Best practices for budget-based design [Mar. 2017]

GEORGIA DOT RESEARCH PROJECT NO. 14-09
FINAL REPORT
BEST PRACTICES FOR BUDGET-BASED DESIGN
OFFICE OF RESEARCH 15 KENNEDY DRIVE
FOREST PARK, GA 30297-2534

1. FHWA-GA-17-1409

2. Government Accession No.: 3. Recipient's Catalog No.:

4. Title and Subtitle:

5. Report Date: March 2017

Best Practices for Budget-Based Design

6. Performing Organization Code:

7. Author(s): Baabak Ashuri, Ph. D., DBIA, CCP, DRMP Minsoo Baek Yan Li 9. Performing Organization Name and Address: Economics of Sustainable Built Environment (ESBE) Lab Georgia Institute of Technology 280 Ferst Drive, Atlanta, GA 30332-0680
12. Sponsoring Agency Name and Address: Georgia Department of Transportation, Office of Research 15 Kennedy Drive, Forest Park, Georgia 30297-2599

8. Performing Organ. Report No.:
10. Work Unit No.: 11. Contract or Grant No.:
0013167
13. Type of Report and Period Covered: Final; May 2014 March 2017 14. Sponsoring Agency Code:

15. Supplementary Notes:

16. Abstract:

State Departments of Transportation (State DOTs) encounter difficulties in establishing feasible and

reliable project budget early in the project development. The lack of a systematic process for establishing

baseline budget with the consideration of potential issues (risks) that negatively impact project cost

throughout project development presents a major challenge for State DOTs. The overarching objective of

this research project is to develop a set of cost estimation and management practices for budget-based

design that can aid GDOT project managers and engineers throughout the plan development process

(PDP). To achieve the research objective, this report conducted three major tasks: (1) Reviewing state of

practice of cost estimation process in other State DOTs; (2) Reviewing state of practice for fixed budget-

best value procurement method in other State DOTs; and (3) Conducting statistical analysis to identify

important variables capable of explaining variations in submitted bid prices for highway projects in the

State of Georgia. Cost estimation and control processes in Minnesota, California, Texas, Ohio, and

Washington State DOTs are provided as examples of best practices in establishing reliable baseline cost

estimates. Procurement process in Utah, Colorado, and Michigan DOTs are presented as examples of

successful utilization of fixed budget-best value procurement method. The results of multivariate

regression analysis show that 12 variables, including quantity of the bid item, housing market index,

Georgia asphalt cement price index, total bid price, project length, 12-month percent change of Georgia

asphalt cement price index, 12-month percent change of Gross Domestic Product (GDP) of the Georgia

construction industry, unemployment, total asphalt volume of resurfacing and widening projects, number

of bidders, project duration, and number of nearby asphalt plants have explanatory power to explain

variations in submitted bid prices for major asphalt line items in the State of Georgia's highway projects.

It is also found out that 5 variables, including unemployment, 12-month percent change of Georgia asphalt

cement price index, quantity of the bid item, total bid price, and Turner construction cost index have power

to explain variations in submitted bid prices for projects in the Transportation Investment Act (TIA)

regions.

17. Key Words: Budget-based Design, Risk Management, Regression analysis

18. Distribution Statement:

19.Security Classification
(of this report): Unclassified

20. Security Classification (of this page): Unclassified

21. Number of Pages: 146

22. Price:

1

GDOT Research Project No. 14-09
Final Report
BEST PRACTICES FOR BUDGET-BASED DESIGN
Prepared by: Baabak Ashuri, Ph.D., DBIA, CCP, DRMP
Minsoo Baek Yan Li
Georgia Institute of Technology
Contract with Georgia Department of Transportation
In cooperation with U.S. Department of Transportation Federal Highway Administration
March 2017
The contents of this report reflect the views of the author(s) who is (are) responsible for the facts and the accuracy of the data presented herein. The contents do not necessarily reflect the official views or policies of the Georgia Department of Transportation or the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.
2

TABLE OF CONTENTS
ACKNOWLEDGMENTS ................................................................................................ 8 EXECUTIVE SUMMARY .............................................................................................. 9 CHAPTER 1 INTRODUCTION ................................................................................... 14 CHAPTER 2 DEVELOPING AND CONTROLLING BASELINE COST ESTIMATE DURING PROJECT DEVELOPMENT PROCESS............................. 17 2.1. Introduction .......................................................................................................................... 17 2.2. State of Practice of Cost Estimation Process in Minnesota, California, Texas, Ohio, and Washington State Departments of Transportation .................................................................. 23 2.2.1. Minnesota Department of Transportation (MnDOT) ...........................................................23 2.2.2. California Department of Transportation (Caltrans) ............................................................30 2.2.3. Texas Department of Transportation (TxDOT)....................................................................38 2.2.4. Ohio Department of Transportation (ODOT).......................................................................50 2.2.5. Washington State Department of Transportation (WSDOT) ...............................................57 2.3. Summary of the Recommended Best Practices for Cost Estimation and Control ......... 66 CHAPTER 3 FIXED BUDGET-BEST VALUE PROCUREMENT METHOD ...... 72 3.1. Introduction .......................................................................................................................... 72 3.2. State of Practice for Fixed Budget-Best Value Procurement Method in Utah, Colorado, Idaho, and Michigan Departments of Transportation............................................................. 74 3.2.1. Utah Department of Transportation (UDOT) .......................................................................74 3.2.2. Colorado Department of Transportation (CDOT) ................................................................79 3.2.3. Idaho Transportation Department (ITD) ..............................................................................83 3.2.4. Michigan Department of Transportation (MDOT)...............................................................85 3.4. Summary of the Recommended Best Practices for Fixed Budget-Best Value Procurement Method.......................................................................................................................................... 89 CHAPTER 4 STATISTICAL ANLYSIS FOR EXPLAINING VARIATIONS IN SUBMITTED BID PRICES FOR ASPHALT LINE ITEM IN HIGHWAY PROJECTS...................................................................................................................... 93 4.1. Introduction .......................................................................................................................... 93 4.2. Methodology.......................................................................................................................... 97 4.2.1. Modeling the Variations of the Submitted Bid Prices ..........................................................98 4.3. Dataset Development.......................................................................................................... 102 4.3.1. Data Compilation Process ..................................................................................................102
3

4.3.2. Submitted Unit Prices for Asphalt Line Items....................................................................103 4.3.3. Factors affecting Variation in the Submitted Bid Prices ....................................................104 4.4. Results of the Multiple Regression Modeling................................................................... 120 4.5. Results of the Multiple Regression Modeling for Projects in TIA Regions................... 129 CHAPTER 5 CONCLUSIONS.................................................................................... 138 REFERENCES.............................................................................................................. 142
4

LIST OF FIGURES
Figure 2-1 Ability to Influence Changes in Cost of a Project .................................................. 18 Figure 2-2 MnDOT Cost Estimate Review and Approval Gates ............................................ 26 Figure 2-3 MnDOT Cost Estimation and Cost Management Process .................................... 27 Figure 2-4 Template of Variance Reports................................................................................. 30 Figure 2-5 Template for QA/QC Documentation ..................................................................... 37 Figure 2-6 Sample of TxDOT's APRA Tool ............................................................................. 44 Figure 2-7 APRA Application Points......................................................................................... 45 Figure 2-8 A Sample Hydraulic Element in the Design Summary Report ............................ 49 Figure 2-9 WSDOT Cost Estimating Process for Each Phase of PDP.................................... 59 Figure 2-10 Timing of CRA/CEVP Workshops........................................................................ 63 Figure 4-1 Data Compilation Process for Multiple Regression Analysis ............................. 103 Figure 4-2 Georgia Terrain Map ............................................................................................. 109 Figure 4-3 Georgia District Map.............................................................................................. 110 Figure 4-4 Locations of Active Asphalt Plants in Georgia in 2016 ....................................... 112 Figure 4-5 Histogram of Standardized Residuals for the Regression Model ....................... 127 Figure 4-6 Normal P-P Plot of Standardized Residuals for the Regression Model ............. 128 Figure 4-7 Scatterplot of Standardized Residuals vs. Predicted Values............................... 129 Figure 4-8 Three Regional Commissions for TIA .................................................................. 130 Figure 4-9 Histogram of Standardized Residuals for the Regression Model Developed Based on Projects in TIA Regions ....................................................................................................... 135 Figure 4-10 Normal P-P Plot of Standardized Residuals for the Regression Model Developed Based on Projects in TIA Regions ......................................................................... 136 Figure 4-11 Scatterplot of Standardized Residuals vs. Predicted Values for the Regression Model Developed Based on Projects in TIA Regions ............................................................. 137
5

LIST OF TABLES
Table 2-1 MnDOT Cost Estimation Milestones........................................................................ 24 Table 2-2 MnDOT Key Inputs for Cost Estimates ................................................................... 25 Table 2-3 Caltrans Cost Estimation Milestones........................................................................ 32 Table 2-4 Caltrans Key Inputs for Cost Estimates................................................................... 33 Table 2-5 Roles and Responsibilities for QA/QC Process........................................................ 35 Table 2-6 Critical Factors for QA/QC process ......................................................................... 36 Table 2-7 TxDOT Cost Estimation Milestones ......................................................................... 39 Table 2-8 TxDOT Key Inputs for Cost Estimates .................................................................... 40 Table 2-9 APRA Sections, Categories, and Elements............................................................... 42 Table 2-10 APRA Review Gates and Purposes......................................................................... 46 Table 2-11 ODOT Five Project Paths ........................................................................................ 51 Table 2-12 ODOT Cost Estimation Milestones......................................................................... 52 Table 2-13 ODOT Key Inputs for Cost Estimates.................................................................... 53 Table 2-14 ODOT's Alternative Selection Methods ................................................................. 54 Table 2-15 Components of ODOT's Alternative Evaluation Report...................................... 56 Table 2-16 WSDOT Cost Estimation Milestones...................................................................... 58 Table 2-17 WSDOT Key Inputs for Cost Estimates................................................................. 61 Table 2-18 Roles and Responsibilities of CRA and CEVP Workshop Team......................... 64 Table 3-1 Example of Evaluation Factors and Category ......................................................... 77 Table 3-2 Relating Project Goals and Values to Best Value Scoring Parameters ................. 80 Table 3-3 Adjectival Evaluation and Scoring Guide ................................................................ 81 Table 3-4 CDOT Design-Build Alternative Algorithms to Determine Total Evaluation Score ....................................................................................................................................................... 82 Table 3-5 Example of Fixed Budget-Best Value Projects in State of Idaho ........................... 85 Table 4-1 Descriptive Statistics of Submitted Unit Prices for Hot Mix Recycle Asphaltic Concrete ..................................................................................................................................... 104 Table 4-2 Summary of Potential Explanatory Variables ....................................................... 105 Table 4-3 Descriptive Statistics of the Submitted Bid Prices based on Georgia Terrain Types ..................................................................................................................................................... 108 Table 4-4 Descriptive Statistics of the Submitted Bid Prices based on Seven Districts ...... 110 Table 4-5 Annual Number of Asphalt Plants in Georgia....................................................... 111 Table 4-6 Results of Outlier Inspection ................................................................................... 120 Table 4-7 Coefficients of the Final Regression Model ............................................................ 123 Table 4-8 Model Summary of the Final Regression Model ................................................... 126 Table 4-9 ANOVA of the Final Regression Model ................................................................. 126
6

Table 4-10 Coefficients of the Final Regression Model Developed Based on Projects in TIA Regions ....................................................................................................................................... 132 Table 4-11 Model Summary of the Final Regression Model Developed Based on Projects in TIA Regions ............................................................................................................................... 134 Table 4-12 ANOVA of the Final Regression Model Developed Based on Projects in TIA Regions ....................................................................................................................................... 134
7

ACKNOWLEDGMENTS
The research reported herein was sponsored by the Georgia Department of Transportation through Research Project Number 14-09. The authors acknowledge and appreciate the help from Mr. Kelvin Mullins, the Administrator of the Georgia Department of Transportation's Transportation Investment Act (TIA) program, and Mr. Tim Matthews, Senior DesignBuild Project Manager. Their invaluable assistance and insight were critical to the completion of this research. Mr. Shrujal Amin, TIA Program Manager, and Mr. Eduardo Gamez from AECOM shared their knowledge and experience on the application of cost estimation and risk management practices throughout the course of this project. The authors would also like to thank Ms. Supriya Kamatkar, Ms. Lisa Myer, Mr. Troy Patterson, and Dr. Peter Wu at the Georgia Department of Transportation who have helped us throughout the progress of this research project. Especially, Ms. Kamatkar's oversight and administrative support has been critical in the success of this research project.
8

EXECUTIVE SUMMARY
State Departments of Transportation (State DOTs) encounter difficulties in establishing feasible and reliable project budget early in the project development. The major challenge that several State DOTs have faced is the lack of a systematic process for establishing baseline budget with the consideration of potential issues (risks) that negatively impact project cost throughout project development. As design advances through various phases, there are several factors affecting the cost estimate, such as right-of-way acquisition, utilities coordination, and scope changes. If these risk factors go unnoticed early in the project development because of limited information related to scope, project site, and design, they can impose a considerable burden on State DOTs and cause cost escalation during project (plan) development process. Considering the new focus by GDOT and other State DOTs on enhanced cost estimation and management process, there is a significant need to make cost estimation a priority and set baseline budget estimate during the planning and concept development phases. This research project will focus on major components of a reliable baseline budget model and their information requirements.
The overarching objective of this research project is to develop a set of cost estimation and management practices for budget-based design that can aid GDOT project managers and engineers throughout the plan development process (PDP). To achieve the research objective, current state of practices in cost estimation and control in several State DOTs
9

are reviewed and best practices are identified in cost estimation and strategies for cost control. Also, current state of practices in fixed budget-best value procurement method are reviewed and best practices are identified in utilization of this innovative contracting method. Finally, statistical analysis is conducted to investigate the impact of macroeconomic, construction market, and oil market conditions on highway construction costs by analyzing submitted bid prices for major asphalt line items in the State of Georgia's highway projects.
The following recommendations are found out to be effective for enhancing the practice of defining and maintaining the established budget for highway projects:
State DOTs should establish an integrated process for cost estimation and cost management to establish accurate, reliable, and consistent estimates thorough project development process.
State DOTs should establish key milestones for estimating, updating, and approving cost estimates as project definition/design advances.
State DOTs should capture any changes in estimating assumptions to track the basis of cost estimate and control estimated project cost.
State DOTs are recommended to establish an automated information system to help them maintain, update, and share project information, cost estimates, and changes in project scope, cost, and schedule.
10

State DOTs should consider potential issues (risks) that may cause cost escalation during developing baseline cost estimates. Risk analysis tools and inputs from key project stakeholders are necessary for identifying critical risk factors for the project.
Finally, State DOTs should utilize a quality assurance/quality control (QA/QC) process to verify the final engineer's cost estimate before a project is advertised.
The following recommendations are found out to be effective for enhancing the practice of delivering projects using this innovative contracting strategy:
State DOTs may consider a fixed budget-best value procurement approach when the full project scope exceeds the baseline cost estimate for the project. For a fixed budget-best value procurement approach, the agency should define the basic configuration scope and should allow the proposers to include the maximum amount of work or additional scope elements in their proposals while staying within the fixed budget.
A fixed budget-best value approach can be utilized in several project types, such as corridor expansion, bridge deck preservation, and seal coating projects. State DOTs should clearly define additional scope elements beyond the base scope for each project type.
State DOTs should clearly establish the evaluation criteria to select the proposers for a project. Since the price is fixed for all proposers and this approach allows higher flexibility in proposing design and construction solutions, the agency should 11

establish rigorous evaluation criteria (e.g., cost, time, and design alternatives) and the weights for the criteria to evaluate the proposals based on the project goals.
Lastly, statistical analysis is conducted to identify important variables capable of explaining variations in submitted bid prices for major asphalt line items in the GDOT's highway projects. Multiple regression analysis is utilized to examine the impact of project characteristics, macroeconomic variables, construction market condition indicators, and oil market parameters on highway construction costs. The main purpose of this research is to examine the effects of several factors representing construction market, economic, and oil market conditions on submitted bid prices. The goal is to develop a regression model with explanatory power to describe variations in submitted bid prices. It is worth noting that the developed regression model can be used for forecasting bid prices for asphalt line items but prediction was not the main objective of this study. Therefore, the results should be used with caution as the forecasting error might be significantly large. An explanatory model is developed for the State of Georgia's highway projects using multiple regression analysis. Several important variables are identified to have power to explain variations in submitted bid prices for major asphalt line items. The identified variables, in descending order of importance, are: quantity of the bid item, housing market index, Georgia asphalt cement price index, total bid price, project length, 12-month percent change of Georgia asphalt cement price index, 12-month percent change of Gross Domestic Product (GDP) of the Georgia construction industry, unemployment, total asphalt volume
12

of resurfacing and widening projects, number of bidders, project duration, and number of nearby asphalt plants. Among these significant explanatory variables, quantity of the bid item, project length, unemployment, number of bidders, project duration, and number of asphalt plants have negative relationship with submitted bid prices while holding all other variables in the model constant. All other variables have positive influence on submitted bid prices. Multiple regression analysis is repeated for identifying significant factors that affect variations in submitted bid prices in the regions included in the Transportation Investment Act (TIA). The identified important variables, in descending order of importance, are: unemployment, 12-month percent change of Georgia asphalt cement price index, quantity of the bid item, total bid price, and Turner construction cost index. Among those significant factors in the explanatory model developed for projects in the TIA regions, quantity of the bid item has negative relationship with submitted bid prices while holding all other variables constant. All other variables have positive relationship with submitted bid price.
13

CHAPTER 1 INTRODUCTION
The Transportation Investment Act (TIA) Referendum was passed in 2012 by Georgia voters in the regions of Central Savannah River Area, Heart of Georgia Altamaha, and River Valley. These three regions will implement a one percent regional sales tax over a ten-year period to fund transportation improvements. The Office of TIA in Georgia Department of Transportation (GDOT) is responsible for the management of the budget, schedule, execution and delivery of all Projects contained in the Approved Investment Lists. The unique nature of funding for these projects presents significant challenges for the Office of TIA to develop the listed projects according to pre-determined budgets. In other words, the Office of TIA needs to work with the established budgets and schedules for these projects and ensure that these established budgets and schedules remain unchanged throughout the course of project design development. Hence, the Office of TIA needs to develop and utilize a systematic approach for establishing baseline budget and schedule models for its program. The application of this systematic approach is not just limited to the Office of TIA. The budget-based design approach can be considered in the delivery of GDOT's other projects since it can enhance the agency's ability to deliver transportation projects according to their original estimated budgets and schedules. This issue is becoming more and more important with the ever-increasing funding pressure on the agency.
14

State Departments of Transportation (State DOTs) encounter difficulties in establishing feasible and reliable project budget early in the project development. The major challenge that State DOTs have is the lack of a systematic process that explains required steps and key inputs for establishing baseline budget throughout project development. In addition, as design advances through various phases, there are several factors affecting the cost estimate, such as right-of-way acquisition, utilities coordination, and scope changes. If these risk factors go unnoticed early in the project development because of limited information related to scope, project site, and design, they can impose a considerable burden on State DOT and cause cost escalation during project development process. Thus, State DOTs should have a risk-based approach to identify potential risks and establish strategies to minimize their risks on project cost. Considering the new focus by GDOT and other State DOTs on enhanced cost estimation and management process, there is a significant need to make cost estimation a priority and set baseline budget estimate during the planning and concept development phases. This research project will focus on major components of a reliable baseline budget model and their information requirements.
The overarching objective of this research project is to develop a set of cost estimation and management practices for budget-based design that can aid GDOT project managers and engineers throughout the plan development process (PDP). To achieve the research objectives, the following tasks have been done, and the report is structured as follows:
Chapter 2- Review state of practice of cost estimation process in other State DOTs 15

In the first step, the academic/professional literature is reviewed to identify the state of knowledge about establishing baseline budget, developing a project cost management system, and describing a process for project cost control. Current practices of selected State DOTs are reviewed to understand how cost estimation and control are conducted in these State DOTs.
Chapter 3- Review state of practice for fixed budget-best value procurement method in other State DOTs
A critical scanning process is conducted on the FHWA and State DOTs websites to determine their execution process and case studies related to a fixed budget-best value contracting strategy. This procurement method provides an effective strategy to deliver projects under strict budget.
Chapter 4- Conduct statistical analysis to identify important variables capable of explaining variations in submitted bid prices for highway projects
Multiple regression analysis is conducted for investigating the impact of project characteristics, macroeconomic variables, market condition indicators, and oil market parameters on highway construction costs by analyzing the submitted bid prices for highway construction projects in the State of Georgia.
Chapter 5- Conclusions
A summary of research findings is presented. 16

