Pump or System Problem?

,

Summer arrives, and it is time to forget about pumps. You go on holiday, but just when you think you have gotten away from the pump world, a friend asks you to look at his cabin’s water system, and so it begins.

Tom isn’t getting any water from the lake to his cabin.  He understands that a centrifugal pump is just an impeller inside an enclosure; if the impeller is rotating and you can get water in, chances are you will get water out. The problem is he feels he has water in, the pump is rotating, but nothing is coming out. I tell Tom, as I have with many others, that 95% of the time pump problems are actually system problems, so let’s look at your system first. Tom assures me that it is a broken pump but indulges me as we go to look at the system.

It is based around a small horizontal pump that is electrically driven by an integral ½ HP motor. The suction side consists of an HDPE pipe dropping about 2 ft to the lake surface, then out into the lake, with a strainer/foot valve at the very end. On the discharge side, a transducer is mounted just above the pump, and an HDPE pipe runs up about 8 ft into the single-story cabin. A small fill port on the pump outlet is used for priming, as it cannot self-prime.

While reviewing the situation, Tom tells me that if he pours water into the priming port just before opening a tap in the cabin, the system works, but an hour later he has to re-prime to obtain water flow. Based on this, I advise Tom that the first thing to check would be the foot valve, as it is probably the culprit. I explain that water is probably leaking back into the lake through the faulty valve, leaving the impeller spinning in air and unable to create any pumping pressure. In the pouring rain, I leave him to fix the valve and hope that resolves his issue.

Two days later, Tom is back with the same problem, so we take a boat out to look at the suction pipe. A good portion of the pipe is floating, but the end is weighted down. A new, shiny strainer sits on the lake bottom. Tom has changed out the old strainer. Before proceeding any further, I attach a float to lift the strainer 2 ft from the bottom of the lake and attach weights to submerge the floating line. This should avoid air locks in the suction pipe, keep boat props away from the pipe, and prevent solids from entering the system and/or jamming the foot valve.

Back on shore, we fill the pump through the priming port and go to lunch. After lunch, sure enough, the water level has dropped out of sight, so we disconnect the suction pipe, bend it upward and top it up with water. To our surprise, the level does not drop in the pipe, so Tom asks where the priming water is going and why the water in the cabin does not flow back to keep the pump full of liquid.

The pump in question was very similar to the illustration to the right and I thought Tom needed a quick lesson on the two common pump systems used to feed cabins before we went any further. Here is a quick review.

Electronic pressure transducer systems eliminate the need for a water tank. Instead, an inline pressure sensor monitors the water pressure downstream from the pump. When a tap is opened in the cabin the pressure in the system drops and the transducers electronics switch on the pump. When the tap is closed the pressure in the pipe climbs and the transducers electronics switch off the pump.

This type of system normally utilizes an integral check valve located within the transducer assembly. This check valve eliminates back pressure on the pump and allows the priming port to be opened while the cabin pipes stay charged with water.

Note: If the pump is not equipped with a self priming feature, the system relies on a good foot valve and no leaks in the system to function.

 

Pressure tank systems have a tank installed in the discharge line. Inside the tank there is a rubber bladder that can be pre-charged with pressurized compressed air (usually  around 20 psi).

When the pump starts, with all taps in the cabin closed, the pump starts to fill the tank compressing the bladder/air. As the back pressure created in the pump reaches a preset level a small diaphragm switch attached to the pump shuts the pump off. When a tap is opened in the cabin the pressure in the tank’s air bladder  forces water into the cabin. As the bladder expands, back pressure on the pump drops and the diaphragm switch attached to the pump turns on the pump and it maintains pressure until the tap is closed and the pressure tank is again charged with water.

Note:  This system must have a tight sealing foot valve or the system will bleed pressure and repeatedly cycle on and off.

After the quick pump lesson Tom understood that he had a transducer based system and that the integral check valve was why the water within the cabin was not backflowing and keeping the pump primed.

Knowing that the suction pipe was not leaking and understanding why the pump was not getting resupplied with water from the cabin left us with only one open question: where was the water in the pump going?  To answer this, I got down on my hands and knees in the pump pit with a mirror and found traces of water on the bottom of the pump casing. This meant that either the casing had a small hole in it, the back plate gasket had failed, or the mechanical seal in the pump had failed.

At this point Tom told me that a small rusty 15-year-old water pump was not worth any more trouble and he would just replace it.

In summary. When the pump was properly primed, the cabin received water. When the taps were closed for more than an hour, water drained out through the damaged pump, lowering the water level in the casing to a point where the pump could not develop enough pressure to open the check valve in the transducer.

At the end of the day we sat and had a beer. Tom couldn’t resist telling me that his pumping problem turned out to be a pump problem not a system problem and maybe my 95% rule was overstated.  I disagreed and replied “nope, Tom as always you are just part of the oddball 5% ”

 

Cheers

RJ

Defining Impeller Specific Speed

Specific speed is defined as the speed in revolutions per minute at which a geometrically similar impeller would operate if it were of such a size as to deliver one gallon per minute against one foot head.

