By Cezary Goch

Over the last ten years, I have participated in the planning and execution of 42 turnarounds (TAs) and capital projects in refining and chemical plants. During this time, I have seen many different philosophies in project planning and scheduling. While most clients acknowledge the value a well-defined schedule brings to a project, their drive to invest the time and resources to improve their own schedules varies significantly. This variance occurs not only between companies and locations, but also between project teams working turnarounds versus capital projects. Some adopt the “schedule is our bible” rigor while others are more freewheeling, suggesting their schedule “is only a general guide and nothing more”.

I will argue, those who follow a quality schedule have a greater chance to finish on-time with an uneventful (and rather boring) turnaround, versus an adrenaline pumping, heart-racing reactive ‘schedule as my guide’ approach.

In this article, I will outline what I believe are a few conditions that must be fulfilled to create a quality schedule.

But first, a question:  whose schedule is it?

The schedule will be affected by who is building it and the responsibility they have been given. Three questions will help define this:

1. Is it the company’s own scheduler or a General Contract (GC) scheduler?
2. Is the position of the scheduler and planner split, or are those the same people conducting both jobs?
3. Is the philosophy of building the schedule set and provided to schedulers, or are they allowed to build it in accordance to their individual preferences?

The TA management team should try to avoid relying on schedules which are built by the GC’s schedulers/planners. While most of the contractors are honest and hardworking, building a schedule that best suits the contractor’s philosophy of execution is unavoidable. It pays off in the long run to have in-house schedulers as employees, or contracting them directly for your own use.

Turnarounds and Capital Projects within the TA Window

Philosophy

Primavera scheduling software is very versatile but also a complicated tool; with it, the same schedule can be built up in many ways. For example, one can incorporate a lot of hard logic links and/or many hard constraints that will produce a rigid, difficult to adjust schedule. Alternatively, one can choose to lock down critical path and make everything else floating without priorities. which will make schedule very flexible but will be difficult for contractors to follow. Organizational consistency is important; therefore, organizations are encouraged to set parameters for the schedule so it is built the same way no matter who does it. This is commonly known as a scheduling guideline or scheduling specification.

Specifications typically include topics such as:

• Scheduling philosophy in general
• What methodology will be used
• Defined level of detail for activities
• System setup
• Project setup
  • settings, options and defaults
  • codes and udf’s
  • calculations
• How to structure
  • OBS
  • EPS and WBS
  • hierarchical Codes
  • cost accounts
• Use of Resources and Roles
  • structure
  • types
  • named or generic
• Progress controls
  • procedure
  • timelines
  • %-complete vs. remaining duration
• Reporting
• process (who gets what report at what time with which content)
• update cycle, refresh period, due date and data cut off
• defined requirements
• layouts and level of detail
• summarization, grouping and sorting
• level of detail
• earned value

Drive for awareness and clarification will promote consistent understanding and use across the planning and scheduling teams.

Schedule Detail

An effective schedule must be detailed enough for both planning and execution teams to clearly understand the project status and readily update progress in the field during execution. The schedule needs to include inspections, hold points, quality checks and other breaks in actual work. It must also be easy to optimize. This means that work teams and scope contributors need to be able to understand the sequence of tasks with enough clarity to spot opportunities and risks, while not becoming bogged down under too much detail. A rule of the thumb is to keep durations longer than the schedulers’ ability to process them (the time it will take the scheduler to process the daily work into a schedule update report) 0.5-2hrs and no longer than one shift of work (8-12hrs). Durations by shift length are easiest to monitor because % progress reporting is reduced to minimum necessary.

Collaboration

The schedule is the backbone which everything in the event is connected to; hence, a good schedule affects almost every piece of the execution. For a schedule to be effective it needs to be credible and achievable in the minds of all participants. It should be built together with the contractor executing the work through interactive schedule building sessions, schedule quality review (SQR) sessions, schedule risk analysis and simulation, Theoretical Maximum Performance™ workshops, and optimization and constructability sessions. Each of those sessions has its place at a different stage of schedule maturity. Building the schedule together with the contractor, but retaining control within the company or third-party schedulers and planners, is a first step to buy in and agreement on durations, labour requirements and work sequencing logic by all parties involved. It is also a first step in making sure that the schedule will be followed.

One problem commonly encountered between companies and contractors is misunderstanding about schedule detail expectations. There exist category definitions for increasing levels of schedule detail but they are inconsistent between companies. For our purpose, we recognize four levels of detail – note that we are typically trying to get to Level 3 through our use of SQR before conducting TMP™ schedule optimization sessions.

