Quantitative Risk Management and Decision Making in Construction

Risk management planning requires recognizing, understanding, and addressing all potential risks through the course of a project. The risk management plan documents and plans this process and includes six steps: risk identification, risk assessment analysis, risk control, risk cost allocation, risk monitoring, and risk planning. First, risk identification identifies and categorizes the potential risks associated with a project. Second, risk analysis analyzes the risks qualitatively and quantitatively to identify those risks of high importance. Third, risk control applies various control strategies to maximize the project’s benefits and minimize losses. Fourth, the cost allocation process allocates the risks to the appropriate parties. Then a monitoring program tracks the progress of the project and its risks and updates the risk management plan whenever a new risk develops.

This chapter introduces a method for bidders to determine a bid price with confidence and without cost overrun. The chapter presents an example problem that utilizes the technique of evaluating bid contingency with triangular distributions, and it exhibits how the owner can use the project’s cumulative probability profile to estimate bid price and discusses risk transfer and risk absorption strategies. Various projects demonstrate that risk analysis techniques are abundantly beneficial, unlike some major construction projects that suffer from substantial cost overruns and delays owing to nonconsideration of risk. The triangular-distribution method is a simple way to evaluate the bid price and the probabilities of cost overrun and underrun for small projects or simplified large projects.

This chapter presents the basics of contingency cost analysis. Once risk has been identified, a contingency fund can be allocated to mitigate risk items and unforeseen events and prevent potential losses. It may also benefit the project with money left over after project completion. The first step in evaluating a work package for risk is to determine whether the costs of that work package are critical to the risk or not, using the “criticality rule.” A safe bid with 100% confidence level will eliminate risks of an overrun while ensuring profit, but will have a tremendously low probability of winning the bid in a competitive environment.

This chapter examines an important quality management tool called cause and effect diagrams (CEDs), also known as Ishikawa diagrams and fishbone diagrams. A CED serves as a systematic tool to determine all possible causes of a problem or outcome and organizes similar causes into groups. Three types of CEDs are dispersion analysis, production process classification, and cause enumeration. The chapter discusses CEDs through applications, examples, and practice problems and explains their construction. The cause enumeration type of CED results in the same diagram as the dispersion analysis type. The difference between cause enumeration and dispersion analysis is the procedure involved when drawing a CED. The production process classification CED resembles an assembly line process where one keeps adding causes, step by step, in a specified order, as product completion advances.

Value management encompasses various proven methods to analyze and value processes, products, and projects. It has directly influenced the business world—and more profoundly manufacturing—by optimizing methods for cost reduction while retaining essential product functionality. This chapter discusses the functional analysis system technique process (FAST) method. It explains how value management can be used to structure the operation of tasks and reduce uncertainty. The chapter describes FAST’s development arc and presents applications showing basic principles in value management. FAST may be used in any setting or situation as long as the functionality is defined. FAST provides an advantage in function analysis in that it can bring together a team, regardless of their knowledge, education, and technical background, and enable a fuller understanding of the problem or project at hand.

This chapter describes decision trees and how to implement them. It explains different situations where a decision tree may be implemented and the benefits and restrictions of this graphical tool. After reviewing several example problems, the reader should be prepared to complete the exercises at the end of the chapter without guidance. Decision trees are helpful in business, marketing, accounting, science, and engineering, which encounter quantitatively complex problems with many potentially different solutions. The decision tree diagram always begins with the base node (or root). Through probabilistic and quantitative analysis, the chapter makes a logical decision about a problem with the aid of a decision tree. The final product appears as a sideways connection of branches that vaguely resembles its clever name, a tree of decisions.

Payoff matrices are important tools in risk analysis and decision making that are used to identify risk in both everyday and multibillion-dollar business decisions. The payoff matrix method breaks the decision process down into decision alternatives and states of nature. When making important decisions in risk management in engineering, the expected monetary value (EMV) method is recommended whenever possible. Overall, the payoff matrix method is great for any type of risk taker. Risk takers, risk avoiders, and anyone in between can use this technique to come up with an answer that suits their needs. Each method can also be used with one of the other methods to help ease the decision maker. A decision is obvious if it has the same outcome as maximin, maximax, minimax regret, and EMV.

The chapter explains the derivation of Bayes' theorem from conditional probability and provides example applications to demonstrate understanding of the theorem. Bayes’ theorem was derived from conditional probability through the understanding of events and sample space, the multiplication rule, and the law of total probability. Probabilities refer to the occurrence or nonoccurrence of events. Bayes’ theorem can produce posterior probability distributions that are biased by prior probabilities. The chapter describes the application of Bayes’ theorem through simple examples of real-world situations, which show that the theorem is an excellent tool for dealing with risk analysis in the decision-making process. Bayes’ theorem can be applied to various industries and fields, such as construction, medicine, industry, organizational management, and international relations.

