The Relationship Between Transit Asset Condition and Service Quality (2018)

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.

26 Guidance for Calculating Effects of Changes in Asset Condition on Transit Service Quality Overview This chapter provides guidance on calculating the effects of changes in asset condition on transit service quality. The measure, Effective Journey Time (EJT) (introduced in Chapter 3) is used as a summary measure of transit service quality for this calculation. The guidance includes a set of steps for transit agencies to use to calculate a baseline value for EJT, as well as the change in EJT that may result from an improvement in or worsening of asset conditions. The guidance is intended to be used by transit asset owners and other transit professionals involved in asset management, planning, and operations to better understand the relationship between asset condition and transit service quality, as well as to support investment decision making. The EJT measure generated through the calculations combines actual journey time for dif- ferent components of a passenger’s journey with adjustment factors for each component of the journey. These adjustment factors vary as a function of passenger perceptions related to the dimensions of service quality described in Chapter 3. The guidance details five basic steps for any condition/service quality analysis: • Step 1: Determine Scope of Analysis—establish the base and investment cases that will be analyzed, including the assets and route that will be modeled and the analysis timeframe. • Step 2: Assess Available Data—review what data are needed to support the calculations, as well as what data are available and what estimates can be made to account for missing data. • Step 3: Select Calculation Approach—determine whether to use the simplified or compre- hensive EJT model, based on available data and the case to be analyzed. • Step 4: Perform Calculations—use one of the calculation tools detailed in Chapter 6 to cal- culate EJT for the base and investment cases. • Step 5: Summarize Results—compile the results of the calculations and compare them to help characterize the effects of changes in asset condition on transit service quality. Two specific EJT calculation approaches are supported in the guidance to accommodate dif- ferent levels of complexity in the scenarios to be analyzed and different levels of data availability: a simplified approach, and a comprehensive approach. The following sections provide additional information on key concepts and assump- tions, detail the steps to be performed for any analysis, and provide additional step-by-step instructions for each of the three calculation approaches. Spreadsheet tools are provided to help support both calculation approaches. Documentation on the tools is provided in Chapter 6. C H A P T E R 5

Guidance for Calculating Effects of Changes in Asset Condition on Transit Service Quality 27 Key Concepts and Assumptions EJT Example Figure 3-1 showed an example of a baseline calculation of EJT, and an example of the improve- ment in EJT that might result following improvements in asset condition from an SGR invest- ment. The figure shows three basic components of EJT: buffer time, wait time, and IVT. In concept, the calculation can be extended to include other trip components, such as the time required for conveyance within transit stations. In the example, an average EJT is shown per passenger. The results of an analysis can also be expressed in terms of EJT per passenger mile, total EJT, or total cost (obtained by multiplying EJT by a value of time). Measuring Asset Condition Changes in asset condition affect EJT in two basic ways: • Customer perception. Changes in asset condition can affect customer perceptions. These effects may not result in a change in actual journey time, but do affect EJT; and • Likelihood of delay. Changes in asset condition can affect the likelihood of delay (e.g., from a vehicle or guideway failure). Such effects result in changes to actual journey time, and by extension, to EJT. Many measures of asset condition have been defined, but in the United States, a commonly used measure for transit assets is the 5-point scale established by FTA. Using this scale, a new asset has a rating of 5, and an asset is deemed to be beyond its useful life, either due to physical condition or obsolescence, if its rating drops to a value of 2 or 1. Applying this scale to customer- facing assets, one might expect the adjustment factor for wait and/or IVT to increase in cases where assets are in poor condition, or have a value of 2 or less on this scale. To predict changes in actual journey time one must predict the delay that results from incidents where an asset fails and requires immediate maintenance. The modeling framework presented in TCRP Report 157 (Spy Pond Partners et al. 2012) has been applied for predicting how the failure rate of an asset will change in the future based on its condition, age, or mileage. This framework can be used to predict the current failure rate, as well, but it is preferable to obtain the failure rate directly. In the case of vehicles, transit agencies routinely track MDBF, so vehicle failure rates are readily attainable. At the same time, it is difficult to predict MDBF based simply on a vehicle’s age or mileage due to differences between transit agencies, modes, transit lines, in-vehicle operating characteristics, maintenance practices, standards for what constitutes a “failure,” and other factors. Asset failure rates can be used to predict changes in the components of journey time resulting from periodic failures of vehicles, guideway components, or other assets that may cause delays if they fail. Generally speaking, the EJT model can accommodate different approaches to defining vehicle, guideway and other asset failures, provided the various model parameters (e.g., failure rates and consequences) are set with a consistent understanding of how failures are defined. Key Assumptions The following are key issues and assumptions related to calculation of EJT: • Journey time is different for each combination of origin and destination, as well as by time of day. Thus, obtaining the most accurate calculation of EJT requires calculating this measure by time of day for each origin-destination (O-D) pair of each route or line a transit agency operates. However, the basic effects of changing asset condition tend to be observed regardless of O-D pair or time of day.

