Risk in Mega Projects Seminar and Workshop Mega

Risk in Mega Projects Seminar and Workshop Mega Projects: Characteristics, Challenges and Causes for Cost and Schedule Overruns 5 - 7, 2007 Trondheim, Norway Dr. George Jergeas PEng. Professor of Project Management University of Calgary, Alberta 1

Agenda ■ My Objectives ■ Introduction ■ Challenges and Characteristics ■ Reasons ■ Planning and Execution phases ● Lots of questions ■ Feasibility Study ■ Trend System ■ Some conclusions 2

My Objectives ■ Analysis of our current practice ● Not to blame anybody ● Lessons to be learned ■ More questions than answers ■ Momentum for further discussions ■ Provide some initial recommendations 3

Mega-Projects ■ Cost > $1 billion? ■ Significant interfaces / complexity? ■ Most significant issues & risks must be managed at a level above the project team ■ Fast tracked ■ Many players with different interests and motives The challenge is PREDICTABILITY! 4

5

Mega Project Challenges ■ No major problems re quality, health and safety, regulatory and environmental ■ Projects running in excess of design capacity ■ Making huge profits ■ No unskilled or unprofessional conduct by Engineers or professionals ■ Significant cost/schedule overruns ● Many reasons ■ Blame unfairly placed on workers 6

Mega Project Challenges ■ Small vs. Mega project ● Overrun on small projects are not recognized ● 50% overrun on $100 K ● 50% overrun on $4 billion ■ 20 - 100% range equates to hundred of millions of dollars, in supplemental funding and foregoing many months of productive operational activity 7

Mega Project Challenges ■ Define and schedule work for 10, 000 people every day ■ Organize, order, store and retrieve 80, 000 material items ■ Manage worker turnover that can reach 300% annually 8

Mega Project Challenges ■ On $2. 5 Billion project ■ Engineering effort ● 3. 5 million workhours at a cost $100/hr ● 40 - 50, 000 design drawings ● 10 - 20, 000 vendor and shop drawings ■ Construction effort ● 15 million construction hours ● Labour force of 10, 000 workers with a turnover of 30, 000 people ● Supported by 500 - 800 staff personnel 9

Mega Project Challenges ■ Field labour approx. 5000 workhours per million dollars of capital ie. 10 - 15 million workhours at $85 to $100 per hour. ■ If labour production is not carefully managed this could easily double as it has done historically. 10

Mega Project Challenges ■ The task of managing a craft mix of 10, 000 workers working in pairs doing at least two different activities per day results in a never ending 100, 000 individual jobs in a 10 day shift ■ Each job requires a combination of the correct, materials, location, access, tools, equipment, scaffold, safety, quality, rigging, consumables, welding, x-ray and many other inputs to allow the worker to get his job done. ■ This task belong to management which to date has not been able to plan, organize or execute 11

Causes for Cost and Schedule overruns 1. Unrealistic or overly optimistic original (AFE) cost estimate and schedules ● The under appreciation of project complexity, interfaces, interdependencies and risks. ● Underestimating the cost to attract and maintain the labor. ● Underestimating the direct and indirect costs of overtime. ● Craft wage increases to attract personnel to the location. ● Regional and national demands on labor, including that from other mega projects. ● Under estimation of the labor productivity loss associated with working in cold weather climates and locations with severely shorter daylight hours in northern regions. ● Shortages of skilled labor and lower than anticipated labor productivity due to mismanagement of the construction phase. ● High labor turnover mainly due to the harsh working environment and competition between employers attracting labor. ● Transportation costs (including custom cost) are generally underestimated for permanent materials, construction equipment, personal, staff, etc. ● Environmental and regulatory compliance costs are not given sufficient consideration. 12 ● Material cost for both permanent facilities and temporary facilities are not sufficiently escalated during the project budget development phase. .

