Newsletter July 2015

Welcome to the mid-2015 issue of “RiskIntegral” - an occasional newsletter about the analysis and management of risk, mainly in projects, by Risk Integration Management Pty Ltd (RIMPL). In this issue our feature article concerns Project Time and Cost Contingency, a prime deliverable of our QRA methodology and a contentious topic in the realm of project management.

As usual, we round out the newsletter with some news of our recent activities.

• PROJECT TIME AND COST CONTINGENCY

• RIMPL News

PROJECT TIME AND COST CONTINGENCY

Introduction

A couple of recent discussions on LinkedIn forums demonstrated that there are many different opinions about Contingency and a fair bit of confusion as well.

• On the Association for the Advancement of Cost Engineering International (AACEI) Members and Friends Network forum, a discussion titled “Why Contingency Reserves Need to be Owned by Senior Management” attracted vigorous opposing views (If you have access, the link is here).

• On the Schedule and Cost Risk Analysis forum, a link to a post by Santosh Bhat “Can a Risk Adjusted Schedule use P80 dates?” stimulated some divergent opinions about where schedule contingency belongs (If you are able to access it, see here). This discussion was asking whether it is valid to build a schedule based on the conservative probabilistic dates (say P80) from a Schedule Risk Analysis. Santosh correctly noted that the timing information could be useful but that it is invalid to present the tasks with P80 dates logically linked as per the deterministic project plan.

Dr David Hulett, respected author of two very important books (“Practical Schedule Risk Analysis” and “Integrated Cost-Schedule Risk Analysis”) disagreed with a post (responding to Santosh’s topic) asserting that it was OK to use P80 dates for a risk-adjusted schedule by stating: “The only reasonable place to put a provision for risk that has been developed from an integrated schedule risk analysis using Monte Carlo techniques is at the end, between the final milestone and its predecessor(s).”

As we had recently completed an Integrated Cost & Schedule Risk Analysis engagement for a client where distributed contingency through the schedule was an essential part of the analysis, it seems timely to discuss the subject of project time and cost contingencies, hopefully to clarify some of the confusion.

What is Contingency and why does it matter?

Confusion about Contingency starts with its definition. PMI call it Contingency Reserve in PMBoK Guide 5th Edition: “Cost estimates may include contingency reserves (sometimes called contingency allowances) to allow for cost uncertainty… Contingency reserves are often viewed as the part of the budget intended to address the ‘known-unknowns’… Contingency reserves are part of the cost baseline and the overall funding requirements for the project.” (PMBoK 5 section 7.2.2.6)

AACEI has a similar view of Contingency, but steers clear of including the word ‘reserve’. In its Recommended Practice 10S-90 ‘Cost Engineering Terminology’, AACEI has a broader definition than PMI but also tells us what Contingency is not: “CONTINGENCY- (1) An amount added to an estimate to allow for items, conditions, or events for which the state, occurrence, or effect is uncertain and that experience shows will likely result, in aggregate, in additional costs… Contingency usually excludes: 1) Major scope changes… 2) Extraordinary events such as major strikes and natural disasters; 3) Management reserves; and 4) Escalation and currency effects… Contingency is generally included in most estimates, and is expected to be expended.”

AACEI provides a second meaning for Contingency, specifically relating to the ANSI Earned Value Standard (widely used for Defence projects): “(2) In earned value management (based upon the ANSI EIA 748 Standard), an amount held outside the performance measurement baseline for owner level cost reserve for the management of project uncertainties is referred to as contingency.”

Management Reserve – a special kind of Contingency

AACEI separately defines (in RP 10S-90) another provision outside of the project funding, which, like Contingency, has a separate special meaning within the ANSI Earned Value Standard:

“MANAGEMENT RESERVE – (1) An amount added to an estimate to allow for discretionary management purposes outside of the defined scope of the project, as otherwise estimated. May include amounts that are within the defined scope, but for which management does not want to fund as contingency or that cannot be effectively managed using contingency.
(2) In earned value management according to the ANSI EIA 748 standard, an amount held outside the performance measurement baseline to handle unknown contingency at the total program level.”

PMI defines Management Reserve (MR) as being part of the overall project budget and funding requirements, in agreement with the ANSI Earned Value standard.

