Equipment and plant shutdowns are critical events that over time provide a fertile ground for growing a culture of continuous improvement. At the core of continuous improvement is the Deming cycle: plan, do, check, act. Embedding this cycle into all aspects of work is a fundamental element of any continuous improvement culture.
The Deming cycle starts with a plan, as should every shutdown you execute at your plant. Without a written plan that the execution team attempts to follow, there is no learning. The robustness of your planning process is a significant factor in your ability to learn from its execution. This means how you create your plan is critical to your ultimate success. A written list of work to be performed does not constitute a plan. To enable learning, your plan must be developed with the content, sequence, timing and outcome specified. It must be presented as a hypothesis to be tested as an experiment. If you do this content in this sequence, then the work will be completed in this timing with this outcome. Running this “experiment” will facilitate analysis and learning.
Remember, all plans are not created equally, so the method you use to build your shutdown plan is critical.
Most project plans are developed using the critical path method (CPM). If you utilize Microsoft Project to develop your project plans, you are using CPM. If you’re utilizing CPM, then you are in luck. You have room for significant improvement just by changing your planning method. In the 1990s, Eliyahu Goldratt described a far better method for planning projects that will enable you to improve your shutdowns in very short order. The improved method for project management is called critical chain project management (CCPM). This is not an article on the differences between CPM and CCPM, so I strongly suggest you do research on CCPM and learn why you should be applying it in all projects in your business.
There are two significant problems with CPM that have been addressed in CCPM. The first problem is with the definition of task dependencies. CPM only recognizes the dependencies where one task has to be completed before the next task can start. Tasks can be independent based on this definition, but actually dependent due to a shared resource, such as a skilled individual, a piece of equipment like an overhead crane, or even physical proximity (above or below, for example). The second problem with CPM is its failure to separate common cause and special cause variation that drive task time estimates. The net result of these two flaws in CPM is that the critical path (the set of tasks that determine project duration) changes during the execution of the project, reprioritizing tasks and creating chaos with its resultant firefighting. The chaos destroys your ability to learn.
If a shutdown task has historically taken between three and 10 hours to perform, it would be common to plan this task for eight hours when using CPM. You would naturally want to protect the project plan from the task variability. CCPM has you examine the data, or ask those who do the work, to determine two estimates – a 50 percent estimate and a 90 percent estimate. Ask the craftsmen who perform the above task, “If everything goes as it should, how long would you say it should take so that you had a 50-50 chance of completing it in that time?” This would likely be an estimate of about four hours. Then ask how long the task would take if things go wrong but still have a 90 percent chance on completing on time. The likely answer would be eight hours. The variability in this task around four hours is what is known as “common cause” variation – the normal differences between people and conditions. The four additional hours allow for “special cause” variation – when “stuff” happens – wrong part, missing tool, etc.
Critical chain project management uses the 50 percent estimates to build the project plan, then pools the special cause variation and positions it at the end of the project to protect the project, because “stuff” does happen. This creates a more robust schedule that does not keep changing during execution.
CCPM also takes into account resource dependencies when creating the project plan, addressing the second problem with CPM, making the resulting project plan even more robust. Critical paths inevitably keep changing during project execution. Critical chains almost never do.
Below is a typical project plan using CPM.
Figure 1. Project plan using the critical path method (CPM)
Because there is extra time built into every task estimate, the planned task time is not taken seriously, resulting in task start times not being honored. Imposed on the project plan below is what the actual work time on each task might look like:
Figure 2. When tasks actually get worked
The project plan for this same project is built using CCPM, and the more optimistic 50 percent times are shown below:
Figure 3. Project plan using critical chain project management
The project plan is compressed but still protected by the project buffer, which combines the safety from all the individual tasks into one project buffer that protects the outcome of the shutdown.
