Construction

Performance Specifications Strategic Roadmap: A Vision for the Future

[Archived] Chapter 2. Performance Specifications

The Umbrella

Performance specification (PS) is an umbrella term incorporating performance related specifications (PRS), performance-based specifications (PBS), and warranties. In the broadest terms, a performance specification defines the performance characteristics of the final product and links them to construction, materials, and other items under contractor control. Performance characteristics may include items such as pavement smoothness or strength, bridge deck cracking or corrosion, chip seal stone retention, embankment slope stability, etc.

Snapshot. When future performance of a product can be estimated using key construction tests and measurements linked to the original design via modeling and life cycle costs, the specification structure is commonly described as performance-related or performance-based.³ When the condition of the product is measured after some predetermined time, the specification structure is commonly known as a warranty. When the final product is described in terms of component materials, dimensions, tolerances, weights, and required construction methodology-equipment type, size, speed, etc.-the specifications are commonly known as method or prescriptive specifications. Currently, method specifications are the predominant specification type used in U.S. highway construction.

PRS

In softened technical terms, PRS are specifications that use quantified quality characteristics and life cycle cost relationships correlated to product performance. In management terms, a PRS is the bridge between design, construction quality, and long-term product performance. So how does one determine that a specification is performance related? Some fundamental and suggested requirements are offered to sort this out.

PRS: Fundamental Requirements

Quality Characteristics and Accountability. Critical quality characteristics should be readily measurable and clearly tied to product performance. Construction contractors should be held accountable only for those quality characteristics under their control.

Performance Predictions. Prediction tools, including modeling and databases, should be verified, calibrated, validated, and otherwise made appropriate for local conditions.

Life Cycle Cost Analyses (LCCA). Life cycle cost analyses should be used to compare the as-designed product section to the as-built section. The LCCA should be based on a clear, well-documented, and realistic preservation, rehabilitation and maintenance decision tree.

Acceptance Plans. Acceptance plans should be statistically based with clearly defined risks. If necessary, pay determination should be made in a timely fashion to allow for prompt corrective action. Sampling and testing plans should properly address material, operator, and testing variability and improve confidence in the results.

Simple and Clear Language. Performance-related specifications should be written simply, clearly, and succinctly for today's busy construction workforce.

PRS: Suggested Requirements

Add Performance and Subtract Method. As PRS end-result criteria are added to a contract for a specific quality characteristic, they should be accompanied by a corresponding reduction in prescriptive or method elements, giving the contractor more freedom to innovate, improve quality, and clarify roles and responsibilities. Add density and eliminate roller requirements, for example. Or add in-situ smoothness and eliminate concrete paver string line requirements.

Quick and Timely Testing. Testing should incorporate standardized tests using nondestructive techniques to measure the product in situ, better quantifying the quality characteristics and enhancing 24-to-48-hour, if not instant, turnaround of information. This also could be the driver to harness computer technology, such as PDAs (personal digital assistants), wi-fi (wireless fidelity) networks, voice recognition, and high-speed linkage to asset management systems.

Process Control. The contractor should be given reasonable latitude to develop and implement a process control plan that can be verified by the transportation agency, especially for those quality characteristics included in the acceptance plan.

Mechanistic Models. Performance prediction techniques used in PRS should be based on mechanistic models and be the same models used in the design process. Asset management systems should track the same assumptions used in the design and construction process.

LCCA and User Costs. User costs should be considered in developing appropriate pay factors. The impact can be high, however, and will require sound judgment when applied. Both the owners and the contractors need to understand the impact on customer satisfaction.

Warranties

Warranties can be divided into two areas: materials and workmanship (M&W) warranties and product performance warranties. M&W warranties call for contractors to correct defects in work elements within their control. The M&W concept is referenced in many State regulations and codes, but it is not directly referenced in highway specifications and has been invoked rarely.

The performance warranty is a recent concept and requires the contractor to correct defects if the product does not perform to some desired quality level over a certain time in service. Product performance warranties are somewhat controversial, exponentially so as the length of the warranty period extends beyond three years. The controversy stems from the concept of risk allocation and the financial burdens that accompany partial or complete product failures.