28 The Relationship Between Transit Asset Condition and Service Quality • The analysis approach is intended for use in predicting EJT for frequently operated fixed- route service for buses, bus rapid transit (BRT), and rail (light rail, heavy rail, and/or com- muter rail). • Changes in service quality resulting from changes in asset condition can be presented either in terms of changes in average EJT or by describing examples (e.g., consequences to passengers on a given line as a result of a vehicle or guideway failure). Examples are particularly helpful if the average effects of asset condition changes are modest or, alternatively, if the effect of a failure is particularly dire. The guidance and supporting spreadsheet tool support calculation of average effects and examples of changes in asset condition. • Although one can convert EJT into a journey cost, other costs are associated with maintaining, rehabilitating, and replacing transit capital assets not addressed through this guidance. The guidance described here is specifically concerned with predicting effects on service quality from changes in asset conditions. Other resources address the full set of transit agency, user, and social costs related to transit assets. • All physical transit assets can conceivably be included in the calculation of service quality impacts, but, practically speaking, effects from changes in condition are expected only for cases where an asset is customer-facing (and thus may affect perceptions) and/or can cause delays in transit service (and thus affect journey time). For other assets, there may be a tie to service, but it is more indirect. For instance, a transit agency that fails to maintain its service vehicles may face severe service impacts in the event of a winter storm, but such events are not explicitly captured in the EJT calculation methods. The following sections detail the calculation guidance based on these assumptions. Analysis Steps Step 1 – Determine the Scope of the Analysis The first step in performing an analysis of EJT is to determine the scope of the analysis. Specifi- cally, one must answer the following questions: • What is the future case scenario? Typically an analysis involves comparing current or base- line conditions to conditions in some assumed future case. Three types of analyses one may wish to perform are – Effects of deterioration—This scenario describes the effects if existing assets are allowed to deteriorate and are not maintained in a state of good repair. – Effects of improvement—This scenario describes the benefits that can be obtained, in terms of journey time, when improvements are made to assets. – Illustrative scenario—This scenario is used to illustrate the worst case or best case in asset condition and/or failures. This could describe the effects on the system of events, such as track closures or vehicle breakdowns that have major effects on service quality. Although the focus of this guidance is on determining the effects of changes in asset condition on service quality, the same basic framework can be repurposed for other applications, such as understanding the service quality effects of investments in new infrastructure (e.g., new bus shelters or infrastructure improvements that result in shorter trip times) or operational improvements. • What assets are being modeled? Vehicles should be included in any analysis, and for bus systems this may be the only asset that bears consideration. However, where a bus system

