Commentary from the Final Rulemaking for 23 CFR Part 637
Published June 30, 1995
Quality Assurance Procedures For Construction
Summary: The FHWA is revising its regulations that establish general requirements for quality assurance procedures for construction on Federal-aid highway projects. The rule provides more flexibility than the existing regulation. The rule allows the use of contractor test results in making the acceptance decision and allows the use of consultants in the independent assurance program and verification sampling and testing. The regulation requires testers and laboratories to be qualified. However, it gives the States the flexibility to establish those qualifications. The revisions will clarify existing policy and procedures and provide additional guidance on the use of contractor-supplied test results in acceptance plans.
Effective Date: July 31, 1995.
The current regulations on sampling and testing of materials and construction appear in 23 CFR Part 637, Construction Inspection and Approval. These regulations were last revised in January 1987. The regulations were written using the concept of the State performing all the sampling and testing, which had been the traditional approach to sampling and testing. The regulations do not address the use of contractor testing. As a result, a number of questions arose in those States which were using contractor testing in their quality control/quality assurance (QC/QA) programs.
The existing regulations do not recognize the use of contractor testing results in an acceptance program. An acceptance program is the process of determining whether the materials and workmanship are in reasonably close conformity with the requirements of the approved plans and specifications. In 1992, the FHWA studied the ramifications of using contractor-performed sampling and testing results. The results of its study are reported in "Limits of Use of Contractor Performed Sampling and Testing," dated July 1, 1993. (A copy of the report is available in the docket for inspection and copying.) One of the report's recommendations was that contractor sampling and testing may be used in acceptance programs, provided adequate checks and balances are in place to protect the public investment. The revisions to part 637 made in this final rule would implement the committee's recommendation.
This final rule provides more flexibility to the States in designing their acceptance programs than currently exists. Acceptance of materials and construction will not be based solely on any one set of information. Each State's verification sampling and testing will be used to ensure the quality of the product. In addition, the rule will permit the use of data from the contractors' quality control sampling and testing programs in acceptance programs if the results from the States' verification sampling and testing programs confirm the quality of the material. The verification sampling and testing must be performed on independent samples obtained by the State or designated agent to verify the quality of the material. If the results of a State's verification sampling and testing program do not confirm the quality of the product, a dispute resolution system must be used to determine payment to the contractor.
The requirement for an independent assurance (IA) program will remain in place. The rule will provide the States more flexibility in designing their IA program. The IA program will allow the use of witnessing, split samples, proficiency samples, and equipment calibration as an independent check of the field sampling and testing procedures and equipment to assure that the testing is being performed properly by both the State and the contractor personnel.
Comments To The Docket
A notice of proposed rulemaking (NPRM) was published in the Federal Register on July 12, 1994 (59 FR 35493), in which the FHWA proposed to revise 23 CFR Part 637, Construction Inspection and Approval. A total of 50 commenters responded to the NPRM as follows: 35 State highway agencies, 1 local agency, 1 toll authority, 10 construction industry associations and contractors, and 3 Subcommittees of the American Association of State Highway and Transportation Officials (AASHTO). The major comments and the FHWA's response thereto are summarized as follows.
Supportive of Change
Twenty-six commenters expressed their support for the revisions to the regulation. Fifteen commenters provided comments without indicating support or opposition to the NPRM. The remaining nine commenters were generally opposed to the proposed rule.
Use of Contractor Test Results
Commenters expressed three related concerns over the required system of checks and balances employed when contractor test results are used in the acceptance decision: 1) requiring the use of independent samples instead of allowing either independent samples or split samples; 2) requiring the use of the F-test and the t-test (which are standard statistical tests for comparing the variances and means of two sets of data) because of the complexity of using the statistical tests; and 3) the perceived duplication of effort between the verification sampling and testing and the testing required by covering the contractor sampling and testing program in the IA program.
The overall intent of the program is to provide adequate assurance that the public is receiving the desired quality in the product produced by the contractor. The first level of assurance is provided by qualifying laboratories and testing personnel. This assures that the equipment and personnel are capable of performing the tests properly. The second level of assurance is provided by the IA program. This level assures that the testers and equipment remain capable of performing the tests properly. The third level of assurance is provided by verification sampling and testing. This level assures the quality of the product.
