Some Misconceptions about the Selection and Application of Centrifugal Pumps

Purchasing a high-quality centrifugal pump that meets manufacturing standards is merely the starting point, not the end goal, for ensuring long-term trouble-free system operation. If the equipment is not designed to be compatible with specific application scenarios, even the most impressive performance parameters will inevitably lead to problems adapting to the actual environment. Each centrifugal pump has its preset optimal operating point and permissible operating range. Once the actual operating conditions deviate from the design boundaries-whether in terms of flow rate, net positive suction head (NPSH), media characteristics, or start-up and shutdown methods-it will directly threaten the pump unit's hydraulic stability, mechanical reliability, and service life.

This article will, based on engineering practice experience, explain and clarify some common misconceptions in centrifugal pump selection and application to avoid users making poor product selection and application decisions.

 

 

• Some common misconceptions in product selection

 

The higher the efficiency, the better

"Efficiency" is the core indicator for measuring the hydraulic performance of a pump and is also the most direct starting point for users to focus on energy consumption costs. However, if the goal of maximizing pump efficiency is pursued unilaterally, multiple compromises often need to be made in the design phase, which may affect the robustness of the equipment and the reliability of its long-term operation.

High efficiency, which naturally translates to lower operating energy consumption, is both a common user expectation and a primary driving force behind continuous innovation for pump manufacturers. It's considered a "goal" because in critical operating conditions requiring long-term continuous operation, such as nuclear power, thermal power, and petrochemical plants, the safety and stability (i.e., reliability) of pump units are always paramount.

Achieving high-efficiency design typically involves several technical features: smaller operating clearances (wear rings, balancing mechanisms, etc.), longer and more flexible pump shafts, stringent tolerance requirements, and the smoothest possible flow path profile. While these designs improve hydraulic efficiency to some extent, they also make the equipment more sensitive to media cleanliness, fluctuations in operating conditions, and installation alignment, thus increasing maintenance frequency, raising technical barriers, and raising spare parts costs for users.

Manufacturers, based on different market positioning and product strategies, have different focuses in their pump design philosophies: some prioritize energy efficiency, some regard reliability as a core principle, and others pursue ultra-long service life under extreme operating conditions. In the 1980s, China's thermal power generation market developed rapidly, attracting numerous internationally renowned power plant pump suppliers. Most manufacturers used safety and reliability as technical barriers, but one foreign company entered the market with "the highest efficiency in the industry" as a differentiating advantage. Although its initial measured efficiency was indeed 1% to 2% higher than its competitors, it frequently experienced shaft breakage accidents after commissioning and was eventually driven out of the Chinese power market. This case profoundly illustrates that centrifugal pumps are not necessarily better the higher the efficiency; a design strategy that sacrifices reliability for efficiency is tantamount to drinking poison to quench thirst in critical operating conditions.

 

The lower the required net positive suction head (NPSH) of the pump, the better.

For users, the lower the required net positive suction head (NPSHR) of the pump, the better, as this can significantly reduce the height of the installation (i.e., the net positive suction head NPSHA) and effectively reduce investment costs.

In most pump systems, NPSHA tends to decrease with increasing flow rate, while NPSHR tends to increase with increasing flow rate. Therefore, before system design, the pump manufacturer's recommendations and application experience should be considered, and a sufficient safety margin should be provided within all expected operating flow rates. At the same time, when determining the net positive suction head, the buyer and seller should clarify the relationship between the minimum continuous stable flow rate and the pump's suction specific speed. Generally, the minimum continuous stable flow rate of the pump increases with increasing suction specific speed… When selecting the suction specific speed and the NPSH safety margin, existing industrial standards and manufacturer experience should be considered.

