Decision Making for Repair Versus Replacement of Highway Operations Equipment (2023)

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.

P A R T I Research Report

7   C H A P T E R   1 1.1 Purpose of the Research Report The Research Report (Part I) summarizes the research team’s approach to developing the resources in NCHRP Project 13-08, “Guideline for Decision-Making for Repair vs. Replacement of Highway Maintenance Equipment.” The Research Report is intended to be used in conjunc- tion with the Guide (Part II), User Manual (Part III), and the Repair, Replace, Rebuild, Retire Tool (the 4R Tool). 1.2 Summary The purpose of NCHRP Project 13-08 was to develop resources for state transportation agency fleet managers and others that (1) help them to evaluate options and to document decision making when faced with a downed piece of equipment, and (2) assist them to judge which options represent the best value for the fleet and the agency. This research included a literature review, survey, and interviews to identify the current prac- tices for managing downed equipment. The research showed that state DOTs employ a wide variety of approaches and processes to address downed equipment decisions; there is no single process that has been widely accepted and integrated across state agencies. The variation of data analysis practices and procedures revealed a need to develop a guide to integrate both qualita- tive and quantitative metrics to support the decision to repair, rebuild, replace, or retire downed equipment. DOT fleet managers need a resource to evaluate the pros, cons, and situational merits when addressing (1) the management decisions relevant to the planned repair versus replacement decisions, and (2) the decisions relevant to unexpected catastrophic equipment failures and loss. The research team synthesized available data from a literature review, surveys, and interviews to establish a framework. From this framework, the research team developed the 4R Tool; “4R” stands for repair, rebuild, replace, or retire. The 4R Tool has two independent modules: (1) the Decision Module and (2) the Economic Analysis Module. Each is described in detail in Part III of this report. 1.3 Organization of the Research Report Chapter 2 describes the research approach, including the approach to the survey, interviews, literature review, and development of deliverables. Chapter 3 describes the outcomes of the research project. Introduction

8 2.1 Overview The research team’s approach to NCHRP Project 13-08 included data collection activities, framework development, and tool development. Data collection activities included a compre- hensive literature review, surveys, and interviews. The purpose of these activities was to identify current practices, understand methodological gaps, and ultimately to ensure that the tool and resources developed in this project aligned with the users’ needs. Based on the literature review, surveys and interviews, the research team developed a framework for decision making and an Excel-based tool known as the 4R Tool. 2.2 Literature Review Approach A central goal of the literature review was to identify the criteria and methods that fleet managers use to determine what to do with a downed piece of equipment. The research team reviewed more than 40 resources, seeking to identify the most applicable and relevant sources. These included the following: • Academic publications and journal articles • Government guidebooks and handbooks • Conference presentations • Industry guidebooks and publications • Fleet software and consulting websites • Blog articles The research team developed an annotated bibliography (Appendix D), which summarizes the most relevant literature reviewed. 2.3 Survey Approach The research team developed a short online survey using the SurveyMonkey platform and sent the survey to fleet managers in state DOTs. The survey included 10 questions, which are shown in Appendix A. Questions covered topics such as the quantitative versus qualitative nature of decision making in the repair versus replace decision, staff responsibility for making the decisions, and factors that influence the decisions. The survey went to all DOT fleet contacts on the contact list that was maintained through the Equipment Management Technical Services Program. The list of respondents is shown in Table 1. Twenty-eight completed survey responses were received. Results from the survey are shown in Appendix B. Research Approach C H A P T E R   2

