Falcon03:

I would tend to believe your data sheets on the pumps and say that you are running them over capacity.  Whether or not that third pump will help, we can't tell, you didn't give us enough information.  The relationship between pumps and power depends a great deal on the type of pump you are using.  Often times the electrical system is designed with greater capacity than the requirements for the pumps and shouldnt be used as a basis for design worthiness of the pumps.

BobPE

The motor current should roughly follow BHP.

BHP vs flow is in general not a monotonic increasing curve and depends on pump type as Bob says.

The shape of the BHP vs flow curve for various types of pumps was discussed at:
Thread407-27055

If a single-stage axial flow in general you will draw less current as you increase flow.  Many mixed flow multi-stage pumps have a peak in BHP vs flow near BEP, and will decrease above that.

If you are that part of the curve than there is no concern for your motor. But there may be concern for operating near runout condition. Does it cause excessive vibration, does it cause cavitation etc.

Thanks guys,

To answer your question electicpete: No we have not noticed any problems with our pumps with the current operation windows. No vibration and cavitations. I just checked the AMP meter in our SUB it shows 210 amp vs 254 design. I will try to find more information and will get back to you. Thanks again

If the flow is equally split between the two running pumps,then they're both operating at 110.7% of design flowrate which is well below the 120 to 130% maximum flow generally recommended for centrifugal pumps that do not have a specified maximum flowrate (usually presented as a minimum system resistance curve for a plant that has varying, unthrottled flowrates). Reduced current draw above design flowrate just means that the peak current is closer to the design flowrate which may be at or near the design best efficiency point (bep) flow. The biggest concerns for very high flowrates are poor incidence angle-induced impeller channel backflow (recirculation) instabilities and, for fluid film unidirectional, toward-impeller loaded fluid film thrust bearings, the possibility of thrust "crossover" to toward-motor unloading which causes shoe flutter damage in Kingsbury-type thrust bearings.Dual-directional Kingsbury thrust bearings always have one set of shoes unloaded and prone to flutter damage and so are not a recommended design feature.

vanstoja  - I know you have talked at length before about the complex causes of thrust. It's hard for me to imagine decreasing thrust with increasing flow.  I always pictured the change in momentum of the fluid was the biggest component of the thrust.

... I guess I forgot about differential pressure across impeller which works in the opposite way. (increasing flow corresponds to reduced pump dp as we move along curve).

vanstoja:

they make a lot of pumps without specified flow rates!!!! I see them in peices all the time and they dont work well....   Seriously, consider the original question, they are trying to increase capacity and are sacrificing the pumps as usual to accomplish this.  The only way I can see increasing the flow like that is knowing you need more flow and changing the system curve with two pumps runing to get it.  You won't hear signs of cavitation too easily in crude, its a great masking fluid for sound.  The 15 to 20 percent range of operation is for the one pump running, with two in service, this range no longer holds.  You would have to look at the combined pump curve with his system curve to see where he stands with the prolonged survival of the pumps.




BobPE

Falcon03:

27336 M3/D is over 173000 BFPD or 5000 GPM.  You have big pumps!!!

I would like to add few things to all of the help offered so far that you may want to consider:

1)
Pump designs include calculations for both the fluid gradient and fluid viscosity.  If your actual operating conditions differ from the design data it will cause errors in the predicted operating conditions.  As an example if you’re pumping a low gravity crude the gradient was assumed at a given temperature.  If your actual fluid temp is different than the design considerations both the gradient and fluid viscosity will change therefore the actual assumed operation will change.

1A)
If you do have a low gravity crude you may also have entrained gas trapped in the fluid that will add to a change in the fluid gravity.  If there is gas it may be breaking out or expanding in your flow line after the pump generating lift that was not accounted for in the original design.  This would be a facility separator or chemical type thing that you should consider.  Personally I have had sub-pumps in wells with both low gravity oil and a low GOR that required gas separation in the wells to prevent gas in the pipe or sales line.  This phenomena occurred after both a heater treater and storage tank.  Makes me wonder whose gas correlations are correct!    

2)
If you are running on a VFD or have power factor correcting capacitors you may actually be operating at the predicted kW load, but the amps you're reading could be deceiving you.  It is my understanding that on the input of a VSD, where you have a full wave rectifier the power factor will appear as unity or 1.00.  This would also be the case with well designed power factor correcting capacitors too. It then depends on where you are actually reading your amps at as to what you should expect to see.  I'm sure that someone like electricpete could explain this better than me, but the fact is that it takes about 746 kW per HP for an electric motor and kW may be a more accurate way for you to consider your problem.    