Although that definition of specific speed may be a bit confusing, it is best to think of it as simply an index number that can predict certain pump characteristics. Pump designers use it to compare the efficiency of designs across a large range of pump models and types. Pump users reference it to know what efficiency can be expected from a particular pump design. The efficiency of pumps with the same specific speed can be compared, providing a starting point for the user or designer during the selection process. ( a benchmark for pump selection). There are four basic categories of impellers; Radial flow, Francis vane, Axial flow and Mixed flow.  Each category tends to have overlapping ranges of Impeller Specific Speeds.

Impeller categories image

 

Before we get into the range of each impeller type lets look at how Specific Speed is calculated. As stated before it is a non-dimensional design index , although the formula is the same between the metric system and the imperial system. The metric specific speed is different than the imperial specific speed and therefore care must be taken not to mix metric and imperial specific speeds. Imperial specific speed is normally represented as ”Ns” where the metric equivalent is sometimes represented as  “Nsm”. In either case care must be also taken to use the correct units. ( see below)

The general formula is:       Specific speed (Ns) formula

The “imperial units”  to use are as follows:

  • N = Pump speed in RPM
  • Q = Capacity in gpm at the best efficiency point
  • H = Total head  in feet per stage at the best efficiency point

The “Metric units”  to use are as follows:

  • N = Pump speed in RPM
  • Q = Capacity in meters cubed per second at the best efficiency point
  • H = Total head per stage in meters at the best efficiency point

The graph below shows how larger pumps with the same specific speed generally have higher efficiency ratings .I am showing this primarily to illustrate how the imperial specific speed shown on the lower horizontal axis differs from the values for the metric specific speeds in m3 & seconds shown on the upper horizontal axis.

Graphic pump size x speed

You will find that the majority of graphs and charts currently published reference specific speed are based on the imperial system. This may be as a result of the metric system to date not being standardized on the units used to calculate the specific speed . Should you need to at any time convert from imperial to metric or visa versa shown below are the various conversion factors.

  • To meters cubed/hour (m3/h) and meters: multiply Imperial (Ns) by 0.861.
  • To liters/sec (l/s) and meters: multiply Imperial Ns by 0.437.
  • To meters cubed/sec (m3/sec) and meters: multiply Imperial Ns by 0.017

Regarding the range of each type of impeller it is best defined by the illustration below.( note Ns values are in imperial units) From this illustration it is easy to see that there is no defined line between the different impeller designs but more so a blending from one category into another. At the lower end of the scale we have Radial flow impellers that develop head principally through centrifugal force and are generally low flow, high head designs. At the other end of the scale we have Axial flow impellers that develop head principally through axial forces and are generally high flow, low head designs.  An axial flow or propeller pump, with a specific speed of 10,000 or greater, generates it’s head exclusively through axial forces. In between we have Francis vane and Mixed flow impellers filling in the middle area  providing more of a compromise between the high head or high flow offered by radial or axial designs.

A simple and rough way of defining an impeller type is by comparing  outlet diameter to inlet .  As the specific speed increases, the ratio of the impeller outlet diameter, D2, to the inlet or eye diameter, D1, decreases.

comparing outlet diameter to inlet Radial vane impellers generally have ratios of over 4, while this ratio becomes 1.0 for a true axial flow impeller.

 

Before I leave you today I should add, specific speed values of less than 500 are associated with positive displacement pumps.

I hope I have provided some definition of impeller specific speed.  If you are interested in how it can be used as a tool, stay tuned it will be covered next time.

Cheers

RJ

Packing Failure and Replacement

Packings are a very simple and cost-effective way to seal a pump. Incorrect selection of material type, poor installation practices and misaligned or bent shafts can all cause premature seal failure. However, there are two additional problems that are at the root of the majority of packing failures.

1) Solids contamination is a primary source of failure for many packed boxes. Whether it’s from clear process liquids that mistakenly contain some solids, contaminated seal flush water or the simple loss of pressure in a properly designed slurry flush system, the result is the same. Solids enter the packing box, become embedded in the soft packing material and form a very effective grinder that quickly destroys the shaft sealing surface as it rotates.

typical damage on a worn shaft sleeve

As the illustration to the right shows, constant adjustment of the packing can fill the gap formed as the shaft is ground away, but premature failure is unavoidable.

 

2) Glazed or burnt (carbonized) packing material will not form an effective seal. Packing glands rely on process fluid or external flush/quench fluids for cooling. If for any reason the flow across the packing material is interrupted, the surface of the packing material quickly overheats and becomes glazed or carbonized. The glazed surface seals in the impregnated lubricants that are an integral part of the packing. Without this critical source of lubrication, sealing fail and repairs are required.

Pumps that have no flush/quench system installed can lose the flow across the packing gland due to dry running of the wet end or to a compression gland that is simply over tightened. Note: Pumps with properly engineered flush/quench systems also  lose cooling and fail due to faulty regulators or operators inadvertently shutting off the cooling water.