CHAPTER 2 DEVELOPING AND CONTROLLING BASELINE COST ESTIMATE DURING PROJECT DEVELOPMENT PROCESS
2.1. Introduction
Preconstruction activities conducted as parts of the GDOT's plan development process (PDP) have profound impacts on final project cost and schedule. The ability of the Department to influence project cost and schedule is reduced as the project moves along through the PDP. Critical decisions made in the design development process have direct impacts on the final project cost. Therefore, it is important to establish and create a proper baseline budget model as early as possible during the initial phases of the PDP. This baseline budget model should be monitored and maintained throughout the PDP to guide project managers and engineers to stay within the predetermined budget and schedule (Anderson et al. 2007).
The initial stages of the project development process are critical for establishing the baseline for project budget and schedule (Oberlender and Trost 2001). As the project preconstruction phase advances, the owner's influence on modifying the design and tightening the budget to alter project costs is reduced. Figure 2-1 shows that as project advances through different phases and information becomes available, the ability to change project cost is reduced (Chou 2009). If the project development process is not properly
17

managed, there is a good chance that the owner experiences cost overrun and schedule delay for the project. The major root causes of cost overrun and schedule delay include an unreliable baseline cost estimate, a failure to be aware of uncertainty early in the project development, and lack of appropriate risk-related management practices and analysis tools for managing and controlling the established budget and schedule (Shane et al. 2009; Molenaar et al. 2010).
Figure 2-1 Ability to Influence Changes in Cost of a Project State Departments of Transportation (State DOTs) strive to establish feasible and reliable project budget and schedule early in the project development considering project type and other project-specific characteristics. To establish feasible and reliable project budget, it is critical that an appropriate estimating technique is selected based on the level of project scope definition, the project type, and the complexity of the project (Anderson et al. 2007).
18

The American Association of State Highway and Transportation Officials (AASHTO) Technical Committee on Cost Estimating (TCCE) provides detailed information about several cost estimating techniques, including conceptual estimating, historical bid-based estimating, cost-based estimating, and risk-based estimating methods.
A conceptual estimating technique is typically used in the planning or early scoping phase where minimal project definition is available. To develop the project cost, this technique requires two key inputs: (a) historical cost data; and (b) project information (e.g., definition, location, and characteristics) matched to cost data.
A historical bid-based estimating technique, as the most common estimating method, is adapted in all phases of project development process, except the planning phase. With this estimating technique, the estimator should identify items, determine item quantities, and select appropriate historical bid prices based on available project information. The key inputs of this technique are project information, historical bid data, and macroenvironmental and market condition.
A cost-based estimating technique is usually utilized in preparing the engineer's estimate and at the Plans, Specifications & Estimate (PS&E) phase. Since this technique uses relatively well-defined project information (e.g., a work statement and a set of drawings and specifications) and the latest costs of materials, equipment, and labor, the estimator can expect much more accurate project costs than other methods. Based on identified quantities and assumed production rates, the direct costs of labor, materials, and equipment
19

are calculated using the latest cost data. The key inputs of this technique are historical bid data, labor cost, equipment cost, material cost data, subcontract items, and macroenvironment market conditions.
A risk-based estimating technique is an integrated mechanism of risk identification and uncertainty analysis. This technique allows the estimators to estimate an expected value and a range of project costs with consideration of the impact of project risks and uncertainty on the project cost. This estimating technique is utilized in the planning, scoping, and early design phases. The key inputs of this technique include a definition of project complexity and a list of design and estimating assumptions and concerns (Molenaar et al. 2011).
In addition, Harper et al. (2014) concluded that monitoring the cost estimates and the contingency amount has a significant impact on the accuracy of the cost estimates throughout project development process. Through utilizing the questionnaire and interviews, the authors defined five important performance measures for monitoring cost estimates as the following:
1) Bidding accuracy- monitoring the differences between the final plans, specifications, and estimates (PS&E) estimate and previous estimate (i.e., the planning and design estimates) or low bid amounts
2) Estimating accuracy- monitoring the differences between a current cost estimate to previous estimates (i.e., conceptual estimate) or newer, more-developed design estimates (i.e., the final PS &E estimate)
3) Competition effects- monitoring the number of bidders per project let
20

4) Estimating process- monitoring the total time to complete an estimate at each project development phase where an estimate takes place
5) Contingency amount- tracking the amount or percentage of contingency included in an estimate for a project
The authors also claimed that monitoring cost estimates and contingency amount is critical for controlling cost escalation and avoiding design scope creep (Harper et al. 2014).
Besides establishing feasible and reliable project budget and schedule, State DOTs also face a challenge with respect to controlling project budget and schedule throughout project development process. To control project budget and schedule, State DOTs should identify cost escalation factors early in the project development process. Numerous studies have identified cost escalation factors to increase awareness of the causes of project cost escalation. For example, Arditi et al. (1985) investigated the causes of cost escalation for the public construction projects through conducting a survey. This study identified the important causes of cost escalation as the following: inflation due to economic conditions, availability of materials, project delays due to shortages in resources (e.g., labor and equipment), and deficiencies in cost estimates.
Another study by Akinci and Martin (1998) identified factors causing cost overruns during project development process of construction projects. The authors focused on the external factors that are not controllable by owners and contractors. The factors affecting cost estimates of projects include estimator's biases, vagueness in scope, design complexity,
21

and project size. Then, they urged that these factors should be identified and considered during developing cost estimates in project development.
More recently, Shane et al. (2009) identified factors that cause cost escalation during project development process through a combination of an in-depth literature review and intense interviews with over 20 transportation agencies. The authors identified twelve cost escalation factors and classified the factors into internal (e.g., bias, delivery/procurement approach, project schedule changes, engineering and construction complexities, scope creep, poor estimating, and inconsistent application of contingencies) and external factors (e.g., local concerns and requirements, effects of inflation, scope changes, market conditions, unforeseen events, and unforeseen conditions). The authors also found out that the identification of the cost escalation factors aids agency/owners, and contractors in developing more accurate cost estimates and establishing better risk mitigation strategies.
With awareness of the cost escalation factors, it is also necessary to monitor and update the project budget and schedule baselines and update them as design advances throughout various phases while controlling and managing cost escalation factors. Any failure in this process can lead to cost underestimation or designing over budget. The major issue is that while projects are being developed, their budgets and schedules often do not meet the expectations (Chou 2010). The underlying reasons for cost overrun and schedule delay include lack of proper procedures for establishing, monitoring, and updating conceptual cost and schedule estimates within the project development process (Anderson et al. 2007).
22

A well-structured procedure integrated with risk analysis tools is essential for developing more accurate cost estimates and controlling cost escalation during the project development process.
2.2. State of Practice of Cost Estimation Process in Minnesota, California, Texas, Ohio, and Washington State Departments of Transportation
2.2.1. Minnesota Department of Transportation (MnDOT)
2.2.1.1. MnDOT Cost Estimation Process
The Minnesota Department of Transportation (MnDOT) uses a well-structured cost estimation system to prepare cost estimates for highway projects. As shown in Table 2-1, the cost estimation milestones for developing the cost estimate consist of four steps including planning, scoping, design, and letting. Table 2-1 includes major tasks and parties responsible for developing the cost estimate. In the planning phase, MnDOT defines the needs of transportation system improvement and establishes a base for documenting the estimate basis of a project. During the scoping phase, it is critical that all potential issues affecting the project cost and schedule are identified. The baseline cost estimates are developed as an output of the scoping phase. The major tasks in the design and the letting phases are to identify any change in scope, cost, and time, and update and review the baseline cost estimates (MnDOT 2008).
23

Table 2-1 MnDOT Cost Estimation Milestones

Planning (Conceptual Estimating) (1%15%)

Scoping (Scope Estimating) (10% to
30%)

Design (Design Estimating) (30%
to 90%)

Letting (PS&E Estimating) 90%-100%)

PDP

Major Tasks

Identify the needs and deficiencies in the transportation system.
Prepare a Project Planning Report at the end of the Planning Phase.

Extensively investigate all potential issues that could affect the cost and schedule of a project.
Hold a meeting to discuss the project definition.
Identify and analyze project alternatives
Prepare a final Scoping Report. Develop project cost estimates
(baseline cost estimate).

Identify changes in scope, cost, time throughout the Design process.
Update the project cost estimate (update baseline cost estimate).

Prepare the Engineer's Estimate and evaluate contractor bids in relation to the Engineer's estimated cost.
Review Construction Cost Estimates.

Who

Lead
Project Manager
Assistant District Engineer

Support
District Engineer State Estimation
Office Functional Groups External
stakeholders

Lead
Project Manager
Assistant District Engineer

Support
Commissioner's staff
District Engineer State Estimation
Office Functional Groups External
stakeholders

Lead
Project Manager
Assistant District Engineer

Support
Commission er's Staff
District Engineer
State Estimation Office
Functional Groups
External Stakeholders

Lead
Central Office Estimator

Support
District Engineer
Assistant District Engineer
District Estimator
Project Manager
Functional Groups

MnDOT puts efforts into developing a reliable estimate basis for more-accurate cost

estimation throughout the project development. To develop the accurate cost estimate, a

well-defined estimate basis is essential. Table 2-2 describes key inputs for the estimate

basis, which are critical sources in developing the cost estimate for each phase of the

milestones. As design proceeds, these key inputs should be defined, updated, and

documented by the key personnel (e.g., the project manager, functional groups, and the

district engineer) for updating cost estimates throughout project development (MnDOT

2008).

24

Table 2-2 MnDOT Key Inputs for Cost Estimates

Cost Estimation Milestone

Key Inputs

Planning

Planning Estimate Basis- the accumulated information on project requirements necessary for completing a Planning estimate
Project Characteristics- description of the type of project and complexity of the project related to the concept, including site location information (e.g., urban versus rural) and/or data that are relevant to preparing the cost estimate
Historical Data- cost data from previous project used as a basis for pricing different components of the Total Project Cost Estimates
Functional Group Input- cost estimates provided by different Functional Groups

Scoping

Project Estimate File- containing the estimate basis that includes project definition requirements and Scoping summary sheet
Project Characteristics- description of the type of project and complexity of the project, including sitespecific information and/or data
Historical Data- cost data from previous projects Functional Group input- cost estimates provided by different functional groups Market Conditions- understanding of the potential market impact on costs for a project in a given location

Design

Approved Baseline Cost Estimate Package- including cost estimate summaries, cost estimate details, estimate project definition basis, estimate assumptions, estimate calculations
Updated Project Cost Estimate File- containing the updated estimate basis with specific emphasis on changes in the project requirements
Project Characteristics- description of the type and complexity of the project, including site-specific information
Historical Data- updated cost data from previous projects Functional Group Input- updated cost estimates provided by different functional groups Market Conditions- understanding of the potential market impact on costs for a project in a given location
in terms of changes from when the baseline cost estimate was prepared

Letting

Engineer's Estimate Basis File- containing all pertinent information used to prepare an estimate, including the item schedule with quantities
Project Characteristics- description of the type of project and complexity of the project, including sitespecific information and/or data
Historical Data- cost data from previous bids and labor, material, and equipment data for different items Market Conditions- understanding of the potential market impact on construction costs for a given
location

25

MnDOT uses cost estimate review and approval gates to achieve consistent and accurate cost estimates. Figure 2-2 depicts review and approval gates during project development process. At each of these gates, the appropriate management staff (e.g., a project manager, a district engineer, or a Central Office Estimator) should provide an estimate approval before moving the project to the next phase of project development.
Note: HIP= Highway Improvement Plan; STIP= State Transportation Improvement Program
Figure 2-2 MnDOT Cost Estimate Review and Approval Gates 2.2.1.2. MnDOT's Practices for Developing and Controlling Baseline Cost Estimates Cost Estimation and Cost Management System The major purpose of an integrated cost estimation and cost management process is to prepare accurate, reliable, and consistent estimates throughout project development process. This practice allows MnDOT to achieve standardization and documentation of
26

project estimating and cost management activities and deliverables, from the planning phase through the letting phase. Figure 2-3 illustrates a work breakdown structure for the cost estimation and cost management process during project development process (MnDOT 2008).

Perform Cost Estimation and Cost Management

Perform Cost Estimation and Cost Management During Planning

Perform Cost Estimation and Cost Management during Scoping

Perform Cost Estimation and Cost Management during Design

Perform Cost Estimation and Cost Management during Letting

Determine Estimate Basis
Prepare Base Estimate

Determine Estimate Basis
Prepare Base Estimate

Update Cost Estimate for Construction Management
Assess Potential Changes for Construction Management

Determine Estimate Basis
Prepare Base Estimate

Determine Risk and Set Contingency

Determine Risk and Set Contingency

Determine Risk and Set Contingency

Review and Approve Estimate
Determine Estimate Communication Approach

Review and Approve Estimate
Determine Estimate Communication Approach

Review and Approve Estimate
Determine Estimate Communication Approach

Figure 2-3 MnDOT Cost Estimation and Cost Management Process The following two steps are performed in all Phases of the MnDOT's project development process:

Step 1. Identify all the existing and desired functions and sub functions relevant to MnDOT cost estimation and cost management process Step 2. Integrate the desired cost estimation and cost management functions with the preconstruction phases of MnDOT project development process
27

The entire MnDOT organization (e.g., the commissioner's staff, transportation program committee, district engineer, and central office estimator) provides inputs into the cost estimation process.
One of the unique features of the MnDOT's cost estimation and cost management process is that the process contains risk analysis to develop the cost estimates at each phase of project development process. The activity of "Determine risk and set contingency" allows the project team to:
a. acknowledge uncertainty, risk, and associated contingencies early in the project development process
b. determine the extent of risk analysis based on the project's complexity, local impacts, and other considerations
In addition, the cost estimation and cost management process includes review and approval procedures for the cost estimates. These processes increase the accuracy and completeness of the cost estimates by verifying cost estimate package and data.
Lastly, the activity of "Determine estimate communication approach" enables to establish accountability for all cost estimates and avoid miscommunications about the cost estimates between entities involved in the project.
The following are the major benefits and advantages of the MnDOT's costs estimation and cost management process:
a. A standard process for estimating, managing, and controlling costs 28

b. Reliable and accurate estimates c. Improved communication and credibility with external stakeholders d. Clear accountability for cost estimating and management Variance Reports on Cost and Schedule
Variance reports are used to capture changes in cost and schedule and provide a mechanism for budget control through tracking changes and alerting project personnel of changes. Three steps are performed in preparation of variance reports: Step 1. Compare a current estimate with baseline cost estimate Step 2. If needed, note and explain deviations from the baseline Step 3. Generate variance reports at key project milestones or when significant changes in the project occur Variance reports are prepared by the project personnel, particularly the project manager, throughput all phases of project development, most importantly during the design phase. One of the unique features of a variance report is that it is a transparent notification system for alerting project personnel of deviations in project baseline costs and/or schedule. As shown in Figure 2-4, the variance report can provide the detailed project information by using the columns. With this report, the project manager can acknowledge the difference between the baseline cost estimate and the updated cost estimate and track the differences. In addition, the variation report provides an explanation for the cost increase or decrease so that the project manager can efficiently allocate resources including the personnel.
29

Figure 2-4 Template of Variance Reports The main benefits/advantages of variance reports are:
a. Early identification of differences in project cost and schedule b. Proper resource allocation 2.2.2. California Department of Transportation (Caltrans)
2.2.2.1. Caltrans Cost Estimation Process
The California Department of Transportation (Caltrans)'s cost estimates are classified into two major estimates: (1) project planning cost estimates; and (2) project design cost estimates. Since project planning cost estimate becomes the baseline cost estimate for a project, the estimator/project team should clearly determine the cost estimate. This cost
30

estimate will be used for project justification, analysis of alternatives, approval, and programming at the planning phase. The following cost estimates are project design cost estimates, which are developed after project approval and with more detailed information of a project (e.g., engineering and environmental studies). As shown in Table 2-3, the two major cost estimates are further divided into several sub cost estimates, which are developed as major tasks are completed at each phase of cost estimation milestones. Caltrans utilizes two major information sharing systems, the Project Report (PR)/Project Study Report (PSR) and Basic Engineering Estimating System (BEES), to maintain, update, and approve project information, cost estimates, and changes in project scope, costs, and schedule. Caltrans prepares critical information, such as an environmental document, consideration of public comments, and the selection of a preferred alternative, in the project report. This information becomes the estimate basis in developing cost estimates. The BEES assists the project engineer/estimator in preparing the cost estimate in the design phase. The BEES contains critical information, such as contract items, supplemental work, and contingencies used for developing the final engineer's cost estimate (Caltrans 2007a). Through these systems, key stakeholders collaborate to develop the accurate cost estimates during project development process. The key inputs for cost estimating are described in Table 2-4 (Caltrans 2007a).
31

Table 2-3 Caltrans Cost Estimation Milestones

1st

Project Planning Cost Estimates

Major Tasks PDP

Project Feasibility Cost
Estimate

Project Initiation Cost Estimate

Draft Project Report (PR) Cost
Estimate

Project Report (PR) Cost Estimate

Prepare project cost information planning studies

Obtain Additional information (e.g., existing and forecasted traffic volume)
Evaluate and validate the project cost estimate and assumptions

Calculate the cost estimate for each project alternative
Complete Environmental and hazardous waste studies

Complete the public hearing process, selection of the preferred alternative, and completion of the environmental document

Who

Lead

Support

Lead

Project Engineer
Project Manager

Headquarters Divisions
District Right-ofway
District Director
External stakeholders

Project Engineer
Project Manager

Support

Lead

Headquarters Divisions
District Right-of-way
District Director
External stakeholders

Project Engineer
Project Manager

Support

Lead

Headquarters Divisions
District Right-ofway
District Director
External stakeholders

Project Engineer
Project Manager

Support
Headquarters Divisions
District Rightof-way
District Director
External stakeholders

2nd

Project Design Cost Estimates

PDP

Preliminary Engineer's Cost Estimate

Final Engineer's Cost Estimate

Major Tasks

Prepare cost estimates using Basic Engineering Estimating System (BEES)
Update frequently during the design phase as the project construction details, specifications and plans are finalized into a contract document

Complete the final engineer's cost estimate Certify that the estimate is completed and accurate,
reflects the true scope of work, and accounts for current market trends

Who

Lead
Project Engineer Project Manager

Support
Headquarters Divisions District Right-of-way District Director External stakeholders

Lead
Project Engineer Project Manager

Support
Headquarters Divisions District Right-of-way District Director External stakeholders

Note: PDP= Project Development Process

32

Table 2-4 Caltrans Key Inputs for Cost Estimates

Cost Estimation Milestones

Key Inputs

Project Feasibility Cost Estimate
Project Initiation Cost Estimate
Draft Project Report Cost Estimate
Project Report Cost Estimate
Preliminary Engineer's Cost
Estimate
Final Engineer's Cost Estimate

Project Information- each function group (e.g., materials, structural design, traffic, and right-of-way) provides the information for developing project information
Existing facilities High cost items that have impacts on the cost estimates Contingencies- between 30 and 50% Appropriate mapping- having accurate maps leads to the reliable estimate basis (e.g.,
environmental studies) and cost estimates (e.g., the right-of-way cost estimate). Project Information- including existing and forecasted traffic volume, geotechnical
design information, materials and pavement structural section design information, advance planning studies for new structure and modifying existing structures, hazardous waste assessment, potential environmental issues and mitigation, right-ofway and utilities data sheets, traffic handling and transportation management plans, and utilization of existing resources Contingencies- 25% Cost estimates for each competing project alternative Updated Project Information from the functional groups (e.g., materials, structural design, traffic, and right-of-way) Contingencies 20 % Public Hearings Preferred alternative Environmental document- including the purpose and need for the transportation improvement, project alternatives, public hearing, etc. Contingencies- 15% Project Construction Details Specifications and Plans Contract Item prices Contingencies- up to 10% All contract items with Quantities Certification of Engineer's Estimate for projects with an engineer's estimates greater than $5 million (District Director) Comparison with Contractor bids received Contingencies- 5% or less

To have the reliable estimate basis, Caltrans encourages functional groups to provide the

information that requires developing cost estimates during project development. In

addition, Caltrans has predefined contingency amounts for cost estimates for each cost

33

estimation milestone. Caltrans should certify the accuracy and completeness of the final engineer's estimate for projects greater than $5 million. 2.2.2.2. Caltrans' Practice for Developing and Controlling Baseline Cost Estimates Quality Assurance (QA)/Quality Control (QC) Certification Practice The main purpose of the Caltrans' QA/QC certification practice is to certify that the contract cost estimate is complete and accurate while the cost estimates reflect the true scope of the work and account for the most current market trends (Caltrans 2007b). The QA/QC process contains four steps to develop the best estimate possible throughout the Plan Specification and Estimate (PS&E) development phase. Step 1. Identify roles and responsibilities for QA/QC process. Step 2. Conduct major activities of the QA/QC process, such as calculating unit quantities, calculating working dates, and verifying all estimates, shown in Table 2-5. Step 3. Verify all factors used in developing the cost estimates, shown in Table 2-6. Step 4. Obtain approval of key stakeholders such as the project engineer, the design senior, the central region estimate specialist, and the project manager. A sample template for approval of key stakeholders is shown in Figure 2-5.
34

Project Engineer (PE) Assistant PE
District Estimator Construction Representative Design Senior
Specification Writer Project Manager Single Focal Point District Director

Table 2-5 Roles and Responsibilities for QA/QC Process

A=Accountable I=Input S=Signature QA=Quality Assurance QC=Quality Control

Major Activity

Responsibility

Calculates Unit Quantities

A QA

QC

Calculates Working Days

A QA I

I

QC

Determines Unit Prices

A

I/QA I

QC I

Ensures project estimate reflects
A QA I
work required by the plans and specs

QC QA

Ensures project estimate is not

A

QA

QC

inflated or constrained

Ensures project scope and estimate is
I
within budget Prepares Estimate Certification
I/QA
(Supporting Data) form

QC

A

I

QC

Q A
A

Prepares Certification Memo for

A

S

estimates > $5 Million

Verifies all estimates

I

I

I

A/S

I

Table 2-5 represents all entities involved in the QA/QC process and describes their roles

and responsibilities. The unique features of the QA/QC certification process allow the

project to prepare complete and accurate project estimates corresponding with project scope and budget by verifying critical factors that are considered in developing cost

estimates. Table 2-6 lists the factors with the detailed information. In addition, the outcome

of this process is a certification of project cost estimates.