Causes for Cost and Schedule overruns 2. Incomplete scope definition or inadequate Front End Loading ● Mainly due to the fast-tracking nature of mega projects an ● Ongoing changing customer requirements. ■ Lack of understanding of the cumulative impact of scope changes on project cost and schedule. ■ Fast-tracking: The client’s and engineering firms’ practice of pushing work to the field early puts construction under an unrealistic compressed schedule with increased overtime requirements and often with little or no cost consideration for the field cost. 13

Causes for Cost and Schedule overruns 3. Inappropriate project strategies for the mega oil sands environment ■ Some project strategies deployed do not properly consider the level of scope definition, the fast track nature of the mega project environment, market condition, owner participation, owner control and owner risk. ■ Improper or late consideration of the following project strategies adds to cost overruns (ECI, 2002): ● Project management strategies such as risk management, project control, change control, communications, organization and responsibilities. ● Contract strategies relating to management, design, construction and commissioning services. ● Design strategies such as contributions from client business, operation, project team, contractors and suppliers. ● Procurement strategies including preferred suppliers, progressing, inspection and expediting, receipt, storage and management, spares and documentation. ● Construction strategies including site management and organization, site layout, power, utilities and drainage, work breakdown structure, construction method, off-site prefabrication and assembly, schedule and milestones, industrial relations, and pre-commissioning. ● Commissioning strategies including responsibilities, schedule and integration with construction, resources, training and validation, engineering and trade support, and provision of operating materials. 14

Causes for Cost and Schedule overruns 4. Mismanagement of the construction phase: ● Later than anticipated engineering, vendor data, equipment and material deliveries. ● Poor project controls. Nobody on the project has single point responsibility except the client who does not control much of the work. ● Inadequate plan of execution and poorly defined tasks and division of responsibility. ● Lack of knowledgeable leadership in the engineering, procurement, construction and start-up of mega/major facilities. ● Inexperienced or poorly equipped project management personnel and supervisors coupled with the inability to understand, plan, adapt, implement project management procedures or systems. ● Lack of standardization and fit-for-purpose including inadequate use of shop fabrication, modularization strategy and constructability reviews. ● Poor communication, team work and alignment between the players leading to adversarial relationships and protracted disputes. ● Poor site organization and layout leading to excessive time wastage and productivity loss during construction. ● Joint venture (JV) of project partners, contractors and engineering firms that are not aligned or not set up to work effectively. 15

Project Development & Execution Process (PDEP) AFE PHASE 1 IDENTIFY & Assess Opportunities Determine Project Feasibility and Alignment with Business Strategy 1 PHASE 2 2 PHASE 3 SELECT from Alternatives DEVELOP Preferred Alternative Select the Preferred Project Development Option Finalize Project Scope, Cost and Schedule and Get the Project Funded AFE = Appropriation For Expenditure 16 3 PHASE 4 EXECUTE (Detail EPC) Produce an Operating Asset Consistent with Scope, Cost and Schedule 4 PHASE 5 5 OPERATE & Evaluate Asset to Ensure Performance to Specifications and Maximum Return to the Shareholders

Mega Project Performance Factors ■ Front-end Planning Stages ■ Execution Stage 17

Investigation Needs ■ Study factors affecting project performance during: ● Front-end Planning stages ● Execution stage 18

Front-end Planning Stages 19

Front-End Planning ■ The front-end period up to the point of official endorsement (project sanction) to proceed, where AFE for full budget funding occurs, and contract ratification with a major EPC contractor for project execution takes place: 1. What accuracy of estimating and % of engineering definition? 2. How “Targets” are established? 3. How “Targets” are agreed to by the Owner and EPC contractor? 4. Suitability of methods employed? 20

Front-End Planning 5. Are project cost-schedule targets reliable? How verified, by whom? 6. What basis is specified for required EPC definition to support project budget approval, schedule, and sanction? Who sets? 7. Are the required deliverables and performance expectations for work during execution stage clearly defined, understood, accounted for in the targets, and mutually accepted? Are exceptions noted & quantified for potential risk/impact? 21

