National Academies Press: OpenBook

Consequences of Delayed Maintenance of Highway Assets (2017)

Chapter: Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets

« Previous: Chapter 2 - Framework for Quantifying Delayed Maintenance of Highway Assets
Page 19
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 19
Page 20
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 20
Page 21
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 21
Page 22
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 22
Page 23
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 23
Page 24
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 24
Page 25
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 25
Page 26
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 26
Page 27
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 27
Page 28
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 28
Page 29
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 29
Page 30
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 30
Page 31
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 31
Page 32
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 32
Page 33
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 33
Page 34
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 34
Page 35
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 35
Page 36
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 36
Page 37
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 37
Page 38
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 38
Page 39
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 39
Page 40
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 40
Page 41
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 41
Page 42
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 42
Page 43
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 43
Page 44
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 44
Page 45
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 45
Page 46
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 46
Page 47
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 47
Page 48
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 48
Page 49
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 49
Page 50
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 50
Page 51
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 51
Page 52
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 52
Page 53
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 53
Page 54
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 54
Page 55
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 55
Page 56
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 56
Page 57
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 57
Page 58
Suggested Citation:"Chapter 3 - Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets." National Academies of Sciences, Engineering, and Medicine. 2017. Consequences of Delayed Maintenance of Highway Assets. Washington, DC: The National Academies Press. doi: 10.17226/24933.
×
Page 58

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.

19 Delayed maintenance affects an asset’s future condition and its ability to meet the performance objectives established by the highway agency. The procedure to quantify the consequences of delayed maintenance of highway assets involves comparing changes in the asset condition and other performance measures under delayed maintenance scenarios. Chapter 2 introduced the fundamental principles and general framework within which procedures have been prepared for quantifying the consequences of delayed maintenance of assets such as pavements, bridges, cul- verts, guardrails, lighting, pavement markings, and signs. This chapter describes the procedures for each highway asset group, incorporating preservation policies, types of maintenance activities, performance objectives, decision criteria, and analysis of scenarios. The procedures have been conceived for use by agency personnel responsible for managing individual highway asset groups. The ultimate goal, however, is to integrate the procedures into an overall asset management process (Figure 4). 3.1 Introduction In this research project, delayed maintenance has been defined as work needed to preserve the highway system but postponed in the agency-defined maintenance program. Maintenance prac- tices can differ significantly within and between highway agencies due to the asset size, overall condition, and the management policies affecting each asset group. Determining the consequences of delayed maintenance requires that the highway agency first articulate its preservation policy, including criteria to determine when to schedule maintenance activities. The agency must define the needed maintenance activities in the preservation program, how frequently or under what circumstances to perform the activities, and how the maintenance activities will contribute to the overall preservation program. This exercise is valuable to the highway agency and critical for quan- tifying the effects of delayed maintenance. It also may help identify resource allocation gaps, even without supplemental analysis of the consequences of delays in maintenance activities. Modeling the effects of delayed maintenance scenarios includes delays caused by postponing maintenance activities over a specified extent of time or by limiting the budget for maintenance. Performance models are required to forecast the asset condition or service life, to determine maintenance and budget needs, and to conduct scenario analyses to quantify the consequences of delayed maintenance. In general terms, performance models can be deterministic, probabilistic (e.g., Markovian), Bayesian, or subjective expert-based. The type of model chosen will depend on the nature of the data, the method and tools used to develop the model, and the ability to produce results that can be compared to the outcomes defined by agency policies. Analytical tools adopted by highway agencies typically use deterministic models to predict condition, service life, and performance measures based on asset age and other factors that may C H A P T E R 3 Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets

20 Consequences of Delayed Maintenance of Highway Assets affect performance over time (e.g., material type, traffic, environment). Development of deter- ministic models requires reliable asset condition and treatment data so that statistical methods can be applied to model asset performance. Given the complexity of asset performance and the number of factors that affect it, probabilistic models (which express the likelihood of achieving a certain condition or state) are alternatives to deterministic models. In the absence of suffi- cient or reliable data, Bayesian models use subjective information in model development, and expert-based models are the best approach when field data are not available. The selection of the performance model approach depends on the available data but also on the requirements of the analytical tool used to perform the analyses. Quantifying the consequences of delayed maintenance scenarios involves comparing the outcomes of a preferred agency preservation policy or baseline scenario with the outcomes of the delayed maintenance scenarios. Life expectancy, failure mechanisms, and other factors that affect asset performance differ across highway asset groups. In spite of these differences, how- ever, the consequences of delayed maintenance typically are reflected through: Step 2: Determine Maintenance and Budget Needs Integrated Asset Management Database Step 3: Conduct Delayed Maintenance Scenario Analyses • Quantify the consequences of delayed maintenance • Report the results Step 1: Define the Asset Preservation Policy • Types of maintenance • Performance measures and objectives • Decision criteria for interventions Bridges Other Highway Assets Culverts Guardrails Signs Pavement markings Asset Management Goals and Objectives Maintenance Program Feedback Budget Needs Performance Models Condition Assessment LightingPavements Figure 4. Integrating the procedures into the asset management process.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 21 • Deterioration of an asset’s condition over time, • Decreases in an asset’s remaining life, • Increases in an agency’s future costs to recover the desired level of service (LOS), • Increases in backlog costs over time, and • Decreases in the asset group’s value over time. This chapter presents procedures to quantify delayed maintenance for pavements, bridges, culverts, guardrails, lighting, pavement markings, and signs. The procedures require specifying what it means to delay maintenance, what is being delayed, and for how long. An agency may defer all maintenance for an extent of time and then resume following the same policy, or rede- fine the maintenance strategy to perform less maintenance. For each highway asset group, fur- ther details about policies, recommendations for the development or selection of performance models, and details for the specific examples of the needs and scenario analyses are provided in Appendices C through I. 3.2 Pavements The pavement preservation program should define the types of treatments conducted by the highway agency, network-level performance measures and objectives, and maintenance decision criteria. This section describes the procedure to quantify the consequences of delayed maintenance of pavements. Appendix C provides an example and more details on agency poli- cies for maintenance of pavements, along with performance models, needs analysis, and analy- sis of each scenario. 3.2.1 Pavement Preservation Policy 3.2.1.1 Types of Maintenance Activities Chapter 2 included FHWA’s definitions for preservation and types of maintenance. In practice, these definitions may vary by highway agency; the following lists cover typical pavement main- tenance treatments (AASHTO 2007, Peshkin et al. 2004): • Asphalt pavements: – Chip seals, – Cold in-place recycling, – Cold milling, – Crack filling or sealing, – Fog seal, – Hot in-place recycling, – Microsurfacing, – Patching, – Profile milling, – Thin asphalt overlays, – Scrub seals, – Slurry seals, – Ultra-thin asphalt overlay, – Ultra-thin bonded wearing course, and – Ultra-thin concrete overlay. • Concrete pavements: – Crack sealing, – Diamond grinding, – Diamond grooving,

22 Consequences of Delayed Maintenance of Highway Assets – Dowel bar retrofit, – Full-depth concrete patching, – Joint resealing, – Partial-depth concrete patching, – Thin asphalt overlay, and – Ultra-thin bonded wearing course. 3.2.1.2 Performance Objectives Pavement performance objectives are targets, defined by performance measures and commonly expressed in terms of agency-specific treatment trigger values. Treatment trigger values typically refer to a pavement’s condition (e.g., roughness, rut depth, faulting, cracking) or age (e.g., years since last treatment or years by which a maintenance, rehabilitation, or reconstruction activity should be applied). Pavement condition affects the functional, structural, and safety performance of the pavement. Typically, pavement condition data characterize surface characteristics, distress, and structural capacity. Surface characteristics data relate to pavement smoothness and surface texture, distress data refer to observations of the visible condition of the pavement surface, and structural capacity data refer to the ability of the pavement to withstand traffic loads (AASHTO 2012). Performance targets, set as objectives, often refer to the threshold value for the performance measure. Performance measures are used to set up objectives and to follow the outcomes (results) from the agency’s preservation programs. Preferably, each performance measure serves as a quantitative indicator of the LOS (e.g., quality of ride, safety, system condition) that has been established by the highway agency. Appendix C describes pavement performance measures in detail. Examples of performance objectives include: • Network-level pavement performance objectives: – Maximum International Roughness Index (IRI) of the pavement network, – Minimum pavement condition of the pavement network, – Minimum Present Serviceability Index (PSI) of the pavement network, – Minimum Remaining Service Life (RSL) of the pavement network, – Minimum percentage of the pavement network in good condition, – Maximum percentage of the pavement network in poor condition, – Minimum Skid Number (SN) of the pavement network, and – Minimum International Friction Index (IFI) of the pavement network. • Project-level individual pavement distress objectives: – Maximum percentage of cracking allowed for a pavement section, – Maximum amount of rutting allowed for a pavement section, and – Maximum amount of faulting allowed for a pavement section. 3.2.1.3 Decision Criteria for Maintenance Activities Pavement maintenance activities are applied at pre-scheduled time intervals or once the pave- ment condition declines to an established trigger value. Pre-scheduled maintenance activities or treatments based on time intervals are common practice for pavement preservation. Pavement performance measures such as the IRI, PSI, or individual distress objectives (e.g., cracking, rut depth, rutting, faulting) are used to set values to trigger maintenance activities. 3.2.1.3.1 Pre-scheduled Maintenance Based on Time Intervals. NCHRP Report 523: Opti- mal Timing of Pavement Preventive Maintenance Treatment Applications provides recommenda- tions about treatment timing intervals (Peshkin et al. 2004). The treatment timing cycles in Table 3 are based on estimates of expected life; in practice, these cycles will require revision based on agency experience and local maintenance practices.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 23 3.2.1.3.2 Maintenance Activities Based on Condition Trigger Values. Table 4 shows a sample set of default trigger values for IRI, cracking, rutting, and faulting used in the Pavement Health Track Tool (PHT) (O’Toole et al. 2013). If the amount of distress is equal to or worse than any of these trigger levels, the pavement section is a candidate for maintenance. Appendix C includes a list of other analytical tools for pavements. 3.2.2 Pavement Maintenance and Budget Needs 3.2.2.1 Assess Current Pavement Condition The method selected to assess the pavement condition depends on the performance measures used by the agency. Each agency has its own method to assess pavement condition. Usually, assessments are based on individual distresses that are used to calculate a pavement condition index. Common methods for assessing pavement condition include the Long-Term Pavement Performance (LTPP) Distress Identification Manual (Miller and Bellinger 2014) and ASTM D 6433, Standard Practice for Roads and Parking Lots Pavement Condition Index Surveys. 3.2.2.2 Performance Models for Forecasting Pavement Condition Performance models are required to predict pavement condition over the analysis period. Forecasts of pavement performance commonly use deterministic models that predict a single Pavement Type Bituminous-surfaced PCC-surfaced Treatment Crack sealing Fog seals Scrub seals Slurry seals Microsurfacing Chip seals Ultra-thin friction course Thin overlays Joint and crack sealing Diamond grinding Recommended Year of Initial Treatment 1–3 0–3 1–6 2–6 3–7 2–5 2–6 5–8 4–10 5–10 Treatment Timing Cycle 2–6 years 1–2 years 1–3 years 3–5 years 4–7 years 4–7 years 7–10 years 7–10 years 7–8 years 5–10 years Source: Adapted from Peshkin et al. (2004) Table 3. Examples of time intervals for maintenance treatment cycles. Surface Type Class IRI (in./mi.) Cracking Rutting (in.) Faulting (in.)Percent Length (ft./mi.) Flexible, Composite Interstate 80 0 250 0.25 N/A Primary 100 0 1,000 0.25 N/A Secondary 125 5 1,000 0.25 N/A Rigid Interstate 100 0 N/A N/A 0.10 Primary 100 0 N/A N/A 0.10 Secondary 125 0 N/A N/A 0.10 Source: O’Toole et al. (2013) Table 4. Default maintenance trigger values in PHT.

