Fuel Usage Factors in Highway and Bridge Construction (2013)

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.

53 Chapter 3 presents the processes and results of the data collection phase. This phase involves the application of the three research methodologies: the engineering analysis, the contractor fuel usage survey, and the BidTabs statistical analysis. The study team designed each of these approaches to provide independent calculations of fuel use for highway construction activities. Each section of this chapter provides a sequential description of the research process undertaken and the results for each work item examined. 3.1 Contractor Fuel Usage Surveys This report section details the project team’s efforts to collect fuel use consumption informa- tion from the highway contracting community. This effort aided in the identification of heavy fuel use activities and allowed the project team to establish current levels of fuel use across a variety of construction activities and project conditions. The project team utilized several sur- veys, including an Excel spreadsheet tool and several industry-segment-specific SurveyMonkey surveys, to elicit contractor responses. In order to maximize contractor participation, the project team strived to ensure the cooperation of several industry organizations. This section contains four subsections. The first subsection describes the survey effort methodology, including survey design and dissemination, as well as the industry collaboration process. The second subsection displays respondent biographical information. The third sub- section presents the acquired survey data. The fourth subsection summarizes the chapter and offers conclusions. 3.1.1 Survey Methodology This section describes the methodology employed by the study team to design and disseminate the contractor fuel usage survey. It describes industry cooperation, survey design, survey review and approval, initial survey dissemination, and efforts to improve the survey response rate. Initial Industry Cooperation This survey effort benefitted from the support of several industry organizations. Soliciting support from industry organizations was a tactic that was strongly recommended by both the project review panel and several contractors during the initial survey effort. The American Road & Transportation Builders Association (ARTBA), the Associated General Contractors of America (AGC), the National Asphalt Pavement Association (NAPA), and the American Concrete Pavement Association (ACPA) each agreed to cooperate with the project team and aid in survey review and dissemination. Exhibit 3-1 displays the contacts within each organization, their titles, contact information, and statements of support for the project. C h a p t e r 3 Findings and Applications

54 Fuel Usage Factors in highway and Bridge Construction The project team also contacted two other organizations whose contact information was provided by a member of the NCHRP project panel. These organizations are the National Association of Minority Contractors (NAMC) and the Associated Minority Contractors of America (AMCA). The executive director of NAMC responded that the NAMC membership would likely not respond in large enough numbers to be statistically significant and declined to participate. The project team attempted to contact AMCA several times but did not receive a response. Survey Design The project team originally planned to conduct the contractor survey of fuel usage using SurveyMonkey. However, the survey design envisioned by the project team was found to be impracticable using this tool. The survey design necessitated sorting pay items by the contractor’s primary state of operation as well as major areas of work, adding a level of complexity to the survey design. SurveyMonkey could only handle this complexity if respondents manually entered work item information, fuel consumption quantities, and units. The survey returns would then have to be manually compiled by the project team. This realization led the project team to a new survey template. The survey, constructed in a user-friendly Excel format and entitled the Contractor Fuel Usage Survey (CFUS), contained the following features: • An introductory page; • A contact page; • A page for background information (name of firm, state, areas of work, etc.); • A fuel consumption information page; and • A submission page. The introductory page provided a brief description of the project’s goals and background. It also provided a link to the official project description on the TRB website. The project team took special care to emphasize that individual firms will remain anonymous in all publicly available research products. Organization Liaison Title Statement of Support AGC Senior Director, Highway and Transportation Division "AGC believes the Fuel Usage Factors Survey is a very worthwhile project that will provide useful information for the construction industry as well as state departments of transportation. AGC is willing to disseminate the survey to our contractors involved in highway, bridge, transit, and other transportation infrastructure construction. Once the results are received, AGC is equally committed to disseminating the results to our state chapters and contractor members." ARTBA Vice President of Policy and Senior Economist Verbal commitment NAPA Vice President of Legislative and Regulatory Affairs "NAPA will be happy to help in any way possible to help ensure the success of this project including assistance with reaching the industry during the survey process." ACPA Vice President of Highway and Federal Affairs "We appreciate the opportunity to provide input to this important effort." Exhibit 3-1. Initial industry contacts and commitments.

Findings and applications 55 The contact page contained contact information for a member of the project team (subject matter questions) and the NCHRP Project Officer (study background, legitimacy, and other concerns). The background information page was designed to collect information similar to the initial contractor survey. This page inquired about the respondent’s name, position, firm name, state, size, and whether a firm works in urban, suburban, or rural areas, among other data points. The respondent’s selections of work category(s) determined the work items available on the fuel consumption page. The next page in the survey was the fuel consumption information page. Based on the work category(s) selected, the respondent was able to supply fuel consumption information for particular work items. Respondents were able to provide fuel consumption information on as many items as they wish. Units of measure were fixed, although the respondent had the option to fill in an alternative unit of measure. Exhibit 3-2 provides a screenshot of a portion of the fuel consumption information page of the survey. The darker areas were locked and could not be altered. The lighter areas allowed the respondents latitude in their responses. A “Notes” column was present to the right of the “Gallons of Fuel Use per Unit” column. The final page was the submission page. This page thanked the respondents for their time and effort. It also provided instructions for submitting the completed survey. Respondents were then asked to save their completed survey and email it to an email address dedicated to this survey collection effort. The project team was able to download survey responses from this email and organize them into a more manageable Excel database. Draft Survey Review and Approval The project team submitted a draft of the Contractor of Fuel Usage Survey (CFUS) and an accompanying memorandum to the NCHRP project officer on May 27, 2011. The project officer set the deadline for comments as June 20, 2011. The panel provided two comments. In response to a comment regarding bridge demolition work items, the study team renamed the bridge work category “Bridge Construction and Demolition” and added two additional work items: complete structure demolition and deck removal. The other comment was an inquiry of a technical nature regarding how to properly use the survey tool, which was resolved shortly thereafter. ARTBA reviewed the survey, provided an email sharing approval, and offered to distribute the final version to ARTBA members. An email sent to the project team dated June 20, 2011, read in part, “. . . I think that looks good and we are happy to help out. Just let us know about distribution when the time comes—I am happy to forward this to our contractor members.” Work Item Description Unit of Measure Alternative Unit of Measure (if different) Gallons of Fuel Use per Unit Clearing - heavy Acres Clearing - medium Acres Clearing - light Acres Structure demolition Acres Pipe removal Each Pavement Removal Linear feet Clearing Exhibit 3-2. Fuel use consumption screenshot.

56 Fuel Usage Factors in highway and Bridge Construction AGC likewise reviewed the survey, provided an email indicating their support, and offered to distribute the survey to AGC members. An email to the project team dated June 20, 2011, read in part, “. . . I think the survey is fine and will forward to our members when finalized.” The study team also worked with NAPA and several of their member contractors to improve the survey form. Specifically, NAPA input led the study team to split the asphalt work category into two separate production and hauling/placing work categories. The study team also utilized NAPA members to conduct a test of the survey. As suggested by NAPA, the project team also contacted the American Concrete Pavement Association (ACPA). The ACPA signaled their support of the project effort and committed to distributing the survey to ACPA members. Initial Survey Dissemination The final survey form was distributed through the Oman Systems contact database and the above industry associations on July 11, 2011, with a deadline for submission of July 29, 2011. The study team subsequently extended the submission deadline to August 15, 2011, in an effort to increase the number of survey returns. Unfortunately, this effort resulted in only 16 survey returns. In an effort to elicit a greater number of survey responses, the study team then acquired the ARTBA official membership list on August 23, 2011. This list contains over 2,600 members. The project team devoted significant staff resources to calling as many ARTBA members as practicable. Between September 12, 2011, and September 30, 2011, calls to more than half of the ARTBA membership resulted in eight additional survey returns. In total, the initial dissemination effort and subsequent phone drive effort yielded a total of only 24 survey returns. Subsequent Efforts to Improve Response Rate In a further effort to increase survey participation, project staff consulted with officials from NAPA. The project team and NAPA decided to test a simplified survey directed toward a single industry segment employing SurveyMonkey. Both the project team and NAPA felt that an industry-specific survey would remove the need for a more sophisticated design, such as in the Excel version of the survey, by significantly limiting the number of work categories and items. The project team submitted an asphalt-specific SurveyMonkey survey to NAPA on November 9, 2011. The NAPA survey was distributed on November 15, 2011, and garnered 89 responses within 3 days of release and 151 responses by January 9, 2012. With the encouraging results from the NAPA survey in mind, the project team reached out to ARTBA, ACPA, and the National Ready Mixed Concrete Association (NRMCA) to inquire if they would release similar SurveyMonkey surveys. Although NRMCA did not believe that their members would respond in large numbers, they did provide the project team with internally conducted survey data regarding concrete hauling and delivery fuel use. This information contains responses from 84 concrete contractors. ARTBA and ACPA committed to disseminating SurveyMonkey surveys. Like the Excel survey, the ARTBA SurveyMonkey survey contained each work item. The work categories and items were listed sequentially. ARTBA and ACPA disseminated their versions of the survey on January 9, 2012, and January 19, 2012, respectively. 3.1.2 Respondent Company Information A total of 186 contractors replied to the Excel and SurveyMonkey solicitations conducted for this effort, in addition to the 84 NRMCA respondents. In addition to inquiring about their task-specific fuel usage, these two survey forms also allowed respondents to provide information about their companies. This section presents relevant company metrics, including region, size, whether they are located in urban or rural locations, and methods of fuel use calculation.

Findings and applications 57 The survey of fuel usage was advertised to contractors across the country. The respondents hail from 37 states. When sorted by the U.S. Census Bureau’s official regional designations, the group of respondents includes 75 contractors from the South, 62 from the Midwest, 27 from the Northeast, and 22 from the West. Exhibit 3-3 displays regional and subregional locations of responding contractors. Participating firms vary widely in size from 50 or fewer employees to over 500 employees. Exhibit 3-4 displays the number of employees for the responding firms. More than two-thirds of respondents report company sizes between 50 and 500 employees. 20 100- 42 0-499 46 500+ 24 1-49 33 50-99 41 199 Exhibit 3-4. Size of responding construction firms. Exhibit 3-3. Regional and subregional locations of contractors. 12 50 4 23 38 9 28 7 15 0 10 20 30 40 50 60 W es t N or th C en tr al Ea st N or th C en tr al N ew E ng la nd M id dl e A tl an ti c So ut h A tla nti c W es t So ut h C en tr al Ea st S ou th C en tr al M ou nt ai n Pa ci fic Midwest Northeast South West

58 Fuel Usage Factors in highway and Bridge Construction The responding contractors perform work in a variety of terrain types. Exhibit 3-5 displays how many contractors perform in urban or rural areas, or both. Nearly half of the respondents (83) perform work in suburban areas or a mix of urban and rural environments. Respondents also provided information on how they calculate fuel costs during the estima- tion process. More than three-quarters of the respondents (142 out of 186) calculate costs using equipment-specific consumption rates. Exhibit 3-6 displays contractor methods for calculating fuel costs. 3.1.3 Fuel Consumption Data Collected This section presents an overview of the data collected from all of the survey efforts and survey instruments. In total, respondents provided fuel consumption information for over 40 different activities. As stipulated in the outreach efforts to highway construction contractors and organi- zations, this report provides results as an average of the valid responses for each activity and does not provide information reported by individual respondents. The fuel consumption estimates represent the simple mean (average) of all of the responses that met two criteria. The first criterion was that the respondent provided the estimate in either the Subur Rural 54 Urban 45 ban/Both 83 Exhibit 3-5. Typical project locations for responding contractors. Fuel Consumption Rates (by equipment) Percentage of Total Cost Use DOT Fuel Use Factors Percentage of Equipment Cost Other 5 6 7 0 2 26 0 40 60 80 100 12 142 0 140 160 Exhibit 3-6. Contractor methods for calculating fuel costs.

