Centrifugal Pumps in the Process Industry How Operators of Centrifugal Pumps Avoid Trouble by Choosing the Right Pumps and the Right Configuration

Author / Editor: Hans-Jürgen Bittermann / Jörg Kempf

Selection and configuration of centrifugal pumps — Even though it doesn’t make happy reading for manufacturers of displacement pumps, centrifugal pumps are generally regarded as the workhorse of the process industry — with an estimated market share of around 80 %. And workhorses need to be properly looked after, meaning that operators can reduce trouble down the line with a more careful selection and configuration of centrifugal pumps. Here are some ideas and suggestions for this …

Related Vendors

No hassle with the pump: To prevent this from happening, the article reveals useful tips.
No hassle with the pump: To prevent this from happening, the article reveals useful tips.
(Source: ©Pixel-Shot - stock.adobe.com)

Do you remember the famous “B-pump”? In the past a redundant second pump (or “B-pump”) was often a mandatory requirement for critical processes and was key to ensuring that the responsible production engineer got a good night’s sleep. Today, we have different technical solutions at our disposal. Smart monitoring tools, which — in the ideal scenario — issue a warning long before any actual problem occurs with the pump, are already the standard in many cases or can be retrofitted at relatively low expense. But this can lull operators into a false sense of security. Monitoring is not a panacea — if the pump is designed incorrectly or poorly installed, even the best monitoring systems will only be able to offer very limited help.

Gallery with 5 images

“The pump is fine, it was just designed incorrectly or is being operated incorrectly” — for years this has also been the consensus of the speakers at the PROCESS Pump Forum (www.pumpen-forum.de). Alongside grave operating errors, most incidences of pump failure are caused by changes to the medium being pumped or by changes to the method of operation.

So what goes wrong when it comes to choosing the right pump? If the purchase decision focuses primarily on the acquisition costs of a pump, this will ignore the lifecycle costs and not take into account all investment, operating and maintenance costs. It is much more economical to design the pumps in the best possible way from the start. This requires a precise definition of the parameters such as the material being pumped and the operating conditions, as well as detailed information about the pumping task.

Exact information about the composition and properties of the product that will be pumped is absolutely essential. Missing or incorrect (outdated) basic information is often also due to the fact that a plant is constantly modified and modified again during the planning process. But at some point the equipment (pumps etc.) needs to be ordered to ensure that a plant can be commissioned on schedule. If revised planning data is not received by the pump manufacturer then operating problems are inevitable.

This is one of the reasons why over-dimensioning is still common practice — driven by the common desire to avoid problems down the line. After all, if a pump turns out to have been under-dimensioned it will be plainly obvious that the design was incorrect — making over-dimensioning the preferred alternative in this way of thinking. But this is a clear misconception, not only because of the energy wastage involved; in partial-load operation, many of the pump’s components are also subjected to increased strains and loads (among other things due to vibrations).

Frequency Converters? Well ...

Looking at the large number of publications, you could be forgiven for assuming that solely the combination of a motor with a frequency converter (FC) will ensure that the supplied energy is used in the best possible way. But of course that is not true: A frequency converter doesn’t run on thin air alone — it also “consumes” energy itself. As a result, it is important to start by clarifying whether or not a FC is actually required. Under dynamically varying load conditions the answer to this question is generally a “yes”, but depending on the specific industry there are other aspects that need to be considered (soft start, thermal transfer into the medium).

Looking at this the other way around, a pump that runs under constant load with a non-varying rotational speed at or close to its operating point is the most energy-efficient solution in all cases. Instead of connecting a FC upstream of an over-dimensioned motor, in this case it may be more cost-effective to select a more efficient, smaller drive.

However, despite all the advantages of a frequency converter, they can also present problems, particularly if they are used to virtually compensate for or neutralize wear-related drops in performance. Operators often won’t even notice this — until the wear causes failure of the pump.

Additional Information
Connectivity as a Selection Criterion for Pumps

The traditional approach when selecting a pump is to look at the demand in terms of flow rate and pump head, then calculate the expected pressure losses (the piping characteristic curve takes into account the frictional resistance of the pipe as well as pressure losses due to heat exchangers, valves/fittings and measuring points), and choose the preferred pump design (inline pump, standard pump) — at which point the configuration tool of the preferred manufacturer will offer a selection of available products. Connection to the process control engineering is performed subsequently on-site by colleagues from the world of “process automation.”

But in the era of massive digitalization in process engineering, is this still the best way to do things? Particularly the connection of the pump to the existing BUS system can, in some cases, lead to a major (i.e. expensive) programming workload, with lots of potential for associated faults if anything is not clearly defined. Would it not make more sense to start from the connectivity of the pump to the process engineering?

Integrating a pump in the shortest possible space of time into an existing BUS system and therefore into a plant concept — that is a task for system integrators who work across different sectors in industry or water resources management or building services engineering and deal with all aspects of electrical instrumentation and control systems (I&C).

