The Gwinnett County (GA) Department of Water Resources is a publicly owned utility, committed to providing excellent water, wastewater and stormwater services at the best possible value to the citizens of Gwinnett County. The department is widely recognized for service excellence in the operation of two water production facilities, three water reclamation facilities, and 218 pump stations. In addition, DWR maintains 3,667 miles of water lines, 2,654 miles of sewer lines, and 1,325 miles of stormwater pipe. Treatment facilities provide drinking water and wastewater service to approximately 900,000 residents within the county.
GCDWR aims to be widely recognized as a leader in the water industry, and in many areas they have achieved this vision. Gwinnett takes water service and public health seriously. To continue to be a leader in this space, Gwinnett pushes continues improvement through innovation, education, efficiency, and personal commitment.
Among recent innovations adding value to Gwinnett is the recent improvement of the organization’s maintenance planning and scheduling program.
2011: The Beginning
GCDWR first created planner positions in 2011. The planners started mostly with generating PMs, and there was limited documentation of previous work, whether preventive or corrective. Each planner was assigned to a different facility and reported to that facility plant manager. Communication was a problem with everything existing in silos and different planners engaged in different tasks.
By the end of 2011, all of the planners were spending the majority of their time on other tasks. Bad habits included helping technicians procure parts and tools on demand, actually bypassing the warehouse, and even leaving the site to pick up items. Any scheduling was only informal. By all accounts, Gwinnett could not tell that the program was providing much benefit.
2014: Refocus on best practices
In 2014, GCDWR embarked on a number of utility improvement initiatives, including revamping its planning and scheduling program. Working with Doc Palmer, the author of McGraw-Hill’s Maintenance Planning and Scheduling Handbook, Gwinnett implemented new practices and saw a dramatic improvement in work-order productivity. Nevertheless, because the new practices were based on principles that were somewhat unusual from common thinking in industry, Gwinnett created a new supervisor position that had oversight over planners. The trades coordinator of maintenance planning and scheduling performs a dual role. Not only does this individual ensure that planners follow best practices, but also he or she acts as a liaison with the maintenance supervisors (also called trades coordinators) that are over the trades themselves.
Proper planning and scheduling is management-driven, and the trades need continued encouragement to accept the help that the program provides. Generally speaking, working on things before they break to keep them from breaking is not something that operators call for. It’s therefore imperative for management to demand and expect that enough proactive work will be completed, supported through proper planning and scheduling program.
The following sections of this article identify the principles of proper planning and scheduling, along with Gwinnett’s practices to implement and sustain them.
Best practices planning
Palmer’s first principle of planning is simply to protect the planners. Often, companies will begin to use planners as backup trades or backup supervisors. They also begin to place more and more administrative duties on planners. Soon planners do not have enough time to plan new work as it comes in.
To protect the planners from non-planning roles and duties, Gwinnett ended the practice of having planners report to local facility managers. Rather, the trades coordinator of maintenance planning and scheduling now supervises all of the division’s planners. This supervisor keeps them focused on proper planning duties and insulates them from picking up interfering duties and assignments.
The second principle of planning is that planners should focus on future work. Often, in the little time available for planning, planners plan fewer jobs than they could for fear of being blamed for a poor plan. Instead, planners should be running a “Deming Cycle” of continuous improvement, planning the best job they can but not at the expense of planning all the work. And instead of trades demanding help from planners on jobs already started, they should execute to the best of their ability on their own and report feedback to improve future plans.
Gwinnett recognized that planners cannot plan the perfect plan. The new trades coordinator of maintenance planning and scheduling actively supports planning more of the new work, with the goal of making “better” plans over the years. The planning and scheduling trades coordinator also works closely with the other trades coordinators, encouraging them to solicit craft feedback on completed work and, as needed, explaining why planners should not help jobs in progress at the expense of planning new work coming in. (Of course, planners can help in true “all hands on deck” emergencies.)
The third principle of planning is to make component-level files for assets. That means planners should make and keep plans for specific assets rather than have generic job plans. Gwinnett started using the Maximo job-plan module to make living plans for assets rather than plan directly on the work order itself. The trade coordinator of maintenance planning and scheduling ensures that all planners have CMMS “rights” to this module and that planners are planning in the right place. The planners update these living plans over the years as they have more time, especially with craft feedback.