24 Consequences of Delayed Maintenance of Highway Assets value. Alternatively, probabilistic models can be used to generate a range of values that express the likelihood of occurrence of a certain condition state. Bayesian models combine objective and subjective data to predict future condition states, and expert-based models rely on expert opinion (AASHTO 2012). These performance models are used mainly at the network management level and are incorporated into PMSs. All these models are based on initial construction or design data; they need to be adjusted to reflect individual performance trends observed over time, including the effects of a treatment action on the condition and RSL of a specific pavement section. 3.2.2.3 Needs Analysis An agency performs a needs analysis to determine maintenance treatment and budget needs over the analysis period. The needs analysis assumes that sufficient funding exists to implement the desired preservation plan, and it is considered the baseline scenario. Agencies use perfor- mance models to predict pavement condition over time. A treatment is assigned based on a deci- sion tree or selection matrix that considers pavement condition or other performance measures, and treatment costs are calculated along with the expected improvement in pavement condition and extension of remaining life due to the assigned treatment. This process is repeated for all years included in the analysis period. The outcome of this process is a list of pavement sections in need of treatment and the corresponding budget. 3.2.3 Consequences of Delayed Maintenance of Pavements 3.2.3.1 Formulate Delayed Maintenance Scenarios In combination with the performance models, a pavement network inventory and current condi- tion assessment are needed to perform the scenario analysis using a variety of analytical tools. Table 5 summarizes the key elements needed to analyze delayed maintenance scenarios for pavements. Data Performance Models Maintenance Scenarios (Length of analysis: 20 years *) Results Pavement Network Inventory with Condition Assessment Deterministic Probabilistic Bayesian Expert-based 1. All needs (baseline) 2. Do nothing 3. Delayed maintenance with treatments delayed: a. By a set period of time (e.g., 2 years) b. Until pavement condition degrades beyond an established trigger value (treatment condition category) 4. Budget driven with limited funds for maintenance, such as: a. 40% of annual baseline maintenance budget b. 80% of annual baseline maintenance budget Analytical tools: • Use the agency-adopted PMS to analyze the scenarios (recommended). • Appendix C describes additional databases and analytical tools. Reports: • Agency costs over time • Effect on pavement network condition • Change in deferred maintenance costs over time • Effect on remaining life • Changes in the pavement network value and Pavement Sustainability Ratio (PSR) * A 20-year analysis period is suggested because it is typical for many highway agencies. Table 5. Key elements to analyze delayed maintenance scenarios for pavements.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 25 Delayed pavement maintenance scenarios are compared to the baseline scenario to quantify the consequences of the delay. The analysis period can be 5 years, 10 years, 20 years, or longer, depending on the agency’s planning period. The research team suggests extending the length of the analysis through the first rehabilitation period after the initial design life. 3.2.3.2 Delayed Maintenance Scenario Analyses When maintenance is delayed, the pavement condition deteriorates and the LOS decreases. The consequences of the delayed maintenance scenarios on pavement condition are quantified by comparing the results of each to the baseline (“all needs”) scenario. The analysis of the vari- ous scenarios can be conducted using the agency’s PMS. As an alternative, databases and ana- lytical tools developed for FHWA and other organizations are available for scenario analysis, as described in Appendix C. The appendix includes an example that details the scenarios formulated in Table 5. Table 6 summarizes the results of this example, including for each scenario the total cost to the agency for all work performed over the 20-year analysis period and the backlog cost at the end of the analysis period. At the start of the analysis period, the percentage of pavements in very poor condition is 2.2%. 3.2.3.3 Report the Consequences of Delayed Maintenance To quantify the consequences of delayed maintenance, the results of the delayed maintenance scenarios are compared to the baseline (“all needs”) scenario. Reporting the consequences of delayed maintenance is important to justify the effectiveness of timely treatments in terms of both monetary values and pavement condition. The scenario analysis results can be used to quantify the consequences on: • Future pavement condition, • Future RSL, • Future budget needs and agency costs, Scenario Description Total Agency Cost 1,2 Backlog Cost 1,2 Percentage of Pavements in Very Poor Condition 1 1 a. All needs $325 million $0 0% b. Preserve current condition $181 million $234.2 million 24.5% 2 Do nothing $0 $593.5 million 45.1% 3 Delayed maintenance, with treatments delayed by 2 years $192 million $209.7 million 18.6% 4 Budget driven with limited funds: a. 40% of annual baseline maintenance budget b. 0% of annual baseline maintenance budget $170 million $181 million $274.7 million $310.9 million 34.1% 35.8% 1 At the end of the analysis period 2 Total cost using 3% interest and inflation rate Table 6. Summary of results, pavement maintenance scenario analysis.

26 Consequences of Delayed Maintenance of Highway Assets • Backlog costs, and • Asset value of the pavement network. The specific results and comments reported in the example detailed in Appendix C are sum- marized as follows: • Scenario 1.a, “All needs,” requires a total agency expenditure of $325 million to implement the agency-desired maintenance program and preserve the highway network over the 20-year analysis period. • Scenario 1.b, “Preserve current condition,” preserves the current condition of the pavement network over the 20-year analysis period and results in an agency cost of $181 million, but by the end of analysis period the backlog cost is $234.2 million and 24.5% of the pavement network is in very poor condition. • Scenario 2, “Do nothing,” allows the pavement network to deteriorate because of the unavail- ability of funds, such that by the end of the analysis period almost half (45.1%) of the pave- ment network is in very poor condition. Also, this scenario results in the highest backlog cost ($593.5 million). • Scenario 3, “Delayed maintenance,” delays maintenance for 2 years, requires an agency expenditure of $192 million, and results in a $209.7 million backlog cost while leaving 18.6% of the pavement network in very poor condition. • Scenario 4.a, “Budget driven with limited funds” using 40% of the annual baseline main- tenance budget, shows the effect of funding only 40% of preventive maintenance needs; the scenario results in a backlog cost of $274.7 million, with 34.1% of pavements in very poor condition. Reducing the allowable maintenance budget by 40% reduces total agency costs only by $11 million in comparison to Scenario 1.b, but it adds approximately $40 million to the backlog costs and adds approximately 10% more pavements in very poor condition at the end of the analysis period. • Scenario 4.b, “Budget driven with limited funds” using no funds from the maintenance bud- get, shows the consequences of performing no pavement maintenance. This scenario results in a total annual cost of $181 million and a backlog cost of $310.9 million, and yields 35.8% of pavements in very poor condition. The scenario’s total agency costs match those of Scenario 1.b; however, this scenario’s backlog costs are approximately $70 million higher than those in Scenario 1.b, and this approach yields approximately 11% more pavements in poor condition. Appendix C provides additional information about the example and the detailed analysis of these scenarios, and presents samples of plot charts for reporting the consequences of delayed maintenance. 3.3 Bridges The bridge preservation program should define the types of treatments conducted by the agency, network-level performance objectives, and decision criteria (i.e., trigger values) for maintenance. This section describes the procedure to quantify the consequences of delayed maintenance of bridges. Appendix D provides an example and more details on agency policies for maintenance of bridges, along with performance models, needs analysis, and analysis of each scenario. 3.3.1 Bridge Preservation Policy 3.3.1.1 Types of Maintenance Activities Bridge preservation includes cyclical preventive maintenance, condition-based maintenance, and rehabilitation (FHWA 2011). Cyclical preventive maintenance activities are scheduled follow- ing “a pre-determined interval” and are intended to “preserve existing bridge element or compo- nent conditions. Bridge element or component conditions are not always directly improved as a

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 27 result of these activities, but deterioration is expected to be delayed” (FHWA 2011). Table 7 shows examples of cyclical preventive maintenance activities. FHWA’s Bridge Preservation Guide (FHWA 2011) describes the activities listed in Table 7, which can be performed as needed based on bridge condition. Condition-based preventive main- tenance activities are “performed on bridge elements as needed and identified through the bridge inspection process” (FHWA 2011). Activities include cleaning and resealing deck joints, installing deck overlays, replacing edge beams or expansion joints, patching spalls, painting structural steel, pressure washing or painting concrete members, installing scour countermeasures, conducting fracture-critical retrofitting, and constructing deck overlays, deck hydro-demolition, or full deck replacement (Georgia DOT 2013). Rehabilitation “involves major work required to restore the structural integrity of a bridge as well as work necessary to correct major safety defects” (FHWA 2011). Often, if needed rehabili- tation work is not performed, a bridge may become structurally deficient (SD) and/or it may become necessary to replace the bridge. The definitions of various maintenance activities incorporate consideration of both the nature and the intent of each activity. An activity such as a deck overlay may therefore be described as “preservation,” “cyclical preventive maintenance,” “condition-based preventive maintenance,” or even “rehabilitation,” depending on the motivation for performing the overlay and the extent of other work being performed concurrently on the bridge. 3.3.1.2 Performance Objectives Highway agencies establish performance target objectives when formulating their preserva- tion programs. When selecting bridge performance measures it is important to consider the dif- ferent factors that contribute to bridge performance, such as structural condition, functionality, structural integrity, and costs to agency and users (Hooks and Frangopol 2013). Appendix D describes commonly used bridge performance measures. The research team suggests that the performance measures selected to express the objectives include at least one measure of physical condition (e.g., percentage of bridges classified as SD), and at least one measure incorporating investment needs (e.g., the increase in overall bridge backlog costs caused by delaying maintenance). The following performance measures are recommended for establishing performance objectives: • Percentages of bridges in good, fair, poor, or severe condition, as defined based on National Bridge Inventory (NBI) condition ratings; Cyclical Preventive Maintenance Activities (examples) Commonly Used Frequencies Wash/clean bridge decks or entire bridge 1–2 years Install deck overlay on concrete decks using, for example: • Thin bonded polymer system overlays • Asphalt overlays with waterproof membrane • Rigid overlays (e.g., silica fume, latex modified) 10–15 years 10–15 years 20–25 years Seal concrete decks with waterproofing penetrating sealant 3–5 years Zone coat steel beam/girder ends 10–15 years Lubricate bearing devices 2–4 years Source: FHWA (2011) Table 7. Cyclical preventive maintenance activities for bridges.