Packing Replacement 

Assuming that the reason or reasons for seal failure have been rectified it is now time to replace the damaged seals.  The 9 steps listed below will guide you through installation and adjustment of common packing.

1) Remove all of the used pump packing rings and the lantern ring. Thoroughly clean and inspect the stuffing box. Replace worn shaft sleeve if needed.

packing rings installation

2) Cut the required amount of packing rings, making sure to use a mandrel equal to the shaft size and  cut at 45 degrees.

3) Coat each new packing ring with an appropriate lubricant.  Liquid soap is often recommended for common packing rings.  Never use Anti-Seize or any metallic based compound on the pump packing!packing rings inside

4) Install each packing ring individually, indexing joints 90 degrees and carefully seating each packing ring in the bottom of the stuffing box. Use a split bushing or a tamping tool. NEVER use a screwdriver or a sharp tool.

(If a flush is used don’t forget the lantern ring and position it to line up with the flush port. )

5) Check to make sure the lantern ring is properly located. This is done by inserting a small object through the flush port and feeling the lantern ring.

6) After the proper amount of rings have been installed, install the gland follower and run the gland nuts up finger tight only.

7) Install any flush set-up required, and start flushing the stuffing box. It is also important to start the flush before allowing fluid to enter the inlet to the pump.

8) Start the pump and allow the packing to leak freely at start-up. During the break-in period, adjustments should be made gradually (one flat at a time), allowing 5 to 10 minutes between adjustments

ID of the packing9) After the break-in period, a typical leakage rate is to 10 to 12 drops per minute per inch of shaft sleeve diameter. (All leakage being from the ID of the packing.)

 

 

As always should you need any advise or help with a Toyo seal box please feel free to contact our very competent engineering team.

Cheers

RJ

What is a Pump’s Packing

No, it’s it has nothing to do with guns or firearms. Packing is a type of material used within a seal box in a pump to keep fluid away from the pump’s bearings and or electrics. Originally, packing was formed from materials such as leather, rope or flax. These materials have long since been replaced by specialized materials often impregnated with graphite or other lubricants to improve performance, but the principle is still the same.

Material called packing

The shaft or shaft sleeve is measured to determine its circumference, and a length of packing material is cut to match that circumference. The ends are typically cut at 45 degrees, and when wrapped around the shaft or shaft sleeve, they fully overlap at the point of union.

As the illustration below shows, it is generally installed with multiple rings, each forming an independent ring but compressed as a unit by a single gland.

Packing placement on a pump

The packing in the seal box or packing box as it is commonly referred to, forms an effective seal when the compression or packing gland is tightened. This gland effectively squeezes the packing material around the shaft to form a tight seal when the pump is not running. However this tight seal creates friction when the shaft begins to rotate during normal pump operation. If a form of lubrication and or cooling were not employed, the friction would  heat the packing material causing it to burn or carbonize in very short order. This would result in seal failure and liquid leaking into sensitive components.

The most common way to avoid friction and dissipate heat in clear liquid applications is to allow a small amount of product to leak by the seal during operation. This provides a level of lubrication but more importantly it also carries away any heat created by friction.

Packing tips on pumpsFlush water injection descriptive image

Flush water is used to preclude solids in slurry applications, or in chemical applications it is used to provide a liquid that can be acceptable to vent to the environment. It must be 100 % solids free. Any particles, no matter how few or how small will embed themselves in the packing material and form a grinder to destroy the shaft or shaft sleeve.

The rule of thumb is flush water pressure be set to pump shut off pressure, plus 10% or plus 20 psi, which ever is greater.

As flush water is always introduced into the seal box at a pressure that is higher than the pumps discharge pressure it will naturally migrate in both directions along the shaft axis. The water heading toward the impeller side serves to flush solids away from the packing material as well as providing lubrication and cooling . The water heading toward the bearing side of the box is solely for lubrication/cooling of the packing on that side of the injection port. A metal ring, often made of brass , acts a spacer between the multiple rings of packings. This lantern ring as it is called, has grooves and holes in it to provide liquid around the complete circumference of the shaft.

 

If you are interested in the causes of packing gland failure and how to replace and adjust the compression gland, I will cover that under a separate lesson soon.

Cheers

RJ

Cutless Bearing Vertical Pumps

If you have a pump application in a sump that is deeper than 7 or 8 feet, you probably cannot use a cantilever pump due to shaft deflection issues, and using a vertical column pump or a submersible may be your only option. Yes, tail pipes and limiting pump down levels can provide a workaround, but if you need to use the full depth of a sump and/or be able to start the pumps at any fluid level, cantilevers are out of the question. Similarly, if you require mechanical agitation by the pump near the sump floor, column pumps or submersibles are the only options. The question then becomes which is best, a long vertical column pump or a submersible pump?

