. Table 57 provides the estimated service life of a number of common coating systems for atmospheric exposure conditions. As apparent from this table, NACE distinguishes coating systems by five items: type, material for primer/intermediate/top coat, surface preparation, number of coats, and DFT. Again, the service life estimates provided by this table for each exposure class correspond to an estimate of 5-10% coating breakdown (and not necessarily the point where they must be replaced). Similarly, Table 58 provides the estimates service life of a number of coatings deemed appropriate for immersion during their service life.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 80 Table 57. Estimated service life for coating systems with atmospheric exposure (Helsel and Lanterman 2018). Coating System No. Type Coating Systems for Atmospheric Exposure (primer/midcoat/topcoat) Surface Preparation N um be r o f C oa ts D FT M in im um (m ils ) Practical Maintenance Time M ild (r ur al )/C 2 M od er at e (i nd us tri al )/C 3 S ev er e (h ea vy ) in du st ria l)/ C 5- I S ea co as t H ea vy In du st ria l/C 5- M 1 Acrylic Acrylic Waterborne/Acrylic WB/Acrylic WB Hand/Power 3 6 12 8 5 5 2 Acrylic Acrylic Waterborne/Acrylic WB/Acrylic WB Blast 3 6 17 12 8 8 3 Alkyd Alkyd/AlkydAlkyd Hand/Power 3 6 12 8 5 5 4 Alkyd Alkyd/Alkyd/Alkyd (AWWA OCS-1C) Blast 3 6 17 12 8 8 5 Alkyd Alkyd/Alkyd/Urethane Alkyd Blast 3 6 18 13 9 9 6 Alkyd Alkyd/Alkyd/Silicone Alkyd (AWWA OCS-1D) Blast 3 6 20 14 10 10 7 Epoxy Surface Tolerant Epoxy (STE) Hand/Power 1 5 11 6 4 4 8 Epoxy Surface Tolerant Epoxy/STE Hand/Power 2 10 17 12 9 9 9 Epoxy Surface Tolerant Epoxy/STE Blast 2 10 21 15 12 12 10 Epoxy Surface Tolerant Epoxy/Polyurethane Hand/Power 2 7 17 12 9 9 11 Epoxy Surface Tolerant Epoxy/Polyurethane Blast 2 7 21 15 12 12 12 Epoxy Surf-Tolerant Epoxy/STE/Polyurethane Hand/Power 3 12 21 15 12 12 13 Epoxy Surf-Tolerant Epoxy/STE/Polyurethane Blast 3 12 24 18 14 14 14 Epoxy Epoxy 100% Sol Pent Sealer/Epoxy Hand/Power 2 6 13 8 5 5 15 Epoxy Epoxy 100% Solids Penetrating Sealer/Polyurethane Hand/Power 2 4 12 7 4 4 16 Epoxy Epoxy 100% Solids Penetrating Sealer/Epoxy/Polyurethane Hand/Power 3 8 17 12 9 9 17 Epoxy Epoxy/Epoxy Blast 2 8 20 14 11 11 18 Epoxy Epoxy/Epoxy/Epoxy Blast 3 12 23 17 14 14 19 Epoxy Epoxy/Polyurethane Blast 2 6 20 14 11 11 20 Epoxy Epoxy/Polysiloxane Blast 2 9 22 16 12 12 21 Epoxy Epoxy/Epoxy/Polyurethane (AWWA OCS-5) Blast 3 10 23 17 13 13 22 Epoxy Epoxy Waterborne/Epoxy WB/ Epoxy WB Blast 3 9 20 14 11 11 23 Epoxy Epoxy Waterborne/Epoxy WB/ Polyurethane WB (AWWA OCS-7) Blast 3 9 21 15 12 12

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 81 24 Epoxy Zinc Epoxy Zinc/Epoxy Blast 2 7 24 17 11 11 25 Epoxy Zinc Epoxy Zinc/Epoxy/Epoxy Blast 3 11 29 20 14 14 26 Epoxy Zinc Epoxy Zinc/Polyurethane Blast 2 6 24 17 11 11 27 Epoxy Zinc Epoxy Zinc/Epoxy/Polyurethane Blast 3 10 29 20 14 14 28 Epoxy Zinc Epoxy Zinc/Epoxy/Fluorinated Polyurethane Blast 3 10 34 24 18 18 29 Organic Zinc Organic Zinc/Acrylic Waterborne/ Acrylic WB (AWWA OCS-3) Blast 3 7 21 15 11 11 30 Organic Zinc Organic Zinc/Epoxy/Polyurethane (AWWA OCS-6) Blast 3 10 26 18 13 13 31 Organic Zinc Organic Zinc/Polysiloxane Blast 2 8 29 19 14 14 32 Organic Zinc Organic Zinc/Epoxy/Polysiloxane Blast 3 12 30 21 15 15 33 Organic Zinc Organic Zinc/Polyaspartic Blast 2 8 26 18 13 13

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 82 Table 58. Estimated service life for coating systems for immersive exposure (Helsel and Lanterman 2018). Coating System No. Type Coating Systems for Atmospheric Exposure (primer/midcoat/topcoat) Surface Preparation N um be r o f C oa ts D FT M in im um (m ils ) Practical Maintenance Time P ot ab le W at er F re sh W at er Im m er si on S al t W at er Im m er si on 1B Epoxy Coal Tar Epoxy Blast 2 16 - 17 14 2B Epoxy Epoxy/Epoxy Blast 2 8 12 9 8 3B Epoxy Epoxy/Epoxy (AWWA ICS-1) Blast 2 6 10 8 6 4B Epoxy Epoxy/Epoxy/Epoxy (AWWA ICS-2) Blast 3 12 15 12 11 5B Epoxy Epoxy 100% Solids (AWWA ICS-3) Blast 1 20 18 16 14 6B Organic Zinc/ Epoxy Organic Zinc/Epoxy/Epoxy (AWWA ICS-5) Blast 3 10 16 13 12 7B Epoxy Phenolic Epoxy Phenolic/ Epoxy Phenolic Blast 2 12 - 14 12 8B Metallizing Metallizing/Epoxy Blast 2 9 20 17 15 9B Metallizing Metallizing/Epoxy/Epoxy Blast 3 13 24 20 18 10B Misc Polyurethane 100% Solids (AWWA ICS-4) Blast 1 25 18 16 14 11B Misc Vinyl Ester/Vinyl Ester Blast 2 20 - 14 12 12B Misc Polyester (composite filled) Blast 2 25 - 14 12 13B Misc Polyurea Blast 1 25 18 16 14 In addition to these estimates of service life, Helsel and Lanterman (2018) present a series of studies that examine the costs of shop versus field applied paint, paint versus galvanizing, and total maintenance cost for paint systems. A general life cycle cost analysis methodology that computes the net present value for each candidate coating system is also provided. To permit this calculation, a series of unit costs associated with each coating system (inclusive of surface preparation) are provided. Although multipliers to reduce unit costs based on the total amount of coating required are provided, it is unclear from the report how the unit costs were obtained or the potential regional variability of such costs. Eurocode The Eurocode does not prescribe specific coating systems based on exposure class, but rather provides a set of general guidance related to corrosion protection of structural steel. For example, the following provision, which was taken from EN 1993-2 (2006) Section 7.11, summarizes the guidance related to corrosion protection of structural steel: All parts should normally be designed to be accessible for inspection, cleaning and painting. Where such access is not possible, all inaccessible parts should either be effectively sealed against corrosion (e.g. interior of boxes or hollow portions) or they should be constructed in steel with improved atmospheric corrosion resistance. Where the environment or access provisions are such that corrosion can occur during the life of the bridge, a suitable allowance for this should be made in the proportioning of the parts.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 83 In addition to this general guidance, EN 1994-2 (2005), which addresses the design of composite steel and concrete structures, requires that steel flanges supporting concrete decks should “extend into the steel- concrete interface at least 50 mm.” Eurocode also references ISO 12944-3, which provides more specific guidance related to detailing steel structures for accessibility and drainage. Australian Bridge Design Code AS 5100.5 (2004) provides general guidance related accessibility of steel details, drainage, sealing, narrow gaps, and section loss. The following is a summary of the guidance provided. • General – Steel surfaces should be accessible for inspection, cleaning, and painting, or should be effectively sealed against corrosion. When this is not possible, a protective coating system should be applied to the surface coupled with an additional steel thickness, or a corrosion resistant steel should be used. • Drainage – Provide drainage wherever water may collect. Design the drainage to carry and discharge water below adjacent parts of the structure. • Sealing – Seal inaccessible box members and other hollow sections against corrosion. Provide internal corrosion protection unless provisions are taken to ensure these members are airtight. • Narrow Gaps and Spaces – To allow for inspection and maintenance, the clear space between parts not in contact should not be less than one-sixth of the width of the face of the smaller part, or 10 mm, whichever is the greater. Alternatively, steel packing or sealant may be used to fill the space. • Inaccessible Steel Surfaces – As described in the first bullet, additional thickness should be provided for steel surfaces not effectively sealed against corrosion. For a design life of 100 years, the following values may be used as a guide for extra thickness for main members: – 6 mm for industrial or marine sites. – 4 mm for all other inland sites. – 1 mm where free drainage cannot be ensured, in addition to the excess given in the first two bullets above. • The importance, criticality, and inspectability of members should be accounted for when determining the extra thickness. Sealed hollow sections should not be less than 5 mm thick, or 4mm thick for such sections with a design life not greater than 50 years. For steel piles in contact with soil, fresh water or sail water, corrosion allowance is in accordance with AS 5100.3. In addition to this general guidance, AS 5100.5 also provides the following quantitative estimates of section loss for unprotected steel in various environments: • 1.5 mm total for the life of the structure for each face in contact with soil, above and below ground water, provided the soil is undisturbed or comprises compacted, well-graded, chemically neutral, structural fill. • 0.025 mm per year for each face in contact with open-graded or rubble fill, or sands and gravels that have moving ground water. • 0.05 mm per year for each face exposed to fresh water and not in contact with soil. • 0.08 mm per year for each face exposed to seawater, except in the splash zone where twice this rate shall be used. Canadian Standards Association CSA (2014) prescribes coating types for specific elements within different environmental exposure conditions. In particular, CSA defines three primary exposure classes: “No direct chlorides”, “Airborne chlorides or light industrial atmosphere”, and “Heavy industrial atmosphere”. Within each class there are three levels within each class: “Wet, rarely dry”, “Dry, rarely wet”, and “Cyclic wet/dry”. An additional