Findings and applications 59 default unit suggested in the survey or an alternative unit that the project team could convert to the base unit with a conversion factor. For example, for most activities, a subset of respondents reported results in terms of gallons consumed per hour. Conversion of these estimates to gallons per unit of work was not possible without assuming a production rate. The second criterion was that each individual response included in the estimate had to be within a range that the engineering staff judged to be reasonable. In some instances, respondents provided estimates that varied from the majority of estimates by a factor of 10 or more. For example, one respondent provided fuel consumption per unit estimates for the six types of milling that ranged from 8.8 to 9.5 times greater than the mean estimate for all of the respondents. In each case, this respondent’s estimate was at least 6.6 times higher than any other estimate. These out-of-range estimates were not included in the calculation of mean values. Exhibit 3-7 provides the summary of the mean quantities of fuel reported per unit of activity. The first column of the exhibit describes the general category of work, such as clearing, grad- ing, milling, and asphalt or concrete paving. The second column describes the specific item, such as grading items that vary according to on-road and off-road, short and long haul, and soil type and milling items that vary according to depth and haul length. The third column presents the mean estimate of gallons of fuel consumption, while the forth column lists unit of measurement, such as gallons per cubic yard. The final column provides the number of observations in the survey sample. In total, survey respondents provided over 500 individual fuel usage observations. Comparisons of fuel usage across activities were difficult because the units are not comparable. For example, the estimate for clearing was 183.33 gallons per acre while the estimate for pipe removal was 1.75 gallons per linear foot. Since the two estimates were in different units (acres and linear feet), the gallons of fuel use were not directly comparable and, therefore, the exhibit did not rate one as more fuel intensive than the other. However, some comparisons within categories were possible and, in general, the observable differences followed expected patterns. For example, grading in rocky conditions is generally more fuel intensive than grading sandy or dirt soils. Milling to greater depths and using longer hauls is more fuel intensive. Warm mix asphalt requires less fuel than hot mix. Therefore, one may conclude that the survey results were internally consistent as fuel intensities, where comparable, followed expected patterns. 3.1.4 Conclusion This chapter presents the efforts of the research team to engage the contracting community and ascertain their fuel usage for highway and bridge construction activities. In total, the research team invited over 10,000 contractors to participate. Further efforts to maximize participation include collaborating with several contractor organizations, creating multiple versions of the survey form, and directly calling over 1,000 highway contractors. The contractors that responded to these efforts specialize in heavy construction activities and encompass a wide variety of locations, working conditions, and firm sizes. In studying the results of the surveys it became clear to the team that certain types of informa- tion are easier to collect than other types. The response rates for general questions that can be answered with little or no additional effort were very high. However, the more detailed questions that required additional analysis resulted in a low response rate. Based on the examination of the comments and discussions with selected respondents, the study team concluded that the reason for lower-than-expected response rates on these questions generally related to the type of data collected by the contractors. The survey questions were

60 Fuel Usage Factors in highway and Bridge Construction designed to capture “fuel use per unit of work.” A large percentage of contractors collect internal data based on equipment usage (gallons per hour), but they do not calculate or retain data based on units of work (per ton or per cubic yard). This made it difficult for respondents to supply meaningful data to the more general question of “fuel use per unit of work,” since they look at each project (and tasks within each project) with the specific set of requirements for that project. The number of observations was sufficient to constitute a valid sample for most work items. With the exception of several outlying responses that would have skewed the calculated averages, Clearing (all 01ercA333.381)sepyt Pipe 3.F.L057.1lavomeR Pavement 3.Y.S576.2lavomeR Off-Road Short 8.Y.C283.0)triD( On-Road Short 7.Y.C603.0)triD( Off-Road Long 9.Y.C503.0)triD( On-Road Long 8.Y.C273.0)triD( Off-Road Short 9.Y.C253.0)kcoR( On-Road Short 7.Y.C114.0)kcoR( Off-Road Long 8.Y.C044.0)kcoR( On-Road Long 7.Y.C294.0)kcoR( Borrow 5.Y.C096.0)kcoR( Strip Topsoil 7.Y.C244.0)triD( Respread Topsoil 7.Y.C814.0)triD( Roadway Finishing 9.Y.S291.0)triD( 0-1" (0-5 mile 9.Y.S820.0)luah 0-1" (6-15 mile 9.Y.S630.0)luah 0-1" (>15 mile 9.Y.S640.0)luah 2-4" (0-5 mile 21.Y.S250.0)luah 2-4" (6-15 mile 9.Y.S470.0)luah 2-4" (>15 mile 9.Y.S890.0)luah Small Pipe 5.F.L006.1werC Medium Pipe 6.F.L233.2werC Large Pipe 5.F.L803.3werC Hot Mix Structural (Place and Compact) 0.970 Ton 27 Hot Mix Surface (Place and Compact) 0.989 Ton 25 Hot Mix Leveling (Place and Compact) 1.026 Ton 25 Warm Mix (Place and Compact) 0.772 Ton 20 61noT573.0gniluaH Prime and 71noT490.0kcaT Plant 41noT489.1)leseiD( Plant (Natural Gas - BTU) 268806.308 BTU 15 Plant (Support 31noT111.0)tnempiuqE Concrete 7.Y.C003.0gnivaP Concrete 48.Y.C041.1gniluaH Concrete elciheV009.2gniluaH Hour 63 Structural Concrete Structural 01.Y.C035.6etercnoC 516 Asphalt Concrete Paving Total Observations Item Number of Observations Milling Storm Pipe Gallons of Fuel Consumption Unit Clearing Grading Category Exhibit 3-7. Survey-based estimates of mean quantities of fuel per unit of activity.

Findings and applications 61 the fuel usage estimates provided by the contracting community were within a reasonable range of accuracy as determined by the research team’s engineering experts. Results within categories demonstrated consistency as well. The survey results provided utility throughout the remainder of the project, especially as a means to complement and verify the engineering results. 3.2 Engineering Analysis of Fuel Usage The objective of the engineering analysis was to estimate the fuel usage of construction activ- ities using engineering cost estimating techniques. The results of this effort, in conjunction with the statistical analysis and CFUS, allowed the project team to formulate new and updated fuel usage factors. Building on the results from the initial engineering analysis, which aimed to identify high fuel use activities, the project team extended the analysis to calculate the fuel use per unit for each work task. Using the initial phase calculations as well as estimated quantities of work for a typical project, the project team was able to estimate a fuel usage factor for each work task. This report section is divided into six subsections. The first subsection describes the expert panel used to develop data elements throughout this effort. The second subsection describes the creation of the list of typical construction equipment and the tabulation of equipment-specific fuel consumption rates. The third subsection describes the creation of the list of construction tasks for which fuel use was estimated. The fourth subsection describes the process of assigning equipment and crew production rates for each work task. The fifth subsection describes the process of calculating per unit fuel usage rates and presents the results. The sixth and final subsection provides conclusions and next steps. 3.2.1 The Expert Panel As in the initial engineering analysis, the study team utilized an expert panel of four construction engineers and estimators. Each panel member employed their industry expertise to compile a list of construction activities, assign equipment and crews to work tasks, and calculate production rates. Panel members A, B, and C each independently calculated fuel use per unit for each work task. Panel member D acted as a mediator during this effort and investigated discrepancies, resolved differences in calculations, and compiled the results. 3.2.2 Equipment Fuel Use The first step in the data development process was the compilation of fuel use by equipment type. For this step, the study team first created a general list of construction equipment commonly used in highway construction. Key data sources for this effort include the 40th edition of the Caterpillar Performance Handbook, which estimates the performance of a wide variety of construction equipment, and historical contractor data. Fuel usage estimates for other equipment were developed using the expert panel. The fuel consumption rates are listed in “gallons per hour” and are for “average” working conditions. These values are derived from manufacturers’ operating handbooks for the major pieces of equipment as well as estimator experience for the minor equipment. The equipment list is based on typical construction practices. The list of equipment that is available to a contractor can have an impact on the crew makeup, production rates, and the ultimate cost of a work activity. The project team created a list of equipment that is generally used within the heavy construction industry and avoided specialty equipment where possible. Exhibit 3-8 displays the items of equipment, fuel types, and fuel consumption in gallons per

62 Fuel Usage Factors in highway and Bridge Construction Equipment Description Fuel Type Fuel Consumption (GPH) Source A 0.32 leseiD 01-D rezoD A 0.0 leseiD YBDNATS 01-D rezoD A 2.2 leseiD 3-D rezoD A 0.4 leseiD 5-D rezoD A 0.5 leseiD 6-D rezoD A 0.5 leseiD LATNER 6-D rezoD A 0.9 leseiD 7-D rezoD A 0.21 leseiD 8-D rezoD A 0.21 leseiD LATNER 8-D rezoD A 0.61 leseiD 9-D rezoD A 0.61 leseiD LATNER 9-D rezoD A 0.61 leseiD rotcarT hsuP 9-D rezoD A 0.5 leseiD 513 taC rotavacxE A 0.7 leseiD 423 taC rotavacxE A 0.11 leseiD 633 taC rotavacxE A 0.51 leseiD 543 taC rotavacxE A 0.0 leseiD ERAPS 543 taC rotavacxE A 0.5 leseiD 613 taC T/R rotavacxE A 0.11 leseiD mareoH /W rotavacxE A 0.04 leseiD 0315 taC levohS tnorF/rotavacxE A 0.6 leseiD wk 051 taC rotareneG A 0.3 leseiD wk 53 taC rotareneG A 0.7 leseiD noT 52 detalucitrA kcurT luaH A 0.8 leseiD noT 03 detalucitrA kcurT luaH A 0.11 leseiD noT 04 detalucitrA kcurT luaH A 0.32 leseiD noT 001 digiR kcurT luaH A 0.21 leseiD noT 05 digiR kcurT luaH A 0.61 leseiD noT 07 digiR kcurT luaH A 0.4 leseiD 839 T/R redaoL A 0.3 leseiD 419 taC T/R redaoL A 0.8 leseiD 089 taC T/R redaoL A 0.13 leseiD C299 taC T/R redaoL A 0.4 leseiD 059 taC T/R redaoL A 0.3 leseiD reetS/dikS redaoL A 0.7 leseiD 359 taC kcarT redaoL A 0.31 leseiD 379 taC kcarT redaoL A 0.31 leseiD LATNER 379 taC kcarT redaoL A 0.3 leseiD 614 taC eohkcaB/redaoL A 0.5 leseiD 034 taC eohkcaB/redaoL A 0.4 leseiD 83TI taC reirraC looT/redaoL A 0.6 leseiD 21 taC redarG rotoM A 0.9 leseiD SPG/w 41 taC redarG rotoM A 0.11 leseiD 61 taC redarG rotoM A 0.7 leseiD rotcapmoC lioS 518 relloR A 0.4 leseiD )nwodkaerB( tlahpsA relloR A 0.4 leseiD )hsiniF( tlahpsA relloR A 0.4 leseiD )eriT rebbuR( tlahpsA relloR A 5.01 leseiD rotcapmoC lioS 528 taC relloR A 0.42 leseiD 726 )niwT( reparcS A 0.03 leseiD 736 )niwT( reparcS A 0.14 leseiD 756 )niwT( reparcS A 5.8 leseiD nogaW retaW 316 reparcS A 0.41 leseiD 126 reparcS A 0.41 leseiD LATNER 126 reparcS A 0.81 leseiD 136 reparcS A 0.81 leseiD LATNER 136 reparcS B 5.0 saG llamS rotareneG B 0.2 saG naV regnessaP 9 B 0.3 leseiD renruB niatruC riA B 0.51 leseiD tnalP tlahpsA B 0.5 leseiD redaerpS redluohS enotS esaB B 0.0 leuF oN xoB redaerpS enotS esaB B 0.7 leseiD revaP tlahpsA 0023FP xonKwalB Exhibit 3-8. Construction equipment and fuel consumption rates.