The fourth principle of planning is to quickly estimate job labor hours. A large purpose of planning and scheduling is to help increase productivity, not necessarily to have perfect time estimates. A quick judgment on the part of a planner is generally “good enough” for assigning work and building schedules. The trades coordinator directs planners to make reasonable estimates that hopefully are not too high or too low, and see if they can be improved over the years with trade feedback.
The fifth principle of planning is to trust the skill of the crafts and add more details over time. There is a trend in industry to want to identify simple jobs that should not be planned at all and simply left to the “skill of the crafts.” This misguided notion springs from trying to decide which jobs could benefit from perfect plans and trying to free planners to focus on them.
Again, in this situation planners become doomed to frustration because no plan can ever be perfect. Planners can plan more of the work if they realize their plans do not have to be perfect because skilled craftspersons can properly execute work with the plant that exists. And simple jobs do benefit from planners running the Deming Cycle to improve them—e.g., “This time take a longer ladder, two gaskets instead of one, and metric wrenches.” A simple job plan with simple details can help a skilled mechanic not waste valuable time, especially on a repeated job.
Gwinnett guides its planners to plan all jobs with an “average” skilled tradesperson in mind rather than a brand-new tradesperson. This strategy allows for planning more of the work, and as more details are added to more plans over the years, eventually the plans not only are a great reference for skilled trades, but also they are helpful to new tradespersons.
The sixth principle of planning is that wrench time (if measured) is properly measured with a statistical study. Repeated studies of maintenance forces throughout industry shows that without proper planning and scheduling, typical workforces have about 35% wrench time—time spent actually, actively making progress on a job. This consistency occurs largely because 35% is the point of feeling busy. Proper planning and scheduling should boost this time to 55%.
Gwinnett does not measure wrench time at all. The concept of wrench time merely explains why productivity improvement is possible. Gwinnett uses an internal “craft utilization” metric that is a measure of available time vs. captured time. This is a measure of total craft time related to performing maintenance that is entered into the CMMS. This is not only “wrench time”; it also incorporates other related time such as time spent gathering information, coordinating activities, traveling to and from the job site, etc.
Implementation and supervision of planning practices represent only half the challenge of establishing a successful planning and scheduling program. Planning allows for improving the quality of maintenance work over time through the collection of tips and guidance that can help crafts avoid delays—e.g., a reminder that a 10-foot ladder is required for a specific job. But even with potential delays identified, wrench time does not rise without scheduling. Parkinson’s Law holds that “the amount of work assigned expands to fill the time available.” Thus, a large part of planning serves to support scheduling to ensure Gwinnett assigns enough work.
Best practices scheduling
Scheduling also plays a vital role in the productivity of maintenance by replacing the feeling of being busy (35% wrench time) with a sense of mission (55% wrench time). That’s why shutdowns, turnarounds, and outages are so productive. Everyone is trying to complete a certain amount of work rather than simply making sure everyone has something to do.
The first principle of scheduling is that scheduling requires job plans to have estimated labor hours for the work. Again, The trades coordinator of maintenance planning and scheduling encourages planners to plan enough work so that schedules can be filled.
The second principle of scheduling is to have a credible priority system. The maintenance force must know when something cannot wait or can wait and for how long. If good maintenance forces expect 20% reactive work and 80% proactive work, three overall levels is not enough. Three levels would mean Emergency, Urgent, and Routine. Defining reactive work as “now” or “this week” means that 80% of all the work would be “routine.” There must be some expansion beyond routine to allow sorting out the 80%. Four overall levels would allow routine to be “next week” or “beyond next week.” Five overall levels seems to be even better, allowing work to be completed “this month” and “beyond this month.”
Gwinnett had a five-level system, but three levels required action within a week, and two allowed waiting more than a week. Thus, there were three levels for 20% of the expected work and only two levels for 80% of the expected work. Gwinnett modified its system so that only two levels were reactive (work performed within a week) and three levels allowed proactive action beyond the current week. This change made the priorities more useful for scheduling.
In addition, the trades coordinator of maintenance planning and scheduling regularly encourages requestors to write fewer reactive work requests by carefully considering if the request could really wait and also write work orders for minor problems that maintenance could address in time to head off larger problems. Many times operators do not want to bother maintenance with little problems, knowing that the maintenance workforce is busy. But they need reminding that maintenance wants to know about little things and can be productive enough not to ignore them.