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 84 exposure class, “Marine”, is defined as a standalone class. The required coating systems as per CSA for superstructure and other components for these exposure classes are provided in Table 59 and Table 60, respectively. In addition, all superstructure elements within 3000 mm of the ends of girders and all girder areas subjected to runoff form the deck must be coated, even in cases where Table 59 indicates uncoated weathering steel is permissible.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 85 Table 59. CSA required corrosion protection for superstructure components (CSA 2014). Component Environmental exposure condition No direct chlorides Airborne chlorides or light industrial atmosphere Heavy industrial atmosphere Marine Wet, rarely dry Dry, rarely wet Cyclical wet/dry Wet, rarely dry Dry, rarely wet Cyclical wet/dry Wet, rarely dry Dry, rarely wet Cyclical wet/dry All superstructures (minimum) Coat Uncoated weathering steel Uncoated weathering steel Coat Uncoated weathering steel Uncoated weathering steel Coat Investigate Investigate Coat Structure with clearance of less than 3 m over stagnant water or less than 1.5 m over fresh water Coat Coat Coat Coat Coat Coat Coat Coat Coat Coat Structure over depressed roadways with tunnel effect Coat Coat Coat Coat Coat Coat Coat Coat Coat Coat Open grid decks Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Structure supporting open grid decks Coat Coat Coat Coat Coat Coat Coat Coat Coat Coat Faying surfaces of joints — — — — — — — — — — Cables, ropes, and strands Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 86 Table 60. CSA required corrosion protection for other components (CSA 2014). Component Environmental exposure condition No direct chlorides Airborne chlorides or light industrial atmosphere Heavy industrial atmosphere Marine or deicing runoff In fresh water In ground water Wet, rarely dry Dry, rarely wet Cyclical wet/dry Wet, rarely dry Dry, rarely wet Cyclical wet/dry Wet, rarely dry Dry, rarely wet Cyclical wet/dry Substructures Coat Uncoated weathering steel Uncoated weathering steel Coat Uncoated weathering steel Uncoated weathering steel Coat Investigate site conditions Investigate site conditions Coat Uncoated Uncoated Sheet piling Coat or increase section thickness Uncoated Coat or increase section thickness Coat or increase section thickness Uncoated Coat or increase section thickness Coat or increase section thickness Coat or increase section thickness Coat or increase section thickness Coat or increase section thickness Uncoated Uncoated Light poles, luminaires, and sign support structures Galvanize Uncoated weathering steel Galvanize Galvanize Galvanize Galvanize Investigate site conditions Investigate site conditions Investigate site conditions Investigate site conditions — — Deck drains Galvanize Uncoated weathering steel Uncoated weathering steel Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize — — Expansion joints Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize Galvanize or metallize — — Bearings (excluding stainless steel and faying surfaces) Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat Galvanize, metallize, or coat — — Faying surfaces of bearing assemblies (excluding stainless steel and Teflon®) Coat Coat Coat Coat Coat Coat Coat Coat Coat Coat — — Moving components or rockers, roller bearings, and pins Grease Grease Grease Grease Grease Grease Grease Grease Grease Grease — — Railings Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize — — Utility supports and hardware Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat Galvanize or epoxy coat — — Components of mechanically stabilized earth structures, bin walls, and gabions Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize Galvanize

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 87 From a detailing standpoint, CSA provides general guidance related to the avoidance of corrosion traps and drainage. Specifically, CSA requires the following: • details that provide free air circulation for all above ground components • details that minimize exposed surface area and are free from pockets, crevices, recesses, reentrant corners, and other locations that collect and retain water, debris, and moisture. Further, areas that will be inaccessible following construction must be marked on plans and provided with a “protective coating” before construction. Additional guidance related to the use of drip bars on the bottom flanges of girders near expansion joints, the orientation of angles and tees such that their vertical leg/web extends downward, and detailing to minimize the effects of salt spray for bridges that cross roadways with a 70 km/h or greater speed limit is also provided. AASHTO AASHTO (2017a) provides limited guidance related to protection of structural steel from corrosion. General guidance provided states that steel “shall be self-protecting, or have long life coating systems or cathodic protection”. From a drainage perspective, AASHTO (2017a) requires that cavities within structures that are likely to retain water must be drained at the lower point. Within the foundations section of AASHTO (2017a), additional guidance related to the corrosion protection of steel piles is also provided. This guidance focuses primarily on identifying the soil conditions for which corrosion is likely to occur. Designers are required to consider corrosion of steel pile foundations located in fill soils, soils with low pH values, and marine environments. Quantitative metrics for defining these conditions is provided as follows: • Soil conditions indicative of potential pile corrosion – Resistivity less than 2,000 ohm-cm, – pH less than 5.5, – pH between 5.5 and 8.5 in soils with high organic content, – Sulfate concentrations greater than 1,000 ppm, – Landfills and cinder fills, – Soils subject to mine or industrial drainage, – Areas with a mixture of high resistivity soils and low resistivity high alkaline soils, and – Insects (wood piles). • Water conditions indicative of potential pile corrosion – Chloride content greater than 500 ppm, – Sulfate concentration greater than 500 ppm, – Mine or industrial runoff, – High organic content, – pH less than 5.5, – Marine borers, and – Piles exposed to wet/dry cycles. FDOT Structures Manual FDOT (2017) specifies the use of weathering steel for new I-girder and box girder bridges where the site conditions permit. The environmental exposure condition for which weathering steel is permitted is summarized in the following:

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 88 • Structures four or more miles from the coast, regardless of superstructure exposure class, with vertical and horizontal clearances that meet the following criteria: – For structures over water  Minimum 12 ft. vertical clearance over mean or normal high water for water with chloride concentrations less than 6000 ppm  Minimum 25 ft. vertical clearance over mean or normal high water for water with chloride concentrations above 6000 ppm – For structures adjacent to water  25 ft. minimum horizontal clearance for water with chloride concentrations less than 6000 ppm  100 ft. minimum horizontal clearance for water with chloride concentrations greater than 6000 ppm • For structures within four miles of the coast, weathering steel may be considered if all of the following criteria are met: – The maximum airborne salt deposition rate is less than 5 mg/m2/day (as determined by ASTM Test G140 over a 30 day period). – The maximum average concentration for SO2 is less than 60 mg/m2/day (as determined by ASTM Test G91 over a 30 day period). – The yearly average Time of Wetness (TOW) does not exceed 60% (as determined by ASTM Test G84). For environmental conditions that do not meet the above criteria, inorganic zinc coating systems are specified. In these cases the use of “high performance coating systems” may also be used but they require written approval from the Chief Engineer. Further, FDOT requires special consideration for coating systems on bridges that are located (a) where the pH of the rainfall or condensation is less than 4 or greater than 10, (b) in areas subjected to salt spray or salt laden runoff, and (c) in areas subjected to concentrated pollution. In addition to the type of coating to be used, FDOT also guides designer to specify details that minimize the retention of water and to use box girders over plate-girders in extremely aggressive environments. ODOT Bridge Design and Drafting Manual ODOT (2016) also specifies the use of weathering steel where site conditions permit, advising caution for specific environmental and location conditions. ODOT (2016) advises caution for the environmental and location conditions outlined in FHWA Technical Advisory T 5140.22 (1989) on the use of uncoated weathering steel (see later section for discussion). In addition, ODOT recommends the designer review FHWA (1989) for recommended detailing practices. Other coating systems are discouraged due to the maintenance costs of recoating. Weathering steel and galvanizing are stated as the preferred options although ODOT does permit coating systems when weathering steel or galvanizing is not applicable. GDOT Bridge and Structures Design Manual GDOT (2017) specifies that all new structural steel shall be painted regardless of the bridge location in the state. ISO ISO 12944-3 (1998) provides guidance related to the protective coating of steel elements with paint to avoid premature corrosion and degradation of the paint system. This standard guides designers to minimize irregularities (corners, overlaps, narrow gaps, lap joints, blind crevices, etc.) which may act as “corrosion traps”. In addition, from a corrosion standpoint, welded connections are preferable to bolted or riveted connections due to the smoothness of their surface. In

