Sliding Scale Contingencies

1
Sliding Scale Contingencies
Transportation Estimators Association Annual
Conference Omaha, Nebraska September 24, 2009

• For Transportation Projects

A Delphi Study
Niyi Olumide Stuart Anderson Texas A M
University
2
Outline

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• Contingency Matrices
• Results
• Applications of Sliding Scales
• Conclusions

3
Introduction
What is a Base Estimate?
The most likely project cost estimate in any
phase at any time, which normally includes all
estimated known project costs, but does not
include project contingency
What is Contingency?
An estimate of costs associated with identified
risks, the sum of which is added to the base
estimate.
4
Cost Estimate at Any Phase

• Total Project Estimate Base Contingency

5

• Introduction
• Background and Motivation
• NCHRP 8-60 Risk Analysis Tools and Management
  Practices to Control Transportation Project Costs

6
Background
NCHRP 8-60
Published Definition of Contingency 48 of 52
SHAs responding
7
Motivation

• Contingency
• Provides costs for uncertainties associated with
  projects cost estimates
• Reduces over planning and project development
  time cycle, therefore, so does the amount of
  contingency
• Enhances accuracy of cost estimates
• Improves communication to stakeholders

8

• Introduction
• Background
• Contingency Estimating
• Contingency and the Project Development Process
  (PDP)
• Risk in the Cost Estimation Process
• Contingency in the Highway PDP

9
Contingency Estimating and PDP

• The Project Development Process (PDP)
• Project 8-49 (Report 574)

10
Risk in the Cost Estimation Process
Phases of Project Development
Compare
Uncertainty
Contingency
Re-evaluate
11
Cost Estimates and Contingency (w/Total Project
Cost Estimate Baseline Estimate)
Baseline Estimate Total Project Cost Estimate
Project Cost
Cost Range
Contingency
Contingency
Contingency
Base Estimate
Base Estimate
Base Estimate
Base Estimate
Planning
Preliminary Design
Final Design
Programming
Project Development Process
12
Cost Estimates and Contingency
Total Project Cost Estimate Baseline Estimate
Baseline Estimate Total Project Cost Estimate
Contingency
Contingency
Project Cost
Base Estimate
Base Estimate
Base Estimate
Preliminary Design
Final Design
Programming
Total Project Cost Estimate lt Baseline Estimate
Baseline Estimate
Contingency
Total Project Cost Estimate
Contingency
Project Cost
Base Estimate
Base Estimate
Base Estimate
Programming
Preliminary Design
Final Design
13
Contingency in the Highway PDP

• SHA methods of setting contingency
• Standard Predetermined Contingency across
  projects
• Unique Project Contingency (Estimators/Engineers)
• Formal Risk Analysis and associated contingency
• NCHRP 8-60
• 16 of 48 SHAs stated use of standard
  predetermined contingencies
• Majority indicated that they use unique project
  contingencies
• Formal risk analysis not as widely used

14
Sliding Scale Contingency Ohio DOT Approach
15
Contingency in the Highway PDP

• Major Problems in Contingency Estimation
• Inconsistencies in application of contingencies
• Poor definitions of what constitutes contingency
• Inappropriate application of contingency amounts
  to cover other issues instead of the intended
  purpose
• Contingency estimation methods One method may
  not necessarily serve all project types
• Rationale for selecting method may not
  sufficiently address major issues like project
  complexity

16

• Introduction
• Background
• Contingency Estimating
• Research Questions

17
Research Questions

• How does the application of contingency change
  across the PDP?
• How do the ranges of contingency change across
  the PDP in the Highway Industry?
• How do factors such as the following impact
  contingency estimates?
• Project Size/Complexity
• Estimation method
• Level of scope definition

18
Research Questions

1. When using contingency scales, do the contingency
   bands narrow linearly or decrease exponentially
   or in some other form across the PDP?
2. What is the relationship between uncertainty,
   risk and contingency?