Especially for this target group, Grundfos has set up a special area on its website that includes (among other things) an electrical I&C selection tool. With this tool, the electrical I&C technology supplier offers active support during the implementation of Pump 4.0 solutions in the desired BUS system. The function blocks for Siemens S7, which are available free of charge as programming examples, are particularly attractive, as they can be used to significantly reduce the engineering outlay (and time) for system integrators in I&C technology.

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy.

Unfold for details of your consent


Vibrations are a root cause of damage that is often underestimated. By design, displacement pumps operate with greater or lesser pulsations; this is a generally known characteristic and is normally remedied with pulsation dampers. It is less well known that centrifugal pumps also pulsate. This is because pressure oscillations occur when the system and the pump interact — causing pressure surges in the worst case. If a pump becomes defective without an obvious reason, these pressure surges may be the reason, as the resulting forces that are generated can be high enough to destroy bearings.

Hydraulic Faults

Hydraulic faults like dry operation or cavitation repeatedly cause serious damage to pump systems. On centrifugal pumps, it is always important to ensure that the pump is filled with the medium being pumped — dry operation is generally an absolute “no-go” for mechanical seals. If a pump starts to rattle like the sound of sand trickling onto a tin roof, the pump operator needs to be at maximum concentration: This is the sound of cavitation — and cavitation is always a sign that the pump is suffering. The main victims are — because of material removal — the impellers, the extent of the damage depends on the material properties. Stainless steel is more resistant to cavitation than bronze, and bronze is more resistant to cavitation than cast iron. In addition, cavitation also causes loud noises and vibrations, which in turn can cause damage to bearings, shaft seals and welded joints.

Apropos welded joints: In principle this is a critical point in conventional centrifugal pumps, as it is in the nature of the system that a small amount — generally not visible — of the liquid being pumped leaks at the shaft seal. Leak-tight pumps like canned motor pumps and magnetic coupling pumps offer reliability, as they do not use a shaft seal.

But which design offers greater reliability? If you need the safest and most reliable solution (critical media!) then a canned motor pump is the — albeit more expensive — technology of choice. Another (likewise older) statistic from a German refinery shows that the MTBF value (Mean Time Between Failure) for canned motor pumps is around 180 months, while that of centrifugal pumps with double-acting mechanical seals is around 50 months. Canned motor pumps thus offer an MTBF that is approximately four times higher.

Reduction in Diversity

The people who need to plan complex industrial plants (whether internally at the operator or externally at an engineering service provider) usually do not have an excellent understanding of modern pump technology. Would it not make sense to restrict ourselves to just a few designs, and to limit these to a smaller number of variants? Would it not make sense to restrict ourselves to just a few designs, and to limit these to a smaller number of variants? The standardization of impellers, materials, O-rings and mechanical seals also contributes to simplified, uncomplicated plant operation.

Additional Information
Six Tips for Operators

Material: If it is to be feared that the fluid will change over time, the more stable material should be prophylactically selected. Attention: Often failures occur, because e.g. cleaning procedures with an acid or a lye have not been observed.

Speed: Pumps in series can swell to similar speeds — a probate solution is then to choose a larger speed difference

Dry running: Install a dry run protection, which will shut down the pump in case of failure.

Cavitation: Make sure that the temperature of the liquid in the pump is not too high or the suction pressure of the pump is not too low.

Standard pump: An advantage of standardization is that the operator can conclude a framework agreement with the manufacturer: The operator then only calls up previously agreed pump types, which not only saves time but also offers an attractive purchasing advantage.

Exchange of experiences: is a must! At the PROCESS Pump Forum, this always comes first. The management of operator companies should be aware that participating employees are returning to operations with comparatively cost-effective problem solutions. You have to take some money in order to save a lot of money when planning and maintaining pumps.

Is there a tried-and-tested exclusion procedure? Or, to look at this the other way around: Is there a single pump design for all purposes? The short answer: No! But there are solutions that cover a wide range of applications. The objectives are clear: The lower the number of different series used, the fewer different spare parts need to be kept in stock. And the service technician on-site will have a better understanding of a standard pump than of a more complex variant or of a different pump type altogether. By the way, this also applies to the operating personnel (many pump failures are based on incorrect starting-up and shutting-down procedures for a pump).


It is rare that the sufficiently precise data required for correct design is already available at the time when a pump is ordered — and this is something we can probably do very little about. The planners are at the mercy of time pressures, which do not permit any delays. Since pump manufacturers are also unable to intervene arbitrarily via the lever “delivery time,” we recommend the following way out: Select a speed-controlled pump with a higher material quality. An inductive flow regulator can help to ensure that the delivery rate neither falls short of nor exceeds the permitted range. And a reliable monitoring system helps with getting a good night’s sleep!

* The author is freelancer at PROCESS.