Research Approach 9 2.4 Interview Approach The research team conducted 11 interviews with fleet managers as shown in Table 2. Inter- views allowed the research team to probe for deeper responses on certain topics than the surveys could provide. The guide that was used for these interviews is provided in Appendix C. The research team conducted interviews with DOT fleet managers and with one private sector fleet entity in order to obtain some comparative information. 2.5 Framework Development Approach In developing the framework, the research team considered two possible approaches based on the literature review, survey, and interviews: (1) a qualitative decision tree and (2) a purely quantitative approach. The research team evaluated these two approaches using the following considerations: • Ability to properly reflect the complexity of the decision, stemming from the wide range of influencing parameters. • Ability to meet the specifications, criteria, and expectations of equipment use. • Ability to maintain flexibility to meet the needs of a diverse set of DOTs. • Ability to provide a consistent and defensible methodology to support implementation. Connecticut Department of Transportation (CTDOT) Northeast Delaware Department of Transportation (DelDOT) Northeast Florida Department of Transportation (FDOT) Southeast Idaho Transportation Department (ITD) Midwest Iowa Department of Transportation (IOWADOT) Midwest Kansas Department of Transportation (KDOT) Midwest Kentucky Transportation Cabinet (KYTC) Southeast Louisiana Department of Transportation and Development (LaDOTD) Southeast Maryland Department of Transportation (MDOT) Northeast Minnesota Department of Transportation (MnDOT) Midwest Montana Department of Transportation (MDT) Midwest Nebraska Department of Transportation (NbDOT) Midwest Nevada Department of Transportation (NDOT) West New Jersey Department of Transportation (NJDOT) Northeast North Carolina Department of Transportation (NCDOT) Southeast Oregon Department of Transportation (ODOT) West Pennsylvania Department of Transportation (PennDOT) Northeast South Carolina Department of Transportation (SCDOT) Southeast Tennessee Department of Transportation (TDOT) Southeast Texas Department of Transportation (TxDOT) Southeast Utah Department of Transportation (UDOT) West Vermont Agency of Transportation (VTrans) Northeast Virginia Department of Transportation (VDOT) Southeast Affiliation Region Alabama Department of Transportation (ALDOT) Southeast Arizona Department of Transportation (ADOT) West Arkansas Department of Transportation (ARDOT) Midwest California Department of Transportation (Caltrans) West Colorado Department of Transportation (CDOT) West Table 1. Survey respondents.

10 Decision Making for Repair Versus Replacement of Highway Operations Equipment The selected framework integrates useful elements of both qualitative (e.g., slider bars with user-chosen inputs) and quantitative (e.g., numerically derived estimates such as lifetime cost) approaches in a single methodology. This framework was used to develop the 4R Tool. 2.6 Tool Development Approach The 4R Tool was developed using an iterative process. Version 1.0 of the 4R Tool was designed using the initial model framework, which used a linear structure for evaluation of four possible outcomes: repair, rebuild, replace, and retire. Based on testing and validation of the Version 1.0 tool, it was determined that new model architecture was needed to improve the user experience and the adequacy of the guidance generated by the tool’s analysis. Version 2.0 of the 4R Tool included key changes that improved the tool’s functionality. The architecture of the 4R Tool was redeveloped, moving from a linear repair, rebuild, replace, or retire decision framework to a gated framework. Rather than producing a repair, rebuild, replace, or retire recommendation based on a set of inputs, the user is instead directed to make a repair versus no repair determination. This is followed by a recommendation of either “rebuild or repair” or a “replace or retire” decision. This structure more closely aligns with the decision path that fleet managers take in real-world applications. Table 2. List of interviewees. Affiliation Region Alabama Department of Transportation Southeast American Transportation Research Institute of the American Trucking Association Nationwide Arizona Department of Transportation West California Department of Transportation West Connecticut Department of Transportation Northeast Kansas Department of Transportation Midwest Kentucky Transportation Cabinet Southeast Minnesota Department of Transportation Midwest Pennsylvania Department of Transportation Northeast Texas Department of Transportation Southeast Utah Department of Transportation West