19

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives

20
Study Objective

• Main Objective
• To develop sliding scale contingencies that can
  be used in the estimation and management of the
  cost of highway projects taking into account the
  effects of project complexity, level of project
  definition, estimation methodology, and phase of
  project development

21

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• The Delphi Technique
• Panel of Experts

22
The Delphi Technique

• What is the Delphi Technique?
• An iterative method used to gather opinions from
  a group of qualified individuals to solve a
  complex problem
• The method relies mainly on the judgment of
  experts to solve problems
• It lends its application to solve problems where
  there is little or no empirical evidence
• It is conducted through the application of a
  series of questionnaires called rounds typically
  a minimum of 2 rounds
• The main aim of the Delphi technique is to
  achieve consensus among the experts

23
The Delphi Technique

• Key Features
• Anonymity minimizes the effects of bias,
  conflict, individual status and intimidation
• Effectively structures group communication
• Eliminates the draw backs of face-to-face
  participation
• Controlled feedback to participants at the end of
  each round enables participants make fully
  informed decisions

24
The Delphi Technique

• Caution!!!
• Repeated rounds may lead to fatigue or loss of
  interest of the panel members
• May take a substantial amount of time and
  participant effort to complete if questionnaires
  are not carefully structured
• Panel members should represent several relevant
  disciplines to guarantee a wide knowledge base

25
The Delphi Technique

• Round 1 Group Response Analysis
• Controlled Feedback in round 2

• Round 2 Group Response Analysis
• Controlled Feedback in subsequent rounds

26
Panel of Experts

• Profile
• Invitations sent to approximately 80 potential
  participants
• 23 professionals agreed to participate and made
  up the panel
• Majority had experience
• Estimating (5 to 20 years experience)
• Risk Assessment (2 to 5 years experience)
• Majority of participants were from SHAs

27
Panel of Experts

• Expertise
• Project Managers
• Project Engineers
• Risk Coordinators
• Cost Estimators
• Construction Managers
• Program Managers
• Design Engineers
• Contract Managers
• Project Planners

28

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• Contingency Matrices
• Complexity Classifications
• Representative Risks
• Matrices

29
Contingency Matrices

• Participants were provided 3 contingency matrices
  for 3 complexity scenarios across the phases of
  project development
• Complexity definitions from NCHRP 8-49 (Report
  574)
• Non-Complex (Minor) Projects
• Moderately Complex Projects
• Most Complex (Major) Projects
• Typical risks were associated with the complexity
  levels

30
Complexity Classifications

• Levels of Complexity are distinguished using
• Roadway attributes
• Traffic control approaches
• Structures
• Right of way
• Utilities
• Environmental requirements
• Stakeholder involvement

31
Complexity Classifications

• Examples of complexity classifications (full
  definitions in Report 574)

Project Types based on Complexity Project Types based on Complexity Project Types based on Complexity
Most Complex (Major) Moderately Complex Non-Complex (Minor)
New highway major relocations New interchanges Capacity adding/major widening Major reconstruction (4R 3R with multi-phase traffic control) Congestion Management Studies are required 3R and 4R projects which do not add capacity Minor roadway relocations Certain complex (non-trail enhancements) projects Slides, subsidence Maintenance betterment projects Overlay projects, simple widening without right-of-way (or very minimum right-of-way take) little or no utility coordination Non-complex enhancement projects without new bridges (e.g. bike trails)
32
Representative Risks

• Examples of the Representative Risks for the 3
  project complexity scenarios

Project Type Most Complex Moderately Complex Non-Complex
REPRESENTATIVE RISKS Unresolved constructability issues Geotechnical Issues Contractor delays
REPRESENTATIVE RISKS Design Complexity Changes in materials/ foundation Changes in Program priorities
REPRESENTATIVE RISKS Political Factors Delays in permitting process Errors in cost estimating
REPRESENTATIVE RISKS Complex environmental requirements Bridge redesign/ analysis Inaccurate Technical assumptions
33
Contingency Matrices

