Construction

Performance Specifications Strategic Roadmap: A Vision for the Future

[Archived] Chapter 1. Examining the Issues

A Brief History

In the early 1900s, the idea of mobility was paramount in the minds of the American people. As travel modes transitioned from ships to wagons to trains to automobiles, the road became the focal point of transportation. Toll roads connected major cities and industrial areas. Public road-building jurisdictions were small, numerous, and unconnected. In the art of road building, little was known about factors that contributed to the success or failure of the road. Under these circumstances, the first option for governing agencies was to require a maintenance guarantee. The contractor promised to do any needed road maintenance and repairs for a specified time period after construction.

Formation of the American Association of State Highway Officials (AASHO) in 1914, followed by a general uprising from the contracting community over warranties and proprietary items, led to the development of a new order of road-building specifications and brought about a certain level of uniformity in State specifications. This was the birth of method specifications. In a method specification framework, work was done in a prescribed way, with maximum control in the hands of the agency. The contractor followed the script - provided materials, equipment, and followed directions.

In the mid-1960s, after noting rather high construction and materials variability in the controlled AASHO Road Test¹, industry leaders determined that method specifications by themselves did not properly control the construction process. Practitioners asked researchers to come up with a new way of addressing these issues. FHWA tackled the specification issue, noting that a properly crafted specification should clearly answer five questions:

• What do we want?
• How do we order it?
• How do we measure what we ordered?
• How do we know we got what we ordered?
• What do we do if we don't get what we ordered?

While most practitioners believed that method specifications reasonably captured best practices at the time, they did not outline an effective sampling and testing program to determine overall compliance. With equipment and material requirements defined, many believed that method specifications inhibited innovation and could not deal with rewarding a contractor for "better-than-minimum" practice. Finally, they also believed that method specifications as written could not consistently deal with work that was outside the bounds of "reasonably close conformance."

Could a new specification with an "end-result" approach lead to more innovation? Could it lead to better handling of non-complying material and a more accurate assessment of in-place quality? As practitioners asked themselves those questions, the construction research approach of the 1970s became, "Don't tell them how to do the job; tell them what you want and let them go."

In theory, an end-result specification should allow contractors more freedom to implement their own procedures, choose their own equipment, and conduct site-specific process control programs. The transportation agency would allow this freedom, but establish a more structured sampling and testing program on the in-place product. And so the journey toward performance specifications began.

While the concept appears simple, researchers posed many difficult questions as a prelude to their work:

1. What is an in-place property? Thickness? Density? Modulus? Stiffness? All of them?
2. What in-place properties most directly influence product performance?
3. What tests best measure these properties? How many? With what variability?
4. What is an in-place product? The base or the pavement? The paint or the stripe?
5. What is product performance? Distress? Remaining life? Failure?
6. What is the value of the product if one or more of the tests show non-compliance? Rework? Replace? Accept at a reduced value? And what if the tests show exceptional results-much better than specified?
7. What elements of construction are contractor-controlled versus agency-controlled?
8. What elements of construction are totally beyond the contractor's ability to control or influence?

A major output from this research was statistically based quality control specifications. They addressed the issues of testing and test variability, sample size, lot size, estimates of the total population, percentage within limits, and pay factors. While most agreed that this new approach did a better job of addressing contractor compliance, it did not necessarily address product performance. Why? For the most part, the specifications measured what COULD be measured, not what SHOULD be measured. The drivers of product performance and the test procedures needed to measure the performance characteristics did not exist.

The critical connection to product performance came about in the early 1980s. What if it were possible to connect product performance to a life cycle cost analysis (LCCA)? LCCA establishes performance relationships between the designed product, costs, and future preservation, maintenance and repair strategies. For pavements, this appeared to researchers to be a key approach. If the design calculations were examined closely, they should include factors that focus on key performance characteristics-strength, thickness, modulus, etc. Would it be possible to recalculate the LCCA using as-built test results and then compare them to the original LCCA? This would provide a ratio that linked costs and time-to-rehabilitate. For example, if the pavement was built one inch shy of the design thickness, the as-built LCCA would show that the projected pavement life was reduced by a certain percentage, triggering an earlier rehabilitation strategy. Researchers were now on their way to determining analytically the performance aspect of the pavement. The ratio of the as-built costs to the as-designed costs also gave them a more rational approach to pay factors. And, of course, this puts pressure on the designers to assure that they are theoretically correct with formulas and assumptions proven over time.