The main benefits/advantages of QA/QC certification practice are:

35

Quality Control

a. Minimal cost changes by reviewing and verifying quantities and the unit
costs in comparison with recent bid openings and market trends
b. The higher accuracy of the final cost estimates by verifying assumptions
and contingencies Table 2-6 Critical Factors for QA/QC process
Factors considered in developing cost estimates Assumptions: How dis assumptions about location (e.g., terrain, distance to construction site, etc.), relative availability of materials, weather conditions, etc., influence the cost estimate? What other elements influence the estimate? Source of Unit Prices: What factors were considered to determine unit prices of major items? Provide expenditure authorizations (EAs) of projects considered, unit prices and quantities used. Add specialty items and costs as appropriate. Traffic Management Plan: Identify lane closure windows and assumptions about traffic control costs and elements (e.g., number of signs, public outreach, component, night work, etc.). Risk Management Plan: Identify risks relating to the development and use basic engineering estimating system (BEES). Escalation Factors Used: Explain forecasted variables and assumptions used. Demonstrate forward estimating method. Contingencies: Is 5% contingency adequate to address each risk factor? If not, why not? How much more is needed? DES Structure verification of Estimate and Quantities: List who prepare calculation data, and verify calculation and data. Constructability Review: What is the assumed construction method and what risks are associated with that method? Indicate when reviews occurred and major findings. DOE Cost Estimate Review: List Completion data and conclusions of the review. Value Analysis Performed: List completion data and any alternatives that impact cost. DES Structural Liaison Review: List date and conclusion of Review and name of reviewer Independent Estimate Performed: List completion data and variance if any, from Caltrans estimate. If variance, explain how resolved. Variance from Programmed Funds (%): Compare current program cost to PS&E BEES. Next Cost Estimate Update: List projected date (three weeks before the California Transportation Commission vote).
Note: DES= Division of Engineering Services; DOE= Division of Equipment
36

Quality Assurance

Status

Note: OE= Office Engineer; CR= Central Region; PJD= Project Development
Figure 2-5 Template for QA/QC Documentation
37

2.2.3. Texas Department of Transportation (TxDOT) 2.2.3.1. TxDOT Cost Estimation Process The Texas Department of Transportation (TxDOT) has five phases of cost estimation process as shown in Table 2-7. In the planning and programming phase, TxDOT develops a preliminary cost estimate, which becomes the baseline cost estimate for a project. Throughout the design and PS&E developments, the preliminary cost estimate will be updated until the agency finalizes the project cost estimate. The preliminary design phase consists of three major design developments, including preliminary schematic, geometric schematic, and value engineering, for review and approval of cost estimate (TxDOT 2014). TxDOT stores all information of the estimate basis and cost estimates in the Design and Construction Information System (DCIS), an automated information system for planning, programming, and developing projects (TxDOT 2006). Through the DCIS, TxDOT updates and shares the cost estimate and the project information with all stakeholders.
38

Table 2-7 TxDOT Cost Estimation Milestones

PDP

Planning and Programming (10%)

Preliminary Design (30%) (Preliminary
Schematics)

Preliminary Design (30-50%) (Geometric
Schematics)

Preliminary Design (3050%) (Value Engineering)

PS&E (100%)

Major Tasks

Gather preliminary information Use Advance Planning Risk
Analysis (APRA) tool to measure project scope definition and identify potential risks Use AASHTO cost estimation program Estimator Prepare construction and ROW cost estimate Prepare preliminary estimate (baseline cost estimates)

Organize design concept conference & prepare design summary report
Prepare preliminary pavement designs
Update construction and ROW cost estimates and corresponding DCIS data
Update cost estimate

Prepare pavement design report
Perform preliminary hydraulic analysis/design
Determine right of way and access needs
Identify potential utility conflicts
Update project scope Update cost estimate

Lead

Support

Lead

Director of

District staff

Project

Transportation Planning and Development

Project Manager District Planner Functional

Manager

Who

groups Transportation

Planning and

Programming

Division

Note: ROW= Right of Way; VE= Value Engineering

Support
Environmental coordinator
Design engineers
Roadway design engineer
Functional groups

Lead Roadway
design engineer Project Manager

Support Hydraulic
engineer District
pavement engineer

Gather project team experts
Consider redesign of alternatives if needed
Conduct VE study Make necessary design
changes Document design changes Revise design based on VE study findings Update cost estimate

Finalize roadway design
Finalize project design, Review environmental
studies, ROW plans, and utilities relocation requirements Update APRA Prepare final engineer's estimate

Lead Roadway
design engineer Project manager

Support
VE Coordinator Design Division
Engineers & Staff District Engineer District Executive Decision Team

Lead Roadway
design engineer Project Manager

Support District design
engineer Hydraulic
engineer

39

Table 2-8 TxDOT Key Inputs for Cost Estimates

Cost Estimation Milestones
Planning and Programming
Preliminary Design (Preliminary Schematics)
Preliminary Design (Geometric Schematics)
Preliminary Design (Value
Engineering)
PS&E

Key Inputs
An accurate scope of work- including type of work proposed, proposed typical section, existing geometry, earthwork and retaining walls or sloped embankments, drainage issues and possible solution, etc.
The critical elements of the project scope identified with the Advance Planning Risk Analysis (APRA) tool
Recent unit bid prices for similar projects Contingencies- 6% to 11% Geographic location (i.e., remoteness) and proximity to material sources Recent bid prices on similar projects Anticipated difficulty of construction Presence of restricted work areas or schedules Project size relative to previous project sizes Proposed project schedule Expected construction staging Changes in the project scope- including design refinements, route or design
alternative selection, utility conflicts, environmental mitigation measure, public involvement, or value engineering analysis findings Geographic location (i.e., remoteness) and proximity to material sources Recent bid prices on similar projects Anticipated difficulty of construction Presence of restricted work areas or schedules Project size relative to previous project sizes Proposed project schedule Expected construction staging Revised design from Value Engineering Study findings Geographic location (i.e., remoteness) and proximity to material sources Recent bid prices on similar projects Anticipated difficulty of construction Presence of restricted work areas or schedules Project size relative to previous project sizes Proposed project schedule Expected construction staging Additional or update data- including the right-of-way maps, as-build construction plans, traffic data, site information, the completed schematic design and project scope

During the TxDOT's cost estimation milestones, several key inputs are required for

developing project cost estimates. Table 2-8 describes the key inputs (TxDOT 2014).

40

As the key information for the cost estimates are obtained at the preliminary design phases (preliminary schematics, geometric schematics, and value engineering), the project manager/estimator should pay careful attention in updating baseline cost and schedule during these phases. 2.2.3.2. TxDOT's Practices for Developing and Controlling Baseline Cost Estimates Advance Planning Risk Analysis (APRA) A Spreadsheet-based Tool The APRA assists the project team/key stakeholders in measuring project scope definition and identifying potential risks/elements that may impact project cost and schedule. TxDOT uses the APRA as a comprehensive checklist to identify potential risks by looking at critical risk elements listed in Table 2-9. The APRA consists of three main sections that include 12 categories that are further broken down into 59 elements (Caldas et al. 2007).
41

Table 2-9 APRA Sections, Categories, and Elements

Section 1 Categories
Elements
Section 2 Categories
Elements
Section 3 Categories
Elements

Basis of Project Decision

A. Project Strategy

B. Owner/Operator Philosophies

C. Project Requirements

A1. Need & Purpose Documentation

B1. Design Philosophy

C1. Functional Classification & Use

A2. Investment Studies & Alternatives

B2. Operating Philosophy

C2. Evaluation of Compliance Requirements

Assessments

B3. Maintenance Philosophy

C3. Survey of Existing Environmental

A3. Programming & Funding Data

B4. Future Expansion & Alteration

Conditions

A4. Key Team Member Coordination

Consideration

C4. Determination of Utility Impacts

A5. Public Involvement

Basis of Design

C5. Value Engineering

D. Site Information

E. Location & Geometry

G. Design

H. Installed

F. Structures

Parameters

Equipment

D1. Geotechnical Characteristics

E1. Horizontal &

F1. Bridge Structure G1. Provisional

H1. Equipment List

D2. Hydrological Characteristics

Vertical Alignment

Elements

Maintenance

H2. Equipment

D3. Surveys & Planimetrics

E2. Control of Access

F2. Hydraulic

Requirements

Location Drawings

D4. Permitting Requirements

E3. Schematic Layouts Structures

G2.Constructability

H3. Equipment

D5. Environmental Documentation

E4. Cross-Sectional

F3. Miscellaneous

Utility Requirements

D6. Property Descriptions

Elements

Design Elements

D7. Ownership Determinations

D8. Right-of-Way Mapping

D9. Constraints Mapping

D10. Right-of-Way Site Issues

Execution Approach

I. Acquisition Strategy J. Deliverables

K. Project Control

Project Execution Plan

I1. Long-Lead Parcel & Utility

J1. Computer Aided

K1. Right-of-Way & Utilities

L1. Environmental Commitments

Adjustment

Drafting and Design

Cost Estimates

&

Identification

(CADD)/Model

K2. Design & Construction Cost Mitigation

I2. Long-Lead/Critical Equipment & Requirements

Estimates

L2. Interagency Coordination

Materials Identification

J2.

K3. Project Cost Control

L3. Local Public Agency

I3. Local Public Agencies Utilities

Documentation/Deliverables K4. Project Schedule Control

Contractual

Contracts & Agreements

K5. Project Quality Assurance Agreements

I4. Utility Agreement & Joint-Use

& Control

L4. Interagency Joint-Use

Contracts

K6. Safety Procedures

Agreements

I5. Project Delivery Method &

L5. Preliminary Traffic Control

Contracting

Plan

Strategies

L6. Substantial Completion

I6. Design/Construction Plan &

Requirements

Approach

I7. Procurement Procedures & Plans

I8. Appraisal Requirements

I9. Advance Acquisition

Requirements

42

The APRA is a spreadsheet-based tool. Figure 2-6 outlines the project score sheet for project scope evaluation. The following steps are followed in the implementation of the APRA: Step 1: Use the project assessment sheet and read its corresponding description Step 2: Discuss issues and review documents if needed Step 3: Choose only one definition level (0, 1, 2, 3, 4, or 5) for each element. All elements have six pre-assigned scores, one for each of the six possible levels of definition Step 4: Add each of the element scores within a category to produce a total score for that category Step 5: Add the three section scores to achieve a total APRA score
43

Figure 2-6 Sample of TxDOT's APRA Tool 44

The APRA is used in planning and programming, and preliminary design phases. The APRA is applied on four gates depicted in Figure 2-7.
a. The first assessment is typically held for projects at the Feasibility and Scoping meeting.
b. The second assessment is typically held for project at a Preliminary Design Conference.
c. The third assessment is typically held for project before preceding to the Plans, Specification, and Estimates development phase.
d. The forth assessment is typically held at the end of the Plans, Specifications, and Estimates development phase, prior to letting.
Figure 2-7 APRA Application Points
45

Table 2-10 APRA Review Gates and Purposes

Gates

1st Gate

2nd Gate

3rd Gate

4th Gate

When

Before proceeding the feasibility and scoping phase /at the feasibility and scoping meetings

Before proceeding the preliminary design phase/at Preliminary Design Conference

Before proceeding the PS&E phase

At the end of PS&E phase

Purpose

To align the team with project objectives;
To ensure good communication among the decision makers and the project development team; and
To highlight stakeholder expectations to facilitate reasonable engineering estimates.

To align project objectives and stakeholders' needs;
To identify high priority project deliverables that need to be completed;
To help to eliminate late-project surprises;
To facilitate communication across the project development team and stakeholders.

To identify risk issues To develop mitigation
plans for risk issues

To reviewing all risk elements by all stakeholders
To resolve all major issues
To control any residual risk elements

Table 2-10 describes the sub-objectives of the APRA tool at each gate. All major

disciplines (e.g., right-of-way, utilities, environmental, design, and planning and

programming) participate in the APRA assessment to identify the critical elements of the

project scope.

The APRA has several unique features for project development as follows:
A checklist that a project team can use for determining the necessary steps to follow in defining the project scope
A listing of standardized scope definition terminology throughout the transportation construction industry
An industry standard for rating the completeness of the project scope development to facilitate risk analysis and prediction of escalation, potential for disputes, etc.
A means to monitor progress at various stages during the advance planning phase and the project development process

46

A tool to aid in communication and to promote alignment between owners (e.g., TxDOT), design contractors, and other stakeholders by highlighting poorly defined areas in the project scope
A means through which project team participants can reconcile differences using a common basis for project evaluation
A training tool for organizations and individuals through the industry A benchmarking tool for organizations such as TxDOT to use in evaluating the
completion of scope development versus the performance of past projects, both within their organizations and without, in order to predict the probability of the success of future projects.
The main benefits/advantages of the APRA are:
a. The improvement of project performance in terms of both cost and schedule b. The identification of the project requirements in all major disciplines by
quantifying, rating, and assessing the level of scope development Design Summary Report (DSR) including the Design Conference
The main purpose of the DSR is to document project information and ensure that the project does not overlook potential critical issues in the project scope development (TxDOT 2014).The DSR documents the agreed upon fundamental aspects, concepts, and design criteria of a project and serves as the definition of the project's scope. Key stakeholders need to update design summary throughout all phases of project development process. As project progresses, the DSR needs to be revised and the agency needs to obtain approval from entities and share information with. Project manager and function groups (e.g., rightof-way, utility, and design offices) are responsible for the execution of the ARPA.
47

As stated earlier, the DSR summarizes information of the key elements that are necessary in developing cost estimates and identifying risks for a project. Figure 2-8 shows a sample of proposed hydraulic elements in the DSR. The key elements of the DSR are as follows:
Programming and Funding Data Existing Elements Advanced Project Development Element Proposed Right of Way & Utility Elements Proposed Geometric Design Elements Proposed Bridge Design Data Proposed Hydraulic Elements Proposed Pavement Structure Elements Proposed Traffic Operations Elements Proposed Miscellaneous Elements Accelerated Construction Procedures
48

Figure 2-8 A Sample Hydraulic Element in the Design Summary Report
The main benefits/advantages of the DSR are: a. The DSR provides reliable basis of estimate for design development and cost estimates based on existing conditions and design elements of a project b. The DSR aligns key stakeholders to input design criteria and elements c. The DSR tracks changes in critical elements (ROW, Environmental, Traffic control, etc.) for the project development
49

2.2.4. Ohio Department of Transportation (ODOT) 2.2.4.1. ODOT Cost Estimation Process The Ohio Department of Transportation (ODOT) develops cost estimates along with cost estimation milestones shown in Table 2-12. The cost estimation milestones consist of four phases, including planning, preliminary engineering, environmental engineering, and final engineering/right-of-way (ROW), before letting. The baseline cost estimate is developed in the planning phase and updated during the rest of project development process (ODOT 2013). To efficiently develop projects, ODOT classifies transportation projects into five paths to have customized scoping process depending on project size, project complexity, and/or potential impact to the environment. These classifications allow ODOT to have the flexibility in adjusting required tasks to address project needs. The customized scoping process ensures that only necessary work is conducted and that the project is properly planned and developed (ODOT 2013; ODOT 2016). For example, since feasibility study and alternative evaluation report are not necessary in selecting the preferred alternative for a simple improvement project, ODOT can expedite in scoping process by selecting the alternative early in the project development process, at the planning phase. Table 2-11 describes five project paths in detail.
50

Table 2-11 ODOT Five Project Paths

Paths

Description

Path 1

Path 1 projects are defined as "simple" transportation improvements generated by traditional maintenance and preventative maintenance. They involve minor structure and roadway work with no ROW/utility impacts. These are typically NEPA exempt or CE Level 1 NEPA documents.

Path 2 projects are also simple projects (similar to Path 1- minor structure and minor roadway work),
Path 2
however, these jobs can involve ROW/utility impacts. These jobs are typically CE Level 1 documents.

Path 3

Path 3 projects involve a higher level of complexity than projects in Path 1 or 2. They involve projects such as: moderate roadway and structure work, intersection and minor interchange upgrades, minor realignments, reconstruction, median widenings, etc. They can involve utility and ROW impacts including relocations.

Path 4

Path 4 projects involve complex roadway and structure work that may add capacity such as: highway widening, new alignments in suburban or rural settings, reconstruction, access management, complex bridge replacement and/or multiple intersection/interchange alternatives. They may have high utility and/or ROW relocations/impacts. These are typically CE Level 3 or higher level NEPA documents.

Path 5

Path 5 projects have the highest complexity and involve projects like: new capacity-adding alignments in complex urban centers, major highway widenings, reconstructed interchange or new interchange. These projects will have high ROW relocations/impacts, complex utility issues, multiple alternatives and access management issues. These projects are typically higher level NEPA documents and will require additional scoping reviews before acceptance.

Note: CE= Categorical Exclusions; NEPA=National Environmental Policy Act

51

Table 2-12 ODOT Cost Estimation Milestones

PDP

Planning (10%)

Preliminary Engineering (Stage 1
Design)

Environmental Engineering (Stage
2 Design)

Final Engineering/ROW (Stage 3 Detailed
Design)

Major Tasks

Develop project concept Develop accurate cost

and scope

estimates for all feasible

Identify environmental,

alternatives

ROW, utility, and design, Conduct feasibility study

geotechnical, engineering and NEPA studies

issues in the preliminary Analyze alternatives and

information package

identify a preferred

Conduct field review

alternative(s) using the

Prepare base cost estimate Alternatives Comparison

including construction,

Matrix

utility, and ROW

Conduct VE study

Involve

Establish and develop the

public/stakeholder

design parameters to

generate an accurate scope,

schedule, and budget

Involve public/stakeholder

Obtain NEPA and permit approvals
Refine the level of impacts associated with the preferred alternative
Conduct VE study Prepare environmental
mitigation cost estimates Develop preliminary ROW
and utilities plans and refine estimates Update cost estimates and milestone dates Involve public/stakeholder

Finalize design package Perform ROW/Utility
acquisition Update construction, right-
of-way acquisition, and utility reimbursement cost estimates Involve public/stakeholder

Lead
Project manager

Support
District staff Appropriate
specialists (functional groups)

Lead
Project manager

Support
Design engineers
All functional groups (e.g., ROW, Utilities, and Environmental)
VE Coordinator

Lead
Project manager

Support
VE Coordinator All functional
groups (e.g., ROW, Utilities, and Environmental) Design engineers

Lead
Project manager

Support
District Environmental coordinator
All functional groups (e.g., ROW, Utilities, and Environmental)
Design engineers Office of Contracts

Who

Note: PDP= Project Development Process; ROW= Right of Way; VE= Value Engineering; NEPA= National Environmental Policy Act

Table 2-13 provides key inputs for developing cost estimates throughout project

development. In the planning phase, ODOT defines the needs of deficiencies in the

transportation system through gathering information from functional groups. To update

the baseline cost estimate, ODOT uses several major inputs, such as feasibility study, the

52

Alternative Evaluation Report (AER), and the Value Engineering study, during the

preliminary engineering and environmental engineering phases. Lastly, ODOT prepare the

PS&E package for developing the final engineer's estimate (ODOT 2013).

Table 2-13 ODOT Key Inputs for Cost Estimates

Cost Estimation Milestones

Key Inputs

Planning (10%)

Determination of the problem (the transportation problems, and existing and future conditions)
Information from technical studies, the environmental secondary resource review, site visits, and engineering reviews
Definition of the scope work Goals and objectives for the project Scope for the preliminary engineering phase (schedule, deliverables,
and budget)

Preliminary Engineering (Stage 1 Design)

Goals, roles, and responsibilities for all project team members Technical studies Waterway permit determination Preferred alternative Feasibility study Alternative Evaluation Report (AER)

Environmental Engineering (Stage 2 Design)

Goals, roles, and responsibilities for all project team members Technical studies Environmental field studies and impact analyses NEPA document Environmental commitments summary Waterway permit application Value Engineering Study Constructability review Right-of-way plans

Final Engineering/ROW (Stage 3 Detailed Design)

Final right-of-way plans and tracings Completed and submitted plan package The PS&E package Final legislation All pre-bid questions and issues Federal authorization

Note: NEPA= National Environmental Policy Act; PS&E= Plans, Specifications, and Estimates

53

2.2.4.2. ODOT's Practices for Developing and Controlling Baseline Cost Estimates

Alternative Evaluation Report (AER)

The AER is used to document all the alternatives and their evaluation to select a preferred alternative. The Alternative Evaluation Report (AER) is applicable in complex Path 3, Path 4, and Path 5 projects as shown in Table 2-14 (ODOT 2016). The AER is prepared when the preferred alternative cannot be defined/chosen in Feasibility study.

