Full CFD analysis can improve the operation of high-maintenance pumps.
by Richard Martinez, Standard Alloys
August 29, 2013

The facility tried to use the existing low-flow design and run with a two-pump operation to meet the system requirements. However, that solution meant that two pumps were being operated to the left of the BEP. The plant returned to the original impeller and the one-pump operation. The engineer then began to look for a service provider that could offer a true, cost-effective solution to his oversized pump and reduce the maintenance costs, which were driving his LCC through the roof for this pump.

One of the independent service providers offered a unique approach to this problem by applying the analysis method described in this article. A careful evaluation of the system curve and the overall pump design was performed. A CFD model was developed for the original impeller and the low-flow impeller. This would give critical information to properly evaluate the suction neck and discharge volute characteristics relative to any new design. Several hydraulic layouts were developed and input into the CFD model to see if an optimum design could be generated.

Proven Results

It was determined that a new impeller could be designed to operate within the existing casing that would be a much better hydraulic fit. The CFD simulation indicated that, with this new design, the new efficiency at the standard operating point would be 71 percent—almost an 11 percent increase in efficiency. Notice that the new BEP is five points lower than the original design. This reduction in maximum efficiency is partially because of the new pump specific speed and also because of the losses associated with the large throat area in the casing.

3-D model of CHS impeller—half showing the impeller and half showing the RCT core

The new impeller was ordered and manufactured in conjunction with the next pump failure. A certified performance test was not required by the engineer based on the results provided by the CFD simulation. It was agreed that the field instrumentation associated with this pump would generate sufficient data to validate the pump’s operation.

After startup, the pump capacity was as predicted. The vibration, which had a previous low reading of 0.25 inches per second, was only 0.05 inches per second. The amp load on the motor coupled with the capacity also indicated that the 71 percent predicted efficiency was obtained within the accuracy of the field equipment. The higher efficiency generated an annual energy savings of more than $15,000 for this typical API process pump. This savings alone offered an eight-month return on investment of the new impeller. This pump is now off the bad actor list and the new MTBR is expected to be at least six years.


Pump efficiency matters. It is a clear indication of the off-design operation of the pump. Modifying the pump design to improve the efficiency can often generate enough savings to justify the new design and the subsequent improved reliability of the pump. When evaluating a pump failure or the bad actor list, reliability engineers need to fully evaluate where the pump is operating. They should also challenge the pump efficiency available to that hydraulic condition and not settle for what is traditionally supplied or what was designed 10, 20, 30 or more years ago. The ability to generate accurate CFD models makes these solutions feasible, timely and practical in today’s energy conscious world.