11   3.1 Introduction This chapter presents the findings of the literature review, survey, and interviews. Additionally, the chapter discusses how the information was refined to develop a framework and prototype 4R Tool. Finally, it presents the process for testing and validation of the prototype tool to produce the final 4R Tool and the Guide provided as Part II of this report. 3.2 Literature Review The literature review revealed that most academic methodologies assess decisions about a downed piece of equipment using a purely quantitative approach. The literature focuses on logi- cal outcomes, objective data, and unconstrained time for analysis. The literature ignores many real-world constraints, such as funding availability, return to service time limits, and replace- ment availability. The literature review focused on understanding the existing approaches fleet managers use when faced with a downed piece of equipment. Two dominant approaches include the “economic life- time” approach and the “repair limit” approach. The Guide in Part II provides more information about these two approaches as well as a third approach used in the 4R Tool (the “hybrid” approach). The literature review also highlighted the lack of consistency in approaches across fleets. For example, Lauria and Lauria (2014) surveyed 38 state DOT fleet managers and asked about fleet replacement management practices. Despite the intention to use purely quantitative methods in life cycle cost analyses (LCCAs), the survey found that DOTs often rely on fleet manager judgment when making replacement decisions. This practice was linked anecdotally to DOTs’ keeping of assets beyond their target economic lifetime. For example, 50 percent of surveyed fleets reported keeping assets for over 20 years. Accordingly, the study concluded that there are no standard decision support tools used to justify decisions, nor are there standard methods to secure funding. Additional findings from the literature review appear in Part II of this report. 3.3 Surveys and Interviews Through an online survey and interviews, fleet managers provided information on the cur- rent structure and practices of repair versus replace decision making for highway equipment. The goal of the survey and interviews was to understand relevant criteria that fleet operators use in their current decision-making practices, whether a formal process was followed, who was involved in that process, centralization of decision making and structure, guidelines and tools that were employed, and how data were managed and used to inform decision making. Research Outcomes C H A P T E R   3

12 Decision Making for Repair Versus Replacement of Highway Operations Equipment Each of the agencies interviewed indicated that they integrate some level of data or infor- mation into their decision-making processes for repairing and replacing highway maintenance equipment. Some agencies track and store data quantitatively in a desktop asset-related system; others base their decisions on qualitative experiences. Examples of tracking systems used include AgileAssets, AssetWorks (FleetFocus M5 or FA), IBM Maximo, FASTER, SAP, and PeopleSoft. Agencies use these tracking systems to view historical data points that have previously been logged (such as the amount of spending per piece of equipment) and to extract data or create reports for analysis, among other uses. Utilization rate of a particular item of highway maintenance equipment or vehicle was con- sistently mentioned in the interviews as a key variable influencing the decision about whether repairing, rebuilding, replacing, or retiring an asset was needed. Repair history and the amount of money previously put into maintaining a given equipment item as compared to other equipment in a similar equipment class or year was another key data point that some agencies considered. Other data related to an asset that were considered by agencies in decision-making processes included the following: • Asset age • Mileage • Stage in the equipment life cycle relative to replacement criteria • Comparative expenses against similar units • Proximity of the equipment to a repair shop • Estimated rental costs for a substitute unit while the asset is undergoing repair • Cost of a replacement unit The research team found that some states had a less formal process for integrating data into decision making, noting that a shop supervisor’s “gut reaction” played a significant role in deci- sions; however, they cautioned that such judgments are subject to bias without the use of objective data. Some agencies noted that they use information such as salvaged equipment values from the Kelley Blue Book. Some agencies also reported basing decisions on cost comparisons for repairing a piece of equipment versus replacing it entirely, or on the costs of maintaining an asset. UDOT noted that limited budgets constrained decision making. CTDOT noted that if the repair cost exceeds the cost of purchasing new equipment, then the agency may replace the equipment and put the damaged equipment out of service. Other performance metrics are also used. For example, UDOT uses an asset’s repair history and its current condition, both of which impact repair, rebuild, replace, or retire decisions. CTDOT considers the cost of the equipment’s downtime as part of its repair, rebuild, replace, or retire algorithm. KYTC’s performance measure is related to utilization: each item of equipment needs should log at least 6,000 miles per year. This agency makes the decision to replace the piece of equipment when its cumulative repair costs exceed its purchase price. Some agencies also noted that metrics are used on a case-by-case basis depending on the type of machinery. Other agen- cies do not use performance metrics related to repair, rebuild, replace, or retire decisions, even though they use metrics (e.g., unit life cycle) for their fleets. At least one agency found that using quantitative measures enabled them to advocate for more funding for replacement and repair of equipment. Process of Decision Making The process of decision making for repairing and replacing equipment varied across every agency interviewed. Some agencies are highly centralized with decisions made at the agency