3)
If your system was designed for all three pumps to operate at the same time and you are only running two the pressure or head will be lower causing a higher flow at this time.  You may want to compare both your current system operating curve and your original system design curve.

4)
I have to believe that you have some big pumps and high HP motors.  If there is an actual failure you will expect the manufacture to at least consider warranties.  I would recommend that you go to the manufacture now and get their input, in writing, on your current operation.  This may save you headaches down the road.    

4B)
Before you contact your pump manufacture you need to get the fluid temp at the pump intake and pump discharge as well as a produced fluid chemical analysis.  You have a crude oil that should have very good lubricating properties, but some minor chemical additives like amine, polymers etc. require self lubricating bearings made of materials like graphite.  

4C)
Another down side to your good lubrication is the specific heat of the oil.  Oil by comparison to water does not transfer heat very well.  By changing your operating or efficiency point you may have heat/pump seal problems that needs to be addressed reasonably quick.

5)
During the design phase of your project your pump supplier may have believed they know more about your operation than you do.  They may have coated the stages of your pump with something that you're not aware of.  As an example if they expect that you will have entrained H2S or CO2 they may have coated the stages with some form of coating to protect the staging metallurgy.  If this coating has a thin, more smooth finish than that of the original metal your efficiency will improve or conversely your load amps will go down.

6) FYI
For the short term a data acquisition system may represent more of an expense than value, but after some run history by getting volts, amps, flow rates, pressures, temperatures etc. you may be able to change your pump replacements into pump repairs.  Just food for thought!

The big seven)
I wish I would have sold these pumps!!!!

I hope this helps!!!
David

Falcon03,

Sorry to jump in so late on this thread, but thought that some basic pump info may be of help.

The first recommendation that I would make is that you obtain a copy of the pump performance curve.  On this size of pump (> 5,000 gpm), you must have some witnessed, "Certified Performance Curves" available.  If you don't have these, the pump supplier can provide copies.

Secondly, pump brakehorsepower is based on flow rate and differential head (not flow rate only), so depending on the pump specific speed and operating point on the curve, the BHP can drop with increasing flow rates.

BHP = Q (flow rate) x Hd (differential head) x Specific Gravity / 3960 x Hydraulic Efficiency.

As a general rule: low Specific Speed (Ns) pumps (radial flow pumps with Ns of less than 1,000) have continuously increasing BHP requirements with increasing flow rates.  Medium Ns pumps (mixed flow pumps with Ns above 1,000 to 4,000)have increasing BHP requirements up to about bep, and then reducing BHP to runout.  High Ns pumps (axial flow with Ns greater than 6,000) can be considered to have continously lower BHP requirements at higher flow rates.

If your pumps are relatively high specific speed then the lower bhp that you have observed makes sense.  

Kawartha - I am happy to see that your description of the general pattern of BHP vs flow curve matches what I have pieced together with assistance from folks here (as described in the link above).  I have not used the concept of specific speed, but my observation was that:

-pure radial flow usually have continuously increasing BHP vs flow
-pure axial usually have continuously decreasing BHP vs flow
-mixed flow have more complex pattern which often peaks at BEP and decreases both directions from there.

The other part I mentioned was that the pure radial and pure axial flow seem to appear only in single-stage pumps... the mixed flow seems to occur in multi-stage pumps.

Electricpete,

Sorry, I didn't mean to denigrate your comments.  Just thought that the thread should get back on track.

However, the full range of specific speeds apply equally to single and multistage pumps.  Multi-stage axial flow pumps are not uncommon, and radial flow designs are extremely common on multistage pumps such as deepwell designs (high head / low flow).  Likewise, most of the end suction (ANSI and API) pumps are mixed flow design.

Regards

Hi Kawartha. I didn't view your comments as negative at all. I was just looking to bounce my understanding off of you.  All of the multi-stage pumps I have encountered in my limited experience have been mixed flow.  Thanks.

Electricpete,
   Generally momentum thrust change is not governing for axial hydraulic thrust loading. It depends on the angle of the flow turn which is 90 degrees for pure radial (low specific speed) impellers and zero for axial flow impellers (high specific speed). The thrust curve change with flowrate tends to match the head-flow curve trend particularly on the high side of bep flow since it is the product of static head minus pressure losses therough the leakage path times the unbalanced front-to-back impeller area.