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 89 addition, they state that discontinuous welds and spot welds should be avoided unless the corrosion risk in the vicinity is negligible. The standard goes on to state that the protective paint system should be accessible to allow for application, inspection, and maintenance. For example, notches in stiffeners should be large enough (at least 50 mm) to permit their painting. In cases where it is not accessible, the paint system should be designed for the service life of the structure. Structures should be designed to promote drainage and avoid areas that may trap water. In particular, it is recommended that surfaces be inclined for chamfered, that open sections along the top of their arrangement be avoided, and pockets/recesses should be avoided. FHWA Weathering Steel Technical Advisory The FHWA, in Technical Advisory 5140.22 (1989), provides engineers with guidance on the proper application of uncoated weathering steel in highway structures. The FHWA advises caution in using weathering steel when a structure will be subjected to the following environmental- or location-related effects: • Environment – Marine coastal areas. – High rainfall, high humidity, or persistent fog. – Industrial Areas where concentrated chemical fumes may contact the structure. • Location – Grade separations in tunnel-like conditions. – Low-level water crossings that are:  Ten feet or less over stagnant, sheltered water.  Eight feet or less over moving water. FHWA also suggests specific design details and maintenance actions when using weathering steel in highway structures: • Design Details. For uncoated steel in bridges and other highway structures, the following items should receive careful consideration: – Eliminate bridge joints where possible. – Expansion joints must be able to control water that is on the deck. Consider the use of a trough under the deck joint to divert water away from vulnerable elements. – Paint all superstructure steel within a distance of 1 1/2 times the depth of girder from bridge joints. – Do not use welded drip bars where fatigue stresses may be critical. – Minimize the number of bridge deck scuppers. – Eliminate details that serve as water and debris "traps". – "Hermetically seal" box members when possible, or provide weep holes to allow proper drainage and circulation of air. – Cover or screen all openings in boxes that are not sealed. – Consider protecting pier caps and abutment walls to minimize staining. – Seal overlapping surfaces exposed to water (to prevent capillary penetration action). • Maintenance Actions – Implement maintenance and inspection procedures designed to detect and minimize corrosion. – Control roadway drainage:

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 90 – Divert roadway drainage away from the bridge structure. – Clean troughs or reseal deck joints. – Maintain deck drainage systems. – Periodically clean and, when needed, repaint all steel within a minimum distance of 1 1/2 times the depth of the girder from bridge joints. – Regularly remove all dirt, debris and other deposits that trap moisture. – Regularly remove all vegetation which can prevent the natural drying of wet steel surfaces. – Maintain covers and screens over access holes. PennDOT Design Manual, Part 4 PennDOT (2015) provides general guidelines for steel paint, galvanization, and metallization, as well as the use of weathering steel. All new bridges are required to be painted with an inorganic zinc paint system; alternatively, galvanization or metallization may be used. Unpainted weathering steel is also permitted for certain rural regions with low air pollution and little to no salt spray or application of deicing chemicals. Unpainted weathering steel is not permitted: • without approval from the District Bridge Engineer; • in acidic or corrosive environments; • in locations with salt water spray or fog; • in depressed roadways (less than 20 ft. of clearance) where salt and other pollutant spray can be trapped; • if there is low under clearance where the steel is either less than 5 ft. above normal water, continuously wet, or buried in soil; • for bridge types where debris (e.g., salt, dirt) can accumulate, such as trusses and inclined-leg bridges, unless provisions for painting are provided; • for expansion dams, or • for members under open steel decking. For situations where unpainted weathering steel is permitted, the following must be met: • minimize the number of expansion joints; • avoid details that retain water and debris; • paint the steel for a length of at least 1.5 times the web depth or a minimum of 5 ft. on each side of expansion joints; • provide drip plates; • protect substructures from staining using drainage details or protective coatings on reinforced concrete surfaces; • use weathering steel and stainless steel fasteners (ASTM A325 and A490, Type 3) and do not use cadmium galvanized carbon steel fasteners, and • the use of load indicator washers is not recommended. On existing uncoated weathering steel bridges, PennDOT requires cleaning and painting of beam ends 5 ft. from leaking joints and for areas subject to salt water spray. Drainage Within current codes and standards, the guidance related to drainage is general in nature. The following sections summarize and/or provide excerpts of this guidance from various codes and standards. In some cases, drainage guidance is provided in relation to structural steel coatings or to the protection of structural steel from corrosion. Provisions of this type were discussed in the section focused on structural steel coating above.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 91 Eurocode As part of the discussion of service limit states within EN 1993-2 (2006), the following guidance related to drainage is provided: • All decks should be waterproofed and the surfaces of carriageways and footpaths should be sealed to prevent the ingress of water. • The layout of the drainage should take into account the slope of the bridge deck as well as the location, diameter and slope of the pipes. • Free fall drains should carry water to a point clear of the underside of the structure to prevent water entering into the structure. • Drainage pipes should be designed so that they can be cleaned easily. The distance between centers of the cleaning openings should be shown on drawings. • Where drainage pipes are used inside box girder bridges, provisions should be made to prevent accumulation of water during leaks or breakage of pipes. • For road bridges, drains should be provided at expansion joints on both sides where is appropriate. • Provision should be made for the drainage of all closed cross sections, unless these are fully sealed by welding. Australian Bridge Design Code AS 5100.5 (2004) provides the following general guidance related to drainage: Provision of Drainage – Drainage should be provided wherever water may collect and should be designed to carry the water to a point clear of the underside of adjacent parts of the structure. Canadian Standards Association CSA (2014) provides the following general guidance related to drainage: • Downspouts for deck drains shall be located in such a way that runoff water is discharged away from any part of the bridge. • Downspouts shall extend at least 150 mm below adjacent members. • Wherever practical, deck drains shall not pass through the box girders. • Box girders shall be made watertight at their ends and adequately drained so as to reduce the potential for moisture entrapment and accelerated corrosion. • Pockets and depressions that could retain water shall have effective drain holes or an alternative means of drainage. • Measures shall be taken to prevent erosion from the discharge of drainage water. AASHTO The following guidance is provided by AASHTO (2017a) related to the drainage of bridge deck surfaces: The bridge deck and its highway approaches shall be designed to provide safe and efficient conveyance of surface runoff from the traveled way in a manner that minimizes damage to the bridge and maximizes the safety of passing vehicles. Transverse drainage of the deck, including roadway, bicycle paths, and pedestrian walkways, shall be achieved by providing a cross slope or superelevation sufficient for positive drainage. For wide bridges with more than three lanes in each direction, special design of bridge deck drainage and/or special rough road surfaces may be needed to reduce the potential for hydroplaning. Water flowing downgrade in the roadway gutter section shall be intercepted and not permitted to run onto the bridge. Drains at bridge ends shall have sufficient capacity to carry all contributing runoff. In those unique environmentally sensitive instances where it is not possible to discharge into the underlying watercourse,

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 92 consideration should be given to conveying the water in a longitudinal storm drain affixed to the underside of the bridge and discharging it into appropriate facilities on natural ground at bridge end. In addition, AASHTO (2017a) provides guidance related to deck drainage systems. Specifically, such systems are to be designed to direct surface water (from either the bridge deck or roadway) away from bridge superstructure and substructure elements. The following specific guidance is provided for consideration for the design of deck drainage systems: • A minimum 4.0-in. projection below the lowest adjacent superstructure component, • Location of pipe outlets such that a 45º cone of splash will not touch structural components, • Use of free drops or slots in parapets wherever practical and permissible, • Use of bends not greater than 45º, and • Use of cleanouts. Additional guidance provided relates to the drainage for cavities within structures, the avoidance of ponding on decks (especially near joints), and drainage of water at the interface between concrete decks and stay-in-place forms. FDOT Structures Manual FDOT (2017) requires drip grooves in accordance with the details shown in Figure 4. These details not only define the geometry of the drip groove, but also the clearance required from superstructure elements. Source: FDOT (2017) Figure 4. FDOT drip groove details (FDOT 2017, Figure 4.2.12-1) ODOT Bridge Design and Drafting Manual ODOT (2016) requires drainage to be carried by drain pipes to 3 inches below the bottom of the superstructure to prevent exposure of the superstructure to the drainage. GDOT Bridge and Structures Design Manual GDOT (2017) requires the design engineer to ensure that the bridge deck will freely drain, in order to minimize ponding through a combination of sloping and drainage systems. GDOT states that scuppers should be spaced along the bridge as required from a hydraulic study. GDOT discourages “low-points” in a bridge design but states that “when a low point is located on a bridge, it shall not be located within 10 feet of the back face of paving rest (BFPR) or centerline of bent, and scupper spacing shall be reduced to 2’-6” within 10 feet of the low point.” For open deck drainage, GDOT (2017) states:

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 93 For bridges with a 2.0% normal crown, the designer shall detail a 4” diameter hole through the bridge deck spaced at 10 feet along the gutter line. Deck drain spacing may need to be reduced to 5 feet to assure adequate drainage on super-elevated structures. However, deck drains shall not be located within 5 feet of the BFPR or centerline of bent, over non-rip-rapped end fills, railroads, or traffic lanes. When the top flange of the exterior beam interferes with the placement of deck drains, use a 3”x6” open slot in the bottom of the barrier in lieu of deck drains. When drainage of the deck is required and cannot be accommodated by conventional scuppers or barrier openings, a deck drainage system is required (GDOT 2017). PennDOT Bridge Design Manual, Part 4 PennDOT (2015) supplements AASHTO (2017a) in regard to roadway drainage by specifying additional requirements for cross slopes. The rate of deck cross slope must be 4% for water table widths 6 ft. or less and 3% for water table widths greater than 6ft. Additional slope requirements are given for certain conditions. Element Replacement and/or Future Expansion Table 61 summarizes the elements specified in each of the domestic and international standards to be designed for ease of replacement. Bearings, joints, and barriers/guardrails were the most common elements specified.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 94 Table 61. Replaceable elements called out in domestic/international design standards. Domestic/International Standard Replaceable Elements Specified AS 5100.5 Bearings Guardrails Joints Light poles Sign structures AASHTO LRFD Bridge Design Specifications Bearings Guardrails Joints Drainage elements Canadian Highway Bridge Design Code Bearings Cable Stays Joints Eurocode Asphalt layer and other surface protection Bearings Drainage devices Expansion joints Guardrails, parapets Noise barriers Stays, cables, hangers Wind shields FDOT Structures Manual Anchorages Barriers Bearings Cable stays Guardrails Joints Tendons ODOT Bridge Design and Drafting Manual Bearings Guardrails Joints The extent of deemed-to-satisfy guidance for replaceable elements is mostly limited. Often, the guidance provided is nothing more than to specify that the element be designed for future replacement or ease of replacement. The following sections detail some of the more extensive guidance for replaceable elements provided by different design standards. Australian Bridge Design Code AS 5100.5 (2004) calls out joints and bearings to be designed for “long service lives”, but also notes that the design should be such that these elements can be easily replaced. Designers should ensure that bridge components that are subject to movement, impact and wear, such as bearings, guardrails and expansion joints, can be readily replaced. Where possible, bolted attachment is preferable to permanent fixing. Sockets or bolts cast into concrete should be highly resistant to corrosion to ensure re-use. Provision should be made for jacking or similar to replace bearings and the practice of locating bearings in step joints beneath continuous deck sections should be avoided.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 95 Light poles, minor roadside sign structures and noise walls may be manufactured more economically with a shorter design life. Ease of replacement, public safety and consequence of failure are major factors in determining a suitable design life. Major sign gantries erected over roads should be designed for a 100 year life. Canadian Highway Bridge Design Code CSA (2014) states the following for replaceable elements: Where it can reasonably be expected that components will have to be replaced or modified during the design life of the structure, methods of replacement shall be investigated to ensure the feasibility, acceptable cost, and duration of the work and, where appropriate, the availability of alternative routes or detours for traffic. This investigation shall ensure the availability of access and the integrity of the structure during the work (CSA 2014). The replaceable elements specified in CSA include bearings, joints, and cable stays. For bearing replacement, design for jacking locations is required. Cable stays shall be designed to be easily replaceable (CSA 2014). Deck joints shall be designed to operate with a minimum of maintenance. They shall be replaceable (except for elements permanently attached to the structure) and accessible for inspection and Maintenance (CSA 2014). Bearings shall be designed to operate with minimal maintenance. They shall be accessible for inspection and maintenance and replaceable without damage to the structure or removal of anchorages permanently attached to the structure. To facilitate their placement, bearings shall be detailed so that they can be removed by jacking the superstructure by an amount not exceeding the vertical relaxation recovery of the elastomeric material within the bearing plus 5 mm (CSA 2014). Eurocode Eurocode specifies that components that cannot be designed to achieve the design working life of the bridge should be replaceable (EN 2006). These may include: • stays, cables, hangers • bearings • expansion joints • drainage devices • guardrails, parapets • asphalt layer and other surface protection • wind shields • noise barriers ODOT Bridge Design and Drafting Manual ODOT (2016) advises designers to consider the potential of bearing replacement during the life of the structure. Consideration of bearing replacement in the design of crossbeams, bents, and provisions for jacking locations are recommended.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 96 PennDOT Bridge Design Manual, Part 4 PennDOT (2015) states in Article 2.5.7.1 that future widening of a structure should be considered in design: “The load carrying capacity of exterior beams shall not be less than the load carrying capacity of an interior beam, unless specifically approved by the District Bridge Engineer.” The commentary to this section further indicates that the “stiffness of the interior and exterior beams should be relatively equal.” fib Benchmarking of Deemed-to-Satisfy Provisions fib Bulletin 76 (2015) aimed to benchmark deemed-to-satisfy rules for chloride-induced corrosion as given in the national codes of several countries. The countries examined in this benchmark were: Portugal, Great Britain, The Netherlands, Germany, Denmark, Norway, Spain, the United States, and Australia. European Standards EN 206 (BSI 2014) and EN 1992-1-1 (2004) were used as the basis for benchmarking the provisions of each of the nine countries considered. The exposure classes of each of the nine countries were equated to the XD and XS exposure classes listed in Table 3. Presented are the reliability ranges (spectrums) regarding the chloride-induced depassivation of rebar that can be expected under the deemed-to-satisfy rules of each country for a reference period of 50 years. The reliability analyses were carried out using the probability-based design for chloride-induced corrosion provided by fib Bulletin 34 (2006). The probabilistic calculations were carried out using input from short- term and rapid laboratory test data. The upper and lower bounds of the ranges are due to the diversity of specifications for concrete cover, cement type (chloride diffusion resistance), W/CM ratio, etc. These bounds represent the most/least critical results (i.e. worst case/best case, termed “favorable/unfavorable” given the different cement types in each of the exposure classes). Figure 5 and Figure 6 give the specified W/CM ratio, the nominal cover, and the reliability spectra of each country for XD and XS exposure classes, respectively.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 97 Source: fib (2015) Figure 5. Favorable and unfavorable types of cement, with maximum w/c-ratio and nominal cover (left) and reliability spectra for a design service life of 50 years (right), for exposure classes XD1 to XD3 (calculations according to data sheet in Appendix A of fib (2015)).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 98 Source: fib (2015) Figure 6. Favorable and unfavorable types of cement, with maximum w/c-ratio and nominal cover (left) and reliability spectra for a design service life of 50 years (right), for exposure classes XS1 to XS3 (calculations according to data sheet in Appendix A of fib (2015)). An independent assessment of existing structures was also carried out. These existing bridges had exposure classes, XS2 and XS3. The results were compared to the probability-based results as shown in Figure 7 below. Source: fib (2015) Figure 7. Spectra of reliabilities provided by deemed-to-satisfy rules for a design service life of 50 years, determined by reliability design (bars) and by assessing existing structures (dots), with numbering of dots based on Table 3-9 of fib (2015). The numbers refer to study cases listed in Table 3-9 and Appendix B of fib (2015).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 99 The results of this benchmark show that the reliability spectrum of the deemed-to-satisfy rules for each of the countries is always broad and independent of exposure class. It was found that the lower reliability level was a result of selecting low-resistance types of cement, suggesting that the widespread reliability within each exposure class is a result of differences in material performance. When compared to the target reliability indices the reliability level provided by the deemed-to-satisfy rules was generally lower. Summary and Knowledge Gaps Although some full probabilistic and partial factor approaches to service life design exist, in practice, these are limited to addressing the corrosion of reinforcement within uncracked concrete. As a result, the vast majority of the guidance provided by current codes and standards are of the deemed-to-satisfy or avoidance type. These generally address material selection/specification, dimensions and detailing, and general guidance related to access and ease of component replacement. To recognize the influence of environmental exposure, many codes and standards define a set of exposure classes, which reflect both global (e.g. climate) and local (e.g. component location) factors. As the exposure class becomes more severe, the deemed-to-satisfy provisions become more stringent. Although there has been some research aimed at validating deemed-to-satisfy provisions through comparison with full probabilistic approaches, it is limited due to the scarcity of generally accepted full probabilistic methods. In addition, there has been limited validation of deemed-to-satisfy provisions through comparison with the field performance of in-service bridges. As a result, while such provisions are easy to use and supported by general experience, their effectiveness for meeting specific service life goals and their relationship with more mechanistic approaches has not yet been quantified. For the development of specifications for service life design, what is required is data that ties design decisions to changes in the service life of components and systems. Determining the deterioration rate of an existing bridge component is not sufficient; knowledge of how design decisions affect the deterioration rate is needed in order to predict service life based on variations in design. As an example, a concrete deck on a multi-girder bridge can be considered. Design variables that significantly affect service life would include the properties of the concrete (chloride diffusion rates, air content, resistance to early-age cracking), the depth of concrete cover over the reinforcing, the type of reinforcing (bare bar, epoxy coated, galvanized), maximum tensile stress under live load, among others. Field data on deterioration needs to also record these variables in order to be useful in developing a deterioration model connecting design variables to target outcomes. The most well-developed deterioration model at present time that has been implemented as a tool in service life design is the chloride diffusion model described in fib Bulletin 34, and implemented in Bulletin 76, and others. As implemented for the chloride limit state, it ties chloride diffusion rates of concrete and concrete cover to service life. Measuring the chloride diffusion rate is somewhat complicated, and the correlation to mix designs is not clear. Perhaps the biggest gap in the method is the assumption of uncracked concrete. To quote directly from Bulletin 34: The corrosion rates in the regions of cracks crossing the reinforcement are extremely dependent on the micro climatic conditions at the concrete surface and the orientation of the concrete surface. Most severe conditions occur in case of horizontal concrete surfaces and both cracks and chloride attack from the top. For usual service lives more than 10 years and frequent chloride attack (e.g. parking decks in regions where deicing salts are used) special protective measures are necessary to avoid the rapid penetration of chlorides to the reinforcement (e.g. linings or crack bridging coatings). In case of vertical surfaces and horizontal surfaces with chloride spray from the bottom side and chloride containing water not leaking