Research Outcomes 13 headquarters level; others are more decentralized with decisions made locally; still others have a hybrid model. The costs of repairing, rebuilding, replacing, or retiring equipment sometimes fall on districts, in which case the decision might fall on that level of the agency; in other agencies costs are carried by the central office and decisions are made at the central level. Some larger fleets are structured hierarchically, in which case the decision-making process is coordinated upward across offices. Typically, agencies that are more centralized rely on the expert opinions of individuals in their central offices (such as the fleet managers or division directors). Some variation was noted between centralized and decentralized agencies. Agencies that are more decentralized tend to have their field offices make most of the repair, rebuild, replace, or retire decisions, but might seek expertise from the central offices for a second opinion or if there are larger financial implications associated with their decisions. Numerous agencies noted that sometimes the decision on whether to repair, rebuild, replace, or retire a piece of equipment came down to immediacy of need for the equipment and whether it was cheaper to repair the equipment to get it back into service or to replace it, which may be the most economical decision in the long run. Some states have specific rules that dictate their repair, rebuild, replace, or retire decision-making processes. For example, UDOT is required to spend 80 to 85 percent of its budget on snowplow trucks. Insights gained from the survey and interviews include the following: • ADOT pays a large risk insurance premium; equipment repair or replacement is often covered by the agency’s insurance program. • Some state agencies (such as MnDOT) are decentralized and allow their individual shops to make decisions on repair and replacement of equipment; this means that some of the shops within the state may vary in their decision processes. Local offices tend to have a green light to make basic repairs when assets are updated for routine maintenance. • UDOT decides whether to replace a piece of equipment using a 10-point system that includes criteria such as miles traveled, hours traveled, and the equipment’s age. Utilization of the equipment, among other factors, plays an important role in whether the equipment gets replaced. This system allows the agency to rightsize its equipment fleet. UDOT also keeps extra snowplow equipment on hand. The agency evaluates its trucks across the state and will shuffle equipment around to ensure that only the worst equipment is disposed of when one district needs a new truck. • One agency interviewed does not have a firm rule about what to do when an asset’s repairs reach a certain cost or when a decision may need supervisor approval. • Seasonality can impact repair, rebuild, replace, or retire decisions, especially for states that experience heavy snowfall in the winter. Sometimes the timeline of the season will drive the decision, based on whether the asset is immediately needed. CTDOT’s districts will make basic decisions but will feed financial decisions through the central office in a formalized pro- cess. CTDOT’s decision is influenced by available funding sources, with repair and replace- ment budgets coming from separate pools. • KYTC’s local shop technicians or equipment supervisors will obtain an estimate of repair costs and then send that figure to the agency’s procurement group, which pulls information on the asset from the agency’s database. They then assess the asset’s fleet service life and history of repair costs to determine whether it should be repaired or replaced. The amount of downtime for the equipment sometimes factors into repair, rebuild, replace, or retire decisions. • Almost half of survey respondents indicated that decisions to repair, rebuild, replace, or retire are made both with input from the central or regional level and qualitatively, depending on the circumstance. This and other survey responses are illustrated in Figure 3 and Figure 4. More survey results can be found in Appendix B.

14 Decision Making for Repair Versus Replacement of Highway Operations Equipment Figure 3. Survey results showing whether the decision process is qualitative, quantitative, both, or neither. Figure 4. Survey result showing the staff involved in decision making.

