Post on 23-Oct-2020
Advanced Features of Schedule Risk Analysis using the Risk Driver
Method David T. Hulett, Ph.D.
Hulett & Associates, LLC Construction CPM Conference
San Diego, CA January 16, 2015
1
Agenda
• Add uncertainty to the schedule, correlation • Demonstrate the Merge Bias • Using Categories to install reference ranges of
Uncertainty • Adding discrete risk events as Risk Drivers • Probabilistic Branching • Probabilistic Calendars • Inflation • Risk Prioritization • Integrated Cost-Schedule Risk Analysis
2
Simplified Progression of Capabilities
• Initially, just applied probability distributions to the activity durations (e.g., PERT) to represent all duration risk
• Within the last 10 years have integrated discrete risk events commonly found in Risk Registers – Relegated 3-point estimates to represent only
inherent uncertainty, estimating error / bias • Added capabilities such as probabilistic
branching, probabilistic calendars and prioritization
3
Inserting Uncertainty in the Schedule
4
Simple 2-path schedule. Added triangular distribution uncertainties at .9, 1.05 and 1.2 representing the only risk considered here
Results for One Path, No Correlation
5
The P-80 value for one path (Test 2) is 15 June 2015
Comparison with and without Correlation = 1.0
6
The P-80 value for one path Without Correlation is 15 June 2015
The P-80 value for one path With Correlation = 1.0 is 22 June 2015
Evidence of the Merge Bias
7
The P-80 value for each path (without correlation) is 15 June 2015
The P-80 value for the Two Path Schedule 20 June 2015
Introducing the Gas Production Platform Schedule
8
Applying Different Uncertainty to Categories of Tasks as Reference Ranges
9
Each type of activity may have different levels of uncertainty, called “reference ranges”
Risk on the Offshore Gas Production Platform - Reference Range Uncertainties
10
With Uncertainty representing: • Inherent variability • Estimating error • Estimating bias • By category of task
The CPM date is 23 March 2017 The P-80 date is 13 June 2017
Introducing the Risk Driver Method for Causing Additional Variation in the Simulation
11
Using the simple 2-path schedule. Four risks are specified. The first is a general risk about engineering productivity, which may be under- or over-estimated, with 100% probability. It is applied to the two Design activities
Risk Driver’s Effect on Design Duration
12
With a 100% likely risk the probability distribution of the activity’s duration looks like a triangle. Not any different from placing a triangle directly on the activity
Risk Driver with Risk at < 100% likelihood
13
With this risk, the Construction Contractor may or may not be familiar with the technology, the probability is 40% and the risk impact if it happens is .9, 1.1 and 1.4. It is applied to the two Build activities
With a 40% Likelihood, the “Spike” in the Distribution Contains 60% of the Probability
14
Here is where the Risk Driver method gets interesting. It can create distributions that reflect: • Probability of occurring • Impact if it does occur Cannot represent these two factors with simple triangular distributions applied to the durations directly
Risk Drivers Models how Correlation Occurs
• Correlation can be caused by identifiable risks that are assigned to two different activities – If the risk occurs it occurs for each activity – If the risk impact multiplier is X% it is X% for each
activity • We are not very good at estimating correlation
coefficients, so generating them within the simulation is a better approach
• There still may be correlations among uncertainty (3-point estimates)
15
Risk Drivers Generate Correlation between Activities (1)
16
Risk 1: Probability 100% Impact .9, 1.05, 1.3
Activity 1 Activity 1
Correlation (Activity 1, Activity 2) = 100%
Risk Drivers Generate Correlation between Activities (2)
17
Risk 1: Probability 100% Impact .9, 1.05, 1.3
Activity 1 Activity 1
Adding uncorrelated uncertainty reduces correlation (Activity 1, Activity 2) to 86%
Uncertainty Not Correlated: .85, 1, 1.2
Risk Drivers Generate Correlation between Activities (3)
18
Risk 1: Probability 100% Impact .9, 1.05, 1.3
Activity 1 Activity 1
Correlation (Activity 1, Activity 2) = 64%
Risk 2: Probability 40% Impact .9, 1.1, 1.4
Risk 2: Probability 65% Impact .9, 1.15, 1.5
Activities Can be Influenced by More than One Risk Driver
19
An Organizational Risk has been added to the mix, assigned to all activities in the Two Path schedule
Adding Organizational Risk to Every Activity
20
With 4 risks including the Organizational Risk added to all 6 activities the P-80 result is 14 July 2015 (Without that risk the P-80 result was 26 June 2015)