The third principle of scheduling is to schedule for a week at a time, largely as a simple batch of work to provide a sense of mission. Setting the weekly schedule as the fruit of backlog research is a great help to supervisors and productivity. Instead of digging through the backlog to find work for persons to keep them busy, the scheduler does the digging to find enough work for the crew. The scheduler simply comes up with enough work for the crew to provide a sense of mission. The week period is long enough to even out the inaccurate estimates of individual jobs and also to allow bundling of related work on common equipment, a great service for the crew supervisors. The week period is also short enough to allow crew supervisors to try and protect it against some new requests that could wait at least a week.
Gwinnett had not been doing any weekly scheduling prior to 2014. Instead, crew supervisors tried to complete PMs each month and otherwise take care of operators. Gwinnett now uses a fairly simple spreadsheet and macros to sort planned work orders that are dumped out of the CMMS and then match them against available crew labor hours. The trades coordinator of maintenance planning and scheduling helps the planners use the scheduling tool properly.
The fourth principle of scheduling goes hand in hand with the weekly scheduling and insists that weekly schedules should be loaded with 100% of the available labor hours. Only fully loaded schedules can defeat Parkinson’s Law. Underloading schedules, say to 65%, to allow for expected reactive work does not provide enough of a sense of mission to move the trades productivity to 55% wrench time.
Nothing strikes fear so much in the heart of crew supervisors as handing them a full schedule. But again, along with this principle is the notion that it is OK to break the schedule. This unusual concept requires ongoing management as well as supervision to keep it in place. Companies have a great ability to stop doing unusual practices and retreat to traditional ones.
Perhaps this requirement alone makes the case for having a full-time trades coordinator of maintenance planning and scheduling who knows that full loading is the right thing to do and keeps the planners doing it. (At Gwinnett, the planners create the weekly schedule at the end of each week.) In addition, the full loading drives other best practices. The increased productivity requires that more work be planned by planners whose time is protected for the task at hand.
The fifth principle of scheduling is that supervisors make daily schedules. Daily scheduling is supervisors simply assigning jobs to individuals and coordinating with operations on a day-to-day basis. Gwinnett simply left the daily assigning of work to the crew supervisors.
The sixth principle of scheduling involves measuring schedule compliance and being mindful that schedule compliance should be between 40% and 90%. Schedule compliance is actually the measure of whether the weekly schedule was fully loaded or not. Scores above 90% generally mean that the schedule was not fully loaded and so the productivity is probably at 35% wrench time. Scores between 40% and 90% suggest the schedule was fully loaded. Scores below 40% might mean that the supervisors are ignoring the schedule or that the scheduler is not successful at finding appropriate work.
Gwinnett generally scores about 80% schedule compliance, although the scores vary from week to week and from crew to crew. The Gwinnett version of the “Red-Green Report” shows the percentage of scheduled work orders that were completed highlighted in green and the number of nonurgent completed work orders that broke the schedule highlighted in red. Trying to have fewer red work orders encourages supervisors not to break the schedule for work that can wait.
It takes a lot of management maturity to accept less than 90% schedule compliance, because everyone wants an “A” (90% to 100%). But the trades coordinator of maintenance planning and scheduling at Gwinnett continually reminds everyone, management and crews, that proper scheduling is more like bowling. If a person could consistently bowl 200 out of 300, that would still be a good bowling score.
The scheduling principles are as unusual from common industry practice as the planning principles are. Generating enough planned work to schedule, having a credible priority system, using a week as the right timeframe, and fully loading it only to expect less than 90% compliance take ongoing supervision to keep them in place. It has to be OK to break a fully loaded schedule.
Factors for success at GCDWR
Gwinnett’s work-order completion rate rose from a traditional 35,000 work orders per year before the maintenance planning and scheduling overhaul was put in place to more than 58,000 in 2014. Now, for the past five years, the average completion rate is 65,000 per year. In addition, the labor hours to accomplish the work is down to 90,000 from 111,000, mostly from a reduction in overtime. Maintenance planning is definitely the “productivity piece” of maintenance.
Gwinnett attributes its results to a commitment to do the following: Implement best practices; insulate the planners; grow detailed job plans over time; get trades feedback to improve the plans; configure the CMMS where needed to support the way the utility wants to run maintenance; monitor and use the right KPIs in the right way; and seek an appropriate (not perfect) level of schedule compliance. Above all, trades coordinators have bought into Deming's concept of having the freedom to fail. You don’t have to be perfect to get better.