Submersible PumpsHNS submersible slurry pump series from Hevvy Toyo Pumps

Submersibles are portable and can be easily moved from one sump to another often providing for some level of standardization.  Lead times can be short as these pumps are sometimes available as off the shelf or stocked items by suppliers. Submersibles are often installed utilizing a flexible hose and electrical plug connectors making them simple to install and disconnect as required. The biggest concern with a submersible however is the issue of submerged bearings, seals and electrical windings. A seal failure can result in a need to rewind the stator. The motor in any submersible pump is integral to the pump so when a motor winding failure occurs it must be repaired by a shop with specialized equipment and skills. This is often not available within the owners maintenance facility, necessitating off site repairs and exposing the user to extended down times.

Vertical Column Pumps

The advantages of a vertical column pump are the inverse of the weaknesses of the submersible. Having just detailed the issues with the submersible in the paragraph above, I will simply say that the motor on the vertical pump is mounted high and dry above the baseplate, and the ability to use standard NEMA framed motors is a significant advantage. Furthermore, fixed rigid piping is normally associated with vertical column pumps and, when used, eliminates messy flexible hoses, making for a cleaner installation.

One disadvantage of a column pump is its shear length. This can often make the installation or removal procedure for maintenance an issue. In some cases, headroom above the sump may be a limiting factor.  Many column pumps, like submersibles, also have seals to protect submerged bearings. As such, a seal failure doesn’t damage the motor, but it will damage the lower bearings, resulting in unwanted repairs.  There is, however, a style of vertical column pump that eliminates the need for submerged seals. This statement leads me back to the subject title for today, “Cutless  Bearing Pumps.”

Cutless Bearings

The first Cutless bearing is said to have been made by a mining engineer in the 1920s. The story goes that a pump bearing failed, and the resourceful engineer pressed a piece of steam hose inside a sleeve to get them by. It worked well, and the Cutless bearing was created. Today, it is most commonly used to stabilize propeller shafts in boats, but it still has a home in the pump industry.

Hevvy Basic Cutless bearing photo

Basic Cutless Bearing

A Cutless bearing is a water-lubricated sleeve bearing used to support the rotating shaft. In a vertical column pump, it is located at the bottom end of the column just above the pump’s wet end. It is there to radially support the shaft, minimizing deflection and any resulting vibration. These bearings typically consist of a rigid outer shell lined with a grooved rubber insert.

 

Lubrication to prevent friction between the shaft and the rubber insert while being used in clear liquid applications is generally provided by the pumpage. In slurry applications, lubrication is provided by an external liquid source, in most cases, just clean water. Whether external or provided by the pumpage, the lubrication acts to minimize wear on both the bushing and the shaft.  The internal grooves aid this by forming a thin hydrodynamic film, physically separating the shaft from the bearing surface. In addition, the grooves provide a pathway to expel any solids that may find there way into the bearing.

The Cutless bearings primary function of supporting the shaft against radial loads is augmented by its ability to dampen any deflection that may be periodically induced. The resilient rubber compound absorbs shock and dampens vibrations, contributing to quieter operation and a longer lifespan for the pump’s structural components.

Common materials used for the outer sleeve include stainless steel, brass, and composite/phenolic, while the bushing is usually nitrile rubber or polyurethane. As mentioned, grooves in the bushing help to expel contaminants. Some manufacturers shape the grooves in a spiral configuration to aid in this process. Some suppliers can also offer split-shell designs that eliminate the need to disassemble the pump’s lower end when replacing the bearing.

Axial and Radial Bearings

A Cutless pump, like any piece of rotating equipment, has both axial and radial bearings. The difference in a vertical Cutless pump is that the conventional bearings are all above the base plate and so are kept far from the liquid that would destroy oil- or grease-lubricated bearings. Back-to-back ball bearings are commonly used in the top end and provide the radial support for the top end as well as securing the shaft axially. The Cutless bearing provides the radial support required at the wet end and accommodates any axial shaft growth due to heat expansion. This ability to absorb axial growth is a key advantage over conventional submerged bearing pumps that often have to incorporate expensive features to compensate for this issue.

As a side note: Using process liquid or clean water as a lubrication medium negates the need for grease or oil lubrication inside the sump and eliminates bearing lubrication as a potential source of contamination of the pumpage. 

Repair Shops​

​As soon as the motor is separated from a cantilever, it becomes a purely mechanical device. As such, a standard maintenance department has no issues accepting it for repair. Submersibles, on the other hand, are often thought of as an electric motor and sent to the electric shop. The electric shop, however, sees it as a pump and refuses to work on it. This problem is generally resolved by sending the submersible off-site for repairs, both costly and consuming. Another point for cantilevers!​

​Cutless Pump Construction

As discussed above, a conventional bearing housing utilizes two radial bearing locations, which is not required in a Cutless pump. As such theres often just a” bearing hat” containing the top conventional bearings and the discharge piping visible above the base plate. With most Cutless pumps being direct driven, even the bearing hat is hidden by the motor stool.

The Cutless bearing is not usually deemed as suitable for applications with high radial loads. As such, impeller/casing designs that provide for balanced hydraulic loads are most commonly used for cutless pumps.