• For each complexity scenario, participants were
  asked to input appropriate ranges of contingency
  (Low, Most Likely Estimate, High) in the matrices
  based on
• Phase of Project Development
• Level of Project definition
• Type of Estimate
• Historic Data
• Representative Risks for the different project
  types

34
Matrices
35
Project Type/ Complexity No. of Phases Phase of Project Development Phase Description Level of Definition Estimate Type Historic Data
Non-Complex (Minor) 5 Planning 10 to 20 yrs from letting 1 - 3 Parametric with Historical Percentages Cost per Lane mile, Past Projects
Non-Complex (Minor) 5 Programming/ Preliminary Design 5 to 10 yrs from letting 5 -15 Bid based (80/20 rule) with other Recent Bids, Past Projects
Non-Complex (Minor) 5 Design 1 4 yrs or less from letting 15 - 40 Bid based with 75 line items identified Recent Bids
Non-Complex (Minor) 5 Design 2 less than 4 yrs from letting 40 - 70 Bid based with 90 Line items identified Recent Bids
Non-Complex (Minor) 5 Design 3 less than 4 yrs from letting 70 - 100 Bid based, Cost based. All items (Pay) Recent Bids and/or Labor, Material, Equipment Costs
Moderately Complex 5 Planning 10 to 20 yrs from letting 4 - 7 Parametric with Historical Percentages Cost per Lane mile, Past Projects
Moderately Complex 5 Programming/ Preliminary Design 5 to 10 yrs from Letting 15 - 25 Bid based (80/20 rule) with other Recent Bids, Past Projects
Moderately Complex 5 Design 1 4 yrs or less from letting 25 - 35 Bid based with 75 line items identified Recent Bids
Moderately Complex 5 Design 2 less than 4 yrs from letting 35 - 70 Bid based with 90 Line items identified Recent Bids
Moderately Complex 5 Design 3 less than 4 yrs from letting 70 - 100 Bid based, Cost based. All items (Pay) Recent Bids and/or Labor, Material, Equipment Costs
Most Complex (Major) 4 Planning 10 to 20 yrs from letting 7 - 15 Parametric with Historical Percentages Cost per Lane mile, Past Projects
Most Complex (Major) 4 Programming/ Preliminary Design 5 to 10 yrs from letting 15 - 35 Bid based (80/20 rule) with other Recent Bids, Past Projects
Most Complex (Major) 4 Design 2 less than 4 yrs from letting 35 - 75 Bid based with 80 Line items identified Recent Bids
Most Complex (Major) 4 Design 3 less than 4 yrs from letting 75 - 100 Bid based, Cost based. All items (Pay) Recent Bids and/or Labor, Material, Equipment Costs
36
Round 1 Matrices
37
Round 2 Matrices
38

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• Contingency Matrices
• Results
• Overview of Rounds
• Contingency Inclusions and Exclusions
• Results

39
Overview of Rounds

• Round 1 formed the basis for subsequent rounds
  contingency ranges were very wide
• Contingency ranges provided were between 0 and
  500 contingency across the categories
• In Round 2 ranges tightened between 0 and 400
  contingency across the categories
• In round 3, majority of participants did not make
  any significant changes made to earlier
  assessments

40
Overview of Rounds

• Mean was used as the main feedback to
  participants at the end of each round
• Participants were provided all summary statistics
  from previous rounds as part of the feedback at
  the end of each round
• Response rate was 100 for all three rounds all
  23 participants responded
• Stability was achieved in the results after 3
  rounds

41
Contingency Inclusions

• Some participants indicated a few of the major
  items included in the contingency
• Cost Overruns
• Change orders during construction
• 5-10 for minor items
• 5 for supplemental work not identifiable at time
  of estimate preparation

42
Contingency Exclusions

• Some participants indicated a few of the major
  items NOT included in the contingency
• An allowance for cost escalation
• Allowances for items known to be required as part
  of the base project but not yet quantified in
  cost estimate
• Funds available for cost adjustments driven by
  predetermined market factors and incentives

