U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-04-046
Date: October 2004
The majority of State highway administrations (SHAs) now employ statistical
quality assurance (QA) specifications to some degree. These specifications
contain statistical acceptance plans that serve to inform a contractor of
at least three items: (1) the quality level that the agency desires, (2) how
the contractor's submitted quality level will be determined (i.e., estimated),
and (3) the consequences for the contractor when the submitted quality level
estimate is below or above the desired level. Whether the acceptance plan
leads to simple pass/fail decisions or adjustments in contract payment, its
proper development is critical for the plan to be effective.
The development of statistical acceptance plans for highway construction requires a good understanding of statistics, materials and construction variability, and the product quality/performance/cost interrelationship. Currently, however, acceptance plan designers must make some design decisions relying more on intuition than on established engineering and economic principles. One problem, of course, is that there is much that is still unknown about highway quality, variability, performance, and cost. Some long-term research into these topics is currently underway. Nonetheless, designers must make do with whatever information, instructions, and resources are currently available when developing new, or revising existing, acceptance plans.
Although much good information exists for acceptance plan design, there is also much misinformation and confusion. This has resulted in a general lack of uniformity among highway agency acceptance plans; use of acceptance plans that range from totally ineffective to impractically severe; difficulties in evaluating the effectiveness of some nonstandard acceptance plans; and, presumably, general dissatisfaction as evidenced by frequent revisions in agency acceptance plans (these major revisions often have a significant impact on construction). It would help greatly if the acceptance plan designer had clear, specific, supported, and comprehensive guidance for developing acceptance plans based on best practices and on what is currently known. The guidance should also, where possible, take the subjectivity out of acceptance plan design and replace it with rational and defensible scientific procedures.
As part of a pooled fund study, 18 SHAs and 1 Canadian Province provided funds for this project. The agencies that provided funding for this study are shown in figure 1.
The objective of the project was to develop a comprehensive manual, supported by scientific evidence and statistical theory, that provides step-by-step procedures and instructions for developing effective and efficient QA specifications.
Figure 1. States that provided funding for the study.
For the project, a team was assembled that had extensive experience in the
area of QA and statistical applications to highway materials. The team included
individuals with extensive experience in the development and analysis of QA
and acceptance procedures. A statistician was also included to assist in the
statistical analyses of the various procedures to be considered. The members
of the project team are identified in table 1.
To oversee the project, a panel (subsequently referred to as the panel) was formed that consisted of one representative from each of the agencies that contributed funding for the study. The contracting officer's technical representative (COTR) for the Federal Highway Administration (FHWA) was also a member of the panel. The panel member for a few of the States changed over the course of the project. Table 2 shows the members comprising the panel that approved the QA specifications manual that was developed.
A number of tasks were completed in the process of accomplishing the goal of a step-by-step manual for developing QA specifications. The methodologies for accomplishing these various tasks are discussed in the following sections.
Table 1. Project team.
|Member||Experience||Areas of Expertise|
|James L. Burati, Jr., Ph.D.
Project Principal Investigator
|25 years||QA Specifications, Evaluation of Asphalt Concrete (AC) Mixtures, AC Pavement Construction, Total Quality Management (TQM), Statistical Applications in Highway Construction, Computer Simulation|
|Hoke S. Hill, Jr., Ph.D.
Project Statistical Consultant
|23 years||Sampling Theory, Statistical Analysis, Statistical Graphics, Statistical Computing, Experiment Design|
|Richard M. Weed, P.E.
Project QA Consultant
|40 years||Statistical QA, Computer Simulation, Construction Specifications Development, Applications of Statistical Theory to Engineering Problems, Portland Cement Concrete (PCC) Specifications|
|Charles S. Hughes, P.E.
Project QA Consultant
|40 years||AC Mix Design, QA Specifications, Pavement Performance, Risk Assignment and Assessment, Asphalt Pavement Construction|
The project team members have written a good deal of the highway materials
and construction QA literature. So, the resumes of the project team were reviewed
and a list of their publications relating to the proposed research was compiled.
In addition, a number of computer database searches were conducted. The computer
search capability of Clemson University's Cooper Library and Web-based search
approaches were employed.
The abstracts obtained from the database searches were reviewed and the full publications were obtained for those abstracts that appeared to be most appropriate for further study for the current project.
In addition, all SHAs were contacted to request copies of their specifications for hot-mix asphalt concrete (HMAC) and portland cement concrete (PCC). Specifications were obtained from 31 agencies, including most of the pooled fund States for this project.
