The catchphrase “out of sight, out of mind” too often can be used to describe the culverts located under our nation's roadways. Once installed, many lie forgotten and inconspicuous, conveying discharges from one ditch or waterway to the next. However, any component of the infrastructure neglected long enough will eventually make its presence known— many times in the form of a catastrophic failure and frequently resulting in large expenditures of emergency funds.
To gain a better understanding of the economic impacts of these failures, the Transportation Research Board (TRB) of the National Academies arranged for a study of culvert failures in North America. The resulting report, “The Economic Costs of Culvert Failure,” was written in November 2003 by Joseph Perrin Jr. and Chintan S. Jhaveri of the University of Utah. The American Concrete Pipe Association provided funding for the report.
By surveying U.S. and Canadian agencies, researchers were able to investigate current culvert design and selection practices, and analyze costs related to several documented culvert failures in order to quantify their costs, identify the need for documentation, and determine if risk of failure influences culvert selection.
Reviewing Existing Literature
To determine existing standards of practice for pipe design life, Perrin and Jhaveri reviewed current literature related to culvert design and selection. Referencing a U.S. Army Corps of Engineers report from March 1998, the researchers summarized their findings as follows:
Service life: For major infrastructure projects, designers should use a minimum project service life of 100 years when considering life-cycle design.
Concrete: Most studies estimated product service life for concrete pipe to be between 70 and 100 years. Of nine state highway departments, three listed the life as 100 years, five states stated between 70 and 100 years, and one state gave 50 years.
Steel: Corrugated steel pipe usually fails due to corrosion of the invert or the exterior of the pipe. Properly applied coatings can extend the product life to at least 50 years for most environments.
Aluminum: Aluminum pipe is usually affected more by soil-side corrosion than by corrosion of the invert. Long-term performance is difficult to predict because of a relatively short history of use, but the designer should not expect a product service life of greater than 50 years.
Plastic: Many different materials fall under the general category of plastic. Each of these materials may have some unique applications where it is suitable or unsuitable. Performance history of plastic pipe is limited. A designer should not expect a product service life of greater than 50 years.
The American Association of State Highway and Transportation Official's (AASHTO) 1991 Model Drainage Manual also was referenced, and recommendations for culvert design and selection presented:
Material selection shall include consideration of service life that includes abrasion and corrosion.
Culverts shall be located and designed to present a minimum hazard to traffic and people.
The detail of documentation for each culvert site shall be commensurate with the risk and importance of the structure. Design data and calculations shall be assembled in an orderly fashion and retained for future reference.
Culverts shall be regularly inspected and maintained.
The material selection shall consider replacement cost and difficulty of construction as well as traffic delay.
The selection shall not be made using first cost as the only criteria.
Select an alternative that best integrates engineering, economic, and political considerations.
The chosen culvert shall meet the selected structural and hydraulic criteria and shall be based on:
- construction and maintenance costs
- risk of failure or property damage
- traffic safety
- environmental or aesthetic considerations
- political or nuisance considerations
- land use requirements.
Because existing literature provides for a wide range of estimates for design life and method of culvert selection, the authors stress a need for documentation of failures and pipe performance on a national level as a means to better predict pipe life and costs.
Developing a Cost Analysis
Although initial culvert installation cost is usually considered in a cost analysis, Perrin and Jhaveri found that only a small number of the responding agencies applied some type of life-cycle cost analysis (LCCA) when designing and selecting culverts.
Neglecting to include future replacement costs of the culvert will result in an inaccurate representation of the actual cost. The authors proposed an equation to accurately determine the actual cost of a culvert over a 100-year time horizon. This equation takes into account costs related to initial installation, replacement costs, and user delay costs.
Perrin and Jhaveri sent out surveys to 57 agencies in the United States and Canada. Following are the author's observations based on the opinions submitted in agency responses:
Only four of the 25 agencies that responded to the survey stated that the agency performs a least-cost analysis for pipe material selection.
Different agencies assumed different life cycles for each material. Assumed life of reinforced concrete pipes and non-reinforced concrete pipes varied from 50 to more than 100 years, assumed life of corrugated metal pipes varied from 35 to 50 years, assumed life of high-density polyethylene pipes varied from 30 to 100 years, and assumed life of polyvinyl chloride pipes was 50 years. None of the respondents used vitrified clay pipe so no data were available.
Five of the 25 agencies provided data regarding cost of pipes. Results indicated that cost varied from agency to agency and also varied by material and size of the pipe. Some agencies also mentioned that the cost of the pipe was negligible as compared to the cost of earthwork, traffic control, and installation/labor costs.
Only two of the responding agencies considered the risk of failure while performing a cost analysis as recommended by the 1991 AASHTO Model Drainage Manual.
Only five agencies considered emergency replacement costs. Seven agencies considered user costs that could be incurred in the event of a culvert failure. This is confusing, as only four states responded that they perform a LCCA.
Fifteen of the 25 responding agencies document culvert failures occurring within their jurisdiction. However, the level of detail in documenting these failures greatly varied from agency to agency. Most written documentation of the specific culvert failure was unavailable.
Some agencies did not allow the use of a certain pipe type above a particular annual average daily traffic (AADT) level. From the responses, 42% of the agencies stated that AADT was a consideration in pipe material selection, often only allowing rigid pipe on high-AADT roads. This identifies that risk and traffic delay is indirectly considered in pipe selection, often without conducting an LCCA.