AEP moves toward implementing a reliabilitybased foundation design approach.
The design of foundations for transmission line structures provides many challenges. A transmission line is a complex system of interacting components: structures, foundations, conductors, insulators, shield wires and hardware. A line may consist of many structure types, which can be supported by an equally wide variety of foundation types. Additionally, a line may traverse terrain that changes from flat to rolling hills or mountains. A line also may traverse different geologic zones, which can present unique geologic challenges, such as karst areas, unstable slopes, and mined and unmined areas. Thus, the foundation design process involves the application of both technology and art in each step of the process. This is especially true in the development of the geotechnical design parameters needed for foundation design.
Allowable Stress Design
As is the case with all technologybased design processes, research and development are performed on a continuous basis to bring new and innovative concepts and improvements to the design process. This is exactly what happened to the process for designing foundations for bridges and what is happening to the process for designing foundations for transmission line structures. This closely parallels similar transitions in the concrete and steel building industries.
Over the past 30 years, the Electric Power Research Institute (EPRI) sponsored many research projects to advance the foundation design. Prior to EPRI's research efforts, the most common design method used in practice was the allowable stress design (ASD) approach.
In the ASD approach, foundation designers assume a foundation size, compute the ultimate design capacity of the foundation and determine the nominal working capacity by dividing by a safety factor. If the working capacity is less than or greater than the applied load, the designer assumes a new foundation size and then repeats the process until an acceptable design is achieved.
The biggest challenge facing the ASD designer is the uncertainty in establishing a safety factor because there is no systematic method for establishing safety factors. The selection of safety factors is dependent on the foundation design engineer's background and experience.
There are a variety of foundation design models, including MFAD, caisson, LPILE and Hansen. EPRI recently conducted a survey that showed that factors of safety used in practice today varied from 2.0 to 4.0 for the MFAD, 1.0 to 2.5 for the caisson, 2.0 to 3.0 for the LPILE and 1.1 for the Hansen foundation design models. These models are used the most to design drilled shafts for tubularsteel single poles. The range of safety factors results in significant differences in foundation construction costs and reliability. Thus, in accordance with the normal evolutionary process of improving technology, EPRI and the Federal Highway Administration moved forward with the development of the reliabilitybased (RBD) approach for the design of foundations for transmission line structures and bridges, respectively; and in recent years, American Electric Power (AEP) has begun a similar transition.
Calibrate Design with Testing
In simple terms, the RBD approach provides foundation designers with a rational decisionmaking framework with several options:

Consider the variability of foundation loads and foundation strengths

Coordinate the level of reliability between foundations and other line components

Optimize foundation costs by establishing a uniform level of foundation reliability.
Among the current RBD options, AEP is adopting one in which the 5% lower exclusion limit foundation strength must be equal to or greater than 50year return period load events, such as extreme wind, wind on ice covered wires, and structure conditions. In turn, the 5% lower level exclusion limit design strength for a given foundation design model is equal to the foundation nominal strength times a calibrated resistance factor. The question then becomes, “How do foundation designers establish resistance factors for each foundation design model they are currently using in practice?” The answer is each foundation design model must be calibrated against the results of fullscale foundation load test results. The calibration procedure provides a consistent and statisticsbased method for establishing design model strength factors:

Collect data on available fullscale foundation load tests, including insitu soil and rock properties, laboratory test data and field data of applied load versus displacement. A database of at least 20 to 25 fullscale load tests is needed for acceptable statistical correlations.

Using the data collected in Step 1, determine the nominal loadcarrying capacity of each test foundation.

For each test foundation, plot the applied groundline load versus the predicted capacity, as determined by the foundation design model being calibrated.

Perform a least square fit to the data plotted in Step 3 and determine the slope of the leastsquare fit line and the coefficient of variation.

Based on the data determined in Step 4, compute the resistance factor for the design model being calibrated.
Fortunately, EPRI has performed much of the research work needed to establish resistance factors for singlepole transmission line structure foundations. Resistance factors, used to design drilled shafts for tubularsteel single poles, have been established by the aforementioned calibration process for the following commonly used foundation design models for the design of singlepole drilled shafts:
 MFAD 5.0 has a factor of 0.63
 Caisson has a resistance factor of 0.43
 Hansen has a resistance factor of 0.71
 LPILE 5.0 has a resistance factor of 0.84.
In essence, the RBD approach uses resistance factors that are based on an evaluation of fullscale foundation load tests, while the ASD approach uses factors of safety that are subjective and, for a given foundation design model, can vary significantly depending on the foundation design engineer. Thus, the implementation of the RBD approach will result in a relatively uniform level of reliability and optimized foundation designs and construction costs.
Cost Implications
ASD and RBD foundation designs can vary significantly from one another for a given set of design loads and subsurface profiles. For comparison, consider a 6ft (1.8m)diameter drilledshaft foundation for a steel single pole using both the RBD and ASD approaches and the MFAD 5.0 design model. The foundation performance design criteria are as follows:

Allowable topofconcrete displacement of 4 inches (102 mm)

Allowable nonrecoverable topofconcrete displacement of 2 inches (51 mm)