Findings and applications 63 B 0.1 leseiD kcurT noitcepsnI egdirB B 0.3 leseiD moorB B 0.2 saG nadeS raC B 0.6 leseiD 008 rosserpmoC B 5.1 leseiD 581-58 rosserpmoC B 0.2 leseiD rehsiniF kceD egdirB etercnoC B 0.1 saG waS etercnoC B 0.7 leseiD revaP mrofpilS etercnoC B 0.01 leseiD relwarC noT 001 enarC B 0.5 leseiD LATNER relwarC noT 001 enarC B 0.3 leseiD kcurT noT 21 enarC B 0.3 leseiD ciluardyH noT 81-51 enarC B 5.3 leseiD relwarC noT 52 enarC B 0.6 leseiD ciluardyH noT 03 enarC B 5.6 leseiD ciluardyH noT 04 enarC B 0.5 leseiD rotoM noT 54 enarC B 0.5 leseiD relwarC noT 05 enarC B 0.7 leseiD ciluardyH noT 05 enarC B 0.8 leseiD relwarC noT 57 enarC B 1.1 leseiD tfilkroF B 0.5 leseiD 088 lladarG B 0.1 leseiD tfilnaM S006 GLJ B 0.1 saG wk6 tnalP thgiL B 0.1 leseiD )leseiD( remmaH eliP B 0.0 leuF oN LATNER )ciluardyH( teehS ,remmaH eliP B 0.4 leseiD B0075 rebruC rewoP B 0.2 leseiD "01 pmuP B 2.0 saG "2 pmuP B 5.0 saG "4 pmuP B 5.1 leseiD "6 pmuP B 0.01 leseiD ygguB elttuhS ceTdaoR B 5.6 leseiD biV 45PS RI relloR B 0.6 leseiD biV 06PS RI relloR B 5.1 leseiD hcnerT xaM maR relloR B 0.4 leseiD biV noT 21 opmaT relloR B 0.3 leseiD noT 53-52 opmaT relloR B 2.0 saG tcapmoC etalP gnitarbiV relloR Screening/Crushing Plant (P B 0.01 leseiD )elbatro B 0.2 saG 4X4 VUS B 5.4 leseiD 073MCE llirD kcarT B 3.5 leseiD )6002( 095MCE llirD kcarT B 5.6 leseiD 927MCE llirD kcarT B 0.2 leseiD goH hsuB htiw rotcarT B 0.0 leuF oN reliarT ytilitU lortnoC ciffarT B 5.01 leseiD 555-T reemreV rehcnerT B 0.2 saG noT 2/1 kcurT B 0.3 *leseiDxT noT1 kcurT B 0.3 leseiDxT redwoP noT1 kcurT B 5.3 leseiDxT debtalF noT2 kcurT B 5.2 leseiDxT noT 4/3 kcurT B 0.2 saG elciheV ytilitU 4x4 kcurT B 0.2 leseiDxT rotubirtsiD kcurT B 5.5 leseiDxT YC 41 pmuD kcurT B 0.4 leseiDxT leuF kcurT B 0.4 leseiDxT cinahceM/ecivreS kcurT B 0.5 leseiDxT retaW kcurT B 0.6 leseiDxT reliarT yobwoL & rotcarT kcurT B 0.01 leseiD enihcaM gnilliM 0002TM ovloV B 5.1 leseiD pmA 002 redleW A: CAT Performance Handbook (Ed. 40) B: Other (Historical Contractor Data) *TxDiesel is taxed diesel fuel. Tax is applied to construction equipment that travels on roads, primarily dump trucks. Exhibit 3-8. (Continued).

64 Fuel Usage Factors in highway and Bridge Construction hour, as well as the data source for each item’s fuel consumption rates. In total, fuel consumption rates are provided for 122 different pieces of equipment. The top fuel consumer is a twin scraper, which consumes 41 gallons of diesel fuel per hour. 3.2.3 Work Tasks The four members of the expert panel worked collaboratively to compile a list of construc- tion work tasks. This list includes work tasks that would be common across geographic areas as well as topographic conditions. Unlike the analysis of specific pay item data in previous efforts, this list contains specific work tasks not always unique to a pay item; there may be multiple work tasks within a single pay item on a project. For example, the excavation pay item on a project may include short- and long-haul dirt as well as rock excavation and stripping topsoil. In other cases, work tasks may relate to many different pay items. For example, there may be hundreds of pay items related to storm water structures, but only one work task for these pay items. The difference between each storm water structure is mostly due to the design of the structure and the materials used in its construction. Because this effort focused on the equipment needed to accomplish the work task, the different material or structure design would not have any effect on fuel consumption. Exhibit 3-9 displays the 66 work tasks compiled by the project team and the units used to measure the work tasks. Note that several additional work tasks were compiled following an analysis of the statistical analysis and contractor survey. These additional items are introduced and explained in Chapter 4. Nine distinct units are used to measure quantities. 3.2.4 Equipment and Production Rates The panel of estimators then used their construction experience and expertise to create a list of equipment needed to accomplish each work task. Because there are varying possible combinations of equipment used and production rates, each member of the expert panel assembled a crew that they believed would be the most efficient to accomplish the task based on the above equipment list. Assigning production rates to each task can be a subjective exercise when dealing with non- project-specific, generic activities. Each panel member utilized their own experience to establish the average production rates based on the equipment selected for the task. As with the development of fuel consumption factors, the goal in this effort is to establish production rates for average conditions that apply across many different project scenarios. The selected production rate used for each task is an average of the estimators’ evaluations. In some cases there was some relatively large variance between the estimators’ equipment choices and selected production rates. Further consultation and discussion between the estimating panel and the moderator (Reviewer D) resulted in modifications to equipment selections and production rates. Factors that affected the different production rates and were considered in establishing the agreed production rates were equipment type, average topography, hauling distances, soil conditions, and industry standards. Listed below are the major categories of work items. A chapter subsection is dedicated to each of these categories and includes a discussion of the project conditions and equipment, as well as a table listing the work task, required equipment, unit of measure, and production rate. The major categories of work items are • Clearing and removal, • Grading, • Base stone,

Findings and applications 65 Task Description Unit noT enotS esaB ercA thgiL - gniraelC ercA muideM - gniraelC ercA yvaeH - gniraelC )toof raenil( .F.L reirraB naideM etercnoC Concrete Pavement (</= 6” )dray erauqs( .Y.S )kcihT .Y.S )kcihT ”6 >( tnemevaP etercnoC .F.L rettuG dna bruC )dray cibuc( .Y.C serutcurtS eganiarD hcaE setaG ecneF .F.LHeight) '6 revO( gnicneF .F.L ' Height)6 ot pU( gnicneF .Y.C luaH gnoL - daoR ffO - triD - gnidarG .Y.C luaH trohS - daoR ffO - triD - gnidarG .Y.C luaH gnoL - daoR nO - triD - gnidarG .Y.C luaH trohS - daoR nO - triD - gnidarG .Y.C luaH gnoL - daoR ffO - kcoR - gnidarG .Y.C luaH trohS - daoR ffO - kcoR - gnidarG .Y.C luaH gnoL - daoR nO - kcoR - gnidarG .Y.C luaH trohS - daoR nO - kcoR - gnidarG ercA )gnideeS ordyH( gnissarG hcaE stsoP liardrauG Hot Mix Asphalt - Leveling Course (0-5 Mile Haul) Ton Hot Mix Asphalt - Leveling Course (5-15 Mile Haul) Ton Hot Mix Asphalt - Leveling Course (Over 15 Mile Haul) Ton Hot Mix Asphalt - Structural Course (0-5 Mile Haul) Ton Hot Mix Asphalt - Structural Course (5-15 Mile Haul) Ton Hot Mix Asphalt - Structural Course (Over 15 Mile Haul) Ton Hot Mix Asphalt - Surface Course (0-5 Mile Haul) Ton Hot Mix Asphalt - Surface Course (5-15 Mile Haul) Ton Hot Mix Asphalt - Surface Course (Over 15 Mile Haul) Ton hcaE )enaL 2( noitazilangiS noitcesretnI hcaE )enaL 4( noitazilangiS noitcesretnI .F.L )epiP "63 >( werC epiP egraL .F.L )epiP "63 ot "81 >( werC epiP muideM .Y.S )luaH eliM 5-0( )"2<( gnilliM .Y.S )luaH eliM 51-5( )"2<( gnilliM .Y.S )luaH eliM 51 revO( )"2<( gnilliM .Y.S )luaH eliM 5-0( )"4-2( gnilliM .Y.S )luaH eliM 51-5( )"4-2( gnilliM .Y.S )luaH eliM 51 revO( )"4-2( gnilliM .Y.C tlahpsA – lavomeR tnemevaP .Y.C etercnoC – lavomeR tnemevaP .F.L seziS llA - lavomeR epiP )dnuop( .B.L leetS gnicrofnieR .F.S llaW gniniateR .Y.S gnihsiniF debdaoR Rock Drilling and Blasting (Only) (No Haul) C.Y. ercA noitaraperP debdeeS .F.L )htpeD '4 revO( eniL reweS .F.L )htpeD '4 ot pU( eniL reweS .F.L klawediS )retem raenil( .M.L gnikraM tnemevaP pikS .F.L )epiP "81 =/<( werC epiP llamS .M.L gnikraM tnemevaP diloS .Y.S gniddoS diloS .F.L smaeB leetS .F.L liardrauG leetS .Y.C liospoT pirtS hcaE noitilomeD erutcurtS .Y.C etercnoC erutcurtsbuS .Y.C etercnoC erutcurtS repuS .F.L )htpeD '4 revO( eniL retaW .F.L )htpeD '4 ot pU( eniL retaW hcaE selohnaM reweS/retaW .F.L liardrauG elbaC/eriW Exhibit 3-9. Highway construction work tasks.

