Preventive Maintenance Intervals for Transit Buses (2010)

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.

13 INTRODUCTION This chapter examines PM intervals, scheduling methods, pro- cedures for determining when parts or components require replacement, and factors that most influence the setting of PM intervals such as manufacturers’ specifications and local oper- ating environment. Information provided in the chapter is based on material obtained from survey responses. PREVENTIVE MAINTENANCE FREQUENCY PM intervals are separated into three classifications: inspec- tions, repairs, and overhauls. The Fleet Profile of agencies responding to the survey, from which data on PM intervals were obtained, is shown in Appendix A. It includes agency name, location, fleet makeup in terms of bus propulsion type (e.g., diesel and CNG), bus quantities, and combined fleet mileages. Inspections Daily Service Line Inspections Eighty-eight percent of those agencies responding to the sur- vey conduct daily service line inspections. The high number is expected given that the time needed to refuel buses pro- vides an excellent opportunity to check essential fluid levels and make other vital inspections. One agency conducts these inspections bi-weekly, whereas another does so every time the bus is in the workshop, or every 20 days. PMIs PMIs are typically based on the minimum interval established by the OEM to change engine oil and provide other lubrica- tion services. The most common suggested oil change inter- val is 6,000 miles. Some newer engines with more advanced emissions controls, especially those with exhaust gas recircu- lation (EGR), require more frequent oil change intervals at 3,000 miles. Of those responding to the survey, 92% conduct PMIs at intervals that fall within 2,000 (one agency) to 6,000 miles, with 71% falling right at the 6,000 mile mark. The remaining 8% conduct PMIs at intervals of 6,500 to 7,500 miles. The primary reason for using miles to schedule bus PMIs appears to be that agencies find it more convenient to use miles rather than hours of service and other time-based interval options recommended by the OEMs. One exception is Seat- tle Metro, which a few years ago changed its PMI interval for the electric trolley bus fleet from its standard 3K/6K/12K/ 24K mile interval to a 28/56/168/336 day schedule. The agency switched its PM interval policy because the trolleys, although not logging many miles, experience large passenger loadings owing in part to the free ride zone where they operate. The base PM interval for the trolley fleet originally was 3,000 miles, which was not an accurate indicator of how these vehicles were being used. Because Metro wanted the trolleys to receive similar PM attention as its traditional diesel fleet, and the diesel fleet accumulates approximately 3,000 miles per month, the agency established a 28-day PM interval for its trolley fleet. The move also leveled out the inspection load in that it could now perform PM inspections on 25% of its trolley fleet each week, thereby cycling the entire trolley fleet through PM inspections every month. Metro is currently evaluating whether the 28-day PM cycle should be extended to 56 days by testing the interval on a sam- pling of 20 trolley buses. The agency closely monitors service interruptions and unscheduled maintenance, both of which have decreased since instituting the 28-day PM inspection cycle. Metro will decide on extending the PM intervals to 56 days based on how well the test fleet performs compared with the baseline buses. Ride On in Montgomery County, Maryland, which cur- rently uses mileage-based PMI intervals, is investigating basing those intervals on fuel consumption, but at this time had not yet concluded its study. Additional information regarding its approach is included later in this chapter. Despite a message posted on the Bus Maintenance listserve asking for other examples of agencies using something other than mileage to establish PMI intervals, none were provided. When asked if agencies use different intervals for other buses in their fleets, 57% noted using distinct PMI inter- vals, whereas the remaining 43% use the same PMI interval for all buses. The primary reason given for conducting more frequent inspections (i.e., those in the 3,000 to 4,000 mile range) pertain to hybrid buses (one response), smaller para- transit vehicles (10 responses), and buses with EGR (three responses). One agency extended its bus PMI intervals from CHAPTER THREE SURVEY RESULTS: PREVENTIVE MAINTENANCE INTERVALS