Bob,
   I'm used to seeing pump designs operating around +/- 20% of design best efficiency flowrate because the powerplant system designers simply couldn't predict the targeted system resistance curves to any better accuracy...repeatedly on several plant designs in succession. Consequently, I'm likely to view a 10% off-design condition as not too disturbing. I'm used to pressurized closed systems with prescribed large cavitation prevention margins built in by conservative minimum suction pressure operating requirements. This does not hold up well if there is only a small cavitation margin at bep flowrate. Then the inlet

 flow incidence angles on the blade leading edges can turn "recirculatin" secondary flows into vortex strings breeding cavitation vapor bubbles in their coreswhich is decidedly bad news as you suggest

vanstoja / BobPE

Your comments regarding recirculation cavitation for pumps operating outside of the recommended envelope are very helpful.  However, they may not be applicable to Falcon03's requirement.  Falcon03 advised the design flow rate and the actual flow rate conditions.  I don't think that he provided the best efficiency flow rate for the pumps, so without this information we can not determine the recommended flow envelope for the pump.

Be very careful when establishing maximum/minimum flows relative to bep.  A very good reference on this issue is ANSI/HI 9.6.3-1997 "Centrifugal and Vertical Pumps - for allowable operating region".  The preferred operating range (POR) varies considerably based on pump specific speed, the pump suction energy, impeller tip speed, and design of the pump.

As a general rule, HI recommends a POR between 70% and 120%.  Depending on max. impeller speed, POR can range from 25% to 120% of bep.  For higher specific speed pumps (axial flow design), the POR is much narrower than for mixed or radial flow pumps.  I have had experience with some low horsepower (3HP), multistage, radial flow pumps operating at zero (0) flow for years - without damage.

Design engineers should always obtain the POR recommended by the pump manufacturer, as warranty will obviously be impacted by operation of the pump outside of recommended range.

Thank you all of your help.

Kawartha, I have checked your equation above and I have applied but I got different results. I would like to make sure that the above equation is based on English unit not Metric. Because we do metric unit. Thanks again.

Falcon03,

U.S. Customary Units.
To calculate Pump BHP:  BHP = Q x Hd x S.G. / 3960 x Eff.

BHP - in HP
Q - flow rate in usgpm
Hd - differential head in feet of liquid
S.G. - Specific Gravity (dimensionless)
Eff. - Pump efficiency at Q (from pump performance curve)

Metric Units.
Formula for Metric, I believe is as follows.
kW = Q x Hd x S.G./ 366 x Eff.

Q - cubic M / hour
Hd - M

Regards,

Kawartha:

I think our intrepretation of the problem stems from information supplied by Falcone3.  he stated that they are over capacity as indicated on the pump data sheets.  From this, one can speculate the problems associated with this overcapacity.

I can't agree with you more though on POR.  This is the most widely misunderstood and neglected range associated with a pump.  In my opinion, it is so important that I have it stamped on the Pump plate in my specifications.

BobPE

BobPE

Thanks for your comments.  I can understand the confusion on this issue.  It would have been more helpful if Falcon03 had provided the pump bep, but he only provided the design and operating flow rate.  My suspicion is that the pump is a relatively high specific speed pump, and that the design point is probably close to or slightly on the right of bep.  This would help to explain the observed lower operating flow hp vs. design flow hp.  

Your comments regarding possible recirculation problems is pertinent and Falcon03 should do further investigation into this potential problem.  As you suggest they may be close to the maximum POR.  Since a lot of the answers will be provided on the pump performance curve - I suggested that Falcon03 obtain a copy.   

If some one is welling to get the pump data sheet I will be welling to send it over to you. What I need is the Fax No. or e-mail address so I can scan the data sheet to you.

Thank you again and sorry If I wasn't clear enough in my original question.

To be frank with you guys I didn't get so much help from our pumping machine people as you provided to me, since I'm Chemical eng.

I'm sure I will need you assistance in the future.

Bye.

Falcon03

You can email me at rburri@pathcom.com
Forward the pump data sheets and also the pump performance curve.  I don't know if you are familiar with performance curves, but the curve represents the performance of the pump based on the x-axis - flow rate and y-axis - the differential head.  The performance curve will also provide pump BHP from 0 flow to runout.  NPSHr curves may also be included on the performance curve.  The curves also may include minimum and maximum recommended flow rates (POR).

Regards,

Richard

Falcon03

A simple answer to your question is yes there is a relationship between pump amps and flow.  In fact most deep well multi-stage centrifugal pumps used in the petroleum industry are controlled by “underload” or “overload” devices that since the motor current.  To determine the setting point for these devices you need to use the pump curve as Kawartha explained.

In the sub-pump industry the pump curves are very clear.  The x axis is flow, the Y axis is head and the second Y axis is HP and Efficiency.  We don’t plot multiple stage sizes on a single curve to maintain clarity.  

