The question may be asked, How much can I expect to spend on maintenance?” This question is easy to ask, less straightforward to answer, and much more difficult to understand. We all tend to agree that it is equally unhealthy business to pay excessively as it is to pay insufficiently for maintenance, but what amount is just right? We also like to rely on easy to use rules of thumb, described as “a means of estimation created specifically as a quick and easy practical guide, not based on science or exact calibrations.” This is frequently confused with benchmarking. However, benchmarking means more exact measurement and the existence of at least some data related to the underlying cause and effect relationship. To achieve excellence, we must benchmark and learn from it.
The question that automatically follows is: “What are applicable benchmarks for maintenance costs?” Most people can use the tools and data easily available to estimate where spending will be next year. However the foremost question for a business concerning maintenance disbursements is not where will I end up on my current path, but what should be my target? What level of spending would be optimum for my business? The discussion that follows describes not only how you can evaluate, using benchmark data, where you are, but more importantly where you ought to be, and also give you some indication of how you can get there.
Of paramount importance is to have a grasp of the factors that affect the optimum cost level for your business. The 3 major factors that affect your maintenance cost at any time are your asset life cycle cost strategy, planned asset utilization, and your organization’s behaviours.
1) Your Asset Life Cycle Cost Strategy. Are your assets being used up over time (wasting assets), being maintained in an optimum running state and basically repairable level (neither improving nor deteriorating and performance is at least acceptable), or are you having to revive your assets to a repairable level due to a period of neglect or lack of maintenance?
If your assets are not at a steady-state, reparable condition or do not have adequate replacement cost reserves built into the budget, for a brief amount of time your costs will either be way over or below an estimated standard value. Why a brief period of time? If you are reviving abused or neglected assets, there is a planned end in sight to the restoration or replacement. Then steady-state maintenance takes over. Rebuilding an entire wasted plant is not in the scope of this discussion, however, a brand new one certainly is. On the other hand, if you continue in under-maintaining your assets, then there is also an end in sight – the end of your business. It can be either a deliberate and planned business exit strategy or it can be an unintentional route to go out of business. It is the end of your job, and your business too, when operating as a wasting asset is done repeatedly either unintentionally or as a way to force “the numbers to look good”. Let me repeat this: Wasting your assets is a business exit strategy, whether an intentional act or not.
2) Planned asset utilization also has a little impact on disbursement, but not as vital as you would initially assume. In general, if your facility falls within a typical industry level of planned utilization of seventy-five percent or better, then the approach of calculating target costs we will discuss is valid. If your site runs in a campaign mode only six months annually or less and sits idle for the rest (as in some food processing plants and peaking power generation stations), then the results are a bit different. Likewise, if it is a “corner drug store” production operation with product lines that run periodically at fifty percent or less utilization, (such as tolling operations and some seasonal products) the costs can be slightly lower.
If a cluster of large assets runs only 1/2 the year, should the yearly cost to maintain them in a non-wasting condition be one-half? Surely not! Several maintenance costs don’t really go away just because the asset is non-operational unless it is mothballed and placed in long term storage. If you plan to take care your idle assets so that they do not deteriorate and are perpetually prepared for service on demand, the maintenance costs hardly go down at all. And if you do not maintain them while they are down, you will most certainly have to spend excess money to repair them when you try to put them back into service. There is some limited data that point to the fact that variable maintenance costs for a well-maintained idle asset are 30% of the running cost of the asset and fixed costs are not reduced. Therefore utilization inside a reasonable range has some, but not a major, impact on your optimum maintenance cost target. Poor utilization requirements can mask a wasting asset strategy for a while, but not indefinitely.
3) The overwhelmingly dominant factor in the cost of maintenance (as well as the other side of the coin, operational efficiency) is your organization’s behaviours and the way they affect the adoption of best practices concerning reliability and maintenance. Best practices are really a set of standardized, validated behaviours. For conciseness, we will simply use the term “behaviours” going forward. What type of behaviours drive cost down? Here’s a partial list of behaviours that have been documented time and time again:
A. The presence of the right culture, that is linked to good decision making. Are you reactive or proactive?
B. The presence of effective and economical PM and PdM programs.
C. Use of effective and economical processes for identifying, planning, scheduling, and executing work and the CMMS system necessary to support them. If processes are ineffective or not used, the result is worse than not having them.
D. The presence of a strong partnership between the operations and maintenance organizations.
E. Good metrics are used to drive improvement and are tied to performance management.
F. Solid and effective failure elimination and loss elimination programs are in place and functioning.
G. Good design practices based on life cycle cost, not just first cost, are standard practice. H. Good operating procedures and standard practices exist that prevent unnecessary asset damage.
When looking at maintenance costs, a vital concept that you should bear in mind is that your organization’s behaviours change the resulting cost numbers. Forcing the cost numbers won’t modify behaviours. And really it is not possible to “force” the cost numbers (or reliability numbers for that matter) to change on a sustained basis without changing either the basic assets themselves or changing your behaviours. If you change the numbers but not your behaviours it will soon result in
When looking at maintenance costs, a vital concept that you should bear in mind is that your organization’s behaviours change the resulting cost numbers. Forcing the cost numbers won’t modify behaviours. And really it is not possible to “force” the cost numbers (or reliability numbers for that matter) to change on a sustained basis without changing either the basic assets themselves or changing your behaviours. If you change the numbers but not your behaviours it will soon result in exorbitant cost, wasted assets, or unintended consequences for asset reliability levels. And inevitably the equilibrium will re-establish itself. However, it might be under the direction of a new maintenance manager who has been given the task to fix it at all costs. The takeaway concept here is that cost is the dependent variable in the equation; behaviours are the independent variables.
Next, to an even bolder statement: if your assets are held at a relatively consistent state of maintainability then the age of your plant/equipment and the type of industry you are in have virtually no impact upon how you estimate what your maintenance spending could or should be. What? The age does not matter? Not really. Old plants/equipment that have been maintained continuously, using “good” behaviours, often have lower overall maintenance costs than newer plants/equipment. Unlike people, plants do not have to wear out. Often their life span is determined by process economics, product obsolescence, and economies of scale, not wear out.
Doesn’t industry type matter? For the most part, not really. It makes only a small difference, only about ten percent or less than all of the variation introduced by behaviours. Industry type really does matter only in as much as it defines the initial cost of the equipment. Well then, how about location? Once more it matters only so much as it drives the delivered, installed, and ready-to-run cost of the plant/equipment. Of course, this is assuming that the asset group or facility has been designed for the intended service and environment. And if it has not been properly designed and installed? Well, that is a behaviour, isn’t it? Item G above, as I recall.