5,000 to 6,500 miles as a forced cost-saving measure, know- ing that doing so would increase road calls. In this case, the agency disregarded an essential PM tenet to ensure reliable bus service for the sake of improving customer satisfaction. Once a standard PMI interval is established (i.e., every 6,000 miles), PM activities requiring less frequent intervals are typically worked into the base inspection. Although the indus- try lacks standardized nomenclature, the term “A” inspection is generally used to denote the base inspection where engine oil is also changed and the chassis lubricated. Additional activ- ities can then be added at multiples of the base “A” inspection, such as “B” inspections done at 48,000 miles to also service the transmission, or a “C” inspection at 60,000 miles to per- form other tasks. Intervals vary depending on equipment needs and local operating requirements. Grouping additional PM activities into regularly sched- uled events is certainly more efficient. However, in some cases agencies cannot keep buses out of service for extended periods of time to perform all needed tasks and bring them in separately. In other cases, OEM specified intervals do not always coincide with existing intervals. Bus systems requir- ing additional inspections, whether at regularly scheduled intervals or separately, include: • The heating, ventilation, and air conditioning (HVAC) system; • Wheelchair lifts and ramps; • Cooling system; • CNG tanks; • Fare collection; • Articulation joint; • Electrical equipment, including radios and the automatic vehicle location (AVL) system; • Hydraulic system; and • Brakes and air drier equipment. The intervals at which additional bus systems are inspected and serviced drew varied survey responses, strengthening the view that no one PM inspection program is suitable for all. Except for HVAC and wheelchair lift and ramp inspections, survey responses to specialized PM inspections were too diverse to classify. Tables 3 and 4 summarize survey responses regarding additional inspections added to address HVAC and wheelchair lifts and ramps. Preventive Maintenance Repairs When asked to list repair activities done as preventive mea- sures based on time, mileage, or other condition (other than failure), survey responses were also varied and difficult to classify. However, the specialized repair activities and related intervals summarized in Table 5 from the survey responses provide several examples for comparative purposes or for agencies to adopt into an existing PM program. 14 Preventive Maintenance Overhauls/Refurbish Overhaul and refurbish activities done as a preventive mea- sure based on time, mileage, or other condition, not because of failure, received responses in six categories. Four re- sponders rebuild engines in full-size buses (35 ft and larger) at 300,000 miles; two perform rebuilds at approximately 450,000 miles. One agency replaces Detroit Diesel Corpo- ration (DDC) S50 engines at 350,000 miles, Cummins model ISC engines at 375,000 to 400,000 miles, and Cummins model ISB engines at approximately 250,000 miles. Regarding transmissions, one agency rebuilds them at five years, another at 300,000 miles. One agency rebuilds wheel- chair lifts at 10 years. When it comes to refurbishing full-size buses, one agency does it at 6 years and another at 7 years. One agency does complete body and paint between 7 and 10 years, whereas another repaints buses every 3 years. Interval Survey Response 6,000 Miles Annually 13 (33%) 6 (20%) Seasonally/Bi-Annually 5,000 Miles 7,500 Miles 12,000 Miles 30,000 Miles 44,640 Miles (72,000 km) Bi-Weekly 60 Days 180 Days 4 (13%) 1 1 1 1 1 1 1 1 TABLE 3 HVAC INSPECTION INTERVALS Interval Survey Response 6,000 Miles 7,500 Miles Annually 4,000 Miles 5,000 Miles 7,000 Miles 12,000 Miles 24,000 Miles 30,000 Miles Bi-Weekly 60 Days 90 Days Bi-Annually 18 (47%) 2 2 1 1 1 1 1 1 1 1 1 1 TABLE 4 WHEELCHAIR LIFT AND RAMP INSPECTION INTERVALS

15 METHODS FOR ESTABLISHING PREVENTIVE MAINTENANCE INTERVALS Scheduling Techniques All agencies responding to the survey use some form of computer-based program to ensure PM intervals are conducted on time; three supplement this process with an Excel spread- sheet, one supplements the process with manual recordkeeping. The variety of programs used is overwhelming; 38 respon- ders reported using 22 different programs. A slight majority (56%) indicated that they are pleased with their PM software program, whereas 24% are not; 20% did not answer. Of all responding agencies, 82% claim the programs are flexible enough to make changes, 8% say they are not; 10% indicated the question is not applicable because intervals are set by manufacturers’ specifications and changes are not needed. In addition to using a computer-based program to establish PM intervals, 87% also use it to guide and track the actual PM activities, whereas 3% do not; 10% said the question was not applicable. Although scheduling programs may differ, the concept is to provide an indication of when buses are due for the next PM activity. Table 6 shows a program report example that identifies buses closest to requiring a 6,000 mile PM inspec- tion listed in descending order. The full table would include all buses in the fleet and could be tailored to account for buses that require different intervals. Table 6 or a similar report would be reviewed daily to schedule PMs. Indeed, intervals would be done at the exact interval. Actually, PMs are done slightly before or after the scheduled interval, which has a cumulative effect over time. Agencies were asked if PM