28 Consequences of Delayed Maintenance of Highway Assets • Percentage of bridges classified SD; • Percentage of bridges classified SD or functionally obsolete (FO); • Average Sufficiency Rating (SR); • Average Bridge Health Index (BHI); and • Number of posted bridges. 3.3.1.3 Decision Criteria for Maintenance Activities The decision criteria should specify the activities needed based on (1) bridge/element condi- tion, (2) cost of activities, and (3) treatment timing. The AASHTO Manual for Bridge Element Inspection (2013) defines condition states and feasible actions for all national bridge elements (NBEs) based on distress severity. Table 8 shows activities typically defined for NBE condi- tion states. In addition to condition-based maintenance actions, cyclical preventive maintenance activi- ties (e.g., deck washing) may be performed regardless of bridge condition. 3.3.2 Bridge Maintenance and Budget Needs 3.3.2.1 Assess the Bridge Network Condition The agency must determine what methodology to use for assessing the bridge condition. All U.S. highway agencies must report NBI data for their highway bridges and recently began reporting element-level data for bridges on the National Highway System (NHS). Agencies must determine whether to assess condition based on general condition ratings or more detailed, element-level data. The Recording and Coding Guide for the Structure Inventory and Appraisal of the Nation’s Bridges (FHWA 1995b) explains how to assess condition using general condition ratings; and the AASHTO Manual for Bridge Element Inspection (2013) details how to assess condition for each National Bridge Element (NBE). FHWA’s requirements for element-level inspections still strictly apply to NHS bridges. Updates on NBI requirements can be accessed online at https:// www.fhwa.dot.gov/bridge/nbi.cfm. 3.3.2.2 Performance Models to Forecast the Bridge Condition Common practice for bridge components or element-level models is to define probabilistic models that specify the likelihood of transition from a rating value (if using condition ratings) or condition state (if using element data) depending on what activity (if any) is performed (Nebraska DOR 2011). The set of probabilities describing all the possible rating/condition values, feasible activities and their probabilities of transition to other condition states is called the transition prob- ability matrix. When an agency uses a BMS to support the analysis, the BMS typically is populated with a default set of probability matrices; but these default values should be carefully reviewed for NBE Condition State 1 Description Commonly Employed Feasible Actions 1 Good Preventive maintenance 2 Fair Preventive maintenance or repairs 3 Poor Rehabilitation 4 Severe Rehabilitation or replacement 1 NBI condition ratings involve more levels than AASHTO’s NBE condition states; however, an FHWA translator computer program can be used to convert inspection data based on NBE condition states to NBI condition ratings. Appendix D includes more information on these rating systems. Sources: AASHTO (2013) and FHWA (2011) Table 8. AASHTO condition states and feasible preservation actions.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 29 reasonableness. If the agency does not use a BMS to support the analysis, additional work will be required to define a set of transition probability matrices, or to develop an equivalent set of models to predict bridge performance. Another approach is to develop a deterministic deterioration curve to predict component or element condition as a function of age. 3.3.2.3 Needs Analysis The agency is expected to use the previously determined set of feasible preservation activi- ties with the performance models to determine treatments that should be performed assuming optimal maintenance. In a BMS, an optimization process determines the “optimal preservation policy” based on a set of activities in order to minimize life-cycle costs. If a BMS is not used, an agency can test various strategies to confirm the best maintenance strategy for its bridges. Once the policy/strategy has been established, it is applied to the bridge inventory to estab- lish initial bridge maintenance needs, creating the preferred or baseline scenario that is used to quantify the costs, bridge condition, and other performance measures when maintenance is performed as scheduled. The example provided in Appendix D includes a series of bridge analysis scenarios and describes how the National Bridge Investment Analysis System (NBIAS) can be used to evaluate each scenario (see Appendix D, Section D.3). In the example, the needs analysis serves as a base- line scenario in which the optimal preservation policy is performed to minimize life-cycle costs. 3.3.3 Consequences of Delayed Maintenance of Bridges 3.3.3.1 Formulate Delayed Maintenance Scenarios Table 9 summarizes the key elements needed to analyze the delayed maintenance scenarios for bridges. 3.3.3.2 Delayed Maintenance Scenario Analysis Analysis of each delayed maintenance scenario is performed using the steps in the following process: 1. Determine the bridge condition at the beginning of each year of the analysis period to establish needed maintenance work. 2. Prioritize what maintenance work will be performed based on available funds. 3. Determine the effect of funded work on bridge condition, including maintenance work and any other rehabilitation/replacement work. 4. Determine the effect on the bridge network condition. 5. Tabulate the performance measures selected by the agency. 6. Calculate the investment needs. 7. Calculate the gap between predicted and targeted performance. Appendix D provides information on databases and tools developed for FHWA, AASHTO, and other organizations that can be used to perform delayed maintenance scenario analysis. Appendix D also provides an example of scenarios analysis using the NBIAS. Table 10 summa- rizes relevant information from this example, including agency costs, user benefits, and benefit/ cost ratio (BCR). The BCR for performing needed maintenance is calculated as the net benefit of performing maintenance work (sum of the agency’s cost increase for total work done and reduction in user benefits from maintenance work) divided by the agency’s increased cost for performing (and not deferring) the needed maintenance. In Table 10, the net benefit of the baseline scenario (all needs) versus a 10-year deferral scenario totals about $6.4 billion and the net benefit of the baseline versus

30 Consequences of Delayed Maintenance of Highway Assets Data Performance Models Maintenance Scenarios (Length of analysis = 20 years *) Results Bridge network inventory with condition assessment NBI data for all 50 states NBIAS default costs Probabilistic— Markov models NBIAS default deterioration models NCHRP Report 713 (deck, super- structure and substructure deterioration models by state) 1. All needs (baseline) 2. Do nothing 3. Delayed maintenance: a. Condition-based treatments delayed by 10 years b. Condition-based treatments delayed by 20 years c. Cyclical preventive maintenance delayed by X years 4. Budget driven with limited funds for maintenance 5. Maintenance performed only on selected bridge systems 6. Maintenance performed only if treatment activity meets a selected benefit/cost ratio (BCR) or if asset condition falls below threshold condition Analytical tools: • NBIAS (used by many DOTs) • In-house BMS available at the agency • Appendix D describes other databases and analytical tools Reports: • Agency costs over time • Effect on bridge network condition • Change in deferred maintenance costs over time • Changes in the bridge network value and Bridge Sustainability Ratio (BSR) • Changes in sustainability and users’ costs Notes: Maintenance policies should be developed considering the full life-cycle of a bridge, regardless of the analysis period. When the initial needs analysis is performed, it should include bridges with a range of ages. * NBIAS can be run for periods up to 30 years; however, 20 years is recommended because FHWA uses 20 years for the analysis of bridge performance. Table 9. Key elements to analyze delayed maintenance scenarios for bridges. MR&R = Maintenance/repair and reconstruction (or replacement). Deferral Period Agency Cost Increase for Total Work Done Reduction in User Benefits Obtained from MR&R Net Benefit of Baseline vs. Deferral Agency Cost Reduction in MR&R Work Done Benefit/Cost Ratio (BCR) of Baseline Relative to Deferral 10 years (Scenario 3.a) $6,355 million $108 million $6,463 million $659 million 9.8 20 years (Scenario 3.b) $12,338 million $160 million $12,498 million $1,483 million 8.4 Table 10. Comparison of delayed maintenance scenarios to baseline (all needs) scenario.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 31 a 20-year deferral scenario totals about $12.5 billion. Dividing these amounts by the relative cost reductions ($659 million for the 10-year deferral and $1,483 million for the 20-year deferral) yields the BCRs (9.8 and 8.4) for not deferring the work by 10 years or 20 years, respectively. The BCR is lower for the 20-year deferral because the NBIAS models do not consider increased needs for bridges that require reconstruction or replacement. This feature of the system is useful for isolating the costs of a shorter delay in needed maintenance, but tends to underestimate the costs that result from an extended deferral. This occurs because, with a long deferral of main- tenance, many additional risks that are not modeled in NBIAS can arise from having bridges in poor condition. 3.3.3.3 Report the Consequences of Delayed Maintenance Delaying maintenance activities on bridges can have serious consequences in condition, life-cycle cost, and safety. For this reason, it is crucial to report the consequences of delayed maintenance on bridges to account for them in the investment decision-making process. The scenario analysis results can be used to quantify the consequences using measures including, but not limited to: • Future bridge network condition (e.g., BHI), • Percentage of bridges classified SD, • Remaining life of the bridge network, • Future budget needs and agency costs, • Backlog costs, and • Bridge network value and sustainability ratio. Various types of reports demonstrate the consequences of delaying maintenance. Some reports are built into existing BMS software; others are customized reports. Examples of these reports are included in Appendix D. The following findings were obtained through the process of preparing and analyzing the results of the example scenarios in Appendix D: • NBIAS can be used to analyze the effects of delaying maintenance; however, depending on the deferral scenario being analyzed, it may help to simplify interpretation of the results by nar- rowing the analysis to bridges that are candidates for maintenance work, omitting any bridges for which reconstruction or replacement is already planned. • The most straightforward deferral scenarios to analyze using NBIAS are scenarios in which all maintenance work is deferred for a set time period. The analysis described in Appendix D illustrates 10-year and 20-year deferrals. • For deferral periods longer than 10 years, some risks (e.g., bridge closure) may not be captured by NBIAS. In practice, therefore, a deferral period of no more than 10 years is suggested. • The example analysis estimates the BCR of performing needed work to be 9.8 relative to delay- ing needed maintenance for 10 years. This estimate accounts for factors modeled by NBIAS, including increased costs needed to reconstruct or replace bridges as a result of delaying main- tenance, and the loss of user benefits of maintenance work. Appendix D provides additional information about the example and the detailed analysis of these scenarios, and presents samples of plot charts for reporting the consequences of delayed maintenance. 3.4 Culverts Deferred maintenance of culverts can result in culvert failures and increased cost for rehabilita- tion, which leads to unplanned financial burdens. Culverts that are in poor condition are a haz- ard; they can cause potholes or total collapse and failure of pavement, which present safety risks

32 Consequences of Delayed Maintenance of Highway Assets along with traffic disruption and time delays from road closures. Public safety and risk reduction are priorities in culvert management, followed by preservation activities to reduce life-cycle costs (Markow 2007). This section describes the procedure to quantify the consequences of delayed maintenance of culverts. Appendix E provides an example and more details about agency poli- cies for maintenance of culverts, along with performance models, needs analysis, and analysis of each scenario. 3.4.1 Culvert Preservation Policy 3.4.1.1 Types of Maintenance Activities Common maintenance or preservation activities for culverts can be grouped as follows. • Emergency maintenance: Activities taken in response to unforeseen events that affect culvert performance (Najafi et al. 2008). • Preventive maintenance: Activities that aim to prevent more serious problems in the future (Najafi et al. 2008). “Typical activities include joint sealing, concrete patching, mortar repair, invert paving, scour prevention, and ditch cleaning and repair” (FHWA 1995b). • Routine maintenance: Scheduled activities that aim to maintain the culvert in working condition by addressing deterioration issues. For example, during the scheduled mainte- nance the entire drainage structure is inspected to define maintenance activities. Routine maintenance includes work such as cleaning, debris removal, and realignment. “If the routine maintenance activities are not enough to solve a problem in a culvert and replace- ment is not a feasible option, then some of the repair techniques should be employed” (Najafi et al. 2008). • Rehabilitation: Activities that restore a culvert’s condition to its initial state and renew culvert service life (Wyant 2002 and Najafi et al. 2008). Rehabilitation methods include “repair of basically sound endwalls and wing walls, invert paving, repair of scour, slope stabilization, streambed paving, addition of an apron or cut-off wall, improving the inlet configuration to enhance culvert performance, or installing debris collectors” (FHWA 1995b), as well as slip lining, cured-in-place pipes, and pipe bursting (Najafi et al. 2008). • Replacement: Replacing an existing culvert with a new one, usually by cutting it open or using a trenchless method (Wagener and Leagjeld 2014). 3.4.1.2 Performance Objectives Common culvert performance measures are shown in Table 11. Culverts with openings 20 feet or greater in size are included in the NBI. Examples of performance objectives for culverts include: • Percentages of culverts in good, fair, and poor condition (Venner 2014), • Culvert age and RSL (Venner 2014), • Culvert condition by material (aluminum, corrugated metal pipe, reinforced concrete pipe, various plastic) (Vermont Agency of Transportation 2011), • Culvert condition by route (Vermont Agency of Transportation 2011), and • Condition by year constructed (Vermont Agency of Transportation 2011). 3.4.1.3 Decision Criteria for Maintenance Activities The decision criteria to trigger culvert maintenance activities should be based on the culvert condition, RSL, and costs. Table 12 shows examples of decision criteria for culvert maintenance activities.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 33 Performance Measure Description Source NBI Culvert Rating 0–9 rating similar to the deck, superstructure and substructure ratings for bridges FHWA (1995b) FHWA Federal Lands Highway Condition Rating Good, fair, poor, critical, unknown Hunt et al. (2010) HydInfra Condition Rating 1 = Like new 2 = Fair 3 = Poor 4 = Very poor 0 = Can’t be rated Wagener and Leagjeld (2014) New York State DOT Condition Rating 1 = Totally deteriorated 3 = Serious deterioration 5 = Minor deterioration 7 = New condition 8 = Not applicable 9 = Condition/existence unknown (Ratings of 2, 4, and 6 are used to shade between 1 and 3, 3 and 5, and 5 and 7.) New York State DOT (2006) Ohio DOT Condition Rating Excellent, good, fair, poor, failure/critical Culvert performance zones: Satisfactory, monitored, critical Najafi et al. (2008) Western Transportation Institute Rating System 0-1-2 rating system for degree of scour, failure, corrosion, inverts, joint separation, and damage. 0 = No issue 1 = Minor issue 2 = Major issue Wall (2013) Table 11. Examples of common performance measures for culverts. Decision Criterion Based on Culvert condition Maintenance (clearing, cleaning) and repair activities (Hunt et al. 2010) NBI condition rating (e.g., perform maintenance for an NBI rating of 4–6 and replace if rating is less than 4) Distresses with action options (Najafi et al. 2008) Intervention cost Statistical formula Software (e.g., Pontis *) * Now AASHTOWare Bridge Management Table 12. Examples of decision criteria for culvert maintenance activities.