Following is a step-by-step process for developing a warranty:

1. Establish what gain is expected and how success of the program will be measured.
2. Define the product service life.
3. Establish a warranty period and describe the condition of the product at the end of the warranty, including expected remaining service life.
4. Describe the sampling and testing plan that will be used to monitor quality during construction and measure quality at the end of the warranty period.
5. Eliminate method or prescriptive requirements that conflict with performance requirements or intent. This includes material selection, mix designs, etc.
6. Establish some thresholds where warranties are invalidated-traffic, weather, inadvertent maintenance, etc.
7. Establish a contract bonding, insurance, or retainer requirement to hold the contractor financially accountable.
8. Establish a repair protocol should the product show early distress.
9. Establish a mediation board to resolve conflicts.
10. Pay according to a pre-determined pay schedule, including incentives and disincentives.
11. Monitor, measure, and feedback into the performance models.

M&W warranties of less than three years generally require the contractor to focus on construction quality. With a performance warranty, the contractor may have more latitude in selecting materials, processes, and design choices. This requires the contractor to have much more than a working knowledge of the product. This means sorting through various combinations of materials or manufactured products and pricing alternate products. Should the contractor provide a higher-cost, longer-life, more-than-meets-the-warranty threshold product or a lower-cost, shorter-life, just-meets-the-warranty product? What is the risk versus costs? What impact will this have on contract award? Price obviously matters in a low-bid contract, but it also matters in emerging procurement options such as design-build and best-value contracting.

This process is the reverse of the PRS process, in which the transportation agency makes the decisions on material type, layer requirements, etc. Not surprisingly, however, both parties need a working knowledge of what drives performance. The fundamental approach in PRS may be applied by a contractor in response to a warranty requirement as well.

PRS and Warranties

The comparison between performance-related specifications and warranties is a natural. Both address product performance and improvement in contractor end product compliance and innovation, and both have an impact on the interrelated issues mentioned previously. The impacts on the contractor and the transportation agency, however, are different in each scenario. The following chart outlines the issues and the requirements under each.

In the future, agencies will have two different approaches to address quality. Each method has its positives and negatives.

Method Specifications

One of the most difficult issues facing the adoption of performance specifications is the impact they have on method or prescriptive specifications. A recent review of select transportation agency standard specifications showed that use of method specifications remains common, with more than 400 prescriptive requirements in the standard specification book. They vary from minimum tire pressure in a rubber-tired roller to paver string line requirements. The difficulty comes when the specification includes both a prescriptive and end-result requirement-a roller specification and an end-result density requirement, for example.

Method specifications have been a mainstay in transportation construction for many years. What is the most commonly accepted principle behind a method specification? If the contractor follows the prescription, then the work product has a high probability (if not a sure bet) of being accepted by the agency and a good probability of performing well in service. What are some of the other impacts of method specifications?

Decision Aids. A method specification tells the contractor exactly what the agency has decided about a certain topic.

Knowledge Tools. Method specifications tell both parties what is considered good practice and, by omission, what is not good practice.

Minimum Acceptable Values. Terms like "no less than" or "at least" show the lowest allowable value that will be accepted by the agency.

Restrain Decision Makers and Force Fair Treatment. Method specifications give BOTH parties protection over arbitrary decision making. In fact, they serve to prevent arbitrary decision-making by the agency as much as the contractor.

Difficult to Change. Method specifications are difficult to change once imposed and set into practice, which is both good and bad. It is good in that training, equipment procurement, and testing programs can be developed around the concepts, but it is bad in that an obviously minor or insignificant method specification is often difficult to remove.

Unintended Negative Consequences. It may be that the agency wants to allow flexibility but is constrained by the method requirements. The contractor, in turn, may want to introduce an innovative concept but is inhibited by having to address each method specification point by point.

Red Tape. While one method specification may be judged as a safeguard to both parties, when does a series of method specifications become overbearing-the definition of red tape?

