Condition-based maintenance is a relatively modern method which does not replace parts until they are in danger of breaking down. While this implies that businesses need to invest less in spare parts, they run the risk of creating an uncoordinated and inefficient maintenance process. Together with eight process industry companies, COPE will try and find the best possible combination of condition-based and preventive maintenance.
The maintenance strategy which businesses traditionally pursue is corrective in nature: when a part has broken down, a mechanic will hurry to the scene to solve the problem. This has the disadvantage that the machine is at a standstill, which costs time and money. That is the reason why many businesses have shifted to preventive maintenance: this means that components are changed before they break down - for instance, after a predetermined number of operating hours - to keep the machinery from coming to a standstill unexpectedly. An additional advantage is that maintenance jobs can be coordinated, so that maintenance costs fall.
A major drawback of preventive maintenance, though, is that components are replaced when they are not broken at all. Condition-based maintenance provides the solution. If certain conditions, including temperature, pressure and vibrations are constantly measured, it will be increasingly easy to predict when components will break down. If businesses replace these components at exactly the right moment, the machine's standstill will be minimal and they will optimally benefit from the components' life. Just as with corrective maintenance, though, businesses run the risk of running from incident to incident, because today one component may announce the end of its life and tomorrow it's another one's turn to give up. This means that there is no longer any coordination of maintenance tasks.
Together with eight businesses and two other universities, COPE is participating in a large-scale research project which aims to find ways to combine the best aspects of preventive and condition-based maintenance. The focus of this research study, which is partly financed by Dinalog, is on the process industry - the industry in which all eight businesses are active. The process industry operates complex, expensive machinery which runs 24 hours a day, 7 days a week. If these facilities are at a standstill, it immediately affects profitability in many instances. Another problem is that it is often difficult to switch them on and off. Therefore, smart maintenance planning is crucial for this industry in particular. It can be created by introducing an integrated approach, using elements from both preventive and condition-based maintenance. The research comprises three subprojects:
COPE specifically studies clustering options for maintenance tasks. This involves complex considerations, which the research study will have to shed more light on. Imagine that a condition-based maintenance process indicates the replacement of a component next week, while preventive maintenance has been planned in three months. Will it be feasible to postpone the replacement of the component by three months, for example, by reducing the speed of the facility?
Among the eight participating businesses are three companies which operate complex and costly process facilities. They will make use of this research to reduce standstills and operational costs.
In addition, four maintenance companies are participating in the project. They are learning how to cooperate for the purpose of improving the services which they provide to their customers. They will also receive the models that have been developed for the purpose of coordinating maintenance planning. Finally, the research will provide scan and demo tools. Companies can use them to find out how they can benefit from the newly-developed concepts and models.
Suppose you have a machine which consists of two components, one of which must be replaced. Would it be wise to replace the other one too? Minou Olde Keizer tried to answer this question in her graduation research assignment into condition-based maintenance. 'We already had a reasonably good model of a system with two components connected in series. I made the model suitable for a system with two components connected in parallel.’
The model shows in what conditions it is sensible to replace the second component together with the first. 'By clustering maintenance tasks that way, companies will not only be able to minimize standstill risks but reduce maintenance costs as well. In addition to that, I conducted research into the best possible time planning for inspections. By taking into account that new parts don't require as many inspections at the start of their life cycles as at the end, maintenance costs can be reduced even more', Ms Olde Keizer, who studied Econometrics in Groningen, said.
The research was conducted at the request of planning expert ORTEC, a partner in this research project, who plans to use Ms Olde Keizer's model to calculate how much businesses can save if they adopt condition-based maintenance. As a PhD student, Ms Olde Keizer is currently conducting full-time research into this subject. 'Initially, my aim was to develop a model for a system with more than two components. I am very happy to be able to work on that, after all.'