34 Consequences of Delayed Maintenance of Highway Assets 3.4.2 Culvert Maintenance and Budget Needs 3.4.2.1 Assess the Culvert System Condition To evaluate the culvert condition, the following types of inspections are suggested (Ohio DOT 2014): • Inventory inspections that are conducted upon construction, • Routine inspections that are performed regularly to identify any physical or functional changes, • Damage inspections that are performed on culverts with known defects after major floods and storms to identify any damage that would require load restrictions or road closures, • Interim inspections that are conducted upon expert decision to perform an inspection on culverts that have known defects, and • Storm sewer inspections that can be either inventory or routine checks on storm sewers. Culverts with openings of 20 feet or greater are included in the NBI. For these culverts, over- all condition is rated using the same 0 to 9 scale described previously for inspecting bridge decks, superstructures, and substructures. A culvert is deemed to be in good condition if it has a rating of 7, 8, or 9 on this scale; in fair condition if it has a rating of 5 or 6; or in poor condition if it has a rating of 4 or less. FHWA’s Culvert Assessment and Decision-Making Procedures Manual for Federal Lands Highway (FLH) defines five condition categories for culverts: good, fair, poor, critical, and unknown (Hunt et al. 2010). Appendix E provides additional information on condition assessment. 3.4.2.2 Performance Models for Forecasting Culvert Condition Culvert performance can be modeled based on the culvert’s condition, age, or a combination of both factors. A condition-based approach requires periodic condition assessment to develop deterioration models, whereas an age-based approach estimates the remaining life from histori- cal construction records. A hybrid (combination) approach is recommended that updates the performance deterioration curve after each inspection. 3.4.2.2.1 Culvert Deterioration Model from Condition Data. In this research, culvert deterioration has been modeled by specifying the probability of transitioning from one condi- tion to another each year using a Markovian distribution. Table 13 shows deterioration prob- ability parameters for culverts. These probabilities were matched empirically to the estimates Rating Deterioration Probability 0 0.0% 1 5.0% 2 10.0% 3 6.3% 4 4.8% 5 4.8% 6 7.0% 7 10.0% 8 9.0% 9 50.0% Table 13. Example of culvert rating deterioration probabilities by rating.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 35 of culvert life from NCHRP Report 713: Estimating Life Expectancies of Highway Assets (Ford et al. 2012a) in combination with the results of analyses performed using state-level NBI data. Figure 5 shows the corresponding average condition rating over time using the probabilities from Table 13. A culvert with a rating of 9 quickly deteriorates to 8; it deteriorates linearly afterwards, reaching a rating of 3 at approximately 75 years. Theoretically, the culvert reaches a rating of 1 at approximately 120 years; in practice, however, replacement is performed before the culvert reaches this point. In the culvert model, the predicted condition rating values take into account needed work relative to deferred work as described in the needs analysis. 3.4.2.3 Needs Analysis The culvert model identifies maintenance and budget needs based on condition. Maintenance activities are set for each condition level with their costs, effect, and priority. The data required for the needs analysis include: • Culvert inventory with condition rating (on a scale 0 to 9), • Deterioration probability for each condition rating, • Effect of maintenance work on culvert condition, and • Cost of culvert maintenance work. Portions of the specified budget are allocated in order of priority. Default priorities for mainte- nance activities at each condition level are established through a Markov modeling approach, with the probability of transition from one condition rating to another determined empirically to match the estimated times to a rating of 3, 4, and 5 published for culverts in NCHRP Report 713. The Figure 5. Predicted culvert condition rating over time.

36 Consequences of Delayed Maintenance of Highway Assets defaults in the model are to perform maintenance work on a culvert when it has a rating of 4 or 5, and to replace a culvert when it has a rating of 3 or less. When no maintenance work is performed on a culvert, its deterioration is predicted probabilistically using the values specified in Table 13. The process followed by the model for each year of an analysis is as follows: 1. For each culvert, the needed work is established based on the culvert’s rating, and the cost of this work is calculated. For this example, culvert replacement was estimated to cost $180 per square meter of roadway area, and maintenance was projected to cost $30 per square meter. 2. A priority is assigned to each recommended action. In the example, highest priority was assigned to maintenance work, followed by rehabilitation/replacement of culverts in poor condition. 3. The future condition of the culvert in the next year is predicted for two scenarios: (a) if work is performed and (b) if work is deferred. In the example, maintenance work was assumed to raise the rating of the culvert to a value of 7, whereas rehabilitation or replacement was assumed to restore the culvert to a condition value of 9. 4. Funds are allocated in priority order until the budget is spent, or until insufficient funds remain to perform the next recommended action. 5. The culvert’s rating for the next year is calculated based on whether or not maintenance or rehabilitation/replacement work is projected to have occurred. 6. The outputs from each year serve as the inputs to the next year’s calculations. This model can easily be reconfigured to use different treatments, different condition ratings, or RSL as alternative approaches. 3.4.3 Consequences of Delayed Maintenance of Culverts 3.4.3.1 Formulate Delayed Maintenance Scenarios Table 14 presents the key elements needed to analyze delayed maintenance scenarios for culverts, including the performance models, a brief description of the set of scenarios, and expected results for culvert maintenance. Data Performance Models Maintenance Scenarios (Length of analysis = 20 years) Results Culvert Inventory with Condition Assessment NBI Data for all 50 States NBI Data on Bridge-length Culverts Probabilistic Markov Models 1. All needs 2. Do nothing 3. Delayed maintenance with a. 5-year cyclical delay of maintenance treatments b. Maintenance treatments deferred but rehabilitation/replacement treatments performed 4. Budget driven with limited funds a. 50% of annual baseline maintenance budget b. 25% of annual baseline maintenance budget Analytical tools: • NBIAS • Spreadsheet-based model to forecast the culvert condition Reports: • Effect on condition due to delayed maintenance • Agency costs of performing deferred work • Agency costs of emergency/reactive maintenance Table 14. Key elements to analyze delayed maintenance scenarios for culverts.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 37 3.4.3.2 Delayed Maintenance Scenario Analysis Appendix E provides a detailed example that includes analysis of the scenarios listed in Table 14. Table 15 summarizes the results of the scenario analysis, including the total costs over a 20-year analysis period and total costs discounted at a rate of 7%. Also shown are the backlog costs (unmet need) at the end of the analysis period, percentage of culverts in poor condition, and average rating. 3.4.3.3 Report the Consequences of Delayed Maintenance To quantify the consequences of delayed maintenance, the results of delayed mainte- nance scenarios are compared to the baseline scenario from the needs analysis. The scenario results clearly demonstrate the effects of delaying needed maintenance to culverts: increased backlog costs over time and increased numbers of culverts in poor condition. The specific results and comments reported in the example detailed in Appendix E are summarized as follows: • Scenario 3.a, “Delayed maintenance” using a 5-year deferral, slightly reduces the allocated budget or agency costs, but also results in an increase in the percentage of culverts in poor condition to 7.65 at Year 5, and a 5.92 condition rating at the end of the deferral period. The culvert system value is reduced to $284 million, and the remaining asset life falls to 43.8 years at the end of the 5-year deferral period. • Scenario 3.b, “Delayed maintenance” in which maintenance activities are not performed but culverts in poor condition are replaced, reduces up front agency costs by $18.0 million, but these savings come at the cost of worse condition and significant backlog costs ($29.2 million). In this example, 16% of the culvert system is in poor condition by Year 20 and the average condition rating is 5.42. The system value has decreased to $235 million, and the remaining asset life has decreased to 36.8 years. • Scenario 4.a, “Budget driven with limited funds” using 50% of the annual baseline mainte- nance budget, decreases the culvert system value to $291 million and the remaining asset ser- vice life to 44.7 years by the end of Year 20. Also by Year 20, the backlog costs are $13.5 million and the condition rating is 5.98. • Scenario 4.b, “Budget driven with limited funds” using 25% of the annual baseline mainte- nance budget also reduces spending, but at the cost of worsened condition and even greater backlog costs in meeting needs over time. By Year 20, the backlog costs are $35.5 million and the condition rating is 5.46. Also by Year 20, the culvert system value has decreased to $240 million, and the remaining asset service life has been reduced to 37.4 years. Appendix E provides additional information about the example and the detailed analysis of these scenarios, and presents samples of plot charts for reporting the consequences of delayed maintenance of culverts. 3.5 Guardrails Delaying maintenance and necessary replacement of the guardrail system affects the highway system’s overall condition, which directly affects the agency’s future maintenance and replace- ment costs. Guardrail systems in poor condition are considered a roadside hazard; proper design and proper placement greatly affect the intended performance of guardrails in redirecting vehicles or assisting vehicles to come to a complete stop (Sicking et al. 2009). This section describes the procedure to quantify the consequences of delayed maintenance of guardrails. Appendix F provides an example and more details on agency policies for mainte- nance of guardrail systems, as well as performance models, needs analysis, and analysis of each scenario.

Scenario Description Total Agency Cost 1 Discount Agency Cost Backlog Cost 1 Percentage of Culverts in Poor Condition Average Culvert Condition Rating End of Year 20 Critical Year End of Year 20 Critical Year 1 All needs $45.2 million $27.3 million $0 0.0 3.7 (Year 1) 6.34 6.28 (Year 1) 2 Do nothing $0 $0 $57.8 million 25.37 25.37 (Year 20) 4.93 4.93 (Year 20) 3 Delayed maintenance a. 5-year cyclical delayed b. Rehabilitation/replacement $46.0 million $45.7 million $27.2 million $25.6 million $21.0 million $15.2 million $0 $0 $29.2 million 0.0 0.0 16.36 7.65 (Year 5) 11.59 (Year 10) 16.36 (Year 20) 6.35 6.4 5.42 5.92 (Year 5) 5.6 (Year 10) 5.42 (Year 20) 4 Budget driven with limited funds a. 50% of annual baseline maintenance budget ($1.5 million/year) b. 25% of annual baseline maintenance budget ($0.75 million/year) $29.4 million $15.0 million $16.7 million $8.4 million $13.5 million $35.5 million 1.17 10.63 4.63 (Year 20) 10.63 (Year 20) 5.98 5.46 5.98 (Year 20) 5.46 (Year 20) 1 At end of Year 20 Table 15. Summary, scenario analysis results for culverts.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 39 3.5.1 Guardrail Preservation Policy 3.5.1.1 Types of Maintenance Activities Guardrails’ strong association with public safety means repairs are normally performed immediately after an incident report or notification is received. Depending on the damage and the available resources, various maintenance activities can be applied to the guardrail system. Because W-beam guardrails are the most widely used in the United States, the research team chose them as the example to develop procedures for guardrails. Typical preventive and routine maintenance activities for W-beam guardrails involve the following activities: • Preventive maintenance: These activities usually include – Washing and – Tightening of fittings (e.g., screws and bolts). • Routine maintenance: These activities usually include – Replacing post attachment bolts, – Realigning posts damaged by snowplowing, – Applying herbicides along roadside barriers to avoid difficulties involved in mowing grass and weeds along and under the barrier, – Removing dirt build-up, debris, and soil along the guardrail system, and – Removing litter that might interfere with the performance of the guardrail system (Minnesota DOT 2010a). 3.5.1.2 Performance Objectives Performance objectives for specific guardrail systems will depend on the agency’s preserva- tion policy and the performance measures the agency has selected. NCHRP Synthesis of Highway Practice 470: Maintenance Quality Assurance Field Inspection Practices mentions that perfor- mance measures are quantifiable measures “of performance to determine progress toward spe- cific, defined organization objectives based on statistical number evidence” (Zimmerman 2015). Sample measures include the percentage of damage to a guardrail system, and the percentage of the guardrail system rated below standard. LOS and numerical ratings are “expressed as a tangible, measurable goal against which achievement can be compared” (Zimmerman 2015). In practice, most maintenance activities are formulated based on field surveys that are con- ducted to evaluate if the guardrails are still in functional condition. The Utah Department of Transportation (Utah DOT) uses a grading scale based on the percentage of deficient features, as shown in Table 16. Percentage Deficient Grade Percentage Deficient Grade 0.00–3.43 A+ 26.82–30.00 C- 3.44–6.83 A 30.01–33.40 D+ 6.84–10.02 A- 33.41–36.79 D 10.03–13.42 B+ 36.80–39.99 D- 13.43–16.82 B 40.00–43.39 F+ 16.83–20.01 B- 43.40–46.78 F 20.02–23.41 C+ 46.76–30.00 F- 23.42–26.81 C Source: Utah DOT (2012) Table 16. Guardrail system grading scale.

