Do you need help with a pump problem and want to know how other pump users would solve it? Post your problem on the Pump Chat forum of our website,, so your industry colleagues can share their insights with you. Here are some recent question-and-answer excerpts from our Pump Chat message board. 

Pump Flanges

Q. We have an ANSI pump and the back of the flanges on the suction and discharge are not flat (or parallel with the front of the flange.) This is causing our studs to bend when we torque down on them. At first I thought we had pipe stress, but actually the back of the flange is angled, causing our bolts to be angled.

The supplier salesman says that it is a drafting angle and as long as it is within 5-deg everything is okay. I don't think that is correct and, although this particular application is not very challenging, I would like my company to keep a design like this out of our plant. It doesn't look right and I question the safety of bending studs.

Is there a common pipe specification that details ductile iron flanges and flatness on the back of the flange? If so, what is it?

A. You could have the flange faces “back faced” or “spot faced.” Back facing involves machining the entire back side of the flange, and spot facing involves machining the area where a nut would contact the back of the flange.

I'd be pretty sure all manufacturers would offer this as an option, i.e. you'll have to pay for it and you'll have to wait for it, since it’s not a standard. You can also have the flanges themselves tapped with threads so that you won’t need to use a nut. You've probably seen this on very small pumps with 1-in and 1.5-in flanges, since these small sizes just don’t leave a lot of room to work a nut onto the back of a bolt.

Energy Losses Related to Misalignment

Q. Does anyone have any information related to misalignment of shafts and energy losses? Perhaps a typical number like ‘there’s a 2 percent increase in energy use if the shaft is misaligned’ or something along those lines. I'm trying to do a simple study to show it is worth the manpower to realign the shafts on some pumps.

A. Don’t try to relate misalignment to power lost by reduced efficiency. If you could manage to do this, it won’t amount to anything in relation to the cost incurred due to the number of times the misaligned pumps will go down for repairs, the increased maintenance time putting it back together, and the process shut down costs. Misalignment causes reduced Mean Time Between Failures (MTBF).

Check the number of times your misaligned pumps went down vs. the pumps that are correctly aligned. If you have a sufficient number of data points, you CAN prove it. If you don't have the data . . . you’ll just have to wait for it.

The MTBF database numbers for pumps are based upon correct alignment. Get this figure for the pump type you have, then reduce it by 25 to 50 percent, depending on how badly misaligned they are. Then see how your calculation looks.

Bearing Housing Oil Replacement Period

Q. I am working at a refinery. I would like to hear your experience for the period of replacing the bearing housing oil of centrifugal pumps (Oil Sump).

A. Several responses, including:

1. Most pump maintenance manuals will recommend a change interval. Actual operating hours between changes are best determined by Oil Analysis, which also warns of pending bearing failures, as does Vibration Analysis. Using a high quality synthetic bearing oil typically doubles or triples the change interval while providing better protection for bearings. Contamination is the key – if the bearing assembly is exposed to dirt or water, change the oil. Otherwise, pull a sample for analysis every month or two and change the oil based upon the Viscosity Index and actual contamination.

2. Most pump companies have recommended hours for bearing housing lubrication. I would suggest finding the instruction manual for your unit and follow their recommendations.

3. This depends on many factors: temperature, moisture in the air, contaminants in the air, operating hours per year, etc. Your best bet is to continue operating as you are and get an oil analysis – check to see if it is discolored or if it has big chunks of metal in it. Then go from there.

4. It depends on your application. However, in general, centrifugal pump oil can be replaced once or twice a year.

5. I also work at a refinery. We have found that if the oil doesn’t have any extra contaminants in it, we can change it every six months for the best efficiency out of our pumps. Right now our time between repairs in our area is 8 years.

Pump Repair Cost

Q. Does anyone know of an industry study that has been done on the total cost of repair of an ANSI pump? I have been told that they range between $3,500 and $5,200. This would include not only the mechanics’ time and parts, but also the paperwork as well as the decontamination of shipped parts, repair request, lock-out, transfer to sister pump, ordering parts, etc.

A. Several responses, including:

1. Average cost = $3700.

2. I own a shop that repairs ANSI and Process Centrifugal pumps for Pulp & Paper, Forest Products, Mining and Power Generation. From my experience, the range is anywhere from $950 to $5,500, depending upon what parts are replaced and what upgrades are performed to increase the MTBF.

3. I’d suggest you assume something using a percentage of the cost of the “average” pump, such as 5 to 10 percent or something similar for routine repairs; 15 to 20 percent for major repairs, etc.

4. A database of “average” pump repair costs wouldn’t really be worth anything because of the differences between pumps and how those differences can significantly affect repair costs. It simply doesn’t make sense to include such widely varying items.

For example, a seal replacement for a 5000-hp pump at $50,000 dollars might be grouped with a seal replacement for a 50-hp pump, which might cost $1500. The average of this is $25,750. So would you then assume the next pump needing repair would cost $25,750? That would look pretty silly if your customer said he only paid $1,450 for the same work three years ago.

In other words, your database would be composed of apples and oranges, unless you carefully itemized each pump according to its specific attributes. There is no use having a database that includes apples, oranges and pears, because apples, oranges, and pears all have different repair costs. Unless you tab the species of each one – which by definition would make it NOT a database of all average units – you defeat your purpose. 


Pump Vortex at Suction

Q.  I was looking to see if anyone has a method to calculate the level at which a vessel will vortex. We have been having a problem with pump seals failing due to dry running, and the only thing we can come up with is vortexing in the vessel.

We want to know if someone has a program that will calculate the level at which the vessel will vortex depending upon the pump size. Can you help?

A.  Vortexing is caused by fluid rotation. Have you considered installing a vortex breaker to reduce the rotation of the fluid at the discharge nozzle?

If it’s a relatively straightforward transfer pump, maybe you can install a VFD and reduce the motor speed as the liquid falls below some volume level (i.e. such as 10 percent full). Finally, install a power monitor to shut the pump down when the power level drops significantly (as happens when the dry run condition begins to develop).

Some of these ideas may give you some form of the calculation tool you seek (I don’t think a specific one exists), but at least they will address the problem. I sincerely hope this helps!


Pumps & Systems, December 2006