NUMERICAL BIAS OF CONSTRUCTABILITY/RISK TEAMS VS PROJECT ACTUAL BID RESULTS
By David A. Hamilton, PE, CVS-Life, CCP, ENV-SP
Hamilton Risk+Value Consultants, LLC
Are risk analyst and team members on constructability/risk studies perpetually pessimistic or does something happen between the time of the risk study and bid day? True, those of us involved with risk analysis have a critical perspective and survey our surroundings looking for the “what if’s” of this world to arise and complicate our projects. Sometimes it’s a fact of remembering a past construction project that went awry or a combination of facts that may lead to either cost or schedule complications. Working in our favor is the fact that risk is now on the radar screen of designers and construction managers, and owners are pressing their teams hard to manage these risk proactively in order to minimize the threat of potential change orders during construction.
This is a prudent approach since the statistics for mega project (>$500M) success are quite dismal, noting that more than 60% of projects exceed their original budgets and 70% miss their targets for schedule completion. Owners are demanding better results from design teams and CM firms to identify project risks earlier and develop mitigation measures to maintain schedules and budgets. This heightened risk awareness may result in team members on constructability/risk teams being hyper vigilant when it comes to risk identification and their respective impacts on schedule and cost. The positive side of this approach is that more and more risks are being identified early in the design process and mitigation measures are being considered prior to project bidding.
Constructability/Risk Study Example
A good example of numerical bias is on a recent constructability/risk study performed in March 2015 on a 105 mgd water treatment plant in the mid-west. The team first brainstormed potential risk items in a risk register, then followed up with the establishment of risk-ranges (high, most likely, and low) numbers for both cost and schedule durations. The tornado diagram shown in Figure 1 presents the work breakdown structure (WBS) and estimated costs based upon the 95% design submittal construction cost estimate of $22.3M.
This data was loaded into Safran Risk software and a Monte Carlo simulation was performed on a resource loaded schedule. Figure 2 presents the probability density and cumulative probability curves for the 95% design submittal estimate.