66 Fuel Usage Factors in highway and Bridge Construction • Asphalt, • Milling, • Structures, • Miscellaneous concrete/concrete pavement/retaining wall, • Drainage pipe and structures/water/sewer, and • Specialty items (fencing/guardrail/landscaping/pavement marking/signalization). Clearing and Removal Clearing and removal activities can vary widely from project to project. The general assumptions to develop the equipment and production rates for these tasks relate to the density and type of materials to be removed from the site. Exhibit 3-10 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the clearing and removal work tasks. Clearing (Light) would be in areas that have only a minimal growth of trees and brush. This would generally be related to projects that are widening or where existing roads are being reconstructed. In addition, light clearing areas would contain little or no general clearing items such as fence rows or other debris. Clearing (Medium) would be in areas where the trees and brush are only moderately dense. An example of these areas would be in residential areas where trees and open areas are mixed. Work Task Unit of Measure Production Rate Production Rate Unit of Measure Clearing – Light Acre 0.225 Acre/Hour Truck 1/2 Ton (1) Dozer D-5 (1) Excavator Cat 336 (1) Tub Grinder (1) Clearing - Medium Acre 0.175 Acre/Hour Truck 1/2 Ton (1) Dozer D-6 (1) Excavator Cat 336 (1) Tub Grinder (1) Clearing - Heavy Acre 0.15 Acre/Hour Truck 1/2 Ton (1) Dozer D-5 (1) )1( 8-D rezoD Excavator Cat 336 (1) Tub Grinder (1) Structure Demolition Each 1.00 Each/Hour Truck 1/2 Ton (1) Truck Dump 14 CY (3) Loader R/T Cat 980 (1) Excavator Cat 336 (1) Pipe Removal - All Sizes L.F. 24.00 L.F./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (1) Dozer D-3 (1) Excavator Cat 336 (1) Pavement Removal - Asphalt C.Y. 50.00 C.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (4) Volvo Milling Machine (1) Broom (1) Pavement Removal - Concrete C.Y. 66.00 C.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (3) Loader R/T Cat 950 (1) Excavator W/ Hoeram (1) Exhibit 3-10. Clearing and removal summary table.

Findings and applications 67 Clearing (Heavy) would be in areas that are densely populated with trees and brush and in more virgin area projects where there are no current roads. Removal Items. The largest cost related to most removal items relates to the distance required to haul the debris. Removal items are not generally “production” type items and cycle times are not calculated in the same way grading items are calculated. The assumption is that the crew will include sufficient trucks to cycle within a 10-mile radius of the project site. Note also that the asphalt pavement removal item is separate from the major work category milling that is described later in this section. Structure Demolition includes the demolition and removal of buildings, homes, or small-to- medium-sized bridges. The range of possible time for removal and hauling of structures is much wider than for most of the other items in the study. Grading and Excavation The largest on-site consumers of fuel on highway projects are the grading items. These items are also the most variable from project to project and even within a project. The equipment utilized to perform the grading activities can also vary from contractor to contractor depending on the techniques employed by the contractor and the equipment that is available. Exhibit 3-11 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the grading work tasks. The grading activities have been separated into tasks that would require different equipment and production rates. Within a single project, one or more of these tasks may be used in the development of the excavation pay item. Based on each estimator’s experience and background, they each developed different equipment lists and production rates to accomplish each task. The end result, however, was that the fuel consumption rates for each activity were very consistent for each activity. Exhibit 3-11. Grading and excavation summary table. (continued on next page) Work Task Unit of Measure Production Rate Production Rate Unit of Measure Grading - Dirt - Off Road - Short Haul C.Y. 215.32 C.Y./Hour Truck 1/2 Ton (1) Truck Water (0.5) Dozer D-7 (1) Haul Truck Articulated 25 Ton (2) Excavator Cat 345 (1) Motor Grader Cat 12 (0.5) Roller 815 Soil Compactor (1) Grading - Dirt - On Road - Short Haul C.Y. 233.38 C.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (6) Truck Water (0.5) Dozer D-8 (1) Excavator Cat 345 (1) Motor Grader Cat 12 (0.5) Roller 815 Soil Compactor (1) Grading - Dirt - Off Road - Long Haul C.Y. 285.60 C.Y./Hour Truck 1/2 Ton (1) Truck Water (1) Dozer D-8 (1) Haul Truck Rigid 50 Ton (3) Excavator Cat 345 (1) Motor Grader Cat 16 (1) Roller Cat 825 Soil Compactor (1)

68 Fuel Usage Factors in highway and Bridge Construction Exhibit 3-11. (Continued). Rock Drilling & Blasting (Only) (No Haul) C.Y. 250.00 C.Y./Hour Truck 1/2 Ton (1) Truck 1 Ton Powder (1) Loader/Backhoe Cat 416 (1) Track Drill ECM590 (2006) (1) Strip Topsoil C.Y. 120.00 C.Y./Hour Truck 1/2 Ton (1) Dozer D-5 (1) Scraper 621 (1) Roadbed Finishing S.Y. 400.00 S.Y./Hour Truck 1/2 Ton (1) Dozer D-5 (1) Scraper 621 (1) Motor Grader Cat 14 w/GPS (1) Grading -Rock - Off Road - Short Haul C.Y. 215.32 C.Y./Hour Truck 1/2 Ton (1) Truck 1 Ton Powder (1) Truck Water (0.5) Dozer D-7 (1) Haul Truck Articulated 25 Ton (3) Loader R/T Cat 980 (1) Motor Grader Cat 12 (1) Track Drill ECM590 (2006) (1) Grading - Rock - On Road - Short Haul C.Y. 140.00 C.Y./Hour Truck 1/2 Ton (1) Truck 1 Ton Powder (1) Truck Dump 14 CY (4) Truck Water (0.5) Dozer D-7 (1) Loader R/T Cat 980 (1) Motor Grader Cat 12 (1) Track Drill ECM590 (2006) (1) Grading - Rock - Off Road - Long Haul C.Y. 240.00 C.Y./Hour Truck 1/2 Ton (1) Truck 1 Ton Powder (1) Truck Water (0.5) Dozer D-7 (1) Haul Truck Rigid 70 Ton (3) Loader R/T Cat 980 (1) Motor Grader Cat 12 (1) Track Drill ECM590 (2006) (1) Grading - Rock - On Road - Long Haul C.Y. 140.00 C.Y./Hour Truck 1/2 Ton (1) Truck 1 Ton Powder (1) Truck Dump 14 CY (11) Truck Water (0.5) Dozer D-7 (1) Loader R/T Cat 980 (1) Motor Grader Cat 12 (1) Track Drill ECM590 (2006) (1) Work Task Unit of Measure Production Rate Production Rate Unit of Measure Grading - Dirt - On Road - Long Haul C.Y. 233.38 C.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (18) Truck Water (0.5) Dozer D-8 (2) Excavator Cat 345 (1) Motor Grader Cat 16 (1) Roller 815 Soil Compactor (1)

Findings and applications 69 Base Stone Unlike clearing and grading items, the base stone category will have a more standard crew. The largest variable in the base stone task is the haul distance from the quarry to the project site, which can vary widely from project to project and state to state. In this study we have assumed a moderate haul distance of 10 to 15 miles. The equipment used for placing and compacting the stone is much more consistent from project to project. Exhibit 3-12 presents the selected work task, units of measure, production rate, and production rate unit of measure for the base stone work task. Asphalt The equipment list for the asphalt category is relatively standard from contractor to contractor. The specific types of pavers, rollers, and other support equipment vary from contractor to contractor, but the overall fuel consumption would not vary significantly. The two main variables in asphalt activities relate to the project conditions and the haul distance from the plant to the project site. Exhibit 3-13 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the asphalt work tasks. The primary project conditions that can affect production rates for lay down operations are traffic conditions, pavement depth, pavement width, and lengths of runs. In this exercise, the project team assumed “general” conditions for each of these factors. Projects with long uninterrupted runs will exceed the listed production rates and projects with high traffic interference and many intersections will fall short of the listed production rates. The most variable cost of asphalt operations is the haul distance from the plant to the project. In order to minimize this effect on the fuel use, the project team broke down each of the three main asphalt activities (structural, surface, and leveling courses) into three haul distance ranges. Each of the three haul distances (0-5 miles, 5-15 miles, and over 15 miles) increases the number of trucks required to service the lay down crew and increases the amount of fuel consumed. Leveling course asphalt has the lowest production rate of the three types of asphalt operations. This task typically involves smaller quantities and larger areas, resulting in slower lay down operations. Structural course asphalt has the highest production rate and is typically completed with larger quantities and thicker courses than the other mixes. Surface course asphalt requires more attention to the finished surface and is typically done with thinner courses (1″–2″) and consequently is slower to place than structural courses. Milling The milling category will have the most standardized crew among the work categories examined. Although there are different sizes of milling machines and the production rates can vary based on the material being milled, all the equipment lists and production rates were similar across all estimators. The largest variable in calculating the production rate for a milling item is the haul Work Task Unit of Measure Production Rate Production Rate Unit of Measure Base Stone Ton 217.00 Ton/Hour Truck 1/2 Ton (1) Truck Dump 14 CY (10) Truck Water (1) Dozer D-5 w/Spreader Box (1) Motor Grader Cat 14 w/GPS (1) Roller Tampo 25-35 Ton (1) Screening/Crushing Plant (Portable) (1) Exhibit 3-12. Base stone summary table.

70 Fuel Usage Factors in highway and Bridge Construction Work Task Unit of Measure Production Rate Production Rate Unit of Measure Hot Mix Asphalt - Structural Course (0-5 mile haul) Ton 200.06 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (6) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Surface Course (0-5 mile haul) Ton 150.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (6) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Leveling Course (0-5 mile haul) Ton 130.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (6) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Structural Course (5-15 mile haul) Ton 200.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (11) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Surface Course (5-15 mile haul) Ton 150.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (8) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Leveling Course (5-15 mile haul) Ton 130.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (8) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Exhibit 3-13. Asphalt summary table.

