Your questions answered: How to Read a Pump Curve

Jim Swetye, senior technical trainer at Grundfos Pumps Corp. in Ohio, tackled unanswered questions from the Jan. 24, 2017, webcast on how to read a pump curve.

By Jim Swetye, Grundfos January 30, 2017

Jim Swetye, senior technical trainer at Grundfos Pumps Corp., hosted the Jan. 24, 2017, webcast "How to Read a Pump Curve," and addressed questions afterward that were not covered during the live event on topics such as fixed system curves, variable frequency drives (VFDs), temperature and viscosity, and more. 

Question: A fixed system curve is seldom true. A system can have a variable configuration and various system curves. Should we select the pump that best fit the range of system curves?

Jim Swetye: Yes. But bear in mind that you will probably be required to be able to achieve the maximum flow that the system ever requires. Trying to achieve this while being able to achieve some potentially very low flows with a single pump can be problematic at best. This is where multiple pumps with variable or fixed speed-or possibly just one pump with variable speed-can be good options. Every application requires a complete evaluation to make the best choice.

Q: Please discuss how the system curve is affected by operating pumps in series and in parallel.

Swetye: In one sense, the system curve is theoretical and is not changed in any way by changes in pumps or their configurations. It will always follow the same arc-shaped progression toward infinity. However, changing anything to do with the pumps’ hydraulic capabilities, including changing from a single pump to multiple pump series or parallel operation will have a definite impact on flow and head within the piping system. Many good courses are available on series and parallel pumping.

Q: Can you estimate the gallons/minute (gpm) just by seeing the pressure drop across the pump?

Swetye: Yes. Assume we are working with clean cool water. Take a pressure gauge reading in psi on the suction side, and then a separate reading on the discharge side. Subtract suction pressure from discharge pressure, and multiply the result by 2.31. The result is feet of head that the pump is working against. Refer to the pump curve for that pump at that speed with that impeller trim. Find that head on the left. Read to the right until you reach the intersection of the pump curve. Read down until you find the flow rate at that head.

Q: Do you need a variable frequency drive (VFD) to obtain variable speeds of a pump?

Swetye: Using a VFD probably is the most common method. However, there are other adjustable speed devices used in conjunction with pumps.

Q: How do temperature and viscosity change the curve?

Swetye: Virtually all published curves are based on clean, cool water. If you are pumping anything else, actual performance of the pump likely will change. A few degrees of change-for example, from 72°F to 78°F-of clean water will have no discernable effect. But going to 210°F can have a huge impact, especially on net positive suction head required (NPSHr). And going from water to oil with a viscosity of 1,200 Saybolt seconds universal (SSU) (SSU is a unit of kinematic viscosity), for example, there will be a radical impact on many pump performance characteristics. Anytime you are pumping anything other than clean, cool water, seek professional assistance.

Q: How is the runout flow determined? Can it be calculated?

Swetye: Runout flow cannot be calculated. It is determined by testing in a laboratory, and then displayed on the pump curve.

Q: I used to work at a utility that did not want engineers to select trimmed impellers (with the knowledge that a trimmed impeller is not as efficient) so that pumps of the same model from the same manufacturer were interchangeable. Is this a good practice?

Swetye: It certainly helps to avoid the risk of installing a replacement impeller with the incorrect diameter in a pump. This also would help to hold down inventory investment if spares are stocked. A downside is that you could be giving up some pump efficiency and thus increasing power consumption. So, this is a judgement call. At a utility, where plentiful power is presumably available, it might a wise decision.

Q: Is there a rule-of-thumb to determine the preferred operating region if the manufacturer doesn’t provide it? What’s a good metric of how far you can go from the best efficiency point (BEP)? Is it a percentage range around the BEP?

Swetye: I suggest you read the Hydraulic Institute Standard on allowable operating region. It discusses the preferred operating region, which can vary by pump type. Generally, if you stay within 20% of the BEP-which would be between 80 and 120 gpm for a pump with a BEP at 100 gpm-with most centrifugal pumps, you will be in an acceptable range. However, there are exceptions. So, always check with the pump manufacturer if it is not displayed on the curve.

Q: Is there a webcast that discusses the effect of VFDs on pump flow/head?

Swetye: Yes. Check the prerecorded Grundfos Technical Institute training website at

Q: How quickly can/does fluid temperature rise in a casing when the pump is operated under a shutoff condition?

Swetye: It varies with the size of the pump. I have seen some very small circulators that have run near shutoff for extended periods, but it is not recommended. Always consult the circulator manufacturer. And by the way, I heard the story of a large split-case pump that operated not only against a closed valve at the discharge side, but also a closed valve on the suction side. It built up so much heat that could not be dissipated fast enough so that the pump literally exploded. Shrapnel from the volute destroyed the pump house, and would have killed anyone had they been near. Respect the shutoff head warnings.

Q: Is NPSHr expressed in feet of water at a specific gravity (SG) of 1?

Swetye: Yes. And at the temperature stated by the pump manufacturer.

Q: Is the power based on water as the pumped fluid at an SG of 1?

Swetye: That is correct. And at a temperature stated by the manufacturer.

Q: What is MCSF? Why do some curves have it and others don’t?

Swetye: MCSF is the "minimum continuous stable flow," which is an area to the left of the curve down to where the pump can safely be operated, although it may or may not do so with reasonable hydraulic and mechanical efficiency. You should not operate to the left of the MCSF for any but a short time or damage to the pump may be imminent. Whether (and how) the MCSF is displayed on the curve or not is based on some complex-decision making by the pump manufacturer. If you suspect you may operate back in that area and do not know the MCSF, always consult the manufacturer before proceeding.

Q: Please define efficiency. Is it a percentage of theoretical calculated percentage?

Swetye: The efficiency displayed on the curve normally is the hydraulic efficiency of the pump. It is based on data collection in a laboratory. In the lab, based on calibrated equipment, flow rates are metered, pressure/head is determined, and power consumption is recorded. A series of calculations are then made, and the various efficiency points are established and displayed on the pump curve. So, in this sense, they are based on empirical data, and are not theoretical. For the best explanation, see the Hydraulic Institute’s standard on pump testing.

Q: We used to specify the motor horsepower just above the horsepower required for the system flow and pressure loss. We found that often, people would open control valves. Then, the backpressure drops, the flow increases, and the motor overloads. We now specify the motor horsepower at maximum flow for the impellor diameter. Is that a good idea? Is there a better way?

Swetye: The conservative way, which is promoted by many pump professionals, is to specify the motor horsepower nameplate rating to be non-overloaded at the maximum point on the pump brake horsepower (BHP) curve. There are some applications when it is certainly legitimate to do otherwise, but should be done only after full investigation.

Q: What effects can be noticed when the pump runs to the right of the runout curve?

Swetye: Noise and vibration will increase significantly. With a really big pump, you will feel that you should not stand near it if the problem is severe enough.

Q: What is the optimal efficiency range?

Swetye: The absolute optimal range is a single point known as the BEP. However, the preferred operating region is generally accepted as a very good range.

Q: Where is rpm on the curve? Where is speed identified on the curve?

Swetye: Usually, speed is found in one of the corners, but sometimes it does not appear on the curve. In that case, you will need to consult an associated data table.

Q: What software are you referring to for system curve modeling?

Swetye: Virtually all pump manufacturers use software with similar functionality. The Grundfos Express platform was used for today’s webcast.