Management reserves are an amount of the project budget withheld for management control purposes and are reserved for unforeseen work that is within the scope of the project. MRs are intended to address the “unknown-unknowns” that can affect a project. The MR is not included in the cost baseline but is part of the overall project budget and funding requirements. When an amount of MR is used to fund unforeseen work, the amount of MR is added to the cost baseline, thus requiring an approved change to the cost baseline.” (PMBoK 5 section 7.2.2.6)

In our view, the AACEI approach, of explicitly stating that Contingency is expected to be expended and excluding Management Reserve from the project budget and funding requirements, reflects industry practice and makes more sense than the approach taken by PMI.

How is Contingency determined?

Time and Cost Contingencies have been assessed in the past by rules of thumb or judgment based on past experience (also called heuristic methods), but are increasingly determined by statistical or simulation methods, as these offer the best way of scaling the amount of contingency to the uncertainty characteristics of the project.

In their RP 40R-08 “Contingency Estimating – General Principles”, AACEI list a number of principles that should be addressed by the methodology chosen to quantify contingency, including that it should:

• Meet client objectives;

• Be part of an effective risk management process;

• Be ‘Fit for purpose’;

• Start by identifying the risk drivers with input from appropriate sources;

• Clearly link the risk drivers to the cost-schedule outcomes;

• Employ empiricism, experience and competency; and

• Provide probabilistic estimating results that supports effective decision making and risk management.

In addition AACEI recommends that the quality of the methodology should minimise the introduction of further risk through poor performance of the contingency quantification process.

AACEI RP 40R-08 goes on to detail the above principles and note that “The definition of contingency and how to estimate it are among the most controversial topics in cost engineering. While there is consensus among cost engineers on what contingency is, there is much less consensus on how to estimate it.” RP 40R-08 itemises four classes of methods for estimating time and cost contingency “that can respect the basic principles”. Where AACEI has prepared a corresponding Recommended Practice it is listed after the method and brief comments below:

• Expert judgment - highly dependent on the quality of the expertise, which must be strongly based in experience and normalised against bias by a diversity of opinions.

• Predetermined guidelines – with varying degrees of judgment and empiricism, ranging from set percentages of base cost or schedule duration depending on project estimate class or project type to complex tables using elements of parametric modelling (see below). Benefits include simplicity and consistency. Disadvantages include inability to incorporate unique riskiness characteristics of the project. Works best early in project development process (eg, concept phase) when systemic risks dominate.

See AACEI Recommended Practices 17R-97 “COST ESTIMATE CLASSIFICATION SYSTEM” and 18R-97 “COST ESTIMATE CLASSIFICATION SYSTEM – AS APPLIED IN ENGINEERING, PROCUREMENT, AND CONSTRUCTION FOR THE PROCESS INDUSTRIES”

• Simulation Analysis - primarily expert judgment incorporated in an analytical model then used in a Monte Carlo simulation process. There are two sub-classes:

Ο Range Estimating (This entails building an analytical model used in a Monte Carlo simulation process to generate probabilistic output. It is the starting point of the class method we use and recommend, based on our more sophisticated Integrated Cost & Schedule Risk Analysis [IRA] methodology and incorporating elements of the Expected Value methodology)

Ο Expected Value (This method focuses on the risk events in the project risk register with significant time and cost impacts. The probability of each risk times the impact is the “expected value”. The probability and impacts are replaced by distributions assigned by the team based on their understanding of the risks and correlations are assigned between risks and costs or schedule activities. Monte Carlo simulations are then run to produce cost or schedule probability distributions.

Relevant AACEI RPs are:

41R-08 “RISK ANALYSIS AND CONTINGENCY DETERMINATION USING RANGE ESTIMATING”

44R-08 “RISK ANALYSIS AND CONTINGENCY DETERMINATION USING EXPECTED VALUE”

57R-09 “INTEGRATED COST AND SCHEDULE RISK ANALYSIS USING MONTE CARLO SIMULATION OF A CPM MODEL”

65R-11 “INTEGRATED COST AND SCHEDULE RISK ANALYSIS AND CONTINGENCY DETERMINATION USING EXPECTED VALUE”

• Parametric Modelling - empirically based algorithm, usually derived through regression analysis, with varying degrees of judgment used. Based on historical cost data for previous completed projects, relating selected risk drivers to cost growth or schedule slippage. For example percent scope definition versus actual cost growth. Best for early stages of project definition, such as Concept or Pre-feasibility, when Systemic risk factors dominate over project-specific ones. Results must be interpreted with expert judgment.