There appears to have been some misunderstanding of the total level of effort required. The rule as adopted gives the States wide latitude in designing the acceptance program. The system approach to IA assures the capabilities of all equipment and testers regardless of the number of projects or material quantities involved. A broad interpretation of the existing regulations would allow the system approach to IA. However, the final rule explicitly allows the system approach to IA. In those States that are performing a significant amount of testing on split samples and no testing on independent samples, testing on split samples would remain as IA sampling and testing; however, some verification testing on independent samples would be required to confirm the quality of the product. In addition, the verification of the quality of the material can be performed on a mix design or grading of material from a given source and is not limited to project-specific data.
Eleven commenters expressed concern over requiring the use of independent samples for the verification sampling and testing program. The commenters recommended that the use of split samples be permitted for the verification sampling and testing program. The commenters are concerned about the potential problems that may arise with differences in testing results caused by sampling errors.
There are three sources of differences between two test results, differences in the material, differences in test procedures and differences in sampling procedures. Split samples will only address the differences in test procedures and will only provide assurance that the contractor is performing the tests properly. In a balanced system it is also necessary to assure that sampling of materials is performed properly. It is our intent that the verification sampling and testing program be used to independently validate the quality of the material. Using independent samples will insure that all sources of differences are measured. The FHWA recognizes the need to ensure that each contractor performs the tests correctly; that is the reason for extending laboratory and testing personnel qualification requirements and IA program requirements to the contractor if the contractor's test results are to be used in the acceptance decision. The FHWA expects the testing variability between the contractor and the State to be held to a minimum by requiring the contractor's testing program to be covered by an IA program and requiring the testing personnel and laboratories to be qualified. The FHWA has changed the definition of "verification sampling and testing" and § 637.207 (a)(1)(ii)(B) to clarify the fact that the verification sampling and testing program is being used to validate the quality of the material.
Eight commenters objected to requiring the use of the F-test and t-test for verifying a contractor's test data. The commenters were concerned about the complexity of the F-test and t-test which would have to be used by field personnel and the lack of flexibility in allowing other comparison systems. The commenters requested that the regulation be revised to allow other types of comparison systems. The FHWA agrees with the concerns and has removed the requirement for a specific comparison procedure. Each State will have the latitude to develop its own verification system.
Three commenters -- two State Highway Agencies and one local highway agency -- objected to including contractors' testers in States' IA programs. The commenters are concerned over the additional resources involved in extending the IA program to contractor testing. If a contractor's test results are to be used in the acceptance decision, assurance must be provided that the contractor's testers and equipment remain capable of performing the tests properly. Some States are currently performing split sampling and testing on project sites to validate the contractor's test results. This split sampling and testing would meet the requirements for an IA program on contractor testing. This proposed requirement has been retained in the final rule.
Qualified Sampling and Testing Personnel
Four commenters specifically supported the concept of certifying testing personnel.
Two commenters wanted to change the term certified personnel to qualified personnel. The FHWA agrees with the comments since the goal of the FHWA is to have qualified personnel perform the testing. The term "certified" was deleted from the definition of qualified testing personnel.
Sixteen commenters expressed concern about the cost, specific requirements, and/or two-year implementation period for establishing qualification programs for testing personnel. To allow adequate time to develop qualification programs, we have extended the implementation time from two years to five years. If a State chooses to use a certification program as its qualification program, the FHWA is developing training material that can be modified for State use. The FHWA will also assist the States in adapting the material for their use.
Independent Assurance Program
Thirteen commenters objected to the proposal to remove the requirement that State highway agency (SHA) personnel perform IA testing. The States wanted to continue to perform IA testing as a means to maintain expertise in the materials sampling and testing area and maintain the credibility of their materials programs. Since materials sampling and testing are an essential part of determining the quality of the product that is obtained from the use of Federal-aid funds, the FHWA has an interest in maintaining the States' expertise and credibility. However, in cases where States are using contractor test results in acceptance decisions, the FHWA believes it is important that the States have the option of using consultants to perform IA testing. It is important to note that the final rule does not require a SHA to use consultants in the IA program, but simply gives SHAs the option to do so. The FHWA has added § 637.205(b) which requires States to maintain an adequate, qualified staff with the capability of overseeing the entire quality assurance program and specifically requires the States to maintain a central laboratory. This requirement is consistent with 23 U.S.C. 302 which requires each State to maintain an adequate highway department.