40 Consequences of Delayed Maintenance of Highway Assets The calculated percentage of deficient guardrail length to total guardrail length determines the grade. If 0% to 10.02% of guardrail length in the system is deficient, the system falls into the “A” grade range (with 89.98% to 100% percent of guardrail length in acceptable condition). If 10.03% to 20.01% percent of the guardrail length is deficient, the system falls into the “B” grade range (with 79.99% to 89.97% of guardrail length in acceptable condition). Table 16 shows the remaining grades by percentage range of deficient guardrail length. The guardrail system grading scale approach has been used for the example provided in Appendix F to illustrate the consequences of delaying maintenance of guardrail systems. 3.5.1.3 Decision Criteria for Maintenance Activities Highway agencies typically conduct “periodic review, inspection, and maintenance of in-service traffic barriers” to ensure that the guardrail system functions as intended (Minnesota DOT 2010a). Specific maintenance repairs are identified after conducting field inspections. Maintenance activities for guardrails are tied to a LOS system, and decision criteria to trigger specific activities are based on the agency’s preferred LOS for guardrails. Typically, a guardrail system’s LOS is described in terms of three damage condition states: • Minor: Although the guardrail system may not be aesthetically pleasing, damage to the system is minor, and it will perform its intended function. • Moderate: Despite obvious damage, the guardrail system maintains its structural integrity and will work for most traffic conditions. • Severe: Damage is so severe that the guardrail system no longer functions as designed or has become a hazard to the traveling public (SCDOT 2010). The W-Beam Guardrail Repair Guide (FHWA 2008) mentions that guardrails are replaced if any existing guardrails do not comply with the current U.S.DOT or AASHTO MASH guidelines (AASHTO 2016). Also, when a guardrail terminal is damaged, it should be investigated to deter- mine whether the guardrail terminal needs to be repaired, replaced, or eliminated. 3.5.2 Guardrail Maintenance and Budget Needs 3.5.2.1 Assess the Guardrail System Condition A few highway agencies use a management system that incorporates condition data and cost for maintenance work. Guardrail inspections do not measure asset performance and only indicate whether the guardrail condition passes the performance criteria specified in the agency’s guidelines and specifications. DOTs perform annual visual inspections for height requirements, lateral offset, and completeness of installation of guardrails. Hard copies of the inspection forms typically are kept at maintenance yards, and some maintenance yards may keep a manual tracking system for their inspection forms. In most cases, centralized maintenance records and condition data are lacking for these highway assets; however, a few highway agencies use a management system that incorporates condition data and cost for maintenance work. The Utah DOT uses an Operations Management System (OMS) to track condition data and guardrail maintenance costs. Semiannual inspections identify the percentage of assets that are deficient within a station (section of highway). Based on this percentage, the station is given a Level of Maintenance (LOM) grade of A, B, C, D, or F. No statewide performance target exists, but a target grade (A, B, or C) is established for each maintenance activity. “Once a target LOM is established, the goal is to meet that LOM as closely as possible, neither falling short of the target nor exceeding it” (Utah DOT 2012).

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 41 3.5.2.2 Performance Models for Forecasting Guardrail Condition Guardrail assets are maintained based on condition. For example, the Utah DOT uses regres- sion curves based on performance versus the annual cost to maintain the asset. The regression curves are used to formulate the budget needed to improve or maintain the guardrail system’s performance. The Utah DOT also is starting to consider an evaluation of the safety benefits of guardrail, pavement marking, and sign assets (Utah DOT 2012). Transition matrices are used to model the performance of the highway guardrail system by simulating deterioration or improvement in the system’s condition. Condition categories (the LOM grades A through F) are based on the percentage of deficient guardrail lengths in a sector of highway. The parameters for the transition matrices are obtained from statistical analysis using historical data. Appendix F describes the step-by-step process used to develop the transition matrices. 3.5.2.3 Needs Analysis The needs analysis determines the maintenance activities and budget required to preserve the guardrail system in an acceptable condition. The model identifies needs related to maintenance and replacement of deficient guardrails, and forecasts the guardrail condition for the highway system over the analysis period. Needs are identified based on the guardrail group condition and decision criteria. Transition condition matrices are used to model the change in condition over time. Differing maintenance activities take place depending on the types of damage and length of deficient groups, or sections, in the guardrail system. The model proposed in this project consid- ers that a certain percentage of the guardrail in condition B, C, D, and F receives maintenance as shown in Table 17. Table 18 lists recent costs per foot for guardrail maintenance and replacement activities. Percentage of Guardrail System Rated as Deficient Grade Preservation Activity 0.00–10.02 A No maintenance 10.03–20.01 B 100% maintenance 20.02–30.00 C 60% maintenance, 40% replacement 30.01–39.99 D 20% maintenance, 80% replacement 40.00–100.00 F 100% replacement Table 17. Guardrail system grading scale and preservation activities. Parameter Value (average) Data Source Maintenance cost by lineal foot $7 Utah DOT (2012) Replacement cost by lineal foot $103 U.S. Department of Agriculture (2013) and Texas DOT letting and bids tabulations (2016) Table 18. Costs of preservation activities for highway guardrail.

42 Consequences of Delayed Maintenance of Highway Assets Data Performance Models Maintenance Scenarios (Length of analysis: 10 years) Results Guardrail System Database with Inventory and Condition Assessment Deterioration models based on transition condition probability matrices to model the increase/decrease in deficient guardrails 1. All needs (baseline) 2. Do nothing 3. Delayed maintenance, treatments delayed by: a. 1-year cyclical delay b. 3-year cyclical delay 4. Budget driven with limited funds: a. 80% of annual baseline maintenance budget b. 55% of annual baseline maintenance budget Analytical tools: • Spreadsheet-based model to perform scenario analyses Reports: • Effect on condition due to delayed maintenance • Agency costs over time • Changes in the guardrail system value and sustainability ratio Table 19. Key elements to analyze delayed maintenance scenarios for guardrails. 3.5.3 Consequences of Delayed Maintenance of Guardrails 3.5.3.1 Formulate Delayed Maintenance Scenarios Table 19 describes the key elements needed to analyze delayed maintenance scenarios for guardrail systems. 3.5.3.2 Delayed Maintenance Scenario Analysis Appendix F provides a detailed example and analysis of each scenario. Table 20 summarizes the results of the scenario analysis and shows the 10-year agency costs, backlog in the last year of analysis, and percentage of guardrail groups for which more than 40% of guardrail lengths are deficient (Grade F). 3.5.3.3 Report the Consequences of Delayed Maintenance To quantify the consequences of delayed maintenance, the results of delayed maintenance scenarios are compared to the baseline scenario from the needs analysis. The scenario results shown in Table 20 clearly demonstrate the effects of delaying needed maintenance to the guard- rail system on the condition and the agency costs of future work. The specific results and com- ments reported in the example included in Appendix F are summarized as follows: • Scenario 1, “All needs,” results in a total agency cost of $14.86 million and 8% of guardrail groups rated Grade F (condition category) at the end of Year 10. • Scenario 2, “Do nothing,” results in a cumulative backlog cost of $95.79 million and 100% of the guardrail system rated Grade F at the end of Year 10. • Scenario 3.a, “Delayed maintenance” with activities delayed by 1 year, results in $31.03 million in agency costs for total work performed and $23.56 million in backlogged costs, with 71% of the system rated Grade F at the end of Year 10. • Scenario 3.b, “Delayed maintenance” with activities delayed by 3 years, results in $19.73 million in agency costs for total work performed and $59.7 million in backlogged costs, with 98% of the system rated Grade F at the end of Year 10.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 43 • Scenario 4.a, “Budget driven with limited funds” using 80% of the baseline budget, results in $11.65 million of agency costs for total work performed and $30.84 million in backlogged costs, with 38% of the system rated Grade F at the end of Year 10. • Scenario 4.b, “Budget driven with limited funds” using 55% of the baseline budget, results in $7.96 million of agency costs for total work performed and $55.13 million in backlogged costs, with 62% of the system rated Grade F at the end of Year 10. Appendix F provides additional information about the example and the detailed analysis of each scenario, and presents samples of plot charts for reporting the consequences of delayed maintenance. 3.6 Lighting The purpose of roadway lighting is to provide better nighttime visibility conditions for drivers to improve safety and reduce the risk of nighttime car crashes (Lutkevich, McLean, and Cheung 2012). Delaying maintenance on the lighting system not only affects future agency maintenance and replacement costs, but also affects safety, increasing the likelihood of nighttime crashes. This section describes the procedure to quantify the consequences of delayed maintenance of lighting. Appendix G provides an example and more details on agency policies for lighting maintenance, as well as performance models, needs analysis, and analysis of each scenario. 3.6.1 Lighting Preservation Policy 3.6.1.1 Types of Maintenance Activities Common maintenance activities for lighting can be grouped as follows: • Preventive maintenance: These activities usually involve switchgear, control cabinets (Markow 2007), or cleaning. For luminaires, the frequency of cleaning is calculated based on the luminaire Scenario Description Total Agency Cost 1 Backlog Cost1 Percentage of Guardrail Groups with More than 40% of Guardrail Rated Deficient (Grade F 1) 1 All needs (baseline) $14.86 million $0 8 2 Do nothing $0 million $95.79 million 100 3 Delayed maintenance, treatment delayed by: a. 1-year cyclical delay b. 3-year cyclical delay $31.03 million $19.73 million $23.56 million $59.69 million 71 98 4 Budget driven with limited funds: a. 80% of annual baseline maintenance budget b. 55% of annual baseline maintenance budget $11.65 million $7.96 million $30.84 million $55.13 million 38 62 1 At the end of Year 10 Table 20. Summary of results, guardrail system scenario analysis.