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 100 through cracks high quality of concrete cover (cover thickness ≥ 50mm, low permeability concrete, w/c ≤ 0.5) and ordinary crack width limitation (wk,cal ≤ 0.3 mm) ensures sufficiently long service life (≥ 50 years) without extra protection. Thus, for horizontal concrete surfaces with cracks and chloride attack from the top (which describes the condition of bridge decks), Bulletin 34 indicates that high quality concrete and cover is insufficient to avoid rapid penetration of chlorides, and thus corrosion. The most significant barrier to improved service life design is the current lack of quantitative validation (inclusive of any necessary refinements) that has been performed for the current deemed-to-satisfy provisions. This validation, in general, should include consideration of all relevant deterioration mechanisms and quantify the probability that specific service lives can be achieved with specific deemed- to-satisfy provisions. To meet this goal, the following future research, beyond the scope of the current work, is suggested: • The development of a rational and quantitative definition for the end of bridge service life. Current standards generally describe this in terms of a design parameter (i.e. “anticipated” or “required” service life) as opposed to providing quantitative criteria for defining when the service life of a bridge has been exceeded. Without defining measurable criteria, research aiming to validate deemed-to-satisfy provisions or quantify the probability of achieving a specific service life is difficult to envision. • Further development of full probabilistic methods for a broad range of deterioration mechanisms beyond the corrosion of reinforcement within uncracked concrete. This research will need to primarily focus on the development and validation of mechanistic models to simulate specific deterioration mechanisms (freeze-thaw, corrosion-induced cracking of concrete, chemical attack, etc.) and, importantly, the interactions between these deterioration mechanisms. Even if such models are never fully adopted by codes and standards, they provide a powerful and necessary framework to both validate deemed-to- satisfy provisions and to develop partial factor approaches. • A coordinated and quantitative documentation of the performance of current deemed-to-satisfy provisions related to the observed service life of bridges. The data required for this research is currently being collected as part of FHWA’s flagship LTBP Program. This program is actively involved in the collection of various levels of quantitative performance data across broad populations of bridges (approximately 1,500 bridges nationwide). Such data should be used to establish a connection between materials/dimensions/details, environmental exposure classes, and service life. • The design and execution of controlled accelerated testing to examine and establish the means of interaction between various deterioration mechanisms. In service, bridges are exposed to environmental inputs, deicing chemicals, and live actions in concert; however, most of the research into deterioration mechanisms has focused on only a subset of these inputs. Such interactions are certainly implicit within the field data collected from operating bridges but may be difficult to isolate. As a result, experimental research capable of controlling and perturbing specific inputs is needed. A summary of data needs is listed below. The deterioration modeling and other data needed are essentially identical for each service life; the difference is in how the designs are adjusted to achieve the target service life. Below is a listing of the data needed, in bullet format. Concrete/Decks: • Include the effects of cracking on chloride migration, including the influence of: – Crack orientation with respect to the top layer of steel – Crack widths – Crack depth

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 101 – Extent of cracking • Determine effects of epoxy coating, galvanizing, and other alternate reinforcing types on service life. In particular, the material effects on the chloride threshold value are needed. • Determine effect of deck thickness on service life, perhaps also span-to-thickness ratio. • Determine effect of average daily traffic (ADT) and average daily truck traffic (ADTT) on service life of decks. • Deterioration of concrete due to processes not related to reinforcing corrosion or freeze-thaw cycling. • Unification of different approaches taken to concrete cracking thresholds in superstructure/substructure/foundations. Environment: • Determine the variation in surface chloride content as a function of: – Distance above roadway deck – Distance below roadway deck – Lateral distance from deck – Combinations of the above – Distance above water (brackish, salt, fresh) – Speed of traffic – Roadway configuration (tunnel-like conditions) – Distance from coast/saltwater – Surrounding medium (buried vs. submerged vs. atmospheric) Steel: • Corrosion rates of unpainted weathering steel as a function of surface chloride concentrations Overlays • Determine chloride migration rates for various types of overlays, including effects of cracking noted above Substructures and Foundations • Steel pile degradation rates and effect of soil chemistry • Concrete footing deterioration • Applicability of current standards – Resistivity testing required in submerged conditions – Not reasonable for Mechanically Stabilized Earth (MSE) walls • Native vs. fill soil effects on deterioration rates (shallow vs. deep foundations) Survey of Industry Practice A questionnaire was developed to gauge industry tacit knowledge and documented practices in relation to service life design. The questionnaire was sent to Bridge Engineers at all State DOTs as well as other selected transportation agencies. The list of responding agencies and raw response data are included in Appendix B. Several of the questions were included in the 2017 AASHTO Subcommittee on Bridges and Structures (SCOBS) Annual State Bridge Engineers Survey, while the remaining questions were sent out by the research team via an online questionnaire. 45 agencies responded to the SCOBS survey and 36 agencies responded to the online questionnaire.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 102 The questionnaire is grouped in three parts: 1. Bridge system 2. Bridge elements 3. Design for durability The following sections present the results of the questionnaire. Bridge System The following definitions were provided to respondents for reference when completing the questionnaire. • Design life: the period of time on which the statistical derivation of transient loads is based. • Service life: the period of time the bridge is expected to be in operation (given proper routine maintenance). 1. Considering the provided definitions, how does your agency define the end of service life for its bridges? [SBE survey]  Level of deterioration  Cost-benefit study on repair/rehabilitation vs. replacement  Structure has reached a life equal to the duration of the design limit state (e.g., 75 years)  Functional obsolescence  Other Figure 8 shows that nearly 80% of agencies surveyed use the level of deterioration to determine the end of service life of its bridges. Several agencies also use cost-benefit studies, while only a few use the design life or functional obsolescence as criteria for end of service life. One respondent provided structural deficiency as an additional factor. Figure 8. Factors used to define end of service life.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 103 2. Rank the following environmental factors in the order your agency considers to have the most significant impact on bridge service life. (1 = most significant; 5 = least significant) [SBE survey]  Precipitation  Temperature  Proximity to the coast or other body of water  Pollution  Other The ranking and associated distribution of each environmental factor is shown in Figure 9. Precipitation was the highest ranked factor by respondents. Temperature and the proximity to water were ranked second and third, respectively, and had nearly identical ranking distributions. Other factors were ranked fourth by respondents followed by pollution. *based on a total of 42 respondents Figure 9. Impact of environmental factors on bridge service life. 3. Rank the following design factors in the order your agency considers to have the most significant impact on bridge service life. (1 = most significant; 10 = least significant [SBE survey])  Span length  Joints/structural continuity  Superstructure type  Deck type (e.g., grid, concrete orthotropic)  Deck type (non-composite vs. composite)  Girder stiffness  Number and spacing of girders  Skew  Bridge profile (clear distance from bottom of bridge to top of water)  Other Design factor rankings and distributions are shown in Figure 10. Joints/structural continuity was by far the highest ranked factor, with nearly 70% of respondents ranking it first. In comparison, the next highest ranked factor, superstructure type, was ranked first by about 12% of respondents.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 104 *No. of respondents varied from 32 to 44 Figure 10. Impact of design factors on bridge service life. 4. Rank the following loading factors in the order your agency considers to have the most significant impact on bridge service life. (1 = most significant; 4 = least significant [SBE survey])  Frequency (ADT, ADTT)  Axle weights/spacing  Speed  Other Respondents ranked axle weights/spacing and load frequency as the two highest loading factors that affect bridge service life, as shown in Figure 11. Speed and other factors were consistently ranked third or fourth by the majority of respondents.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 105 *No. of respondents varied from 29 to 44 Figure 11. Impact of loading factors on bridge service life. 5. Rank the following owner actions in the order your agency considers to have the most significant impact on bridge service life. (1 = most significant; 10 = least significant) [SBE survey]  Selection of materials  Application of deicing salts on bridge  Application of deicing salts below bridge  Type of deicing salts applied  Frequency with which deicing salts applied  Frequency of maintenance  Frequency of cleaning  Load permitting  Bridge preservation activities  Other The ranking of owner actions by respondents is shown in Figure 12. Factors related to deicing salt usage, in terms of application and frequency of use, were ranked highest by respondents. Owner actions pertaining to maintenance of bridges were also considered to have a significant impact on service life, as frequency of maintenance and bridge preservation activities were ranked third and fourth respectively.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 106 *No. of respondents varied from 31 to 44 Figure 12. Impact of owner actions on bridge service life. 6. What is the average age (years) of a bridge that is replaced in your agency primarily due to deterioration? [SBE survey] A histogram of the average replacement age is shown in Figure 13. The majority of respondents indicated average replacement ages between 55 and 70 years. A significant number of agencies (6) also indicated average replacement ages greater than 75 years.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 107 Figure 13. Average age of bridge replacement due to deterioration. 7. Does your agency observe different durability problems depending on the environments (e.g, location, climate) within your State/jurisdiction? If yes, please explain the types of environment dependent durability problems. The majority of respondents (31) indicated that their agency sees different durability problems depending on the environment, while only five agencies answered no. For those who answered yes, durability problems associated with deicing activities caused by winter weather conditions were indicated the most by respondents, as shown in Figure 14, followed by marine environments and freeze-thaw conditions. Other environmental conditions specified by respondents included moisture areas and splash zones, scour, temperature, and urban and industrial environments. Figure 14. Agency observed environmental conditions associated with durability problems. 0 2 5 10 4 9 4 6 0 2 4 6 8 10 12 < 45 46-50 51-55 56-60 61-65 66-70 71-75 > 75 Fr eq ue nc y Age (years) No. of Respondents = 40

