By Evangelia Ieronymaki, Ph.D., P.E., Associate Director and Senior Lecturer in the M.Sc. Project Management, SPS
Urban construction increasingly depends on deep excavations for foundations, transit infrastructure, underground utilities, and other critical projects. As cities grow denser, these excavations often take place only a few feet away from existing buildings and infrastructure, making the performance of excavation support systems a critical safety and risk management concern.
In a recent study published in the International Journal of Geotechnical Engineering, my co-author Antonio Marinucci, Ph.D., P.E., and I set out to better understand how these systems behave in practice. Our goal was to bring together the fragmented body of case history data on Support of Excavation (SOE) walls and synthesize it into a single empirical dataset that engineers and project managers can use to better evaluate excavation performance.
The study compiles and analyzes more than 1,500 documented excavation case histories from projects reported in the technical literature. By assembling this large dataset, we created one of the most comprehensive empirical resources to date on excavation wall displacements and ground settlements. Our hope is that this dataset helps practitioners better understand what “normal” performance looks like during excavation and how it varies depending on soil conditions, wall systems, and project context.
Why Excavation Performance Matters in Urban Construction
Support of Excavation systems are retaining structures that stabilize the surrounding soil during excavation. They are commonly used for projects such as building foundations, subway stations, and underground infrastructure. Although these systems may be temporary, their performance is extremely important. Excessive wall movement or ground settlement can affect adjacent buildings, utilities, and transportation infrastructure.
One of the challenges engineers face is that predicting excavation behavior is inherently complex. Soil conditions vary widely from site to site, and even sophisticated numerical models often require empirical validation. Much of the knowledge about excavation performance comes from individual case studies reported in the literature, but these studies are typically examined in isolation.
Bringing Together a Global Body of Case Histories
Antonio and I wanted to step back and look at the bigger picture. By synthesizing hundreds of published studies and technical reports, we were able to analyze deformation patterns across a much broader range of projects and ground conditions than is usually possible in individual investigations.
In the study, we focused on two key performance indicators that are widely used in excavation design and monitoring: maximum horizontal wall displacement and maximum ground settlement behind the wall. Because excavation projects vary significantly in size, we normalized both measurements by excavation depth. This approach allows engineers to compare deformation behavior across projects with different excavation heights.
We also categorized the data according to four generalized soil environments: fine-grained soils, coarse-grained soils, mixed soil conditions, and soft soils over rock. Each of these ground conditions tends to influence excavation behavior in different ways, and examining them separately allowed us to identify patterns that might otherwise be obscured.
Data-Driven Benchmarks for Expected Deformation
One of the most useful outcomes of the analysis was the development of empirical benchmarks for expected deformation. Across the dataset, the majority of normalized wall displacements fell below about 2% of the excavation depth, while most normalized ground settlements remained below roughly 1% of the excavation depth. These values provide practical reference points for engineers evaluating whether field measurements during construction fall within expected ranges.
We also observed how different wall systems and soil conditions influence deformation behavior. Stiffer wall systems generally tend to control ground movements more effectively, while more flexible systems may exhibit larger displacements. Soil conditions also play a major role. Excavations in coarse-grained soils or soils overlying rock often produce smaller and more predictable movements, while fine-grained or mixed soils may show greater variability.
Supporting Better Risk Management in Construction Projects
Beyond the technical insights, the broader contribution of this work is that it provides a data-driven perspective on excavation performance. Engineers and construction teams rely heavily on monitoring data during excavation, but interpreting that data requires context. Having empirical benchmarks derived from hundreds of projects helps engineers and project managers evaluate whether observed movements are typical or whether they indicate potential risk.
This kind of information is valuable not only for geotechnical engineers but also for project managers responsible for risk mitigation on complex construction projects. Excavation performance can directly affect construction sequencing, adjacent property protection, schedule management, and overall project risk.
Another important aspect of the work is that the dataset can serve as a reference for validating numerical simulations. Many design approaches rely on modeling tools to estimate soil movements and wall behavior. By comparing those predictions to documented field performance, engineers can better calibrate their models and improve forecasting accuracy for future projects.
What Does This Mean for the Future of Urban Construction?
For me personally, this research reflects something I value deeply in engineering practice: the ability to learn systematically from past projects. As cities continue to expand vertically and underground, deep excavations will remain a central part of urban infrastructure development. The more we can learn from previous projects, the better we can design and manage the ones that follow.
Our hope is that this synthesis of excavation case histories provides a useful reference for engineers, researchers, and project managers alike, helping to improve design decisions, strengthen monitoring strategies, and ultimately support safer and more resilient construction in complex urban environments.
About the Program:
The Columbia University Master of Science in Project Management program equips individuals with the strategic, analytical, and leadership skills essential for a successful career managing complex projects across industries and borders.
Available full-time or part-time, the M.S. in Project Management is designed for professionals who want to advance into leadership roles or formalize their project management experience with a strong academic and practical foundation. Students can opt for the general Project Management program or choose from one of the four specialized concentrations: Construction, Sports Management, Sustainability Management, and Technology Management.
Taught by scholar-practitioners and enhanced by Columbia’s location in New York City, the curriculum integrates emerging digital tools and AI-driven practices to help graduates make data-informed decisions and improve operational efficiency. Graduates will be prepared to lead high-stakes projects with confidence and clarity, and return to the job market with a competitive edge.
The application deadline for the M.S. in Project Management program is June 1. Learn more about the program here.