intervals are based on fixed points (i.e., every 6,000 miles, every 30 days, etc.) or are they reset based on the last actual interval. Of those responding, 61% reset intervals based on previous intervals regardless if it was early or late, whereas 39% conduct PMs at fixed- point intervals. For those who reset intervals based on the previous one, subsequent inspections could be thrown off even if the schedule shows it to be “on time.” For example, if scheduled 6,000 mile “A” inspections were done at an average of 6,500 miles, the 24,000 mile “B” inspection would actually be done at 26,000 miles. Because each “A” inspec- tion was done within an acceptable window of 10% of the scheduled interval, the “B” inspection would also be con- sidered “on time,” even though it was actually performed 2,000 miles beyond the required interval. Setting PMs at specific intervals mitigates the cumulative effect. Ensuring On-Time Preventive Maintenance Performance Agencies also use a variety of means to ensure PMs are done on time. Of responding agencies, 42% use an accept- able window of 10%, 5% do not use an interval window, Repair Application No. Response/Interval Service Air System All w/air brakes 1—24,000 miles 1—48,000 miles 1—60,000 miles 1—Annually 1—Bi-annually 1—Every 1–2 years Change Air Valves All w/air brakes 1—60,000 miles Replace Rear Brake Chambers All w/air brakes 1—150,000 miles Replace Brake Application Valve All w/air brakes 1—150,000 miles Replace Belts/Tensioners - Cummins engines - All buses in fleet 1—100,000 miles 1—48,000 miles Replace Spark Plugs CNG buses 1—15,000 miles Valve Adjustment All buses in fleet 1—36,000 miles Engine Tune-Up 30 ft and larger buses 1—50,000 miles 1—350,000 miles Replace Engine Harmonic Balancer 40 ft buses 1—125,000 miles Check Diesel Particulate Filter (DPF) Backpressure All w/DPF 1—6,000 miles Replace DPF All w/DPF 1—24,000 miles Replace Crankcase Breather Filter All w/filter 1—24,000 miles Service Turbocharger All w/turbocharger 1—50,000 miles Repair/Replace Radiators 35 ft buses Every 8 years Replace Cooling System Thermostats 35 ft–45 ft buses 60,000 miles Replace AC Fan Motors 40 ft buses 1 – Every 4 years Adjust/Replace Front Wheel Bearings Not specified 1—24–26,000 miles Service Driver Seat All buses in fleet 1—Every 6 years Replace Fire Suppression Squib All w/squib 1—Every 3 years Adjust Doors All buses in fleet 1—18,000 miles Repair Farebox All w/fareboxes 1—6,000 miles 1—48,000 miles Replace Shock Absorbers 30 ft–60 ft buses 1—60,000 miles Replace Ultra-Capacitor Fan Motors Hybrid buses 1—48,000 miles TABLE 5 PM INTERVALS FOR SPECIFIC REPAIRS

and 53% use other methods. Of those using other methods to determine if PM intervals are done on time, 65% use some type of a mileage-based system that falls outside the 10% window. Those using an on-time acceptability window greater than 10% risk the cumulative effect of having sub- sequent inspections done beyond the suggested interval even though on paper they appear to be punctual. Computer-Based Systems to Guide Preventive Maintenance As indicated earlier, agencies reported using a wide variety of computer-based programs to schedule their PMs. When asked if these programs are also used to guide and track actual PM activities (i.e., follow-up repairs, parts, costs, etc.), 87% say they do. More capable programs are part of a larger Manage- ment Information System (MIS) that tracks and helps manage many maintenance-related activities, including repairs, costs, parts inventory and purchasing, fuel and lubricant dispensing, vehicle history files, vehicle availability, timekeeping, pay- roll, account payable, facilities maintenance, and others. These systems also have the capability of distinguishing between scheduled and unscheduled PM activities. Several agencies indicated that their programs generate a unique checklist based on bus type to guide technicians through the inspection process. The work order, common in virtually every automotive maintenance operation, continues to be the basis for issuing PM work to technicians, providing historical information regarding past work, and for techni- cians to note defects and parts usage. The Whatcom Transportation Authority, Bellingham, Washington, uses a vendor-developed MIS program. It gen- erates a work order that identifies the maintenance tasks to be performed on a given bus, provides historical data for that bus, tracks parts and labor based on information pro- vided by technicians, and manages warehouse inventories. Included with the work order is a listing of parts needed to perform each PM activity along with a specialized checklist based on the specific bus type. Work orders are all bar-coded and parts inventories automatically readjusted. The system also separates those activities done as scheduled events as opposed to those done on an unscheduled basis, which is essential for tracking PM performance. 