Relevant AACEI RPs are:

42R-08 “RISK ANALYSIS AND CONTINGENCY DETERMINATION USING PARAMETRIC ESTIMATING”

43R-08 “RISK ANALYSIS AND CONTINGENCY DETERMINATION USING PARAMETRIC ESTIMATING – EXAMPLE MODELS AS APPLIED FOR THE PROCESS INDUSTRIES”

Note that all the AACEI Recommended Practices are available from their website www.aacei.org, either at no cost, or for a nominal fee.

Who should be responsible for the use of Contingency?

Recently (June 2015), a discussion was initiated on the AACEI LinkedIn forum “Why Contingency Reserves Need to be Owned by Senior Management”. The post was made by someone clearly influenced by the PMI view of and nomenclature for Contingency, both in the choice of terminology and advocacy of ownership of contingency NOT to be in the hands of the Project Manager.

The responses from regular commentators were consistent: the Project Manager should be responsible for Contingency (the portion of the budget that takes the Base Estimate to the agreed probability level defining the limit of sanctioned expenditure – usually P50, including all assessed risk events). If the PM is not responsible for contingency usage decisions, he or she is better described (as noted by Dr Paul Giammalvo) as “a project controls lead with diplomatic skills”.

What about Management Reserve?

Any provision beyond what is expected to be spent – say to P80 or P90 – that may be included in the economic analysis of the project, but excluded from the scope of the estimate, is classified as Reserve, which is often held by the Project Sponsor and only available to the PM through a specific event occurring and requires formal application and approval.

The principle followed is that accountability must be matched by responsibility and control. PMs accountable for project budgets must have control of the contingency provisions that go with them.

An example of a project aspect not within the PM’s control and thus falling logically into Management Reserve is the funding of currency fluctuations, normally managed by the organisation’s Treasury function.

Another may be the cost of funding the project, which may fluctuate substantially such as during periods of financial turmoil. In extreme circumstances, the requirement may grow to the point of requiring scope to be deferred or reduced.

Probabilistic Cash flow projections including these financing aspects as risks could help anticipate growth in Management Reserve.

Sometimes organisations go beyond the sanctioned limit that has been economically analysed – covering what may be called the “known-unknowns” in the Risk Register - to attempt to assess what the project could cost if unforeseen events were to occur – the so-called “unknown-unknowns”. This extra provision is, by definition, not assessable by conventional analytical means involving project-type risks, but may be able to be estimated partly through application of systemic risks. This Management Reserve “last resort” provision has been shown to be necessary in a disturbingly significant proportion of mega-projects which have substantially exceeded their sanctioned budgets of cost and time.

John K Hollmann, Editor of AACEI’s Total Cost Management Framework and author of several of the RPs listed here, has delivered many papers on this, based on data collected on completed projects by Independent Project Analysis, Inc. (IPA) (for which he previously worked) and others. In his paper “Risk Analysis at the Edge of Chaos”, presented to the AACEI Annual Meeting in New Orleans in June 2014, Hollmann noted that “none of the AACE Recommended Practices (RPs) are able to predict the reality that 10% of large projects overrun their budgets by 70% or more. Overruns of that scale for a mega project can cause significant financial damage to a company, project financed or not.” This may be a major reason for even large multi-national corporations “selling down” their equity in a mega project. In Australia, the incidence of such Mega Project overruns in recent years seems to have exceeded 10%.

Hollmann’s paper drew on Chaos and Complexity theories to propose that large and complex projects can be pushed into a non-linear state of chaos where adding more resources, rather than shortening the project, may lengthen it. He proposed that hierarchies of Complexity and Stress risk factors act on such projects to increase the probability of the project falling into a chaotic state, exacerbated by the occurrence of risk events, as shown in the accompanying diagram.

He has developed a traffic light system to forecast the probability of the project tipping into the chaotic state where the contingency required may be 5-6 times the P50 contingency. Further discussion is beyond the scope of this article, but the paper, published with the proceedings of the AACEI Annual Meeting of 2014, can be provided by RIMPL on request.

Where should Schedule Contingency be placed in the project schedule?