I have looked at other pump curves that may have several impellor sizes listed on a single curve.  In some cases this type of curve confuses me.  In a lot of cases with this type curve I will make a pump curve in EXCEL just using the specific stage size that I require.

I would like a copy of your data for two reasons.  For me I would like an example of what not to do with curves.  If the curve is difficult for the end user it is not a good curve.  Second I will plot a single curve that will be simpler for you to use.

If you get a minute please send me a copy of your data.

davidlsus@yahoo.com

Thanks
David

d23 - if you are saying there is always continuously increasing current with increasing flow, then I think you will find disagreement from the folks who have posted above. I apologize if I have misunderstood you.

electricpete

I am not saying that all, but you do make a very good point that I was not clear with my responce.  In fact I have argued your same point in other post.  

From my perspective:

Most "deep well" pumps in fact do not continue to increase BHP, amps etc... as the flow increases.  In fact we have one pump in our product line that has a higher BHP requirement at shut-in than any other point in the curve.  I was only saying that in the case of deep well pumps they are typically controlled by current.  

I saw another post where someone was talking about recommended pump ranges.  Pumps do have a recommended range.  There are exceptions, but if they are operated outside that range pump or equipment damage can occur.  It does not matter if the current (BHP required) goes up or down the pump needs to shut down to prevent damage.  Using both overload and underload devices can shut the pump down or prevent pump damage.     

In the case of multi-stage deep well pumps they are operated in a vertical position.  If you pump a well off the pump needs to shut down to prevent caviation or running a pump dry.  If the well is trying to flow on it's own and you start a pump the impellors will try to climb out of the pump housing.  It does depend on the specific pump, but normally I would expect both cases to operate with a low BHP requirment or low current.  Both cases will lead to a premature pump failure and need to be avoided.

The question on this post was is there a relationship between pump amps and rated flow.  An over simplified answer would be yes.  If we had the pump curve and motor spec a more specific answer could be offered.

David

d23

Just a couple of comments.

First, there is a simple relationship between amps and flow and it is defined by the formula:   BHP = Q x H x S.G. / 3960 x Eff (in U.S. Customary Units).  Refer to my posts above.  This applies to all centrifugal pumps, whether submersible or not.  

Secondly, I have had extensive experience in the submersible "deep well" pump industry, and consider many of your comments regarding submersible pumps as incomplete and in some cases not typical.  For example you wrote that submersible pumps are operated in the vertical position, when in fact many "deep well" pumps are installed in the horizontal position.  These are used extensively to boost water pressure in municipal water distribution systems (installed in booster cans), in ships to pump ballast water, etc.

Perhaps your "deep well" pump comments should be the topic of another thread.

Kawartha:

Your BHP required formula is true.  We need to remember that the pump affinity laws apply to centrifugal loads.  If you slow the speed of the pump down the HP required changes faster than the head or flow changes.  The same laws are true if you reduce the area of the impellor(s).  By lowering the speed or trimming the impellor enough the pump curve on the right hand side (low head-high flow) will eventually require less HP at 0 flow than it does at its BEP point.  This can be verified by finding a pump that offers several different size impellors for the same pump.  The smaller the impellor gets the more flat the BHP required curve gets.

If Falcon03 is looking for some linear correlation between flow and amps to be used on all centrifugal pumps it doesn't exist.  If the question is "do amps increase with flow" they may or may not depending on the centrifugal pump design.  If he is asking if amps can be predicted using a pump curve the answer would be yes in most cases.  The reason I say most cases if you assume a pump has a 10 HP load, but a 25 HP motor is being used due to availability then the motor characteristics will be such that predicting amps based on pump curve information (BHP Required) won't happen.

David

d23,

If you apply the affinity laws you'll find that the head-capacity curve shrinks towards the 0 point, as does the bhp curve.  The shape of a trimmed impeller curve after trimming the impeller is consistent with the curve prior to trimming.  In other words, if the bhp curve is continuously increasing with increasing flow (typical of a low specific speed pump), this shape will also be reflected in the trimmed impeller curve.  The reason that the curves (flow/head or bhp) appear to flatten is because the scale is reduced.

Keep in mind that there is a limit to the amount of trimming which is possible, and this limit varies significantly based on pump design.  It is not unusual to trim typical impellers by 25% and some even to 50% (of maximum diameter), but more serious trims result in unpredictable performance because of significant alterations in the geometry of the impeller vane discharge angle, the relationship of the impeller periphery to the casing (cutwater or diffuser vanes), etc.