Table 2-14 ODOT's Alternative Selection Methods

Project Milestone

When is information prepared to define the Preferred Alternative?

Path 1

Path 2

Path 3
Non-complex Complex

Path 4

Path 5

Project Initiation

Project's description, method, and
footprint

Project's description, method, and
footprint

Feasibility Study

Project's method and
footprint

Project's description, method, and
footprint

Project's description
and footprint

Project's description
and footprint

Project's description
and footprint

Alternative Evaluation
Report

Project's footprint

Project's method and
footprint

Project's method and
footprint

Note: Description (what will my project involve and where will it be located?); Method (what design standards will apply, how will we build it, and how will traffic be maintained?); Footprint (what are the limits that should be used for environmental clearance, will there be temporary impacts?)

The following steps are conducted by project manager in the district and central offices to develop the AER during the Preliminary Engineering Phase:

54

Step 1. Discuss environmental and design issues of all the alternatives Step 2. Summarize the advantages and disadvantages for each alternative Step 3. Use the combination of both design and environmental factors that contribute to the selection of the preferred alternative ODOT summarizes all the alternatives in the AER. In Table 2-15, major components of the AER are listed. ODOT evaluates several alternatives based on technical analysis, costs, long-term versus short-term solutions, and stakeholder involvement. To effectively evaluate several alternatives, ODOT uses an alternative comparison matrix to eliminate alternatives and select the feasible alternative among proposed multiple alternatives.
55

Table 2-15 Components of ODOT's Alternative Evaluation Report

Alternative Evaluation Report Executive Summary

Typical Alternative Evaluation Engineering Element

Introduction/Background Alternatives Traffic Analysis Roadway Assessment Drainage Assessment Geotechnical Assessment Right of Way Assessment Utility/Railroad Assessment Environmental Analysis Public Involvement Alternative Comparison Recommendation

Field Survey and Mapping Typical Sections including Lanes, Curbs,
Sidewalks, Trees, Lawns, and Shoulder Widths Alignments (Horizontal and Vertical) Clearances (Horizontal and Vertical) Field Survey Information including Topography, Bridges, Utilities, Channels, and Railroads Geological and Soil Boring Data (Highway/Bridge Foundations) Drainage Conceptual Drainage Design including Preliminary Culvert Sizes Traffic Control Signal Warrants Maintenance of Traffic Conceptual Maintenance of Traffic Utilities Right-of-Way, Construction, and Utility Reimbursement Cost Estimates Locate Waste and Borrow Areas Railroad Coordination Value Engineering

The main benefits/advantages of the AER are:

a. ODOT expects to mitigate funding constraint issues by allowing project managers to choose a cost effective alternative
b. ODOT expects to minimize scope changes or issues with the selection of an optimal alternative by considering impacts of technical, engineering, and design issues

56

2.2.5. Washington State Department of Transportation (WSDOT) 2.2.5.1. WSDOT Cost Estimation Process The Washington State Department of Transportation (WSDOT) develops the cost estimates along with cost estimation milestones as shown in Table 2-16. Although the planning phase estimate is developed with limited project information, WSDOT puts efforts into developing the accurate cost estimate by utilizing information collected from project stakeholders, historical projects, and field review. In the scoping phase, WSDOT sets the baseline cost estimates/the project budget. Throughout the design and the PS&E phases, WSDOT updates the cost estimate and refines risks, uncertainty, and assumptions that are used in preparing the cost estimate (WSDOT 2015).
57

Table 2-16 WSDOT Cost Estimation Milestones

Major Tasks

PDP

Planning
Determine the full scope of the project
Develop a comprehensive list of all of the components for the project
Conduct a project field review and document potential high-cost items (costs of mitigating hazardous waste and other environmental impact, utility relocation, etc.)
Review the unit price to reflect current trends.
Review planning estimates and assumptions

Scoping
Set the baseline cost estimates Document assumptions and scope
definitions Choose the correct unit costs for
items in current dollars Justify any changes in the cost of
the project Ensure major risks to the project Conduct a site visit Understand the current design
standards and their project impact Identify the major items of work Review the project constructability Verify traffic control strategy

Design
Conduct geometric review, general plans review, and preliminary contract review
Track changes in the estimated cost Compare the current budget for the
project cost and schedule to the new estimate Document each update and provide a written explanation of any significant changes Get a quote for materials sources Coordinate with the appropriate entities for the review of the cost estimates

PS&E
Conduct the final independent QA/QC checks of calculations, prices, and assumptions
Review the basis of estimate for completeness, accuracy, and clarity
Check quantities of major items and cost drivers
Review special group estimates for scope and cost
Review contract special provisions Evaluate the potential impact of staging,
materials storage, hauling of materials, location of batch plants, and other constructability related issues

Lead

Support

Lead

Support

Lead

Support

Lead

Project

Region Planning

Project

Region Plans Office Project

Assistant State Design Project Manager

Manager

Manager

Manager

and Programming

Manager

Engineer

Estimator

Estimator Project Engineer

Estimator

Project Engineer

Estimator Project Engineer

Designer

Who

Region Plans Office

Region Plans Office Designer Project Development

Peer Review Team

Peer Review Team

Engineer

Specialty Group

Specialty Group

Region Plans Office

Peer Review Team

Specialty Group

*PDP= Project Development Process; PS&E= Plans, Specifications, and Estimates; QA/QC= Quality Assurance/Quality Control

Support Construction staff Specialty Group Region Plans Office Peer Review Team Project Engineer Project Development
Engineer

58

WSDOT has a rigorous process to develop the cost estimates for each phase of cost estimation milestones. Figure 2-9 depicts the steps of cost estimating process. For developing the cost estimate, this process allows the estimators to have the reliable estimate basis by using historical databases and internal subject matter experts (SMEs).
Note: HQ= Headquarters
Figure 2-9 WSDOT Cost Estimating Process for Each Phase of PDP 59

The potential benefit of this process is implementing risk assessment for each phase cost estimate. WSDOT considers factors of market conditions and inflation for the cost estimate, as well as inputs from external and internal SMEs to capture the effect of uncertainties on project cost and schedule (WSDOT 2015). WSDOT defines the key inputs for developing cost estimates as shown in Table 2-17. WSDOT emphasizes on documenting the estimate basis (e.g., project description, risks, and changes in scope/design) to communicate with key stakeholder and convey key information from one phase to the next of project development (WSDOT 2015).
60

Cost Estimation Milestones

Table 2-17 WSDOT Key Inputs for Cost Estimates Key Inputs

Planning

Description of the project including lead agency, responsible person, estimating processing software, etc.
Purpose, timing, and location of the project The basic scope of the project (e.g., mission/design,
estimate type, project type, and wetlands issues) The primary estimating methodology for the cost estimate Design basis including information that describe the types
and status of engineering and design deliverables for preparing the estimates, as well as any design assumption

Planning basis including descriptions of the project management, engineering, design, and construction approaches Cost basis including methods and sources for determining listed item pricing
Description of allowances, assumptions, exclusions, exceptions
Risks (i.e., all threats and opportunities) in preparing the cost estimate
Estimate quality assurance Reconciliation for any differences in the cost estimate List of all parties in preparing the cost estimate

Scoping

Description of the project Purpose, timing, and location of the project The basic scope of the project The primary estimating methodology for the cost estimate Basis of design, planning, and cost

Description of allowances, assumptions, exclusions, exceptions
Risks (i.e., all threats and opportunities) in preparing the cost estimate
Estimate quality assurance Reconciliation for any differences in the cost estimate List of all parties in preparing the cost estimate

Design PS&E

Description of the project

Risks (i.e., all threats and opportunities) in preparing

Purpose, timing, and location of the project

the cost estimate

The basic scope of the project

Estimate quality assurance

The primary estimating methodology for the cost estimate Documentation reconciliation for any differences in

Basis of design, planning, and cost

the cost estimate

Description of allowances, assumptions, exclusions, List of all parties in preparing the cost estimate

exceptions

Description of the project

Risks in preparing the cost estimate

Purpose, timing, and location of the project

Estimate quality assurance

The basic scope of the project

Reconciliation for any differences in the cost estimate

The primary estimating methodology for the cost estimate List of all parties in preparing the cost estimate

Basis of design, planning, and cost

Quantities of major items and cost drivers

Description of allowances, assumptions, exclusions, Contract special provisions

exceptions

Potential impact of staging, materials storage, hauling

of materials, location of batch plants, and other

constructability related issues.

61

2.2.5.2. WSDOT's Practices for Developing and Controlling Baseline Cost Estimates
Cost Risk Assessment (CRA) and Cost Estimate Validation Process (CEVP) Workshop The CRA and CEVP are systematic project review and risks assessment processes to identify and describe cost and schedule risks associated with the project (WSDOT 2012). CEVP is an intense workshop in which a team of top engineers and risk managers from local and national private firms and public agencies examine a transportation project and review project details with WSDOT engineers. The CEVP can be used in all projects in excess of $100 million. The CRA has the similar process and objective, but less intense than the CEVP. The CRA can be applicable in all projects in excess of $25 million. Sevenstep process of the CRA and CEVP consists of the following: Step 1. Select the project and method. Step 2. Structure the project team effort. Step 3. Define and evaluate the base cost estimate and schedule. Step 4. Assess uncertainty and risk. Step 5. Quantify uncertainty in the project cost and schedule. Step 6. Apply probabilistic analysis and document. Step 7. Implement and measure risk response actions, monitor, and control.
62

The CRA and CEVP are applied in the late-planning phase and early stages of the PS&E phase as represented in Figure 2-10. Participants in CRA and CEVP workshops are listed in Table 2-18.

Plannig

Scoping

Design

Typical Timeframe for CRA/CEVP Workshops

PS&E

Figure 2-10 Timing of CRA/CEVP Workshops

63

Project Team Members

Table 2-18 Roles and Responsibilities of CRA and CEVP Workshop Team

Workshop Team

Roles and Responsibilities

Project Manager Estimator Scheduler Lead Designer Key Technical Experts Project Team Experts Agency Experts (HQ et al.) Other Stakeholders

Prepare project resource (decision maker) Prepare and document project estimate Prepare and document project schedule Prepare primary resource for design questions Specialty groups as needed Internal SMEs work with external SMEs to review and validate project cost and schedule estimates. They provide objective review and comment regarding project issues, risks and uncertainty. At the end of the workshop, the SMEs should provide a brief (i.e. one page) summary of their thoughts about the workshop.

External Consultants

Subject Matter Experts (SMEs)

Cost-Risk Team Members

Risk Lead
Risk Lead Assistant
Cost Lead
Cost Lead Assistant Cost Risk Estimating Management (CREM) Workshop Coordinator

Conducts risk elicitation and manages meeting during risk elicitation; performs, or directs the performance of the statistical analysis. Assists with risk elicitation and meeting management during risk elicitation Conducts Base Cost and Schedule Review and validation; manages the meeting during the review If needed, assists the cost lead position, as appropriate. Coordinates the agenda and participants' discussions, works with the project manager to insure the success of the workshop

The CRA and CEVP methods have the following unique capabilities:
a. Define and review or validate cost and schedule base estimates using a Lead Cost and Schedule Reviewer, Subject Matter Experts, and WSDOT specialists.
b. Document assumptions and constraints used in developing the estimated project cost and schedule range.
c. Replace (or greatly reduce) the traditional project "contingency" with key identifiable risks that can be more clearly understood and managed.

64

d. Under the direction of a risk lead, identify and quantify key events in a project that can cause a significant deviation from the base cost or schedule. This identification and quantification should begin prior to the workshop through Advance Elicitation meetings.
e. Perform a Monte Carlo simulation analysis to model the collective impact of base and risk issues for the complete project as a system to produce an estimate of a reasonable range and distribution.
f. Discuss and develop concepts for responses to risks to the schedule that could impact the cost of the project. Promote pro-active risk management by project teams. Provide the project team with actionable information on risk events that allow them to manage the risks (threats/opportunities) on an on-going basis, via mitigation strategies to better control project costs and schedules.
g. Perform "post-mitigation" analysis to ascertain the effectiveness of planned and/or implemented risk response actions.
The main benefits/advantages of CRA and CEVP are:
a. The improved communication with the public b. The improved team communication c. The increased ability to quantify risks and develop strategic risk
management plans for the identified risk. d. The more realistic cost ranges
65

2.3. Summary of the Recommended Best Practices for Cost Estimation and Control
Changes in estimated project cost is one of the major concerns of State DOTs throughout project development from concept to completion of a project. This chapter reviewed the current state of practices in cost estimation and control in several State DOTs and identified best practices in cost estimation and strategies for cost control. The following recommendations are found out to be effective for enhancing the practice of defining and maintaining the established budget for highway projects.
State DOTs should establish an integrated process for cost estimation and cost management to establish accurate, reliable, and consistent estimates thorough project development process. A cost management process is essential for managing project costs, evaluating the basis of project estimate, and identifying and analyzing project issues (risks) at any phase of the project development process. A systematic process should be in place to collect critical inputs from key stakeholders and validate key information items to arrive at a reliable basis for all components of project cost estimate. For example, Minnesota DOT enhances the accuracy and completeness of the cost estimates by verifying cost estimate package and data. The MnDOT's practice also contains risk analysis to identify risk factors and establish contingency. The MnDOT's project managers utilize this integrated cost management system to mitigate risk and refine contingency for project cost estimates at each phase of the project development process.
66

State DOTs should establish key milestones for estimating, updating, and approving cost estimates as project definition/design advances. Establishing key milestones for cost estimates are indispensable for improving the consistency and accuracy of the cost estimates. For example, Caltrans identifies several cost estimation milestones in different phases of project initiation, preliminary design, and detailed design development. Roles and responsibilities are clearly defined for different project participants from project management team and functional groups to provide information needed for updating and controlling project estimate. Example of specific tasks assigned to the project team are: visiting a project site, selecting a preferred alternative, and preparing project plans and specifications. Minnesota DOT has also developed cost estimate review and approval gates consisting of 7 gates for reviewing and approving cost estimate with approved project information items (e.g., a scope report) from appropriate project management staff.
State DOTs should capture any changes in estimating assumptions to track the basis of cost estimate and control estimated project cost. During the project development process, State DOTs often experience cost increase or decrease. It is recommended that all changes are documented and any deviations from the baseline cost estimate are justified. Information about the change in cost estimate and reasons behind it should be described and shared with other key stakeholders. For example, Minnesota DOT utilizes a reporting system called "Variance Report" for documenting the difference between baseline and updated cost estimates and alerting project participants about deviations in project definition and
67

estimated cost. Variance reports allow project managers to explain cost increase or decrease based on important cost items and scope elements of the project. Variance reports help project managers reconcile deviation in the baseline project definition and budget.
State DOTs are recommended to establish an automated information system to help them maintain, update, and share project information, cost estimates, and changes in project scope, cost, and schedule. As highway construction projects require handling massive amount of project information and data provided by numerous parties, it is difficult to handle the project information and data during a relatively long time span of project development. Thus, a database management system is essential for maintaining, updating, sharing project information and cost estimates. For example, Caltrans has developed the Project Report (PR)/Project Study Report (PSR) and Basic Engineering Estimating System (BEES) for sharing project information. With the PR/PSR, Caltrans is able to summarize key information points about the project, such as scoping study, cost, and overall impact of alternatives to obtain approval for the project to proceed and support development of costs estimates. The Caltrans' BEES program helps project manager/estimator prepare more reliable cost estimates in the design phase. Texas DOT has also developed the Design Summary Report (DSR) that helps the agency prepare a reliable basis for cost estimation by documenting key elements of fundamental aspects, concepts and design criteria of the project. In addition, the DSR allows project managers and estimators to track critical elements (e.g., right-of-way, environmental, and traffic control) throughout the project
68

development. Oregon DOT uses a similar concept called the Alternative Evaluation Report (AER) to describe all design alternatives and explain the evaluation process to select the most preferred alternative. Detailed information items about utilities and site conditions, right-of-way, and environmental assessment are available in the AER that help Oregon DOT's project managers prepare a more reliable basis for cost estimation.
State DOTs should consider potential issues (risks) that may cause cost escalation during developing baseline cost estimates. Risk analysis tools and inputs from key project stakeholders are necessary for identifying critical risk factors for the project. For example, Texas DOT has utilized the Advance Planning Risk Analysis (APRA) system for measuring project scope definition and identifying potential risk factors that may impact project cost and schedule. The APRA helps Texas DOT monitor progress of scope development and identify potential issues and risks in major components of a project at various stages during the project development process. Washington State DOT uses a couple of systematic project review and risk assessment workshops called the Cost Risk Assessment (CRA) and the Cost Estimate Validation Process (CEVP) workshops to identify and describe cost and schedule risks of a project. WSDOT uses these workshops to assess and quantify the impacts of the identified risk factors on project cost and schedule. CRA and CEVP workshops enable the project team to establish proactive risk management for controlling baseline project budget.
69

Finally, State DOTs should utilize a quality assurance/quality control (QA/QC) process to verify the final engineer's cost estimate before a project is advertised. The final validation process is necessary for checking if the cost estimates reflect the true scope of the work and the most current market conditions. For example, Caltrans utilizes a QA/QC certification practice to evaluate project cost estimates and maintain consistency of cost estimates at the program level. The Caltrans' QA/QC process integrates with value analysis and constructability review. Project managers coordinate with other stakeholders to verify all information items and risk factors used in developing project cost estimates, including assumptions, sources of unit price, and a risk management plan.
It is expected that State DOTs realize significant improvements in their cost estimation process by utilizing the identified best practices described above. The major benefits of the identified best practices are as follows.
1. Uniformity and consistency of cost estimates with a well-structured cost estimation system
2. Enhancement of the project scope definition and identification of critical issues/risk areas in major components of a project
3. Maximization of information and the knowledge of a multidisciplinary team through various levels of the review process
4. A closer estimation of the contingency amount by using risk management tools that may differ by project size and complexity 70

5. Verification of the final engineer's cost estimate through a final quality assurance/quality control (QA/QC) process
71

CHAPTER 3 FIXED BUDGET-BEST VALUE
PROCUREMENT METHOD
3.1. Introduction
The highway construction industry in the United States faces significant challenges in rehabilitating aging infrastructure and meeting growing traffic volumes with limited funding. Thus, delivering projects within available funds become far more critical in the highway construction industry. Amid increasing complexity of projects and funding constraints, State Departments of Transportation (State DOTs) have utilized best value procurement methods, such as fixed budget-best value, to maximize the value of dollars expended for their projects. The best value procurement methods can include several evaluation criteria, such as price, schedule, and technical factors, in request for proposals (RFPs). With best value procurement methods, State DOTs can select key factors that match or meet the project's specific requirements (Scott et al. 2006). Based on the key evaluation factors, State DOT selects the proposal that most closely meets or exceeds the owner's expectations and project's requirements.
Fixed budget-best value, also known as "design-to-costs", allows State DOTs to generate the greatest amount of work while achieving the best value for dollars expended (FHWA 2013). This approach encourages the proposers to submit the proposals with the best value that they are able to achieve while staying within the defined budget. As a variation of best
72

value procurement methods, a fixed budget-best value approach provides State DOTs a choice for selecting evaluation and selection criteria, such as project scope, qualifications, and schedule that meet or exceed their project requirements. The National Cooperative Highway Research Program (NCHRP) Project 10-61 provides an example of the fixed price-best value algorithm as follows:
: , T=Technical Score P=Project Price
Using this algorithm, State DOT selects the proposal that obtains the maximum technical score while fulfilling the premise of the fixed budget. The technical score can be calculated based on several types of parameters (e.g., time, qualifications, and design) that the owner requires for the project goal (Scott et al. 2006). As the fixed budget-best value evaluates the proposal by using project scope, qualifications, schedule, and non-cost factors, this approach can be typically applied on design-build projects. As many highway construction projects have suffered from significant cost overrun, this approach provides an attractive alternative for procuring a project with a tight budget.
73