Guidance for Calculating Effects of Changes in Asset Condition on Transit Service Quality 29 includes station facilities, or in the case of rail systems, one must decide whether to include station facilities, as well. Further, for rail systems one may or may not wish to consider poten- tial guideway failures. • What lines or routes are being modeled? EJT calculations are specific to a given line or route. If an analyst is calculating EJT changes for a fleet of vehicles used for multiple routes or lines, it may be necessary to select typical routes/lines. • What is the timeframe for the analysis? Particularly if one is modeling effects of gradual deterioration in asset conditions, it may be appropriate to analyze service impacts by “turning the clock” on asset conditions by 10 years or more. • What results are required from the analysis? Typically, an analysis will yield EJT calculations for the baseline and future case, as well as the difference between them. One may also wish to express the results in terms of EJT per passenger, EJT per passenger mile, or in terms of total cost. Table 5-1 provides examples of how one might define the analysis scope for different situ- ations. Carefully considering the analysis scope in this step will help guide decisions made in later steps. Step 2 – Assess Available Data Once the scope of the analysis has been determined, the next step is to determine what data are available or may reasonably be obtained. All available data will be used to help guide the decision on the appropriate calculation approach in Step 3. Further, if key data needed for an analysis are unavailable, the scope of analysis may need to be reconsidered. Table 5-2 identifies Description Example 1 – Deterioration of a Bus System Example 2 – Effects of Guideway SGR Investment Example 3 – Illustrating Potential Consequences of Underinvestment Future Case Resulting conditions in the future, assuming the current fleet is kept in service Improved conditions for guideway and/or stations following SGR investment to reduce slow zones or improve station conditions Comparison of baseline conditions to a worst case example in which stations are deteriorated and vehicles are prone to frequent failure Assets Buses Rail vehicles, guideway Bus/Rail vehicles, guideway (for bus rapid transit and rail), stations Line/Routes A set of representative routes Selected rail line Single representative routes Analysis Timeframe 15 years 5 years 25 years Results Generated Average change in EJT per passenger Average change in total annual EJT for the line Change in EJT per passenger Table 5-1. Analysis scope examples.

30 The Relationship Between Transit Asset Condition and Service Quality data requirements for the simplified and comprehensive approaches. Data items required for the specified calculation approach are indicated with a check mark. The calculation tools come populated with default values for all parameters. As shown in Table 5-2, basic information on the vehicles used for service is required for both approaches (i.e., the vehicle type, vehicles per consist [number of cars per train], average fleet age, useful life, and MDBF). Additional items required for the comprehensive approach include seats per vehicle, headway standard deviation, time and distance between each station, daily passengers using each station, and directional splits for passengers boarding at each station. The simplified model requires additional data items not required in the comprehensive model: total route length, average vehicle speed, percentage of fleet typically under repair, and the spare ratio. Other factors and parameters are optionally included in the calculations. These are discussed further in Chapter 6. Step 3—Select Calculation Approach The next step is to select the calculation approach. The two calculation approaches are as follows: • Simplified Approach. The Simplified Approach is intended for use in estimating EJT for selected O-D pairs. This approach is most appropriate for testing effects of changes in fleet condition or to obtain a general sense of effects of changes of condition on service quality. • Comprehensive Approach. The Comprehensive Approach provides the most complete prediction of EJT. This approach is recommended where EJT calculations are required considering each O-D pair on a route/line or for cases where one is analyzing in-station conveyance time. Table 5-3 presents a comparison of the two approaches. Data Item Calculation Approach Simplified Comprehensive Vehicle type Vehicles per consist (1 for bus) Seats per vehicle Vehicle headway Headway standard deviation Route length Average vehicle speed Trains/buses per day Distance between each station Travel time between each station Daily passengers using each stop/station Directional split of passengers using each stop/station Average fleet age Vehicle useful life MDBF Percentage of fleet typically under repair Spare ratio Table 5-2. Data requirements by calculation approach.

Guidance for Calculating Effects of Changes in Asset Condition on Transit Service Quality 31 Data Item How Addressed in Each Approach Simplified Comprehensive Vehicle failures Initial MDBF is directly entered, or approximated based on vehicle type and fleet age. Change in MDBF is approximated based on TCRP Report 157 models. Guideway failures Not modeled. Default failure rate specified. Change in failure rate approximated based on TCRP Report 157 models. Buffer time Calculated as a multiple of the standard deviation of total journey time. Variances from different trip components are summed. Wait time Calculated assuming uniform passenger arrivals. For infrequent service this may overestimate wait time – in these cases, wait time should be removed from the results. In-vehicle time Calculated based on route length and vehicle speed. Calculated based on time required between each station. Includes adjustment for crowding based on the TCRP Report 165 model. In-station conveyance time Not modeled. Optionally can calculate boarding and alighting times, including increased time from elevator and escalator failure, if applicable. Headway Increased based on failure rate and/or if insufficient vehicles are projected to be available to meet the headway requirement. May be adjusted for vehicle and/or guideway failures. User must consider whether the headway is realistic for future cases. Headway standard deviation Can be entered by the user, or is approximated assuming headways are distributed as a gamma distribution. Can be entered by the user, or is approximated assuming headways are distributed as a gamma distribution. Deviations in travel time can be captured through the standard deviation of the headway or time between stations. Adjustment factors Defaults values are listed in Chapter 3 – the user can override these. The vehicle adjustment factor is calculated based on fleet age and useful life. Table 5-3. Comparison of EJT calculation approaches.