Dual volute/discharge or recessed impeller designs would be prime examples of hydraulically balanced wet ends.

The question of “what is the practical pump setting length for this design?” is a key one. In answer to that, the simplicity of cantilever pumps often makes Cutlass pumps unnecessary in settings less than 6 ft.  How long you can go is application-dependent, but lengths around 15 to 20 ft are not uncommon.

Summary

Cutless bearing pumps can provide a very attractive alternative to conventional bearing vertical pumps. Hevvy/Toyo does manufacture several product lines utilizing this technology. For additional information or help with a specific project, please feel free to contact the application department at head office or any of our sales team.

Cheers,

RJ

Taking A Look At The Specificities Of A Cantilever Agitator Pump

Our last discussion centered around the definition of a cantilever pump and how shaft deflection issues can determine maximum setting length. Today, I would like to focus on agitator pumps and leave non-agitator pumps for another day. 

Agitator wet end ​
(Drive assembly not shown)​

For those readers who are unfamiliar with the term “agitator pump”, it is simply a pump design where the main pump shaft extends past the impeller hub so an agitator can be mounted just in front of the suction inlet. The agitators is there to resuspend solids so they can be pumped along with the surrounding liquid.

Agitator pumps are generally associated with solids suspension and pumping slurry. Within that definition, cantilever agitator pumps are used anywhere a vertical pump, with a setting length of approximately. 7 ft or shorter is required. This covers a wide range of applications, but by far and away the most numerous is a collection sump. Prior to the advent of cantilever pumps, vertical column pumps or submersible pumps were the only choice for sumps. 

Submerged Bearings and Seals​

Conventional grease or oil-lubricated bearings must be isolated from water and solid contamination. This means that the bearings used to support the rotating elements in conventionalvertical or submersible pumps must have seals to protect them from contamination. Submergedseals have a finite operating life, especially in slurry applications. When seals fail users are subject to downtime and expensive repairs. Cantilever pumps require no submerged bearings and therefore require no submerged seals. This is a key feature of cantilever pumps!!!

Short Vertical Column Pumps​​

With cantilever agitator pumps being limited to settings in the area or 7 ft, it is only reasonable to compare them with vertical column pumps of similar setting length. The much smaller shafts and support structures of the column pump make them lighter. The cost savings in steel are somewhat offset by the extra cost of the seals in their wet end. However, any slight advantage either pump may have in capital cost is a non-issue, as capital cost is greatly overshadowed by the reliability and associated cost saving of the cantilever design. This, of course, stems from the cantilever design eliminating all submerged seals or bearings. As suc,h I will, for the balance of today, drop the vertical column pump as an alternative to a cantilever pump.​

​Motor Availability ​

Cantilever pumps have separate detachable motors. Generally speaking, the motor is an off-the-shelf standard “T” or “TC” frame. This allows for quick and easy re and re of the motor. It also may allow the user to increase the motor size if operating conditions call for more power. Any increase in motor size would, of course, have to be done within the original design restraints of the pump assembly. Submersibles have no such ability. ​

Variable Speed​

​Both submersible and direct drive cantilever pumps require a variable frequency drive to modify pump speed to adjust for any change in operating conditions. Cantilever pumps are often purposely built as a V-belt driven pump, knowing that speeds may easily be altered with a quick and inexpensive change of sheaves and belt. Advantage cantilever!

Repair Shops​

​As soon as the motor is separated from a cantilever, it becomes a purely mechanical device. As such, a standard maintenance department has no issues accepting it for repair. Submersibles, on the other hand, are often thought of as an electric motor and sent to the electric shop. The electric shop, however, sees it as a pump and refuses to work on it. This problem is generally resolved by sending the submersible off-site for repairs, both costly and consuming. Another point for cantilevers!​

​Repair Quality

Assuming the windings on a submersible are okay, it is possible for any maintenance shop to facilitate a repair. The personnel working on the pump, however, need to be the type of individual who takes pride and care with the repair. Mistakes as simple as a pinched o-ring will result in a complete failure when the pump is reinstalled. A cantilever pump, on the other hand, is a simple, robust piece of equipment that anyone can repair should the need arise.

Down time

Submersible pumps all have submerged seals, and often the first sign of seal failure in a submersible is the pump’s electrics tripping out as water reaches the motor windings. This failure, happening with no warning, can lead to a facility scrambling to counter rising water levels. Conversely, most cantilever failures/problems occur over time and, as such, can be scheduled for maintenance as time allows ​

​Capital Investment​

​Submersible pumps are clearly an alternative to cantilever pumps, and in small horsepower applications, the low initial cost can make them quite attractive. This advantage, howeverdisappears as the size of the unit increases. Somewhere around 10 hp, a crossover happensand a cantilever unit requires less initial investment than an equivalent submersible agitator slurry pump.​

​Adaptability

The ease with which cantilever pumps can be adapted to meet different conditions (with belt drives) allows one pump size to be used in different applications requiring different flows and heads. This standardization minimizes spare pumps and or parts inventory, and allows the pump manufacturer to batch manufacture and pass on the savings in the form of lower pump prices.