44 Consequences of Delayed Maintenance of Highway Assets dirt depreciation (LDD) factor, which accounts for characteristics such as “luminaire type, mounting height, environment of the luminaire location (i.e., urban or rural setting), traffic volume, and roadway offsets” (AASHTO 2005). • Immediate maintenance: Also called remedial maintenance, these activities often are per- formed to resolve emergency safety hazards such as knockdowns, cable breaks, and switch gear problems (Markow 2007). • Corrective maintenance: These activities usually address fixture failures and any problems that occur with the lamp or ballast (Markow 2007). • Worst-first maintenance: These maintenance activities are performed on “underground breaks from deteriorated systems resulting in failures from salt water and freeze-thaw in winter” (Markow 2007). • Group replacement/routine maintenance: Lighting systems can be scheduled for replacement at specified intervals. For example, in the European Union, where lighting is of high quality and experiences very few outages, the lighting systems typically are relamped every 3 or 5 years (Wilken et al. 2001). As part of group replacement/routine maintenance, older lighting systems may need to be updated to accommodate current light sources, energy conservation standards, and wiring (Illinois DOT 2013 and Minnesota DOT 2010b). For example, there is an incen- tive to replace high pressure sodium vapor (HPS) luminaires with light-emitting diode (LED) luminaires, which are more energy efficient and last longer. 3.6.1.2 Performance Objectives The performance objectives for lighting systems largely depend upon the performance mea- sures that the agency has established. Common performance measures used to establish objec- tives for lighting systems are: • Percentage of lighting in a certain condition, • Lighting age, and • Degree of lighting material degradation. 3.6.1.3 Decision Criteria for Maintenance Activities The decision criteria that trigger lighting maintenance activities fall in two major groups: proactive maintenance and reactive maintenance, as follows: • Proactive maintenance: Agencies that take this approach not only fix failed lamps, but also fix lamps that have a greater probability of failure because another part of the system has fallen below a specified condition threshold. For example, a DOT can plan for rewiring of old direct-bury wires to reduce future failures and proactively retrofit lighting fixtures to LED. • Reactive maintenance: Agencies that take the approach of fixing lamps only when they have failed. Agencies may implement reactive maintenance in different ways. For example, Region 5 of the Colorado Department of Transportation (Colorado DOT) regularly assesses the condi- tion of a percentage of the lighting asset inventory and replaces any lights that are not in good condition. The Texas DOT monitors lighting asset condition remotely on an ongoing basis by monitoring voltage use; a drop in voltage use suggests a knockdown or a burnt-out lamp, trig- gering a maintenance action. In this study, the lighting model simulates reactive maintenance by allowing the user to specify what percentage of failed lamps will be replaced each year. It would be ideal to replace 100% of failed lamps, but in practice the percentage may be lower if maintenance is delayed. The model also simulates proactive replacements by allowing the user to specify the prob- ability threshold at which lamps are proactively replaced. For example, if the user enters a value of 90%, then in any given year, any lamp that has or exceeds a 90% chance of failure will be replaced. The user also can specify what percentage of conventional HPS fixtures will be converted to LED each year.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 45 3.6.2 Lighting Maintenance and Budget Needs 3.6.2.1 Assess the Lighting System Service Life Lighting service life is usually determined based on agency experience, professional judgment, and manufacturer’s data; however, assets often are “repaired or replaced as soon as they fail with- out regard to service life” (Markow 2007). Group relamping based on a lamp mortality curve based on manufacturer’s data is a common maintenance method (Colorado DOT 2006). DOTs perform nighttime drive-by inspections looking for problems such as flickering or knockdowns, usually at intervals less than 3 months (Markow 2007). Highway lighting is monitored more frequently (e.g., every 2 weeks). The median life expectancy for lighting ranges between 25 to 30 years for structural components, 1 to 4 years for lamps, 7 years for ballast, 18 years for control panels and 16 years for luminaires (Thompson et al. 2012). 3.6.2.2 Performance Models for Forecasting Lighting System Service Life Lighting system performance can be estimated based on condition or age. A condition-based approach requires periodic condition assessment inspections to develop deterioration models. An age-based approach estimates the remaining life from historical records of construction and reconstruction. For lighting systems, an age-based approach is frequently the only viable approach; it is often impractical to establish a condition assessment program for lighting, and visual inspection of key components such as lamps can be difficult. Popular performance models used to forecast lighting system service life include: • Exponential functional form (Szary et al. 2005) and • Weibull distribution (Ford et al. 2012a and 2012b). In this study, the lighting model predicts the percentage of lights in operation at a given point. When lights fail, this event reduces energy costs but increases accident costs. The model predicts an increase in the nighttime crash rate of 33% at a given location when lighting is not functioning, which is roughly equivalent to a 25% decrease in nighttime crashes if lighting is added to a previously unlit location. This crash rate estimate is based on a synthesis by Wilken et al. (2001) on the effects of lighting on safety. To use the model, targets are specified for proactive relamping, reactive replace- ment of failed lamps, and conversion of lamps from HPS to LED. The model predicts needs for both the reactive replacement of failed lamps and the funds necessary to achieve the target values during the analysis period. Appendix G provides more information about the model for the lighting system. 3.6.2.3 Needs Analysis Using the lighting model developed for this research, the procedure for the needs analysis is performed as follows: 1. HPS lamps and LED fixtures are grouped by age in years, and calculations are made for each 1-year age bin (e.g., 1-year-old LED lamps, 2-year-old HPS lamps). 2. Three types of needs are considered: (1) needs for replacing failed HPS lamps and LED fixtures; (2) needs for proactively replacing HPS lamps or LED fixtures with a specified probability of failure, and (3) needs for conversion from HPS to LED lamps. 3. The model predicts the number of failed lamps and fixtures for each 1-year age bin as described in Appendix G. The percentage of failed lamps/fixtures replaced is specified as input, as is the average amount of time between the initial failure and lamp/fixture replacement. 4. For each age bin, the model predicts the likelihood of failure within the next year for the HPS lamps and LED fixtures that did not fail within the current year. If the failure likelihood exceeds a specified percentage, then these lamps/fixtures are replaced. 5. The model predicts the number of HPS lights converted to LED. The percentage converted is an input specified by analysis year, and the same value is applied regardless of age. 6. Agency costs are tabulated for reactive replacements, proactive replacements, and conversions to LED.

46 Consequences of Delayed Maintenance of Highway Assets 7. Energy costs are tabulated, accounting for the savings in energy costs from not operating failed lights and the reduced energy costs of operating LED lamps relative to HPS lamps. 8. Crash costs due to failed lights are tabulated, accounting for both the number of failed lights and failure duration. (The model predicts increased crash costs from nighttime crashes based on assumptions detailed in the next section.) 9. Lamp/fixture ages are increased by 1 year, and the analysis is repeated for the next year. The process is repeated until the end of the analysis period. 3.6.3 Consequences of Delayed Maintenance of Lighting 3.6.3.1 Formulate the Delayed Maintenance Scenarios Table 21 summarizes the key elements needed to analyze the set of scenarios recommended for lighting maintenance. Data Performance Models Maintenance Scenarios (Length of analysis: 10 years) Results Lighting System Database Inventory Weibull models for predicting likelihood of lamp or electrical failure Alternatively, straight-line service life (based on original design life) 1. All needs (baseline) a. Failed lamps/fixtures replaced in 2 weeks; HPS replaced with LED over 10-year period; no additional proactive replacements performed b. Failed lamps/fixtures replaced in 2 weeks; HPS replaced with LED over 10-year period; additionally, lamps/fixtures proactively replaced when failure probability exceeds 90% c. Failed lamps/fixtures replaced in 2 weeks; no proactive replacements performed and no additional HPS fixtures converted to LED 2. Do nothing All lamp/fixture replacements deferred for 10 years 3. Delayed maintenance with All lamp/fixture replacements deferred for 5 years; after deferral period, failed lamps replaced in 2 weeks; no proactive replacements performed, and no additional HPS fixtures converted to LED 4. Budget driven maintenance with limited funds: a. Only 90% of failed lamps/fixtures replaced; no proactive replacements performed; no additional HPS fixtures converted to LED b. Only 75% of failed lamps/fixtures replaced; no proactive replacements performed; no additional HPS fixtures converted to LED c. Only 50% of failed lamps/fixtures replaced; no proactive replacements performed; no additional HPS fixtures converted to LED (Replacements completed in 2 weeks) Analytical tools: • Spreadsheet- based model that incorporates probability of failure Reports: • Effect on condition due to delayed maintenance • Agency costs of scheduled and unscheduled maintenance • Agency costs of converting HPS to LED where applicable • Agency energy costs • Increased user accident costs from loss of lighting Table 21. Key elements to analyze delayed maintenance scenarios for lighting.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 47 3.6.3.2 Delayed Maintenance Scenario Analysis Appendix G presents a detailed example that includes analysis of the various maintenance scenarios. Table 22 summarizes the results of all the scenarios with the 10-year agency total cost and percentage of the lighting system in service. 3.6.3.3 Report the Consequences of Delayed Maintenance To quantify the consequences of delayed maintenance to the lighting system, the results of the delayed maintenance scenarios are compared to the baseline scenario from the needs analysis. The scenario results clearly demonstrate the effects of delaying maintenance to the lighting sys- tem, affecting both the system’s condition and the agency costs. Delaying maintenance results in more lighting fixtures out of service. This reduces energy costs, but these savings are more than offset by increased crash costs. The specific results and comments reported in the example detailed in Appendix G are summarized as follows: • Scenario 1.a and Scenario 1.b show that proactive replacement of lamps or fixtures that are likely to fail increases agency costs slightly but reduces overall costs because of the sav- ings in crash costs. Agency replacement costs increase by approximately $0.7 million over 10 years in Scenario 1.b relative to Scenario 1.a, but Scenario 1.b achieves a reduction in overall costs of $0.4 million. • Scenario 1.a and Scenario 1.c, when compared, show that conversion from HPS fixtures to LED fixtures increases costs over a 10-year period. (Over a longer period, the conversion would be more beneficial, given the increased life of LED fixtures compared to HPS; however, this analysis was intended to demonstrate the effects of delaying maintenance rather than the effects of investing in LED conversion.) • Scenario 1.c, Scenario 3, and Scenario 4 together demonstrate that delaying maintenance always results in increased costs in the scenarios evaluated, whether the delay takes the form of increas- ing time to respond to failures, imposing a deferral period, or reducing the percentage of failed lamps/fixtures that are replaced. In Scenario 1.c, failures are replaced but no proactive work is performed, and no HPS fixtures are converted to LED. In Scenario 3, no maintenance work is performed for 5 years, then the same policy is performed as for Scenario 1.c. Scenario Description Agency Replacement Costs 1 Excess Crash Cost 1 Energy Cost 1 Total Cost 1 Discounted Total Cost Percentage in Service Reactive Proactive LED Min. End 1 All needs a. HPS replaced with LED b. Proactive replacement c. 100% of failures replaced $3.64 million $1.80 million $8.95 million $0 $2.47 million $0 $13.53 million $13.53 million $0 $1.95 million $0.88 million $5.03 million $5.50 million $5.50 million $6.22 million $24.64 million $24.20 million $20.21 million $19.13 million $18.73 million $14.78 million 100 100 100 100 100 100 2 Do nothing $0 $0 $0 $305.66 million $3.05 million $308.71 million $215.06 million 55 55 3 Delayed maintenance (5 years) $5.34 million $0 $0 $144.24 million $4.82 million $154.41 million $123.44 million 56 100 4 Budget driven with limited funds: a. 90% of failures replaced b. 75% of failures replaced c. 50% of failures replaced $8.56 million $7.87 million $6.35 million $0 $0 $0 $0 $0 $0 $17.53 million $39.65 million $89.07 million $6.09 million $5.85 million $5.31 million $32.19 million $53.38 million $100.74 million $23.37 million $38.49 million $72.02 million 98 95 87 98 95 88 1 At end of Year 10 Table 22. Summary of results, lighting system scenario analysis.

48 Consequences of Delayed Maintenance of Highway Assets • Scenario 2 simulates a 10-year deferral period, which reduces the agency’s costs by approximately $8.9 million and energy costs by approximately $3.2 million; however, deferring maintenance increases overall life-cycle costs by approximately $200.3 million because of increased crash costs. Also, by the tenth year of deferral, the lighting system value decreases by $13.8 million, as almost half of the lighting system is not in service. • Scenario 3 simulates a 5-year deferral period, which reduces the allocated budget or agency costs of the work performed by approximately $5.5 million and saves approximately $1.4 mil- lion in energy costs. However, deferring maintenance increases the overall life-cycle costs by approximately $108.6 million because of increased crash costs. Also, by the fifth year of deferral, the lighting system value decreases by $13.5 million, as almost half of the lighting system is not in service. Appendix G provides additional information about the example and the detailed analysis of these scenarios, and presents sample plot charts for reporting the consequences of delayed maintenance. 3.7 Pavement Markings The purpose of pavement markings is to provide “guidance and information for the road user” for safer driving conditions (FHWA 2009). Pavement markings include longitudinal lane, shoul- der, and center lines; raised markers; and symbols, guidance, and warning messages found on the surface of the roadway. This section describes the procedure to quantify the consequences of delayed maintenance of pavement markings. Appendix H provides an example and more details on agency policies for maintenance of pavement markings, as well as performance models, needs analysis, and analysis of each scenario. 3.7.1 Pavement Marking Preservation Policy An agency’s preservation policy for the pavement marking system usually is formulated by a central office that provides policies for maintenance, specifications for materials, and criteria to allocate funding. FHWA has also established standards in the Manual on Uniform Traffic Control Devices (MUTCD) regarding minimum retroreflectivity requirements for pavement markings (FHWA 2010). 3.7.1.1 Types of Maintenance Activities Most maintenance activities for pavement markings fall into two categories—routine or corrective—depending on the type of material used to create the markings: • Routine maintenance: These activities are performed on a routine basis to preserve the LOS of the pavement marking system. • Corrective maintenance: These activities include repairs and replacement of individual ele- ments. The decision to replace a pavement marking is mainly condition-driven, but it also can occur in response to a change in design requirements. 3.7.1.2 Performance Objectives The performance objectives for maintenance of pavement markings depend on the pave- ment marking standards, warrants, and design criteria that have been established in the MUTCD (FHWA 2009; FHWA 2010). Additional information can be found in the FHWA Roadway Delin- eation Practices Handbook (Migletz et al. 1994). Typically, pavement markings are assessed for ret- roreflectivity following the MUTCD standards (Markow 2007). The markings can be categorized into three general groups, defined as follows (FHWA 2010):