Findings and applications 71 distance from the project site to the disposal site. As mentioned previously, these distances can vary dramatically from project to project and state to state. In this study the researchers assumed a moderate haul distance of 10 to 15 miles. The equipment used for milling and hauling is consistent from project to project. Other factors that affect the production rates for milling activities relate to specific project conditions related to length of runs, number of turnouts, width of pavement, and traffic conditions. This exercise assumed an average of all these factors. Exhibit 3-14 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the milling work tasks. Structures Activities related to structures vary widely from project to project and state to state. The project team identified four main activities that are common to many structures. The estimating panel then identified the equipment needed to perform each activity. The equipment lists were fairly consistent among the estimators. The largest difference in equipment is the size of the crane that each estimator used in the calculation. There is also a large variance in the cranes that would be used by different contractors. The largest variance in the estimates is the production rates for each item. This is consistent with the idea that each structure on each project would also be unique to that project. There are many factors that can have an impact on the productivity for each of these work items. These factors include location, size, design, height, width, span, and type. The production rates used are also average productivity across the duration of the task. The concrete structure items are based Exhibit 3-13. (Continued). Hot Mix Asphalt - Structural Course (Over 15 mile haul) Ton 200.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (12) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Surface Course (Over 15 mile haul) Ton 150.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (12) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Hot Mix Asphalt - Leveling Course (Over 15 mile haul) Ton 130.00 Ton/Hour Truck 1/2 Ton (1) Truck Distributor (1) Truck Dump 14 CY (7) Truck Water (1) Roller Asphalt (Breakdown) (1) Roller Asphalt (Finish) (1) Roller Asphalt (Rubber Tire) (1) BlawKnox PF3200 Asphalt Paver (1) RoadTec Shuttle Buggy (1) Asphalt Plant (1) Work Task Unit of Measure Production Rate Production Rate Unit of Measure

72 Fuel Usage Factors in highway and Bridge Construction on the cubic yards of concrete poured. Although the actual pouring of the concrete takes place relatively quickly, the production rate accounts for the preparation, forming, pouring, wrecking, and finishing of the concrete. Exhibit 3-15 presents the work tasks, units of measure, production rates, and production rate units of measure for the structures’ work tasks. Miscellaneous Concrete/Concrete Pavement/Retaining Wall The items that make up this work category are relatively standard and the estimators were in general agreement on the necessary equipment and production rates for this section. Although concrete curb specifications can vary from state to state, the equipment required and production rates are relatively consistent. Another factor that can have an impact on the equipment used, as well as the production rate, is the ability to use a machine to slip-form the item. Some projects can have unique circumstances that require hand forming and pouring of the concrete instead of using a paver. This exercise assumed the use of pavers to perform the majority of the work. Exhibit 3-16 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the miscellaneous concrete, concrete paving, and retaining wall work tasks. Work Task Unit of Measure Production Rate Production Rate Unit of Measure Milling (<2") (0 - 5 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (3) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Milling (<2") (5 - 15 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (4) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Milling (<2") (Over 15 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (7) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Milling (2-4") (0 - 5 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (3) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Milling (2-4") (5 - 15 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (11) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Milling (2-4”)(Over 15 Mile Haul) S.Y. 6,250.00 S.Y./Hour Truck 1/2 Ton (1) Truck Dump 14 CY (20) Truck Water (1) Dozer D-5 (1) Volvo MT2000 Milling Machine (1) Broom (1) Exhibit 3-14. Milling summary table.

Findings and applications 73 Work Task Unit of Measure Production Rate Production Rate Unit of Measure Substructure Concrete C.Y. 10.00 C.Y./Hour Truck 1/2 Ton (1) Ready-Mix Truck (6) Loader/Backhoe Cat 430 (1) Excavator R/T Cat 316 (1) Crane 100 Ton Crawler (1) Pump 4 (1)" Generator Small (1) Superstructure Concrete C.Y. 10.00 C.Y./Hour Truck 1/2 Ton (1) Ready-Mix Truck (5) Loader/Backhoe Cat 416 (1) Crane 100 Ton Crawler (1) Compressor 85-185 (1) Generator Small (1) Concrete Bridge Deck Finisher (1) Reinforcing Steel L.B. 2,000.00 L.B./Hour Truck 1/2 Ton (1) Crane 30 Ton Hydraulic (1) Steel Beams L.F. 100.00 L.F./Hour Truck 1/2 Ton (1) TruckTractor & Lowboy Trailer (1) Crane 100 Ton Crawler (1) Exhibit 3-15. Structures summary table. Work Task Unit of Measure Production Rate Production Rate Unit of Measure Concrete Pavement (</= 6 Thick) S.Y. 60.00 S.Y./Hour Truck 1/2 Ton (1) Ready-Mix Truck (6) Loader/Backhoe Cat 416 (1) Concrete Slipform Paver (1) Concrete Pavement (>6 Thick) S.Y. 45.00 S.Y./Hour Truck 1/2 Ton (1) Ready-Mix Truck (6) Loader/Backhoe Cat 416 (1) Concrete Slipform Paver (1) Curb & Gutter L.F. 100.00 L.F./Hour Truck 1/2 Ton (1) Ready-Mix Truck (1) Loader Skid Steer (1) Gomaco Commander GT3200 Curber (1) Concrete Median Barrier L.F. 70.80 L.F./Hour Truck 1/2 Ton (1) Ready-Mix Truck (6) Loader Skid Steer (1) Power Curber 5700B (1) Sidewalk L.F. 100.00 L.F./Hour Truck 1/2 Ton (1) Ready-Mix Truck (6) Loader Skid Steer (1) Power Curber 5700B (1) Retaining Wall S.F. 24.00 S.F./Hour Truck 1/2 Ton (1) Ready-Mix Truck (1) Loader Backhoe Cat 430 (1) Crane 30 Ton Hydraulic (1) Exhibit 3-16. Miscellaneous concrete/concrete pavement/retaining wall summary table.

74 Fuel Usage Factors in highway and Bridge Construction Storm Drainage/Water/Sewer Pipe crews are generally consistent from project to project and generally vary by pipe size and depth. The estimating panel produced generally consistent equipment lists and production rates. This exercise assumed standard open conditions with standard specification depths for pipe. These production rates would not be for urban areas where site conditions limit the work area and for unusual depth requirements. Exhibit 3-17 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the storm drainage, water, and sewer work tasks. Specialty Items The equipment lists for most of the specialty items are much shorter than for many of the previous items. Labor and material costs make up a much larger percentage of the cost for these items. In addition, most of these items are performed by companies that specialize in the items listed and are not performed by the average highway contractor. Although the equipment lists are generally used, the production rates for many of these items can vary for each subcontractor depending on a number of project-specific factors. For example, signalization installations can vary from one intersection to the next within the same project. The estimating team relied on information from specialty subcontractors for much of the information in this section. Exhibit 3-18 presents the selected work tasks, units of measure, production rates, and production rate units of measure for the specialty item work tasks. 3.2.5 Calculation of Per Unit Fuel Use The last step in this process was the calculation of the “Fuel Use Factor” for each task. Each estimator created an estimated task quantity for each task. For example, a quantity of 1,000 L.F. of pipe was assigned for each of the pipe items. Using an estimated quantity for each crew, a total time for completing the task was established. Total fuel consumption was then calculated by factoring in the total time for the crew and the required equipment for each crew. Creating a sample quantity for each task eliminated rounding errors that occur when entering the calculations for large production rates. In addition, using larger quantities allows the estimator to better visualize the results. Exhibit 3-19 displays a sample computation for a small pipe crew. Exhibits 3-20 through 3-28 display the final per unit fuel use for the selected work tasks. The results are presented by category in order to group similar tasks. Under the clearing and removal category, the heavy clearing work task uses more fuel than light and medium clearing. Asphalt pavement removal (1.397 gallons per cubic yard) is nearly three times more fuel intense than concrete pavement removal (0.562 gallons per cubic yard). Exhibit 3-20 displays fuel use per unit among the clearing and removal work tasks. Exhibit 3-21 displays the fuel use per unit for the grading work tasks. The work tasks within dirt and rock do not demonstrably differ in fuel intensity, with the one exception of the work task involving dirt and rock with a short on-road haul. Rock takes longer to load, and the loading operation will be a larger percentage of the cost with shorter hauls. Exhibit 3-22 displays the fuel use per unit for the base stone work task. This is the only work task under the base stone category. The base stone task is estimated to use 0.406 gallons of fuel per ton of base stone. Exhibit 3-23 displays the fuel use per unit for the work tasks under the asphalt category. The leveling course work tasks have the highest average fuel use per unit, and the over 15-mile leveling course is the most fuel intensive work task.

Findings and applications 75 Work Task Unit of Measure Production Rate Production Rate Unit of Measure Small Pipe Crew (</= 18" Pipe) L.F. 24.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Medium Pipe Crew (>18" to 36" Pipe) L.F. 16.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Large Pipe Crew (> 36" Pipe) L.F. 8.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-6 (1) Loader R/T 938 (1) Excavator Cat 345 (1) Roller 815 Soil Compactor Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Drainage Structures Each 2.00 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Ready Mix Truck (0.1) Loader/Backhoe Cat 416 (1) Excavator Cat 324 (1) Roller Ram Max Trench (1) Water Line (up to 4' depth) L.F. 20.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench Roller Ram Max Trench (1) Water Line (over 4' depth) L.F. 10.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Sewer Line (up to 4' depth) L.F. 20.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Sewer Line (over 4' depth) L.F. 10.00 L.F./Hour Truck 1/2 Ton (1) Dozer D-3 (1) Loader R/T 938 (1) Excavator Cat 336 (1) Roller Vibrating Plate Compact (1) Roller Ram Max Trench (1) Water/Sewer Manholes Each 2.00 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Loader/Backhoe Cat 416 (1) Roller Ram Max Trench Exhibit 3-17. Storm drainage/water/sewer summary table.

76 Fuel Usage Factors in highway and Bridge Construction Work Task Unit of Measure Production Rate Production Rate Unit of Measure Fencing (up to 6' height) L.F. 200.00 L.F./Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Loader Skid/Steer (1) Fencing (over 6' height) L.F. 200.00 L.F./Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Loader Skid/Steer (1) Steel Guardrail L.F. 300.00 L.F./Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Truck Guardrail Install (1) Generator Small (1) Wire/Cable Guardrail L.F. 100.00 L.F./Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Truck Guardrail Install (1) Guardrail Posts Each 25.00 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Guardrail Install Truck (1) Solid Sodding S.Y. 500.00 S.Y./Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Loader Skid/Steer (1) Hydro Seeding Acre 3.00 Acre/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Truck Hydroseeder (1) Seedbed Preparation Acre 0.50 Acre/Hour Truck 1/2 Ton (1) Tractor w/Disk (1) Solid Pavement Marking L.M. 2.00 L.M./Hour Truck 1/2 Ton (2) Truck, Thermoplastic Paint (1) Skip Pavement Marking L.M. 2.00 L.M./Hour Truck 1/2 Ton (2) Truck, Thermoplastic Paint (1) Intersection Signalization (2 Lane) Each 0.50 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Crane 12 Ton Truck (1) Intersection Signalization (4 Lane) Each 0.25 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Crane 12 Ton Truck (1) Fence Gates Each 2.00 Each/Hour Truck 1/2 Ton (1) Truck 2 Ton Flatbed (1) Loader Skid/Steer (1) Exhibit 3-18. Specialty items summary table.

