Over the years use of the term “minimum flow” has evolved. Decades ago industrial centrifugal pump manufacturers quoted a single, relatively low value for minimum flow intended to prevent users from running their pumps to destruction. The term “minimum flow” generally meant the lowest continuous flow the pump was permitted to operate, without reference to duration, vibration level or other criteria. Today we have minimum flow values for continuous operation, for intermittent operation and for permissible temperature rise.
The accompanying chart Pump phenomena and minimum flows shows the relationships among the various off-design pump phenomena and minimum flow conditions. The head versus rate of flow curve with indicated phenomena is a variation of S. Gopalakrishnan’s from his well-cited paper titled, “A New Method for Computing Minimum Flow,” Proceedings of the Fifth International Pump Users Symposium; Texas A&M University, May 1988, pp. 41-47. As an aside, I recall Gopal (everyone knew him by that name) had made a local technical presentation using the now well-known chart, before it was published. Evidently the chart was copied from a handout of the overhead slides and was quickly pirated by another, and then others. Copies or variants of this chart are now found widely in papers and presentations on pumps.
The quoted minimum flow for continuous operation is usually called “Minimum Continuous Stable Flow” or its more common abbreviation “MCSF.” It is the flow below which the pump should not be operated continuously. The usual purpose of MCSF is to achieve satisfactory bearing and seal life; however MCSF may be based on other considerations. Any of the following factors may be considered in establishing the MCSF:
- manufacturer’s experience
- rule of thumb
- calculated onset of suction recirculation or discharge recirculation
- radial thrust
- temperature rise
- cavitation erosion intensity
- maximum permissible pressure rise (for system purposes)
- maximum permissible power rise (high specific speed and axial flow pumps)
- a combination of the above factors or others not listed
For hydrocarbon process industry API 610 specification pumps, the value of MCSF is normally coincident with the lower flow limit of the “Acceptable Operating Range” (refer to chart titled “Vibration limits for Allowable Operating Range and Preferred Operating Range”) where a specified vibration limit is not to be exceeded.
MCSF is a value that can range from roughly 10% to 80% of Best Efficiency Point flow depending on pump size and type, operating speed, impeller suction geometry, liquid density, and other factors. A size 2” (50mm) discharge single-stage process pump may have an MCSF as low as 10% of BEP flow. MCSF is often in the range of 30% to 60% of BEP flow for process pumps with discharge sizes 3” (75 mm) and larger. Large mixed flow vertical pumps and very high head-per-stage centrifugal pumps may have an MCSF greater than 60% of BEP flow. Axial flow pumps have a power curve that rises toward shutoff and minimum flow may be limited by the power rating of its driver.
On certain high energy pumps the minimum flow is governed by cavitation erosion damage. Minimum continuous flow for 40,000-hour impeller erosion life is where the system NPSH Available curve intersects the pump’s NPSH Required curve, at lower-than-BEP flow.
Intermittent minimum flow, when specified, is usually given as a percentage of MCSF. On some applications the governing value may be based on temperature rise. On large high energy pumps the value of intermittent minimum flow could be, for example, “70% of MCSF and not to exceed 100 hours per year.”
For some applications a thermal minimum flow or “Minimum Continuous Thermal Flow” is specified based on permissible liquid temperature rise. MCTF is usually, but not necessarily, lower than MCSF. While a pump thermal minimum flow is not always specified, the end user can readily calculate its value based on input mechanical power heating up the liquid. The limiting temperature rise is based on a safe margin to prevent flashing of the pumped liquid to vapor, potentially causing pump seizure.
Thermal minimum flow is not normally a concern at pump start-up as long as the closed discharge valve is set to begin opening right away. If the margin of system NPSHA above pump NPSHR is minimal, then the temperature rise conditions at pump start-up should be checked carefully.
A few pump applications, such as a vertical turbine jockey pump for maintaining pressure in a large fire sprinkler system, can potentially operate continuously at shutoff while pump suction recirculation mixes with the water in the sump in which it operates. The sump acts as a heat sink and a minimal water temperature rise is not a problem. This example is a rare exception to an almost invariable stricture on operating the pump continuously at shutoff.