Summary

Today I think we have shown a pretty strong case for opting to a vertical cantilever pump over a submersible. That is not to say that all short setting sump applications must be cantilevers, but it is probably fair to say that cantilevers would be a good choice in many cases. Before leaving today… I am sure someone will ask, “Well what about deep sumps??” Stay tuned, next time we will answer that question.​​​​

​Cheers

RJ

What Are Vertical Cantilever Pumps?

“Vertical cantilever pumps”, as the name implies, are pumps designed to operate with the shaft running in a vertical plane with some sort of cantilever component in their design. The shaft running in a vertical plane is self-explanatory, but the term “cantilever component” may require a quick explanation.

Impellers between bearings ​

​Rotating shafts/assemblies are normally supported by two bearing points. Some pumps have an impeller mounted between the bearing points, while some impellers are mounted outboard of the bearing points.

Cantilever overhung design ​

When the impeller is mounted outboard of the bearings but in close proximity to them, the design is referred to as an“overhung” design.

If the impeller is purposely mounted with a significant overhung length (distance away from the bearing points), it is still an overhung design but is generally referred to as a cantilever design. 

So why build a vertical pump with the impeller mounted so far from the support points? Surely it increases all the loads within the pump. 

Standard overhung design ​

Well, it does, but it also distances the bearings from the pumpage and therefore allows the elimination of the seals in the wet end, which is a significant source of failure for standard overhung pumps. This is particularly true with slurry pump,s where the abrasive nature of some slurries can lead to extremely short seal life.

Loads

In the previous paragraph, I mentioned increased loads. Well, yes, the further a force is from the support the greater the mechanical advantage, and therefore the greater the forces on the entire structure. In a rotating assembly, vibration is a killer and often a result of shaft deflection, causing an imbalance or just plain balance issues. The long overhang associated with vertical cantilever pumps makes this design of pump very susceptible to shaft deflection issues, and engineers pay particular attention to this aspect of design. To quantify this, the load increases to the fourth power as the overhang increases. That means if you double the overhung distance, you increase the deflection by a factor of 16 and very quickly create a destructive imbalance.​​

Defection​

Engineers control the deflection and vibration in a cantilever shaft by adjusting the distance between the bearing points, increasing the diameter of the shaft, altering the material the shaft is made of, and, of course, specifying operational parameters.​

The selection of material is pretty easy. You generally avoid materials that stress relieve over time or have a low modulus of elasticity (bend easily). This means opting for materials like 4140 for general applications or 174ph for some more corrosive applications is usually a good choice.​

Increasing the diameter of a shaft to control defection is the simplest solution, but you can only go so far before problems arise. With deflection increasing by the fourth power, you quickly require a very thick and heavy shaft. As the diameter increases, it becomes impossible to find bearings that can support a very large shaft and sustain any decent operating RPM. In short, the peripheral speed of the bearings becomes too fast for the rating of any bearing strong enough to handle the load.

Shortening the distance between the bearings can minimize deflection between the bearings and therefore greatly alter the deflection at the impeller. However, as the simplified load sketch below illustrates, halving the distance between the bearings increases the radial load on both bearings. These same bearings may also be under duress from the heavy weight of the shaft and the high peripheral speeds discussed in the previous paragraph.​

​Last but not least, the design engineer can control deflection and vibration by restricting operational parameters so as to avoid critical speeds and any overspeed condition.​

​How Deep Can You Go??

The elimination of submerged seals and the overall simplicity of the design make a cantilever pump ideal for sump applications, but how deep can you go? Generally speaking, the deflection issue discussed above limits setting lengths (distance from the base plate down to the pump end) to around 7 ft for agitator pumps. For non-agitator pumps, a suction tail pipe can be added to increase the “pump down distance”. When adding tail pipes, engineers must also consider any structural limitations of the support assembly and the NPSA limitations of the wet end. Tail pipe applications will, of course, require the sump level at startup to be high enough to ensure a flooded wet end and thus proper pump priming.​

On that note, it seems like today’s discussion is exposing a need for more information on cantilever applications and the many advantages of this type of pump. Stay tuned, we will get to those points next time!​​

​Cheers

RJ

The Basics of Variable Frequency Drives, Part 3

In Parts 1 & 2, we discussed the basics of how VFDs work, looked at how to size them for centrifugal pumps and the pitfalls that need to be avoided. Today, I would like to discuss the advantages of VFDs.

Ease of Varying Speeds

It goes without saying that VFDs are great for varying the speed of a pump. Simply pressing a button can raise or lower a motor/pump’s RPM. Do keep in mind that selected speeds may be, or need to be, “locked out” to avoid a pump’s critical speeds. This is particularly applicable to cantilever pumps that often have low critical speeds.

Power Savings

VFDs offer huge energy savings for fans and pumps. Adjusting the operating speed of the driven unit to match the required output drastically reduces overall power requirements.