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 108 8. Based on the provided definitions above, what should be the target service life for each of the following service life categories of highway bridges? a) Normal service life (i.e., typical highway bridges) b) Enhanced service life c) Maximum service life (i.e., maximum achievable under current design standards and utilizing current materials) A histogram of the responses for target service life is shown in Figure 15 with corresponding median and standard deviation values listed in Table 62. Bridges categorized as normal were assigned the lowest target service life. The enhanced and maximum service life categories had identical median target service lives, but their corresponding standard deviations varied significantly. Figure 15. Target service lives of highway bridges. Table 62. Median and standard deviation of target service life responses. Service Life Category Normal Enhanced Maximum Median (years) 75 100 100 Standard Deviation (years) 25.6 73.6 21.0 9. Based on the provided definitions above, what should be the target service life for the following replaceable bridge components? [SBE survey] a) Joints b) Bearings c) Deck d) Paint e) Parapet Histograms of the target service lives for replaceable bridge components provided by the respondents are shown in Figure 16(a) through (e). As shown in the distributions, joints and paint were given the lowest target service lives, with median values of 20 and 25 years respectively. Bearings, decks, and parapets display similar target service life distributions, as shown in Figure 16(b), (c), and (e), with respondents assigning a median service life value of 50 years for all three components.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 109 (a) (b) (c) (d) (e) Figure 16. Target service lives for (a) joints, (b) bearings, (c) decks, (d) paint, and (e) parapets.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 110 Bridge Components 10. Rank the following bridge components in the order of most frequently repaired/replaced by your agency (1 = most frequent; 9 = least frequent) [SBE survey]  Deck (including parapet/railing)  Parapet/railing (excluding deck, except as minimally necessary to replace parapet/railing)  Superstructure (complete replacement from bearings up)  Partial superstructure (cut out and replace or reinforce beam end)  Piers, walls, and/or abutments  Foundations  Joints (including minimal deck work, as required)  Bearings  Other The ranking of repair and replacement frequency by respondents for the bridge components listed above is shown in Figure 17. Respondents indicated that joints are the most frequently repaired or replaced component, as over 60% of respondents ranked them first. Decks are also repaired or replaced often, with about 65% or respondents ranking decks either first or second. On the other hand, nearly 90% of respondents ranked foundation repair or replacement seventh or lower. *based on a total of 43 respondents Figure 17. Frequency of bridge component repair or replacement.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 111 11. What is the most common durability problem experienced by your agency in relation to concrete superstructure girders, beams, truss members, etc.? (check all that apply) [SBE survey]  Corrosion of reinforcement  Freeze-thaw attack  AAR  Chemical damage  Abrasion  Poor detailing (leading to poor performance or difficult maintenance and inspection)  Impact damage  Fire damage  Other More than half of respondents indicated corrosion of reinforcement as the most common durability problem of concrete superstructures, as shown in Figure 18, followed by impact damage, freeze-thaw attack, AAR, poor detailing, chemical damage, and fire damage. Figure 18. Concrete superstructure durability problems indicated by respondents. 12. What is the most common durability problem experienced by your agency in relation to concrete piers, walls, and abutments? (check all that apply) [SBE survey]  Corrosion of reinforcement  Freeze-thaw attack  AAR  Chemical damage  Abrasion  Poor detailing (leading to poor performance or difficult maintenance and inspection)  Impact damage  Fire damage  Other

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 112 As shown in Figure 19, corrosion of reinforcement was the most common response, followed by freeze- thaw attack, AAR, impact damage, abrasion, poor detailing, chemical damage, and fire damage. One respondent answered “unknown”. Figure 19. Concrete pier, wall, and abutment durability problems indicated by respondents. 13. What is the most common durability problem experienced by your agency in relation to concrete foundations? (check all that apply) [SBE survey]  Corrosion of reinforcement  Freeze-thaw attack  AAR  Chemical damage  Abrasion  Poor detailing (leading to poor performance or difficult maintenance and inspection)  Impact damage  Fire damage  Other Concrete foundation durability problems indicated by respondents are shown in Figure 20. Corrosion of reinforcement, poor detailing, freeze-thaw attack, and abrasion were all chosen by respondents. “Other” responses were comprised of scour (7), settlement (1), undermining (1), slope stability (1), and construction quality (1). Five respondents said concrete foundation durability was not an issue or not applicable.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 113 Figure 20. Concrete foundation durability problems indicated by respondents. 14. What is the most common durability problem experienced by your agency in relation to concrete bridge decks? (check all that apply) [SBE survey]  Corrosion of reinforcement  Freeze-thaw attack  AAR  Chemical damage  Abrasion  Poor detailing (leading to poor performance or difficult maintenance and inspection)  Impact damage  Fire damage  Other In order of the number of responses, the most common deck durability problems indicated were corrosion of reinforcement, freeze-thaw attack, chemical damage, abrasion, poor detailing, impact damage, AAR, and fire damage, as shown in Figure 21. The two “Other” responses were cracking (1) and poor quality control during construction (1).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 114 Figure 21. Concrete bridge deck durability problems indicated by respondents. 15. What is the most common durability problem experienced by your agency in relation to steel superstructure girders, beams, truss members, etc.? (check all that apply) [SBE survey]  Corrosion  Fatigue  Chemical damage  Abrasion  Poor detailing  Impact damage  Fire damage  Other As shown in Figure 22, nearly 90% of respondents indicated corrosion to be one of the most common durability problems of steel superstructures. In addition, multiple respondents listed both fatigue and impact damage as problems. Chemical damage, abrasion, poor detailing, and fire damage received one response each.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 115 Figure 22. Steel superstructure durability problems indicated by respondents. 16. What is the most common durability problem experienced by your agency in relation to steel foundation piles, shaft casings, etc.? (check all that apply)  Corrosion  Fatigue  None  N/A  Other The majority of respondents indicated corrosion to be the most common durability problem of steel foundations, as shown in Figure 23. “Other” answers were abrasion (1) and pile stability due to channel degradation (1). Figure 23. Steel foundation durability problems indicated by respondents.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 116 17. What is the most common durability problem experienced by your agency in relation to MSE and geosynthetics? (check all that apply)  Corrosion of MSE reinforcement  Chemical attack of geosynthetics  None  N/A  Other The most common MSE and geosynthetic durability problem indicated by respondents was corrosion of MSE reinforcement, as shown in Figure 24. “Other” problems mentioned by respondents included loss of wall fill (2), settlement (2), deterioration/failure of facing (2), poor construction (2), loss of material at joints (1), and concrete deterioration (1). Figure 24. MSE and geosynthetics durability problems indicated by respondents. 18. What is the most common durability problem of joints experienced by your agency? (check all that apply)  Leakage  Debris in joints  Material failure  Damage  Jamming and/or locking  None  N/A  Other In terms of joint durability problems, leakage was chosen the most by respondents, followed by debris in joints, material failure, damage, and jamming and/or locking, as shown in Figure 25. Fatigue (3), misalignment (2), header failure (1), and compression set (1) were mentioned as “Other” answers.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 117 Figure 25. Joint durability problems indicated by respondents. 19. What is the most common durability problem of bearings experienced by your agency? (check all that apply)  Deterioration due to leaking joints above bearings  Freezing and/or locking  Steel corrosion  Material failure  None  N/A  Other As shown in Figure 26, deterioration due to leaking joints above bearings was the most frequent answer among respondents, followed by steel corrosion, freezing and/or locking, and material failure. “Other” bearing durability problems included movement (2), over tilted rocker bearings (1), floating bearings (1), misalignment (1), walking (1), over rotation (1), leaking pot joints (1), and deterioration of the reinforced seat (1). Figure 26. Bearing durability problems indicated by respondents.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 118 Design for Durability 20. For the bridge components listed below, please provide initial design strategies (e.g., specific material, detail) that your agency uses to achieve and/or increase the service life. a) Superstructure girders, beams, truss members, etc. b) Piers and abutments c) Foundations d) Concrete bridge decks e) Joints f) Bearings Common design strategies identified by respondents are shown in Figure 27 through 32. Figure 27 shows service life design strategies used by agencies for superstructures, the most selected answer being the use of a protection system; the second most popular strategy involved the use of weathering steel or HPS. “Other” responses included accounting for the magnitude of pre-stressing in design (2), avoiding fatigue and fracture critical details (1), avoiding details that trap moisture (1), design using Strength II for permit vehicles (1), improved grouting procedures (1), the use of neoprene bearings (1), using steel over highways and concrete over water (1), and avoiding truss bridge designs (1). One respondent answered with “nothing / N/A”. *Includes coating, painting, sealing, galvanization, and metallization †Includes stainless steel, epoxy coated, and composites ‡Includes high performance concrete, admixtures, and consolidation Figure 27. Initial design strategies for superstructures indicated by respondents. Design strategies utilized for piers and abutments are shown in Figure 28; the most selected strategy being the use of corrosion/weather resistant material. “Other” strategies indicated by respondents included the use of precast and pre-stressed concrete (2), QA/QC procedures (2), and designing for impact (1).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 119 *Includes coating, painting, sealing, galvanization, membranes, and repellants †Includes stainless steel, epoxy coated, and composites ‡Includes high performance concrete, low permeability concrete, high strength, admixtures, SCMs, and water/cement ratio Figure 28. Initial design strategies for piers and abutments indicated by respondents. Figure 29 shows the design strategies indicated by respondents to increase the service life of foundation elements. “Other” responses were comprised of the grade of steel used (2), the use of deep foundations (2), QA/QC procedures (2), the use of drilled shafts (1), providing sufficient frost depth (1), the use of sacrificial anodes (1), the seismic design of all bridges (1), proper surface preparation (1), and designing on a case- by-case basis (1). “Nothing” or “N/A” was answered by five respondents. *Includes coating, painting, galvanization, metallization, and coal tar epoxies †Includes stainless steel, epoxy coated, and composites ‡Includes high performance concrete, admixtures, and SCMs Figure 29. Initial design strategies for foundations indicated by respondents.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 120 Common design strategies used by agencies for concrete bridge decks are shown in Figure 30, with type of reinforcement and concrete mix being the most selected design strategies. “Other” design strategies included pre-stressed concrete decks (3), evaporation control and protection (2), concrete finishing techniques (2), increasing the deck thickness (2), reinforcement orientation (2), the use of the empirical design method (1), avoiding the empirical design method (1), designing for site specific loads using weigh- in-motion data (1), proper drainage details (1), and QA/QC procedures (1). *Includes LMC, overlays, membranes, and sealants †Includes stainless steel, epoxy coated, composites, and galvanized ‡Includes high performance concrete, low permeability concrete, low cracking concrete, admixtures, SCMs, and fiber reinforced concrete Figure 30. Initial design strategies for concrete bridge decks indicated by respondents. Figure 31 shows design strategies utilized to increase the service life of joints. Jointless/joint elimination was clearly the preferred answer, as over 65% of respondents selected the design strategy. “Other” strategies indicated by respondents were fatigue design for modular joints (1), requiring a joint leak test at construction (1), promoting communication between the designer, contractor, and subcontractor (1), improve header materials (1), use of armoring (1), and prayers (1). “Nothing” was specified by one respondent.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 121 *Includes coating, galvanization, and metallization Figure 31. Initial design strategies for joints indicated by respondents. Service life design strategies for bearings provided by respondents are shown in Figure 32. Jointless/joint elimination and protection systems were the most common answers among respondents. Responses in the “Other” category included installing keeper bars (2), using steel reinforced pads (2), using weathering steel (1), proper construction techniques (1), and utilizing seismic bearings as required (1). *Includes galvanization and metallization †Includes stainless steel Figure 32. Initial design strategies for bearings indicated by respondents. In addition to providing design strategies for bearings, a large number of respondents indicated specific bearing types that were used to achieve and/or increase service life. The specific bearing types mentioned are shown in Figure 33, with the majority of agencies utilizing elastomeric bearings, followed by disc, neoprene and Highload Multi-Rotational (HLMR) bearings.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 122 Figure 33. Specific bearing types indicated by respondents used to increase service life. 21. For the bridge components listed below, please provide any design strategies that your agency has used in the past but no longer uses to achieve and/or increase the service life. a) Superstructure girders, beams, truss members, etc. b) Piers and abutments c) Foundations d) Concrete bridge decks e) Joints f) Bearings Design strategies that are no longer used by agencies are shown in Figures 34 through 39. Figure 34 shows service life design strategies no longer used for superstructures. “Other” responses included no longer using voided concrete box beams (1), reducing the number of truss designs (1), no longer using pin and hanger designs (1), avoiding steel girders over water (1), no longer using aluminum for sign support structures (1), not designing fracture critical members (1), and no longer using Allowable Stress Design (1). Seven respondents answered with “nothing” or “N/A”. Figure 34. Design strategies that are no longer used for superstructures.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 123 Design strategies no longer utilized for piers and abutments are shown in Figure 35. The only designs indicated by multiple respondents were the type of abutment, which included stub and diaphragm types, and the use of timber. “Other” strategies no longer used that were indicated by respondents included the use of deck joints over piers (1), integrating abutments into the roadway surface (1), utilizing MSE wall construction for abutments (1), using epoxy coated reinforcement (1), the use of epoxy resin protective coatings (1), and designing according to Allowable Stress Design (1). Eighteen respondents answered “nothing” or “N/A”. Figure 35. Design strategies that are no longer used for piers and abutments. Figure 36 shows the service life design strategies formerly used by agencies for foundation elements. Strategies with multiple responses included specific types of paints and coatings, the use of shallow and spread foundations due to scour and settlement issues, and designs using timber piles. “Other” responses were comprised of uncoated steel piling (1), the use of epoxy coated reinforcement (1), and Allowable Stress Design (1). “Nothing” or “N/A” was answered by nine respondents. Figure 36. Design strategies that are no longer used for foundations.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 124 Common design strategies no longer used by agencies for concrete bridge decks are shown in Figure 37. Specific types of overlays and certain types of reinforcement were mentioned the most by respondents. In addition, respondents identified several concrete mix design factors as well as deck designs with mixed reinforcement (e.g., epoxy coated reinforcement top mats and uncoated reinforcement bottom mats) as design strategies no longer used. “Other” design strategies included concrete surface treatments (1), using joints on bridge decks (1), the use of precast deck panels (1), and designing according to Allowable Stress Design (1). There were eight “Nothing” or “N/A” responses. *Includes LMC, silica fume, and asphalt overlays †Includes uncoated, epoxy coated, and GFRP reinforcement Figure 37. Design strategies that are no longer used for concrete bridge decks. Figure 38 shows design strategies no longer utilized to increase the service life of joints. The most common answer among respondents was specific types of joints, including compression seals, strip seals, preformed joint seals, aluminum joint seals, finger joints, pourable joints, armored joints, and sliding plate joints. One respondent indicated their agency has reduced the number of simple span bridges in an effort to eliminate joints. “Nothing” or “N/A” was specified by eight respondents. Figure 38. Design strategies that are no longer used for joints.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 125 Service life design strategies no longer used for bearings that were provided by respondents are shown in Figure 39. The majority of respondents mentioned specific types of bearings, including rocker bearings, pot bearings, steel sliding bearings, and fiber reinforced elastomeric pads. One respondent indicated grease fittings as no longer being used by their agency. “Nothing” or “N/A” was answered by seven respondents. Figure 39. Design strategies that are no longer used for bearings. 22. Which of the following requirements related to the reduction of corrosion in bridge decks does your agency have? (Check all that apply) [SBE survey*]  No requirements for reinforcing in bridge decks  Use empirical deck reinforcing to reduce steel in top mat  Use top mat of non-ferrous rebar  Use top mat of coated rebar  Use top mat of stainless clad rebar  Use top mat of corrosion resistant alloy rebar  Use top and bottom mats of non-ferrous rebar  Use top and bottom mats of coated rebar  Use top and bottom mats of stainless clad rebar  Use top and bottom mats of corrosion resistant alloy rebar  Other Requiring the top and bottom mats of reinforcement to be coated was by far the most indicated answer, as shown in Figure 40. Agencies also indicated designs utilizing top and bottom reinforcement mats of stainless clad or corrosion resistant alloys as common requirements. “Other” requirements listed by respondents were increased concrete cover (3), use of a waterproofing membrane with asphalt overlay (2), use of galvanized rebar for the top and bottom reinforcement mats (2), use of epoxy polymer overlays (1), and extending the wet cure to control shrinkage cracking (1). 0 Indicates questions that were included in the 2017 AASHTO SCOBS Annual State Bridge Engineers (SBE) Survey relevant to service life design developed by other researchers