43
Results

• Median ranges were slightly lower than the mean
  ranges for most of the categories
• The median was used as the final results to
  eliminate bias in the results

44
Mean Contingencies
Phase Planning Programming Design 1 Design 2 Design 3
Non-Complex Projects Non-Complex Projects Non-Complex Projects Non-Complex Projects Non-Complex Projects Non-Complex Projects
Ranges () 24-68 21-56 16-38 11-25 5-14
MLE () 41 34 25 17 9
Moderately Complex Projects Moderately Complex Projects Moderately Complex Projects Moderately Complex Projects Moderately Complex Projects Moderately Complex Projects
Ranges () 33-92 27-68 21-50 15-32 8-20
MLE () 59 43 31 22 13
Most Complex Projects Most Complex Projects Most Complex Projects Most Complex Projects Most Complex Projects Most Complex Projects
Ranges () 47-125 36-89 20-48 12-32
MLE () 75 59 31 21

• MLE Most Likely Estimate of Contingency

45
Non-Complex Projects
Phase Planning Programming Design 1 Design 2 Design 3
Ranges () 22-50 20-40 14-30 9-20 5-15
MLE () 35 30 20 15 8

• Median Contingencies
• Ranges Range Estimating
• MLE Deterministic Estimating
• Estimator can pick values within ranges based on
  project risks

46
Non-Complex (Minor) Projects
47
Moderately Complex Projects
Phase Planning Programming Design 1 Design 2 Design 3
Ranges () 30-75 25-60 20-40 13-30 7-19
MLE () 50 40 30 20 10

• Median Contingencies
• Ranges Range Estimating
• MLE Deterministic Estimating
• Estimator may pick values within ranges based on
  project risks

48
Moderately Complex Projects
Planning Programming Design 1 Design 2
Design 3
49
Most Complex Projects
Phase Planning Programming Design 2 Design 3
Ranges () 40-100 34-75 20-40 11-30
MLE () 60 50 28 20

• Median Contingencies
• Ranges Range Estimating
• MLE Deterministic Estimating
• Estimator may pick values within ranges based on
  project risks
• Higher level of definition in the programming
  phase due to complexity of project

50
Most Complex (Major) Projects
Planning Programming
Design 2 Design 3
51

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• Contingency Matrices
• Results
• Application of Sliding Scales
• 5 major steps, Benefits, Limitations,

52
Application of Sliding Scales

• To successfully apply these scales estimators
  must perform 5 major steps
• Remove all contingencies and conservative biases
  from the base estimate
• Classify the project by complexity as most
  complex, moderately complex or non-complex
• Determine current phase of project development at
  time of estimate preparation
• Identify risks as drivers of contingency
• Add appropriate contingency to the base estimate
  consistent with risks
• Repeat the process at each major phase of project
  development

53
Benefits

• Creates consistency in defining and applying
  contingency to projects
• Relatively easy to use
• Implied retirement of contingency across the
  phases of project development

54
Limitations

• Contingencies are not directly tied to risks and
  uncertainties
• Risks should be identified and monitored
• A risk list could provide justification to
  stakeholders for contingency included in estimate
• For larger moderately complex and all complex
  projects
• Analysis of unique project risks is recommended
  in developing contingencies
• Focus on contingency related to construction
  costs only

55

• Introduction
• Background
• Contingency Estimating
• Research Questions
• Study Objectives
• Methodology
• Contingency Matrices
• Results
• Applications of Sliding Scales
• Conclusions

56
Conclusions

• The Sliding Scales provide a defensible top-down
  method for estimating contingency
• Takes into account the effect of major factors
  that affect contingency
• Note!!!
• For successful application, estimators must
  remove all contingencies and conservative biases
  from base estimate before applying sliding scale
  contingencies

57
Questions / Comments?

• Thank You!