3.2. State of Practice for Fixed Budget-Best Value Procurement Method in Utah, Colorado, Idaho, and Michigan Departments of Transportation
Under the provisions of Special Experimental Project No. 14 (SEP 14) (FHWA 2016), several state departments of transportation currently utilize the fixed budget-best value approach to maximize the use of their available funds. To document state of practice of a fixed budget-best value contracting strategy, a comprehensive review of academic and professional literature was conducted. In addition, a critical scanning process was conducted on the FHWA and State DOTs websites to determine their execution process and case studies related to a fixed budget-best value contracting strategy. The results of scanning indicate that use of a fixed budget-best value approach was successfully utilized in several State DOTs, including the Idaho, Michigan, Utah, and Colorado DOTs. 3.2.1. Utah Department of Transportation (UDOT) The Utah Department of Transportation (UDOT) defines the fixed budget-best value/fixed price-best value in the context of the following three objectives (UDOT 2016):
Knowing funding limitation Maximizing scope for the price Encouraging innovation
With this procurement method, UDOT aims to maximize the amount of work under a single contract while spending all authorized funding for the contract. In addition, UDOT
74

encourages the proposers to develop innovative solutions to achieve the State's goal (UDOT 2013). Since the fixed budget-best value approach provides higher flexibility in design and construction methods and techniques than that in traditional procurement methods, such as low bid, UDOT utilizes this method in design-build projects. The selection process of the fixed budget-best value follows a similar process of best value design-build procurement as follows (UDOT 2016):
A. Develop and approve projects goals: The process begins with understanding of the major factors impacting the project based on environmental study information or other known issues. During this process, the project team and Regional Leadership should clearly define the project goals based on scope, schedule, budget, and impacts to the public. Based on the project goals, the project team and Regional Leadership apply relative weights to goals and develop evaluation criterial for each scored goal. The project goals and evaluation criteria should be refined by the Selection Committee throughout project development. Finally, the project team and Regional Leadership request approval of the project goals and evaluation criteria from the Selection Committee.
B. Receive and Evaluate Proposals: Once the project goals and evaluation criteria are approved by the Selection Committee, the proposals will be received and evaluated by the Selection Committee.
75

a. Analysis Committee identifies the added values, risks, strengths, and weaknesses of proposers
b. Evaluation Committee offers one-one meetings with each proposer c. Selection Committee meets with the Evaluation Committee early in the
process to discuss the project and agrees on purpose and objective of the project. Next, through the review of blinded technical and blinded price proposals, the Selection Committee determines overall best value selection and provides a written and blinded justification of best value selection. To measure quantitative and qualitative benefits of proposals, UDOT uses evaluation adjectives, such as "HIGH", "MEDIUM", and "LOW", which indicate the relative significance to UDOT. The example of evaluation factors for project definition is shown in Table 3-1.
76

Table 3-1 Example of Evaluation Factors and Category

Evaluation Factor
HIGH
MEDIUM LOW

Evaluation Category
Number of I-15 lane and shoulder miles added or improved, by type and level of improvement. Number of interchanges reconstructed or improved and level of improvement Operational metrics of mainline, at and between interchanges Operational metrics of mainline transitions to existing facilities Level of improvement to regional mobility associated with mainline improvements using the
results from the transportation demand management (TDM), as listed below: o Vehicle miles traveled (VMT) o Vehicle hours traveled (VHT) o Average speed o Total delay o User costs o Percent VMT with volume-to-capacity (V/C) greater than or equal to 1 (for all links excluding centroid connectors)
Level of improvement of the interchange operations using the results from the microscopic simulation software (VISSIM) models for traffic flow as listed below: o Delay o Speed o Density o Travel time index o Queuing
Other operational improvements including the following: o Number and nature of decision points o Length of weave areas o Width and location of shoulders and refuge areas o Number of bicycle/pedestrian conflicts with traffic o Provision of clear zones
Number of intersections improved and level of improvement For areas between American Fork Main Street and Provo Center Street Operational metrics in cross street transitions to existing facilities Extent and functionality of non-motorized improvements

3.2.1.1. An Example of Fixed Budget-Best Value from UDOT

The first fixed budget-best value project in UDOT was the 24-mile I-15 Corridor Expansion (I-15 CORE) project in 2008. The major challenge of this project was the budget

77

cut from $2.6 billion to $1.7 billion. Using a fixed budget-best value approach and proactive risk management, UDOT was able to deliver all the basic configuration scope with additional elements while spending $1.1 billion, which was less than the State legislature approved budget. I-15 CORE project is an exceptionally successful example of a fixed-price-best value procurement method. All proposers submitted more scope with innovative solutions for design and construction and did not exceed the approved budget. The winning proposal provided fastest schedule, more lane miles, fewer lane closures, and additional inch of pavement that has longer life and lower life cycle costs (UDOT 2013 ;WSDOT 2013).
The evaluation criteria for I-15 CORE were: Technical, must-have requirements Pass/Fail elements Project goals and values
The scores for three categories were given as the following: 60% project definition (scope) 20% maintenance of traffic 20% schedule
Overall, UDOT verified that a fixed budget-best value approach is an effective contracting strategy in maximizing the amount of work while staying within the approved budget.
78

3.2.2. Colorado Department of Transportation (CDOT) The Colorado Department of Transportation (CDOT) also utilizes a fixed budget-best value (or fixed price-best proposal) procurement method when the agency has a budget constraint and wants to maximize the scope of work. This method provides proposers with flexibility in selecting the technical approach and scope for a project within the defined budget. In addition to the basic configurations, CDOT usually defines additional scope elements, known as "Additional Requested Elements (AREs)", so that proposers can have options to select. As more AREs are included in proposals while staying within the budget, the proposers will obtain the higher evaluation score. To achieve the project goal, the agency should carefully define the budget and the AREs for a project. The selection process for a fixed budget-best value approach is as follows (CDOT 2016):
A. Develop Evaluation Procedure: the process begins with determining the project goals. CDOT should determine the project goals by using best value parameters including cost, time, scope, technical design consideration, and construction operation consideration (such as Maintenance of Traffic (MOT) and Public Involvement (PI) parameters). The best value scoring parameters are shown in Table 3-2.
79

Table 3-2 Relating Project Goals and Values to Best Value Scoring Parameters

Project Goals

Maximize operational capacity





Maximize use of available funds





Manage impacts during construction,

or Minimize inconvenience to the

traveling public, or Minimize

inconvenience to the stakeholders



Complete the project on or before a set

date



Provide a high-quality project





Safety of the public and workers



Maximize project durability or



Minimize life cycle costs of project





Possible Best Value Parameters
Project technical approach and commitments AREs AREs Additional Proposal scope commitments MOT approach and commitments PI approach and commitments Time of completion Duration of construction impacts Time of completion Time to obtain key schedule milestones Quality Management Plan approach and commitments Technical approach and commitments Safety Management Plan approach and commitments Maintenance Level of Service commitments Low-maintenance structures Low-maintenance pavement Other low-maintenance designs

B. Receive and Evaluate Proposals: each evaluator reviews and assesses individual statements of qualifications (SOQs)/Proposals using the overall criteria set and records observations using provided evaluation forms. a. Each evaluator determines an adjectival rating for each evaluator category using the adjectival evaluation and scoring guide as shown in Table 3-3. Each evaluator uses a best value evaluation formula to determine total score. Each parameter is then assigned specific scoring criteria. The maximum of the total proposal score is 100 points. Table 3-4 shows Alternative Algorithms to calculate total score.
80

b. The Evaluation Committee and technical advisors meet and discuss the

submitted SOQs/Proposals and the evaluation forms. The Evaluation

Committee then determines the final score for each proposal.

c. CDOT provides the opportunity for one-one meetings for each proposer that

requests a meeting within the allowed time period.

Table 3-3 Adjectival Evaluation and Scoring Guide

Adjective Excellent (E) Very Good (VG)
Good (G) Fair (F) Poor (P)

Description
SOQ/Proposal supports an extremely strong expectation of successful project performance if ultimately selected as the contractor. SOQ indicates significant strengths and/or a number of minor strengths and no weaknesses. Submitter provides a consistently outstanding level of quality. SOQ/Proposal indicates significant strengths and/or a number of minor strengths and no significant weaknesses. Minor weaknesses are offset by strengths. There exists a small possibility that, if ultimately selected as the contractor, the minor weaknesses could slightly adversely affect successful project performance. SOQ/Proposal indicates significant strengths and/or a number of minor strengths. Minor and significant weaknesses exist that could detract from strengths. While the weaknesses could be improved, minimized, or corrected, it is possible that if ultimately selected as the contractor, the weaknesses could adversely affect successful project performance. SOQ/Proposal indicates weaknesses, significant and minor, which are not offset by significant strengths. No significant strengths and few minor strengths exist. It is probable that if ultimately selected as the contractor, the weaknesses would adversely affect successful project performance. SOQ/Proposal indicates existence of significant weaknesses and/or minor weaknesses and no strengths. SOQ indicates a strong expectation that successful performance could not be achieved if ultimately selected as the contractor.

Percentage of Max. Score
90-100%
75-89%
51-74%
25-50% 0-24%

81

Table 3-4 CDOT Design-Build Alternative Algorithms to Determine Total Evaluation Score

Alternative Algorithm

Formula

Result

Technical Score Adjusted by Price

Total Score = Ts x (GMP/Pp)

The highest score determines the

apparent best value.

Proposal Price Score Adjusted by Total Score = Pp/Ts

The lowest score determines the

Technical Score

apparent best value.

Qualitative Technical Score plus Total Score = Ts + (Pmax x Plow/Pp)

The highest score determines the

Quantitative Price Score

apparent best value.

Qualitative Technical Score plus Total Score = Ts + [Pmax ((Pp Plow)/($ per The highest score determines the

Quantitative Price Score (based on Pt))]

apparent best value.

defined dollars per point)

Note: Ts = Technical Proposal score: the sum of all other best value scoring elements, including AREs; Pmax

= Maximum Proposal price points; Pp = Proposal price; Plow = Lowest Proposal price; $ Per Pt Factor = A defined dollar amount per point value; GMP = Guaranteed Maximum Price

3.2.2.1. An Example of Fixed Budget-Best Value from CDOT

CDOT utilized a fixed budget-best value approach in the $1.67 billion TransportationExpansion (T-REX) design-build project in 1999. The scope of this projects is to add 19 miles of double-track light rail, build 13 stations with park-n-Rides, add 13 light rain vehicles to the Regional Transportation District (RTD)'s fleet, and construct a new light rail maintenance facility in Englewood. The project goals of this project are as follows (CDOT 2003):

To minimize inconvenience to the public To meet or beat the total program budget of $1.67 billion To provide for a quality project To meet or beat the schedule to be fully operational by June 30, 2008

CDOT achieved the significant schedule and cost saving because of the innovative funding and design-build/fixed budget-best value approach. The winning proposal was selected
82

based on a best-value evaluation process by looking at technical and price proposals. The Innovative contracting strategy enabled CDOT to complete project within schedule and under the approved budget. 3.2.3. Idaho Transportation Department (ITD) The Idaho Transportation Department (ITD) also started to conduct an experiment with a fixed budget-best value (or fixed Price-best design) approach under the provision, Special Experimental Project No. 14 (SEP 14). ITD uses this contracting strategy in a design-build delivery method to yield a greater amount of work than the low-bid method and not an additional element of work. Thus, ITD selected a proposer who submits maximum scope or quantity of work within the approved budget. The selection process of the ITD's fixed budget- best value approach is as follows (ITD 2014):
A. Develop Evaluation Procedure: the process begins with defining the project goals for the project. Next, the project team needs to develop project scope, estimated cost, and maximum time allowed for the project. Based on the project goals and other information, the evaluation criteria and process need to be developed.
B. Receive and Evaluate Proposals: proposers submit technical and price proposals concurrently. ITD should keep price proposal confidently until technical proposals have been evaluated, scored, and reviewed by higher levels. a. The Evaluation Committee evaluates technical and price proposal by using Pass/Fail and Scored Criteria. Pass/Fail Criteria include formatting, executive summary, legal, and financial aspects of proposals, as well as 83

participant experience. Next, Score Criteria consist of organizational structure, project management, maintenance of traffic, and project-specific technical and quality factors (i.e., design and construction qualifications, innovation, design and construction quality, and time of completion). b. The Selection Committee discusses and reviews the evaluation of technical and price proposal with the Evaluation Committee and documents the results of the evaluation. c. The Contracting Officer approves the evaluation of the technical and price proposal and summary of scores and feedback from evaluators. 3.2.3.1. Examples of Fixed Budget-Best Value from ITD
ITD tried fixed budget-best value with several project types (i.e., bridge deck preservation, resurfacing, and seal coating projects). Table 3-5 provides examples of project types in State of Idaho that the fixed budget-best value procurement method has been utilized in. For example, in 2010, ITD used fixed budget-best value in a bridge deck preservation project. ITD required the bidders to determine the total number of square yards of deck preservation that they could accomplish for the fixed budget of $700,000. ITD selected the bidder who submitted bid with the largest square yardage of 27,641 squad yards. In 2015, ITD had a fixed budget of $651,500 for the roadway resurfacing projects between MP36.783 and MP48.869 in Idaho. The contractors were required to bid a tonnage of crushed aggregate base that is excavated or blasted from the source, crushed, placed, and compacted. The range of the tonnage was between 14,115 and 41,448 tons. ITD procured the contract to the bidder who submitted the biggest tonnage, 41,448 tons. In 2016, ITD also used the fixed budget-best value approach for sealcoating projects in District 4 of the
84

State of Idaho. The bidders bid how many square yards they could sealcoat for the fixed budget of $2,948,000. The range of the square yards is between 1,433,897 and 1,616,228.07 square yards. The winning bid was the bidder who submitted the bid with the 1,616,228.07 square yards. Using a fixed budget-best value, ITD achieves equal to or better than the base concept.

Table 3-5 Example of Fixed Budget-Best Value Projects in State of Idaho

Construction Year
2010

Budget $700,000

2015

$651,500

2016

$2,948,000

Work Type Bridge Deck Preservation
Projects Roadway Resurfacing Projects
Sealcoating Projects

Winning Bid
The largest square yardage (27,641 sq. yd.)
the biggest tonnage (41,448 tons) The largest square yardage (1,616,228.07 sq. yd.)

3.2.4. Michigan Department of Transportation (MDOT) The Michigan Department of Transportation (MDOT) uses a fixed budget-best value (or fixed price-variable scope) to maximize the amount of work within a maximum budget. Thus, the contractor providing the most scope/work for the established budget is awarded the contract. MDOT classifies projects into three types that can be procured by a fixed budget-best value approach (MDOT 2015):
Type 1: projects receive bids by the units of work that can be completed for a State fixed price. The selected contractor is the bidder that proposed the most units of

85

work for the given fixed price. Type 1 has been used for HMA crack seal, chip seal, and fog seal projects, bid by the lane mile. Type 2: projects receive bid by the units of work that can be completed for a maximum price. Contractors bid units of work, and may also bid a price for that work which is below the maximum price. The selected contractor is first determined by the bidder that proposes the most units of work, for their determined maximum price. If two or more contractors propose the same amount of work, then the successful bidder is determined by which of those contractors proposed the lowest maximum price. Type 2 has been used for bridge deck epoxy overlay work, bid by the square yard. Type 3: projects receive bids through a traditional low bid process. The contractor provides unit prices for pay items provided in the schedule of items. The selected contractor is determined by the lowest submitted bid. The project is awarded at the low bid price. With Type 1, the proposer submits the maximum amount work while spending all authorized funding. On the other hand, the Type 2 projects allow MDOT and proposers to adjust the maximum price depending on the maximum amount of work submitted by proposers. The Type 3 project will go through normal low bid process. It allows to add work until final construction costs equal to the engineer's estimate (Youngs 2013).
86

MDOT considers a combination of technical and price factors to select the winning bid in a fixed budget-best value approach. The selection process for a fixed budget-best value method is as follows:
A. Develop and approve projects goals: the project manager prepares a proposal evaluation plan that details the process and criteria to be used during technical proposal evaluation. The Selection Team develops scoring criteria for the technical portion of the evaluation.
B. Receive and Evaluate Proposals: a. The proposals will be reviewed by a Selection team consisting of the project manager, staff from the region/Transportation Service Center, the Innovative Contracting Unit, The Central Selection Review Team (CSRT), as well as other technical experts. b. The project manager and deputy project manager (DPM) review the technical proposals by using the Pass/Fail criteria in the RFP and score the proposals. The project manager provides the Selection Team with the submitted proposals and the results of the technical proposal review. c. The Selection Team reviews the technical proposal and determines the score for each proposal with justification. d. The project manager provides CSRT with the information of final review and approval. The final results are posted after approving the scores.
3.2.4.1. Examples of Fixed Budget-Best Value from MDOT
MDOT also utilized the fixed budget-best value contracting strategy in several projects to achieve the maximum amount of work within the fixed budget for the project. In 2012,
87

MDOT used this innovative approach for crack sealing work in Hillsdale, Ingham, Jackson, and Lenawee counties in the state of Michigan. The scope of this project included a maximum of 103.78 miles of hot mix asphalt crack treatment and Overband crack filing on 15 segments of various roadways in Michigan. Three bidders submitted the bids with the maximum number of roadbed miles of work that could be completed for the established project budget of $387,000. To evaluate proposals, MDOT had two evaluation criteria: past performance and maximum amount of work. MDOT awarded the contract to the bidder who submitted the maximum length of 74.43 roadbed miles, which was longer than the Department's estimate of 70.62 miles (MDOT 2012).
88

3.4. Summary of the Recommended Best Practices for Fixed Budget-Best Value Procurement Method
State DOTs experience crucial funding limitations for delivering much-needed construction and rehabilitation projects that are necessary for maintaining the quality of transportation infrastructure systems. Innovative contracting strategies, such as a fixed budget-best value procurement method, can help State DOTs complete a project within an established budget. This chapter reviewed the current state of practices in fixed budgetbest value procurement method and identified best practices in utilization of this innovative contracting method. The following recommendations are found out to be effective for enhancing the practice of delivering projects using this innovative contracting strategy.
State DOTs may consider a fixed budget-best value procurement approach when the full project scope exceeds the baseline cost estimate for the project. For a fixed budget-best value procurement approach, the agency should define the basic configuration scope and should allow the proposers to include the maximum amount of work or additional scope elements in their proposals while staying within the fixed budget. For example, several State DOTs, including Utah, Idaho, and Michigan DOTs typically let the proposers submit the maximum amount of work that they can achieve with the established budget. State DOTs then procure the contract to the proposer who proposes the maximum amount of work within the established budget. In contrast, Colorado DOT defines specific additional scope elements, known as "Additional Requested Elements (AREs)", in the RFP to enable
89

the proposers to select additional scope elements beyond the base scope of a project within the fixed budget. The agency evaluates whether the proposers have incorporated the additional elements in accordance with project design requirements.
A fixed budget-best value approach can be utilized in several project types, such as corridor expansion, bridge deck preservation, and seal coating projects. State DOTs should clearly define additional scope elements beyond the base scope for each project type. For example, Utah DOT used the fixed budget-best value approach in the I-15 Corridor Expansion (I-15 Core) project, which is addable for the scope of work in terms of lane miles. UDOT selected the proposal that provided more lane miles for the corridor expansion. Idaho DOT used this contracting strategy in bridge deck preservation and sealcoating projects. Idaho DOT procured the contract to the bidder who submitted the largest square yardage for the preservation.
State DOTs should clearly establish the evaluation criteria to select the proposers for a project. Since the price is fixed for all proposers and this approach allows higher flexibility in proposing design and construction solutions, the agency should establish rigorous evaluation criteria (e.g., cost, time, and design alternatives) and the weights for the criteria to evaluate the proposals based on the project goals. For example, Utah DOT uses evaluation adjectives (i.e., "High", "Medium", and "Low") to measure quantitative and qualitative benefits of the submitted proposals for a project. "High" indicates that the proposal covers more scope elements for a project with respect to quality and quantity of a
90

project. Utah DOT assigns higher points to a proposal with a "High" category. Colorado DOT establishes best value scoring parameters based on project goals. For instance, if the project goal is to maximize operational capacity, CDOT uses project technical approach and commitments, as well as AREs as primary parameters to evaluate the proposals for a project.
The fixed budget-best value approach maximizes improvements within the defined budget and provides incentives to proposers to utilize the full budget. This approach increases competition and exploits the budget as much as possible that can result in maximum improvements for the project. The fixed budget-best value has several advantages and disadvantages. The major advantage of this approach is that it can be a good tool for controlling costs and keeping a project within budget. However, the agencies may get less work done than originally planned if the budget is too tight. In addition, this approach may require more time for evaluating proposal and have challenges in selecting the contractor if selection criteria are not clearly defined and defendable. Several advantages and disadvantages of the fixed budget-best value procurement are summarized as follows (Scott et al. 2006; WSDOT 2013):
Advantages: 1. Allows the agency to achieve the maximized amount of work within an established budget.
91

2. Provides proposers with significant flexibility in structuring their proposals; i.e., whether to only submit the basic configuration or to include various levels of additional scope elements.
3. Allows the agency to communicate its desired additional scope through the outcomes it values; for example, additional mainline capacity, direct connections between certain roadways or longer pavement life.
4. Allows proposers to develop highest-value, creative solution for fixed construction budget.
5. Fosters competition and innovation as the value-based approach.
Disadvantages: 1. Provides the agency with challenges in determining and splitting work for the bidding purpose that accurately corresponds to the established budget. 2. Requires the agency to define exactly how it will evaluate additional scope beyond the high/low value definition otherwise provided during meetings with the contractor. 3. Limits the flexibility and range of what is actually proposed due to difficulty in defining more specific evaluation criteria. 4. Limits creativity of the proposers to respond and use innovative approaches that achieve desired goals if the agency specifically prescribes what additional elements it desires beyond the basic configuration. 92