Minimum Quality Equals Maximum Quality. While method specifications clearly define MINIMUM acceptable behavior, they may also, as a result of the low-bid process, define MAXIMUM performance levels as well.⁴

Distinguishing a Quality Contractor. Method specifications (in harmony with low-bid contracting) have a way of grouping all contractors under one quality umbrella. They reduce the ways contractors can differentiate themselves in the quality arena.⁵

Specifications and Contracts

Several key questions need to be answered about how these three types of specifications will work in the future. Will the highway specification book be filled with performance specifications and void of all method requirements? Will the book contain a blend of specifications? Or will it have different types of specifications for different types of contracts-method specifications for less-critical subjects and performance specifications for design-build, for example. Or will method specifications always be used to control those illusive long-term durability issues?

A window to the future might be the European Union (EU) process for improving trade and competition among European countries. The EU is providing the stimulus for the highway industry to develop functional highway specifications for contracts (tenders). Functional specifications are a cross between end-result and performance specifications and define the final in-place product with some specificity. Method specifications gradually are being removed, especially those that relate to material composition and installation procedures. Industry and government are working on many of these specifications and acknowledge to the complexity of the issue.

In addition, many European countries have moved to functional contracts with specific language on performance of the in-place product over time. This includes everything from retroreflectivity of a pavement marking in service to litter pickup in rest area waste containers. The United Kingdom's Highways Agency bases 80 percent of a contract decision on quality factors and 20 percent on cost. In 2003, this will change to a 100 percent quality award. The dollars will be negotiated after the award, and the specifications all will be functional.

Some European countries are increasingly using design-build-operate-maintain (DBOM) contracts that may extend for 20 to 30 years. These contracts are performance based, including eventual turnback to the agency at a required performance standard or benchmark.

The drivers in Europe to move to these types of contracts are the same as those in the United States:

• To pull the private sector into the innovation equation.
• To address the reduction in government personnel.
• To allow the remaining governmental workforce to focus more on performance requirements for the transportation system.

Is everybody in Europe happy about this movement? No. Is everybody in Europe seeing the long-range vision the same way? No. But they are working on the issue and already are seeing fruits of their labor in several key technology areas.

What does this mean to the United States? Is Europe a window to our future? Maybe. Should the United States copy what Europe is doing? Not at all. The European construction industry is structured differently than the U.S. industry, and the social implications cannot be dismissed. But it does mean that the United States has a real-life laboratory to learn about performance specifications and performance contracts. With a watchful eye, the United States could learn from Europe's organization efforts, experiment with its specifications, and dismiss those that would bear little fruit.

Expected Benefits

It makes no sense to start something without clear reasons and expected benefits. Developing and implementing performance specifications offers many potential benefits. The following are some of the most important:

Improved Design-to-Construction Communication. Performance specifications could more directly connect design requirements with construction, assuring that both parties communicate effectively.

Rational Pay Factors. Pay factors could be more accurate, rational and defensible, as they would be based more on processes and less on bartering.

Improved and Focused Testing. Testing would focus on those characteristics that relate to performance.

Improved Tradeoff Analyses. Performance, quality, and costs could be uniquely connected through modeling and life cycle cost analyses with a much better way to analyze tradeoffs.

Improved Understanding of Performance. Performance specifications could lead to a better understanding of those quality characteristics that relate more directly to product performance.

Improved Quality Focus. Performance specifications could lead to improvement in the overall quality of the product in areas that caused problems previously.

Clearer Distinction in Roles and Responsibilities. Performance specifications could help clarify changes in roles and responsibilities between the transportation agency and the contractor, as well as define the levels of risk that each would carry.

More Innovative Environment. By being less prescriptive, performance specifications could create an environment that encourages innovation.

All of these benefits would apply to either PRS or warranties.

³ From this point on in the report, PRS is assumed to include PBS. For all practical purposes, the distinction between the two is not relevant at the program level, but should be considered at the research and engineering level. For more information, see http://gulliver.trb.org/publications/circulars/ec037.pdf. (.pdf, 1 mb)

⁴ This is one reason that incentive clauses were created and added to method specifications.

⁵ This is yet another reason why incentive clauses were created.