The graph below illustrates an example. The yellow line represents a 20 % reduction in flow from a centrifugal pump.  The bold red line shows the reduction in HP when a throttling valve is employed. The yellow line intersects the bold red line at 93% indicating a 7% reduction in power. However, where the yellow line intersects the curved blue VFD HP  line, a 40% reduction in power is indicated.

33% less overall power!

Valve & Pump Wear

Reducing flow by introducing a partially closed valve may be a smaller capital expense than a VFD, but it does add to wear /maintenance costs, especially when handling slurry. Obviously, the valve becomes a wear item (in some slurries, it’s an item with a very short life expectancy); however, it also affects the pump itself.

If the pump was originally sized correctly, that is to say just to the left of BEP, then reducing the flow by throttling will move the operating point away from BEP.  The new operating point will then be in an area on the pump curve that is marked by lesser ISO efficiency lines. Operating at a less efficient point increases overall wear.

On the other hand, reducing the flow by the use of a VFD  moves the operating point down while still staying close to the original ISO efficiency line. This is illustrated in the graph below.

The best efficiency curve (red line) intersects the n2 speed head/flow curve at point #2, indicating a flow rate of q2.  Reducing the flow by slowing the pump motor with a VFD moves the operating point to the lower flow rate of q1. This is done while staying on the BEP line, as marked by point #1.

The arrow marks the lower flow rate achieved when a throttling valve is used, illustrating how the operating point has moved to an area of lower efficiency.

Tip Speed

Tip speed is often used to compare anticipated wear rates.  The lower the impeller tip speed, the less wear. Reducing the flow by using a VFD reduces RPM and, therefore, tip speed.  Regardless of how you size it up, using a VFD to control flow rates is superior to a fixed speed and a throttling valve!

 

Closing

We have covered a lot of territory in this three-part review of VFDs, so before I sign off for today, I would like to leave you with a summary of what I would consider some of the important takeaways.

  • If the pump should not exceed a certain percentage of the first critical speed, has this percentage been programmed as a locked-out maximum speed?
  • Pumps may have been sized at 60Hz to run between critical speeds. Will the VFD pass the pump through those speeds quickly, and are the dangerous speeds locked out from operation?
  • Have minimum speeds been locked in when splash lubrication is required by the equipment (e.g., PD pump cranks)?
  • Over speed may be at the touch of a dial, but are the mechanics of the equipment up for the task? Especially the motor!
  • Can the standard fan on the motor shaft supply enough cooling at less than 60 Hz? A critical question when operating at altitude.
  • Are bearings insulated against current circulation?
  • Are line reactors installed on both sides of the drive?
  • VFDs generate heat and require airflow that may not be available in a dirty environment.

 

Bye for now!

RJ

The Basics of Variable Frequency Drives, Part 2

In Part 1, we covered how VFDs work and also discussed their use in controlling AC motors. Today, I would like to discuss some of the pitfalls associated with VFDs.

Sizing for a driven RPM at less than 60 Hz

There is a common mistake made when sizing a motor and VFD around a pump required to operate with a drive running at less than 60Hz. This is best illustrated by an example.

If we have a pump that requires operation at various speeds between 700 and 875 rpm and plan to use a 1750 motor, then we will need to use a setup capable of running at a maximum of 30 Hz. If we follow the pump hp curve (the red line on the right-hand graph below) up to where it intersects the 30 Hz line, we see that the pump would need approximately 100 HP. Some people would stop there and order a 100 hp motor and VFD. WRONG! You actually need to follow the green line up to where it intersects the 60Hz line. In this case, 200 HP; hence, a 200 HP motor is required, not 100 HP.

To further illustrate the point, the graph above to the right shows the required pump HP (the red curve) significantly above the available HP from a 100 HP motor (the green line) at all frequencies/pump speeds.

  Note: Even with a 200 HP motor, the drive and motor would still be undersized if the operating speed is increased to over 30 Hz.

Some would say that the cost associated with a 200 HP drive for a 100 HP pump is an expensive way to achieve variable speed, and that would be true. The solution to that issue is to use a 2 to 1 reduction belt drive between the motor and the pump. In the example above, this would allow the motor to run at 60Hz, providing the 100 hp required at a pump speed of 875rpm, while the VFD can still provide variable speeds below 875 RPM.

Heat Dissipation

VFDs are ideal for running electric motors at lower than rated RPM for sustained periods of time.  Operators of such equipment are, or should be, aware of the reduced cooling available from the motor fan at these speeds (especially with constant torque applications or at altitude). The picture to the right illustrates the result of not discussing this issue with your motor supplier prior to installation.

 

The VFD itself is also a heat source and must be cooled. This can sometimes be an issue in dirty or dusty environments where forced ventilation would be difficult.

Power spikes and harmonic distortion

Power spikes in the electrical supply can be a common source of “ nuisance tripping” of the drive’s circuit breakers. The solution to this is to install an input line reactor upstream of the controller.