Allowable topofconcrete rotation of 2 degrees

Nonrecoverable topofconcrete rotation of 1 degree.
In the application of both the ASD and RBD approaches for the design of drilled shafts for tubularsteel single poles, the designers must establish foundation performance criteria such as the allowable topofdrilledshaft lateral displacement and rotation. The establishment of reasonable performance criteria is important in that overly stringent performance criteria, of either lateral displacement or rotation, can have a significant effect on shaft size.
A subsurface profile consists of a 7ft (2.1m)thick layer of mediumdense fine sand on top of an 8ft (2.4m)thick layer of stiff to very stiff clay layer, which lays on top of a poor rock layer. Based on this subsurface profile, the RBD approach, using MFAD and a strength factor of 0.63, results in a depth of embedment of 21 ft (6.4 m). The ASD approach, using MFAD and safety factors of 2, 3 and 4 results in depths of embedment of 22 ft (6.7 m), 25 ft (7.6 m) and 27 ft (8.2 m). Thus, depending on the assumed safety factor, the RBD design depth varies from 1 ft (0.3 m) to 6 ft (1.8 m) less than the ASD design depth. Using an estimated construction cost of US$1,000/ft ($3,290/m), the cost savings for a single foundation would vary from $1,000 to $6,000.
RBD Experience
AEP first used the RBD approach in 2007 on a 345kV doublecircuit transmission line project in Kenedy County, Texas, U.S. Since then, the RBD approach has been successfully employed on several other projects, including the AEP Electric Transmission Texas/Competitive Renewable Energy Zone (ETT/CREZ) project.
The ETT/CREZ project is a 400mile (644km) 345kV transmission line being built in Texas and is an excellent example of applying RBD to the design of directembedded poles. Since the available fullscale foundation load test database did not contain any tests for directembedded poles with concrete backfills, the ETT/CREZ team decided to conduct two tests. The results of the two fullscale foundation load tests verified the use of the MFAD 5 foundation design model and its calibrated strength factor of 0.63.
Seize the Advantages
The RBD approach for designing foundations for transmission line structures provides many advantages over the ASD approach, such as optimized foundation designs and a relatively uniform reliability level.
T. David Parrish (tdparrish@aep.com) is a manager in American Electric Power's Transmission Line Standards Group in Gahanna, Ohio, U.S. His group is responsible for all transmission linerelated engineering, material and construction standards, specifications, guidelines and drawings. Parrish is a civil/structural engineer, and has structure and foundation design experience in the utility industry. He holds a BSCE degree from the University of Akron and is a registered professional engineer in 13 states.
J. Kelly Bledsoe is a supervisor in American Electric Power's Transmission Line Standards Group in Roanoke, Virginia, U.S., responsible for structure and foundation design within AEP Transmission. He joined AEP in 1990 and his duties have included transmission line, structure and foundation design. He holds a BSCE degree from Virginia Military Institute and is a registered professional engineer.
Anthony M. DiGioia Jr. (tony@digioiagray.com) is president of DiGioia, Gray & Associates, LLC. He received his BS, MS and Ph.D. degrees from Carnegie Mellon University and is currently an adjunct professor in the civil environmental engineering department at Carnegie Mellon University. DiGioia has managed EPRI foundation research projects and is a leader in the development and implementation of the RBD approach. He also lectures for the University of Wisconsin Continuing Education courses on foundation design.
Foundation Design Models
The four main foundation design models used include MFAD, caisson, LPILE and Hansen.

The MFAD foundation design model was developed by the Electric Power Research Institute (EPRI) for the design of drilled shafts for tubularsteel poles and for the design of direct embedment of tubularsteel poles using granular, cohesive soils or concrete backfills. In the MFAD model, the applied moment, horizontal shear and compression loads are resisted by the combination of soil and rock strata lateral pressures, circumferential verticalside shear resistance, and base lateral shear and moment resistances. The program initially computes nominal capacities. Design capacities are based on a calibrated resistance factor of 0.63. The program also provides deflections, rotations, bending moments, shear force and lateral soil response over the length of embedment and the base of the drilled shaft. MFAD is licensed by DiGioia, Gray & Associates, LLC on behalf of EPRI.

The caisson foundation design model was developed by Alain Peyrot and Tarun Naik of the University of Wisconsin for the design of drilled shafts for tubularsteel poles. In the caisson model, the applied moment and shear loads are resisted solely by soil strata lateral pressures. Caisson is licensed by Power Line Systems.

The LPILE foundation design model was developed by Lynn Reese of the University of Texas. LPILE is a specialpurpose program based on procedures for analyzing a pile or drilled shaft under lateral loads. The applied moment and shear loads are resisted solely by soil and rock strata lateral pressures, which have nonlinear py relationships. The program computes deflections, bending moments, shear forces and soil response over the length of the pile. LPILE is licensed by Ensoft Inc.

The Hansen model was developed by J.B. Hansen in 1961. A detailed description of the Hansen model is contained in his book Earth Pressure Calculation, published in 1961 by the Danish Technical Press. In the Hansen model, the applied moment and shear loads are resisted solely by soil strata lateral pressures. Implementation of the Hansen model can be done in the form of a spreadsheet by programming the lateral pressure equations developed by Hansen in his 1961 publication.
Companies mentioned:
American Electric Power www.aep.com
DiGioia, Gray & Associates, LLC www.digioiagray.com
Electric Power Research Institute www.epri.com
Electric Transmission Texas www.ettexas.com
Ensoft Inc. www.ensoftinc.com
Power Line Systems www.powline.com