By Allan R. Budris
When replacing an old worn pump with a new pump having a different design or head-capacity (H-Q) curve shape, the resulting performance may fall short of expectations. This is typically because of a lack of detailed knowledge of the actual pump and/or system H-Q curves, and how the system conditions interact with the actual pump performance (H-Q curve). A pump will always operate at the intersection between the actual system and pump H-Q curves, as shown in figure 1.
The following specific issues could cause the pump flow rate to fall short of expectations:
1. Pump not meeting H-Q curve:
The use of a “Non-OEM” (Non-Original-Equipment-Manufacturer) replacement pump, casing and/or impeller, could result in a lower H-Q curve. This would then cause the new pump to fall short of the system head at the required flow rate. This failure to meet OEM published performance is due to the fact that some non-OEM pump/parts suppliers either reverse engineer OEM pump parts from worn components, and/or do not appreciate the sensitivity of the impeller vane shape, or casing throat area to the pump performance. The result can be a reduced flow rate when operating on the current system. Even an OEM pump which is not individually factory tested could fall somewhat short of the Price Book H-Q curve. In order to insure that the new pump matches or exceeds the current pump performance, a factory test should be performed on the replacement pump.
2. Different H-Q curve shape:
When the replacement pump is a different model type, speed, and/or has a different specific speed (September 2009 column), the shape of the new pump H-Q curve could be either steeper or flatter than the old pump. The higher the specific speed the higher the shut-off head, and steeper the H-Q curve, while the lower the specific speed the flatter the H-Q curve, which could droop toward shut-off. If the shape of the replacement pump H-Q curve is flatter or steeper than that of the old pump, any system operation at higher or lower than the design system head point will yield a higher or lower than expected flow rate. This means that, in order to avoid an unpleasant surprise, the impact of any change in H-Q curve shape on the full system performance range should be fully evaluated ahead of a new pump purchase.
3. NPSHR too high:
If the replacement pump NPSHR (requirement based on a 3% head drop) is higher than that of the old pump (over any portion of the operational flow region), it could cause the system NPSHa (available) to be less than the new pump NPSHR. This could result in a throttling of the pump flow rate, should changes in the system curve push the pump further out on its H-Q curve, as shown in Figure 2. To avoid this cavitation caused flow loss (when ordering a replacement pump), make certain that, not only does the new pump meet the H-Q curve of the old pump, but that the replacement pump NPSHR is equal to or less than that of the old pump, over the full actual flow range. However, selecting a new pump with a lower NPSHR than the old pump could create a different problem, if it results in a “High Suction Energy” pump. Care should, therefore, be taken to insure that the “Suction Energy” of the new pump has not been moved into the “High of Very High Suction Energy” regions, in order to avoid possible cavitation damage, see October 2008 column.
4. Actual system H-Q curve not known:
The actual current system H-Q curve may be different than the original system design. Once a plant is commissioned and the plant is put in service, the system head begins to change. In the short term, levels change in the tanks and wells, valves open and close, and filter screens become clogged. As maintenance occurs, pipe schedules are changed, equipment is changed and new equipment is added into the system. In the long term, equipment loses efficiency, scale forms on the internal pipe walls and the plant undergoes expansion and contraction. Even when new, the original calculated system curve may differ from the actual system performance due to the assumptions used in the calculation, such as 10 year old pipe. Any pump change should, therefore, start with the development (confirmation) of the true current pumping system “Head-Capacity” curve, as detailed in the writer’s January 2009 Column on: “Creating an Accurate Pumping System Head-Capacity Curve...“ A field test of the pump total developed head at one or more measured flow rates can help determine the actual (current) pump and system H-Q curves. By developing the true system head-capacity curve, an accurate determination of the current and new pump operating conditions can be established.
5. Different local pump pipe fittings:
In some cases, when changing the pump speed or model type, the new pump configuration may have different types and/or sizes of attachment fittings that have higher flow losses than on the old pump. Should the new pump type be a vertical turbine configuration, the price book H-Q curve may not tell the whole story. The bowl assembly H-Q curve, which is typically shown for price book VTP performance, does not include the column pipe and discharge head losses, which must be subtracted from the bowl total developed head, before applying it to the system H-Q curve. To avoid unexpected flow losses due to new/different pump pipe fittings, the detailed performance analysis must include the impact of any new fittings.
In conclusion, if you are considering replacing a critical pump, special care must be taken when: changing the pump vendor, pump type, or pump speed, in order to avoid not meeting the expected flow rate. Since the pump will always operate at the intersection between the actual pump and system H-Q curves, it is essential to know the actual system and pump H-Q curves, taking into account any changes in local pump fittings, all of which may require field and/or factory pump tests. Finally, any differences between the old and new pump NPSHR should be evaluated for possible flow and/or Suction Energy changes.
About the Author: Allan R. Budris, P.E., is an independent consulting engineer who specializes in training, failure analysis, troubleshooting, reliability, efficiency audits and litigation support on pumps and pumping systems. With offices in Washington, NJ, he can be contacted via e-mail at email@example.com.