Risk Drivers can be Applied In Series or In Parallel
• Two or more risks can be applied to the same activities. If they occur together in an iteration they may be in parallel or in series
• We are talking about the impacts of a risk that has occurred – Risk takes most resources or is so important to be
addressed that recovering from others must wait are entered in series (we have been assuming this)
– Risks can be recovered from simultaneously can be entered in parallel
21
Parallel and Series Risks Multiplicative with Risk Drivers
If recovery from two risks can be accomplished simultaneously, they are entered in parallel
Risk 1 1.2 factor
Risk 2 1.05 factor
Use 1.2 Factor, the largest factor, only
Risk 1 1.2 factor Risk 2 1.05 factor Use (1.2 x 1.05 = 1.26) Factor, multiply the two
If these two risks cannot be recovered from simultaneously, they are entered in series
Changing the Risks to In Parallel Reduces the Schedule Risk
23
Putting the Risk Drivers in Parallel results in an earlier P-80 of 8 July 2015 (had been 14 July with Risk Drivers in series) This capability is more important if more risks are assigned to the same activities
Failing the Test may lead to Multiple Activities that are Not In the Schedule
• If the test fails we may need to do: – Examine the Root Cause of the failure – Determine what to do next – Do what is needed to be done to recover – Re-test the article
• All of these activities need to be done, or none is needed – These 4 activities constitute a probabilistic branch,
since the possibility of doing them is probabilistic
(C) 2014 Hulett & Associates, LLC 24
Set up the Probabilistic Branch
(C) 2014 Hulett & Associates, LLC 25
We create a 4-activity probabilistic branch, adding 4 activities: Root Cause Analysis, Plan the recovery, Execute the Plan and Retest Notice that they all have a remaining duration of 0 working days – they will not affect the schedule unless they occur
Give the New Activities Ranges of Impact, if they Happen
(C) 2014 Hulett & Associates, LLC 26
Highlight the new activities in turn an give them uncertainties: • Root Cause Analysis 20d – 40d – 60 d • Plan the Recovery 10d – 20d – 30d • Execute the Plan 10d- 30d- 50d • Retest 20d – 30d – 50d (probably less time than the first test)
With the Probabilistic Branch in Place, Results may show Bi-modal Distribution
(C) 2014 Hulett & Associates, LLC 27
Probabilistic branch develops a shoulder at 60% There can be more than one probabilistic outcome from a node. The probabilities need to sum to 100%. Probabilistic branch can represent more planning than just a single probabilistic activity
Showing the Probabilistic Branch in the Bar Chart
(C) 2014 Hulett & Associates, LLC 28
Notice that the activities in the branch are designated by an orange milestone (duration zero days) and a flag B Finish is not orange indicating that it has two predecessors, Test 1 and Retest, so technically it is not only in the branch
Probabilistic Calendars (1)
• In many applications a weather-related event can impede progress and affect the schedule without regard to the dates permitted by predecessors – Monsoon weather can impede installation of jackets and
topsides for offshore platforms – Freezing weather can stop supplies getting in to a jobsite – Thawing ground may make it difficult to move equipment
• Other calendar-related events can be important for the schedule – Activities such as moving into a building can be
determined by calendar of events
(C) 2014 Hulett & Associates, LLC 29
Probabilistic Calendars (2)
• The static schedule might show that the weather-sensitive activity will take place outside of the weather window
• However, with schedule risk the date of the weather-sensitive activity is uncertain
• Risk on predecessors might push an otherwise safe activity into the weather window, causing it to be unlikely to occur
(C) 2014 Hulett & Associates, LLC 30
Simplify Probabilistic Calendar by using Two Path Schedule
(C) 2014 Hulett & Associates, LLC 31
Using the 2-Path schedule. There are no risks on this schedule
Probabilistic Calendar Event Editor
(C) 2014 Hulett & Associates, LLC 32
> Remove the Winter Weather Window > Select Event > Create the Winter Weather Event January with > Triangular distribution 15d, 20s, 25d in January 2015
Effect of Winter Weather Event
(C) 2014 Hulett & Associates, LLC 33
Build 1 finishes with a probability distribution that is 15 – 20 – 25 days beyond its scheduled finish of 3/17/15 because of the weather event that is 100% likely
Set Probability of Winter Weather Event to 50%