The purpose of minimum flow is generally to prevent undue wear and tear or damage to the pump. In the real environment of a process or utility plant, a pump is operated at just about any condition demanded by the situation at hand. Thus there are different pump minimum flows for different purposes.
For an independent evaluation of a pump minimum flow issue, contact an experienced consulting engineer who can help with your specific application. Please take a look at our services to see our areas of expertise.
Well said Randy.
So, When we decide using minimum flow valve or using motor inverter? thks
Ricky,
The energy costs of using a valve will need to be compared with that of a Variable Speed Drive (inverter). An energy and cost evaluation of the pumping system, the pump, operational cycles of the process and the machinery, etc. will provide a good indication of which type of pump minimum flow control is better. There are benefits to VSD pump control for many, but not all, systems. Benefits such as reduced pump and valve wear and tear, reduced power demand and reduced noise may be more difficult to justify economically, but may help to sway management opinion in favor of approval. I hope this helps.
Randal Ferman
If you can use a VFD that would be the best way to control flow and pressure. It’s more efficient. Long term less costly because you are using less power. A valve restricts flow but the pump is working just as hard. Pumps are dumb animals, if you will. They are selcted by the engineer to run at the best efficdiency point based on the given conditions. You change that with a valve you lose efficiency and could cause some long term added maintnenance.
Head varies as the square of the change in speed or impeller diameter, therefore the installation of VSD pump control depends plant characteristics.
In two locations in the graph cavitation is mentioned,why
Gentlemen,
You are talking about minimum flows like you would like pump to run where ever you
want to satisfy your minimum.
I designed thousand of hydraulic centrifugal pumps with very strict requirements for
given flow and head.
To not go what is pump specific speed and its value (the best is 2500 to 3500 with units of gallons, rpm, feet
in USA) pump has required and available suction heads region and it is very sensitive how one picks discharge angle and speeds and on test stand you measure head vs. flows and find the curve for min and max flow.
Pumps are like centrifugal compressors – at low flows they are unstable so that compressors surge and shake to destruction of itself.
For a given pump impeller at say 100 gpm geometry of impeller is done for that flow
and if you want to run say 50 gpm I always kept OD of impeller same and eye dia same
with diffrent inlet and outlet angles for pump to run satisfactory.
zelen
310.839.1764
Really is it possible ?
How do you change inlet and outlet angles ?
For Zelen’s comment, I think it is best to assume that what he is saying is technically rational. If you read it from that point of view, not necessarily using a literal interpretation, then it makes sense.
very interesting
i just want to know on the set pressure for my valve. my friend said the set pressure is based on the pressure at mscf. does it mean when mscf occured, it indicate that the pressure in the system is in maximum?
Faiz, Yes. A centrifugal pump Head-Capacity (head versus rate of flow) curve rises in pressure from the rated point towards pump shutoff. The MCSF falls somewhere between the rated point and shutoff, so the pressure at MCSF will be the maximum recommended for continuous pump operation. Randal
Nice conversations and brilliant information’s. thank gentleman, i have a question it is all about mechanical MCF and Process MCF ? they are not the same and i need clarification . thank u in advance
Amir,
I have not before seen the term “Mechanical MCF.” But with the word “Mechanical,” I assume it relates to requirements of the pump. The Mechanical MCF would then be the same as either the MCSF or the MCTF.
The term “Process MCF,” I assume, relates to requirements of the process external to the pump. This value may be entirely different than the pump MCSF or MCTF, but might possibly be related by pump temperature rise considerations.
I hope this helps.
Dear Rendal,
Thanks for your prompt reply.
I guess mechanical minimum flow related to the shut of head and piping and equipments downstream of the pumps which should able to work at this pressure. but minimum process continues flow related to the cavitation and minimum flow line of pump. I guess…
please let me know if you find any complimentary information in this case.
Regards,
Amir, The terms, obviously, must be defined. It is possible, but not usual, that pump minimum flow is determined by piping allowable working pressure. This minimum flow value may or may not correspond to the pump’s MCSF or MCTF.
Cycle Stop Valves are used frequently in the irrigation business. What concerns should one have about using this type of valve as it relates to MCSF?
