In Part 1 of this three-part series, published in the July-August 2009 edition of Machinery Lubrication, I talked about what makes a good preventive maintenance (PM) task and how we need to capture explicit knowledge about each task so that it can be executed with consistent and continuity. Key to this is to ensure that the work instructions include sufficient detail such as grease type and quantity to take the guesswork out of lubrication. In Part 2, I will explore the way in which lubrication PMs are actually provided to the technician, mechanic or operator. Specifically, I will outline what I like to call "hybrid" PMs and the problems they create for controlled management of change.
In theory, each lubrication task should have a work order associated with it. The work order is raised, the work is executed and the work order is closed. However, in an average-sized plant, the number of routine lubrication activities that must be executed on an annual basis can total as many as 100,000 or more. This includes tasks such as bearing regreasing, oil changes, inspections, filter changes, breather replacements, oil top-offs and oil samples. Because of the sheer number of tasks involved, it's often unrealistic to suggest that a discrete work order be created for each and every task. In fact, one of my consulting clients estimated that the cost to raise, execute and close out a single work order for every single task would equate to an amount equivalent to 15 percent of their maintenance budget just in administrative overhead!
Instead, we need a way to aggregate tasks into larger blocks of work so that a work order can be issued for a collection of related tasks, thereby reducing the administrative overhead and work order costs. In practice, there are two ways companies do this. The first is to create lube routes, an aggregation of lubrication tasks based on a common frequency, area or task type. The second is to combine lubrication with other time-based maintenance activities into PMs based on a specific machine, group of machines or area of the plant.
Let's examine these two commonly used approaches and talk about the problems they create and, more importantly, how to overcome these problems.
'Classic' Lube Routes
In an ideal world, the lube route is the perfect answer. Group activities based on commonality of frequency, task type (e.g. motor regreasing) or area of the plant and execute in one block. It's efficient, effective and cuts down on paperwork by having just one work order for tens or even hundreds of tasks. The key here is "ideal world" - few if any of us live in this utopia, which causes some very real problems when it comes to executing what I tend to call "macro" lube routes.
First, let's talk about management of change. Let's assume that you want to change the frequency of one or more of the tasks on a route. To do this properly, the only option is to go back into our routes and refactor the lube routes, moving the task(s) with the new frequency to a different route. Or, how about changing the task details (i.e. lubricant type or quantity of grease), which I wrote about in Part 1 of this series. Again, we must delve into the route and edit the work instructions. In reality, this rarely happens, causing a deviation from optimal lubrication over time.
The next issue with macro routes is the addition or subtraction of machines. I recall one plant where only 80 percent of machines were on its lube routes because the addition and subtraction of machines over time had rendered the routing obsolete. Just like changing detail or frequency, any addition or subtraction of machines requires us to edit the route. And since many machines have several different tasks on different routes, we must edit not just one but several routes at once. Again, few plants take the time to modify the routes, resulting in work either being missed or performed at a suboptimal interval.
The final issue is with the tracking of lube PM compliance. With a macro route (one work order for many tasks), if we miss just one task, we're faced with an impossible conundrum: Close out the work order because we completed 99 of the 100 tasks, or leave the work order open until the last task can be completed. In reality, most would choose to close out the work order - after all, 99 out of 100 isn't bad, right? The issue is that the incompletion of one task is usually due to a systemic problem. For example, the machine was not running but needed to be for completion of the task, or vice versa. It's not at all uncommon to find machines that simply haven't been lubricated in plants for this reason.
What's required to resolve all of these problems is a more dynamic approach to lube routing. With dynamic routing, instead of compiling a list of tasks up front (including tasks details, area, lube points, etc.) and saving this in perpetuity as a "lube route", we should generate routes on demand. Whenever the work order system generates the work request, our knowledge management system compiles a list of machines and tasks that match the lube route criteria (currently due, same task, same area, etc.) on demand. The advantage is that any change in the underlying data (change in lubricant, addition of a new machine, change in task frequency) results in an automatic modification to the lube route to reflect the plant's current reality.
The second most common method of aggregating lube tasks is the "combination PM". In this mode, different maintenance tasks (lubrication, mechanical inspections, electrical checks, adjustments, time-based component replacements, vacuum and steam checks, etc.) are combined into one PM for a single machine or group of machines.
Some of the same problems apply to the combination PM as to the lube route. Any change in a single task frequency requires a complete refactor. But now, refactoring affects more than just the lubrication schedule. And with combination PMs often tied to production schedules, the barrier to "do the right thing" is so great that it is almost impossible to adjust the "ideal world" combination PM. This can and does result in suboptimal lubrication and, for that matter, suboptimizes other maintenance tasks.
Combination PMs also require greater depth of skill and knowledge in our maintenance team. Now, instead of handing off a lube PM to a qualified lubrication technician, we require a multi-craft person with skills in many different disciplines. And while this can work, more often than not, we see suboptimal performance due to a lack of knowledge, lack of training or insufficient task details issued with the work order.
Just like lube routes, many of the issues surrounding combination PMs can be resolved with a dynamic approach to work planning, whereby pertinent task details are compiled "on the fly" vs. in a hard document. Where this is done, combination PMs can work for lubrication, although personally I do not like to see combination PMs for lubrication unless accessibility or geography make different PM plans impractical.