(C) 2014 Hulett & Associates, LLC 34
Winter Weather Event only 50% Likely
(C) 2014 Hulett & Associates, LLC 35
Build 1 may finish in March 2015 but may be delayed by a month because of the 50% likely January weather event
Monsoon Calendar Prohibits Offshore Installation (1)
(C) 2014 Hulett & Associates, LLC 36
Monsoon Calendar Prohibits Offshore Installation (2)
• Notice that the installation activities are mostly nominally before the monsoon season that occurs November - February – Exception to this is that Installation CPP Topsides is
already scheduled to complete in November
• With schedule risks on predecessors the other installation activities will occur during monsoon
• This calendar might also affect some pipe laying activities
(C) 2014 Hulett & Associates, LLC 37
Fixed Window of Non-Work
(C) 2014 Hulett & Associates, LLC 38
> Add Monsoon Calendar > Apply to Installation tasks > Window > add the impact and dates > Enable this calendar
Effect of Probabilistic Calendar on Install Drilling Jacket
(C) 2014 Hulett & Associates, LLC 39
Notice that the period from November to the end of February represents no completion because of the calendar. Predecessor activities have usual histograms
Start and Finish of Monsoon Window is Uncertain (1)
(C) 2014 Hulett & Associates, LLC 40
Start and Finish of Monsoon Window is Uncertain (2)
(C) 2014 Hulett & Associates, LLC 41
Notice that the window of finish dates is now narrower because the start and finish of the Monsoon season is uncertain
Calendar Events can have Different Probabilities and Impacts
(C) 2014 Hulett & Associates, LLC 42
Results with Risks, Uncertainties and Three Different Weather Events
(C) 2014 Hulett & Associates, LLC 43
Install Drilling Topsides with 3 Weather Event Windows, each with its own impact
Risk Prioritization Method
• Risks should be prioritized through the project schedule and the Monte Carlo simulation method to inform the risk mitigation exercise
• For management we need to identify those risks by “days saved” if they were fully mitigated so management can do benefit/cost
• For management we should identify “days saved” at the target level of certainty, say P-80
(C) 2014 Hulett & Associates, LLC 44
Two Approaches to Risk Prioritization using Quantitative Methods
• Typical Tornado Diagram with Risks (not activities or paths) as the arguments help to prioritize the risks
• However, with the structure of the schedule the Tornado Diagram is instructive but not definitive – The order of the risks’ importance can change
when one is removed, since that exposes other paths that were “risk slack paths” before
(C) 2014 Hulett & Associates, LLC 45
Use the Offshore Gas Production Platform Project
46 This project has reference ranges by category and 8 Risk Drivers
Standard Sensitivity Tornado
(C) 2014 Hulett & Associates, LLC 47
This standard Tornado shows activities in order of their correlation with finish date. These are not risks and risks cannot be teased out of these results, correlations are hard to understand
Criticality Tornado Diagram based on Percent on the Critical Path
(C) 2014 Hulett & Associates, LLC 48
Using the Criticality Index. These are also not risks but activities sorted by their percentage of iterations on the critical path Shows which paths are the most likely to delay the project, so new information, but not risks
Tornado Highlighting Risks, Not Activities or Paths
(C) 2014 Hulett & Associates, LLC 49
This special risk-based tornado diagram focuses on the entire impact of the risks, including their probability, impact range and the activities to which they are assigned Still, based on correlation concepts It shows Drilling Risk as an Opportunity, negatively correlated with finish date. Correct focus but measure is still correlation
Risk Prioritization Approach • Identify the level of uncertainty desired (P-80) • Simulate the schedule as many times as there are
risks (For the first risk this is = single pass method)
• Then identify the risk that saves the most days when it is eliminated – Eliminating the risk (probability = 0 or “disabled”)
represents complete mitigation, an ideal but impractical goal
– Once the most impactful risk is identified and eliminated, look for the second most-important risk, disable it, then look for the third risk….. This is the Iterative method
(C) 2014 Hulett & Associates, LLC 50
Run Prioritization View the Risk List in Priority Order
(C) 2014 Hulett & Associates, LLC 51
Prioritized Risks in a Table
(C) 2014 Hulett & Associates, LLC 52
The Grid view shows the risks with their “Days Saved.”