Schedule Level	Description
Level 1	Milestones (start, finish) and major phases (engineering, procurement, construction, commissioning)
Level 2	Activities broken down by phase (engineering, procurement, construction, commissioning), area (project specific) and by discipline (process engineering, mechanical engineering, etc., or piping installation contractor, electrical installation contractor), down to the level of grouped activities (develop P&IDs, order long lead items, pipework installation, etc.)
	First pass at networking of activities has taken place and a preliminary idea of what the critical path activities are, has been formed.
Level 3	Activities are broken down by phase (engineering, procurement, construction, commissioning), area (project specific) and by discipline (process engineering, mechanical engineering, etc., or piping installation contractor, electrical installation contractor), down to the level of specific deliverables (P&ID X, issued for construction, etc. or Vessel Y installed, etc.) The activities are all fully networked and the critical and near critical path have been identified. Ideally, the schedule has been resource loaded and resource smoothed.
	Manpower is loaded and task schedule is identified and sequenced
	Manpower profile by work period
	Work periods identified
	Supporting operations manpower requirements identified
	Differentiation between contractor and owner workforces
	Specialty equipment identified and potential sources identified
	Applicable Regulatory timelines identified as a constraint
	Intermediate milestone reviews identified
	Owner acceptance and turnover activities identified
	Quality control trigger/hold/verify/record/data points captured
	Schedule to include Operations activities, max limits for manpower density loading by unit when warranted
Level 4	Activities detailed down to squad task level. Typically, only used for very detailed construction or turnaround day-to-day and hour-by-hour activities. (e.g. open man way on vessel X, inspect trays, close man way, etc.) Resources are leveled, baselined and schedule ready for execution

 

Time Line

Developing a high-quality schedule requires time and constant communication. The following are some Front-End Loading (FEL) deliverables which will have the scheduler equipped with enough timely information to keep the scheduling progress moving and avoid gaps. As a comparison, note that any decent gated planning process has anywhere from 350-600 deliverables and milestones, so this list is only a very brief overview.

Activity	Timing before execution	Notes
Define Turnaround and Capital Projects Integration Requirements Plan	34-30 Months	Traditionally, alignment between the TA and capital projects team requires attention. Even small capital content introduces risk into the TA event.Bulk of it may start at T-24 but it has to start much earlier.
Initiate Formal Scope Collection	30-24 Months	Scope collection must be combined with robust scope challenge process. Challenging scope items “in” versus “out” of scope list will provide more time for estimating, engineering and producing quality work packs.
The list of projects is frozen and the design documents are approved.	26-24 Months	Both engineered expense and capital projects is frozen to allow time for funding, design, and procurement and pre-TA installation
Identify Capital Long Lead Materials & Develop Procurement Strategy	24-12 Months
Perform Preliminary Critical Path Analysis - Turnaround	18-12 Months
Perform Preliminary Critical Path Analysis - Capital Projects	18-12 Months
Capital Projects Coordinator to issue detailed Capital Project Pre-TA plan and schedule to TA Planning/Scheduling	15-12 Months
Issue Capital/Expense Engineered Scope Packages (IFC) to Planning	14-12 Months	Typically, late and will affect quality of schedule if not delivered on time
Communicate Work Breakdown Structure (Schedule level 1)	14-12 Months
Issue Preliminary Shutdown and Startup Schedule	14-9 Months	Needs to be integrated with mechanical window and capital schedule
Scope freeze	13-9 Months	Track scope creep. It is a good practise to set a goal for maximum scope growth. No more than 12% for both: scope creep between scope freeze date and executions and discovery work.
Complete Critical Path Preliminary Constructability Reviews	10-7 Months
Schedule level 3	6-4 Months
Schedule optimization	5-1 Months	TMP sessions, SQRs
Schedule level 4	1 Month - 2 weeks

Schedule quality and detail must be the same for TA and capital projects executed within the same TA window. The attributes and values in both schedules must follow the same procedure to unify vocabulary and provide opportunities for interconnection and interlinks. This is crucial for successful event and is frequently a problem, especially when capital projects and turnaround planning are undertaken by distinct teams and even more so when these teams are separated geographically.

To achieve this, there must be one integrated schedule with single-person accountability, ie. has one person responsible either form TA planning group or projects group. This integration is understood by most clients even though there are struggles with perfect alignment between these two organizations of TA and capital projects. What quite often is neglected or overlooked is the additional integration of shut down and start-up activities into the schedule with similar level of detail. The best practise is to include pre-TA work in the same schedule. Critical and near critical path are identified. The premise of the schedule is to be followed, therefore it has to have a buy-in from all participants and updates of the schedule have to happen every day. Discovery work, if not already accounted for within the schedule (anticipated additional time marked at the task level), is added at the task level as additional time or separated line.

Large Capital Projects

Approach

A capital project schedule, especially large capital construction (execution longer that 1 year), will be different by design from a TA schedule. First, it is based on norms versus historical data. Work that is described in the schedule often has never been done before, at least not on this project in this location with this team. A further big difference lies in the duration of execution. Projects with durations much longer than typical turnarounds, are more difficult to predict and therefore may not be accurate if detailed. An example of these types of projects is ship building which can go on for years.

