Predictive Maintenance for Heat Exchangers

What makes a heat exchanger a crucial component of any industrial plant, is the fact that it can recover and recycle energy.

As all components however, heat exchangers end up wearing and malfunctioning. Damages can be caused by several factors, such as corrosion, or a slowing down of the internal flow.

These types of events are unavoidable, but the speed at which they occur depends on several factors, such as the structure of the exchanger, the type of material it's built of, the compounds it gets in contact with, or operating temperatures.

The malfunctioning of a heat exchanger won't necessarily cause a refining plant to stop production. In fact, modern plants are designed to keep production going even without heat exchangers. The problem however is, without exchangers productions gets less efficient and more expensive, and environmental impacts will take a hit too.

Maintenance is therefore crucial to minimize malfunctioning.

Two classic types of maintenance

Typically, maintenance is performed in two cases: with periodical and regular inspections (preventive maintenance), and any time an actual problem occurs (corrective maintenance).

The classic saying better safe than sorry applies here too, so the first type of maintenance will be safer than the second type. Both however, end up being quite inefficient.

In the first case because maintenance will be probably performed too early (when there is no real problem yet), and in the second case because it's done too late, after the issue came out.

A more efficient maintenance

The ideal type of maintenance is the one that offers the best of both worlds: taking place only when it's reasonably sure that the exchanger will start malfunctioning very soon.

This maintenance model is called predictive maintenance and it consists of a constant monitoring specific parameters, which for heat exchangers could be temperature, pressure or internal flow.

Such values are compared to those benchmarked as normal, and each meaningful deviation is then a signal that an intervention is required.

This type of monitoring requires specific equipment and is more complex and expensive than the classic ones (which are expensive nonetheless), although costs have been constantly decreasing over the last years.