Findings and applications 77 Estimated Quantity: 1,000 L.F. Estimated Production Rate: 24 L.F./Hour Estimated Crew Time: 41.67 Hours (1,000 L.F./24 L.F./Hour) Equipment Fuel: Truck ½ Ton 41.67 x 2.0 Gallons/Hour = 83.34 Dozer D-3 41.67 x 2.2 Gallons/Hour = 91.67 Loader R/T 938 41.67 x 4.0 Gallons/Hour = 166.68 Excavator Cat 336 41.67 x 11.0 Gallons/Hour = 458.37 Roller Vibrating Plate 41.67 x 0.2 Gallons/Hour = 8.34 Roller Ram Max Trench 41.67 x 0.5 Gallons/Hour = 20.84 Total Fuel Consumption: 829.24 Gallons Fuel Use Factor: 829.24 Gallons/ 1,000 L.F. = 0.829 Gallons/L.F. Exhibit 3-19. Sample computation for small pipe crew. Task Description Fuel Use Per Unit Units ercA/snollaG 678.821 thgiL - gniraelC ercA/snollaG 954.171 muideM - gniraelC ercA/snollaG 743.372 yvaeH - gniraelC .Y.C/snollaG 793.1 tlahpsA - lavomeR tnemevaP .Y.C/snollaG 265.0 etercnoC - lavomeR tnemevaP .F.L/snollaG 368.0 seziS llA - lavomeR epiP hcaE/snollaG 000.573 )gnidliuB/esuoH( noitilomeD erutcurtS Structure (Bridge per S.F. .F.S/snollaG 626.0 )kceD fo Exhibit 3-20. Clearing and removal fuel use per unit. Task Description Fuel Use Per Unit Units .Y.C/snollaG 023.0 luaH gnoL - daoR ffO - triD - gnidarG .Y.C/snollaG 362.0 luaH trohS - daoR ffO - triD - gnidarG .Y.C/snollaG 786.0 luaH gnoL - daoR nO - triD - gnidarG .Y.C/snollaG 913.0 luaH trohS - daoR nO - triD - gnidarG .Y.C/snollaG 943.0 luaH gnoL - daoR ffO - kcoR - gnidarG .Y.C/snollaG 852.0 luaH trohS - daoR ffO - kcoR - gnidarG .Y.C/snollaG 786.0 luaH gnoL - daoR nO - kcoR - gnidarG .Y.C/snollaG 214.0 luaH trohS - daoR nO - kcoR - gnidarG .Y.S/snollaG 370.0 gnihsiniF debdaoR .Y.C/snollaG 350.0 )luaH oN( )ylnO( gnitsalB dna gnillirD kcoR Exhibit 3-21. Grading fuel use per unit. Task Description Fuel Use Per Unit Units noT/snollaG 604.0 enotS esaB Exhibit 3-22. Base stone fuel use per unit. Task Description Fuel Use Per Unit Units noT/snollaG 298.0 )luaH eliM 5-0( esruoC gnileveL - tlahpsA xiM toH noT/snollaG 779.0 )luaH eliM 51-5( esruoC gnileveL - tlahpsA xiM toH Hot Mix Asphalt - Leveling Course (Over 15 Mile Haul) 1.061 Gallons/Ton noT/snollaG 085.0 )luaH eliM 5-0( esruoC larutcurtS - tlahpsA xiM toH Hot Mix Asphalt - Structural Course (5-15 Mile Haul) 0.718 Gallons/Ton Hot Mix Asphalt - Structural Course (Over 15 Mile Haul) 0.745 Gallons/Ton Hot Mix Asphalt - Surface Course (0-5 noT/snollaG 077.0 )luaH eliM Hot Mix Asphalt - Surface Course (5-1 noT/snollaG 748.0 )luaH eliM 5 Hot Mix Asphalt - Surface Course (Over 15 Mile Haul) 0.994 Gallons/Ton Exhibit 3-23. Asphalt fuel use per unit.

78 Fuel Usage Factors in highway and Bridge Construction Exhibit 3-24 displays the fuel use per unit for the work tasks under the milling category. The 2- to 4-inch milling tasks have higher fuel use than the corresponding 0- to 1-inch tasks, and the difference is exacerbated as hauling distances increase. Exhibit 3-25 displays the fuel use per unit for the structure work tasks. Substructure concrete and superstructure concrete are particularly fuel intensive tasks, requiring 4.70 and 4.15 gallons per cubic yard, respectively. Exhibit 3-26 displays the fuel use per unit for the miscellaneous concrete, concrete paving, and retaining wall work tasks. Heavy fuel usage tasks include concrete median barrier (0.508 gallons of fuel per linear foot), concrete pavement more than 6 inches thick (0.867 gallons per square yard), and retaining wall (0.729 gallons per square foot). Exhibit 3-27 displays the fuel use per unit for the storm drainage, water, and sewer work tasks. Fuel use increases depending on pipe size and sewer and water line depth. Drainage for structures requires the most fuel per unit at 8.725 gallons of fuel per cubic yard. Large pipe requires three times as much fuel per linear foot as medium pipe. Task Description Fuel Use Per Unit Units .Y.S/snollaG 010.0 )luaH eliM 5-0( )"2<( gnilliM .Y.S/snollaG 110.0 )luaH eliM 51-5( )"2<( gnilliM .Y.S/snollaG 410.0 )luaH eliM 51 revO( )"2<( gnilliM .Y.S/snollaG 310.0 )luaH eliM 5-0( )"4-2( gnilliM .Y.S/snollaG 810.0 )luaH eliM 51-5( )"4-2( gnilliM .Y.S/snollaG 520.0 )luaH eliM 51 revO( )"4-2( gnilliM Exhibit 3-24. Milling fuel use per unit. Task Description Fuel Use Per Unit Units .B.L/snollaG 400.0 leetS gnicrofnieR .F.L/snollaG 081.0 smaeB leetS .Y.C/snollaG 007.4 etercnoC erutcurtsbuS .Y.C/snollaG 051.4 etercnoC erutcurtsrepuS Exhibit 3-25. Structures fuel use per unit. Task Description Fuel Use Per Unit Units .F.L/snollaG 805.0 reirraB naideM etercnoC .Y.S/snollaG 056.0 )kcihT 6 =/<( tnemevaP etercnoC .Y.S/snollaG 768.0 )kcihT 6>( tnemevaP etercnoC .F.L/snollaG 251.0 rettuG & bruC .F.S/snollaG 927.0 llaW gniniateR .F.S/snollaG 063.0 klawediS Exhibit 3-26. Miscellaneous concrete/concrete paving/retaining wall fuel use per unit. Task Description Fuel Use Per Unit Units .Y.C/snollaG 527.8 serutcurtS eganiarD .F.L/snollaG 833.4 )epiP "63 >( werC epiP egraL .F.L/snollaG 184.1 )epiP "63 ot "81>( werC epiP muideM .F.L/snollaG 090.2 )htpeD '4 revO( eniL reweS .F.L/snollaG 540.1 )htpeD '4 ot pU( eniL reweS .F.L/snollaG 928.0 )epiP "81 =/<( werC epiP llamS .F.L/snollaG 090.2 )htpeD '4 revO( eniL retaW .F.L/snollaG 540.1 )htpeD '4 ot pU( eniL retaW hcaE/snollaG 000.5 selohnaM reweS/retaW Exhibit 3-27. Storm drainage/water/sewer fuel use per unit.

Findings and applications 79 Exhibit 3-28 displays the fuel use per unit for specialty items that do not belong to the above work categories. Some heavy fuel use work tasks include skip and solid pavement marking at 4.5 gallons per linear meter each. An average four-lane intersection signalization would require 340 gallons of fuel. 3.2.6 Conclusion The professional engineering estimation described in this chapter was one of three method- ologies considered in the effort to tabulate new and updated fuel usage factors. The fuel use calculations, arrived at through a consensus-building process among the expert engineering panel, provide accurate average fuel use specifications for a variety of work tasks prevalent in the highway construction industry. Although any given estimator might choose approaches and equipment that differ slightly from those presented, the fuel use calculations enumerated above represent realistic baseline numbers for a detailed set of average work tasks. 3.3 Statistical Analysis of Fuel Usage The objective of the BidTabs statistical analysis was to estimate the fuel usage of construction activities using a statistical model that incorporates changing fuel prices and bid prices. This included specification of the model, testing of different combinations and forms of the variables, exploration of lagged variables, evaluation of residuals and error terms, and exploration of different combinations of pay items both within and across states. There are six subsections in this chapter section. The first subsection discusses the develop- ment of a database of unit prices for pay items and BidTabs over time. The second sub- section introduces the KLEEM model, which is used to determine the fuel usage of construction activities. The third subsection provides an example of a one-variable application of the KLEEM model. The fourth subsection discusses the development of the variables used in the model estimation. The fifth subsection presents the results of the two-variable KLEEM model. The sixth subsection provides conclusions and next steps. 3.3.1 Development of the Database The goal of this effort is to determine the fuel usage from construction activities in order to develop updated fuel usage factors. For this purpose, the project methodology included a statistical analysis of state highway construction bid data. The first step in the methodology was Task Description Fuel Use Per Unit Units hcaE/snollaG 052.4 setaG ecneF .F.L/snollaG 340.0 )thgieH '6 revO( gnicneF .F.L/snollaG 340.0 )thgieH '6 ot pU( gnicneF ercA/snollaG 794.3 )gnideeS ordyH( gnissarG hcaE/snollaG 240.0 stsoP liardrauG hcaE/snollaG 000.071 )enaL 2( noitazilangiS noitcesretnI hcaE/snollaG 000.043 )enaL 4( noitazilangiS noitcesretnI ercA/snollaG 000.01 noitaraperP debdeeS .M.L/snollaG 005.4 gnikraM tnemevaP pikS .M.L/snollaG 005.4 gnikraM tnemevaP diloS .Y.S/snollaG 710.0 gniddoS diloS .F.L/snollaG 730.0 liardrauG leetS .Y.C/snollaG 761.0 liospoT pirtS .F.L/snollaG 501.0 liardrauG elbaC/eriW Exhibit 3-28. Specialty items fuel use per unit.

80 Fuel Usage Factors in highway and Bridge Construction to tabulate unit prices for pay items over time. This required the development of a database of unit costs. The original methodology envisioned that the database would contain prices over 3 to 5 years. The study team selected a start date of 1/1/2006 and a database containing 5 years of data. Exhibit 3-29 provides a summary of the number of pay items and the number of bids that were in the database for the selected period. In total, 363,137 separate pay items are available in the Oman Systems BidTabs database. For these pay items, there were more than 4.1 million low bids. Note that low bids are the unit price bid for the item in the winning low bid as opposed to the lowest bid for that item. To prepare the database, the study team excluded records that were not suitable for the analysis. The first step was to exclude non-standard pay items. Non-standard pay items are items that do not have the same definition or units from one project/bid to another. Therefore, for these items, there is no price per unit of work. This means that a comparison of unit price across projects or over time it not possible. Similarly, without a price per unit of work, it is not possible to regress unit price on fuel prices in order to assess the existence of a relationship or correlation between the two prices. Exclusion of these non-standard items reduces the number of pay items by roughly half, but only reduces the number of bids by about 8.5 percent. The second step was to exclude lump sum pay items. Lump sum bid items are items for which the bid quantity is essentially equal to one. For example, a lump sum bid item would be building one bridge or paving one section of road. As with non-standard pay items, there is no price per unit of work for lump sum bid items. Therefore it is not possible to compare unit price across projects or over time. Nor is it possible to regress unit price on fuel prices to assess the existence of a relationship or correlation. The exclusion of non-standard items only reduces the number of pay items by about 15,000, but again only reduces the number of bids by about 8 percent. The third step was to exclude pay items that the issuing state DOT did not put out for bid with much frequency. In this case, the analysis excluded pay items if there were fewer than 100 lettings of that item over the 5-year period or fewer than 20 bids per year. The purpose of excluding bid items with very few bids is that the small sample size may hamper the ability to accurately assess the existence of a relationship or correlation. The exclusion of these non-standard items reduces the number of pay items drastically to about 2 percent of the original number of bid items, but only reduces the number of bids to about half of the original number of bid items. Note that there was only an average of eight bids per pay item for the pay items that were excluded from the analysis. The compiled database has approximately 2.1 million records providing data on 6,835 bid items. There are approximately 308 bids per pay item, on average. Exhibit 3-30 provides a sample of 50 records. The database includes state, pay item number, pay item description, unit, quantity, amount (in dollars per unit), a category identifier developed by Oman Systems, and the bid date. Options Number of Pay Items Number of Records (Bids) 808,721,4731,363 ylnO sdiB woL Also Exclude Non-Standard Pay Items 185,846 3,774,921 486,114,3537,071 smetI yaP muS pmuL edulcxE oslA Also Exclude Items Bid Fewer than 100 Times 6,835 2,106,926 Exhibit 3-29. Number of pay items and bid lettings from 1/1/2006 to 12/31/2010 (5 years).