Front-End Planning 8. Are the cost-scheduling methodologies selected appropriate and capable of generating reliable targets given the level of scope definition & uncertainty? How reconciled? 9. Are scope uncertainties adequately captured & quantified for impact on cost-schedule targets? How done, by whom? 10. Are cost-schedule data representative of real-world conditions, given scope definition, and cost-scheduling methods employed? How applied/verified? 22

Front-End Planning 11. How are contractor/owner cost estimates & schedules reconciled prior to execution? How are deviations resolved/accepted, and resulting targets deemed to be achievable? By whom? 12. Are cost-schedule upper thresholds established, and overrun tolerances identified? What risk management practices are put in place to protect project targets, the owner & contractor? 23

The Feasibility Study ■ A very comprehensive document ■ Considerable amount of work to scope the project, select the technologies to be used and present a business case ■ Lots of time and effort: ■ What Owner wanted to do ■ How much it would cost ■ What the economics of the investment would be and ■ What the risks might be 24

The Feasibility Study ■ No sufficient level of detailed work done to achieve the level of accuracy and confidence that the owners decision makers place in it when a project is approved ■ To achieve a level of accuracy of the estimate of 10 - 15%, we need a 30% engineering completion ■ The right 30% - not just any 30% ● in particular considerable technical information/ engineering is required from equipment vendors 25

The Feasibility Study ■ Limited input from Operations ● May be unknown at this time ● Operations input is invaluable when considering: ■ Plant and equipment layout ■ Equipment selection ■ Determining the ongoing operations costs and staffing needs ■ Laying out the pre-commissioning and commissioning requirements and costs, start-up and early production projections. 26

The Feasibility Study ■ Downplays or overlooks Organizational Performance - both at the Joint Venture level and at the Project level ■ New companies with new and unproven technology in industry new to partners ● Complexity on complexity ■ Benchmarking and Risk Analysis services ● External reviews ● Several concerns are usually raised ● Cost and schedule projections are not aggressively challenged 27

The Feasibility Study ● Not challenged aggressiveness of start-up and the production ramp up curves that follows start-up ■ Optimism and a lack of meaningful operations input contributes to false hopes on the part of the owners ● Overlooking potential impact of new technology and the inefficiencies and problems that could result from a new joint venture and sometimes new players 28

The Feasibility Study ■ It is not easy to determine just how seriously the project team take the concerns that are raised in the external reviews and respond to them. In some instances the concerns are not referred to at all and in others, while there is an acknowledgement of the concerns and the issues, there is only a comment that they are being addressed and that mitigating actions will be taken. 29

Delays in Engineering ■ Early delays in achieving key milestones such as: ● Substantial Completion of Engineering ● Freezing Process Flow Diagram’s (PFD’s) ● P&ID issued for design ■ Delays do not seem to be reflected on final project completion date ● Fast-tracking the fast-track! 30

Delays in Engineering ■ Before the production of construction drawings can be taken, a number of development steps must be completed: ● Production of PFDs - show the logic of the various chemical steps that will be used in the new facility. ● The quantities of each of the process streams will be shown ● Review and approval of owner before the next development step is released 31

Delays in Engineering ■ Develop the Process and Instrument Diagrams (P&IDs): ■ Identify all the pieces of equipment required for the process to work ■ Identify all piping required together with identification number, size, wall thickness and metallurgy, valve locations and type as well as control logic and hookup ● P&IDs contain critical engineering detail that must be agreed upon before the construction drawings can be started 32

Delays in Engineering ■ Series of extensive, time-consuming reviews - by the engineering consultant and the owner before P&IDs are approved and the detailed discipline engineering can begin at earnest. ■ What is worrisome is that the PFD’s and P&ID’s Milestones are some of the first Milestones on the Schedule and it is only once they have been reached that succeeding activities can be started. ■ In fact the Engineering build up can only take place once these early Milestones have been achieved - hence their importance. 33