16 Innovative Approaches to Establishing Preventive Maintenance Intervals The survey sought to identify any unique or innovative approaches used to establish PM intervals. Fluid analysis was mentioned by several responders. Here agencies take samples of vital fluids such as engine oil, transmission fluid, and coolant at the end of the drain interval and send those samples to a laboratory to identify deteriorating conditions and determine if fluid intervals need to be extended or made more frequent. Fluid analysis results showing traces of wear metals and other contaminants also provide early indications that components are beginning to fail, thereby allowing agen- cies to schedule repairs and overhauls before more serious and costly problems develop. Normal fluid analysis reports may be an indication that oil drain intervals could be extended. Before doing so however agencies need to consult their oil analysis lab to become famil- iar with the ramifications of doing so and gain authorization from the component manufacturer to ensure warranty cover- age. Use of synthetic and other lubricants may be allowed by OEMs under certain conditions to extend oil change intervals. When extending fluid change intervals, one must also consider the results of conducting less frequent inspections. Chapel Hill Transit, North Carolina, maintains close com- munications with bus, engine, transmission, and other OEMs to obtain information regarding activities that might be added to the PM program based on their insight and experiences. Agencies such as Rockford Mass Transit, Illinois, speak with others in bus maintenance to learn from their experiences. The San Diego Metropolitan Transit System analyzes statisti- cal data on breakdowns and adds items to its PM program. The Santa Clara Valley Transportation Authority (VTA) uses a committee of maintenance personnel headed by the agency’s engineering department to establish PM intervals, but is care- ful not to exceed OEM specifications. Coast Mountain Bus, Burnaby, British Columbia, Canada, has recently established a QA program for PMIs and uses feedback to fine tune its PM program. Two procedures are used. One is an Inspection Change Request form, attached as Appendix D, which can be completed and submitted by anyone in the maintenance department to suggest changes to the PM program. The agency’s QA team investigates each Bus Number Life Miles Last PM Mileage Miles Since Last PM Miles to Next Scheduled PM 333 239,900 233,924 5,976 24 326 272,406 266,449 5,957 43 343 245,127 239,236 5,891 109 TABLE 6 REPORT SHOWING NEXT PM INTERVAL DUE IN DESCENDING ORDER

17 request to determine whether the PM program needs to be altered. The investigation explores and considers any impacts the change would have on individual groups within the depart- ment (e.g., Material Control, Parts, and Work Planning). Any change made to the PM program ultimately must be approved by the Fleet Technical Support Manager. Another form of feedback used by Coast Mountain to improve its PM program is monthly bus audits conducted by the QA team. The audits are done at least once a month on a recently inspected bus. These audits are reviewed by man- agement to ensure that all defects were identified and proper steps taken to repair them. The audits are also used to deter- mine if new maintenance practices might be investigated. Montgomery County, Maryland (Ride On) monitors mile- age, engine hours, and fuel consumption to establish a mileage- based interval appropriate for each PMI. The agency is also monitoring a test vehicle to determine if PM intervals are bet- ter established using fuel consumption. It is its theory that fuel consumption is a stronger indicator of the severity of operating conditions than mileage alone. The agency’s initial test began with a 2007 International bus model 3200IM with a VT365 engine. Unlike Detroit Diesel or Cummins engines operated by Ride On, International according to Ride On does include an option for changing the engine oil and filter according to fuel consumption. For this particular engine, the specification is 1,000 gallons of fuel, 10,000 miles, six months, or 350 h. When making its calculations using an average miles per gallon of 5.85, 1,000 gallons of fuel consumption averaged 5,850 miles. Because it already uses miles to trigger PM inter- vals the agency rounded it to 6,000 miles to keep it consistent with their existing interval. Going forward, Ride On will continue its investigations and believes that using fuel consumption for setting PM inter- vals is more appropriate for Bus-to-Block operations where buses are assigned to the same route every day and average fuel consumption is more consistent. Table 7 illustrates how buses on different Ride On routes accumulate mileage at dif- ferent rates. Buses used in Interstate commuter service accu- mulate nearly four times the mileage in less than half the time as buses operating on a city route. One agency does some work ahead of scheduled PMIs at 6K and 12K intervals to avoid the 24K and 48K inspec- tions from consuming too much time, thereby fitting work in to suit schedule demands. Miami–Dade Transit, Florida, uses its MIS system to track the history of each major bus