Is it about contractor compliance? Or about product performance?

The process gets more complicated when other pavement performance characteristics-such as density, smoothness, skid, segregation, stripping, durability, and noise-are considered. All of these characteristics are important, although maybe not equally. All have an impact on pavement performance and are interrelated. Some characteristics are inputs into the equation or model used to design the pavement originally. Some are not. Some are clearly under the control and responsibility of the contractor. Some are not. Some can be tested quickly and easily, while others involve test that are slow and late in the process. And some performance properties are tied to agency specifications that require the contractor to use a certain technique or material.

For the past 20 years, FHWA, AASHTO, and contractors have been on a journey to sort out and understand this complicated issue. Performance-related specifications have been identified as a high-priority area in every major research plan developed over the past decade by FHWA, AASHTO, and various industry groups. Much of the current national research has focused on asphalt and concrete pavement systems. But much of the above could relate to bridge decks or other highway products.

While the performance specification framework for pavements is relatively mature, this leads to questions about how this methodology translates to other highway products. Bridges, earthwork, retaining walls, and many other highway products do not have direct connections to a life cycle cost analysis methodology or to a clear design formula or model. And while the performance specification framework for pavements aids in determining values, it does not directly address roles, responsibilities, and ways to create an innovative atmosphere.

In 1991, a milestone asphalt pavement study tour of European countries initiated a U.S. discussion of new contracting mechanisms, including the possible impact that warranties might have on product performance and contractor innovation. In 1995, FHWA clarified its approach to warranties and set the direction for their future application. Warranties appeared to be another way that agencies could address product performance. While PRS concepts were emerging for experimental use in concrete pavements, agencies could, if they wanted to, simply invoke a warranty clause and bypass all the apparent complexities associated with PRS-and in the process transfer risk to the contractor.

Many activities have been initiated over the past several years, further validating interest in new approaches to specifications. In 2000, the National Cooperative Highway Research Program, under AASHTO sponsorship, initiated a contract on performance-related specifications for asphalt mixes, continuing work begun on the WesTrack² Superpave project.

Also in 2000, the Indiana DOT completed the first experimental project based on FHWA's model PRS for concrete pavement. In 2002, an international scanning study focused on the growing use of long-term asphalt pavement warranties in Europe that allow routine maintenance and preservation during the performance period. In 2003, another international scan on superior materials found significant movement toward performance specifications (functional specifications) in several European countries.

Interrelated Issues

All of these factors contribute to a realignment of conventional roles and responsibilities, and provide new opportunities to examine product performance and construction contracting.

Specification Language. The objective of all specification writers is to translate the transportation agency's intentions into clear, legally defensible instructions for the contractor. Today, more than ever, specification writers recognize that this objective also must allow the contractor to exercise ingenuity and creativity in complicated rehabilitation and reconstruction projects. Projects built under traffic conditions complicate the normal sequence of operations, equipment selection, site access, construction speed, and safety to the worker and traveling public. Less-prescriptive specifications could be a way to allow the contractor to exercise more creativity to meet the demands of a particular project. This is even more justifiable for the growing number of unique projects with special quality and performance requirements-integration of work zone management with construction sequencing, for example.

Inspection Staff. State DOTs-indeed, nearly all public agencies-have seen a dramatic decrease in the numbers and experience levels of inspectors and engineers in their workforce. This has led to more contractor quality control programs and agency testing of as-built products.

Construction Duration and Speed. In urban areas where traffic congestion is a major issue, nearly all construction products go into service immediately. This is a radical departure from the practice of allowing 100 percent completion before the project opening. New approaches to construction product testing and construction acceptance have become necessary to satisfy the needs of the traveling public.

Public Response to Highway Quality and Performance. The public, including elected officials, are asking why so many highway facilities always seem to be under construction and need so much attention. They compare highway products with commercial products and insist on more contractor accountability through a warranty or guarantee or even open product testing. "Get in, stay in and do it right, get out, and stay out" will be driving the industry for years to come.

¹ A 7-mile long full-scale test road near Ottawa, IL, aimed at gathering data on significant variables affecting pavements and short-span bridges. The project was designed and managed by the Bureau of Public Roads, the American Association of Highway Officials and the Highway Research Board in the late 50s and 60s and still provides valuable pavement data.

² WesTrack refers to an experimental road test facility constructed in Nevada that continues the development of performance related specifications for hot mix asphalt.