The case for condition-based maintenance

Keeping physical assets in proper working order is essential to any organisation. Today, most maintenance is still performed either according to a set calendar schedule or according to that most ancient of operating procedures: “if it breaks, fix it”.

But in a modern, data-intensive industrial environment we can do much better, and take plant performance to the next level with condition-based maintenance (CBM).

The perils of old-school maintenance

The drawbacks of a reactive, 'break-and-fix' maintenance policy should be obvious. Unplanned downtime kills capacity and bleeds money; emergency parts and labour incur premium costs; and major failures present substantial health, safety and environmental risks.

Clearly we ought to fix assets before they break down, and scheduled maintenance can facilitate this. The drawbacks to this strategy are less spectacular, but no less real, than those of simply waiting for things to start smoking. First, the frequency of scheduled maintenance must necessarily be based on worst-case projections of time-to-failure, but not all assets will be wearing at worst case rates. Scheduled maintenance therefore results in over-maintenance — wasting downtime, parts and people on unnecessary repairs and replacements. And without attention to actual operating asset status, even the most diligent maintenance calendar will be unable to account for abnormal conditions that fall outside its projections — and maintenance is back in reactive mode with a smoking heat exchanger on their hands.

An intelligent approach

Condition-based maintenance is maintenance based on an asset’s actual condition — a philosophy that entails performing maintenance only when there is an impending fault or failure condition. The objectives of condition-based maintenance are prevention of unplanned downtime, making optimal use of maintenance resources and maximising the operational life of plant assets. An online, integrated CBM program first takes real-time data on plant performance (such as from sensors and alarms) and compares that data to engineered or defined parameters to determine the condition of the equipment. By analysing historical real-time data, an accurate prediction of upcoming failure is possible, and maintenance schedules can be adjusted based on these predictions

The benefits of condition-based maintenance are potentially enormous. First, by making real asset health data and solid fault-or-failure predictions available, intelligent plans can be made to proactively head off unplanned shutdowns. Second, with the same intelligence, healthy assets that are running well can be left to run when they might otherwise have been unnecessarily taken offline for scheduled maintenance. A third benefit also presents itself, as the data-gathering and analysis tools that make a CBM program possible can provide the basis for many other plant optimisation efforts.

Your plant is ready for CBM

The most exciting thing about condition-based maintenance is that, in any modern facility, the key resource needed by a CBM program — namely, real-time asset data — is already present in abundance, flowing untapped through your control room. The information critical to the operation of your mill — temperature, vibration and airflow data, etc — could become critical to its maintenance. A condition-based maintenance program takes this data and unlocks its predictive power to add capacity, reliability and profitability to your operation.

Matrikon Inc
www.matrikon.com