Research Outcomes 15 Level of Decision Making The transportation agencies that were interviewed generally had three different decision-mak- ing levels: at the field office level, at the central headquarters level, or a combination of the two. Some agencies make financial decisions at the local district level because the spending of their capital or operating budgets is determined at that level. The decisions that involve a hybrid pro- cess may dictate that their equipment managers and supervisors make decisions with the division director’s final approval, or they may designate field offices to make repair decisions and the central office to make replacement decisions. Finally, some of the transportation agencies rely on central office leadership for decision making on repair, rebuild, replace, or retire decisions. Documenting Decisions Transportation agencies vary in how they document decisions, and at what level. The agencies with less formal decision-making processes may not document their decisions. One example of this is MnDOT—its process is more informal, and staff will only document decisions occasionally. Some agencies document all elements of decision making. These agencies enter important data into their fleet management systems, with key documents attached. CTDOT will only document its decisions in its formal management system when decisions are final. ADOT documents its decisions in email form. CTDOT has found that documenting its decisions has helped the agency to advocate the business case for increasing the capital budget for equipment replacements. Summary of Survey and Interviews The research survey and interviews showed that each fleet varies somewhat in terms of its degree of centralization, funding constraints, repair, rebuild, replace, or retire decision-making structure, level of decision approval, and documentation of decisions. Some agencies rely on centralized decision-making structures; others allow their district offices to make decisions. Key insights from the surveys and interviews include the following: • Many fleets report a strong bias toward making repair decisions in order to avoid major replacement backlogs because of insufficient replacement funding. • DOTs with dedicated fleet funding or that operate using an enterprise approach generally treat replacement as a more viable option than do fleets lacking these factors. • Fleets that have a strong fleet manager or an organizational structure with a central fleet group tend to also centralize most major repair, rebuild, replace, or retire decisions. • The following DOTs indicated that fleet maintenance is borne directly by the operating units: Caltrans, KDOT, MnDOT, PennDOT, and UDOT. These fleets also tend to have more decen- tralized repair, rebuild, replace, or retire decision-making processes. • MnDOT was the only DOT reporting that equipment replacement budgets come from dis- trict-controlled funds. • Most fleets report using a hierarchical process for making repair, rebuild, replace, or retire decisions. In practice, this means that the larger the estimated repair costs, the more those decisions require additional levels of approval. The key insight gained from the surveys and interviews was the importance of real-world constraints and the need for rules of thumb when implementing repair, rebuild, replace, or retire decisions. 3.4 Framework Development The purpose of the framework development was to provide a systematic, repeatable process for making the repair, rebuild, replace, or retire decision. The framework was informed by sur- veys of 28 state DOTs and interviews with nine state DOTs on existing repair versus replace

16 Decision Making for Repair Versus Replacement of Highway Operations Equipment decision tools. Additionally, the research team conducted a literature review to better understand existing frameworks for determining the economic lifetime of assets as well as repair limits. The bullets that follow highlight key decisions made by the research team during the develop- ment of the framework: • Added “rebuild” and “retire” as outcomes. The original scope of NCHRP Project 13-08 was oriented around the repair versus replace decision. The research team determined that two additional outcomes (rebuild and retire) were necessary to encompass the full set of options a fleet manager may consider. • Determined the decision criteria. The research team reviewed responses from surveys and interviews to develop a list of 19 key criteria that impact the repair, rebuild, replace, or retire decision (e.g., age, utilization, and repair cost). • Determined the decision path of the tool. In developing the framework, the research team considered two possible structures: a qualitative decision tree and a purely quantitative approach. The survey and interviews demonstrated that most state DOT fleet managers use a qualitative approach, whereas academic literature typically focuses on a quantitative approach. When properly structured, a qualitative decision tree approach meets the four considerations better than a purely quantitative formula. In addition, a purely quantitative approach is insuf- ficient to incorporate the many qualitative considerations of fleet operators, such as budgetary restrictions, operational urgency and need, and environmental factors. However, a purely qualitative approach (such as a decision tree) would not fully meet state DOT needs in terms of justifying repair versus replace decisions to senior DOT management and politicians. Thus, the selected framework integrates both qualitative and quantitative information. • Added an economic analysis module. The original scope of the NCHRP Project 13-08 was to develop a decision-making framework and tool. The research team determined that a quantitative module that estimated the costs of each of the outcomes relative to one another was needed. This information would allow fleet managers to understand the “cost of the wrong decision.” 3.5 Tool Development The research team converted the developed framework into a computational, Excel-based tool with the following independent modules: • Decision Module. A decision support tool that recommends one of four outcomes: Repair, Replace, Rebuild, or Retire. • Economic Analysis Module. A quantitative economic analysis module that estimates the cost of each outcome into the future. Types of User Inputs The research team developed an interface in Excel that allowed the following types of inputs: • Binary. Answers with only two distinct and mutually exclusive options. An example is war- ranty coverage—the repair either is or is not covered by a warranty. • Gradient. Answers with a range of possible options that can be provided using a slider. • Numerical. Answers that have an exact, quantifiable value, such as age and repair cost. Decision Module The Decision Module provides a score between 0 and 100 for each of the four outcomes: repair, rebuild, replace, and retire. The user begins in an Initiation tab and evaluates the repair vs.