 For those reasons, large construction projects are built with the schedule that has a low level of detail, with a large float attached to every activity, extremely flexible (you can literally grab half of the activities and move them in time from the back to the front of the schedule and it will still make sense). This schedule is used to run many scenarios as it progresses to determine next best approach. It reflects campaign style approach to construction. Thus, this schedule will display the number of hours that remain to finish the job and give a time frame of when that job should be finished. You could have a five-hour job that needs to be finished anytime in the next couple of months and it is up to the contractor to choose the proper sequence with in the campaign.

Also, rework, constructability issues and unplanned activities are typically an issue in those kinds of projects. Increase amount of actual man hours versus the planned during the execution can reach up to 100% and more. With the philosophy of schedule flexibility, it is difficult to plan this kind of work. Often, work cannot be undertaken as scheduled because many resources are diverted deal with rework, scope growth and non-progressive work. This consumes slack time and the man hours are not reflected in the schedule. Work is executed based on the look ahead where details come from resource level. Therefore, it lists all the activities but without the logic attached, not unlike a large task list. A one-week look ahead is produced in the form of the list of the activities that have to be worked on. Actual detail is obtained from the resource level of the schedule and work cards (Work Packs) after that it is up to the contractor supervisors to determine sequence of work, resources and materials needed. With this approach, it is crucial to have at least two week ahead for planning purposes as the material on site is quite often an issue.

During the project, float is consumed and critical path starts to emerge. A decision must be made to freeze the schedule and transition to a shutdown mentality. This means there can be no more changes to the schedule, detail gets added, baseline the schedule and start the process of detailed monitoring. Variances to the critical and near critical paths, and actual vs. plan metrics must be measured and tracked. Productivity metrics for the uncritical volume of work are necessary to understand, if the project is progressing on time and budget.

This is a good opportunity to use Schedule Quality Review and TMP™ sessions. Schedule optimization to adapt the schedule from construction baseline norms to actual sequential times based on in-the-field understanding of task duration and sequencing will produce both savings and improved schedules. Plus, involving contractors will promote a greater understanding of the plan, reduce the risk of unforeseen events and improve cooperation among trades.

As the project comes to the end, there is a tendency to run it without schedule and planning rigour. For the contractor to be successful, schedulers must include more detail and switch from campaign to system completion. Many times, the capital project schedule can not produce that kind of detail so an old-fashioned skyline chart would do the trick.

It is also good practice to include a “System” as an activity code for each task. This allows for a much simpler transition to system focus and can help easily identify tasks remaining for system turnover.

 The construction industry is moving towards “Building Information Management “(BIM) systems which create a 3D model of the completion rate of the project based on the original schedule.  This helps facilitate client communication but assumes that the project build is progressing per the schedule. If the schedule is vague, the updates are questionable at best. Capital projects need to transition from a schedule built on norms to an integrated schedule owned by the discipline and project leaders to achieve success.

Assessing Schedule Quality 

There are two aspects to the schedule quality and they both apply to TA and construction schedules. First, there is the technical side of the schedule, that is, the quantitative data describing the schedule (ie. Number of lines, amount of slack, occurrence of predecessors and successors, missing dates, etc.. There are many ways of addressing this but the Defence Contract Management Agency (DCMA) provides us with a very good standard.

Defence Contract Management Agency Schedule Assessment (DCMA-14)

1.       Logic

The number of activities that are missing a predecessor, a successor or both should not exceed 5% of the activities within the schedule

2.       Leads

This check identifies activities that are carrying a lead (also known as a negative lag). The DCMA require no leads (0%)

3.       Lags

Total number of activities with lags should not exceed 5% of the activities within the schedule.

4.       FS Relations

Total number of activities with Finish to Start (FS) logic links >90%.  Non-typical tasks limited to <10 % of total tasks.  

5.       Hard Constraint

Number of activities with hard or two-way constraints should not exceed 5%

6.       High Float

Number of activities with total float greater than 2 months (44 working days) should not exceed 5%.  (For TAR, I would recommend no more 72hrs)

7.       Negative Float

No activities that are incomplete should have total float that is less than 0 working days.

8.       High Duration

The total number of activities that have a duration longer than 2 months (44 working days) should not exceed 5%. (For TAR, I would recommend 24 hrs)

9.       Invalid Dates

There should not be any invalid dates in the schedule defined as activities with planned work in the past or actual work in the future indicated by all actual dates being prior to the data date and all forecast dates being on or later than this date. 

10.   Resources

All incomplete tasks should have resources (hours/$) assigned. This check verifies that all tasks with durations of 1 or more days have $ or resources assigned.