Input line reactors are electromagnetic devices, with coils that protect variable frequency drives (VFDs) by limiting the rate of current rise, reducing harmonics, and absorbing power line disturbances like spikes and transients. By connecting the reactor in series between the power source and the VFD, it protects the drive from damage and nuisance tripping, improves system reliability, and helps prevent harmonics generated by the VFD from returning to the power grid.

A line reactor filters out pulsed and notched distortion, which can minimize interference with other electronic equipment like computers, PLCs, telecommunications systems, instrumentation, and other VFDs.

VFD Generated Voltage Spikes

A VFD in some applications can generate harmful voltage spikes and electrical noise that can harm a motor. In these applications, an output line reactor, also known as a load reactor, should be installed. As illustrated to the right, this is installed between a variable frequency drive (VFD) and a motor.

The need for a load reactor stems from the operating principle of the VFD. In short, a VFD generates a pulsed output known as a Pulse Width Modulated (PWM) signal to control a motor’s speed. These pulses can interact with the distributed inductance and capacitance of long motor leads, resulting in amplified peak voltages at the motor terminals. The pulses also contain sharp edges and high-frequency noise. The output reactor works by adding impedance to the circuit, which:

  • Filters electrical noise: It suppresses the high-frequency electrical noise and voltage spikes produced by the VFD’s rapid switching.
  • Softens the waveform: The reactor “tempers” or “softens” the sharp voltage pulses, creating a smoother and more reliable current waveform for the motor.
  • Reduces stress on the motor: By conditioning the waveform, the reactor reduces voltage stress on the motor’s insulation, which helps the motor and its lead cables run cooler and extends their lifespan.

 

The question then becomes, when do I need to use line reactors?  Generally, use a 5% impedance load reactor on the load side of a VFD when the motor lead length exceeds 100 feet. Typically, you can use two 5% impedance reactors in series (10% impedance) if the motor leads exceed 1,000 ft in length.

However, whenever possible, discuss the use of a VFD with your motor supplier as VFDs without output reactors can cause motors to run noisier, hotter and have a shorter bearing life.

Having scared you with all the possible problems associated with VFDs, I will leave you today with a promise of a part 3 that explains the significant advantages of  VFDs.

Stay tuned!

 

RJ

The Basics of Variable Frequency Drives, Part 1

Variable frequency drives are becoming more and more common within the pump industry. As such, it is important for pump users to understand the basics of VFDs.

A Variable Frequency Drive (VFD) is a solid-state device that inputs a fixed frequency AC current and outputs an AC current at a user-controlled frequency.

 

VFD

 

This makes a VFD  an ideal speed controller for an AC motor, as the rotational speed of the magnetic field inside an AC motor determines RPM, and the speed of the field rotation is a function of the frequency of the electrical input.

Modern VFDs also vary the voltage along with the frequency; hence, they are sometimes also called VVVF (variable voltage variable frequency) drives.

A VFD has three distinct sections, each of which performs a discrete operation. These are:

  1.  An input converter that changes AC current to DC current.
  2. A DC Bus that filters, or smoothes, the DC voltage.
  3. An output converter (referred to as the inverter) that changes DC to AC at a user-controlled frequency/voltage.

The working of each of these sections is quite complicated and as such will be a topic for another day, so today we will stay on the subject and look first at how a VFD interacts with an AC motor.

 

Current In Rush

Electric motors, when started by a simple on-off switch, draw a considerable amount of current for a short time. This can often overload the electrical supply available and cause a breaker to trip or a fuse to blow.  A VFD can limit starting current to as low as 100% of full load current( FLA), in effect eliminating inrush. To illustrate the magnitude of the problem, a 460 volt 150HP will have a full load current of approximately 180 amps. It may, however, have an inrush current of over 900 amps for over a second with a DOL start (direct on-line start). The graph below compares the inrush that can be expected with various starting methods.

A VFD can manage the magnetic flux of an induction motor. Magnetic flux is directly proportional to the voltage and inversely proportional to the frequency. By keeping the flux constant, the inrush current does not exceed the FLA rating of the motor, and full torque is maintained.

 

Motor Torque & HP Output

A motor controlled by a VFD can provide full torque from 0 to 100 % of rated speed . Anything over 100 % and the torque output falls off.

HP is a function of speed and torque; therefore, HP climbs proportionally with speed until the motor’s rated HP is reached, and then it stays constant.

 

Centrifugals and available HP

The HP required by a centrifugal pump increases with the cube of operating speed, as illustrated by the blue and red lines in the graph below:

The green line illustrates the HP output from a typical VFD.

Clearly, the VFD provides more than enough power from  0 to 100% of the rated speed.

However, at speeds over 100 % the HP available does not increase, but the HP the pump requires continues to climb exponentially. In short , an overspeed application can only succeed if the VFD is rated at a higher output than the required input of the pump.

 

Pretty sure that is enough for today. Stay tuned for next time, when we will look in some detail at the advantages and pitfalls to avoid when applying VFDs to pumps.

Cheers

RJ