sub-component, analyze the data, and includes specific inspection, replacement, or adjust- ment activities to take place at certain intervals based on the analysis. Trend analysis conducted from the data determines the distance in miles between needed actions, and PM inter- vals are established accordingly. Miami–Dade has two types of PM, a 3,000 mile PM referred to as an “O” inspection and a 6,000 mile “A” inspection. The “O” inspection is basically an oil and lubrication service with additional safety inspection items. The 6,000 mile “A” inspection is where the agency adds tasks based on manufacturers’ specifications, regulatory requirements, and collected data. Miami’s PM scheduling software, as with most others, maintains life-to-date mileages on each bus. If their trend analysis dictates that a part should be replaced at a certain time, the PM scheduler is programmed to flag the action. As an example, their trend analysis reveals that the reliability of brake slack adjusters drops off signifi- cantly after 60,000 miles. When buses reach a 60,000 mile interval, the task “Remove and Replace Slack Adjuster” appears as a PM action item. Predicting Parts Replacements In addition to establishing PM intervals, programs are also available to accurately predict part and component replace- ments just before they fail. Most are based on the Weibull analysis, named for Swedish engineer Ernst Weibull (2). It allows managers to make predictions about the life of all products in the fleet by “fitting” a statistical distribution to life data from a representative sample of vehicles. The data set can then be used to estimate important life characteristics of a product such as reliability or probability of failure at a specific time, mean life for the product, and failure rate. Applying these models, however, requires resources to collect accurate data and apply mathematical formulas or run software programs. Even with required resources, deter- mining the life cycle of a particular part is difficult. One survey respondent admitted to understanding the benefits of predictive models, but unfortunately does not have the time or money to use them. DART, on the other hand, has found what it considers a rel- atively easy method to calculate part replacement intervals. The software program is available through Oliver Interactive, Inc, and is called Relcode (13). Using Weibull mathematics, Route Number Service Miles Hours Average Daily mph 61 City 533 64 9 100 Interstate commuter 1,974 31 32.5 TABLE 7 COMPARISON OF ROUTE MILES

the program determines probabilities of component failure and helps the user decide whether to replace parts as a preventive measure (i.e., in advance of failure) or only on failure. Relcode requires two basic user inputs. One is data from agency records, including the life history of each part being analyzed such as how long each bearing, belt, or shaft lasted before it was replaced. The other user input consists of esti- mating (1) the cost of carrying out a preventive replacement of the component being analyzed, and (2) the cost of replacing the component in the event of an in-service failure. Reports are provided in the form of graphs and tables. One example, shown as Figure 1 plots the probability of survival over time for a drive belt. DART uses Relcode software for a variety of tasks such as documenting fleet defects, identifying premature failures, estimating budgets and stocking levels for common replace- ment parts, and to determine optimal PM parts replacement cycles. Concerning parts replacements, DART examines the mileage or time when failures occur, takes the average of all failures identified, and uses the program to determine if it is more cost-effective to replace parts at certain intervals or replace only on failure. The software has allowed DART to identify a variety of failure patterns such as: 18 • Premature failures, • Failures resulting from design defects or issues related to installation, • Failures that occur when the item is not suitable for its intended application, • Random pattern failures, • Failures affected by external factors, • Wear out failure patterns (when item reaches end of its service life) – Gradual increase in failures – Sudden sharp increase in failures. Required inputs consist of the number of parts in service, the date parts were placed in service, cost of the part, cost to replace the part, and date when each part fails. Reports gen- erated by the program are used as a decision-making tool to determine when to replace the part in question. Figure 2 shows that it is more cost-effective to replace a particular brake valve after it fails. Depending on the part, the agency examines the data and weighs whether a failure could result in a road call or other service interruption. Because the part shown in the Figure 2 example affects braking, the agency would replace it at an optimal interval. Not all parts are analyzed for PM replacement purposes, and DART is quick to note that Relcode is not the only tool FIGURE 1 Probability of component lasting to a given age.

