Managing a work-order backlog is not the most exciting of maintenance tasks, but without a complete and up-to-date backlog, important work will be forgotten. Indeed, good backlog management is a prerequisite for effective planning and scheduling.
"Backlog" means different things to different people. There are two common definitions. The first and most common is that a "backlog" is a list of all work that has been approved and will eventually get done. This is the correct definition. It is sometimes measured in trades-hours, but it is better measured in weeks, calculated as the time it would take to complete all the current work in the backlog with the resources that could be applied to this work. This may or may not include PM work.
The second definition is that a "backlog" is just those work orders that have passed their "required by" date. This definition should not be used because it is not logical. Most maintenance departments have a reasonably fixed number of tradespeople who perform work from work orders generated more or less at random.
When a work order is initiated, the date on which the work will be completed depends on its importance relative to the work already in the backlog, which is known, and also the work orders that will be generated in the future, which are unknown. The result is that any "required by" date assigned when a work order is initiated will be just a wild guess and usually wrong. Assigning a "required by" date should be limited to those few work orders that have a genuine deadline. Otherwise, these dates will be in conflict with the objective of always working on those jobs that have the greatest value at any time.
In this article, the first backlog definition will be used. Within this backlog of work orders that have been approved but not yet started, there are sub-groups. These include the "planning backlog," which can be defined as all work orders on which any commitment such as purchasing has been made, and the "ready-to-schedule backlog,” which is made up of those work orders for which all materials and other resources are available so work could start at any time.
Combining all approved work orders into a single backlog can be overwhelming. Instead, it should be filtered into logical components. The following filters are recommended:
Shutdown work must obviously stay in the backlog until the appropriate shutdown is scheduled, which may be a year or more. Leaving this inactive work in the backlog complicates the management of ongoing non-shutdown work, so it should be hidden until the time comes to prepare for the shutdown, when it will be managed on its own. Of course, the preparation work for shutdowns is very important and should be prioritized along with all other non-shutdown work. Separating shutdown and non-shutdown work is also necessary for efficient shutdown planning.
This also would include work for all other categories of maintenance resources, such as area maintenance crews. Remember, the backlog for a maintenance crew should be limited to the work for that crew and must include references to the support required from other crews.
Preventive maintenance work should be pre-planned and pre-scheduled. The instructions for inspections and other routines should be on file and included in preventive maintenance (PM) work orders. The work should be automatically scheduled by the maintenance computer system. Of course, PM work and corrective maintenance require the same limited trades resources and need to be scheduled together, but for the purposes of backlog management, they can be separated. Backlogs are more easily managed if PM work is hidden until the time comes for it to be scheduled.
PM work should be set up in the maintenance computer system so there is a steady workload scheduled for each work day. This allows the manpower assigned to PMs to be constant and considered "untouchable" for corrective maintenance. This way, the scheduling of both preventive and corrective maintenance is simplified.
Considerable discipline is needed to limit work in the backlog to just those jobs that will be completed in the near future. Backlogs should never contain completed jobs, duplicated work orders or low-priority work that no one ever intends to do.
An important part of the discipline for maintaining a clean backlog is to close work orders (or change the status to "physically complete") as soon as the work is done, which should be the same day for non-shutdown work and within a few days for major shutdown work.
The function of maintaining a clean backlog should be included in the job description for a designated maintenance position. It is one non-planning function that is appropriate to assign to a planner. It does not take much time, and the planner is in position to know the status of all work orders in the area.
There is an optimum size for a non-shutdown work-order backlog. If a backlog is too small, it will be difficult to keep tradespeople on priority work. Break-in and unplanned work will increase, and productivity will fall.
If the backlog is too large, a lot of material may be tied up and the backlog will be difficult to control. There will be a loss of confidence that work will be done, and "emotional emergencies" will be encouraged. It can even become easier to submit a new work order than to try to find an existing work order in a large backlog. A large backlog is of little help for work scheduling.
Ideally, the backlog should be of such a size that key maintenance and operations personnel, including the area maintenance supervisor, the operations coordinator and the planner, have a good enough "feel" for what's in the backlog to be able to immediately recognize duplicate work requests.
For a typical 24/7 continuous-process operation, a good starting objective would be to have a "total backlog" of about four weeks, a "planning backlog" of about two to four weeks and a "ready-to-schedule" backlog of one to two weeks.
Note that a "ready-to-schedule" backlog of one to two weeks implies that all the material for this work should be staged somewhere onsite and be ready for use. This kind of materials management has great benefits but will be successful only if a large percentage of the work on the schedule is executed according to that schedule.
Any work added to the backlog should receive some scrutiny from the area decision-makers. A good process is for the area maintenance supervisor and the operations coordinator to review all new work requests each morning. One important function in this step is to decide whether the request is for a "small job" that can be completed immediately and does not justify being planned or scheduled. Allowing anyone to add work to the backlog without review guarantees that it will become disorganized and of little value.
Ideally, your maintenance management system should be used for backlog control, but unfortunately many systems have very weak functionality in this area. Managing a backlog is all about sorting and filtering lists of work. These lists will contain columns for equipment and work identification, scheduling notes, priorities, resources, status, etc. From my observations, there are few, if any, maintenance computer systems that are better at managing work lists than a good spreadsheet. In fact, many organizations cannot effectively manipulate work lists without first downloading them from their maintenance computer system to Excel.
Most maintenance computer systems are excellent at recording maintenance costs against work orders, and should always be employed for this purpose. However, if using them to manage work lists is cumbersome, a well-designed and well-managed spreadsheet process that is integrated with the maintenance computer database should be used.
Backlogs can be easily maintained and kept up to date by utilizing your site's server to manage and share work lists with secured templates and disciplined management.
In conclusion, managers must pay close and frequent attention to backlogs to ensure they are "clean," the estimated times are realistic and that they stay close to the optimum size.