Notice that the project buffer does not extend all the way to the end of the CPM schedule. This is the result of pooling all the safety from the individual tasks into one common buffer. When the “special cause” variations are combined statistically, they require less total time. The easiest way to think about this is to consider that for the buffer to extend all the way to the CPM flag, every task would have to take its worst-case time. While that is possible with an individual task, it is unlikely with the combination of many tasks.
Next, let’s identify the critical chain of tasks, taking into account both typical task dependencies and resource dependencies. The critical chain is shown in red below:
Figure 4. The critical chain of tasks is identified.
Here is what CCPM will do for your next shutdown:
There are three key questions every manager must answer:
Answering these questions has been a challenge for many managers.
The first level of management is responsible for its team’s execution of the systems they have been given to run (and improve) the shutdown. This now includes a project plan built using CCPM. The plan provides clarity in answering the first question – what to change. This clarity was likely missing when you either had a list of work or a CPM plan that had its critical path changed every update meeting.
In a typical shutdown, there are hundreds of tasks to be executed. By using CCPM, you have identified the critical few that, if improved, will shorten your shutdown. This answers the question of what to change.
Figure 5. The critical chain of tasks is “what to change (improve)”
How can you improve this set of nine project tasks – “what to change to?” Find your local business-improvement person and ask if he or she knows the single-minute exchange of dies (SMED) system. Better yet, Google it and see if you cannot improve this sequence of tasks with the help of your team. “How to cause the change” has been answered.
Next, look at each of the critical chain tasks and apply your common sense. Do you have a standard procedure to execute each task? If not, develop one. Call it your “current best practice” (CBP) for doing the job. Remember, this is an experiment, so your CBP should specify the content, sequence, timing and expected outcome. It is the CBP because you want to get your team’s input on how to do it better next time.
The timing you specified in the CBP is based on some assumptions, which include having available and handy all the equipment, tools, materials, supplies and information required by your team to execute the work. If you want ideas on how to create a workplace to facilitate the smooth execution of that work, Google “5-S” or ask your local improvement expert for help. Whatever you do, involve the people doing the work from the very beginning.
Every improvement you make on these critical tasks will shorten the duration of the shutdown, right up until you improve enough so that other tasks currently not on the critical chain start impacting. You will see these coming as you improve and build subsequent shutdown plans, enabling you to continue to focus your critical improvement efforts on those work tasks that actually will make a difference.
First-level managers are responsible for executing the work systems provided, which include following the shutdown plan and holding after-action reviews of the work to capture ideas from the team on how to improve next time. The first-level managers with their supervisors and crew are responsible for executing the systems to the best of their ability and helping the second-level managers identify opportunities to improve those systems.
The second-level manager (often called a department manager) is accountable for the results of those systems and for redesigning the systems when the results are not acceptable. It is this level of manager that is responsible for the planning methodology used to plan and schedule your shutdowns. This manager is also accountable for achieving the results required, such a production rates, overall equipment effectiveness (OEE) and so on.
The shutdown fits within the context of a whole set of systems, including systems that determine what work needs to be performed during the shutdown, systems that estimate the scope of each task, and systems to improve all aspects of the business.
By eliminating the chaos of shutdown-schedule execution using CCPM, the second-level manager will be able to identify those systems that are not delivering the required results. Tasks that exceed their planned times can now be prioritized and the critical few identified. An analysis can also be performed to understand which systems failed to produce the outcome expected of them. What to change? What to change to? How to cause that change? All these questions become far easier to answer.
When you focus your limited resources on managing and improving those critical few tasks that really make a difference to your business, you create a winning atmosphere. Everyone will want to come to work and be successful, seeing their efforts payoff as shutdowns become more efficient, work tasks become more efficient, and the workplace becomes more organized around things that actually matter. Do this and your team will become more successful, feel more successful, and be more engaged and motivated at work. These focused continuous improvement efforts you unleash through your team will catapult your business forward.
This article was previously published in the Reliable Plant 2019 Conference Proceedings.