|Plant Maintenance practices have a large impact on the|
productivity and sustainability of industrial sites
The general role of plant maintenance includes:
- The assurance of safety – this should be the no.1 priority in asset-intensive industries, and for industries like mining, has become a major strategic issue;
- The prevention of functional failures – this is a critical issue, and requires an appreciation of the fact that an asset may be operating, yet may have failed, resulting in risks for product quality, the environment, safety, plant flexibility, throughput and other business objectives;
- To correct failures – not all failures can be prevented, or indeed should be prevented, since the cost of prevention may exceed the cost of allowing the failure to occur and then correcting it after the fact. Also, even a good preventive maintenance programme will not have a 100% success rate;
- To optimise equipment life – note that this is different to maximising equipment life, since obsolete equipment may not deliver competitive performance;
- To ensure that equipment delivers performance as per its design – this is related to prevention of failure, but also includes the prevention of injudicious plant modification;
- To achieve the above at minimum cost – cost here refers to total operating and capital costs, not simply the costs of performing maintenance, and strictly speaking should include the cost of externalities which may come about as a result of failure.
There are a plethora of approaches to the maintenance discipline, and in general, organisations do not subscribe to a single philosophy. Let’s explore individual approaches and what they mean in the context of productivity and sustainability.
Run-to-failure Maintenance (RTF)
A run-to-failure philosophy is one in which failures are allowed to occur and are then dealt with after they have been detected. The philosophy is sometimes referred to as “breakdown maintenance” but the difference between RTF and breakdown maintenance is that, as discussed above, a component may have failed without a breakdown actually occurring i.e. the failure does not stop the equipment from running.
This is essentially the base from which all preventive maintenance philosophies were ultimately developed. This is not a preventive philosophy, but a reactive one. If adopted as the dominant philosophy, it has many risks and can also be expensive due to consequential damage that may occur as a result of component failure, much of which can be far more extensive than that incurred through the original failure itself. Often, since spares cannot be held for every component, the approach leads to long periods of downtime as the site waits for delivery of the required components. Stable operating performance cannot be achieved with this philosophy.
From a sustainability perspective, there are obvious economic drawbacks and also risks to safety and the environment that arise from use of this approach, and it is seldom adopted as the primary philosophy on an industrial site. There is however room for limited application of the philosophy within the context of a wider preventive maintenance programme. This is for the case where the consequences of failure of an individual component or sub-component are very low, and the costs of prevention of the failure exceed the costs of the consequences of failure. A further criterion is that the failure must be readily detectable once it occurs. In this case the component may be allowed to fail and then repaired without incurring risks of multiple failures occurring at the same time, which may have far more serious consequences than the original failure.
Time/age-based Preventive Maintenance (PM)
Preventive maintenance involves taking action before failures occur to prevent the failures from occurring. With time-based PM, components are replaced or reworked at defined frequencies, independent of their condition. The initial philosophy was based on the age of components, and involved replacing components when they reached a fixed period in service. The approach was pioneered in arms manufacturing companies during WW2.
For this approach to be viable, components would have to display distinct wear characteristics that would allow some prediction as to when they would be expected to fail. The approach would then be to replace them as they approached this failure point, which would be determined by their operating hours.
The American airline industry, with its strong focus on safety, pioneered many of the approaches that form the basis for much of modern maintenance practice, and research into component failure conducted in the 1960’s shows that greater than 90% of components in the industry did not exhibit a predictable probability of failure with age. Maintenance tasks driven by time of use hence are not generally effective, and maintenance programmes using this philosophy will therefore not prevent functional failures. Infant mortality represents a simple example of why the philosophy is flawed. Even new components have some probability of early-life failure, and hence replacement of a working used component may in fact reduce the reliability of a piece of equipment. Maintenance-induced failures may also occur due to dis-assembly and re-assembly of complex equipment, particularly by unskilled artisans.
This approach, while potentially better than a run-to-failure philosophy, cannot guarantee improved productivity and sustainability, simply because most components do not wear in a predictable fashion. However, in replacing or reworking components, inspection of equipment is naturally carried out, and other potential problems may be detected and corrected before failure occurs. Reliability should therefore be somewhat improved over an approach comprising RTF.
Reliability Centred Maintenance (RCM)
RCM originated in the airline industry, and was a product of a study commissioned to maximise availability and control costs, with primary consideration for safety. Its roots lie in a steering committee representing airlines and aircraft manufacturers which was formed in 1960 to study the effectiveness of PM. The committee, known as the Maintenance Steering Group (MSG) developed a logic known as MSG-1, which after successive iterations and improvements developed into MSG-3 by 1978. This decision logic, applied with detailed knowledge of failure causes and consequences, is effectively RCM. The technique quickly found application in other industries, notably nuclear power plants (due to its safety focus), manufacturing plants and railroads.