Findings and applications 81 STATE PAY ITEM # PAY ITEM TINUYTITNAUQTINUNOITPIRCSED PRICE CATEGORY BID DATE WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 290.00 205.00 26 (concrete pavement) 10/21/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 144.60 209.00 26 (concrete pavement) 10/21/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 161.00 205.00 26 (concrete pavement) 10/21/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 284.00 300.00 26 (concrete pavement) 10/21/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 453.00 218.00 26 (concrete pavement) 12/02/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 350.00 300.00 26 (concrete pavement) 12/02/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 113.00 270.00 26 (concrete pavement) 12/02/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 200.00 230.00 26 (concrete pavement) 12/02/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 364.00 280.00 26 (concrete pavement) 12/09/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 1707.87 183.95 26 (concrete pavement) 12/09/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 145.00 275.00 26 (concrete pavement) 12/09/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 708.00 175.00 26 (concrete pavement) 12/02/08 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 152.00 260.00 26 (concrete pavement) 02/10/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 126.00 93.00 26 (concrete pavement) 02/24/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 579.00 205.00 26 (concrete pavement) 04/16/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 136.00 225.00 26 (concrete pavement) 04/21/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 174.00 210.00 26 (concrete pavement) 04/21/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 98.00 225.00 26 (concrete pavement) 04/30/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 178.00 200.00 26 (concrete pavement) 06/02/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 163.00 182.00 26 (concrete pavement) 06/25/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 120.00 335.00 26 (concrete pavement) 06/18/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 682.91 162.06 26 (concrete pavement) 06/18/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 132.00 258.28 26 (concrete pavement) 07/07/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 294.00 240.00 26 (concrete pavement) 07/14/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 103.00 250.00 26 (concrete pavement) 07/14/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 400.00 231.00 26 (concrete pavement) 07/23/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 518.00 195.00 26 (concrete pavement) 08/13/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 178.00 190.00 26 (concrete pavement) 08/18/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 143.00 200.00 26 (concrete pavement) 08/18/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 111.00 220.00 26 (concrete pavement) 09/15/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 428.00 205.00 26 (concrete pavement) 09/15/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 237.00 188.00 26 (concrete pavement) 10/27/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 377.00 200.38 26 (concrete pavement) 07/21/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 98.00 210.00 26 (concrete pavement) 07/21/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 251.00 180.40 26 (concrete pavement) 11/17/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 124.00 200.00 26 (concrete pavement) 12/08/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 124.00 200.00 26 (concrete pavement) 12/08/09 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 166.00 275.00 26 (concrete pavement) 01/20/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 187.00 195.00 26 (concrete pavement) 01/20/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 204.00 205.00 26 (concrete pavement) 02/09/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 1125.00 150.00 26 (concrete pavement) 02/09/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 156.00 200.00 26 (concrete pavement) 02/23/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 131.00 234.00 26 (concrete pavement) 06/29/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 118.00 250.00 26 (concrete pavement) 06/29/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 242.00 200.00 26 (concrete pavement) 06/29/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 153.00 210.00 26 (concrete pavement) 06/29/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 142.00 215.00 26 (concrete pavement) 07/13/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 140.00 235.00 26 (concrete pavement) 07/13/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 98.00 225.00 26 (concrete pavement) 07/13/10 WV 502001-012 12 INCH PORTLAND CEMENT CONCRETE APPROACH SLAB S.Y. 208.00 200.00 26 (concrete pavement) 07/27/10 Exhibit 3-30. Fifty sample records from the custom BidTabs data base.

82 Fuel Usage Factors in highway and Bridge Construction 3.3.2 KLEEM Model of Fuel Usage After completing the tabulation of the unit prices for pay items over the 5-year period, the team used the diesel fuel price index from the initial statistical analysis from the first phase as a surrogate for other fuel types. With the individual price items and fuel prices over time, the team proceeded to integrate the KLEEM model into the statistical model. An overview of the modified KLEEM model and the analysis follows. The cost of producing a unit of output is a function of the cost of the inputs required to pro- duce a unit of output. The inputs to the construction industry can generally be classified into five types of inputs: capital (K), labor (L), energy (En), equipment (Eq), and materials and other intermediate inputs (M). The acronym KLEEM is used in reference to the classic KLEM input- output model. The traditional KLEM model includes four factors of production: capital (K), labor (L), energy (E), and materials and other intermediate inputs (M). In the KLEM model, equipment is included in capital costs. The KLEEM model provides separate accounting for equipment (Eq), which may be purchased or leased equipment. The input-output model is primarily used to determine the economic impacts due to a change in final demand. In this application, the input- output model framework is used to estimate the quantity of fuel required to produce a unit of output based on observed prices of the inputs and outputs. The total cost to produce a unit of output in each year is the sum over all inputs of the input price per unit in that year times the quantity of the input required to produce a unit of output. It is assumed that the input quantities required per unit output are constant over time (i.e., change very slowly when compared to the frequency of price changes over the 5-year time period). Then the cost of producing a unit at times t = 1, . . . , T is c t p t qi i i I( ) = ( ) = ∑ . 1 Here c(t) represents the cost of producing a unit of output at time t and the pi(t) terms represent the price of input i = 1, . . . , I. In this application, I = 5 for the inputs K, L, En, Eq, and M. The qi terms represent quantity of input i required per unit output. In matrix form, let the column vectors C_ and Q_ and the matrix P be defined as C c c c T p p = ( ) ( ) ( )       = ( ) ( )1 2 1 11 2   ,P p p p p p T p T p T 1 1 2 1 1 2 1 1 2 2 2 ( ) ( ) ( ) ( ) ( ) ( ) ( )            =       , .andQ q q q 1 2 1  Then the matrix equation C_ = PQ_ is a linear regression model that expresses the cost per unit output at each time as a linear function of the unknown input quantity vector Q_ . If T ≥ I and no two columns of P are collinear, the least squares estimate of the input quantity vector Q_ is Q_ˆ = [P′P]-1 P′C_ where P′ denotes the transpose of the matrix P. 3.3.3 One-Variable Model of Fuel Usage The KLEEM model defined above includes five classes of inputs. The regression model includes five coefficients, one representing each class of input. However, not all classes of input are especially variable over the 5-year period. For example, capital, labor, and equipment costs vary slowly, for the most part, over such a short period. Fuel price, on the other hand, is highly volatile. Much of the variation in the cost of a unit of output may be due to the change in the price of fuel alone.

Findings and applications 83 In this way, the model can be simplified to one variable: fuel. In the one-variable model, the quantities and prices of all other inputs are considered fixed. Consider the following example. A unit of output costs $10 to complete. The price of fuel at that time was $2.50 per gallon. In another period, the same unit of output costs $12.50 to complete. In this second period, the price of fuel was $5 per gallon. In a one-variable framework, fuel price is the only input that varies. It is assumed that the price of all other inputs remained unchanged and the quantity of fuel required to produce the unit of output was constant. The unknown in this example is the quantity of fuel required to produce the unit of output. Recall that the quantity of fuel that is required to produce one unit of output remains constant, regardless of the price of fuel. In this example, it can be determined that one gallon of fuel is required to produce the unit of output and the cost of all other inputs is $7.50. In this simple example, where the bid price and the price of fuel are known, and the prices of all other inputs remain fixed, the amount of fuel required per unit can be calculated. Exhibit 3-31 presents a hypothetical example of the KLEEM model with one variable input: the price of fuel. The graph shows a scatter plot of the change in unit cost ($/Unit) versus the change in fuel price ($/Gallon) for a series of observations of unit output prices and fuel input prices. The slope of the regression line is an estimate of the quantity of fuel required to produce a unit of output. The slope has the dimensions of ($/Unit)/($/Gallon) = (Gallon/Unit). 3.3.4 Development of Independent Variables The original KLEEM model utilizes five variables in determining fuel use per unit of output. Running the five-variable model with the available data, however, revealed a high degree of cross- correlations and unexpected signs that did not fit with the assumptions of the model variables. The two-variable model, which uses fuel and labor costs, gave reasonable results and was selected for this regression. The energy prices for the model are monthly averages of the U.S. average daily Low-Sulfur No. 2 diesel fuel prices ($/Gallon) from the U.S. Department of Energy. The construction employment Exhibit 3-31. Trend line estimate of fuel usage for one-variable model (Slope = ($/Unit)/($/Gallon) = Gallon/Unit).

84 Fuel Usage Factors in Highway and Bridge Construction cost index is provided in Exhibit 3-32. The labor prices for the construction industry were calculated as a 50-day moving average of the quarterly construction wages price series (Decem- ber 2005 = 100) obtained from the Bureau of Labor Statistics. The 50-day moving average is a standard technical indicator used in the analysis of day-to-day price movements of commodities and stocks. The moving average serves to smooth the data, reducing the effect of shocks, in this case fuel price shocks, and identifies the underlying trends in the data. Exhibit 3-33 shows the time series plot of the average daily U.S. No. 2 Diesel Fuel price and its 50-day moving average from November 2004. 3.3.5 Analysis of the Two-Variable KLEEM Model The analysis proceeded in several stages, with a significant degree of data aggregation required at each stage to generate useful results. The data were first screened to eliminate any bids from Exhibit 3-33. U.S. distillate fuel price with 50-day moving average. Exhibit 3-32. U.S. construction employment cost index.

