Framework for Risk-Based Electric System Deep Foundation Design

"Annually thousands of deep foundations are constructed across North America supporting high voltage transmission line structures. However, a recent state of the practice survey by the Deep Foundation Institutes’ Electric Power Systems Foundation Working Group discovered that the electric power utility industry lacks consistent foundation design guidelines. The survey identified a wide variability in design assumptions, a minimal understanding of risk based analysis, and a need to standardize design methodologies. This paper summarizes the state of practice in electric system foundation design and proposes a framework for risk-based foundation design by accounting for uncertainty from both loads and resistances. Loads derived from electric system codes do not necessarily provide a compatible level of reliability for the design of foundations, as they are a mixture of probabilistic and deterministic values. Load uncertainty is derived from variability in weather conditions, structural models and material properties, typically accounted for by code derived overload factors. Resistance uncertainty arises from variability in natural soil properties, resistance models, subsurface construction techniques, material sampling methods, and limited observations. A consistent methodology is offered relating nominal resistance capacity (strength) to structure loads and service state criteria. Discussion also includes the need for risk-based assessments to account for non-structure load influences (fluvial loads, scour ground loss, liquefaction, etc.).INTRODUCTIONNearly 40 years ago the Electric Power Research Institute (EPRI) embarked on a major research effort to optimize reliability and economics of transmission line foundation design. Strategies focused on improvements in design models, site characterization, exploration strategies and testing to reduce uncertainty and variance in analysis. Transmission line engineers realized that safety and reliability advancements in design based on probabilistic approaches, with a focus on the unique nature of transmission line, could better control risk and cost.Some transmission line professionals took full advantage of these advancements, particularly where research led to improved software and models. Yet, even after two decades of extensive research, by 1995 the industry still had made little progress in changing design habits. The majority of designers were still characterizing sites via deterministic methods, using loads that may or may not relate to a consistent probability of occurrence, and applying a traditional global factor of safety approach (Phoon, et al. 1995). The following two decades have seen even greater developments in software and reliability-based analysis methods, but old habits seem to die hard as recent surveys show less than half of design professional quantitatively assess risk in their design process."