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 126 Figure 40. Agency requirements for the reduction of corrosion in bridge decks. 23. Does your agency allow for the use of galvanized reinforcing bars in the construction of concrete deck slabs and concrete barriers? [SBE survey*] A plurality of responding agencies (19 of 44) allow for the use of galvanized reinforcement in deck slabs and barriers, while 14 do not. 11 of the respondents answered “N/A”. 24. What measures are recommended by your agency for the protection of driven piles? (Check all that apply) [SBE survey*]  Corrosion protective coating for steel piles  Over-dimensioning piles for steel piles  Cathodic protection for steel or P/S concrete piles  Increased cover to reinforcing in P/S piles  Corrosion resistant reinforcing and strands (e.g., stainless steel, carbon fiber reinforced polymer) in P/S piles  Other Figure 41 shows the driven pile protection measures used by agencies. The majority of agencies indicated the use of a sacrificial thickness (over-dimensioning) for steel piles, followed by corrosion protective coatings for steel piles. There were significantly fewer agencies that use any form of protection for their pre-stressed concrete piles. “Other” protection measures included coating exposed piling with a zinc rich primer (1), limiting the structural resistance of H-piles to Grade 36 steel (1), use of composite steel/concrete (1), using coal tar epoxy to protect piles in corrosive soil or water (1), use of fiber glass jackets in waterways (1), use of admixtures in pre-stressed concrete piles (1), and the use of a thin lining or Linex coating on the exterior of pipe piles (1). 5 respondents answered “N/A” or “None”.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 127 Figure 41. Protection measures for driven piles. 25. Does your agency account for the type and severity of environment as part of the bridge design phase? If yes, please explain such environment related design strategies. The majority of respondents, 23 of 36, indicated that their agency considers the environment when designing a bridge. Those who answered yes indicated corrosion and/or weather resistant material as the most common design strategy when considering the bridge environment, as shown in Figure 42, followed by adjusting the concrete cover or member thickness, using protection systems, avoiding the use of steel near water, and concrete mix design adjustments. “Other” responses included the use of bridge deck overlays (2), detailing for drainage (2), concrete curing practices (1), reducing the number of deck joints (1), use of pile foundations to avoid scour issues (1), the type of MSE wall reinforcement (1), use of elastomeric bearings (1), QA/QC procedures (1), and designing on a case-by-case basis (1). In addition, three respondents mentioned specific environment related considerations (e.g., bridge location, exposure), but did not provide associated design strategies. *Includes galvanization, metallization, sealing, and coating †Includes stainless steel, epoxy coated, and composites ‡Includes high performance concrete, admixtures, and SCMs Figure 42. Design strategies used by agencies when accounting for the type and severity of the environment.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 128 26. In marine environments, or other corrosion-prone environments, what material does your agency recommend for use in primary superstructure members? (Check all that apply) [SBE survey*]  Pre-stressed concrete beam with carbon steel strands  Pre-stressed concrete beam with stainless steel strands  Pre-stressed concrete beam with carbon fiber strands  Post-tensioned concrete girder with carbon fiber strands and grouted tendons  Post-tensioned concrete girder with stainless steel strands and grouted tendons  Post-tensioned concrete girder with carbon steel strands and flexible filler  Stainless steel (A1010)  Metallized A709/A992 non-weathering grade steel  Galvanized A709/A992 non-weathering grade steel  Other Over 60% of respondents indicated that their agency recommends pre-stressed concrete with carbon steel strands for superstructures in corrosion-prone environments, followed by metallized and galvanized steel, as shown in Figure 43. “Other” responses included the use of epoxy coating reinforcement in decks (2), calcium nitrite corrosion inhibitor added to pre-stressed concrete girders (1), corrosion resistant rebar for stirrups in pre-stressed girders (1), painted A709 steel (1), and not having a particular superstructure type recommendation but requiring other corrosion mitigation measures (e.g., increased cover, lower W/CM ratio, epoxy coated rebar, and use of SCMs for concrete and sacrificial thickness for steel) (1). 6 respondents answered “N/A” or “None”. *grouted tendons †flexible filler ‡non-weathering grade steel Figure 43. Material recommended for superstructures in corrosion-prone environments. 27. In marine environments, or other corrosion-prone environments, what type of reinforcement does your agency recommend for use in substructures? (Check all that apply) [SBE survey*]  Uncoated A615 or A706 reinforcing steel  Epoxy coated A615 or A706 reinforcing steel  Galvanized A615 or A706 reinforcing steel  Stainless reinforcing steel  A1035 reinforcing steel  Carbon Fiber Reinforced Polymer (CFRP) reinforcement