Table 2. Panel members.
|Chris Abadie||Louisiana Department of Transportation and Development|
|Roger Apple||Pennsylvania Department of Transportation (DOT)|
|Ataur Bacchus||Ontario Ministry of Transportation (MOT)|
|Kevin Dayton||Washington State DOT|
|Steve DeWitt||North Carolina DOT|
|Doug Dirks||Illinois DOT|
|Milton Fletcher||South Carolina DOT|
|Steve Gage||Connecticut DOT|
|Jeff Hale||Nevada DOT|
|Kurt Johnson||Wisconsin DOT|
|James Klessig||Minnesota DOT|
|Rick Kreider||Kansas DOT|
|Bill Maupin||Virginia Transportation Research Council|
|Garth Newman||Idaho Transportation Department|
|Thomas Reis||Iowa DOT|
|Deniz Sandhu||New York State DOT|
|Jeffrey Seiders||Texas DOT|
|Ken Stoneman||Oregon DOT|
|Richard Weed||New Jersey DOT|
One of the most important tasks for the project was to identify the necessary
steps (and the specific options available at each step) for a highway agency
to comprehensively develop a QA specification. To help accomplish this task,
the information collected during the literature search was sorted and synthesized.
The project team was also able to call on their many collective years of experience
in the development and analysis of QA specifications.
The entire project team met in Clemson, SC, to develop the structure for the QA specifications development, implementation, and monitoring process. At this meeting, a detailed flowchart was developed and the major discussion points to accompany the flowchart were identified. Subsequently, Charles Hughes developed the initial draft of an executive summary to accompany and describe the elements of the process flowchart. The executive summary and flowchart were distributed to the panel for review. The flowchart was then reviewed and discussed at a meeting attended by the panel members and members of the project team. After a great deal of discussion, some revisions were agreed upon and were subsequently incorporated into the final flowchart.
To a great extent, the approved QA specifications development process flowchart
identified the various analyses that needed to be conducted. A number of decisions
needed to be made to progress through the flowchart. Each of these decisions
had several possible options from which to select. These options were analyzed
to determine which one was the best approach to recommend.
The analysis techniques used depended on the decision that needed to be made. Both analytical and computer simulation approaches were used. In some cases, such as with the risks associated with certain hypothesis tests, an analytical evaluation was relatively simple to use. In many cases, however, a computer simulation approach was the best (and, in some cases, the only) analysis method to use. This is particularly true for issues that are related to multiple quality characteristics.
The specific analyses that were conducted, and the decisions that were made based on these analyses, are covered in detail in the various chapters of this report.
The principal product of this project, as stated in the initial request for
proposal (RFP), is
... a comprehensive manual that a highway agency can use when developing new, or modifying existing, acceptance plans. The draft manual shall provide all necessary instructions and illustrative examples to clearly lead the agency through the entire process of acceptance plan development, including:
It is important to note that two documents have been prepared for the project-a
manual for SHAs and this technical report, which summarizes the procedures
and findings of the project.(1) While the focus and objectives
of these documents are quite different, they are not entirely stand-alone
documents. In preparing the two documents, an attempt has been made to minimize
duplication of coverage. The technical report should be read in conjunction
with and as a companion to the QA specifications manual that also resulted
from the project.
The QA specifications development manual is directed toward SHAs that have a need to develop new QA specifications, or revise or update existing specifications. The focus of the manual is on what to do when developing QA specifications. The reasons for the various steps and possible decisions are explained, and easy-to-follow examples are included to assist in understanding the process that is involved. The manual does not explain what was done during the project, nor what analytical and simulation analyses were conducted unless it was necessary to clarify why certain steps in the process are necessary.
Individual members of the project team were assigned the task of preparing the first draft of specific portions of the manual. The principal investigator then collected the individual drafts and edited them into a cohesive draft manual for review by the COTR and the panel members. A panel meeting was then held to review and discuss the final draft. Comments and suggestions resulting from the panel meeting were incorporated into the final manual.
Chapter 2 presents a brief overview of the literature review, with primary emphasis placed on the specifications obtained from the various agencies that responded to the request for information. Chapter 3 presents the flowchart of the QA specifications development process that guided the specific analyses that were conducted to determine and support the recommendations that are made in the QA specifications manual. The chapters that follow the flowchart present the results and conclusions from the various analyses.