CHAPTER 4 STATISTICAL ANLYSIS FOR EXPLAINING VARIATIONS IN SUBMITTED BID PRICES FOR ASPHALT LINE ITEM IN HIGHWAY
PROJECTS
4.1. Introduction
Highway construction costs are subject to significant variation from project initiation to project completion. Variation in construction cost disturbs transportation agencies in making investment decisions and preparing accurate engineer's cost estimates for their projects. The underestimation of project costs can lead to cost overrun, financial problem, and project delay or cancellation (Peng 2006). The overestimation of project costs results in inefficient budget allocation of public funds that could be used on other needed projects (FHWA 2015). State Departments of Transportation (State DOTs) may face credibility issues with the public if cost estimation problems remain unresolved. Therefore, construction cost variation should be treated properly for delivering projects that have been programmed and committed. Variation in construction costs depends on several factors, such as the current volume of construction in the project location, inflation rate, and the cost of material, labor, and equipment. The impact of these factors varies based on project characteristics (e.g., type of projects, complexity of projects, geographical location of projects, and project size) and its
93

surrounding environment (e.g., macroeconomic, market conditions, and competitive bidding environment). The combined effects of these factors impose considerable uncertainty in pricing construction costs for State DOTs. According to the American Association of State Highway and Transportation Officials Technical Committee on Cost Estimating (TCCE) publication (2011), macroeconomic variables, such as prices of materials and labor, and market condition factors, such as competition and contractors' work volume, are key inputs for preparing cost estimates and structuring letting strategies. According to the Federal Highway Administration (FHWA) (2015), market conditions are major drivers of over- and under-estimation of highway construction costs. This report also noted that most State DOTs have low capability to capture volatile market conditions in their cost estimating process. There is no consistent approach to deal with inflation and escalation factors in preparing costs estimates (Actis 2010). Developing a systematic method to deal with uncertainty in market conditions represents a great challenge in managing variation in construction costs. Major critical factors in regard to not only project characteristics, but also macroeconomic and market conditions, should be identified and properly considered in the cost estimation of construction projects. Several studies identified factors affecting variation in construction costs. Chua and Li (2000) focused on market condition factors affecting contractors' decisions in unit price bidding. Through conducting a survey of 153 top contractors in Singapore, the authors identified four major factors, competition, risk, company's position in bidding, and need for work, as influential factors on bid prices. Wilmot and Cheng (2003) conducted an
94

empirical analysis for testing impacts of project-specific factors and market condition indicators on submitted bid prices to the Louisiana DOT. They identified factors that affect the price of 5 pay items, embankment, concrete pavement, asphalt pavement, reinforcing steel, and structural concrete, through studying 2,827 highway and bridge contracts in the State of Louisiana. The authors used several independent variables, such as project location, quantity of pay items, and labor costs to explain the variation in the submitted bid price for 5 pay items. Through conducting multivariate regression analysis, the study found that the most influential factors are the price of the resources (i.e., materials, labor, and equipment) and the quantity of the pay items. Moreover, contract size, duration, location, and the quarter in which the contract is let were found to have a significant impact on the price of the asphalt pavement pay item. Shrestha and Pradhananga (2010) examined the effect of competitive bidding on variations in bid prices using 435 bids on 113 public street projects in Clark County, Nevada. Through a correlation analysis, the authors concluded that there is a significant negative correlation between the number of bidders and the submitted bid prices. Another study carried out by Shrestha et al. (2014) analyzed the bid data of 151 road projects conducted in Clark County, Nevada, to examine the relationship between the unit price bids and the quantity of the unit item. The results of correlation analysis show that an increase in the quantities lowers the unit price bids. Hegazy and Ayed (1998) identified factors affecting construction costs by using 18 bids submitted by construction contractors. This study found that season, location, type of
95

project, contract duration, and contract size significantly impact construction costs. Damnjanovic and Zhou (2009) examined the impact of the crude oil prices on excavation bid item of 5,180 highway construction projects let in the State of Texas. The authors identified that the crude oil price has the positive effect on the bid prices. Ilbeigi et al. (2015) analyzed submitted bid prices in the State of Georgia to explain variations in construction cost. The authors identified that quantity of the line item, total contract price of the project, and asphalt cement price index are influential factors that explained variations in bid prices submitted to Georgia DOT for asphalt line items. Lastly, Ashuri et al. (2012) conducted Granger causality tests to capture and predict construction cost variations using construction cost index (CCI) published by the Engineering News-Record (ENR). The authors concluded that economic conditions including consumer price index, producer price index, money supply, gross domestic product (GDP), crude oil prices, and construction market conditions including building permits, housing starts, and employment level in construction are the leading indicators of CCI and can help predict future CCI trends. In this study, variation of submitted bid prices for major asphalt line items were investigated. The major objective was to assess the effects of several potential variables on explaining the variation of submitted bid prices. Potential factors were grouped into 4 major categories: (1) Project characteristics; (2) Construction market conditions; (3) Macroeconomic conditions; and (4) Oil market conditions. Historical bid data submitted
96

for major asphalt line items from resurfacing and widening projects in the State of Georgia were used to examine the relationships in this study.
4.2. Methodology
The primary goal of this chapter is to model the variation in the unit prices of asphalt line items (i.e., hot mix recycled asphalt concrete) by using several factors that may have potential to explain the variation. The potential factors are related to project-specific variables, and macroeconomic, construction market, and oil market conditions. This chapter utilizes multiple regression analysis to develop an explanatory model for describing variations of submitted bid prices. Regression analysis is used to establish the nature of the relationship between the unit prices and the potential factors and explain the variation in the submitted bid prices with combination of significant factors. The following steps are followed in the methodology of multiple linear regression (MLR) analysis:
1. Inspect data for identifying outliers 2. Conduct pairwise correlation between submitted bid price as the dependent variable,
on one hand, and each of the potential explanatory variables, on the other hand, to assess if linear correlation (in some cases variable transformation (e.g., logarithmic)) should be performed if the variable transformation better reflects the nature of relationship between the explanatory variable and the submitted bid price 3. Develop a multiple regression model to describe variations in submitted bid prices using the information embedded in the potential explanatory variables
97

4. Interpret the results of regression modeling 5. Examine residual plots to check error variance assumptions in regression modeling

4.2.1. Modeling the Variations of the Submitted Bid Prices 4.2.1.1. Inspect Data for Identifying Outliers

Since large data are typically exposed to uncertainty regarding calculation, writing, or

coding errors, abnormal and suspected outliers should be diagnosed and removed before

developing a reliable regression model. The detected outliers should be omitted from

further consideration in model development. z-scores are calculated for all observations of

submitted bid prices in the original dataset as follows.



=



-



= 1,2, ... ,

where yi is the ith observation in the dataset of submitted bid prices, is the mean value of

all observations of submitted bid prices, and is the standard deviation of submitted bid

prices. If the absolute value of is greater than 2.576 (representing 99% confidence level), the ith submitted bid price will be considered as an outlier and will be removed from

further consideration.

4.2.1.2. Conduct Pairwise Correlation between Submitted Bid Price and Potential Explanatory Variables

Pairwise correlation is performed between submitted bid price as the dependent variable, on one hand, and each of the potential explanatory variables, on the other hand, to assess the degree of linear correlation. This initial correlation assessment is used to examine the significance of association between the submitted bid price and any of the explanatory

98

variables. Scatter plot is also used to assess the nature of the relationship. In addition to linear correlation, other forms of correlation, such as quadratic, cubic, logarithm, exponential, and power relationships might exist between the submitted bid price and the potential explanatory variable. Scatter plots are useful methods to detect these other forms of relationships. Whenever appropriate, variable transformation (e.g., logarithmic) is conducted to better reflect the nature of relationship between the explanatory variable and the submitted bid price. The transformed variable is used in regression analysis. 4.2.1.3. Develop a Multiple Regression Model to Describe Variations in Submitted Bid Prices Ordinary least squares (OLS) regression modeling is used to explain variations in submitted bid prices based on a combination of the explanatory variables. Analysis of Variance (ANOVA) is conducted to determine whether the developed regression model is statistically significant (i.e., at least one of the coefficients of the identified explanatory variables is not zero in the regression model). The fitness of the model is examined by calculating the adjusted R-squared of the model. The higher the R-squared, the greater the model is in explaining the variation of submitted bid prices. Variance inflation factor (VIF) is calculated for each of the identified explanatory variables, in order to assess multicollinearity issues in the regression model (Montgomery et al. 2015). Multicollinearity can result in misinterpretation of the regression modeling results. The significance of the model coefficients is examined using the calculated p-value for each explanatory variable in the model.
99

4.2.1.4. Interpret the Results of Regression Modeling

The effects of potential explanatory variables on the submitted bid prices are examined

using the calculated p-value in the developed regression model. Significant explanatory

variables are identified at the significance level of = 5%. The sign and the magnitude of

the coefficients of the significant variables show the direction of the effects of the

significant explanatory variables on the submitted bid price. The results of regression

analysis are used to examine whether and how a potential variable has the power to explain

variations in submitted bid prices.

Explanatory variables have different measurement units. Thus, there is a limitation in

comparing the relative impacts of explanatory variables on the dependent variable. The

following equation is used to convert all continuous explanatory variables to standardized

variables that have the same scale with the expected value of 0 and variance of 1

(Washington et al. 2010):



=

- ()

= 1,2, ... ,

where is the standardized form of the ith input variable, is the ith input variable,

is the average value of , and () is the sample standard deviation of . MLR

analysis using the standardized variables produces the beta/standardized coefficients for

the explanatory variables. The beta coefficient of each explanatory variable is used to

determine the relative importance of the explanatory variables in the regression model. The

higher the absolute value of the beta coefficient, the stronger the effect of the respective

explanatory variable is on the submitted bid prices.

100

Note that categorical variables will not be considered for transformation. Standardization is not applicable to categorical variables to determine the relative significance of these factors, compared to other non-categorical factors, on submitted bid prices. This issue can be explained as follows. First, categorical variables cannot be increased by a standard deviation because they are the dichotomous variables, coded as 0 and 1. Estimated coefficients of categorical variables show difference in average level of submitted bid prices between two categories (Jacoby 2005). Standardizing categorical variables prohibits us from interpreting the difference between the impacts of different categorical variables. 4.2.1.5. Examining Residual Plots to Check Error Variance Assumptions Once a regression model is developed, the regression assumptions should be checked. The following assumptions should be examined (Field 2009):
Independent errors: the residual terms should not be correlated. If the assumption of independence is violated, the confidence intervals and significant tests of the model will be invalid. The scatterplot of standardized residuals against predicted values is useful for detecting independence.
Homoscedasticity: the variance of the residual term should be constant and evenly dispersed. If the assumption of homoscedasticity is violated, the confidence interval and significance tests are not valid. The scatterplot of standardized residuals against predicted values is also useful for detecting heteroscedasticity.
101

Normality: the residuals in the model should be random, normally distributed variables with a mean of 0. The histogram and the normal probabilityprobability (P-P) plot of regression standardized residuals are useful for testing normality.
4.3. Dataset Development
4.3.1. Data Compilation Process Data compilation process includes extraction of data corresponding to the research problem, critical evaluation of data sources, assessment of data quality, and compilation of data. Figure 4-1 depicts the compilation process of data. Raw historical bid data are retrieved from the BidTabs database of Oman Systems. Historical bid data include information about unique project identification number (project ID), unit price bids for main asphalt line items, volume of asphalt line items, total bid price (contract amount), number of bidders, and winning bids. Several other software systems used by the Georgia DOT for contract administration are also utilized as data sources to: (a) cross-validate the quality of the collected data from the BidTabs; and (b) retrieve further information about project characteristic. Unit price bids retrieved from the BidTabs database are rechecked for accuracy with another sources of bid data, Bid Express , Construction Administration System (CAS), SiteManager (SM), and the GDOT transportation project information (TransPi). Bid Express is an online system containing information about past proposals, bid prices, and bidders of the Georgia highway projects. CAS and SM are software systems used by GDOT for contract administration. In addition, information about the geographical
102

location of the project including the map of the project is accessed via the GDOT TransPi. Project-specific information collected from the above multiple sources is stored in a centralized database.

Bid-Tabs Database
Historical Unit Price Bids
Quantity of the Line Items
Project Locations
Total Bid Prices
Source: Oman Systems

Construction Administration System (CAS) and SiteManager
(SM)
History of Projects
Total Dollars of the Projects
Total Quantities for Each Item
Contract Time Charged to Date
Source: GDOT

ID: Job Number

ID: Job Number

All Contracts Search Report
History of Projects Construction
Contractors Project Locations Contract Time
Charged to Date
Source: GDOT
ID: Job Number ID: Project ID

Bid Express
History of Projects
Project Locations Construction
Contractors

Transportation Project Information (TRANSPI)
History of Projects
Project Locations
Project Types

Total Bid Prices
Source: Bid Express

Construction Contractors Contract Time Charged to Date

Total Bid Prices

ID: Job Number

Project Documents
Source: GDOT TransPI
ID: Project ID

Data Compilation

Database

Engineering News-Records (ENR) Georgia Department of Transportation
(GDOT) U.S. Bureau of Labor Statistics (BLS) U.S. Census Bureau U.S. Inflation Calculator Federal Highway Administration (FHWA) U.S. Energy Information Administration

Data Evaluation
The American Institute of Architects The National Association of Home
Builders (NAHB) Bureau of Economic Analysis InflationData.com S&P Dow Jones Indices Fails Management Institute (FMI) Turner Construction Company

A Pool of Input Variables
Figure 4-1 Data Compilation Process for Multiple Regression Analysis 4.3.2. Submitted Unit Prices for Asphalt Line Items Hot mix recycled asphaltic concrete, as one of asphalt line items, is the most common pavement material used by State DOTs in the United States (Kandhal et al. 1995).
103

Submitted bid prices for asphalt line items from 1,424 projects let in the State of Georgia
between 2008 and 2015 are used in this research. Descriptive statistics of submitted unit
prices are described in Table 4-1. Table 4-1 Descriptive Statistics of Submitted Unit Prices for Hot Mix Recycle Asphaltic Concrete

Number of Projects Mean ($/Ton) Median ($/Ton) Min ($/Ton) Max ($/Ton)

Total
1,424 69.02 67.75 40.25 354.53

Resurfacing
1,310 68.43 67.30 40.25 354.53

4.3.3. Factors affecting Variation in the Submitted Bid Prices

Widening
114 75.74 73.50 51.97 155.37

Table 4-2 provides a summary of explanatory variables that are identified with potential impacts on submitted bid prices. The identified potential explanatory variables are categorized into 4 groups: (1) Project characteristics; (2) Construction market conditions; (3) Macroeconomic conditions; and (4) Oil market conditions.

104

Construction Market Conditions (30 Variables)

Project Characteristics (11 Variables)

Category

Table 4-2 Summary of Potential Explanatory Variables

Factors
Project Duration Quantity of the Bid Item Terrain of the Project District of the Project Total Bid Price Project Length Number of Nearby Asphalt Plants (within 50 miles) Hauling Distance between Asphalt Plant and Project Location Hauling Distance between Quarry and Asphalt Plant Number of Bidders Number of Pay Items Total Asphalt Volume of Resurfacing and Widening Projects Awarded in the Current Month at the Level of the County Total Number of Resurfacing and Widening Projects Awarded in the Current Month at the Level of the County Total Number of Projects Awarded in the Current Month at the Level of the State of Georgia Total Dollar Value of Projects Awarded in the Current Month at the Level of the State of Georgia Total Asphalt Volume of Projects Awarded in the Current Month at the Level of the State of Georgia Architecture Billings Index Building Permits for New Residential Construction ENR Building Cost Index ENR Common Labor Index ENR Construction Cost Index

Unit
Day Ton Terrain Types Districts
$ Miles No. Miles Miles No. No.

Region/ Industry
Project Level Project Level Project Level Project Level Project Level Project Level Project Level Project Level Project Level Project Level Project Level

$

County/Construction

No.

County/Construction

No.

Georgia/Construction

$

Georgia/Construction

$

Georgia/Construction

IDX

South/Construction

No.

South/Construction

IDX

Atlanta/Construction

IDX

National/Construction

IDX

Atlanta/Construction

Source
Bid-Tabs database of Oman Systems & GDOT TransPi Bid-Tabs database of Oman Systems GDOT (2009) Bid-Tabs database of Oman Systems Bid-Tabs database of Oman Systems Bid Express GDOT TransPi & GDOT Office of Materials and Testing GDOT TransPi & GDOT Office of Materials and Testing GDOT TransPi & GDOT Office of Materials and Testing Bid-Tabs database of Oman Systems Bid Express
Bid-Tabs database of Oman Systems
Bid-Tabs database of Oman Systems
Bid Express
Bid Express
GDOT Item Mean Summary
The American Institute of Architects U.S. Census Bureau Engineering News-Record (ENR) Engineering News-Record (ENR) Engineering News-Record (ENR)

105

Construction Market Conditions (Cont'd)

Category

Factors
ENR Material Price Index ENR Skilled Labor Index Equipment Operator Wages, Paving, Mean Hourly Wage, Georgia Georgia Asphalt Cement Price Index Job Opening and Labor Turnover Index (Hires) Housing Market Index National Highway Construction Cost Index Producer Price Index (Construction machinery manufacturing) Producer Price Index (Construction sand and gravel mining) Turner Construction Cost Index Value of Construction Put in Place (Pavement) Value of Construction Put in Place (All construction) FMI Nonresidential Construction Index Labor Productivity Number of Establishments in Private Construction Industry Gross Domestic Product (GDP) of the Georgia Construction Industry 12 Month Percent Change of Georgia Asphalt Cement Price Index 12 Month Percent Change of Gross Domestic Product (GDP) of the Georgia Construction Industry 12 Month Percent Change of Job Opening and Labor Turnover Index (Hires) 12 Month Percent Change of Value of Construction Put in Place (Pavement)

Unit
IDX IDX
$ $/Ton Thousands IDX IDX IDX IDX IDX Thousands of $ Thousands of $ IDX IDX No. Millions of $
%

Region/ Industry
National/Construction National/Construction Georgia/Construction Georgia/Construction National/Construction South/Construction National/Construction National/Construction National/Construction National/Construction South/Construction Georgia/Construction National/Construction National/Construction County/Construction Georgia/Construction Georgia/Construction

%

Georgia/Construction

Source
Engineering News-Record (ENR) Engineering News-Record (ENR) U.S. Bureau of Labor Statistics GDOT U.S. Bureau of Labor Statistics National Association of Home Builders/ Wells Fargo Federal Highway Administration U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics Turner Construction Company U.S. Census Bureau U.S. Census Bureau Fails Management Institute (FMI) U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics Bureau of Economic Analysis GDOT
Bureau of Economic Analysis

%

National/Construction U.S. Bureau of Labor Statistics

%

Georgia/Construction U.S. Census Bureau

106

Macroeconomic Conditions (10 Variables)

Category

Factors
Dow Jones Industrial Average Inflation Rate Average Weekly Wage (all industry) Consumer Price Index (south) Producer Price Index (Gasoline products) Producer Price Index (Steel mill products) Producer Price Index (No. 2 diesel fuel products) Producer Price Index (Crude petroleum products) Unemployment 12 Month Percent Change of Unemployment
Crude Oil Price

Unit
IDX. % IDX IDX IDX IDX IDX IDX No. %
$/Barrel

Oil Market Conditions (4
Variables)

Diesel Retail Prices

$/Gallon

Georgia Fuel Price Index

$/Gallon

12 Month Percent Change of Georgia Fuel Price Index

%

Note: No. = number; IDX. = index; S.Y. = square yard of surface area; %= Percentage

Region/ Industry
National/Industry National/Industry County/Industry South/Industry National/Industry National/Industry National/Industry National/Industry County/Industry County/Industry
National/Industry
South/Industry
Georgia/Industry
Georgia/Industry

Source
S&P Dow Jones Indices USInflationcalculator.com U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics U.S. Bureau of Labor Statistics InflationData.com
U.S. Energy Information Administration
GDOT
GDOT

107

4.3.3.1. Project Characteristics Project characteristics are unique features of each project, such as asphalt quantity of the project, duration, total bid price, and project bidding date. Collectively, these unique features represent project characteristics. Two categorical variables are used to classify projects: (a) terrain type; and (b) Georgia DOT district.
1. Project Duration: It is period between 2 dates: notice to proceed date and completion date.
2. Quantity of the Bid Item: It is volume of the asphalt line items. 3. Terrain of the Project: It is the geographical feature of the project location. Georgia
has four types of terrain, rolling, flat, mountainous, and coastal. Figure 4-2 shows the terrain types in the Georgia map (GDOT 2009). These terrain types are used as categorical variables in the regression analysis. Descriptive statistics of the submitted bid prices based on terrain types are described in Table 4-3. Coastal terrain in Georgia has relatively higher average unit prices compared to other terrain types.

