The Case for Schedule Risk Analysis

Size: px

Start display at page:

Download "The Case for Schedule Risk Analysis"

Lindsey Bradford
5 years ago
Views:

1 The Case for Schedule Risk Analysis Francisco Cruz, PE, VDC Section Meeting February 19, 2015

2 Presentation Outline The Case for Project Risk Analysis Risk Analysis Approach Tools to Treat Risks Final Remarks There are lies, there are damned lies, and then there are deterministic schedules Attributed to Dr. Vivek Puri, PMA s resident simulation guru 2

3 Francisco Cruz - Background Senior Associate at PMA Consultants Based in Tampa, Florida Registered Professional Engineer (P.E.) in FL and PR BS in Civil Engineering, M.Eng. in Construction Management Experience of over 10 years spans pharmaceutical, infrastructure, airports, municipal, and mixed-used projects. Performing Cost and Schedule Risk Analysis on projects since 2007 on over 50 projects in 3 continents, ranging in size from less than $5 million to over $1 billion AACEi member since

4 The New Normal in Project Management Terminology: Risk Management Soup to nuts strategy for identifying and controlling project or program risks. Control (Monitor & Communicate) Risk Management Planning Assess (Identify & Analyze) Treat (Plan & Implement) *TCM Framework (Section 7.6) 4

5 The Case for Project Risk Analysis What are the three most common causes of cost overruns? * What are the three most common causes of delays? * Material Price Escalation 60% Poorly Defined Scope 57% Poorly Defined Scope 55% Design Creep 57% Contractual Disputes 40% Lack of Approvals 36% Time Delay 36% Contractual Disputes 34% Design Creep 32% Weather 21% Achieving Productivities 19% Time Delay 17% Lack of Approvals 6% Achieving Productivities 15% Weather 4% Material Price Escalation 13% Industrial Relations 2% Commissioning Process 2% Others 26% Others 19% * KPMG 2008 Global Construction Survey According to the Bent Flyvbjerg s Study on Large Projects, 9 out of 10 projects have cost overrun; Overrun is found across the 20 nations and five continents. Overrun is constant for the 70-year period covered by the study; cost estimates have not improved over time. 5

6 The Case for Project Risk Analysis What is the current performance in the construction industry? Did you have underperforming projects? Main causes of underperforming projects * KPMG Global Construction Survey

7 Deterministic vs Probabilistic Schedules Deterministic Schedule Consider the following simple 2-activity schedule: Start Activity A (12 weeks) Finish Activity B (10 weeks) 7

8 Deterministic vs Probabilistic Schedules Deterministic Schedule % critical = 100% Start Activity A (12 weeks) Finish Activity B (10 weeks) % critical = 0% Project always to finish in 12 weeks 8

9 Deterministic vs Probabilistic Schedules Probabilistic Schedule OPT MOST PESS Start Activity A (12 weeks) Finish Activity B (10 weeks) OPT MOST PESS Activity B has much more variability to the pessimistic side due to inherent risks of its work 9

10 Deterministic vs Probabilistic Schedules Probabilistic Schedule Start % critical = 63% OPT MOST PESS Activity A (12 weeks) P10: 11.3 weeks P50: 13.1 weeks P90: 15 weeks Finish Activity B (10 weeks) OPT MOST PESS % critical = 37% 10

The New Normal in Project Management Risk Assessment process through probabilistic modeling is the next step in Project Management Brings together and aligns stakeholders Vets and

11 The New Normal in Project Management Risk Assessment process through probabilistic modeling is the next step in Project Management Brings together and aligns stakeholders Vets and improves project scope, schedule and estimate Promotes recognition and transparency in project challenges Risk treatment reduces potential impacts Evaluates feasibility of targets 11

12 How is the Schedule Contingency Typically Determined? Similar Project History Expert Judgment (a.k.a. Well-reasoned Heuristics) Rule-of-Thumb Guidelines Parametric Modeling Statistical Methods through Probabilistic Simulations 12

13 PMA s Risk Assessment Approach Qualitative Assessment Critical & Risk-sensitive sequences Project Risks & Uncertainty Identify Schedule Summary Sequences, Probabilistic Branching Model Durations of Risks & Critical Activities Range OPRA NetRisk Simulate Simulate Summarize Review Prioritize Range Identify Estimate Issues, Possible Risks & Response Plans Cost Model Cost Breakdown, Variables, & or Crystal Ball Develop Ranges for Variables: Scope, Price, Productivity, & Duration Quantitative Assessment * First Introduced at the 2004 AACEi Annual Meeting 13

14 Risk Register Sample Risk Register used to prioritize risks and subsequently input in the quantitative model. 14

15 Risk Thresholds Figure 1: Defined Conditions for Probability Scales of a Risk Very Low Low Medium High Very High Probability 0% 10% 11% to 30% 31% to 50% 51% to 70% >70% Figure 2: Defined Conditions for Impact Scales of a Risk on Major Project Objectives Project Objective Very Low Low Medium High Very High Cost Time Scope Insignificant Cost Increase 0-2 Week Time Increase Scope Decrease Barely Noticeable <2% of TPC 3-6 Week Time Increase Minor Areas of Scope Affected 2-7% Cost Increase 7-12 Week Time Increase Major Areas of Scope Affected 8-15% Cost Increase Week Time Increase Scope Reduction Unacceptable to PA >15% Cost Increase >18 Week Time increase Project Does not Satisfy Original Intent at Completion Project team defines thresholds to classify risks and reduce subjectivity Quality Quality Degradation Barely Noticeable Only Most Demanding Specs Not Met Degradation Requires PA Approval Quality Reduction Unacceptable to PA Major Re-work Required to Complete Project 15