Engineers use matrix analysis to structure important decisions and problems. The tool prioritizes the decision maker’s needs, then evaluates and rates the different solutions. The matrix analysis is basically a list of alternatives to evaluate on one axis, subject to a list of criteria on the other axis. These criteria are weighted based on their importance to the solution. This decision-making model can be applied to almost any decision, engineering related or not. Constructing a decision matrix is easy and follows the criteria, options, weights, and scores (COWS) method. This chapter shows an example of a weighted evaluation matrix and the associated different segments for constructing a concrete foundation. The analytic hierarchy process (AHP) is one way to systematically assign weights, removing subjectivity from assigning the weights.

This chapter focuses on the analytic hierarchy process (AHP), which uses paired comparisons, mathematical procedures, and the subjective judgments of professionals to make decisions. Multicriteria decision making is relevant to many complex, multivariant real-life issues, which is why learning this process is so helpful. The overall method is called multicriteria decision analysis (MCDA), which encompasses “mathematics, management, informatics, psychology, social science, and economics,” creating a stepping stone to arrive at a compromise solution. AHP rationalizes individual judgments by assigning values to them. Individuals generally arrive at a decision using two approaches that represent how they think: the deductive approach and the systems approach. AHP offers a systematic and structured methodology to analyze the behavior of decisions that depend on weight assignments.

The OOPS game explores decision-making uncertainty and its implications in project management and is built on simple relationships and scoring rules. The concepts, lessons, and strategies that the OOPS game can teach reinforce common ideas about successful project management, which in general emphasizes planning and risk taking. The goal of the OOPS game is to construct a project for the least cost. Common playing cards can represent project activities. All the statistics quoted in the chapter were generated by playing the OOPS game every possible way and recording the score. MATLAB was used to develop 71 programs and two user-defined functions to quickly play the game and track the score for each attempt. The heart of the OOPS game is balancing the number of PLANs and BUILDs to construct the project for the least cost.

This chapter studies control charts and their role in tracking performance. Control charts are relevant to risk analysis because they can predict future performance. The predicted future performance, which is based on observed performance, is relatively uncertain and undependable. The chart indicates the operating trend, emphasizing the process average and the variability within the process. The causes for process variations fall into two general categories: chance (or random) and assignable. A run chart, which may also be referred to as a “time plot” or “trend chart,” is a graph that measures data over time. When a run chart takes standard deviation (σ) into account, such as in 3σ or 6σ, it becomes a control chart. This concept of six sigma uses the normal (or Gaussian) distribution, upon which control charts are typically modeled.

The success of a company or business relies on many factors. One is its ability to manage product inventory efficiently, especially when inventory is valued in the millions of dollars and occupies space across many warehouses in many cities. Efficient management ensures that the costs of acquiring and storing inventory are minimized while the amount of products in inventory is maximized. This chapter illustrates an inventory management technique to contain risks and explains how to calculate the typical costs involved. The economic order quantity (EOQ) model is a powerful tool in inventory management. The EOQ model will also suggest that orders are placed at random points. One of the underlying assumptions of the EOQ model is that a constant demand for the product exists. An alternative method to deal with shortages instead of safety stock is backordering.

Risk is the possibility that an action will affect a given project or event. Six main steps in risk management are identification, assessment, analysis, mitigation, allocation and monitoring, and updating. Risk management is a repetitive and cyclical process that continues throughout the design and construction of a project. This chapter describes the six steps of risk management and presents a fictive example, “Roundabout at Kapi’olani Boulevard and Ward Avenue,” in Honolulu. After the risk assessment, the risks deemed to need further analysis were analyzed. Risk analysis combines the identified and assessed risks into an overall project risk estimate. Project owners can determine if a project should progress and, if so, determine cost and schedule contingency values. Risk monitoring and updating aim to track the identified risks systematically, identify any new risks, and manage contingency funds effectively.

This chapter overviews a direct, mechanical process for evaluating contingency funds. It presents four hypothetical work packages: establishing critical and noncritical items based on the 0.5% criteria, using Monte Carlo to generate 10 sets of random numbers, drawing an overrun profile and a contingency drawdown curve that show the contingency funds consumed vs. time elapsed in percentages, and contingency fund vs. time. Random numbers for selecting the appropriate price of each work package in any simulation, per the Monte Carlo Slab, can be generated on an ordinary scientific calculator or in Excel using the RANDOM function. The contractor must ascertain or determine at what confidence level he or she wishes to operate. To distribute the contingency fund of $7 million, one must know how to allocate it to the various work packages.