Table 4-3 Descriptive Statistics of the Submitted Bid Prices based on Georgia Terrain Types

Terrain
Number of Projects Mean ($/Ton) Median ($/Ton) Min. ($/Ton) Max. ($/Ton)

Rolling
999 67.09 66.10 40.25 170.85

Flat
281 72.72 72.00 52.37 116.86
108

Mountainous
70 68.86 68.02 51.59 125.00

Coastal
74 81.21 77.31 56.87 354.53

Figure 4-2 Georgia Terrain Map 4. Districts of the Project: GDOT has 7 district offices throughout the State of Georgia
as shown in Figure 4-3 (GDOT 2014). These districts are used as categorical variables in the regression modelling. In addition, descriptive statistics of the submitted bid prices in different districts are summarized in Table 4-4. Districts 4 and 5 have relatively higher average unit prices for asphalt line items compared to other Districts.
109

Figure 4-3 Georgia District Map

Table 4-4 Descriptive Statistics of the Submitted Bid Prices based on Seven Districts

Districts District 1 District 2 District 3 District 4 District 5 District 6 District 7

Number of Projects
Mean ($/Ton) Median ($/Ton)
Min. ($/Ton) Max. ($/Ton)

219
65.45 64.00 47.41 170.85

241
69.68 68.25 50.66 111.63

242
64.16 36.73 42.00 101.02

237
71.35 70.46 50.00 116.86

195
75.80 74.00 52.42 354.53

126
69.50 68.17 50.24 110.18

164
68.18 67.13 40.25 155.37

110

5. Total Bid Price: It is the lowest total bid price that is submitted by contractors on

the project. Total bid price represents the size of the project.

6. Project Length: It is an approximate length of the project. Project length is another

indicator of the size of the project.

7. Number of Nearby Asphalt Plants (within 50 miles): It is the number of asphalt

plants within 50-mile radius of the center of the project. Since distance from an

asphalt plant to the paving location should not exceed 50 miles (80km) (ODOT

2016), the number of asphalt plants are counted within 50 miles from the project

location. The annual number of certified asphalt plants in Georgia is provided in

Table 4-5. In addition, Figure 4-4 shows the geographical distribution of asphalt

plants in Georgia. Table 4-5 Annual Number of Asphalt Plants in Georgia

Year Number of Certified
Asphalt Plants

2008 118

2010 113

2012 106

2014 108

2016 103

111

Figure 4-4 Locations of Active Asphalt Plants in Georgia in 2016 8. Hauling Distance between Asphalt Plant and Project Location: It is the hauling
distance between project location and the closest asphalt plant to the project location. 9. Hauling Distance between Quarry and Asphalt Plant: It is the hauling distance between the closest asphalt plant to the project location and the closest quarry to the asphalt plant. 10. Number of Bidders: It is the number of bidders submitting bids for the project. This variable is a market indicator that represents the market competitiveness for the project.
112

11. Number of Pay Items: It is the number of pay items in the contract for which the contractor submits a unit bid prices. This variable is an indicator of project complexity.
4.3.3.2. Construction Market Conditions Market conditions for the construction industry can be described by several factors (e.g., prices of construction materials and labor, and construction cost indexes). Rapidly changing market conditions represent significant challenge for public and private sectors in pricing construction cost. The following variables are considered in this study to represent construction market conditions:
1. Total Asphalt Volume of Resurfacing and Widening Projects Awarded in the Current Month at the Level of the County: It is the sum of the asphalt volume of resurfacing and widening projects awarded in each month and in the same county as of the project in the State of Georgia. This factor represents the activity level of paving construction in the area close to the project location.
2. Total Number of Resurfacing and Widening Projects Awarded in the Current Month at the Level of the County: It is the number of resurfacing and widening projects awarded in each month and in the same county as of the project in the State of Georgia.
3. Total Number of Projects Awarded in the Current Month at the Level of the State of Georgia: It is the number of projects awarded in each month and in the same county as of the project in the State of Georgia. 113

4. Total Dollar Value of Projects Awarded in the Current Month at the Level of the State of Georgia: It is the total dollar value of projects awarded in each month and in the same county as of the project in the State of Georgia.
5. Total Asphalt Volume of Projects Awarded in the Current Month at the Level of the State of Georgia: It is the total asphalt volume of projects awarded in the current year at the level of the State of Georgia.
6. Architecture Billings Index: It is a leading economic indicator for forecasting nonresidential construction activity for the future 9 to 12 months. This index is prepared by the American Institute of Architects (AIA) through gathering billings data from architectural firm leaders.
7. Building Permits for New Residential Construction: It is the number of new housing units authorized by building permits, for new privately-owned residential construction.
8. ENR Building Cost Index: This index represents and tracks average cost of skilled labor and materials in the building construction industry over time.
9. ENR Common Labor Index: This index represents and tracks average total wages for laborers, including fringe benefits, in the construction industry over time.
10. ENR Construction Cost Index: This index represents and tracks average cost of skilled labor and materials in the construction industry over time.
11. ENR Material Price Index: This index represents and tracks average cost of major materials, such as structural steel, in the construction industry over time. 114

12. ENR Skilled Labor Index: This index represents and tracks average total wages for skilled laborers (such as carpenters, bricklayers, and iron workers), including fringe benefits, in the construction industry over time.
13. Equipment Operator Wages (Paving): It is a mean hourly wage of an equipment operator, such as asphalt paving machine operators, in the State of Georgia.
14. Georgia Asphalt Cement Price Index: It is an average selling price of asphalt cement that is collected from approved local asphalt cement suppliers as reported in the GDOT's monthly survey.
15. Gross Domestic Product (GDP) of the Georgia Construction Industry: It is the total market value of goods and services provided by the Georgia construction industry.
16. Job Opening and Labor Turnover Index (Hires): It is an index that represents the number of hires during the entire month as a percent of total employment.
17. Housing Market Index: It is an indicator for the single-family housing market. The data are collected through the survey by asking respondents to rate market conditions for the sale of new homes at the present time and in the next six months.
18. National Highway Construction Cost Index: It is a highway construction index to track pure price-changes in highway construction costs.
19. Producer Price Index (Construction machinery manufacturing): It is an index measuring changes in prices received for the output of the construction machinery manufacturing sold to another industry.
115

20. Producer Price Index (Construction sand and gravel mining): It is an index measuring changes in prices received for the output of the construction sand and gravel mining sold to another industry.
21. Turner Construction Cost Index: It is a cost index on a main basis of labor rates, productivity, and material prices in the non-residential building construction market in the U.S.
22. Value of Construction Put in Place (Pavement): It is an estimate of total dollar value of State and local pavement construction work done each month in the U.S. measured in Millions of Dollars.
23. Value of Construction Put in Place (All construction): It is an estimate of total dollar value of construction work done each month in the State of Georgia measured in Millions of Dollars.
24. Fails Management Institute (FMI) Nonresidential Construction Index: It is an index measuring nonresidential construction activity. The index is developed based on the survey of national contractors in several regions or around the U.S.
25. Labor Productivity: It is the ratio of the output of goods and services to the labor hours devoted to the production of that output.
26. Number of Establishments in Private Construction Industry: It is the number of private construction establishments in the State of Georgia.
27. 12 Month Percent Change of Georgia Asphalt Cement Price Index: It is a measure of trends in the Georgia asphalt cement price index by comparing a value in the 116

current month with a value in the corresponding month last year. The 12-month percent change shows how much of the downturn or rise in asphalt cement prices in the Georgia is. 28. 12 Month Percent Change of Gross Domestic Product (GDP) of the Georgia Construction Industry: It is a measure of trends in GDP of the Georgia construction industry by comparing a value in the current month with a value in the corresponding month last year. The 12-month percent change of GDP shows how much of the downturn or rise in GDP of the Georgia construction industry is. 29. 12 Month Percent Change of Job Opening and Labor Turnover Index (Hires): It is a measure of trends in the labor market of the construction industry by comparing a value in the current month with a value in the corresponding month last year. The 12-month percent change of Job Opening and Labor Turnover shows how much of the downturn or rise in hires of the construction industry is. 30. 12 Month Percent Change of Value of Construction Put in Place (Pavement): It is a measure of trends in the total value of construction put in place for pavement projects by comparing a value in the current month with a value in the corresponding month last year. The 12-month percent change shows how much of the downturn or rise in the total value of construction put in place for pavement projects in the South regions (e.g., Georgia, Alabama, and Florida) is.
117

4.3.3.3. Macroeconomic Conditions Macroeconomic conditions can be described by GDP and unemployment. Macroeconomic conditions have significant impact on investment in the construction industry. The following macroeconomic variables are considered in this study:
1. Dow Jones Industrial Average: It is a stock market index that reveals trading activities covering various industries among 30 large publicly-owned companies in the U.S.
2. Inflation rate: It is the rate of general rising prices for goods and services, and falling of the purchasing power of currency.
3. Average weekly wage (all industry): It is an average weekly wage for all industries that covers 98 percent of the U.S. economy. This measure is available at the county level.
4. Consumer Price Index (south): It is an economic indicator of average change of prices for purchasing consumer goods and services.
5. Producer Price Index (Gasoline products): It is an index that measures the average change over time in selling prices of gasoline related products and power by domestic producers of goods and services.
6. Producer Price Index (Steel mill products): It is an index that measures the average change over time in selling prices of steel related products and power by domestic producers of goods and services.
118

7. Producer Price Index (No. 2 diesel fuel products): It is an index that measures the average change over time in selling prices of No. 2 diesel related products and power by domestic producers of goods and services.
8. Producer Price Index (Crude petroleum products): It is an index that measures the average change over time in selling prices of crude petroleum related products and power by domestic producers of goods and services.
9. Unemployment: It is a count of people who are eligible to work but unable to find a job.
10. 12 Month Percent Change of Unemployment: It is a measure of trends in the number of unemployed people by comparing a value in the current month with the value in the same month last year. The 12-month percent change shows how much of the downturn or rise in unemployment is in the State of Georgia.
4.3.3.4. Oil Market Conditions Oil market affects unit bid prices for major asphalt line items. Volatility in oil market should be considered in pricing construction costs and managing risks related to oil/fuel prices (Damnjanovic and Zhou 2009). The following variables represent oil market condition in this study:
1. Crude Oil Price: It is the spot price of unrefined petroleum product measured in Dollars per Barrel.
2. Diesel Retail Prices: It is the spot price of diesel measured in Dollars per Barrel.
119

3. Georgia Fuel Price Index: It is an average statewide selling price of Unleaded

Regular Gasoline and Diesel Fuel.

4. 12 Month Percent Change of Georgia Fuel Price Index: It is a measure of trends in

the Georgia fuel price index by comparing a value in the current month with a value

in the corresponding month last year. The 12-month percent change shows how

much of the downturn or rise in fuel prices in the Georgia is.

4.4. Results of the Multiple Regression Modeling
As the first step, the outliers in the dataset are detected. Based on the z-scores, 33 outliers

(2.3%) are detected as shown in Table 4-6. Thus, 1391 observations are used to develop a

multiple regression model. Table 4-6 Results of Outlier Inspection

Bid Item
Recycled Asphaltic Concrete

Number of Observations
1424

Detected Outliers 33

Percentage of Removed Data
2.3%

Variable Transformation: Scatter plots and Pearson correlation analysis are used to examine whether any variable transformation is helpful to enhance the quality of regression analysis. The correlation of the submitted bid price with each of the potential explanatory variables is calculated, once with the variable in its original form and once with the transformed variables in its logarithmic form. It is found that the following variables are better transformed into their logarithmic forms before they are included in regression analysis because the correlation of these variables with the submitted bid price is higher when the variables are in the logarithmic form: the quantity of the item, project length,
120

building permits for new residential construction, Dow Jones Industrial Average, housing market index, producer price index for No. 2 diesel fuel related products, and unemployment. Multiple regression model is developed for explaining variations in submitted unit prices for major asphalt line items. The entire dataset consists of the information of resurfacing and widening projects let in the State of Georgia from January 2008 to December 2015. Ordinary least squares (OLS) regression is developed with 55 potential factors. Multicollinearity issues are also reexamined using Variance inflation factors (VIF). The results of VIF diagnosis indicated that several factors represent a multicollinearity issue in the regression modeling since the respective VIF values for these variables are greater than 10. The variables that have multicollinearity issues are: architecture billings index, building permits for new residential construction, Dow Jones Industrial Average, ENR building cost index, ENR common labor index, ENR construction cost index, ENR material price index, ENR skilled labor index, equipment operator wages for paving, Georgia fuel price index, Gross Domestic Product (GDP) of the Georgia construction industry, producer price index for gasoline related products, producer price index for steel related products, producer price index for No. 2 diesel related products, producer price index for crude petroleum related products, producer price index for construction machinery manufacturing, producer price index for construction sand and gravel mining, Turner construction cost index, consumer price index, diesel retail prices, value of construction put in place for pavement projects, value of construction put in place for all construction projects, FMI nonresidential
121

construction index, and labor productivity in highway, street, and bridge construction. These 23 variables are, therefore, removed from the final model. The results of this model are presented in Table 4-7.
122

Table 4-7 Coefficients of the Final Regression Model

Categorical Variable

Model

Unstandardized

Standardized

Coefficients

Coefficients

t

B

Std. Error

Beta

Sig. VIF

(Constant)

Flat

Terrain

Mountainous

Coastal

District 2

District 3

District

District 4

District 5

District 6

District 7

Natural Logarithm of Quantity of
1
Bid Item

Natural Logarithm of Housing Market
2
Index

3 Georgia Asphalt Cement Price Index

4

Total Bid Price

5 Natural Logarithm of Project Length

12 Month Percent Change of Georgia
6
Asphalt Cement Price Index

12 Month Percent Change of Gross

7

Domestic Product (GDP) of the

Georgia Construction Industry

8 Natural Logarithm of Unemployment

Total Asphalt Volume of Resurfacing

and Widening Projects Awarded in the
9
Current Month at the Level of the

County

10

Number of Bidders

11

Project Duration

12 Average Weekly Wage (all industry)

Total Number of Projects Awarded in

13 the Current Month at the Level of the

State of Georgia

Number of Establishments in Private
14
Construction Industry

15

Crude Oil Price

54.734 2.289 3.344 5.748 3.660 0.166 4.083 5.654 3.365 1.880 -1.561
4.347 0.023 2.83410-7 -2.223 0.037
0.091
-0.582
9.05510-7
-0.427 -0.003 0.003
0.021
0.001 0.027

5.130 0.565 0.811 1.014 0.665 0.663 0.749 0.886 0.709 0.841 0.180
0.669 0.004 4.06110-8 0.273 0.012
0.034
0.242
2.85610-7
0.127 0.001 0.002
0.011
0.001 0.020

0.107 0.085 0.147 0.161 0.007 0.18 0.228 0.113 0.071 -0.247
0.236 0.234 0.211 -0.193 0.141
0.110
-0.097
0.089
-0.081 -0.069 0.065
0.064
0.064 0.063

10.668 4.052 4.123 5.667 5.501 0.25 5.451 6.378 4.749 2.235

0.000 0.000 0.000 0.000 0.000 0.802 0.000 0.000 0.000 0.026

2.114 1.277 2.026 2.578 2.608 3.279 3.836 1.690 2.991

-8.652 0.000 2.459

6.497 0.000 3.966

6.214 6.979 -8.134

0.000 0.000 0.000

4.277 2.738 1.690

2.984 0.003 6.681

2.659 0.008 5.191 -2.410 0.016 4.834

3.170 0.002 2.392

-3.363 -2.238 1.557

0.001 0.025 0.120

1.724 2.88 5.161

1.927 0.054 3.320

1.827 0.068 3.743 1.319 0.187 6.782

123

Rank

Model

Unstandardized

Standardized

Coefficients

Coefficients

t

B

Std. Error

Beta

Sig. VIF

Number of Nearby Asphalt Plants
16
(within 50 miles)

-0.041

Job Opening and Labor Turnover
17
Index (Hires)

0.005

12 Month Percent Change of
18
Unemployment

0.017

12 Month Percent Change of Job

19 Opening and Labor Turnover Index

-0.029

(Hires)

Total Number of Resurfacing and

Widening Projects Awarded in the
20
Current Month at the Level of the

-0.201

Rank (Cont'd)

County

12 Month Percent Change of Georgia
21
Fuel Price Index

1.99610-4

22

Inflation Rate

0.144

Total Dollar Value of Projects

23 Awarded in the Current Month at the 3.80710-9

Level of the State of Georgia

Hauling Distance between Asphalt
24
Plant and Project Location

0.013

Hauling Distance between Quarry and
25
Asphalt Plants

0.003

26 Monthly GA Total Asphalt Volume

2.26310-9

National Highway Construction Cost
27
Index

-0.351

28

Number of Pay Items

1.08910-5

12 Month Percent Change of Value of
29
Construction Put in Place (Pavement)

0.001

Note: Sig. = Significance Probability (p-value)

0.017 0.003 0.012
0.016
0.126
1.53110-4 0.253
5.65410-9
0.023 0.006 7.65810-9 4.116 1.74810-4 0.024

-0.060 0.057 0.056
-0.046
-0.042
0.027 0.026 0.017
0.011 0.010 0.007 -0.003 0.001 0.001

-2.374 0.018 1.903 1.851 0.064 2.802 1.408 0.159 4.675
-1.801 0.072 1.975
-1.594 0.111 2.103
1.304 0.192 1.321 0.566 0.571 6.524 0.673 0.501 2.024
0.549 0.583 1.308 0.503 0.615 1.188 0.295 0.768 1.517 -0.085 0.932 4.156 0.062 0.950 1.192 0.036 0.971 1.276

The beta coefficient for each variable is calculated to determine the relative importance of

explanatory variables in the regression model. Reviewing the absolute values of beta

coefficients and significance probabilities of each variable in the developed regression

model indicates that the most important factors in explaining submitted bid prices are

124

quantity of the bid item, followed by housing market index, Georgia asphalt cement price index, total bid price, project length, 12-month percent change of Georgia asphalt cement price index, 12-month percent change of Gross Domestic Product (GDP) of the Georgia construction industry, unemployment, total asphalt volume of resurfacing and widening projects, number of bidders, project duration, and number of nearby asphalt plants. The results of the regression analysis also indicate that quantity of the bid item, project length, unemployment, number of bidders, project duration, and number of asphalt plants have negative relationship with submitted bid prices while holding all other variables in the model constant. Georgia has 4 types of terrain, including rolling, flat, mountainous, and coastal terrain. Terrain type is included as categorical variable in the regression model. The results of regression analysis conclude that submitted bid prices in flat, mountainous, and coastal terrain are higher than those in rolling terrain, on average, while holding all other variables constant. Submitted bid prices in coastal terrain are relatively higher than those in other terrain types, on average. GDOT has 7 district offices throughout the State of Georgia. Seven categorical variables are used in regression analysis. The results show that submitted bid prices in all other districts are higher than those in District 1, on average, while holding all other factors in the model constant. Submitted bid prices in District 5 showed relatively higher than those in other districts, on average.
125

The final regression model with selected explanatory variables has adjusted R-Squared of 0.538 with F (38, 1352) = 43.543 and p < .001. As it can be seen in Table 4-8, the regression model can explain approximately 53.8% of variations in submitted bid prices for major asphalt line items. The results of ANOVA tests are provided in Table 4-9. The null hypothesis is rejected at 1% significance level indicating at least one of the coefficients of the identified explanatory variables is not zero in the regression model. Overall, a combination of variables used in the regression model is statistically significant for explaining variations in submitted bid prices.

Table 4-8 Model Summary of the Final Regression Model

R

R Squared Adjusted R Squared Std. Error of the Estimate

0.742

0.550

0.538

5.771535

Model Regression

Table 4-9 ANOVA of the Final Regression Model

Sum of Squares 55116.571

df

Mean Square

F

Sig.

