by Greg Case, Pump Design, Development & Diagnostics, LLC

The last reason, but certainly not the least, is that the pump was poorly selected in the first place. This can be because of poor initial specifications or just a bad choice. A classic error is the “fudge factor.” No one wants the pump to be too small, so everyone adds their +5 percent to the pump size, so when the pump is installed it is 10 percent to 15 percent larger in flow and head than it needs to be. A pump that is 10 percent oversized in flow and head requires 21 percent more energy to run than the properly sized pump.

Initial assessment

For most facilities there are far too many pump systems to do an assessment on each one. So we must reduce this to the few that offer the most promise. If the goal is reliability improvement, the selection may be as simple as a review of your biggest maintenance troublemakers.
Many pumps that chronically suffer from bearing and seal issues are often not running at their BEP. If the goal is to improve system efficiency, the following series of mental filters has proven useful in finding systems with the largest impact on reducing pump system energy costs.

Filter 1. The first filter separates the higher horsepower equipment from the lower. The theory is that the larger equipment will have a larger potential for reducing energy costs. The term higher horsepower is a bit subjective. In the training course for the Pump System Assessment Tool (PSAT) from the US Department of Energy, it is suggested that equipment which uses more than $20,000 per year in energy is a good candidate. This cost can be calculated using the following formula:


E = energy (kilowatt hours)
Q = flowrate (gpm)
H = head (feet)
T = operating time (hours/year)
sg = specific gravity (dimensionless)
5308 = units conversion constant
ηpump = pump efficiency (decimal)
ηmotor = motor efficiency (decimal)
ηdrive = drive efficiency (decimal)
C = Cost of power ($/kilowatt hour)

Filter 2. The second filter separates centrifugal from non-centrifugal pumps. While positive displacement pumps certainly have room for optimization, they are beyond the scope of this article.

Filter 3. This filter separates pumps that have a high number of operating hours per year from those that do not operate very often. Pumps that do not operate very often are less likely to use a significant amount of power. One thing to watch for with large equipment is that even occasional use might set the energy demand charge on your power bill at a much higher rate.

Filter 4. If the pump is currently controlled by a variable frequency drive (VFD), assume that there is enough control over the pump to optimize its operation individually. Only units that are currently not using VFD control will be considered in the first round of energy assessments.

Flow chart

At this point, the filters have reduced all of the pump systems in the facility, down to centrifugal pumps with high horsepower drives, significant operating hours, which do not have VFD control. This should produce a manageable number of pump systems for potential assessment.

The final filter is based on experience with the actual systems. Factors such as high maintenance costs or frequent seal and/or bearing failures can be key indicators of a poorly matched pump and system. When inspecting the system, personnel may find throttled valves as discussed previously.
Another indicator of inefficient operation is a continuously open bypass line. This is constantly pumping fluid, that you have paid to add energy to,  back to the suction side of the system, wasting all that energy. Another indicator of possible inefficient operation is cavitation noise. This can indicate that the pump is running at a flowrate much higher or lower than its designed rate.

In a system with a large range of operational flowrates and multiple parallel pumps, often the same number of pumps is always running. In times of low demand, too many pumps may be running. While none of these symptoms guarantee system efficiency problems, they help to further identify the high potential systems so they can be ranked and assessed.

Once the systems are sorted to a manageable number that can be further evaluated, a system should be selected as a test case. If this is a new part of a company's energy management process, getting a big win on the first project is important. The system should be one that is assessable without a huge expense in equipment and/or consulting. It should also be one with an expected sizable financial impact.