Three commenters requested further clarification on the use of the system approach in performing an IA program. The intent of the system approach to the IA program is to concentrate on assuring that the testing personnel and equipment remain capable of performing the tests properly, regardless of the location or number of projects covered by the equipment and tester. The system approach will permit an SHA to fulfill the requirement for an IA program by implementing a schedule of activities to cover equipment operations and tester competence. The activities may include calibration checks, split samples, proficiency samples, and observations. The schedules and type of activity would be based on the test procedure. In the system approach, the frequency of IA may be independent of the number of tests performed or the quantity of material tested. It is envisioned that the system approach will be especially useful in cases where one tester performs testing for more than one project during a construction season. The previous requirement for IA entailed sampling and testing frequencies based on individual project production. In addition, a State may choose to use the information developed from the IA program in the qualification programs for testers and laboratories. One commenter asked if the NPRM would allow a State to use a hybrid approach, which would include some frequencies based on project quantities and frequencies based on the overall system. This rule as written would allow that approach. It should be noted that the rule does not require a State to use this approach.
One commenter wanted the requirements for the IA program to be less stringent. The requirements in the final rule for IA have been made less prescriptive than the current regulations and give a State more latitude in designing its IA system. The existing regulation requires State personnel to perform the IA sampling and testing. The final rule would allow: (1) the use of accredited consultant laboratories in executing an IA program, (2) a system approach instead of a project approach, (3) proficiency samples instead of split samples, and (4) equipment calibration to cover the testing equipment.
Four commenters supported the proposed requirements for laboratory qualifications.
Eight commenters expressed concerns about the requirements for laboratory qualifications. The NPRM proposed to include by reference two paragraphs from the "Standard Recommended Practice for Establishing and Implementing a Quality System for Construction Testing Laboratories" (R-18) published by the AASHTO in the "Standard Specifications for Transportation Materials and Methods of Sampling and Testing." The commenters believed that R-18 was not appropriate for field laboratories. It was not the FHWA's intent that the entire R-18 standard be used for the qualification of field laboratories. Due to the confusion caused by specifying only a part of R-18, the rule has been revised to specifically list the minimum requirements for field laboratories and delete the reference to R-18.
Eight commenters wanted clarification of the requirements for accreditation of the SHA central laboratory. It is the intent of the FHWA that the accreditation program must meet the guidelines in ASTM E-994. In addition to the guidelines in ASTM E-994, we have two additional concerns: First, regarding the acceptability of the assessors; and second, concerning the scope of the on-site assessment. For an accreditation program to be acceptable to the FHWA, the assessor must be employees of the accrediting body and not employed by a laboratory which may compete for work with the laboratory being assessed. This would avoid any potential conflicts of interest. In addition, the on-site assessment must include a detailed review of the test procedures in which the laboratory is being accredited. The FHWA believes that only one laboratory accreditation program currently meets the above concerns, and that is the AASHTO Accreditation Program. As we understand the operating procedures of other accreditation programs, they allow reviewers to be employees of other testing laboratories and do not require the laboratory to demonstrate all the tests in which the laboratory is being accredited. If other accreditation programs can satisfy our concerns, we will approve them. Any inquires or requests for approval should be directed to the FHWA's Office of Engineering.
Six commenters expressed concern about the cost and implementation time necessary for accrediting an SHA central laboratory. The commenters believe that two years is too short a time in which to become accredited. At this time 30 SHAs are accredited by the AASHTO Accreditation Program (AAP). The FHWA contacted the AAP to obtain data on the average length of time required by the AAP to accredit a SHA laboratory after receipt of an application for accreditation. Based on the information supplied by AAP, the FHWA believes that two years is an adequate lead time for obtaining accreditation. The requirement for accreditation replaces the inspections by the National Reference Laboratories which are required by § 637.205 of the current regulation. The actual cost of accreditation to the SHA is the same as the cost of inspection program that it replaces. However, there will be some costs associated with developing the quality system for the initial accreditation for the SHAs. The rule provides flexibility to the SHAs to designate private laboratories to perform independent assurance tests and dispute resolution testing. Since the SHAs must review the qualifications of designated laboratories, the SHAs need to be qualified at the highest level, which is accreditation. Therefore, this final rule maintains the laboratory accreditation requirements as originally proposed.
Four commenters suggested changes to the definition of quality control. The definition of quality control was adapted from the definition in ANSI 90 and ISO 9000 which are the industry consensus standards for quality assurance. Therefore, the FHWA is retaining the definition as proposed.