Gas Platform-1 - Risk Prioritization (80%)
UID Name Days Saved
8 Organization's engineers may not have the needed experience 50
4 Fabrication Risk Driver 49 2 Engineering Risk Driver 17 7 HUC Risk Driver 16 3 Procurement Risk Driver 8 6 Installation Risk Driver 4 1 Approval Risk Driver 0 5 Drilling Risk Driver 0
Total Risk Drivers Days Saved 144 Uncertainty Responsible for 86 Total Risk Contingency Reserve to the P-80 Level 230
Integrated Cost-Schedule Risk Analysis
(C) 2014 Hulett & Associates, LLC 53
Project Schedule built in Primavera P6. Resources include Time-Dependent (labor) and Time-Independent (materials) assigned to the three activities
Polaris shows the same Total Cost $880,000
Add Uncertainty to the Schedule
(C) 2014 Hulett & Associates, LLC 54
Simulation Details > Templated Uncertainty > Add > specify Schedule Uncertainty .8, 1.1, 1.5 > Replace Existing Distributions > Apply > Run the simulation 5,000 iterations This causes the schedule uncertainty. With Time-Dependent resources it will also cause cost uncertainty proportional to the uncertainty of the durations x daily cost “burn rate” of Labor and Rented Equipment for each activity LOE (indirects, overhead) would increase too if placed on a hammock activity as Time Dependent
Histogram Results Schedule
(C) 2014 Hulett & Associates, LLC 55
Standard Finish Date Histogram CPM date is 3/27/15 which is 8% likely P-80 is 7/30/15
Histogram Results Cost
(C) 2014 Hulett & Associates, LLC 56
Cost Histogram Estimated cost is $880,000 which is 6.7 % since time is the only cost risk P-80 is $1,020,000
Create the Cost-Time Scatter Plot
(C) 2014 Hulett & Associates, LLC 57
View JCL Scatterplot. “JCL” stands for Joint Confidence Level, which is a NASA term for integrated cost-schedule. Polaris was originally developed for NASA
Properties of the Cost – Time Scatterplot
(C) 2014 Hulett & Associates, LLC 58
> Check Regression Line > check Color > Check JCL curve @ 70% > place the crosshairs where the regression line intersects with the JCL curve
Liquidated Damages
• Polaris has flexibility in specifying the structure of Liquidated Damages
• The exposure to Liquidated Damages depends on the contract’s structure and the amount of time the project may be late
(C) 2014 Hulett & Associates, LLC 59
Liquidated Damages Scaling Rule
(C) 2014 Hulett & Associates, LLC 60
Several approaches to calculating Liquidated Damages
Extent of Schedule Overrun in Days
(C) 2014 Hulett & Associates, LLC 61
This gives a number of days. The total planned is 350 days. Note that 350 days is 8% likely Since the P-80 is 425 days and the LD is $1,000 per day the expectation is that the limit of $100,000 will be reached and exceeded, so will be effective
Results for Liquidated Damages
(C) 2014 Hulett & Associates, LLC 62
Note that there is some chance that there will be no LDs, and some chance that the maximum value of $100,000 is reached, in this case
Varying the Time Dependent Activities’ Daily Cost or Burn Rate
(C) 2014 Hulett & Associates, LLC 63
> Continue using One Path Project Resources > Simulation Options > Resources > Rate per unit or day. Notice the current rate is $800 representing a $100 per hour (burdened) times 8 hours per day per person (only one person needed, apparently) – this was put into Primavera P6 before importing to Polaris > Triangular and insert dollar numbers 700, 800, 900 This range may be based on uncertainty in the workforce, wage rates, general uncertainty in estimating, strength of the economy
Cost Results adding Burn Rate Uncertainty
(C) 2014 Hulett & Associates, LLC 64
Schedule results are unchanged The P-80 cost has increased to $1.07 billion
Time – Cost Scatterplot with Schedule and Burn Rate Uncertainty
(C) 2014 Hulett & Associates, LLC 65
Cost – finish date correlation is 71%, reduced from 100% by the addition of burn rate uncertainty. There is cost risk even if the schedule is perfectly predictive
Add Uncertainty to the Total Value of Time-Independent Resources
(C) 2014 Hulett & Associates, LLC 66
Simulation Details > Resources > Rate per unit or day > select Time Independent resource and enter 90, 120, 170. (The value 100 in the database equates to $200,000 spread over 200 days if 10 units per day are assumed in P6)
Cost Risk with added Time-Independent Resource Risk
(C) 2014 Hulett & Associates, LLC 67
P-80 is $1.13 billion with both burn rate and time-independent uncertainty added to schedule risk. Nearly half of the Cost Contingency to the P-80 Target $1.13 B - $.88 B = $.250 B) is from Schedule Risk acting on Time-Dependent resources • Schedule risk contributes
$120,000 • Cost risk on burn rate and
material cost contributes $130,000
Cost Time Scatter with Schedule, Burn Rate and Time-Independent Risk
(C) 2014 Hulett & Associates, LLC 68
Correlation between Time and Cost is reduced to 67.5%. At a 70% JCL (70.62 in this case): Schedule to 8/4/15 Budget to $1.125 billion
Joint Confidence Level (JCL)
(C) 2014 Hulett & Associates, LLC 69
The trend line, fitted to the data, shows the strong influence schedule risk has on cost risk and validates the need to conduct Integrated cost-schedule risk analysis to understand cost risk The Joint cost-schedule confidence level has been set at 70% (NASA target). The curved blue line shows combinations of time and cost with JCL = 70%
Adding Inflation
70
Adding 3% inflation causes the deterministic estimate of cost to increase to $1.69 million from $1.57 million The P-80 costs increase to $1.98 million from $1.82 million
Advanced Features of Schedule Risk Analysis using the Risk Driver
Method David T. Hulett, Ph.D.