JK, A Cycle Stop Valve (CSV) is a pressure regulating device for pumps designed to bypass a minimum flow in the fully closed position. On smaller pumps, this bypass flow is sufficient to prevent serious overheating of the pump. The user should verify that the CSV bypass flow is consistent with the pump manufacturer’s minimum flow recommendations. I suggest checking with your equipment distributor.
Hi, I have a question.
I have a problem with my centrifugal pump. The minimum continuous flow was stated at 265 m3/hr in the datasheet. However when we try to lower the flow to 300 m3/hr the vibration starts to increase and we manage to reach to only 450 m3/hr until it reach the vibration alarm level. We did not go further as it would trip the machine and damage the pump. The pump was running fine at the design operating flow, however due to low demand we have to reduce the pump flow rate.
Can anybody point us to the right direction on the root cause and how can we overcome this..??
Additional info:
The shut of head is 113m, the BEP is 80.2% at 1480rpm, Capacity of 897m3/hr, TDH 94m, and NPSH R 2.99m, NPSH A 5.3m.
MCSF value given by any OEM is subject to control’ valve location from discharge valve. If control valve distance increases, MCSF will also be increased. Assume your pump installation and internals are fine. At the point vibration reaches recommended permissible limit is your pump’s MCSF.
Kirit, Your comment about the control valve and discharge valve positions relative to MCSF is new information to me – interesting. Vibration is often the criteria for MCSF but there can others such as heating of the pumped product or cavitation. Best, Randal
Dear Rendal,
In what conditions should we pay attention to Minimum Continuous Thermal Flow? Or we just have to calculate the values of MCSF and MCTF and then define the larger one to be the minumum continuous flow for pump to operate?
Jaqueline,
Minimum Continuous Thermal Flow (MCTF) can be an issue with high head multi-stage pumps. Liquid heats up as it passes through the pump. Some of this heated liquid passes through the axial thrust balance device and returns to suction. If sufficiently heated, this liquid may now flash suddenly into vapor and cause catastrophic failure of the pump. This is particularly a concern, and definitely an operational constraint, with boiler feed water applications where the pump takes suction directly from an elevated deaerator tank. In such an arrangement, the margin of NPSH Available at the pump suction, above the NPSH Required by the pump, is limited by the height of the deaerator and resistance losses in the piping.
For the majority of pump applications the calculated MCTF will be substantially lower than the Minimum Continuous Stable Flow (MCSF).
Best holiday wishes,
Randal
Knidly tell me the conditions of Minimum flow of centrifugal pumps?
Hi dear Gents
And congrats for this discussion. Please, assume on rated speed we have a clear minimum continuous flow defined by pump vendor, how that minimum flow is changes if the pump speed is changes to by using a VDF?
The lower the pump speed, the lower the minimum continuous flow too?
Many thanks for your feedback
Aurel,
I will take liberty to rephrase this as a question: “How is pump minimum flow affected by a reduction in speed?”
Generally it would be safe to reduce pump minimum flow proportional to the reduction in pump speed. That is, if the minimum flow value provided by the manufacturer is based on methods that take into account the onset of suction recirculation, NPSH margin ratio, and energy level of the pump.
However, the NPSH-Required and energy levels vary exponentially with speed, so that at lower speeds one should be able to run a pump a lower minimum flows than would be indicated by a simple linear extrapolation.
Thank you for bringing up this excellent point.
Randal
Many thanks Randal
Dear Randal,
How to calculate minimum continuous thermal flow ?
Alan,
For reference, I assume you’ve looked at my reply to Jaqueline on 27 November 2013. Calculate the temperature rise of the pump for several low flow conditions. At these low flow conditions adjust the NPSHR upward as necessary to compensate for the temperature rise of the balance leakage. Remember, the return leakage flow is not permitted to flash to vapor as it returns to the reduced pressure suction area. The pump manufacturer’s NPSHR curve may already reflect this requirement. The MCTF condition is established based on having sufficient NPSH margin to prevent flashing.
I hope this helps.
Randal
Minimum flow pumps help protect the pump if the downstream unit is plugged or under any circumstances went down or tripped. On the other hand, the base of every column has a level set point it helps maintaining the level set point not to run out of level which has a high possibility of effecting the on a column distillation to malfunction and also the heating media will rupture the tubes of the re-boiler.