16 Prioritization Probability & Impact Matrix (P/I) Probability Threats VH H M L VL VL L M H VH Impact Probability-Impact Score Legend High = > 9 Med = 5-9 Low = 1-4 Tolerance Thresholds Tolerance is used to define Risk appetite of the organization 16

17 Output Example: Cumulative Probability Distribution Pre-treatment Results What about posttreatment analysis? The project schedule deterministic date of 16-Sep-14 is in the 24 th percentile of the simulation results distribution (including weather effects). 17

18 Tornado Diagram Risks Contemporaneous Method /SRA Duration Sensitivity RD11 - FHWA Design Exception for substandard geometry not received by Q2, % REN4 - NY Environmental permits not received by Q4'12 41% RD9 - Cash Collection is required in toll Lanes 40% RC20 - Restricted Work Hours may delay construction 36% RD2 - Procurement Process w ill take longer than 8 months therefore D/B contract w ill not... 26% RC23 - Modified Toll Plaza collection not accepted for Toll Plaza delaying Traffic 15% RD3 - Program not fully defined by Q1, % RC14 - Relocation of Know n/unknow n Utilities may cause cost and schedule overruns 12% RD1 - Design Package for Design Build w ill not be completed in Q2, % RRW7 - ULURP process w ill affect property acquisition in NY 4% 18

19 Risk Impact Analysis Helps determine impact of each risk on project schedule Distribution Analyzer Removing RD1 - Finish Date Removing RC14 - Finish Date Removing RC23 - Finish Date Removing RD2 - Finish Date Removing RC20 - Finish Date Removing REN4 - Finish Date Removing RD9 - Finish Date Removing RD11 - Finish Date All Risks - Finish Date Variation:13 Variation:128 Variation:19 Variation:144 Variation:82 Variation:73 Variation:111 Variation: % 80% Variation:5 Variation:54 Variation:85 Variation:119 Variation:59 Variation:25 Variation:72 Variation:127 60% 40% Cum ulative Probability 20% 22/ 09/ / 12/ / 04/ / 07/ / 10/ / 02/ / 05/ / 08/ / 12/ / 03/ / 06/ / 09/ / 01/ % 19

20 Output Example: End Node Diagram of Criticality In lieu of an unhelpful table of activities, a graphical representation of the activities most likely to influence project completion is provided. 20

21 Which percentile would you use for your project? Most public agencies and private companies use the P80 through P90 Some practitioners use a combination of two percentiles to set the baseline, schedule margin and management reserve (e.g., P50 and P80) ID Public Agency / Private Company Confidence Level Aerospace 1 NASA 70% Consultant 2 Hulett and Associates (Project Risk Consultants) 80% Energy 3 NorthWestern (Montana, Nebraska, North Dakota) 95% 4 US Energy Information Administration (EIA) 90% Food, Health & Environmental 5 FDA & USDA 95 99% 6 Forest Ecology and Management (Australia) 90% 7 Environmental Protection Agency (EPA) 90 & 95% 8 Environmental and Occupational Health Private Industry 95 98% Government 9 US Corps of Engineers (CoE) 80% 10 US Government Accountability Office (GAO) 70 80% 11 USGS 90% Military 12 Air Force 80% Nuclear 13 US Nuclear Regulatory Commission (NRC) 90, 95, 98% Oil & Gas 14 British Petroleum (BP) 50% (Adjusted Mean)*, 90%* 15 GeoGlobal Resources 90% 16 Marathon 80% Pharmaceutical 17 Johnson & Johnson 75 90% 18 AstraZeneca 80% 19 Siemens HealthCare 80% Software % Transportation 21 DOT 95% 22 FRA 95% 23 FTA 70 & 90% 24 MTA 80% 25 Wisconsin DOT 80% 26 FHWA Mean, Standard Deviation, 90 95% 27 PANYNJ 90% 28 WSDOT 90% 21

22 Comparing Results What Do Pre and Post-Treatment Results have in Common? 22

23 Tools to Treat Risks Value Engineering Constructability Reviews Project Delivery Systems Virtual Design and Construction 23

24 Risk Transfer Through Project Delivery Systems 1. Design/Bid/Build (DBB) a. Multiple-Prime Contracting 2. Design/Build (DB) a. Bridging b. Public Private Partnership (P3 or PPP) 3. Construction Management at Risk (CMAR or CM/GC) 4. Integrated Project Delivery * CMAA s An Owner's Guide to Project Delivery Methods 24

25 Risk Allocation Through Contract Types * RP-67R-11 Contract Risk Allocation 25

26 * 2014 Mortenson VDC Construction Analysis Report 26

27 Risk Monitoring and Control 27

28 Final Remarks Any project or contract schedule that is not risked through its life cycle does not conform to scheduling best practices Any schedule that does not expressly reserve reasonable schedule margin does not conform to best practices either Most deterministic schedules have less than 20% chance of meeting baseline dates; majority have even less than 1% chance. Post-treatment analysis is hardly ever performed; both pre and post-treatment analysis should be performed in tandem The use of VE, VDC, Constructability and Risk Management greatly increases the probabilities of project success 28

29 29

30 THANK YOU Francisco Cruz, PE, VDC Senior Associate PMA Consultants 30

Use of the Risk Driver Method in Monte Carlo Simulation of a Project Schedule

Use of the Risk Driver Method in Monte Carlo Simulation of a Project Schedule Presented to the 2013 ICEAA Professional Development & Training Workshop June 18-21, 2013 David T. Hulett, Ph.D. Hulett & Associates,