In addition, the affinity laws are useful in calculating expected performance at reduced impeller diameters, but as noted above, the pump geometry changes with impeller trim, and actual performance can vary appreciably from calculated results.  This is one reason why pump manufacturers predict performance subject to test.

kawartha:

Please stop marking your own posts as helpful/expert, thats for others to do and it is verry annoying.

BobPE

BobPE,

I have never intentionally marked my posts as helpful/expert. If I am marking them it is unintentional and I don't know how I'm doing it.

Check my Member Profile and you'll find that I have not used the helpful/expert mark on any string.

Your claim is insulting.

BobPE

Checked my last post and mark is there again.  I don't know what the problem is, but it's not intentional.  Doesn't seem to happen on other strings.

BobPE

After my last 2 posts were also marked as helpful/expert, I thought that there must be a reason for the repetitive mark.  In checking other strings, I find that when a post is marked as h/e, then all subsequent posts by that individual are marked as well.

Possibly an apology might be in order BobPE.

kawartha:

Thanks for your input, I will look into fixing the problem...

I do apologize, I perhaps should have worded my initial post to you a little better.  I assumed that the computer is always right, a mistake that I am bound to repeat again in the future I am sure!!

Take care...

BobPE,

I appreciate your comments.  No harm done.

If you are looking into this you may want to ensure that individuals can not vote for themselves, although I suspect that this is already the case.

Kawartha, All

I actually was not trying to confuse the issue by mentioning deep-well pumps.  In one of your responses it seemed that I may have irritated you some.  That was not my intent either!

With deep-well type pumps we have a small OD pump and try to produce high volumes and high head.  In a more industrial world you would consider my pumps to be “Medium flow, high head pumps.”  I have several units operating that produce over 100 GPM with over 10,000 feet of head inside a 5 ½” OD casing.  To add to the confusion in the past I have operated pumps at over 5000 RPM.  This does reduce the MTBF, but the economics of the fluid being produced justified the op-X for my customer.  

We do two things that make most of our pumps require less BHP at low head, high flow than they do at their BEP.  We have very small diameter impellors and we operate them at high speeds.  This is the pump affinity laws in both cases.

I looked up a couple links to the Mc Nally Institute for you that explains centrifugal pumps from my perspective much better than I can.

Quote one from the Mc Nally Institute:

If you are using a low specific speed impeller the pump will require less horsepower if you start with the discharge valve throttled. If you have a higher specific speed impeller the high power requirement comes at higher head so you would want to start with the discharge valve open.

http://www.mcnallyinstitute.com/CDweb/s-html/s083.htm

Quote two from the Mc Nally Institute:

• The steepness of the head-capacity curve increases as specific speed increases.
• At low specific speed power consumption is lowest at shut off and rises as flow increases. This means that the motor could be over loaded at the higher flow rates unless this was considered at the time of purchase.
• At medium specific speed the power curve peaks at approximately the best efficiency point. This is a non-overloading feature meaning that the pump can work safely over most of the fluid range with a motor speed to meet the best efficiency point (BEP) requirement.
• High specific speed pumps have a falling power curve with maximum power occurring at minimum flow. These pumps should never be started with the discharge valve shut. If throttling is required a motor of greater power will be necessary.

http://www.mcnallyinstitute.com/CDweb/s-html/s072.htm

Once again my intent was not to confuse the issue of the original question or to upset anyone.  I was just trying to point out that just because you have a centrifugal pump do not arbitrarily expect the motor current to increase with the flow.

If you would like to see some pump curves that you would consider being strange let me know.

David

Falcon03,

Excuse me for jumping in so late...hopefully I'm providing something new.

You mentioned that you measured 210 Amps, but you didn't mention what your power factor is.

For a 3-phase motor,

E=Voltage
I=Aperage
eff=motor efficiency
pf=power factor

Amps(I)=746*HP/(1.73*E*eff*pf)

or

Horsepower=1.73*I*E*eff*pf/746

If a very conservative power factor was used for design (i.e., 0.82) and the actual power factor is quite good (i.e., 0.92) there can be a significant difference in the result.

In addition to the above, are you sure that your actual flowing density is identical to what you see on the pump curve (remember to use SG at flowing conditions...not standard conditions)?

Also ensure that you have calculated the flowrate at flowing conditions for reading the pump curve.  The flowrates indicated in your control room will likely report at standard conditions...not flowing.


I hope this helps!

Regards,

Bob

Thank you Kbander for your feed back. The power factor I used is 0.85. Regarding the flow rate, we usaully refer to the actual flow.