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 129  Other Fiber Reinforced Polymer reinforcement (e.g., aramid, glass, basalt)  Other The majority of respondents indicated epoxy coated reinforcement is used for substructures in corrosion- prone environments, as shown in Figure 44. Stainless steel, uncoated steel, and galvanized steel were other common types of reinforcement indicated by respondents. “Other” answers included stainless and CFRP reinforcement are only recommended for pre-stressed piles in certain locations (1), calcium nitrite corrosion inhibitor and silica fume are added to concrete (1), and the use of A615 reinforcement (1). “N/A” or “None” was answered by four respondents. Figure 44. Reinforcement recommended for substructures in corrosion-prone environments. 28. What type of specifications does your agency use when designing for durability? [SBE survey]  Agency specific, prescriptive specifications  Performance based specifications  Combination  N/A  Other For those respondents who provided a type of specification, agency specific prescriptive specifications were indicated most often, as shown in Figure 45. Of the agencies surveyed, only three use performance based specifications, while an additional six use a combination of prescriptive and performance based.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 130 Figure 45. Types of specifications used by agencies for durability design. 29. Has your agency used a formalized service life design process? (Check all that apply) [SBE survey*]  No  Yes, as a pilot  Yes, on smaller projects  Yes, on a major signature bridge  Planning to try as a pilot  Would use as a guide specification, if available  Other Nearly 80% of responding agencies indicated that they have not used a formalized service life design process, as shown in Figure 46. Four agencies have either used or are planning to use a formalized process, while 3 indicated they have used service life design on a major bridge project. An additional three agencies were open to using a guide specification for service life design. Figure 46. Agency use of a service life design process.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 131 30. Are there any design provisions related to durability or service life not currently required in Standards used by your agency that you believe should be implemented? If yes, please provide the design provisions you propose. Out of 37 respondents, 9 answered yes, 27 answered no, and 1 answered N/A. Of those respondents who answered yes, a majority proposed design provisions related to corrosion resistant materials, as shown in Figure 47, followed by protection systems, jointless bridge designs, and concrete mix designs. “Other” recommendations included adding location/exposure provisions (1), greater use of disc bearings (1), and increasing the use of bridge deck overlays (1). *Includes galvanization, metallization, sealing, and overlays †Includes stainless steel and other corrosion resistant material ‡Includes high performance concrete, admixtures, and chloride migration coefficient criteria Figure 47. Service life design provisions recommended by respondents. 31. Has your agency experienced issues resulting in reduced durability which could have been avoided if AASHTO or State design specifications included additional durability requirements? If yes, please explain. [SBE survey] Of 41 respondents, 90% believed durability problems could not have been avoided with additional specification requirements. Causes of reduced durability provided by those who answered yes (4 respondents) included staged construction, concrete permeability, and unforeseen circumstances. 32. What durability related practices should be avoided in new designs, based on your experiences? The most common practices mentioned by respondents are shown in Figure 48, with a large number of respondents indicating that joints should be avoided when possible, followed by problem prone details (e.g., bolted plates, crevices, fatigue prone) and concrete mix design aspects. Mixing the type of reinforcement in members, specific reinforcement types, and low clearance and close proximity issues (e.g., steel too close to water, piers immediately adjacent to roadway shoulders) were also common answers. “Other” responses included the use of steel items exposed in the deck (1), insufficient cover (1), steel culverts (1), polymer lining of existing steel culverts (1), protective overlays (1), timber piling (1), high skew (1), and “one size fits all” solutions (1). Five respondents answered “None” or “N/A”.

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 132 Figure 48. Durability related practices to avoid in new designs. 33. What durability related practices should be encouraged in new designs, based on your experiences? The majority of respondents (19 of 35 responses) believed jointless bridge designs should be encouraged, as shown in Figure 49, followed by specific concrete mix design practices and the use of corrosion resistant materials. Protection systems, increased concrete cover and thickness requirements, and proper detailing were additional practices that respondents thought should be promoted. “Other” responses were comprised of the use of precast and pre-stressed concrete (3), designing for maintenance and replacement (3), the use of weathering steel (2), specific bearing types (2), specific joint types (2), hydromilling before overlay application (1), designing well balanced structures (1), improved deck design and placement schemes (1), concrete deck curing (1), the use of concrete culverts (1), evaluate protective levels for specific exposure zones (1), and not designing to the limits of stress (1).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 133 *Includes galvanization, metallization, sealing, coating, painting, and overlays †Includes stainless steel, epoxy coated, composites, and other alternate/corrosion resistant materials ‡ Includes high performance concrete, low permeability concrete, low cracking concrete, admixtures, SCMs, fiber reinforced concrete, self consolidating concrete, strength requirements, and crack control requirements Figure 49. Durability related practices to encourage in new designs. 34. In making decisions regarding extending the service life of existing bridges, what information is gathered to determine what components should be salvaged and/or repaired? What new information, not available now, would be useful? The majority of respondents indicated the condition of bridge elements as criteria used to make salvage and repair decisions, as shown in Figure 50. Additional criteria mentioned by respondents included physical testing, inspection reports and bridge history, cost evaluations, capacity analyses, visual inspections, and the age or remaining life of the component. “Other” criteria listed by respondents included the presence of fatigue and fracture critical details (3), scour analysis (2), the type of reinforcement and protection (2), traffic capacity analysis (1), the type of substructure and foundation, and the presence of seismic issues (1). Four respondents answered “None” or “N/A”. New information that would be useful to agencies, in the opinions of the respondents, included methods to determine the remaining life of elements (3), means to assess the existing capacity and end of life of substructures (2), guidance on repair, rehabilitation, and replacement decisions (1), timeframes for overlay applications (1), a method to determine the severity and extent of corrosion (1), and a precursor to steel cracking (1).

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 134 *Includes load rating, superstructure analysis, substructure analysis †Includes core sampling, chloride content testing, petrographic testing, GPR, delamination testing, half cell potential, carbonation testing, and patina testing ‡ Includes corrosion/section loss, condition state, and overall deterioration/condition Figure 50. Criteria used to make salvage and repair decisions. 35. In your opinion, has your agency collected sufficient element level data that would be useful in developing deterioration models? Of 35 respondents, 16 (44%) believed their agency had collected enough element level data for use in developing deterioration models. 17 respondents (48%) did not think their agency had sufficient data, while 2 respondents responded “N/A”. Summary of Findings The results of the questionnaire provided the industry guidance related to bridge durability problems experienced by transportation agencies, past and present design strategies that were both beneficial and detrimental to service life, and information that would be useful in future service life design decisions. By far the most common durability problem indicated was corrosion of steel, whether it be reinforcement or the main material for superstructure members. Sources of corrosion included application of deicing salts and marine environments. Corrosion problems are accelerated by the number of joints on a bridge deck and their inability to seal the joint opening from water intrusion. Parallels were observed between the service life problems experienced by agencies and the design strategies implemented to combat such problems. Due to the failure of most joints to protect underlying bridge elements from exposure to moisture, agencies actively limit the number of joints or implement jointless designs when possible. Means of joint elimination vary by agency, but common strategies include giving preference to continuous spans over simple spans, integral abutments, and the use of link slabs. In addition to joint elimination, numerous other strategies are used by agencies to protect against corrosion, the most common being increasing concrete cover requirements, designing more durable concrete mixes, the use of protection systems such as coatings, paint, and deck overlays, and the use of corrosion and weather resistant materials including stainless steel and composites. In terms of design strategies no longer used by agencies to increase the service life of bridge components, there were not as many unanimous answers. Some agencies no longer design simple span bridges, do not allow a particular type of

NCHRP Web-Only Document 269: Guide Specification for Service Life Design of Highway Bridges 135 reinforcement, or have altered their concrete mix designs. Rather than providing answers that arrived at a consensus of design strategies, most of the responses were directed toward a specific type of bridge element or material, such as discontinuing the use of a certain joint, bearing, or deck protection system. Respondents also provided insight into what bridge owners would like to see implemented in a service life guide specification that would go above and beyond current standards. Agencies advocated for provisions related to joint elimination, the use of corrosion resistant materials, and higher quality concrete mix designs. Others indicated guidance on remaining service life and end of life of bridge components would be useful information. In addition, most agencies have never used a formalized service life design process, indicating a clear need for a guide specification to facilitate service life design into standard bridge design practice.