Pump professionals across the industry spend a significant amount of time creating and determining the correct design selection for pump applications. Making the right selection is dependent upon many factors, the type of application the component will be installed in, the media that is being run through the pump, and other end user requirements. In addition to this, there are various technical factors that are to be considered, from the law of physics to root causes and lifecycle costs, pump engineers are encouraged to understand the various ways in which pump failure can occur.
By Kristo Naude, Subject Matter Specialist & Senior Engineer.
The Laws of Physics Rule
When equipment engineers consider the possibility of something going wrong with a pump installed in a pipeline, it goes without saying that they cannot opt out of the law of physics. It is entirely up to the pump professionals to determine which laws apply and how they apply.
For example, when a pump project goes wrong and a newly constructed process plant installation had a boiler blowdown tank collecting boiling condensate, which then needed to be pumped elsewhere. From the very first day this will also likely result in oscillating piping, causing hangers to break, and check valves hammering to bits repeatedly, in a very short time.
It is a well-known principle that a chain cannot be pushed. Conversely, the same principle applies to water – it cannot be pulled along. Technically, this means that any liquid has to be pushed onto the impeller of a pump, from where the energy imparted by the impeller, pushes it along further.
For example, an installation resulted in an involuntary attempt to opt out of the laws of physics that specifically govern the world of pumps. When the stamped design documentation of the installation is reviewed, it was found that the design NPSHa was 0. Thus, an entire production unit was designed and constructed with zero NPSH available. No knowledgeable pump OEM would consider an equipment specification such as this.
Typos
Unfortunately, sometimes the project mechanical engineer responsible for pumps may make a human error, and a ‘typo’ slips in. When such typos go by as insignificant and accumulate over time, it can trigger a cascading avalanche, just as many significant events so frequently have. If this happens, the typo ultimately may involve considerable implications.
A relatively simple solution might have been to lower the pump’s first stage in a can configuration similar to condensate pumps. Only in this case, the detection of benzene at the existing pit depth made that option non-viable. This potentially meant that physically elevating the entire production unit by approximately two meters would be the only solution – only in theory of course.
An Unorthodox Reprieve
What happens when a pump gets “vapor locked” and the flow stops or starts pulsating, like in the Zero NPSH case study? An operator is typically requested to go to the pump and vent trapped air from the casing. For example, The Hydraulic Institute Standards include a pump configuration under the topic of NPSH that can help with a few other refinements in attempts to solve this self-made problem.
This particular pump is configured vertically. The top part of the pump case is vented continuously into a chamber directly above the pump volute, which is then vented back to the suction tank. Hence, the “operator” continuously vents the pump to resolve the vapor locking problem. When the typical NPSH3 curve of this pump is compared with that of other rotodynamic pumps, its Total Dynamic Head, (TDH) drop-off is not nearly as abrupt as its more conventional counterparts; it falls off gradually with diminished NPSHa and allows at least some positive delivery down to approximately NPSH3 of 1 meter- at significantly reduced TDH.
Some possible refinements at the installation included raising the tank’s normal operating water level, modifying the pump mounting to lower the pump impeller’s elevation, improving the suction piping, and enabling variable speed operation through a Variable Frequency Drive (VFD).
Root Causes and Counting the Costs
Don’t be fooled – this unorthodox solution does come at a price – the efficiency of the pump used in this example, is 38% at best. Therefore, as a last resort, the user might have to consider using this pump configuration if a typo has created complications. The global offering of this style pump is very slim. Expect maintenance costs to rise accordingly.
One might ask, how do users get into this position? The short answer would be that the application engineer was tasked with defining the process parameters and specifying the appropriate pump for the service, but failed to apply the knowledge typically acquired in Pumps 101.
Pump History, Selection and Lifecycle Cost (LCC)
The world of pumps and pump applications has seen significant developments since the advent of the Archimedes Screw. Much has been written on the subject of pump selection, and almost every application known to man has inspired and found a pump configuration (or sometimes multiple pump configurations) to serve it best.
It should be noted that well-published data establishes the capital cost of pumps between 10% and 25% of its lifecycle cost (LCC). Thus, the majority of LCC comprises maintenance, energy consumption, and production losses. Yet almost always, pump procurement is based on price and delivery, while the expected pump LCC plays virtually no part.
Wisdom Pays Now and Later
While misapplication in the case of a Zero NPSH pump, for example, is certainly extreme, it highlights the importance of the requirement to develop and apply a sound pump selection strategy. Proper pump selection retains the services of a seasoned pump professional to trust with engineering processes and specifying pumps to minimize the Cost of Ownership. Lifecycle costing of pumps also considers the minor component of the initial capital expenditure relative to the total cost of ownership and allows users to invest wisely in pumps that are specified and procured.
About The Author
Kristo Naude obtained a Bachelor of Science in mechanical engineering from the University of Pretoria in South Africa in 1986, and an MBA through the University of Stellenbosch Business School in 1998. After completing his engineering degree from the University of Pretoria, he completed two years years of military service before beginning an engineering position with Eskom, a power utility in South Africa.