A Mathematical Structure for Modeling Uncertainty in Cost, Schedule, and Escalation Factor in a Portfolio of Projects Ali TOURAN, Ph.D., P.E., F. ASCE 1 1 Professor, Northeastern University, Department of Civil and Environmental Engineering, 400SN, 360 Huntington Avenue, Boston, MA 02115. Phone: (617) 373- 5508; Fax: (617) 373-4419; email: a.touran@neu.edu ABSTRACT The past decade has witnessed a surge in the application of formal probabilistic risk assessment on cost, schedule, or both in major capital projects. Depending on the level of detailed information required the sophistication of risk assessment approach ranges from simply considering some important variables to fully integrated cost/schedule risk models. The most common modeling approach is to model cost components as random variables and calculate total cost distribution. For projects or portfolios spanning several years, the effect of cost escalation on budget is profound. Because of this, the variability of the escalation factor should be considered in the conduct of the risk assessment. This paper provides a mathematical framework for modeling of cost uncertainty in a portfolio of large infrastructure projects with a multi-year duration. This framework considers the randomness of cost and escalation factor at the project level. Relevant equations are presented that consider various degrees of probabilistic modeling from basic to complex. The paper can be used for understanding major drivers of uncertainty in portfolio budget and to evaluate the effect of escalation variability on project costs. The concepts and the suggested approach are explained using a numerical example. INTRODUCTION This paper provides an overview of a mathematical framework for modeling of uncertainty in a portfolio of projects. A portfolio is defined as a collection of multiple projects managed by a single management team or organization. The projects can be interrelated and part of a program, or independent from each other but under the supervision of a single organization. The proposed overview is intended for large infrastructure portfolios (such as transportation) where project development can take several years. Under this scenario, the uncertainty may be considered and modeled in three areas: cost, schedule, and escalation factor. Table 1 gives an overview of probabilistic modeling of these three factors. Table 1. Various modeling levels for probabilistic estimating Factors modeled as random variables General approach and input needed (1) Cost, schedule, escalation factor Joint density function of the three variables; or marginals and correlations (2) Cost, schedule Joints density function of the two variables; or 1743 Construction Research Congress 2014 ©ASCE 2014