Hulett & Associates, LLC Construction CPM Conference
San Diego, CA January 16, 2015
71
Advanced Features of Schedule Risk Analysis using the Risk Driver MethodAgendaSimplified Progression of CapabilitiesInserting Uncertainty in the ScheduleResults for One Path, No Correlation Comparison with and without Correlation = 1.0Evidence of the Merge BiasIntroducing the Gas Production Platform ScheduleApplying Different Uncertainty to Categories of Tasks as Reference RangesRisk on the Offshore Gas Production Platform - Reference Range UncertaintiesIntroducing the Risk Driver Method for Causing Additional Variation in the SimulationRisk Driver’s Effect on Design DurationRisk Driver with �Risk at < 100% likelihoodWith a 40% Likelihood, the “Spike” in the Distribution Contains 60% of the ProbabilityRisk Drivers Models�how Correlation OccursRisk Drivers Generate Correlation between Activities (1)Risk Drivers Generate Correlation between Activities (2)Risk Drivers Generate Correlation between Activities (3)Activities Can be Influenced �by More than One Risk Driver Adding Organizational Risk �to Every Activity Risk Drivers can be Applied �In Series or In ParallelParallel and Series Risks�Multiplicative with Risk DriversChanging the Risks to In Parallel Reduces the Schedule RiskFailing the Test may lead to Multiple Activities that are Not In the ScheduleSet up the Probabilistic BranchGive the New Activities Ranges of Impact, if they HappenWith the Probabilistic Branch in Place, Results may show Bi-modal DistributionShowing the Probabilistic Branch �in the Bar ChartProbabilistic Calendars (1)Probabilistic Calendars (2)Simplify Probabilistic Calendar �by using Two Path ScheduleProbabilistic Calendar Event EditorEffect of Winter Weather EventSet Probability of �Winter Weather Event to 50%Winter Weather Event only 50% LikelyMonsoon Calendar �Prohibits Offshore Installation (1)Monsoon Calendar �Prohibits Offshore Installation (2)Fixed Window of Non-WorkEffect of Probabilistic Calendar on Install Drilling JacketStart and Finish of �Monsoon Window is Uncertain (1)Start and Finish of �Monsoon Window is Uncertain (2)Calendar Events can have Different Probabilities and ImpactsResults with Risks, Uncertainties and Three Different Weather EventsRisk Prioritization MethodTwo Approaches to Risk Prioritization using Quantitative MethodsUse the Offshore Gas Production Platform ProjectStandard Sensitivity TornadoCriticality Tornado Diagram based on Percent on the Critical PathTornado Highlighting Risks, Not Activities or PathsRisk Prioritization ApproachRun Prioritization�View the Risk List in Priority OrderPrioritized Risks in a TableIntegrated Cost-Schedule Risk AnalysisAdd Uncertainty to the ScheduleHistogram Results ScheduleHistogram Results CostCreate the Cost-Time Scatter PlotProperties of the �Cost – Time ScatterplotLiquidated DamagesLiquidated Damages Scaling RuleExtent of Schedule Overrun in DaysResults for Liquidated DamagesVarying the Time Dependent �Activities’ Daily Cost or Burn RateCost Results adding Burn Rate UncertaintyTime – Cost Scatterplot with Schedule and Burn Rate UncertaintyAdd Uncertainty to the Total Value of Time-Independent ResourcesCost Risk with added�Time-Independent Resource RiskCost Time Scatter with Schedule, Burn Rate and Time-Independent RiskJoint Confidence Level (JCL)Adding InflationAdvanced Features of Schedule Risk Analysis using the Risk Driver Method