Please comment back if I was mistaken
Dear Moe,
You are correct – pump minimum flow can always serve a process or safety function. In either case, the requirements and limitations of both the pump and the system must be considered.
Thank you for the comment,
Randal
Appreciate if someone in this forum can clarify what happen if in case of failure of a centrifugal multi stage HP BFW pump minimum flow recycle valve and possibility of overpressuring the suction pressure if the minimum flow line is directly connected to the suction of the pump. the below is the system description.
the system consists of a series of 3 LP BFW pumps which are feeding to 3 HP BFW pumps with common suction and discharge header.
the LP pumps discharge are rated for 300# and HP Pumps for 1500# pressure rating. the LP pumps have individual minimum flow recycle back to the suction vessel which is atmospheric BFW tank . the HP pumps minimum flow recycle lines are connected directly to the HP pumps common suction header which is common Discharge header of LP pumps via a cooler which can cool down the minimum flow before it is recycled back to the HP pump suction (so no overheating issue). the dischrge head of LP pumps are 320 m and for HP pumps it is 900 m at their rated flow 130 m3/hr each (normally two pumps running and one in stand-by).
so the question is if the minimum flow recycle valve of one HP pump fails to fully open then what would happen to the suction pressure of that HP pump if only one sets of LP/HP pumps running and same question if two set of LP/HP pumps running.
the question is raised due to consideration of 300# rating of LP discharge header which is connected to 1500# rating of the suction and discharge of HP pumps and to evluate if there is a need for PSV for this case or not.
appreciate your feedback on this.
regards,
Matt
Matt,
The risk is remote, but there are extreme scenarios where the LP discharge header could become over-pressurized by the HP pumps. Imagine, for instance, that LP and the HP discharge valves are closed, or are nearly closed. Now, if one or more HP pumps are operating or are turned on, then one or more of the HP bypass minimum flow valves open up and the LP discharge header becomes over-pressurized. If one wants to reduce the risk of over-pressurization, then one should consider the possibilities by virtue of the connected piping and unplanned operating events. It’s a sort of ‘What’s the worst that can happen?’ way of thinking but this point of view might ultimately save lives.
Thank you for this one.
Randal
Dear,
I’m a piper and involved in a system check on a pipeline with two parallel distribution multistage(12) horz. pumps (Flowserve). (one running.). These are running with a VFD.
Normal 180 M3/hr@dynamic head of 1400m. I ‘m still searching for the setting of the Automatic Recirculation Control Valve. Still searching for this datasheet of this valve 🙁
When the demand in the discharge is less the minimum flow will return to directly to the storage sphere with locked open valves. This will be a new line from this minimum flow valve. The minimum flow is 70M3/hr–>1620m Head. Taken from the pumpcurve equals to 82.5 bar. At deadhead cond. (zero flow) a max. head of 1680 m. This give a max diff. pressure of 85.5 Bar. (Spec. Gravity 0,519) The max pressure in the suction is the max outlet pressure of the sphere. at the bottom. This spere is forseen with a safety stting of 15 Barg. the max. height spere is 18.6m equals 0.91 Bar. The max. suction pressure will be 15.91 Bar. Adding this to the max. operating press. gives 98.4 Barg but adding this to the deadhead pressure of the pump this will give a maximum discharge of 15.91+85.5=101.4 Bar. So the design pressure taken for the minimum flowline is 101.5 Bar @ 50 degC. the max temp that can be reached is corrs. to the max vapour press. in the sphere safety valve @15Bar=38.5degC. The quantity taken out is pumped back to the sphere at the time of no demand to the pipeline. The variable frequency drive will lower the speed at that time but the pump will still running. I don’t think that I need more information of this minimum flow valve since the design of this line is 101.5 Bar @ 50 degC.
Please any comment?
Thank you in advance. Freddy
We want to unload a oil well at offshore and send the flow through the Crude oil export pump to other platform. The pump will take suction from a production separator and first feeds a booster pump and then the Crude oil export pump.