19 used to determine part replacement intervals. In addition to brake valves, it has also been used by DART to study hydraulic fan controllers, automatic voice announcement system controllers, brushless motors, and to project optimal engine overhaul intervals. Another software model that calculates part replace- ments is available from the ReliaSoft Corporation, which also uses Weibull mathematics (14). Figure 3 plots the cost per unit time versus time to determine the minimal cost of replacement. As shown in Figure 3, corrective replacement costs increase as the replacement interval increases. According to the model, the less often a PM action is performed the higher the cor- rective costs. The longer a component is allowed to operate its failure rate increases to a point where it is more likely to fail, thus requiring more corrective actions. This figure is a theoretical example and is not based on any specific bus part or component. The opposite is true for the preventive replacement costs. The longer you wait to perform PM, the lower the costs; if you do PM too often, costs are greater. When both cost curves are compared there is an optimum point that minimizes replace- ment costs. To arrive at this one must strike a balance between the risks (costs) associated with a failure while maximizing time between PM actions. The mathematical model formulated by Reliasoft to deter- mine PM replacements is attached as Appendix E. Another method for calculating mean time between failures is described in a paper by Mondro prepared for IEEE (15). Unlike complex calculations that can be awkward, Mondro presents a much simpler method for calculating mean time between failures without appreciable sacrifice in accuracy. Despite the claim, the model does require what appears to be a detailed calcula- tion. Agencies may find computer-based software tools such as the one used by DART easier to use. Additional information on the application of Weibull math- ematics to predict parts replacements is found in Bus Fleet Management Techniques Guide (4) described in the Literature Review section of chapter two. FACTORS THAT AFFECT PREVENTIVE MAINTENANCE SCHEDULING Manufacturers’ Specifications OEMs have established specified intervals for determining when PM activities take place. Table 8 shows survey responses for the most common OEM specifications responsible for shaping agency PM intervals. Again, the diversity is great. Reasons for the diversity include dissimilar equipment, different operating and environmental conditions, and that OEMs typically offer a choice of specific intervals depending FIGURE 2 Relcode report showing the least expensive replacement interval.