Research Outcomes 17 no repair decision. Based on this outcome, the user is directed to either the Rebuild vs. Repair tab or the Replace vs. Retire tab. The decision tree of the Decision Module is shown in Figure 5. For each criterion, the user enters a binary, gradient, or numerical response. The user can exclude a criterion if it is not applicable in the user’s fleet. For example, a user may exclude the criterion “environmental constraints” if no regulation impacts the fleet. Each user input will, individually and collectively, contribute points toward each of the four possible outcomes. The scoring in the Decision Module is based on the expert judgment of the research team. This scoring was refined over several months of testing a variety of cases, which eventually became the 4R Tool use case examples found in the User Manual. Additionally, the scoring of the slider bars in the Decision Module uses an underlying beta distribution such that only when the user adjusts the slider bar toward the extreme cases at either end of the bar does the score change in a meaningful way. In contrast, when the user adjusts the slider bar within the middle 80 percent of the bar, the scores only change by a small quantity. This aligns with the informal Pareto principle (or 80/20 rule) in which only extreme cases change outcomes. For added flexibility, advanced users can adjust weight factors for each criterion on a scale of 1 to 100. These weight factors are needed because state DOT fleets differ in how important they consider certain criteria to be. For example, a given state DOT fleet may place a larger emphasis than another fleet on environmental considerations and therefore would weight the environ- mental criterion higher. The Output tab summarizes the scores for each of the four possible outcomes and makes a recom- mendation to the user. This tab is set up to be a printable page for documentation purposes. Figure 5. Decision tree in 4R Tool from initiation to outcomes.

18 Decision Making for Repair Versus Replacement of Highway Operations Equipment Economic Analysis Module The Economic Analysis Module estimates the differential cost between the repaired, rebuilt, and replaced equipment through the one-year planning horizon. A planning horizon of one year is reasonable because replacement of the repaired equipment is considered in the next annual cycle. The differential cost is calculated based on the estimated cost of each piece of equipment (the repaired and the replacement equipment) for its estimated use (hours or miles) during the subsequent year. The cost of the replaced equipment is taken as the cost of any other piece of equipment in the class and based on the total owning and operating cost rate ($/h or $/mi) established for the class. The cost rate in the next year for the repaired piece of equipment should be estimated based on the cost history of the piece of equipment and estimated repair cost. The depreciation schedule applied to the piece of equipment and the operating cost data can be used to estimate unit rates for owning (depreciation) and operating it in the next year. The estimated repair cost is included in the cost of operating the repaired piece of equipment because the repair is required for operation over the next year. The input data required for the Economic Analysis Module include the following: • Rate cost for the class • Expected utilization over the next year • Estimated repair cost • Depreciation cost rate for the repaired piece of equipment over the next year • Linear model for operating cost rate by year (intercept and slope values) • Age of the equipment This module is intended for users familiar with LCCA, which accounts for all costs associated with an asset (from purchase to disposal or resale) over time. To obtain the cost rate, the user must provide values for the current equipment age, annual use going forward, current equipment value, and A and B coefficients describing the operating costs as the equipment ages. The form of the equation that describes life-to-date (LTD) operating costs is a second order polynomial: LTDOperating Cost A Age B Age2= × + × The coefficient A is the linear portion of growth in costs over time, and B is the nonlinear portion of the curve, representing how costs accumulate more quickly as the equipment ages. Finalization of the 4R Tool To complete development of the 4R Tool, the research team undertook testing, validation, and finalization for the 4R Tool product. Nine use case examples were developed for the Decision Module, and one use case example was developed for the Economic Analysis Module. These examples are included in the User Manual.