Two commenters wanted to delete the word "accredited" from the definition of "qualified laboratories". There appears to be confusion over the use of the term "accreditation" since the NPRM used the word to describe two different levels of qualifications. The FHWA agrees with the comment because of the apparent confusion. The word "accredited" has been removed from the definition of "qualified laboratories".
Two commenters wanted clarification of the term "vendor." A definition of "vendor" has been added to insure that it includes suppliers of project-produced materials. It was the FHWA's intent that the rule cover only project-produced materials and not manufactured materials.
One commenter suggested changes to the definition of "quality assurance". The definition of "quality assurance" was adapted from the definitions in the ANSI 90 and ISO 9000 standards which are the industry consensus standards for quality assurance. Therefore, the FHWA has retained this definition as proposed in the NPRM.
One commenter suggested requiring random sampling. The FHWA agrees with the comment. In order for test data used in the acceptance decision to be properly analyzed, samples must be obtained on a random basis. Section 637.205(e) has been added to require random sampling.
One commenter was concerned with the wording of the definition for IA, which the commenter interpreted as requiring the IA to be performed by a consultant. As stated earlier, it is the FHWA's intent that the States have the option to perform IA sampling and testing themselves or have a qualified designated agent perform the testing. The definition in the final rule has been revised to reflect our intent.
Eight commenters requested a delay in issuing a final rule. Their major concern was over potential conflicts between this final rule and AASHTO's effort to develop guide specifications for Quality Assurance. The AASHTO effort is related to this rulemaking. However, the "AASHTO Quality Assurance Guide Specification" and the "AASHTO Implementation Manual for Quality Assurance" are in the draft stage and are still being reviewed. It may be some time before these documents receive full endorsement by AASHTO. Since the current regulations do not address the practice of using contractor testing in making acceptance decisions, the FHWA believes that it is necessary to proceed with the final rule. The commenters were also concerned that the SHAs did not have adequate time to comment on the regulation. The NPRM provided a 60 day comment period. All comments that were received by the FHWA, including the eleven received after the closing of the comment period, were considered and included in the analysis. In addition, the FHWA received comments from 35 of the 52 SHAs. Therefore, the FHWA believes that adequate time was provided.
Five commenters provided comments on the dispute resolution system. There were comments on both sides of the issue of whether the dispute resolution system should allow third party involvement. Three commenters were in favor of keeping the system in the State; two were in favor of using third parties. In the NPRM the FHWA proposed to permit the SHAs to determine how they wanted to set up the dispute resolution system. The FHWA is aware of cases where a dispute resolution system has worked well in both cases, so this proposal has been retained in the final rule.
Three commenters requested clarification of the terms "acceptance", "verification" , and "assurance". This rule requires an acceptance program which includes the establishment of qualifications of testers and laboratories and inspection of construction operations and testing performed by the SHA or its designated agent. Verification sampling and testing is used to validate the quality of the product. Independent assurance is used specifically to insure that the testing is performed correctly and that the equipment is in calibration.
Two commenters provided comments on the materials certificate. One commenter requested that the wording on the material certificate be revised from requiring the materials and operations to be in "conformity with the approved plans and specifications" to "reasonably close conformity to the approved plans and specification." The commenter was concerned about the added work of adding the individual material exceptions to the project plans and specifications to the materials certificate. The current regulation requires the material certificate to list all materials that do not meet the specifications. The FHWA reserves the right to review the materials certificate to determine if the materials are in conformity with the project plans and specifications. Therefore, the FHWA has retained the wording as proposed in the NPRM. The other commenter wanted to eliminate the requirement for the materials certificate. Section 637.201 limits the rule to projects on the NHS. In addition, § 637.207(a)(3) further limits the requirement for a materials certificate to projects that are subject to FHWA oversight reviews. This will eliminate the requirement for a materials certificate for the vast majority of projects. Since the cost of materials make up a substantial portion of each project and the information supplied by the materials certificate indicates the quality of the material, it is necessary to have the materials certificate in order to make an informed decision on whether to accept those projects for which the FHWA has retained construction oversight. Therefore, the FHWA has retained the proposed requirement for a materials certificate in this final rule.