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 49 • “Not required to be retroreflective”: Pavement markings that occur where ambient illumina- tion assures adequate visibility or where the markings are needed only during the day (e.g., at a park where access is restricted to daytime hours) “do not require minimum levels of retro- reflectivity.” All other pavement markings must meet some level of retroreflectivity. • “Required to be retroreflective, but not subject to minimum levels”: Some retroflective pavement markings are not subject to the minimum retroreflectivity levels established by the MUTCD (e.g., crosswalk markings, other transverse markings, words, symbols, and arrows). Some longitudinal lines also are exempt from minimum retroreflectivity levels under cer- tain conditions (e.g., presence of continuous roadway lighting or raised retroreflective pavement markers); however, “pavement markings on Interstate Highways are required to be retroreflective.” • “Subject to minimum retroreflectivity levels”: Pavement markings that are subject to the minimum retroflectivity levels specified in the MUTCD include the white and yellow longi- tudinal pavement markings (e.g., center lines, edge lines, lane lines, and channelizing lines) that the MUTCD says “shall or should be used” on roads “above certain volumes or for certain roadway conditions.” The MUTCD defines the minimum levels of retroreflectivity for certain pavement markings based on roadway type, posted speed, and weather conditions, but allows that “during winter months in northern climates, along some isolated horizontal curves, [or] near driveways” the minimum retroreflectivity levels may not be reached. Performance measures for pavement markings can be either physical and measured in units (e.g., mcd/m2/lux for retroreflectivity), or qualitative and assessed on a scale (e.g., “good, fair, poor,” or “A, B, C, D, F”) (Markow 2007). Markow (2007) identifies the following common performance measures that are used in pavement markings to establish objectives: • Loss of retroreflectivity, • Pavement marking age and estimated remaining life, • Percentage of pavement markings at certain LOS, • Percentage of broken or missing raised pavement markers, • Deterioration due to abrasion or wear, and • Customer complaints. 3.7.1.3 Decision Criteria for Maintenance Activities Most pavement marking maintenance activities are formulated based on field surveys that have determined the pavement marking condition or LOS. The decision criteria consider LOS categories based on retroreflectivity measurements. In this report, the decision criteria for pavement marking maintenance activities are presented in terms of a letter-grade system (Table 23). Condition (LOS) Category Lower Retroreflectivity Limit A 319 mcd/m²/lux B 263 mcd/m²/lux C 207 mcd/m²/lux D 150 mcd/m²/lux F Below 150 mcd/m²/lux Source: AASHTO (2007) Table 23. Pavement marking condition categories based on retroreflectivity.

50 Consequences of Delayed Maintenance of Highway Assets Pavement markings below 150 mcd/m2/lux (Category F) are scheduled for repainting to restore retroreflectivity to 375 mcd/m2/lux (well above the lower retroreflectivity limit for Category A), as mandated by FHWA. Maintenance activities for pavement markings are recommended to be scheduled in coordination with pavement treatment activities needed in the same road segment (AASHTO 2007). 3.7.2 Pavement Marking Maintenance and Budget Needs 3.7.2.1 Assess the Pavement Marking Condition The MUTCD provides general guidelines on condition assessment for pavement markings. Visual inspections using handheld or mobile retroreflectometers are performed in daytime and nighttime (FHWA 2009). The service life expectancy of pavement markings usually is determined based on a combi- nation of agency experience, professional judgment, and manufacturer’s data. Several factors affect pavement marking durability and retroreflectivity, including wet thickness of paint during application. Typical life expectancy ranges for pavement markings (Markow 2007) are: • Non-epoxy paint, 6 months to 2 years; • Epoxy paint, 1 to 5 years; • Thermoplastic, 2 to 10 years; • Cold plastic, 1 to 10 years; and • Tape, 5 to 10 years. 3.7.2.2 Performance Models for Forecasting Pavement Marking Condition Pavement marking performance models are based on condition or age. A condition-based approach requires periodic condition assessments to develop reliable deterioration models, and an age-based approach estimates the remaining life of the pavement marking from historical records. Performance models that can be used to forecast pavement markings condition include: • Exponential functional form, • Weibull distribution, and • Straight-line deterioration. In this study, the pavement marking model follows a straight-line deterioration trend and predicts the LOS categories (A, B, C, D, F) based on retroreflectivity measurements or RSL. The model captures the percentage of the system in each category due to a given maintenance preservation policy. In the model, the user specifies whether the analysis is based on condition or age, and specifies the available budget over the period of analysis. The model calculates the budget needs over the analysis period for repainting pavement markings with retroreflectivity values below 150 mcd/m2/lux (condition categories D and F). In the case of age-based analyses, pavement markings with a RSL 20% below the initial service life require new painting. A pave- ment marking median service life of 4 years is assumed in the model based on previous studies (Markow 2007) and current data analysis. The following steps summarize the procedure used in the straight-line pavement marking deterioration model in this study: 1. Extract data from the pavement marking inventory to analyze deterioration and improve- ment trends for all the pavement marking sections in the inventory. For the model described in this chapter, the minimum data include the total length for each pavement marking sec- tion, pavement marking color, pavement marking material, and retroreflectivity. This step is done for all the years in the inventory.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 51 2. The deterioration rate is based on minimum retroreflectivity at the beginning of the service life, retroreflectivity at the end of functional service life (Bahar et al. 2006), and the mean service life. Straight-line deterioration is assumed to model the condition deterioration for six combinations with two colors (white and yellow) and three pavement marking materials (epoxy, polyurea, and thermoplastic). 3. Based on retroreflectivity, the pavement marking system is divided into five condition catego- ries (A, B, C, D, and F). Categories A, B, and C comply with FHWA minimum retroreflectiv- ity levels. Category D indicates that pavement markings will soon not meet the minimum retroreflectivity level. Category F means that a pavement marking is below the minimum retroreflectivity threshold. The following equation is used to model the retroflectivity changes along the service life: ,1 0 R R R R SL n n end = − − where Rn = lower limit for retroreflectivity that classifies as condition Category A (n = 1), B (n = 2), C (n = 3), D (n = 4), F (n = 5), R0 = assumed retroreflectivity at the beginning of service life, Rend = minimum required retroreflectivity by FHWA, and SL = expected mean service life. 4. Each condition category has an associated maintenance activity. If pavement marking groups are in condition categories A, B, or C, no treatment is applied. Pavement marking sections in condition categories D and F are fully repainted. Appendix H provides more information about the model for the pavement marking system. 3.7.2.3 Needs Analysis The needs analysis identifies maintenance and budget needs to preserve the pavement mark- ing system in an excellent condition with high visibility and retroreflectivity during both daytime and nighttime. Maintenance criteria are based on minimum retroreflectivity standards. “Paint” is applied to restore retroreflectivity when a section of pavement marking reaches Category D or F. Table 24 shows a cost estimate for repainting pavement markings. If retroreflectivity data are not available, the RSL approach can be used to determine whether repainting is necessary. RSL is estimated from current retroreflectivity, initial retroreflectivity (375 mcd/m2/lux), and end-of-life retroreflectivity (150 mcd/m2/lux), assuming an expected service life of 4 years. Source: Data provided by North Carolina DOT Pavement Marking Type Cost ($/ft.) White, epoxy 0.39 White, polyurea 0.78 White, thermoplastic 0.61 Yellow, epoxy 0.34 Yellow, polyurea 0.79 Yellow, thermoplastic 0.61 Table 24. Cost of repainting pavement markings.

52 Consequences of Delayed Maintenance of Highway Assets 3.7.3 Consequences of Delayed Maintenance of Pavement Markings 3.7.3.1 Formulate Delayed Maintenance Scenarios Table 25 summarizes the key elements needed to analyze the set of scenarios recommended in this study for pavement markings. 3.7.3.2 Delayed Maintenance Scenario Analysis Appendix H provides an example, together with a detailed analysis of each scenario. Table 26 summarizes the results of these analyses with the 5-year agency costs, backlogged in the last year of analysis, and the percentage of the pavement marking system below the minimum retroreflectivity requirements. 3.7.3.3 Report the Consequences of Delayed Maintenance At the beginning of the analysis, the percentage of length of pavement markings below the minimum retroreflectivity is 29%. Delaying maintenance activities increases the number of pavement markings in condition Category F and affects other performance measures. The specific results and comments reported in the example detailed in Appendix H are summarized as follows: • Scenario 1, “All needs,” involves repainting of pavement markings once they have reached their service life and results in an allocated budget of $25.22 million for total agency costs, a pavement marking system in good condition (42% in Category A), no backlog costs, and a system value of $11.7 million. • Scenario 2, “Do nothing,” results in $16.6 million in backlog costs, a pavement mark- ing system in poor condition (86% in Category F), and a decrease in the system value of $9.1 million. • Scenario 3.a, “Delayed maintenance” with a 1-year cyclical deferral period, reduces the agency’s costs by approximately $10 million, but results in an increase of approximately $4.5 million in the unfunded backlog. • Scenario 3.b, “Delayed maintenance” with a 3-year cyclical deferral period, reduces the agen- cy’s costs by approximately $16 million, but the delayed maintenance results in an increase of approximately $7.4 million in unfunded backlog, and the system value decreases by $2.6 million, Data Performance Models Maintenance Scenarios (Length of analysis: 5 years) Results Pavement Marking System Database with Inventory and Condition Assessment Weibull models for predicting pavement marking retroreflectivity failure Straight-line deterioration model Transition probability matrices to model the increase/decrease of deficient signs 1. All needs (baseline) 2. Do nothing 3. Delayed maintenance with treatments delayed by: a. 1-year cyclical delay b. 3-year cyclical delay 4. Budget driven with limited funds a. 80% of annual baseline maintenance budget b. 40% of annual baseline maintenance budget Analytical tool: • Spreadsheet-based model to forecast pavement marking condition categories over the period of analysis Reports: • Effect on condition due to delayed maintenance • Agency costs over time • Changes in pavement marking system value and sustainability ratio Table 25. Key elements to analyze delayed maintenance scenarios for pavement markings.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 53 as approximately 38% of the pavement system is below required retroreflectivity levels at the end of Year 5. • Scenario 4.a, “Budget driven with limited funds” using 80% of the baseline budget, reduces the agency’s costs by approximately $5.3 million, but unfunded backlog increases by approxi- mately $3.6 million due to delayed maintenance, and 19% of the pavement marking system is below required retroreflectivity levels at the end of Year 5. • Scenario 4.b, “Budget driven with limited funds” using 40% of the baseline budget, reduces the agency’s costs by approximately $15.2 million, but unfunded backlog increases by $10 mil- lion due to delayed maintenance, and 51% of the pavement marking system is below required retroreflectivity levels by the end of Year 5. Appendix H provides additional information about the example and a more detailed analysis of each scenario, and presents samples of plot charts for reporting the consequences of delayed maintenance. 3.8 Highway Signs Highway signs can be categorized based on their function as regulatory, warning, and guide signs. The purpose of highway signs is to “communicate the rules, warnings, guidance, and other highway agency information that drivers, bicyclists, and pedestrians need to safely and efficiently navigate roads and streets” (McGee 2010). Signs that are in poor condition compromise the safety of road users. Good condition of highway signs is crucial for traffic safety, and deferring maintenance can result in highway signs with reduced retroreflectivity. This section describes the procedure to quantify the consequences of delayed maintenance of highway signs. Appendix I provides an example and more details on agency policies for highway sign maintenance, as well as performance models, needs analysis, and analysis of each scenario. Scenario Description Total Agency Costs 1 Backlog Cost 1 Percentage of Pavement Marking System below Minimum Retroreflectivity End of Year 5 Critical Year 1 All needs (baseline) $25.22 million $0 0 2 (Year 2) 2 Do nothing $0 $16.60 million 86 86 (Year 5) 3 Delayed maintenance a. 1-year cyclical delay b. 3-year cyclical delay $15.15 million $14.06 million $4.48 million $6.85 million 22 38 37 (Year 1) 50 (Year 2) 4 Budget driven with limited funds a. 80% of annual baseline maintenance budget b. 40% of annual baseline maintenance budget $19.95 million $10.04 million $3.62 million $9.95 million 19 51 19 (Year 5) 51 (Year 5) 1 At end of Year 5 Table 26. Summary of results, pavement marking system scenario analysis.