on how their equipment is used. In the case of extended trans- mission fluid changes shown in Table 8, the longer intervals (70K to 100K miles) are typically a function of using syn- thetic fluids. Although more costly to purchase, some transmission OEMs specify it. Rockford Mass Transit con- ducted a return on investment study and found that extended transmission fluid drain intervals and labor savings justi- fied the extra purchase cost. Other single-entry responses not shown in Table 8 included differential flushing, engine tune up and adjustments, and inspections of steering, CNG, fire suppression, farebox, and camera equipment. Although adhering to OEM specifications is critical, find- ing a single schedule that represents all bus components can be difficult. A suggestion made by several survey responders is to require bus OEMs to consolidate all PM specifications into one document location as part of APTA’s Standard Bus Procure- ment Guidelines. Another is to consolidate all replacement parts information into a single document and make it available in electronic format. Influence of OEM Warranty Coverage When asked if OEM warranties influence the setting of PM intervals, 74% of survey respondents claimed it does. Not abiding by OEM specifications for PM can be grounds for dismissing warranty claims if the OEM can prove the prob- lem was caused by neglected maintenance. One agency noted that OEMs often look for evidence of improper service 20 or maintenance as a way of voiding warranties and related costs, implying that agencies strictly follow OEM specifi- cations to ensure reimbursement of warranty claims. Although the responses show most agencies meet or exceed OEM requirements to maintain warranty coverage, some are will- ing to extend intervals and risk losing warranty coverage. A sensible approach expressed by some survey responders is to clear any PM deviations with the OEM in advance of vio- lating the specified intervals. Survey responders were also asked if they change PM inter- vals after the OEM warranties expire. Of those responding, 82% continue to follow OEM intervals; 18% change them after warranties expire. Of those that change intervals, some actually increase intervals (i.e., perform them less often) after warranty coverage expires, whereas some make the intervals more frequent. VTA (San Jose), after careful evaluation of several factors including equipment failure rates, found that more frequent maintenance intervals were required after the warranty expires and buses age. One agency increases PM intervals of batteries because of extra loads placed on them by onboard electrical and electronic systems. Others use oil and fluid analysis to extend intervals, whereas another agency extends change intervals for spark plugs and coalescing filters used in CNG engines. Data Driven Factors Changes made to PM intervals as a result of specific operat- ing and environmental conditions are provided in the next four sections. Based on the wide range of preventive measures Time, t Corrective Replacement Costs Preventive Replacement Costs Minimum Cost of Replacement Co st p er u ni t T im e FIGURE 3 Cost curve for preventive and corrective replacement; cost per unit time versus time.

21 PM Activity No. of Responses Intervals Established as a Result of OEM Specifications Change engine oil/filter 19 8 4 1 1 6–6.5K No interval given 3–4K 5K 15K Change transmission fluid/filter 7 7 6 4 3 48–60K 70–100K No interval given 4–12K 24–36K Check brake throw 8 4 2 1 1 1 1 1 5–6.5K No interval given 1.5–2K 3K Weekly Every 2 weeks 40–60K Per state requirement Check brake thickness 6 5 2 1 1 1 1 6–6.5K No interval given 1.5–2K 3K Weekly Per state requirement Every 2 weeks Check wheel nut torque 7 6 3 1 1 1 6–6.5K No interval given (1PM) 3–4K Weekly 24K At alignment or tire change Check tire pressure/depth 8 7 3 2 1 1 1 6–7.5K No interval given Daily 2–3K Weekly Per state requirement Every 2 weeks HVAC 1 1 Bi-yearly 15K (24,000 km) Wheelchair lift/ramp 2 1 6K 30K Articulation Joint 3 6K Fuel Filter 2 12–15K Change hydraulic fluid 1 1 6K 30K TABLE 8 PM ACTIVITIES INFLUENCED BY MANUFACTURERS’ SPECIFICATIONS taken, managers will need to determine which are appropriate for their particular application. Table 9 summarizes specific PM activities put in place because of data and information generated from road call reports, driver pre/post trip reports, service line inspections, and other sources. Environmental Influences When asked what specific PM activities have been put in place because of unique environmental conditions such as extreme heat, cold, humidity, winter salt use, and other environmental factors, agency responses also varied. PM activities, corre- sponding interval and location by geographic location (state) are shown in Table 10. Operating Conditions Table 11 summarizes specific PM activities in place because of unique operating conditions such as duty cycle, terrain, road conditions, and other such factors. Fleet Makeup Table 12 summarizes specific PM activities taken because of agency fleet make-up including diesel buses equipped with