Findings and applications 85 pay items with a range of less than 3 years of data. This step was required to ensure that the prices reflect a wide range of the changes in input prices that occurred over the 5-year time period covered by this study. Approximately 600,000 bids were eliminated in this stage. In the second stage, the pay items in each state were aggregated to pay item groups within each state. The groups were based on the categories shown in Exhibit 3-34. A consistent definition of the unit of production is required for the KLEEM model. However, each category contained bids for pay items expressed in various units. An example of the pay items included in each category/unit grouping for bridges in Ohio is presented in Exhibit 3-35. The group of pay items for bridges with units of cubic yards (C.Y.) includes mainly concrete shapes, while the group for bridges in pounds (Lbs.) includes only bridge steel items and the square yard (S.Y.) category includes mainly organic surface coating. The fourth grouping is linear feet (L.F.) and shows more diversity in the listed pay items, indi- cating that the category/unit grouping within a state does not always result in a homogeneous group of pay items. Some lack of homogeneity within the groupings is an inevitable result of the aggregation process. Exhibit 3-36 defines the work categories and items and presents the sample sizes for each. These items were obtained by selecting the most common units of measure for each pay item category in Exhibit 3-34. Separate tables were prepared for each state with the average quantities and weighted-average unit prices by month of pay items within each selected group shown in bold in Exhibit 3-36. The pay items selected for the state tables were screened to eliminate many bid records with very small or unusually large values of quantity or price from the calculation of the state means. The number of retained records in the second stage generally exceeded 100 per month within each grouping in each state. Category Description 1 GRADING/EXCAVATION 2 BRIDGE 3 ASPHALT 4 BASE STONE 5 DRAINAGE-PIPE 6 DRAINAGE-INLETS/CATCH BASINS 7 CONCRETE-CULVERTS 8 CONCRETE-MISC 9 TRAFFIC CONTROL 10 GUARDRAIL 11 FENCING 12 GRASSING 13 CLEARING 14 EROSION CONTROL 15 RETAINING WALL 16 SIGNALIZATION 17 SIGNS-PERMANENT 18 STRIPING/PAVEMENT MARKING 19 PAINTING STRUCTURES 20 UTILITY-WATER 21 UTILITY-GAS 22 UTILITY-SEWER 23 LIGHTING 24 BUILDINGS/MISC STRUCTURES 25 MOBILIZATION 26 CONCRETE PAVEMENT 27 MISC STONE/RIPRAP 28 ROADWAY LIGHTING/ELECTRICAL 29 UNDERDRAIN 30 EQUIPMENT/LABOR 31 ALTERNATES/BONUS/TIME Exhibit 3-34. Pay item categories.

86 Fuel Usage Factors in Highway and Bridge Construction Category Unit Item Bridge C.Y. 'CLASS C CONCRETE 'CLASS C CONCRETE RETAINING WALL/WINGWALL ABOVE FOOTING' 'CLASS HP CONCRETE 'CLASS HP CONC 'CONCRETE 'POLYMER MODIFIED ASPH EXP JNT SYSTEM' 'QC/QA CONCRETE L.F. 'STEEL PILES HP10X42 'STEEL PILES HP12X53 '12" CIP REINFORCED CONCRETE '12" CIP REINFORCED CONCRETE PILES 'CONC RPR BY EPOXY INJ' 'STRUC EXP JT INCL ELAST STRIP' 'STRUC CXP JT INC ELAST 'SEMI-INTEGRAL ABUT EXP JOINT SEAL 'JOINT SEALER 'SAWING & SEALING BIT CONC JTS' 'POLYMER MOD ASPH EXP JT SYSTEM' 'RAILING (TWIN STEEL TUBE)' 'STEEL DRIP STRIP' LBS 'EPOXY COATED REINFORCING STEEL 'REINFORCING STEEL 'STRC STEEL MEM S.Y. 'SLG OF CONC SURF (NON-EPOXY)' 'SLG OF CONC SURF (EPOXY-URETHANE) 'SEALING CONC BRIDGE DECKS W/HMWM RESIN' 'TREATING OF CONCRETE BRIDGE DECK W/SRS' 'TYPE 2 WATERPROOFING' 'TYPE 3 WATERPROOFING' 'TRTING CONC DECKS W/GRAVITY FED RESIN' 'PTCHNG CONC DECK-TYPE B' 'PTCHNG CONC DECK TYPE C' 'APPROACH SLABS (T=15") 'TIED CONCRETE BLOCK MAT 'QC/QA CONCRETE Exhibit 3-35. Pay item grouping for bridge category in Ohio. Units EACH L.F. S.Y. C.Y. TON S.F. LBS TON-G GAL ACRE MILE L.M. ASPHALT X X 45,659 X 76,888 X X 26,044 23,039 - X X BASE STONE X X X 13,055 15,554 - - X - - X - BRIDGE X 23,330 20,107 22,754 X X 19,638 - X - - - CLEARING X 30,168 20,607 X X X - - - X - X CONCRETE PAVEMENT X 6,107 11,551 X X - X X X - X X CONCRETE-CULVERTS X 1,098 X X - X X - - - - - CONCRETE-MISC X 20,749 18,218 11,765 X X X - - - - - DRAINAGE-INLETS\ CATCH BASINS 55,242 5,471 X X - X X - - - - - DRAINAGE-PIPE X 38,769 X X - - - - - - - - EROSION CONTROL X 42,561 23,347 X X X X X - X - - FENCING X 5,112 X - - - - - - - - - GRADING\EXCAVATION X X 10,582 87,727 X X - X - X X X GRASSING X X 23,678 X X X 22,649 X X 22,904 - X 219,83 620,85 LIARDRAUG X - - - - - - - - - MISC STONE\RIPRAP X X X 16,257 X - - X - - - - MOBILIZATION 19,168 1,633 - X X - - - - X - X PAINTING STRUCTURES - X X X - 2,642 - - - - - - RETAINING WALL - X - 758 - X X - - - - - ROADWAY LIGHTING\ 120,32 120,71 LACIRTCELE - X - X - X - - - - SIGNALIZATION 38,371 27,873 - X - - - - - - - - SIGNS-PERMANENT 36,762 X X X - 29,892 X - - - - - STRIPING PAVEMENT MARKING 85,065 152,255 X - - X - - X - 9,777 10,679 TRAFFIC CONTROL 91,788 41,377 X X X 19,716 - X - - X X 641,21 170,4 NIARDREDNU X X - - - - - - - - UTILITY-SEWER 619 514 - - - - - - - - - - UTILITY-WATER 4,654 497 - - - - X - - - - - X Positive number of records, generally smaller than those shown for each pay item, but not selected for analysis. - No bid records available. Exhibit 3-36. Number of records in pay item groups selected for analysis.

Findings and applications 87 In the third stage, two-variable KLEEM model regressions were estimated for each pay item grouping in each state using the change in weighted-average price per unit as a function of the changes in the mean fuel price and labor cost. In this stage, regressions were estimated for approximately 330 state/pay item groupings. Each regression included at most 60 monthly data points, although fewer than 60 months of data were available in most states. The changes in the independent and dependent variables were calculated by subtracting the minimum value of the mean for each variable over the 60-month period from the mean value in each month. This transformation defines a multivariate origin for the regression where zero is equal to the minimum value of each variable. The origin transformation does not affect the scale or units of the data. In the final stage of aggregation, the regression coefficients for fuel and labor for each pay item group were averaged across all states. Only states with regression coefficients with plausible significance (p < 0.50) were retained for the calculation of the mean fuel coefficient in each pay item group. 3.3.6 Results of the Two-Variable KLEEM Model The two-variable KLEEM model assumes fixed prices and quantities for capital, materials, and equipment, while labor and fuel prices are permitted to vary. The mean fuel coefficient estimated for each pay item group is shown in Exhibit 3-37. Since the cost of labor is expressed as an index, the coefficients estimated for this input are not meaningful. The regression results in Exhibit 3-37 show some degree of consistency. The fuel required for a ton of asphalt is (by a factor of approximately 10) higher than for a ton of base stone. The fuel required for asphalt per square yard is slightly smaller than the fuel required for the pay item grouping of bridges per square yard (mainly organic surface coatings) as described above. Drainage pipe has a higher fuel requirement per linear foot than fencing, which in turn has a higher requirement than erosion control. Guardrails require only slightly more fuel input per linear foot than roadway lighting/electrical. On the other hand, however, several of the estimates generated by this analysis clearly do not appear to represent actual fuel usage. For example, the statistical estimate of fuel usage for grading on a cubic yard basis differs from the engineering and contractor survey estimates by a large factor of magnitude. Despite the high level of aggregation, the number of states is small for many pay item groups. For example, in Exhibit 3-37, seven pay items rely on data for only one state, while 15 more rely on data from only two or three states. State-defined pay items are at a finer level of detail than the pay item groups shown in Exhibit 3-37. As a result, the estimates for these groupings are not as robust as those for pay items that are common to many states. 3.3.7 Conclusion The statistical analysis of bid data described in this section was one of three methodologies under consideration in the effort to tabulate new and updated fuel usage factors. The statistical analysis demonstrated that most highway construction activities consume large amounts of fuel and are fuel intensive. However, the approach does not appear to have generated estimates of fuel usage that would be accurate enough to contribute to the development of the final fuel usage factors. However, in developing these fuel factors, the results of the statistical analysis were considered where it was felt that they might be useful.

88 Fuel Usage Factors in highway and Bridge Construction Pay Item Group [Unit] Estimated Gallons/Unit Number of States 6 83.0 ]laG[ tlahpsA 6 16.5 ].Y.S[ tlahpsA 01 23.12 ]noT[ tlahpsA 6 15.91 ]G-noT[ tlahpsA 1 77.1 ].Y.C[ enotS esaB 2 65.2 ]noT[ enotS esaB 3 59.86 ].Y.C[ egdirB 3 58.51 ].F.L[ egdirB 3 43.0 ].sbL[ egdirB 3 86.7 ].Y.S[ egdirB 5 82.22 ].F.L[ gniraelC 3 89.861 ].Y.S[ gniraelC 1 83.62 ].Y.C[ .csiM-etercnoC 2 91.2 ].F.L[ .csiM-etercnoC 3 44.41 ].Y.S[ .csiM-etercnoC 1 35.1 ].F.L[ tnemevaP etercnoC 2 33.31 ].Y.S[ tnemevaP etercnoC 01 65.823 ]hcaE[ snisaB hctaC/stelnI-eganiarD 1 45.67 ].F.L[ snisaB hctaC/stelnI-eganiarD 7 15.11 ].F.L[ epiP-eganiarD 9 54.1 ].F.L[ lortnoC noisorE 5 88.39 ].Y.S[ lortnoC noisorE 1 59.5 ].F.L[ gnicneF 61 76.763 ].Y.C[ noitavacxE/gnidarG 1 90.0 ].Y.S[ noitavacxE/gnidarG 6 85.291 ]ercA[ gnissarG 2 94.4 ].sbL[ gnissarG 3 07.0 ].Y.S[ gnissarG 21 63.952 ]hcaE[ liardrauG 6 56.2 ].F.L[ liardrauG 3 42.24 ].Y.C[ parpiR/enotS .csiM 1 71.4 ].F.S[ serutcurtS gnitniaP 4 29.091 ]hcaE[ lacirtcelE/gnithgiL yawdaoR 4 71.2 ].F.L[ lacirtcelE/gnithgiL yawdaoR 7 59.217 ]hcaE[ noitazilangiS 7 04.4 ].F.L[ noitazilangiS 8 82.051 ]hcaE[ tnenamreP-sngiS 5 44.2 ].F.S[ tnenamreP-sngiS Striping/Pavement Mark 41 45.41 ]hcaE[ gni Striping/Pavement Mark 12 42.0 ].F.L[ gni 3 38.97 ].M.L[ gnikraM tnemevaP/gnipirtS 3 68.014 ]eliM[ gnikraM tnemevaP/gnipirtS 51 70.143 ]hcaE[ lortnoC ciffarT 4 39.1 ].F.L[ lortnoC ciffarT 5 67.0 ].F.S[ lortnoC ciffarT 2 04.5 ].F.L[ niardrednU Exhibit 3-37. Regression-based fuel use coefficients (gallons per unit output).