Initially the Oil well production will be much much below the rated flow of the COEP. So the plan is to run the COEP on recycle. But as the oil is bound to get heated up beyond pump design temperature, designer has proposed to install heat exchanger of the recycle line and cool the oil that goes back to the suction vessel , production separator.
My query is can the pump system survive if the initial well flow is very low (about 5000 bopd) as against 75000 bopd in normal condition.
Can the min flow recirculation line of such pump have a shell and tube heat exchanger, in which the recycle oil flow passes through the tube side. Can the recycle stream handle such resistances of passing through the tube side of the exchanger and still satisfy min flow requirements adequately?
Please comment on the feasibility, pros-cons of the above proposal.
Dear Rajib Mukhopadhyay,
The minimum flow operation proposed is quite low, as you are obviously aware. Temperature rise is a concern. The pump manufacturer should take a look at this to ensure that the temperature rise does not lead to internal flashing (vaporization). This would include checking flows and temperature rise through the pump’s internal recirculation passages such as wear rings and bushings.
Also, vibration at such low flows will likely exceed API 610 specifications. This may or may not be acceptable and should be discussed with the pump manufacturer. It’s possible, for instance, the manufacturer might suggest a lower speed selection or a stiffer rotor design, to name two possible alternatives.
Concerning the shell and tube heat exchanger, these can be readily customized for the application and there should be no problem selecting the tube side configuration to satisfy the flow resistance requirements.
The proposed conditions, as generally described, are feasible, but the technical details of the pump and the heat exchanger selections are important.
Best regards,
Randal
Does the minimum flow vary with type of pump(centrifugal, reciprocating, diaphragm,gear,screw etc),impeller (axial/radial/mixed flow etc) etc?. Give a link to study/download. If we design a pump whose flow can be adjusted (by varying stroke of pump)can we solve this problem?.
Sinnadurai,
Minimum flow, as an operational constraint, is applicable to rotodynamic (centrifugal or radial, mixed flow and axial flow) pump types. The value of minimum flow for these pumps varies widely and can depend upon a number of factors.
With positive displacement pumps, minimum flow is not so much the issue as is the discharge or back pressure which must not be permitted to exceed a specified value. Thus PD pumps require some means of output control such as variable speed and/or pressure safety valve or pressure regulator with a piping bypass return to the source reservoir.
Regards,
Randal
Minimum flow on reciprocating pumps is often determined by the method of lubrication. If the power end is splash lubricated, it will need to rotate above a minimum speed to supply adequate lube oil to the bearings. If the pump is force lubricated with an external pump, you can slow them down much more.
Detail of the efficient pump operation instruction will lead us achieving
longer pump life.
Inamul,
Yes, the closer we are to the most efficient rate of flow for the pump, at a given operating speed, the longer the pump can be expected to operate trouble-free.
Best,
Randal
Hi all
I have some questions about minimum flow line for centrifugal pumps:
1. Who is responsible for considering minimum flow line for pump? client or pump manufacturer?
2. Is it usual to consider minimum flow line for API OH2 type pump?(capacity 130 m3/hr and head 150 m)
Hoyar,
For an API 610 specification pump, the Centrifugal pump process data sheets have a place under the Performance section for specifying Minimum Continuous Flow, Thermal and Stable. This would be filled in by the pump vendor. It is customarily the responsibility of the process engineer to design and specify the minimum flow piping, valves and means of control based on the vendor provided data sheets and the process system requirements. Offhand I would say that for the pump you mention, a minimum flow line is not normally specified. But the process engineer must evaluate the temperature rise and rate of temperature rise, the margin to prevent flashing of product in the pump and mechanical seal and the system overall to determine whether a minimum flow line or special pump sensors and controls might be needed.
Good luck with this!
Randal
Thank you so much Mr. Ferman for your helpful reply.
Hi Randal,
Firstly, great job. This is an excellent and extremely informative discussion forum.
Now, coming to my query, I work for a licencor in oil and gas sector and so have very limited direct interaction with vendor. While issuing datasheet for pumps, we usually give a general comment like:” Pump vendor to specify minimum flow protection requirement”. However, for P&ID and pressure drop purposes, we do some preliminary estimates on what type of minimum flow circuit will be provided, viz.: RO with a continuous min flow/ Yarway valves/ Minimum flow line with control valve. The selection among these three options is done on basis of power requirement of the pump.