38

1450.436 43.543 0.000

Residual

45035.965

1352

33.311

Total

100152.527

1390

Note: df= Degree of Freedom; F= F Statistic; Sig. = Significance Probability

Figure 4-5 depicts the histogram of standardized residuals for the regression model that is

used to diagnose the regression model for the normality assumption. The histogram of

126

standardized residuals is appeared normal and bell-shaped. This histogram suggests that the normality assumption is not violated in the regression modeling.
Figure 4-5 Histogram of Standardized Residuals for the Regression Model Figure 4-6 depicts the normal probability-probability (P-P) plot based on the calculated standardized residuals. If the residuals are normally distributed, the residuals should fall on the diagonal line. In Figure 4-6, the residual plot (the normal probability plot) is appeared to generally follow a straight line. This result indicates that the normality assumption is met.
127

Figure 4-6 Normal P-P Plot of Standardized Residuals for the Regression Model Figure 4-7 shows the scatterplot of standardized residuals against predicted values. If residuals are homoscedastic, then the spread of the residuals should balance evenly around zero on the Y-axis. The scatterplot has a random pattern centered on the line of zero standard residual value and there is no clear relationship between the residuals and the predicted values. Thus, the assumptions of homoscedasticity and independence are met.
128

Figure 4-7 Scatterplot of Standardized Residuals vs. Predicted Values
4.5. Results of the Multiple Regression Modeling for Projects in TIA Regions
Regression analysis is conducted for modeling variations in submitted bid prices in the three Regional Commissions in the State of Georgia: Central Savannah River Area, Heart of Georgia Altamaha, and River Valley where the Transportation Investment Act (TIA) is implemented for improving the local transportation system. These regions are depicted in Figure 4-8. A specific regression model is created to identify significant factors that can explain variations in submitted bid prices just using the data from the TIA regions. The
129

dataset used in this analysis is submitted unit prices for asphalt line items used in resurfacing and widening projects let in these three regions from 2008 to 2015.
Figure 4-8 Three Regional Commissions for TIA Since this section focuses on TIA regions, categorical variables for Districts are excluded from the analysis. In addition, as TIA regions do not contain any mountainous terrain, this terrain variable is also excluded from the regression analysis. Regression modeling is done to explain variations in submitted bid prices in TIA regions using information from 21 remaining factors. Variable Transformation: Scatter plots and Pearson correlation analysis are used to examine whether any variable transformation is helpful to enhance the quality of regression analysis. The correlation of the submitted bid price with each of the potential explanatory variables is calculated, once with the variable in its original form and once with the
130

transformed variable in its logarithmic form. It is found that Total Asphalt Volume of Resurfacing and Widening Projects is better transformed into its logarithmic form before it is included in regression analysis because the correlation of the variable with the submitted bid price is higher when the variable is in the logarithmic form. Ordinary least squares (OLS) regression is developed with 54 factors. Multicollinearity issues are found in the results (i.e., VIF is less than 10 for all explanatory variables). The variables that contain multicollinearity issues are: building permits for new residential construction, crude oil price, Dow Jones Industrial Average, ENR building cost index, ENR common labor index, ENR construction cost index, ENR material price index, ENR skilled labor index, equipment operator wages for paving, Georgia fuel price index, Gross Domestic Product (GDP) of the Georgia Construction Industry, housing market index, national highway construction cost index, producer price index for gasoline related products, producer price index for No. 2 diesel fuel related products, producer price index for crude petroleum related products, producer price index for construction machinery manufacturing, producer price index for construction sand and gravel mining, consumer price index, diesel retail prices, value of construction put in place for pavement projects, value of construction put in place for all construction projects, FMI nonresidential construction index, labor productivity in highway, street, and bridge construction, and 12month percent change of value of construction put in place for pavement projects. These 23 variables are removed from the final model. The regression results are shown in Table 4-10.
131

Table 4-10 Coefficients of the Final Regression Model Developed Based on Projects in TIA Regions

Model
(Constant) Flat

Unstandardized

Coefficients

B

Std. Error

35.064

17.731

Standardized Coefficients
Beta

7.197

0.724

0.441

t
1.978 9.948

Sig.
0.049 0.000

Terrain

Coastal

1

Unemployment

2

12 Month Percent Change of Georgia

Asphalt Cement Price Index

3

Quantity of the Bid Item

4

Total Bid Price

5

Architecture Billings Index

6

Number of Establishments in Private

Construction Industry

7

Georgia Asphalt Cement Price Index

8

Project Duration

9

Turner Construction Cost Index

10

Job Opening and Labor Turnover Index

(Hires)

Natural Logarithm of Total Asphalt

11

Volume of Resurfacing and Widening

Projects

Total Dollar Value of Projects Awarded in
12 the Current Month at the Level of the State
of Georgia

Total Number of Projects Awarded in the

13 Current Month at the Level of the State of

Georgia

14

Number of Bidders

15

Hauling Distance between Quarry and

Asphalt Plants

12 Month Percent Change of Gross
16 Domestic Product (GDP) of the Georgia

Construction Industry

Total Number of Resurfacing and Widening
17 Projects Awarded in the Current Month at

the Level of the County

18

Inflation Rate

19

Average Weekly Wage (all industry)

12.895 0.001 0.048 -1.78210-4 2.53710-7 0.309 -0.011 0.015 -0.005 0.023 0.009
-0.952
1.78610-8
0.027 -0.500 0.022
0.055
-0.603 -0.341 -0.004

2.260 3.89210-4
0.024 7.41410-5 1.11910-7
0.198 0.007 0.008 0.003 0.012 0.005
0.658
9.97010-9
0.023 0.277 0.012
0.080
0.490 0.497 0.004

0.242 0.342 0.202 -0.197 0.190 0.181 -0.178 0.168 -0.128 0.128 0.113
-0.107
0.097
0.094 -0.084 0.083
0.072
-0.066 -0.065 -0.057

5.706 2.917

0.000 0.004

2.025 0.044

-2.403 2.267 1.564

0.017 0.024 0.119

-1.569 0.118

1.823 -1.408 1.986

0.069 0.160 0.048

1.783 0.076

-1.447 0.149

1.791 0.074

1.163 0.246 -1.805 0.072 1.920 0.056
0.690 0.491

-1.232 0.219 -0.687 0.493 -1.066 0.287

VIF
1.241 1.140 8.711 6.279 4.244 4.449 8.447 8.125 5.361 5.266 2.611 2.557
3.488
1.872
4.108 1.358 1.193
6.929
1.815 5.595 1.815

Rank

132

Rank (Cont'd)

Unstandardized

Model

Coefficients

B

Std. Error

20

12 Month Percent Change of

Unemployment

21

Number of Pay Items

-0.016 0.001

0.023 0.001

Total Asphalt Volume of Projects Awarded
22 in the Current Month at the Level of the

1.39610-8

1.72810-8

State of Georgia

23

Value of Construction Put in Place

(Pavement)

24

Project Length

1.70210-4 2.72610-4

-0.039

0.105

25 12 Month Percent Change of Georgia Fuel
Price Index

-1.39410-4 3.43810-4

26 Number of Nearby Asphalt Plants (within
50 miles)
27 Producer Price Index (Steel mill products)

0.016 -0.006

0.041 0.047

Hauling Distance between Asphalt Plant
28
and Project Location

0.013

0.044

12 Month Percent Change of Job Opening
29
and Labor Turnover Index (Hires)
Note: Sig. = Significance Probability (p-value)

0.001

0.030

Standardized Coefficients
Beta -0.052 0.048
0.040
0.031 -0.023 -0.019
0.017 -0.015 0.013
0.002

t
-0.712 1.141 0.808
0.624 -0.374 -0.405 0.393 -0.131 0.298 0.027

Sig.
0.477 0.255 0.420
0.533 0.708 0.685 0.694 0.896 0.766 0.978

VIF 3.324 1.136 1.553
1.575 2.361 1.336 1.245 8.051 1.272 2.006

Reviewing the absolute values of beta coefficients and significance probabilities of each

variable in the developed regression model indicates that 5 most important factors in

explaining submitted bid prices in descending order of importance are: unemployment, 12-

month percent change of Georgia asphalt cement price index, quantity of the bid item, total

bid price, and Turner construction cost index. The results of the regression analysis also

suggest that quantity of the bid item has negative relationship with submitted bid prices

while holding all other variables constant.

Terrain types in TIA regions are rolling, flat, and coastal terrain. Submitted bid prices in

flat and coastal terrain are higher than those in the rolling terrain, on average, while holding

133

all other variables constant. Submitted bid prices in coastal terrain are far higher than those

in other terrain types in TIA regions, on average.

The final regression model with selected explanatory variables is developed for explaining

variations in submitted bid prices in TIA regions. Table 4-11 shows a summary of

regression modeling in TIA regions. The final model can explain approximately 42% of

variations in submitted bid prices using information from 30 explanatory variables. It is

shown in Table 4-12 that the overall model is found to be statistically significant (F = 9.559,

p<0.001).

Table 4-11 Model Summary of the Final Regression Model Developed Based on Projects in TIA Regions

Std. Error of the

R

R-Square

Adjusted R-Square

Estimate

0.685

0.469

0.420

5.92246

Table 4-12 ANOVA of the Final Regression Model Developed Based on Projects in TIA Regions

Sum of

Mean

Model

df

F

Sig.

Squares

Square

Regression 10393.507

31

335.274

9.559

0.000

Residual

11785.379

336

35.076

Total

22178.886

367

Note: df= Degree of Freedom; F= F Statistic; Sig. = Significance Probability

Figure 4-9 depicts the histogram of standardized residuals for the regression model that is

used to diagnose the regression model for the normality assumption. The histogram of

134

standardized residuals is appeared normal and bell-shaped. This histogram suggests that the normality assumption is not violated in the regression modeling.
Figure 4-9 Histogram of Standardized Residuals for the Regression Model Developed Based on Projects in TIA Regions
Figure 4-10 depicts the normal probability-probability (P-P) plot based on the calculated standardized residuals. If the residuals are normally distributed, the residuals should fall on the diagonal line. In Figure 4-10, the residual plot (the normal probability plot) is appeared to generally follow a straight line. This result indicates that the normality assumption is met.
135

Figure 4-10 Normal P-P Plot of Standardized Residuals for the Regression Model Developed Based on Projects in TIA Regions
Figure 4-11 shows the scatterplot of standardized residuals against predicted values. If residuals are homoscedastic, then the spread of the residuals should balance evenly around zero on the Y-axis. The scatterplot has a random pattern centered on the line of zero standard residual value and there is no clear relationship between the residuals and the predicted values. Thus, the assumptions of homoscedasticity and independence are met.
136

Figure 4-11 Scatterplot of Standardized Residuals vs. Predicted Values for the Regression Model Developed Based on Projects in TIA Regions
137

CHAPTER 5 CONCLUSIONS
This research aims to present a set of cost estimation and management practices for budgetbased design that can aid GDOT project managers and engineers throughout the plan development process (PDP). To achieve the research objective, current state of practices in cost estimation and control in several State DOTs are reviewed and best practices are identified in cost estimation and strategies for cost control. Also, current state of practices in fixed budget-best value procurement method are reviewed and best practices are identified in utilization of this innovative contracting method. Finally, statistical analysis is conducted to investigate the impact of macroeconomic, construction market, and oil market conditions on highway construction costs by analyzing submitted bid prices for major asphalt line items in the State of Georgia's highway projects.
The following recommendations are found out to be effective for enhancing the practice of defining and maintaining the established budget for highway projects:
State DOTs should establish an integrated process for cost estimation and cost management to establish accurate, reliable, and consistent estimates thorough project development process.
State DOTs should establish key milestones for estimating, updating, and approving cost estimates as project definition/design advances.
138

State DOTs should capture any changes in estimating assumptions to track the basis of cost estimate and control estimated project cost.
State DOTs are recommended to establish an automated information system to help them maintain, update, and share project information, cost estimates, and changes in project scope, cost, and schedule.
State DOTs should consider potential issues (risks) that may cause cost escalation during developing baseline cost estimates. Risk analysis tools and inputs from key project stakeholders are necessary for identifying critical risk factors for the project.
Finally, State DOTs should utilize a quality assurance/quality control (QA/QC) process to verify the final engineer's cost estimate before a project is advertised.
The following recommendations are found out to be effective for enhancing the practice of delivering projects using this innovative contracting strategy:
State DOTs may consider a fixed budget-best value procurement approach when the full project scope exceeds the baseline cost estimate for the project. For a fixed budget-best value procurement approach, the agency should define the basic configuration scope and should allow the proposers to include the maximum amount of work or additional scope elements in their proposals while staying within the fixed budget.
A fixed budget-best value approach can be utilized in several project types, such as corridor expansion, bridge deck preservation, and seal coating projects. State DOTs
139

should clearly define additional scope elements beyond the base scope for each project type. State DOTs should clearly establish the evaluation criteria to select the proposers for a project. Since the price is fixed for all proposers and this approach allows higher flexibility in proposing design and construction solutions, the agency should establish rigorous evaluation criteria (e.g., cost, time, and design alternatives) and the weights for the criteria to evaluate the proposals based on the project goals.
Lastly, statistical analysis is conducted to identify important variables capable of explaining variations in submitted bid prices for major asphalt line items in the GDOT's highway projects. Multiple regression analysis is utilized to examine the impact of project characteristics, macroeconomic variables, construction market condition indicators, and oil market parameters on highway construction costs. The main purpose of this research is to examine the effects of several factors representing construction market, economic, and oil market conditions on submitted bid prices. The goal is to develop a regression model with explanatory power to describe variations in submitted bid prices. It is worth noting that the developed regression model can be used for forecasting bid prices for asphalt line items but prediction was not the main objective of this study. Therefore, the results should be used with caution as the forecasting error might be significantly large. An explanatory model is developed for the State of Georgia's highway projects using multiple regression analysis. Several important variables are identified to have power to explain variations in submitted bid prices for major asphalt line items. The identified
140

variables, in descending order of importance, are: quantity of the bid item, housing market index, Georgia asphalt cement price index, total bid price, project length, 12-month percent change of Georgia asphalt cement price index, 12-month percent change of Gross Domestic Product (GDP) of the Georgia construction industry, unemployment, total asphalt volume of resurfacing and widening projects, number of bidders, project duration, and number of nearby asphalt plants. Among these significant explanatory variables, quantity of the bid item, project length, unemployment, number of bidders, project duration, and number of asphalt plants have negative relationship with submitted bid prices while holding all other variables in the model constant. All other variables have positive influence on submitted bid prices. Multiple regression analysis is repeated for identifying significant factors that affect variations in submitted bid prices in the regions included in the Transportation Investment Act (TIA). The identified important variables, in descending order of importance, are: unemployment, 12-month percent change of Georgia asphalt cement price index, quantity of the bid item, total bid price, and Turner construction cost index. Among those significant factors in the explanatory model developed for projects in the TIA regions, quantity of the bid item has negative relationship with submitted bid prices while holding all other variables constant. All other variables have positive relationship with submitted bid price.
141

REFERENCES
Actis, C. M. (2010). "FHWA Risk Based Cost Estimate Reviews", Federal Highway Administration.
Akinci, B., and Martin, F. (1998). "Factors Affecting Contractors' Risk", Journal of Management in Engineering, 14.1(February), 6776.
Anderson, S., Molenaar, K., and Schexnayder, C. (2007). "Guidance for Cost Estimation and Management for Highway Projects during Planning, Programming, and Preconstruction", Transportation Research Board.
Arditi, D., Akan, G. T., and Gurdamar, S. (1985). "Cost overruns in public projects", International Journal of Project Management, 3(19641974), 218224.
Ashuri, B., Shahandashti, S. M., and Lu, J. (2012). "Empirical tests for identifying leading indicators of ENR Construction Cost Index", Construction Management and Economics, 30(11), 917927.
Caldas, C. H., Gibson, G. E., and Le, T. (2007). "TxDOT Best Practices Model and Implementation Guide for Advance Planning Risk Analysis for Transportation Projects", Texas Department of Transportation.
Caltrans. (2007a). "CHAPTER 20 Project Development Cost Estimates", California Department of Transportation.
Caltrans. (2007b). "Report on Quality Control / Quality Assurance (QC / QA) Activities for Project Cost Estimating", California Department of Transportation.
CDOT. (2003). "TREX Fact Book 2003", Colorado Department of Transportation.
CDOT. (2016). "Design-Build Manual". Colorado Department of Transportation, Denver, Co.
Chou, J. S. (2009). "Web-based CBR system applied to early cost budgeting for pavement maintenance project", Expert Systems with Applications, Elsevier Ltd, 36(2), 29472960.
Chou, J. S. (2010). "Cost simulation in an item-based project involving construction engineering and management", International Journal of Project Management.
142

Chua, D. K. H., and Li, D. (2000). "Key Factors in Bid Reasoning Model", Journal of Construction Engineering and Management, 126(5) (October), 349357.
Damnjanovic, I., and Zhou, X. (2009). "Impact of crude oil market behavior on unit bid prices: the evidence from the highway construction sector", Construction Management and Economics, 27, 881890.
FHWA. (2013). "Programmatic Use of Fixed Price Variable Scope Contracting on Capital Preventative Maintenance Projects", Federal Highway Administration.
FHWA. (2015). "National Review of State Cost Estimation Practice", Federal Highway Administration.
FHWA. (2016). "SEP-14 Active Project List", Federal Highway Administration.
Field, A. (2009). "Discovering statistics using SPSS". Sage Publications.
GDOT. (2009). "Georgia Department of Transportation Cost Estimation Review Study, RUCEST and Trns-port CES Training Materials". Georgia Department of Transportation.
GDOT. (2014). "Georgia Counties and GDOT Field Districts", the Office of Transportation Data, Georgia Department of Transportation.
Harper, C. M., Molenaar, K. R., Anderson, S., and Schexnayder, C. (2014). "Synthesis of Performance Measures for Highway Cost Estimating", Journal of Construction Engineering & Management, 30.
Hegazy, T., and Ayed, A. (1998). "Neural Network Model for Parametric Cost Estimation of Highway Projects", Journal of Construction Engineering and Management, 124(3), 210 218.
Ilbeigi, M., Ashuri, B., and Shayegh, S. (2015). "Price Adjustment Clauses and Submitted Bid Prices for Major Asphalt Line Items in Highway Projects", Journal of Construction Engineering and Management, 4015103.
ITD. (2014). "Alternative Project Delivery." Idaho Transportation Department.
Jacoby, W. G. (2005). "Regression III: Advanced Methods", Department of Political Science, Michigan State University.
143

Kandhal, P. S., Rao, S. S., Watson, D. E., and Young, B. (1995). "Performance of Recycled Hot Mix Asphalt Mixtures", NCAT Report 95-01, Auburn, AL.
MDOT. (2012). "Final Evaluation Report Special Experimental Project 14 (SEP-14): Fixed Price Variable Scope Contracting", Michigan Department of Transportation.
MDOT. (2015). "Innovative Construction Contracting Guide", Michigan Department of Transportation.
MnDOT. (2008). "Cost Estimation and Cost Management Technical Reference Manual", Minnesota Department of Transportation.
Molenaar, K., Anderson, S., and Schexnayder, C. (2010). "Guidebook on Risk Analysis Tools and Management Practices to Control Transportation Project Costs", Transportation Research Board.
Molenaar, K., Anderson, S., and Schexnayder, C. (2011). "AASHTO Practical Guide to Estimating", American Association of State Highway and Transportation Official.
Montgomery, D. C., Peck, E. A., and Vining, G. G. (2015). "Introduction to linear regression analysis", John Wiley & Sons.
Oberlender, G. D., and Trost, S. M. (2001). "Predicting Accuracy of Early Cost Estimates based on Estimate Quality", Journal of Construction Engineering & Management, 127(June), 173182.
ODOT. (2013). "ODOT Project Development Process Manual", Ohio Department of Transportation.
ODOT. (2016a). "Feasibility Studies and Alternative Evaluation Reports", Ohio Department of Transportation.
ODOT. (2016b). "301 Asphalt Concrete Base" Ohio Department of Transportation.
Peng, M. (2006). "Item-Level Quantity-Based Preliminary Cost Estimating System for Highway structures and Miscellaneous Construction", Doctoral dissertation, The University of Texas at Austin, Texas
Scott, S., Molenaar, K. R., Gransberg, D. D., and Smith, N. C. (2006). "Best-Value Procurement Methods for Highway Construction Projects", National Cooperative Highway Research Program, Washington, D.C.
144

Shane, J. S., Molenaar, K. R., Anderson, S., and Schexnayder, C. (2009). "Construction project cost escalation factors", Journal of Management in Engineering, American Society of Civil Engineers, 25(4), 221229.
Shrestha, P. P., and Pradhananga, N. (2010). "Correlating Bid Price with the Number of Bidders and Final Construction Cost of Public Street Projects", Transportation Research Record: Journal of the Transportation Research Board, 2151(310).
Shrestha, P. P., Pradhananga, N., and Mani, N. (2014). "Correlating the quantity and bid cost of unit price items for public road projects", KSCE Journal of Civil Engineering, 18(6), 15901598.
TxDOT. (2006). "DCIS User Manual", Texas Department of Transportation.
TxDOT. (2014). "Project Development Process Manual", Texas Department of Transportation.
UDOT. (2013). "OTEC Financing Our Future Roles and Responsibilities on DB Contracts", Utah Department of Transportation.
UDOT. (2016). "Best Value Design-Build Selection", Utah Department of Transportation.
Washington, S. P., Karlaftis, M. G., and Mannering, F. L. (2010). "Statistical and Econometric Methods for Transportation Data Analysis", CRC press.
Wilmot, C. G., and Cheng, G. (2003). "Estimating Future Highway Construction Costs", Journal of Construction Engineering and Management, 129(3), 272279.
WSDOT. (2012). "WSDOT Guidelines for CRA-CEVP Workshops", Washington Department of Transportation
WSDOT. (2013). "Appendix N: Design-Build Delivery White Papers", Washington Department of Transportation.
WSDOT. (2015). "Cost Estimating Manual for WSDOT Projects", Washington State Department of Transportation.
Youngs, C. (2013). "Innovative Contracting", Asphalt Pavement Association of Michigan 2013 Annual Conference.
145