Pumps and Systems, June 2009

Many troubleshooting guides and charts have been developed to help us understand the nature of pump failures. Although beneficial and well constructed, these handy guidelines focus on the physical cause of pump malfunctions and breakdowns.

These tools can can make a user believe that pump failure is always related to the individual physical components of the equipment. Based on this mindset, the user often changes or repairs parts to eliminate or mitigate the incident.

However, premature pump failure is not merely limited to defective parts; it can also occur from poor maintenance strategies or even inappropriate operational practices. The user can fail to consider that inappropriate human interventions-at any point in the process-can be the true causes of an undesirable event.

Design and Operation

Even during the design phase, well-intended actions can often create undesirable outcomes. Based on life cycle engineering studies, most engineers choose pumps that operate 80 to 110 percent of their Best Efficiency Point (BEP), the point on the curve where the pump is most efficient. Since facilities are constantly trying to improve processes and increase the output/throughput of operations, an engineer will often over-design a pump to meet the expected increasing demand that will be placed on the equipment.

Conversely, pumps can be chosen on the other end of their BEP to meet budgetary constraints. All points to the right or the left of the BEP obviously have a lower efficiency and induce premature failure in the equipment.

When centrifugal pumps are operated outside of their BEP (either to the right or left), then their impellers are subject to non-symmetrical forces. The induced forces manifest in many mechanically unstable conditions like high vibration levels, excessive hydraulic thrust, increased temperature, erosion and cavitation due to either starvation (inadequate net positive suction head) or recirculation (improper flow rates at the discharge). This in turn induces premature bearing and mechanical seal failures due to shaft deflection, or an increase in the temperature of the process fluid pumped, which causes cavitation damage and seizing in close tolerance parts.

Maintenance

After the pump is installed, efforts to mitigate problems or enhance the reliability of a plant's rotating equipment can also cause unintended consequences. Consider, for example, reliability centered maintenance (RCM). RCM involves the determination of the types of failure events that can produce an undesired outcome from the equipment in operation, the ways it can fail to perform its function (functional failure) and when key performance indicators (KPIs) reveal that failure has started. Preventive and predictive maintenance strategies are implemented to fix primary (first detectable state) failures before they can manifest themselves catastrophically in the form of secondary failure (pump locks up, shaft breaks, impeller damaged, etc.).

Image 1It does not matter if maintenance strategies are time-based (preventive maintenance) or condition-based (predictive maintenance) activities as long as the limitations and subsequent "built-in" failure mechanisms associated with both types are understood. Well-intended maintenance activities can sometimes be so intrusive in the equipment that they can potentially cause new, yet unseen, failures. Even a complete rebuild has its consequences, e.g., a new "infant mortality" failure rate for the equipment.

Solution

The most practical way to understand the true causes of why pumps fail is to conduct a systematic, fact-based analysis of issues associated with failure events. By doing so, the systemic risks being taken will be completely understood. This can be done proactively at the design phase or reactively at the maintenance stage. In either case, the latent issues must be uncovered to reveal the true causes of how and why the equipment is failing.

Stopping the investigation at the physical cause of failure can best be described as "Shallow Cause Analysis." By limiting ourselves to the component, we cannot produce the type of exponential paybacks required in today's business environment. A structured in-depth approach to analysis is required to reveal the latent issues causing failure.

Simply put, dig deep to find out the origins of excessive forces, determine how they got there and take all necessary steps to eliminate the undesired event and truly understand and address why pumps fail.