Can you suggest a more accurate method to select the type of minimum flow system required in basic engineering phase?
Rajarshi,
Each industry and each pump application is going to have special process, operational, safety and specification requirements to consider. Beyond the pump manufacturer’s data sheet specifications there are many variables in the mix. If one narrows the problem down to a specific industry and a specific process type, then design guidelines or a decision algorithm can be developed – you may already have something along these lines. Power requirement of the pump is the most rational point to start because input power directly affects temperature rise. Conceptually, one might be able to generalize sizing of the piping and the fittings required based on the head and flow of the pump, with different types of minimum flow control arrangements being decided by other factors such as power, process type, etc. But as one attempts to more broadly cover the range of process applications, or cross over into other industry categories, the level of effort is certain to escalate.
Thank you for this excellent question and I’m happy to hear you find this discussion helpful.
Randal
Hi Randal,
Thanks for the detailed explanation. Just to summarize your view, the approach based on power requirement is quite good in basic engineering phase.
Regarding type of industry, we mainly deal with hydrotreating and hydrocracking operations – so gasoil, VGO, diesel, naptha and sour water are the things that we mainly need to pump. And the discharge pressure roughly varies from about 6 kg to 80 kg (140-150 kg in case of hydrocrackers).
Based on this additional information, if you can suggest any other more specific guideline for selection of minimum flow arrangement, it will be very helpful.
Hi Randal,
I have two VS4 pump offers to select from ie one offered with rated flow of 9 m3/hr at 71m head, efficiency of 26%, a fairly steeper rising curve but requiring minimum flow pass flow of 5 m3/hr. Another one with 4 m3/hr at 71 m head rated flow but with 10% pump efficiency and a rising flatter H vs Q curve and without the need of minimum flow bypass. Kindly advice the right selection.
Operating parameters
Rated flow : 4 m3/hr
Head: 71 m
Liquid: open drain sour service
Seal: API Plan 74
Sp gr : 1.16
Application: Open drain pump mounted inside the open drain tank working WITH intermittent operation for tank unloading
Sunil, Even though the input power is greater for the second offering, I believe it would be the better of the two. There might be additional factors which could lead me to recommend the first offering or suggest that either of the two offerings are fine. Good luck, Randal
Thank you for taking the time to share your knowledge.
Hello Sir,
We are facing peculiar problem in our centrifugal pump. The pump is high flow rate pump with rated capacity of 538 m3/hr. The pump is intended to deliver stabilised crude oil from stripper column to Flow suction Tank (crude).
These pumps are new pumps replacing the old pumps. Now the scope was to replace the pumps but there was no scope related to minimum recirculation line updation. We missed to check its adequacy. Now the old pumps had MCSF – 170m3/hr and new pumps have 195m3/hr. The FCV in minimum circulation valve is also design to deliver only 170m3/hr.
We asked vendor to revisit the MCSF and reduce the values. However vendor recommended 30% of BEP value, i.e 195m3/hr. We are unable to start the train currently due to this discrepancy for concerns related to pump safety.
Could you please advise how vendor can make theoretical calculations to arrive at MCF value which will be only intermittent case in extreme scenario of blocked discharged. We are able to maintain the MCSF value during normal operations in all scenarios by limiting the closure of discharge valve to 90%. Hence the flow of 195m3/hr in all scenarios except the startup and discharge valve failure which will rare case for short duration before pump is tripped.
Sandesh,
A common practice for the nuclear electric power generation industry here in the U.S. is to allow for an intermittent minimum pump flow of 70% of MCSF for up to 100 hours per year. I do not know where this is written, except I mention it, not in connection with any specific industry, in my article titled ‘What is the Purpose of Minimum Flow?’ which is posted in the Ekwestrel website. Using this value would yield an intermittent minimum flow of 137 m3/hr.
If the pump vendor does not offer a calculation for this, you have a couple possible options. One would be to test the pump. Another would be to hire a consultant or engineering firm to evaluate the pump. Some measurements of the impeller would probably be necessary.
I hope this helps.
Randal