11.   Missed Activities

Number of activities that have slipped from their baseline dates should not exceed 5%.  Identifies tasks that are supposed to have been completed (prior to the status date) with actual or forecast finishes after the baseline date, OR have a finish variance greater than zero.  

12.   Critical Path Test

This check evaluates the project’s network logic, particularly for the critical path and is a ‘what-if’ test performed directly on the schedule. It’s intent is to identify a current critical path activity, to grossly extend its remaining duration, and note if a corresponding extension occurs to the project completion date. 

13.   CPLI

Calculation of the Critical Path Length Index (CPLI) verifies that the critical path makes sense and that the critical path is "believable" Ratio of critical path length + total float to the critical path length should = 1 (>1 favourable; <1 unfavourable). 

14.   BEI

Calculation of the Baseline Execution Index (BEI). BEI is the ratio of the number of tasks completed to the number that should have been completed by the status date. BEI should be > .95 (>1 favourable; <1 unfavourable).

Schedule Operating Rhythm™

The second criteria of schedule quality evaluates whether there is an Operating Rhythm™ around it. By that I mean, a defined series of interactions where focus, accountability, involvement and response bring the right data to the right people for effective decision making. Specially, a few elements are crucial:

• Schedule optimization and constructability reviews
• Look ahead meetings
• S-curves, KPIs

Look-Ahead Meeting Expectations

A Look-Ahead meeting is a once-per-shift, or daily, meeting where those in leadership, accountability or scheduling roles come together to quickly assess progress and strategize next steps. An agenda may look like this:

1. Safety
2. Schedule Update (can be done by Lead Planner or scheduler)
   1. Summary View Graph/Chart to review overall status
   2. Display a Chart with S-Curves that demonstrates one of the areas (Exchangers, Drums, Electrical, etc.) where we are successful.  This will be reviewed with the entire group and shown what good looks like.  This should be shown every day.  Rotate groups if successful in more than one area.
3. Critical Path Update.  LP asks specific questions to contractor responsible for critical path that will ensure there is mutual understanding of the expectations, such as:
   1. Where will you be in 12 hours? 24 hours?
   2. Do you need any support or help from others to complete on time?
   3. What materials do you still need to undertake your work?
   4. What will prevent QM packages being complete by end of shift?
4. “Around the Room” with the remaining contractors asking questions.
   1. Do you need support/help from others to complete on time?
   2. Are there any materials or equipment that you need?
   3. QM packages complete by the end of this shift?
5. “Around the Room” with client’s personnel; Ops, Engineering, etc.
6. Lessons Learned
7. Recognition.  Take time to recognize good results, actions or behaviors observed.  Allow others time to recognize as well.
8. Close Effectively.  Use this opportunity to wrap-up meeting with positive feedback to the group.

Note:  Action Items to be captured throughout Agenda items 3-6

Chances of succeeding

Typically, probabilistic Monte Carlo type evaluations are adopted. To measure a statistical confidence level for an estimate (schedule), a Tornado sensitivity chart is used.

There are two approaches to how to read the results of the early and late finish probability exercise:

First one where P10 means that there are only 10% chances of succeeding and P90 means 90% chance of succeeding. This is the most common understanding of the risk analyses.  



The second understanding of the probability is somewhat different and not used as often as the first.

“P50 is defined as 50% of estimates exceed the P50 estimate (and by definition, 50% of estimates are less than the P50 estimate). It is a good middle estimate. Mean and Expected (same level of measure just different names) usually lie about the P40-P30 levels in oil field evaluations and are therefore high estimates. P90 and P10 are low and high estimates respectively. P90 means 90% of the estimates exceed the P90 estimate. It does not mean that the estimate has a 90% chance of occurring – that is a very different concept. The central limit theorem indicates that the P50 estimate has more chance of occurring than the P90 and P10 estimates” – (Cooper Energy web site)  

References:

Cooper Energy web site.

http://www.cooperenergy.com.au/our-operations/glossary/p50-and-p90-mean-expected-and-p10

DCMA 14-point Assessment Metrics. White Paper. Mosaic Projects

http://www.mosaicprojects.com.au/WhitePapers/WP1088_DCMA-14-Point.pdf  

Pat Weaver. “Using probability to analyse your project”. Project @ Manager.com. Apr 21, 2015.

http://projectmanager.com.au/using-probability-to-analyse-your-project/

About Us

RLG International partners with organizations who share the belief that capturing the hearts and minds of their workforce will produce significant performance improvements. We bring a powerful set of performance tools with which we target specific operational opportunities, involve people and enhance results. RLG acts as a coach, a catalyst for change and an expert implementer. For over 34 years and 2500 projects, we have delivered step-change results for clients around the globe.

Read our entire story and see our results at www.rlginternational.com.