Technical Program Abstracts

(RISK) Decision and Risk Management

NOTE: Program Subject to Change

(RISK-3346) Producing More Accurate Range of Variables for Monte Carlo Simulations

Author(s)/Presenters(s): John A. Armstrong, PSP; Avi Sharma

Abstract:

Monte Carlo simulations are a useful technique in providing a range of probable costs and completion dates given a known set of plausible risks, scope, and baseline control.  However, the reliability of results is dependent on professionals' abilities to complete 3 necessary steps.  First is to completely identify and understand relevant risk events.  Second is to accurately define the range of random variables associated with each discrete risk event.  Third is to correctly relate risk events with baseline schedule and cost models so that the ranges are appropriately applied.

This paper will evaluate a large sample of CPM baseline and as-built schedules to identify what scopes of work, as defined by CSI division, typically exhibit the smallest and largest amount of slippages.   The paper's findings will be translated into ranges of probable schedule slippage by CSI division and confidence intervals.  The findings of this paper will serve immediate and practical benefit for risk professionals because it will provide them a realistic range of random variables for discrete scopes of work and is based upon empirical evidence.

(RISK-3380) Risk Intelligent Strategy for Managing HILP Events in Major Projects

Author(s)/Presenters(s): Dr. Manjula Dissanayake, CCP; Dr. Christopher Stroemich, Peng

Abstract:

Major capital project execution is plagued with cost and schedule overruns across the globe. Substantial efforts are invested in planning and preparing projects, including the identification and mitigation of risks and uncertainties through qualitative and quantitative risk models. Despite these efforts, residual uncertainties remain and are often a major contributor to overruns.   These include both 'known unknowns' and 'unknown unknowns' that are difficult to incorporate into typical risk models and project controls approaches. These uncertainties tend to be high-impact, low-probability (HILP) events. They are typically overlooked, impact is underestimated and not adequately prepared for.

This paper presents a risk intelligent strategy to plan and prepare for HILP events to improve predictability. It will propose a common taxonomy for uncertainty categories, a critique  of statistical methods and modeling approaches commonly used,  assessment of structure of project organizations, a communication strategy to enable effective decision making and the need to introduce incentive scheme(s) to ensure project organizations remain focused and driven.

(RISK-3399) Lessons Learned from Risk Assessment of Mega-Projects Track Record

Author(s)/Presenters(s): Ario Mirfatahi

Abstract:

Mega-projects often experience significant delays and fail to be performed in line with cost and expected quality. Throughout the industrialized and industrializing nations mega-projects share prevailing factors of failure including unclear scope of work, unrealistic requirements and planning, inaccurate cost estimates, unrealistic financing and unqualified personnel from managers to workers leading to unsuccessful execution of these projects. This paper reviews mega-projects that have some elements of failure, summarizes lessons learned from them and uses standard project models from heavy industries to benchmark and identify principal factors that lead to these failures. Risk assessment, which is essential to deal with these factors, is commonly inadequate on mega-projects. The paper identifies weaknesses of standard models in assessing risk factors; profiles dominant risks based on projects and their industry and then utilizes these factors to build a risk assessment framework to better understand various phases of mega-projects from a different perspective. This framework for project evaluation enables effective decision making, optimizes resource allocation and balances risks to lay out a better strategy for successful delivery of mega-projects.

(RISK-3457) The Lewis and Clark Expedition's Integrated Cost, Schedule, and Risk Analysis

Author(s)/Presenters(s): Robert G. Fatzinger, CEP; Adam James

Abstract:

Would the application of historical data and integrated cost, schedule, and risk analysis have improved the reliability of Thomas Jefferson's initial estimate for the Lewis and Clark expedition in 1803? Using integrated analysis of schedule and cost risk to estimate the appropriate level of contingency reserves on projects continues to demonstrate improved results over the traditional methods. These methods are most often reserved for mega-projects and have not yet 'trickled down' to the majority of those we estimate in industry. This paper seeks to examine the application of Recommended Practice 57R-09 to more common, phase-gate projects of moderate size, scope and complexity through lightweight, flexible Microsoft Excel modeling. We focus on the use of historical data to inform cost and schedule uncertainty, capturing both the known and unknown risks inherent in the data- improving the speed, accessibility, and applicability of integrated cost, schedule, and risk analysis on the common project. To demonstrate the flexibility and applicability of these concepts, this paper explores the application of integrated cost schedule and risk analysis for the historical Lewis and Clark Expedition.

(RISK-3474) The Cost of War: Data and the Defense Industrial Base

Author(s)/Presenters(s): Omar Akbik

Abstract:

Data has become the most valuable commodity for businesses in nearly every industry, and the defense industry is no exception. As product oriented sectors of the defense industry become more visible due to geopolitical events and increased media scrutiny, the need for robust, authoritative insight into the operations of the industrial base has become a priority for stakeholders at all levels of government. This requirement for increased visibility necessitates data at a more granular level than has previously been employed in project management functions. This paper will explore how the collection and utilization of raw financial data, coupled with technical specifications and programmatic metrics, can aid decision makers in the development of both predictive and descriptive models that provide unique insights into the realized challenges, risks, and opportunities facing these weapons programs that are of unique interest to policy makers. We will also explore the utilization of advanced statistical methods to track and forecast the costs and schedules of these challenges as programs progress.