The answer to this depends on the perspective of the questioner and the nature of the contractual relationship between the Owner and the Contractor. If the contract is of the Time & Materials type and does not assign contractual responsibility for schedule slippage to the Contractor, the issue may not arise, because the Owner carries full responsibility. But if the Contractor is required to assume responsibility for schedule slippage, this question is contractually important. The following discussion examines the question from the perspectives of the two parties:

Contractor Perspective

A contract that requires a Contractor to deliver a specific scope by a specified Contract Date for a lump sum fee may include a clause which assigns responsibility for meeting that date. If it is a construction contract, it may also explicitly exclude downtime due to inclement weather as the basis for claims for extension of time. The onus is on the Contractor to make realistic estimates of down time that may be incurred during construction, including wet weather allowances etc., to achieve the Contract Date.

To ensure the Contractor can meet the contractual date, the Contractor’s schedule should set a Target Milestone ahead of the Contractual Date which provides adequate Schedule Contingency between it and the Contract Date. This Schedule Contingency must allow not only for adequate inclement weather allowance, but also for general schedule uncertainty and risk events. The Contractor must then manage the Schedule Contingency to achieve that objective.

Project Owner Perspective

The Owner may assign responsibility for schedule slippage to the Contractor, but the Completion Date to be achieved by the Contractor assumed by the Owner should still allow further slippage beyond that date to ensure that the Owner is able to meet project completion timing commitments provided to the Owner’s board. This is due to the following:

• The Contract cannot assign complete responsibility for schedule slippage to the Contractor – there are likely to be risks that, if they occur, can form the basis for EoTs or Variations due to acts or omissions by the Owner;

• If sufficient delays occur to exceed the insurable limits of the Contractor, the Contractor is likely to claim that the contract has materially changed and requires re-negotiation.

Special Case – highly uncertain weather conditions with seasonal variability

As noted in the Introduction, we recently performed an Integrated Cost & Schedule Risk Analysis (IRA) for an Oil & Gas Explorer, where the project was in a location with very high time-dependent costs and contractors and equipment were on cost-reimbursable contracts. The weather conditions were a dominant cause of schedule slippage and were also highly seasonal. All deliveries of people, equipment and materials to site had to be made by helicopter and there were only 3 months of the year when more than 40% of available daylight hours could be relied on to be usable for flying time. In addition, there was significant uncertainty in site construction durations and multiple risk events affecting construction, logistics and drilling.

The IRA, as a key objective, was to determine the probabilistic timing of all tasks including schedule contingency to P50, to identify when in the year they would occur and also, the level of helicopter resourcing required to ensure that site progress would not be delayed by deliveries to site. In this case, schedule contingency had to be inherently included in all the tasks to model the timing and cost of the scope realistically, as the deterministic schedule excluding probabilistic weather calendars was meaningless.

Conclusions

Contingency and its cousin Management Reserve are project time and cost provisions that are needed in most projects to ensure that commitments made at project commencement can be met at project completion. However, they are contentious and there is always tension between providing too little and too much. In this as in many other areas of project management, hard experience is sometimes the best teacher. The guidance of subject matter experts can provide that experience for those wishing to avoid the pain of missed commitments.

RIMPL NEWS

Quantitative Risk Analyses

Since our last RIMPL Newsletter in December, we have delivered several QRAs for clients:

• An IRA for a substantial manufacturing facility being built on the northern outskirts of Melbourne, which we performed in January;

• An IRA for Santos GLNG project on a significant expansion of their upstream Coal Seam Gas fields in the Roma area, performed in February/March;

• A Schedule Risk Analysis for gas field facilities and pipeline in South West Victoria, performed in March;

• An IRA for the drilling of two exploration wells in PNG for Talisman Energy, performed in April/May.

A further IRA is expected to be performed on a gas condensate project in August.

Project Controls Services

We have also been active providing project planning and forensic scheduling advisory services for manufacturing and commercial building companies:

• Provision of planning services for a substantial goods manufacturing facility being built on the northern edge of Melbourne.

• Ongoing provision of planning services for multiple Australian food manufacturing facility projects.

• Planning and Forensic Scheduling Advisory Services for a Commercial Construction company.

Dispute Resolution Services

• We have recently been appointed to provide Expert opinions in a dispute relating to installation of pipeline gathering systems.

Lecturing

RIMPL’s Managing Director has been delivering lectures and training in Quality and Risk Management to Masters of Project Management students at Victoria University and Risk Management lectures to RMIT University Masters of PM students during the first half of the year.

He is due to deliver similar lectures in the second half of the year to a further intake of students at RMIT University and to an IT-oriented cohort of students at Victoria University.