At the same speed and flow rate, the lower the NPSHR, the higher the pump's suction specific speed. Compared to pump designs with lower suction specific speeds, pumps with higher suction specific speeds are more likely to encounter undesirable vibrations and noise, and the permissible operating range also becomes narrower. Regarding the impact of suction specific speed on the operational reliability of centrifugal pumps, international peers have extensive engineering application experience and have provided maximum limits for suction specific speed, which can be used as a reference when selecting pumps. Among them, the limit for suction specific speed specified in UOP 5-11-7 [2] has been widely recognized and applied globally, and its provisions are as follows: the suction specific speed of the pump shall not exceed 13000 (m3/h, rpm m); when the pumping medium is water or a solution with a water content exceeding 50%, and the single-stage impeller power of the pump exceeds 75 kW, the suction specific speed shall not exceed 11000 (m3/h, rpm, m). With the development of science and technology, today, without increasing the impeller inlet diameter, there are many ways to improve the suction performance of centrifugal pumps, and the limit for suction specific speed has also been increased accordingly. After studying the products of many multinational companies (such as EBARA, KSB, ITT, etc.), the author found that the BB2 type, designed using modern design methods (rather than the traditional method of increasing the impeller inlet diameter), can achieve a suction specific speed limit of approximately 14,400 (m3/h, rpm, m).

Therefore, users should not blindly pursue "low NPSH3" without considering actual operating conditions. Instead, they should work with manufacturers to leave a sufficient and reasonable margin between the unit's net positive suction head (NPSHA) and the pump's required net positive suction head (NPSH3), while carefully assessing the operational stability risks that may be caused by high suction specific speed design.

 

The further the pump's critical speed is from its actual speed, the better.

Critical speed refers to the characteristic speed at which a rotor system resonates, and its value depends on rotor stiffness, support stiffness, and mass distribution. In actual engineering bidding processes, due to an excessive avoidance of resonance risks, buyers tend to require the first-order transverse critical speed to be as far away as possible from the pump's rated speed (for example, early tender documents for conventional island feedwater pumps in nuclear power plants required that "the first critical speed of the pump shaft system in water should be higher than 125% of the speed corresponding to its rated operating point"). A few years later, this percentage was increased to 135%, and in the most recent tender, it was even required to reach 150%, while still meeting current energy efficiency standards.

 

From a dynamics perspective, raising the critical speed to a point far from the operating speed range is entirely feasible – simply by increasing the shaft diameter, shortening the span, or increasing the support stiffness. However, this inevitably leads to increased rotor weight, increased balancing difficulty, and a direct conflict with energy efficiency targets: under the same conditions, the higher the critical speed, the thicker the shaft system, and the greater the mechanical friction loss and disc friction loss, meaning lower pump efficiency.

 

Therefore, the selection of the critical speed is essentially a trade-off between "dynamic safety" and "hydraulic economy." The ratio between the critical speed and the pump's rated speed should be rationally determined based on different pump types and operating conditions.

 

Centrifugal pumps can handle gas-liquid two-phase flow.

The dissolved gas operation of centrifugal pumps involves a highly complex gas-liquid two-phase flow, which has been extensively studied by researchers both domestically and internationally. Experience has shown that with special impeller designs (e.g., a semi-open impeller back cover with a return hole near the impeller inlet on the impeller flow channel), centrifugal pumps can maintain stable operation even with a gas content of 10% (by volume). Ordinary centrifugal pumps can handle liquids containing small amounts of gas (1% to 2% by volume). The presence of a small amount of gas in the liquid can buffer the impact of cavitation bubble collapse and reduce associated noise, vibration, and erosion damage. However, when the gas content reaches 6%, ordinary centrifugal pumps may experience cavitation and gas lock phenomena, leading to a sharp decline in performance (flow rate, head, and efficiency).

 

• Some misconceptions in application

 

Allows for long-term operation below the minimum continuous stable flow rate

The ANSI/API 610 standard defines the minimum continuous steady-state flow rate (MCSF) as the minimum flow rate at which a pump can operate continuously without exceeding the vibration limits specified in the standard. In other words, if a pump operates below this flow rate for an extended period, it will immediately trigger a series of adverse physical processes, including internal backflow, cavitation, additional radial forces (especially seen in single-volute pumps), and increased liquid temperature. These can then induce a chain of failure modes, such as shaft deflection, wear ring wear, mechanical seal end face separation, and abnormal bearing overheating. The consequences are a significant increase in mechanical vibration and noise, which can drastically shorten the service life of mechanical seals and bearings, or even lead to shaft breakage and severe damage to the rotor and mating components.