(RISK-3479) Variability in Accuracy Ranges: A Case Study in the US and Canadian Power Industry

Author(s)/Presenters(s): John K. Hollmann, PE CCP CEP DRMP FAACE Hon. Life

Abstract:

This paper presents a case study of the variability in accuracy ranges for phased project cost estimates in the North American power industry. The study sought to improve the company participants' understanding of risks and estimate accuracy for their major power generation and transmission projects. The study team also sought to verify the theoretical accuracy values in the relevant AACE International Recommended Practices (RPs) for cost estimate classification. The study team analyzed estimated cost by phase (class) and final actual data from xx projects completed from 20xx to 2018 with actual costs from xx million to xxx million (2018$US).  Greenfield and brownfield power generation and transmission projects from across the US and some from Canada were included. Comparisons of the findings is made with other published studies (including two similar Canadian studies for hydropower and overhead power transmission presented at AACE conferences in 2014 and 2017.).

(RISK-3492) Utilizing Scenario Planning on the risk assessment exercise for high complexity projects

Author(s)/Presenters(s): Abbas Shakourifar, PSP; Micah J. Meads

Abstract:

Across many industries, Scenario Planning is regarded as an important strategic planning method with a variety of use cases but in its basic form it is employed to create flexible long-term plans. The planning method combines givens that are assumed about the future with key uncertainties and driving forces, identified by considering external and internal factors. The scenarios include plausible but unexpectedly important situations or complications that are subtle in the present but over time have the potential to create dramatic impacts on future events or strategies. In this paper, we describe the utilization of the Scenario Planning method as an ad-hoc tool that compliments the schedule risk assessment workshops. Following in-line with traditional project management processes, this planning method requires certain decisions to be made prior to identifying key uncertainties and plausible scenarios. Scenario Planning creates the flexibility needed for thinking in a broader range of possible futures, enabling decision makers to understand and consider the full risk profile of programs and strategic plans more broadly.

(RISK-3511) Stochastic and Arithmetic Forecasting of Contingency or Management Reserve Requirements and Drawdown

Author(s)/Presenters(s): Monica R. Shiever; Steven W. Wageman, CCP CEP EVP PSP

Abstract:

Establishing an appropriate amount of cost contingency or management reserves for a project can be a challenging and oftentimes perilous task for many organizations. There are numerous methods of varying complexity currently in use to help determine this elusive value. Arguably, the most robust and therefore defendable method is quantitative Monte Carlo cost uncertainty and risk analysis, which can and should also include an assessment of the cost of potential schedule delays. However developed, cost reserves are ideally time-phased to better understand when in the lifecycle of the project they will most likely be needed. This paper will describe a 'top-down' stochastic model for parametrically evaluating the actual use of cost reserves relative to their forecast, and more importantly for most organizations, to probabilistically forecast remaining reserve requirements and drawdown without conducting an updated 'bottom-up' uncertainty and risk analysis. A simplified arithmetic version of the stochastic model will also be presented for organizations that do not have the capabilities and/or resources to conduct quantitative Monte Carlo analyses. The described approaches will provide managers with higher decision confidence when facing the need to request additional cost reserves, or conversely, to release budgeted but unneeded cost reserves for more productive use in other activities of the organization.

(RISK-3512) Establishing an Appropriate Level of Schedule Detail for a Quantitative Schedule Uncertainty and Risk Analysis

Author(s)/Presenters(s): Steven W. Wageman, CCP CEP EVP PSP

Abstract:

Various professional opinions exist as to whether a summary or detailed-level schedule is more appropriate for a quantitative Monte Carlo Schedule Uncertainty and Risk Analysis (SURA). Which is best? This paper will first outline the basic schedule quality requirements necessary to conduct a meaningful SURA and why those requirements are important, followed by description of the fundamental process steps required to conduct a SURA. The pros and cons of using summary as well as detailed-level schedules will then be identified to help guide the practitioner in making an informed choice. Finally, the criteria, methodologies, and results of parallel SURAs conducted on summary, semi-detailed, and detailed-level schedule scenarios for the same test project will be presented to ascertain if there are meaningful differences in modeling results.

(RISK-3519) Visualizing Quantitative Risk

Author(s)/Presenters(s): J. Gustavo Vinueza

Abstract:

Communicating risk is an art and science.  We often build reports containing dozens of graphics, showing different sides of reality.  We are looking for the perfect slide, searching for that revelation moment. 

It is very common to find ourselves overwhelmed with charts and plots, attempting to hit a moving target: the mind of the decision-makers.  We try to make people understand risk, -an abstract concept-, and barely we see improvements in the field.  This document's goal is to improve the understanding of risk for people.

We will use an integral cost simulation model as an example.  The main outcome will be a set of graphical tools, adequate for well-known risk cases, along with practical tips on how and when to use each.