Delays in Engineering ■ This undoubtedly would put pressure on the completion of the succeeding activities. ■ The delay in Engineering can also be caused by: ● Additional work by the steady stream of trends ■Either adding to original scope or requiring work already completed to be re-done ■Slower build-up of the engineering workforce than planned ■Lower engineering productivity than expected ■ Whatever the reason(s), how do Project Managers react? 34

Multiple Choice: How do the PM react? A) Mechanical Completion and Start-up dates will be changed to reflect the delay in Engineering B) Mechanical Completion and Start-up dates will not be changed C) All remaining activities will be squeezed into a duration less than originally planned D) The overlap of Engineering with Construction activities will be increased 35

How do the PM react? Answer is B, C and D ■ All the remaining activities will be squeezed into less than originally planned, and it is likely that the overlap of Engineering with Construction activities will increase ■ Assuming the overlap in the original schedule was optimal, then the new overlap would be less than optimal. ■ This will almost inevitably bring inefficiencies into the execution process and likely cause additional rework, thus pushing the costs up and putting more pressure on the schedule completion dates. 36

Trend System ■ Normal practice is that all changes to the base (AFE) be documented so that the potential cost and schedule impact can be flagged and estimated. This gives the PM a series of snapshots as the project develops as to whether and to what extent time and cost may be affected. ● History has shown that while the trends flag individual activity changes, they do not give the full picture. 37

Trend System (cont. ) ■ Invariably, subsequent project re-estimates and assessments indicate higher levels of cost variations than would be expected from the trend indications. ● Trends are classified in four ways: ■Design development ■Changes to the estimate ■Estimate Omissions ■Changes in Execution Strategy 38

Trend System (cont. ) ■ When a trend is raised, the originator has to indicate not only the estimated cost of the trend but the schedule impact of the trend on the activity being trended. ■ It is quite likely- especially in the first few months of the project- that the schedule impact resulting from individual trends would not change the Mechanical Completion and Start-up dates ● PM can mitigate the delays, by adding people. . . 39

Trend System (cont. ) ■ However, the PM should regularly be assessing the cumulative effects of the trends on the overall schedule. ■ As the number of trends climbs, one thing is certain - more work and more costs are being added to the project, and more hours must be spent before the project is completed. ■ This means that either the end dates slip or the additional work identified will take away from the flexibility that was originally in the schedule (if any) thus making the probability of achieving the schedule less likely 40

Cost Allowances and Contingencies. ● To cover for design changes, material quantity increases ● Historical and risk based ■ covers for money that is going to be needed, but cannot be allocated to specific activities or areas at this early stage ● The PM should be “running down” the allowances and contingencies according to some agreed upon plan ● The run down of the allowances should match the additional costs being identified through the trend process. ● With the huge number of trends raised at this early stage, it may well be that the allowances are proving to be inadequate for additional costs being identified. ■ Warning signal to the PM that events are not evolving as expected 41

Industrial Capacity and Skilled Labour Availability ■ Feasibility studies usually underplay the skilled labour availability. ● Claim that project would not be affected by labour shortages and project would fit neatly into “lull” between other projects ■Slow reaction to Alberta Workforce Supply/Demand Forecasts - COAA ■Increased labour cost associated with shift changes to make it more attractive to construction workers to come the job ●Additional costs 42

Other Questions ● Are Canadian market & Geography factors under-estimated? Are they unexpectedly influencing project outcome (e. g. remoteness, climate, terrain, operability, supply/logistics) ● Are the availability& resourcing of skilled contractors, services, attrition, competing work opportunities? ● Are project learnings being captured in the EPC market? What barriers to continuous improvement in the PM practices exit, in which areas, and why? 43

Other Questions ● Are new technologies &/ or designs impacting progress/acceptance (e. g. familiarity, operability & risk concerns)? Are increasing specializations complicating PM interfaces, processes and outcomes? ● What are the effects of business partnerships, joint-venture, contractual interfaces … on outcomes? ● Do Mega projects present unique problems/issues? 44