 

In actual operation, two common operational misconceptions arise: First, some users, due to a lack of understanding of the operation and maintenance manual during commissioning, mistakenly believe that centrifugal pumps can operate for extended periods under any conditions (including below the minimum continuous stable flow rate or even zero flow). Second, a few users, due to a misunderstanding of the warranty terms, attempt to artificially create extreme operating conditions during the commissioning phase to verify the equipment's "endurance." Such operations constitute destructive testing for centrifugal pumps, potentially causing not only immediate malfunctions but also the development of hidden defects such as fatigue cracks, and must be strictly prohibited.

 

Starting the pump with the outlet valve fully open means starting the pump when the outlet valve is fully open.

In many systems requiring long-term continuous operation (such as power plant systems), standby pumps are typically installed and are always ready to start (the pump is filled with medium; auxiliary systems such as seals, cooling water, and lubricating oil are operational). If the main pump fails, the standby pump will automatically start. Therefore, the pump outlet valve must be open, commonly known as "open valve start."

 

Many buyers or users have a misconception about open valve start: they believe it means starting with the outlet valve fully open. This is incorrect! "Open valve start" actually means starting with the outlet valve at a specified flow rate (such as the rated flow rate). Typically, the flow rate that the pump outlet valve in a system can handle is much greater than the pump's rated flow rate. If the outlet valve is fully open, the pump will operate at an excessively high flow rate, potentially causing a series of problems such as cavitation, vibration, motor overload, or failure to start the unit.

 

The correct procedure is as follows: Based on the pump performance curve and the system resistance curve, preset the outlet valve opening to a position near the corresponding working flow rate, and then fine-tune it based on the readings of the pressure gauge and ammeter after the pump starts.

 

The packing seal can be directly replaced with a mechanical seal.

Taking the OH1/OH2 type centrifugal pump as an example, the differences between packing seals and mechanical seals lie not only in their sealing type but also in the fundamental differences in the overall structural layout. For packing seals, in order to provide sufficient axial space to accommodate the packing, packing rings, packing gland, and an area for easy manual operation, pump manufacturers have to place the impeller far from the bearing, resulting in a significant increase in the impeller cantilever span L. When the pump starts or the shaft deflects due to the maximum radial force, the packing, due to its certain radial support stiffness, objectively acts as an auxiliary bearing, becoming part of the bearing system and providing support.

 

Simply removing the packing seal and replacing it with a mechanical seal without re-verifying the rotor dynamics will lead to the following risks:

 

Lack of Support: Mechanical seals do not possess the function of packing seals as "part of the shaft support system." Once the damping and stiffness provided by the packing disappear, when the pump starts or the shaft deflects, the shaft end deflection may exceed the mechanical seal's following capability, inevitably leading to seal face separation, leakage, and uneven wear. This reduces component lifespan and may even cause damage or equipment failure.

Spatial Interference: The inner diameter of the stuffing box (sealing cavity) in a packing seal is typically smaller than the cavity size required by the mechanical seal. This narrow space is insufficient for the mechanical seal to use centrifugal force to throw solid particles off the sealing surface, and it also cannot provide sufficient cooling clearance, leading to overheating, coking, or premature failure of the sealing surface.

 

Therefore, for end-suction centrifugal pumps, when considering switching to a mechanical seal, a rotor dynamics assessment must first be performed to calculate the ratio L³/D⁴. Here, L is the distance (inches or millimeters) between the center of the inner bearing and the center of the impeller; D is the shaft diameter (inches or millimeters) at the mechanical seal bushing. This value should be less than 60 (US units), or less than 2.0 in metric units, to be considered within the acceptable range for the shaft end deflection of the mechanical seal.

 

Allow reversal

In many critical centrifugal pump applications, users typically require that the pump can withstand short-term reverse rotation when conditions such as outlet valve failure or check valve malfunction occur, and that the permissible reverse speed reach the rated speed. From the perspective of rotor dynamics torque margin, this is theoretically feasible – provided the shaft diameter, key connection, and lock nut are checked against the peak reverse torque. However, in practical engineering applications, due to constraints from multiple factors such as cost, energy efficiency, safety and reliability, mechanical seal systems, and operating conditions, it is often difficult to simultaneously achieve optimal forward performance while providing the equipment with strong reverse tolerance.