(RISK-3524) Risk assessment of a pipeline construction project in Alberta during El Nino year

Author(s)/Presenters(s): Mohamed Abdelgawad; D. Anthony Payoe; Guy Krepps, Peng

Abstract:

Risk and uncertainty is inherent in all construction projects and can result in affecting the overall project objectives. One of the key challenges of building pipelines in Northern Alberta is weather delays; either due to extreme cold or late winter freeze. Building through muskeg terrain has its own challenges due to the need to have frozen ground to support the weight of heavy construction equipment and to minimize environmental disturbances. Constructing during El Nino year adds further level of challenge attributed to the later winter freeze and early spring thaw. This paper presents a case study of a pipeline construction during El Nino year. This paper demonstrates the value added of using risk analysis to continuously assess the in-service date (ISD). During this study, risk analysis was applied to assist management with assessing the probability of the contractor meeting the ISD and to adjust the execution plan. The same concept can be applied to future projects were probability of meeting the ISD is continuously assessed and the execution plan is revisited to ensure successful completion of projects.

(RISK-3530) Establishing A Relationship Between Systematic Risk Management And The Attainment Of Project Objectives In The Ghanaian Building Construction Industry

Author(s)/Presenters(s): Ali Boateng

Abstract:

The attainment of project objectives has been an elusive target in the construction industry. Several suggestions have been made as to how to attain project objectives, with the implementation of systematic risk management being prime. Systematic risk management implementing requires the commitment of financial, human and other resources in its implementation thus should be expected to yield the intended results. This study therefore aimed to establish a relationship between the implementation of systematic risk management and the attainment of project objectives. The study adopted a quantitative approach. Data was collected through extensive review of relevant literature and administering of structured questionnaires with respondents being construction professionals of the office of Works And Physical Development of Three public Universities in the Kumasi Metropolis as well as contractors in the financial class of D1K1 who have undertaken projects with these Universities in the last three years. Data collected was analysed by using Partial Least Square Structural Equation Modelling. The findings pointed out that there exist a positive relationship between systematic risk management and the attainment of project objectives.

(RISK-3540) Conditional Branching Models how Project Managers Typically React to Schedule

Author(s)/Presenters(s): Dr. David T. Hulett, FAACE; Michael Trumper

Abstract:

It is common to find that project owners, contractors and managers can be counted on to react to prospective schedule overruns by developing a 'recovery schedule' that adds resources to try to make up time.   However, analysts using Monte Carlo simulation typically model the current plan as if the manager will not react even if the schedule is jeopardized. This is not realistic.

When simulating a project schedule the analyst needs to represent the manager's response to delays when there is still time to recover the schedule.  Conditional branching can represent the project manager's response to a schedule event such as the detailed engineering's finishing later than anticipated.

During a Monte Carlo simulation conditional branching can test for missing a key finish date.  Depending on that test, a branch of the schedule that models the management decision as anticipated is chosen. In the case study, a conditional branch is modeled with 2 possible plans. 'Plan A' is the original schedule, and 'Plan B' has more resources but shorter durations.. Using Conditional Branching, if detailed engineering finishes more than 1 month longer than planned, then Plan A is disabled, leaving the recovery Plan B to be followed.  If detailed engineering finishes sufficiently on time, then Plan B is disabled leaving the baseline Plan A in place.  The richness of the possible responses to schedule risk events will be discussed.

(RISK-3561) Multipath Risk Analysis

Author(s)/Presenters(s): Eric Ho

Abstract:

The purpose of performing quantitative risk analysis is to predict all possible outcomes for a project through an analysis of the project’s specific risks and uncertainties.  This is accomplished most often through a stochastic process using the Monte Carlo simulation technique, which results in a range of possible outcomes for cost and schedule.  However, traditional quantitative risk analysis is only evaluating the project through only one base set of assumptions.  This means that the risks and uncertainties are evaluated only to one starting deterministic model for cost and schedule.  Often times, key assumptions regarding the base model would change as the project evolves, rendering the results of the quantitative risk analysis obsolete.  This paper will present the Multipath Risk Analysis, which is an alternative to traditional quantitative risk analysis through the use of evaluating multiple different possible paths that a project could develop through its life cycle.

(RISK-3569) Emergent Project Risk Management Techniques

Author(s)/Presenters(s): Jason Gastelum; Brett Simpson; Casey Spitz; Stephen Unwin

Abstract:

Large projects often rely on Oracle’s Primavera Risk Analyzer™ or Safran Risk™ to perform Monte Carlo analysis, estimating the distributions of cost and duration of their projects given their risk exposure. For straightforward risks, the built-in risk models are usually adequate. Standard risk models allow an already-scheduled task to take longer or cost more on an individual iteration. But what to do when there is no appropriate task in the schedule on which to attach the risk and still maintain schedule fidelity? To address this shortcoming, a risk model method was developed that allows users to add a discrete risk to a new schedule task dedicated to hosting the risk. The host task is inserted into the schedule while ensuring that downstream start and finish dates are preserved. This implementation uses a suite of tools that includes a web-based risk register containing the risk’s parameters and software that automates constructing the risk model.

Meetings & Events