diesel particulate filters (DPF) and other emissions control technologies, CNG buses, paratransit buses, etc. As noted in Table 12, three agencies report inspecting CNG tanks every three years or 36,000 miles. This is a requirement of FMVSS 304, which also includes a provision that CNG tanks be labeled with this interval (16). Needed Tools and Resources Survey responses regarding what information, tools, or addi- tional resources would help agencies do a better job to estab- lish PM intervals include: • More staff and resources to analyze data, failure trends, times, etc.; good supervision; and well-trained personnel (eight responses). 22 • More precise presentation of PM information from man- ufacturers, such as comprehensive PM tables located in one central document (three responses). • Sharing of agency PM information on emissions, engines, and transmissions. • Examples of other transit PM schedules. • Specified life expectancy of equipment. • More reasonable PM schedules and specifications from the OEMs. Most are reasonable; however, sometimes one can detect a few unreasonable intervals that are probably designed to “cover” the OEM or component manufacturer. • A dedicated QA program to analyze failures and sug- gest updates to preventive maintenance procedures. • Software to track actual idle time to achieve a more precise accountability of both engine and transmission life cycles. PM Activity Based on Data/Information Interval Electrical connections/wiring As needed HVAC Spring and Fall, seasonal, bi-weekly Check batteries/charging 6K (3); 12K; 24K Check air system 12K; 18K; air dryer 24K Check belts, tensioners 5K; 6K Bumper mounting bolts Visual 6K, re-torque and paint at 24K Check hoses and cables Various Brake inspection 6K (3); every time bus comes over a pit Camera system inspection 6K; 24K Diesel particulate filter check 6K; 12K; 24–30K replacement Farebox PMI 6K; 24K Automatic vehicle location PMI 12K Engine-related activities 6K Oil change more frequently owing to high soot 3K Fuel management system 6K Check potential chaffing of various lines to previous fires Various intervals Rear door inspection 6.5K Tire pressure 10 days minimum Multiplex system batteries Bi-annual Engine thermostats 60K Tow air check valve 100K Safety inspection Every time bus is in the shop or every 20 days Lift/ramp inspection 6K; bi-weekly Centrifugal oil filter cleaning 24–48K TABLE 9 PM ACTIVITIES INFLUENCED BY DATA AND INFORMATION

23 PM Activity Based on Unique Environmental Conditions Interval State (no. of responses) HVAC PMI HVAC filter change Auxiliary heater inspection Spring and Fall Spring 6 m onths Annual Annual, in addition to 6K 3K Every PM 6.5K 6K 6K PA CA NY, TX, NC, WA (2) MD, IL TX FL (2) FL NM WI IL, WA Wheelchair lift/ram p PMI Daily by drivers 60 days PA FL Air dryer service Air dryer overhaul Drain air tanks Seasonal Bi-annual Bi weekly in Oct.–April WA GA DC Oil and filter change 3K FL Glow plug system service Annual MD Radiator cleaning 6K 6 m onths 30 days WI, CA, MD, FL NY FL Seasonal inspection Spring Fall CA CA, UT Diesel particulate filter service 3K MI Underside and radiator cleaning, and inspecting and main taining electrical syste ms owing to salt environm ent As needed UT TABLE 10 PM ACTIVITIES INFLUENCED BY UNIQUE ENVIRONMENTAL CONDITIONS

24 PM Activity Based on Fleet Make-Up Interval (no. of responses) CNG - replacement spark plugs - drain oil from fuel filters - inspect tanks - clean vent lines - drain CNG filters - check/service methane detection system 24K (1), 36K (1), varies (1) Every PMI (1) 3 years (2), 36K (1) Annually (1) Weekly (1) 12K (1) í07 Diesel - crankcase ventilation/filter - diesel particulate filter service - exhaust gas recirculation service 24K (1) 6K (2), not yet established (1) 3K (1) Use of brake tester PMI, accident or driver complaint (1) Fire suppression system 12K (1) Paratransit PMI 4.5K (1), various (1) PM Activity Based on Unique Operating Conditions Interval HVAC filters Heating system inspection Bi-weekly Annually Suspension 6 m onths Brake inspection Replace brake application valves Replace rear brake cham bers Air dryer inspection/service Replace slack adjusters Every 2 weeks 150K 150K Annual (Oct.–Nov.) 100K Radiator cleaning Radiator/charge air cooler cleaning 1–2 tim es per year as required, owing to plugging. Usually as a result of operating near construction sites or high dust areas 3K (approx.) Tire sidewall check Daily (service line check) Engine oil and filter change at less than OEM rrecommendation 3K 7.5K Change engine air filter 6K Under-carriage inspection 6K TABLE 11 PM ACTIVITIES INFLUENCED BY UNIQUE OPERATING CONDITIONS TABLE 12 PM ACTIVITIES INFLUENCED BY FLEET MAKE-UP