 

Furthermore, "reversal," as a typical fault/accident condition, is often not an isolated event, but is frequently accompanied by a superposition of extreme situations:

"Power Slip" Reverse Reversal Self-Start: When the equipment loses power, the outlet valve malfunctions simultaneously, and the pump is in high-speed reverse rotation, if the power supply is suddenly restored within a short period of time (i.e., the motor accelerates again), it is equivalent to suddenly starting the pump unit during the reverse rotation process. At this time, the pump rotor will be subjected to the vector superposition of positive torque and reverse inertial torque, and will instantly face an impact load of nearly twice the rated torque, which can easily cause shaft breakage or tearing of the elastic element of the coupling.

Mechanical seal flushing failure: For mechanical seal structures equipped with pumping rings (vortex flow), the pumping direction is strictly unidirectional. Under reverse operation, the pumping ring may fail to establish flushing flow, or even generate reverse pressure differential, leading to interruption of the mechanical seal flushing fluid. This, in turn, causes dry friction and thermal cracking on the sealing surface, significantly shortening the mechanical seal's lifespan. Therefore, once the pump reverses, the mechanical seal must be disassembled and inspected after shutdown, and the dynamic and stationary sealing rings should be replaced if necessary.

Failure of anti-loosening mechanism: If the fasteners on the rotor (such as impeller nuts and bearing lock nuts) do not have a reliable anti-loosening design (e.g., relying solely on frictional torque without a mechanical locking structure), the instantaneous impact torque generated by reverse rotation may cause parts to loosen, shift, or even be damaged.

 

• Summarize

Based on the above risks, a tiered control strategy is recommended: For non-critical operating conditions, or for critical scenarios where reverse rotation can be avoided through process interlocking, it is not recommended to force pumps to have reverse rotation tolerance; for critical pumps that must tolerate reverse rotation, the reverse torque index should be clearly specified in the design phase, and the manufacturer should conduct special shaft strength verification and fastener anti-loosening certification to ensure safety margin.

 

The rationality of centrifugal pump selection and application directly determines the life-cycle cost and operational reliability of the pump system. Based on the foregoing analysis, the core conclusions are summarized as follows:

• Higher efficiency is not always better – High-efficiency designs often come at the cost of sacrificing wear ring clearance, shaft diameter margin, and flow channel smoothness. Reliability should be prioritized in critical operating conditions.
• Lower required NPSH is not always better – Low NPSH3 is usually accompanied by high suction specific speed, leading to a narrower operating range and increased vibration risk. Pump selection should match the NPSH of the system, and suction specific speed should be limited according to standards such as UOP and HI.
• Critical speed should be set reasonably, not blindly increased – Increasing the critical speed is negatively correlated with energy efficiency. The critical speed margin for wet rotors should be reasonably selected based on the pump type and standards (such as API 610) to avoid over-design.
• Ordinary centrifugal pumps have extremely low tolerance to gaseous media – Only specially designed pumps can handle 10% gas content. Conventional pumps experience performance degradation above 2%, and may experience airlock above 6%.
• Prolonged operation below the minimum continuous stable flow rate is strictly prohibited – minimum flow protection measures (such as bypass or automatic recirculation valves) must be configured.
• Open valve start-up ≠ full outlet valve start-up – the preset opening degree should match the rated flow rate to avoid overload and cavitation. ● Packing seals cannot be directly replaced with mechanical seals – rotor deflection calculations (L³/D⁴) must be performed, and the dynamic impact of changes to the support structure must be assessed.
• Reverse rotation should be handled with caution – it is not required unless absolutely necessary; if necessary, specific strength checks and anti-loosening designs should be performed.
• Industrial purified water is not suitable as a cooling medium for high-temperature pumps – softened water or steam is recommended, and water quality management and maintenance should be strengthened.

 

Only by abandoning the isolated selection mindset of "parameter supremacy" and establishing a scientific decision-making framework of "system adaptation, operating condition matching, and full life cycle trade-offs" can the above-mentioned misconceptions be fundamentally avoided, and efficient, stable, and long-life operation of centrifugal pump systems be achieved.