Other Questions ● How is project definition best enhanced in the planning stage to reduce uncertainties by ■ Simply extending engineering timeframe ■ Applying concurrent engineering methods based on systems or sub-projects ■ How can cost-estimating and scheduling data be enhanced in reliabilities? Are practices limited to scaled adjustments from historical data, or can real-market information be obtained and relied upon. What level of definition and kind of commercial/contractual relationship would be required to obtain reliable market information to support project scope, cost & timing 45

Managing Mega Projects Execution Stage 46

The Execution Stage ■ The back-end of the project, and the period from the conclusion of the planning stage when the project is sanctioned and AFE funding is granted, to commissioning/start-up of industrial facilities. ● How effective teamwork and PM practices to manage and control execution of scope? ● The role and interaction of both Owner and EPC contractor ● The suitability of the practices employed 47

Execution Stage ■ How effective were PM practices in controlling progress & mitigating risk/uncertainty? 1. Are the cost-estimates & schedule from planning stage suitably structured for management/control purposes during execution? If re-formatting is necessary, how are funds/activities accurately re-packaged, using what methods? How are variations accounted for & resolved? Who decides? 48

Execution Stage 2. Are commitments made & monitored during execution with knowledge of budget provisions? Are controls in place to detect & manage significant variations 3. How well do teams/stakeholders interact. Mechanism to track relationship, manage issues 4. How are deliverables reviewed/ accepted. What is the process for acceptance? How is quality insured? How much how often do delays and/or re-work occur? 49

Execution Stage 5. How well are risks & uncertainties managed & controlled during execution to keep the project on budget/ on-schedule? What methods are used? 6. How promptly and accurately are scope changes identified & captured. Is there an approval process in place? How significant are scope changes & variations on the overall project outcome 7. How is performance managed & progress managed? What practices are employed to keep project on track? How effective are they with the team? 50

Other Questions ● Can cost-estimating & scheduling practices be structured so they carry-over from the planning to execution stage and eliminate the risk of unrecognized gaps in “re-translating” targets into management/control formats? Could a transferable WBS methodology in conjunction with a project commitment plan be used in both stages to define & control budget-schedule targets? ● Can/should owner & contractor integrate more closely in the front-end to jointly determine project approach to planning & control 51

Construction ■ Engineering working on the wrong sequence ■ Owner does not know what sequence they need ■ Construction demand a construction-driven engineering ● Sit down with engineering 52

Construction ■ Use density and complexity to decide which areas to start design ■ Engineering have to design by Systems, then develop General Arrangements drawings by area 53

A Partial Solution ■ Have a Vision ● Get some one who built a similar project ● Vision for construction and commissioning ● Engineer must visulize the plant operating when developing PFD’s and P&ID’s ■ Turn the project by system for commissioning ● 150 turnover packages ● Steam system, condensate system, electrical system, pumps. . . 54

A Partial Solution ■ Study plot plan and its sub-blocks ■ Decide the density and complexity ● Estimate total Manhours or work effort ● Divide total work manhours between the sub-blocks ● Estimate manpower required by dividing area (Square Feet) of sub-blocks by 200 i. e. , 200 SF/man ● Draw a schedule based on density and complexity. ● Level manpower and optimize schedule, using preassembly, overtime, two shifts, weekend work … etc. 55

A Partial Solution #3 Air cooler 25% #2 Heat exchanger-30% #1 Pipe Rack - 15% #4 30% 56

Design by Discipline Process Long lead items Mechanical/vendor print Civil Electrical Instrumentation 57

Design by Discipline ■ 30 manhours per drawings ● Includes meetings ■ Drawings Control ● Use RACI chart to control 58

Class Discussions ■ What strategies currently being applied to deal with risks in project development and successful execution? 59

60