54 Consequences of Delayed Maintenance of Highway Assets 3.8.1 Highway Sign Preservation Policy Sign retroreflectivity preservation policies must correspond to the requirements of the MUTCD (FHWA 2009). ASTM standards differentiate signs as Type I, II, III, IV, VIII, IX, or XI (Carlson and Lupes 2007). In practice, most DOTs combine national standards with their own guidelines to develop sign maintenance practices. 3.8.1.1 Types of Maintenance Activities Maintenance activities for highway signs can be classified as preventive or corrective (Markow 2007). Activities discussed by McGee (2010) can be grouped as follows: • Preventive maintenance: This category includes sign cleaning, vegetation control, anti-theft measures, and sign support adjustments. • Corrective (immediate) maintenance: These activities are applied to signs that need to be repaired or replaced immediately due to events such as vandalism, vehicle collision, damage by natural forces, or a sign having reached the end of its service life. Because the absence or poor condition of regulatory signs (e.g., STOP signs) “could result in or contribute to a severe crash,” they need to be “replaced or repaired within hours of the agency having notice of them missing, down, or damaged.” 3.8.1.2 Performance Objectives The MUTCD’s required minimum retroreflectivity levels vary, depending on the sign’s sheet- ing type and symbol type. The main objective of sign preservation activities is to maintain the sign’s retroreflectivity above the minimum threshold. In selecting sign performance measures, it is important to consider the causes of failure, such as (Markow 2007): • Decrease in retroreflectivity, • Color fading, • Decrease in daytime/nighttime legibility, • Diminished structural condition, • Corrosion, • Dirt accumulation, • Vandalism (e.g., graffiti, bullet holes), and • Age. At the network level, condition indicators for signs can include percentage of signs faded; percentage of signs that are not straight, have damaged posts, or have breakaway devices that are not working; and percentage of signs not readable at night. A condition indicator rating for retroreflectivity readings (mcd/m2/lux) that ranges from 4 (greater than 200 mcd/m2/lux) to 0 (49 mcd/m2/lux or less) can be used to compute a letter grade from “A” to “F” for the mainte- nance LOS (FHWA 2009; see Appendix I). 3.8.1.3 Decision Criteria for Maintenance Activities Based on the Asset Management Data Collection Guide (AASHTO-AGC-ARTBA 2006) main- tenance activities can be grouped as follows: • Preventive maintenance: These maintenance activities are used only for signs in good condi- tion (comply with MUTCD standards and local standards); • Repair: These activities address needs such as missing bolts, signs that are leaning or have damaged posts, or are in need of cleaning; and • Replacement: These activities are triggered when signs have been damaged or become illegible. In this study, the decision criteria for maintenance activities are based on a letter-grade system. Condition categories (A, B, C, D, F) are based on percentage ranges of deficient signs

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 55 within a group or sector (Table 27). A deficient sign is defined as having a problem with the sign face or post that can be fixed by a maintenance or replacement activity. For sign groups in Category A (less than 5% of signs are deficient), no maintenance action needs to be taken. Sign groups in Category B (roughly 5% to 10% of signs are deficient) will require some maintenance to return all signs to a non-deficient state. For sign groups in categories C and D (between 10% and 20% of signs are deficient), significant numbers of signs will require maintenance or replacement. Lastly, sign groups in Category F require the replacement of all deficient signs. 3.8.2 Highway Sign Maintenance and Budget Needs 3.8.2.1 Assess the Sign System Condition and Service Life The MUTCD provides general guidance for condition assessment of signs; however, it does not mention survey frequency (Re and Carlson 2012). Most DOTs do not maintain sign service life information in their databases, and a common management practice is to identify the percentage of deficient signs in a sector without recording details of the condi- tion of individual signs. Appendix I provides additional information on condition assess- ment methods. 3.8.2.2 Performance Models for Forecasting Highway Sign Condition Age, weather conditions, light exposure, and material type all affect the deterioration of signs. To model the probability of failure, a Weibull distribution can be adopted to estimate the signs’ remaining life. Alternative approaches are to use a straight-line deterioration model or use tran- sition condition matrices to simulate the deterioration or improvement in the condition of signs over time. In this study, the performance of the highway sign system has been modeled using transi- tion matrices and condition categories based on the percentage of deficient signs in a sector. The parameters for the transition matrices were obtained from statistical analysis of historical data. Appendix I details the step-by-step process used to develop the transition matrices from condition data. 3.8.2.3 Needs Analysis The needs analysis determines preservation activities and a budget required to maintain the sign system in an acceptable condition. The model identifies needs for maintenance and replace- ment of deficient signs over the analysis period. Needs are identified based on the signs’ condi- tion categories and the decision criteria. Transition condition matrices are used to model the changes in sign groups’ condition over time. Condition Category Percentage of Deficient Signs in a Group Lower Limit Upper Limit A 0.00 5.00 B 5.01 10.00 C 10.01 14.99 D 15.00 19.99 F 20.00 100.00 Table 27. Sign condition categories based on percentage of deficient signs.

56 Consequences of Delayed Maintenance of Highway Assets Each condition category has an associated treatment. Sign groups in Category A receive “Do nothing” treatment. For sign groups in Category B, the associated treatment is that 100% of the deficient signs receive maintenance treatments. For sign groups in Category C, the associated treatment is that 60% of the deficient signs receive maintenance and 40% percent are replaced. For sign groups in Category D, the associated treatment is that 20% of the deficient signs receive maintenance treatments and 80% of the deficient signs are replaced. For sign groups in Category F, the associated treatment is that 100% of the deficient signs are replaced. The average cost of sign maintenance and replacement is shown in Table 28. 3.8.3 Consequences of Delayed Maintenance of Highway Signs 3.8.3.1 Formulate Delayed Maintenance Scenarios Table 29 summarizes the key elements needed to analyze the set of scenarios recommended in this study for the highway signs system. 3.8.3.2 Delayed Maintenance Scenario Analysis Appendix I provides an example and detailed analysis of these scenarios. Table 30 summarizes the analysis results, including the total agency costs in 10 years, the backlog cost at the end of the analysis period, and the percentage of sign groups projected to have more than 20% deficient signs at the end of the analysis period. Parameter Value 1 Average cost of sign maintenance $121 Average cost of sign replacement $249 1 Estimates based on data from the Utah DOT Table 28. Costs of preservation activities for highway signs. Data Performance Models Maintenance Scenarios (Length of Analysis: 10 years) Results Signs Inventory Database with Inventory and Condition Assessment Transition probability matrices to model the increase/ decrease of deficient signs 1. All needs (baseline) 2. Do nothing 3. Delayed maintenance with treatments delayed by a. 1-year cyclical delay b. 3-year cyclical delay 4. Budget driven with limited funds using a. 80% of annual baseline maintenance budget b. 60% of annual baseline maintenance budget Analytical Tools: • Spreadsheet-based model to forecast sign condition categories over the period of analysis Reports: • Effect on condition due to delayed maintenance • Agency costs over time • Changes in sign system asset value and sustainability ratio Table 29. Key elements to analyze delayed maintenance scenarios for the highway sign system.

Procedures to Quantify Consequences of Delayed Maintenance of Highway Assets 57 3.8.3.3 Report the Consequences of Delayed Maintenance At the beginning of the analysis, 42% of the highway system sign groups are in Category A, and 15% percent are in Category F. Delaying maintenance activities results in increased needs over time and increased numbers of signs in Category F. The specific results and comments for the example reported in more detail in Appendix I are summarized as follows: • Scenario 1, “All needs,” results in agency costs of $7.8 million, a sign system in good condition (52% in Category A and 40% in Category B), no backlog costs, and a system value of $18.3 million. • Scenario 2, “Do nothing,” results in $97.2 million in backlog costs and 100% signs in poor condition (Category F) at the end of Year 10. • Scenario 3.a, “Delayed maintenance” using a 1-year cyclical delay, increases the agency’s costs to $13.5 million but results in a significant increase in the percentage of signs in condition Category F (ranging between 17% and 30%) during the analysis period. • Scenario 3.b, “Delayed maintenance” using a 3-year cyclical delay, best illustrates the effects of delaying all investments on signs. In this case, agency costs increase from $7.8 million to $17.3 million, while signs in poor condition increase to 92% at the end of Year 10. • Scenario 4.a, “Budget driven with limited funds” using 80% of the baseline budget, shows results similar to Scenario 3.a. • Scenario 4.b, “Budget driven with limited funds” using 60% of baseline budget, illustrates that a cut in the budget reduces spending to $4.6 million, but at the cost of a $30.4 million backlog and worsening the sign system condition (40% are in Category F at the end of Year 10). Appendix I provides additional information about the example and a more detailed analysis of each scenario, and presents sample plot charts for reporting the consequences of delayed maintenance. 3.9 Suggestions for Implementing the Procedures To implement the procedures described in this chapter, the following actions are suggested: • Recognize the need for this type of analysis and make the decision to proceed. Although the pres- ervation of all assets typically falls under the supervision of the highway agency’s maintenance Scenario Description Total Agency Cost 1 Backlog Cost 1 Percentage of Sign Groups Having More than 20% of Signs Deficient (Category F) 1 1 All needs (baseline) $7.8 million $0 1 2 Do nothing $0 million $97.2 million 100 3 Delayed maintenance with a. 1-year cyclical delay b. 3-year cyclical delay $13.5 million $17.3 million $7.8 million $42.1 million 18 92 4 Budget driven with limited funds using a. 80% of annual baseline maintenance budget b. 60% of annual baseline maintenance budget $6.2 million $4.6 million $17.1 million $30.4 million 19 40 1 At end of Year 10 Table 30. Summary of results, sign group scenario analysis.

58 Consequences of Delayed Maintenance of Highway Assets department or division, responsibility for managing each asset group may fall under different agency offices (e.g., bridges may be managed by a Bridge Office, pavements by a Materials or Design Office). For state highway agencies, regional or local maintenance offices typically are responsible for budget expenditures. The central office establishes the design standards and specifications, and may perform quality control checks. Most agencies also experience resource limitations, often delegating allocation decisions to the regional offices that have the responsi- bility to make maintenance decisions. Implementation of the delayed maintenance procedures need not begin at the top of the agency hierarchy and encompass all asset groups; doing so would be most efficient, but smaller scale pilot studies are beneficial too. • Integrate the agency’s asset preservation policies into an overall set of asset management goals and objectives. The integration of these procedures should improve communication across management levels about the consequences of delayed maintenance. In this context, the research team recommends implementation of an integrated database that includes all the highway asset groups. • Conduct periodic updates of the asset group inventory, condition assessment, and service life data. These procedures rely on up-to-date records with reliable maintenance and inspection data to develop or calibrate the performance models. Agencies often prefer to use their own performance models, given that jurisdictions are subject to differing design standards, regula- tions, product availability, climates, and traffic conditions. Agencies without agency-specific performance models may find it beneficial to use national average performance data until agency-specific information is collected. For developing performance models and assessing the condition of the asset group, the research team suggests that agencies monitor installa- tion dates and dates of performed maintenance, and collect asset condition data on an annual basis. Some highway assets (e.g., guardrails) may be difficult to rate by condition; however, for most highway assets, indices that measure asset condition to varying degrees (i.e., good, fair, poor) are generally more useful than simple “Pass/Fail” adequacy criteria. Guidelines for assessing condition in each major asset group are provided in this report. • Select appropriate analytical tools to perform the needs and scenario analyses. Tools selected by the highway agency should have the capabilities to forecast the asset group condition and/ or service lives, and show the effect on future budget needs and backlog costs. Again, this report provides a discussion of available analytical tools and a sample scenario analysis for each asset group. Appendices C through I provide additional details. • Select reports that facilitate the interpretation of the scenario analysis results. These results should provide agencies with additional information to make better-informed decisions when preparing their preservation programs. Once the preservation program is implemented, peri- odic monitoring of the results will assist in determining if the asset management goals and objectives are being met or if program adjustments are needed.

Next: Chapter 4 - Research Contributions and Suggestions »
Consequences of Delayed Maintenance of Highway Assets Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Research Report 859: Consequences of Delayed Maintenance of Highway Assets presents a process for quantifying the consequences of delayed maintenance of highway assets that considers the asset preservation policy, the maintenance and budget needs, and the analyses of delayed maintenance scenarios. This process considers delayed maintenance caused by the inability to meet the agency-defined application schedule or the unavailability of the funds required to perform all needed maintenance, and expresses the consequences in terms of asset condition and the costs to owners and road users. Detailed descriptions of the use of the proposed process to quantify the consequences of delayed maintenance for seven highway assets are available in appendices from the contractor’s final report:

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!