RCM uses a logical, structured framework for optimising the availability and lifespan of equipment and systems. It ensures that maintenance tasks are effective and economical. The following are core principles of the methodology:
- It seeks to preserve an item’s function – hence any failure which impacts on a component’s ability to function is addressed, not just those which cause plant stoppages;
- It focuses on the entire life (not to be confused with life cycle) of the system or item, making RCM a continuous process;
- It seeks to understand the failure characteristics of the item in question, and uses this to evaluate whether PM is appropriate (i.e. should an item be allowed to fail or not?) , and if so what type of PM ;
- RCM is driven by safety and environmental considerations first, and then by economics;
- RCM acknowledges that an item has “inherent reliability” which cannot be improved through maintenance, but only through re-design.
The primary difference between RCM and time/age-based maintenance is that RCM stresses the examination of the condition of an item, with tasks that are then defined based on this condition. The philosophy also differentiates between evident functions, whose failure can readily be detected by the operator of the equipment, and hidden functions, whose failure may not be readily apparent to the operator. In the case of hidden functions, there is the potential for exposure to multiple functional failures, and this is what the maintenance effort aims to prevent. The consequences of failure determine the options to be followed in terms of preventive maintenance tasks.
The work of F.S. Nowlan and H.F. Heap is considered to be where RCM began. Their book on the subject explores the principles of failure, age-reliability patterns and the step-by-step process that needs to be followed in developing a preventive maintenance programme using RCM. Still relevant, I recommend that you read it, and it can be downloaded here.
Total Productive Maintenance (TPM)
TPM originated in Japan in the 1970’s, and while focused to a large degree on plant reliability, is in fact a broader continuous improvement philosophy and culture encouraging participation at all levels of the implementing organisation. Central to the philosophy is the performance measure used to judge its success, which is called Overall Equipment Effectiveness (OEE). The measure takes into account the so-called “six major losses”: breakdown losses, setup and adjustment losses, speed losses, idling and minor stoppage losses, quality defect and rework losses and start-up losses. It therefore addresses machine-related impacts on throughput and product quality, but importantly also addresses the relationship of the operator to the equipment being operated. Hence operator work practices are a critical aspect of TPM implementation. From a maintenance point of view, TPM also employs operator maintenance tasks (so-called autonomous maintenance – a misnomer in my view but perhaps something that I can discuss in a future post). These generally involve simple aspects of preventive maintenance such as cleaning, lubricating and tightening, which together are considered by the philosophy, if neglected, to be accountable for up to 80% of failures. Cleaning is considered a form of inspection, and hence there are linkages here to RCM, which stresses the performance of tasks on the basis of condition rather than time.
As an organisation-wide philosophy, TPM involves everyone from the shop floor through to senior management. Communication within and across levels is considered vital. TPM implementation generally involves the use of a large number of tools, such as visual management, problem solving, quick-changeovers, 5S, standardised work and others. While many of the tools of TPM are now used in implementing philosophies such as Lean manufacturing, Lean and TPM are not the same thing. The goals of TPM are to:
- Maximise overall equipment effectiveness (OEE);
- Establish a comprehensive PM system spanning the life cycle of the equipment;
- Involve all departments that plan, use and maintain equipment;
- Involve all employees from top management to front-line workers;
- Promote autonomous, small-group activities;
- Promote zero breakdowns;
- Promote zero defects.
TPM seeks to harness the power of the collective. Its success requires good change management and strong leadership, and has to be underpinned by a focus on technical skills at all levels in the organisations that implement it. It is notoriously difficult to sustain, with a large part of the problem being that it is generally considered as a “programme” with steps that have to be followed in order to be able to make the claim that “we have implemented TPM”. Of course, implementing fundamental change is just not that simple, and generally once those driving the programme reduce their involvement, the process is prone to break down. There are however significant benefits to be had from successful implementation, and while I am no change management expert, my advice would be to choose the individual tools that will add value for you and implement those rather than a full-on implementation of a bunch of approaches which may just add bureaucracy to your organisation simply because they are considered part of TPM. More important is the development of a culture of teamwork and a focus on performance.
The above are the primary maintenance philosophies that are practiced. There are further derivatives such as predictive maintenance, opportunity-based maintenance and others, but all of these are based on these core philosophies.
Which then would be the preferred approach for organisations wishing to be sustainable and productive? Clearly run-to-failure and time/age-based preventive maintenance have too many drawbacks to be considered viable. RCM represents a rigorous methodology and is in my view the best approach for determining maintenance tasks. It does not however address the need for participative engagement between plant operators and maintenance personnel, or answer questions as to who should carry out maintenance tasks. TPM attempts to do this, but lacks the rigour of RCM as regards task development. The correct approach is therefore to use a mix of RCM and TPM, with the former used for task development (including the development of autonomous maintenance tasks) and the latter used to build teamwork between plant operators and maintenance staff. The principle of assessing condition before executing tasks can be executed in practice through inspections and the use of condition monitoring technologies.
A maintenance philosophy alone cannot guarantee reliability. Successful implementation of a chosen philosophy requires a significant maintenance infrastructure encompassing comprehensive task development, planning and scheduling, skill development, spares selection and acquisition, execution, performance management and many other aspects of maintenance management. I will explore these in future posts, but all of them, and others, such as development and execution of the business processes needed to support plant maintenance, begin with the philosophy chosen. While the philosophy you choose should draw on best practice, it should ultimately be an approach adapted to your needs and resources.