One commenter indicated that the cost of implementing the regulation was high and a full regulatory review was needed. As noted below the FHWA has determined that this action is not a significant regulatory action under Executive Order 12366, Regulatory Planning and Review, nor significant under DOT Order 2100.5, Policies and Procedures for Simplification, Analysis, and Review of Regulations, and has concluded that a full regulatory evaluation is not required.
Costs to the States. Currently all States must have approved sampling and testing programs which include an IA program. In addition, all States are required to have their central laboratories inspected by the National Reference Laboratories. As indicated in the fee schedule for the AAP, the actual cost of accreditation itself for the SHAs is the same as the current inspection fees. The additional cost to the States for becoming accredited is in developing the quality assurance manuals which are required by the AAP. The justification for requiring accreditation is stated above. Since the vast majority of States have qualification requirements for their subsidiary laboratories, there would be no additional costs for the States that have these requirements. There would be minimal costs to those States that will have to develop qualification requirements for laboratories. There would be some costs in developing qualifications for testers. One aspect of tester qualifications is attendance at training programs. All States have some training for their technicians, but some of this training may have to be upgraded. However, as stated earlier, the FHWA has a training effort that is available to assist the States in setting up certification programs. The certification programs could be used in the States' establishment of tester qualifications.
Costs to the public. There would be no additional costs to the industry if a State chooses not to incorporate contractor tests into the acceptance system. If a State chooses to use contractor tests in acceptance decisions, contractors would be required to hire employees qualified in the appropriate tests and the State would be required to ensure that the contractors maintain a qualified laboratory or hire a qualified laboratory to perform the testing. When a State uses contractor quality control testing results in the acceptance decision, testing performed by the State is reduced. This reduction in testing by the State reduces the overhead costs in the State. However, any additional cost the contractors incur in performing the testing, including costs of obtaining qualified laboratories and testers, will be passed onto the State through higher bid prices. The cost savings by the State due to the reduction of testing by State personnel would be offset by the increase in bid prices charged by the contractors. As a result, the FHWA believes that the additional costs of these actions would be minimal.
One commenter was concerned because its Quality Assurance program is located in several documents and it did not want to consolidate the information into one document. The FHWA does not see the need for all the documentation of a State's Quality Assurance program to be in one document.
One commenter interpreted the NPRM to propose a requirement for a central laboratory and the commenter opposed such a requirement. The NPRM did not expressly propose to require a central laboratory; however, the NPRM did propose to require that each State's central laboratory be accredited by the AAP or a comparable program approved by the FHWA. For the reasons stated above, this final rule now requires a central laboratory.
One commenter was concerned about the effect of these QC/QA regulations on small projects. As indicated in the preamble of the NPRM, it is not the intent of the FHWA in this regulation to require the use of contractor testing in the acceptance decision. In addition, the rule expressly covers only projects on the National Highway system (NHS); projects not on the NHS can use other SHA procedures to accept materials. It is anticipated that the majority of small projects will not be on the NHS.
One commenter was against QC/QA procedures. The rule does not require SHAs to use statistical concepts or to use contractor-supplied test results in the acceptance decision. However, the rule does establish minimum requirements if an SHA chooses to use contractor tests results in the acceptance decision.
One commenter suggested a revision to the portion of § 637.207 concerning inspection to reflect the positive as well as the negative aspects of the quality of the product or construction. The section in the NPRM read, "The SHA shall inspect the product or construction or both for attributes that are detrimental to the performance of the finished product." The FHWA agrees with the comment. Section 637.207(a)(1)(i)(C) has been revised to reflect both beneficial and negative aspects of the quality of the finished product.
One commenter indicated that the regulation was too prescriptive. The rule, however, provides more flexibility than the existing regulation. The rule allows the use of contractor test results in making the acceptance decision and allows the use of consultants in the independent assurance program. Neither of these were allowed by the existing regulations. The regulation requires testers and laboratories to be qualified. However, it gives the States the flexibility to establish those qualifications. In addition, the final rule modified Section 637.207 to remove the requirement for a specific comparison procedure to validate the quality of the material. The rule clarifies existing policy and procedures and provides additional guidance on the use of contractor-supplied test results in acceptance plans.
One commenter questioned the title and purpose of the proposed rule, indicating that the rule covers materials and not construction. Over 50 percent of the cost of construction is the cost of the material. In addition, the rule requires each State to inspect construction to insure that the construction procedures do not adversely affect the properties of the material. Therefore, the title of this rule remains unchanged.