32 The Relationship Between Transit Asset Condition and Service Quality The analyst should review the data requirements and assumptions for each approach, and in this step select which to use for the analysis. As a general rule of thumb, the Comprehensive Approach is required for complex cases, and recommended where sufficient data are available to support it. Nonetheless, the Simplified Approach provides an alternative where either sufficient data are unavailable to support the Comprehensive Approach, or where an approximate value of EJT is required. Step 4—Perform Calculations Once the approach is selected, it is time to perform the EJT calculations. Spreadsheet tools are provided to support both the Simplified Approach and Comprehensive Approach. These tools are documented in Chapter 6. In both of the tools, default values are provided for all parameters, and the analyst can either use the default value or enter a new value. Figure 5-1 shows an example data entry screen, in this case for the simplified version of the tool, with default values shown for each parameter and fields for user input. Both tools provide a prediction of EJT in minutes per passenger for the base and investment case. In the comprehensive tool, one can specify two different investment cases, labeled “typical” and “worst case.” The tool computes an average future case using the results of these two as well. Figure 5-1. EJT calculator – example data entry screen.

Guidance for Calculating Effects of Changes in Asset Condition on Transit Service Quality 33 Step 5—Summarize Results The final step is to compile and summarize the analysis results. Depending on the scope of the analysis and calculation approach, one may obtain one or all of the following metrics for the base and future cases: • EJT per passenger • EJT per passenger mile • Total EJT The analyst may wish to supplement these with • Percentage change in EJT between the base and future scenarios • Unadjusted journey times per passenger • Total journey cost (obtained by multiplying the per passenger results by number of passengers) Table 5-4 summarizes results for various representative scenarios. Chapter 6 includes a set of worked examples illustrating calculation of the above measures. Table 5-4. Representative results for effects of asset condition changes on transit service quality. Mode Headway (minutes) Scenario Change in EJT (minutes per passenger) Notes Bus 10 Fleet age increased 5 years +1.5 Initial age = 5 years IVT = 15 minutes 15% of fleet typically under repair Fleet age increased 10 years +5.1 Fleet replaced -1.1 15 Fleet age increased 5 years +1.7 Fleet age increased 10 years +6.6 Fleet replaced -1.1 20 Fleet age increased 5 years +1.9 Fleet age increased 10 years +8.1 Fleet replaced -1.1 (continued on next page)

34 The Relationship Between Transit Asset Condition and Service Quality Although EJT is valuable as a single measure that summarizes transit service quality, no one measure can properly reflect all of the dimensions of service quality described in Chapter 3. It may be helpful to supplement the presentation of EJT results with a qualitative description of the changes in service quality expected from a predicted improvement in or worsening of asset condition. Mode Headway (minutes) Scenario Change in EJT (minutes per passenger) Notes Heavy Rail 10-15 Fleet age increased 5 years +0.5 Initial age = 5 years IVT = 15 minutes 15% of fleet typically under repair Fleet age increased 10 years +1.6 Fleet replaced -0.5 20 Fleet age increased 5 years +0.5 Fleet age increased 10 years +2.1 Fleet replaced -0.5 Light or Heavy Rail 10 Fleet exceeding useful life replaced -3.5 Initial age = 35 years IVT = 15 minutes 15% of fleet typically under repair Bus 10 Fleet exceeding useful life replaced -3.2 Initial age = 15 years IVT = 15 minutes 15% of fleet typically under repair Light Rail 10-15 Fleet age increased 5 years +0.5 Initial age = 5 years IVT = 15 minutes 15% of fleet typically under repair Fleet age increased 10 years +2.0 Fleet replaced -1.5 20 Fleet age increased 5 years +0.5 Fleet age increased 10 years +2.6 Fleet replaced -1.5 Table 5-4. (Continued).