16 replies to this topic

[farid.k]

Hye team

 

I just confusing with pump vendor statement.

 

I just want to re-confirm my understanding regarding to centrifugal pump. Basically, we are process engineer will specify say; I want a centrifugal pump that can deliver 50 m3/h and can give head 70 meter. By giving the process datasheet, pump vendor will come-out a proposal that a pump that can deliver my flowrate demand and my required head and by right, my required head (70 meter) and flowrate (50 m3/h) is run at pump BEP as stated at pump curve.

 

In other word, when I receive my pump, install it at site, and run the pump, opening all the valves at pump discharge at fully open, I will get flowrate 50 m3/h and head 70 meter isn’t? Let say i close the valve at discharge pump 50%, the flow will be reduced and head will be increased isn’t?

 

I just confuse with typical pump vendor (brand E*ara), he said that, in order to know the flowrate, i need to buy and install the pump at site first. Then what if I don’t get my head and flowrate? By right, pump vendor already trim the impeller and specify the speed so that it can give my flow demand and head isn’t?

 

I never heard that need to install the pump first. Even during testing i.e. factory acceptance test, it must prove that the pump can give my process demand flowrate and head isn’t?

[Zauberberg]

Hi Farid,

 

It is not possible to purchase a pump which will have the BEP (best efficiency point) exactly at the Flow and Head conditions as specified by the client. Pumps are standardized products and none of the suppliers is going to fabricate a new pump just in order to have the specified H/Q conditions at the BEP point, for that particular pump. Because he can sell it only to your company, and then another company would ask for something close to that range but a bit different. This would take them - if they decide to do this for each client and each purchase order - into complete chaos because standardization of design would be lost. Imagine the world where only custom made pumps exist. Not to forget different manufacturing and performance standards for the same H/Q/BEP - so probably not even two identical pumps would exist on the whole planet.

 

You need to ensure that the pump can fully meet requirements of the process - in terms of head and flow variations. And to have this operating envelope decently close to the BEP. This is particularly important for continuous service pumps. For intermittent services, you are not so much concerned about efficiency (who cares if the pump runs 2 hours per month). Ask the supplier to provide you with catalog pump models which are closest to what you are looking for. If the product portfolio is not satisfying, change the supplier.

 

I really do not understand the other part of the post in which you say that the vendor requested installation of the pump in the field, in order to establish the required flows (?). Defining flow and head variations are parts of process design. For big, expensive pumps, factory acceptance test (FAT) is usually performed to verify design and rated conditions of the pump. For smaller pumps I have seen the FAT often cancelled (unless you are buying a lot of pumps from the same manufacturer) because there is no logic to mobilize the entire company crew to witness the FAT of a 5 kW pump. The whole trip would cost 20 times more than the pump itself. For these pumps, the SAT is usually sufficient.

 

Here is the link to Sulzer's Pump handbook. You can read the book online for more information: https://books.google...andbook&f=false

[fallah]

farid.k,

 

When you have already specified the rated capacity of the pump and relevant head at rated capacity, vendor would be responsible to deliver a pump with such capability at  (or close to) BEP and should send the document showing such capability before shipping the pump to  the site. Obviously, without your final approval on pump curve and documents, the pump can't be shipped by vendor and, on the other hand, even if the pump is brought to site after your approval the performance test should be done under vendor supervision and your witness. If the working point would be deviated from the BEP due to changing the system curve, vendor isn't responsible for such deviation...

[breizh]

Farid ,

You should receive from the vendor a performance test report of the pump before installation , should be mentioned in your PO to prevent dispute .

Vendor is responsible to supply the pump as per your specification .

Hope this helps.

Breizh

[Kiran T]

Dear all, 

hlw...

1) I want to install scrubber system in plant how can i calculate load of scrubber, i have 7 nos of 10 kl vessels and each have same exothermic reaction. During reaction sulfuric acid fumes came out and which i want to scrub.

2) If i have scrubber capacity of  1500 cfm then how many nos of reactors i can use with it ???

[fallah]

KIRANTANK1209,

 

Please submit your query in a new thread...

[farid.k]

Hi Farid,

 

It is not possible to purchase a pump which will have the BEP (best efficiency point) exactly at the Flow and Head conditions as specified by the client. Pumps are standardized products and none of the suppliers is going to fabricate a new pump just in order to have the specified H/Q conditions at the BEP point, for that particular pump. Because he can sell it only to your company, and then another company would ask for something close to that range but a bit different. This would take them - if they decide to do this for each client and each purchase order - into complete chaos because standardization of design would be lost. Imagine the world where only custom made pumps exist. Not to forget different manufacturing and performance standards for the same H/Q/BEP - so probably not even two identical pumps would exist on the whole planet.

 

You need to ensure that the pump can fully meet requirements of the process - in terms of head and flow variations. And to have this operating envelope decently close to the BEP. This is particularly important for continuous service pumps. For intermittent services, you are not so much concerned about efficiency (who cares if the pump runs 2 hours per month). Ask the supplier to provide you with catalog pump models which are closest to what you are looking for. If the product portfolio is not satisfying, change the supplier.

 

I really do not understand the other part of the post in which you say that the vendor requested installation of the pump in the field, in order to establish the required flows (?). Defining flow and head variations are parts of process design. For big, expensive pumps, factory acceptance test (FAT) is usually performed to verify design and rated conditions of the pump. For smaller pumps I have seen the FAT often cancelled (unless you are buying a lot of pumps from the same manufacturer) because there is no logic to mobilize the entire company crew to witness the FAT of a 5 kW pump. The whole trip would cost 20 times more than the pump itself. For these pumps, the SAT is usually sufficient.

 

Here is the link to Sulzer's Pump handbook. You can read the book online for more information: https://books.google...andbook&f=false

Thanks Mr. Zauberberg for the detail response.

 

Yea, I get your point. Regarding the point that you don’t understand, I need to clarify my actual problem so that you understand my point. One of my colleague already purchased a small submersible pump. This pump is to transfer effluent from sump to treatment pond. The point from the sump to the treatment pond is quite far (about 1 km) and when i refer to the purchased pump datasheet, it capable to produce head about 16 meters at flowrate 30+ m3/h. I quite doubt and I think the pump is undersize. Then I was asked to do verification calculation to check the pump is capable or not to pump from sump to treatment pond. Then in order to do the adequacy check, I need to do pressure drop calculation, and I did requested what is the normal flowrate of the pump. I asked my colleague for the design flowrate for the pump and he said 30m3/h. Then I try to cross check with the pump curve for the design flowrate and also the BEP. Unfortunately vendor didn’t provide the curve yet. Then I called vendor to ask the rated flowrate of the pump, and he said it depends on the system. I start confusing when he mentioned that. I asked again how about the design flow at BEP for this pump, he said he can’t answer my question. He said depends on my system. I asked again, what will be the flowrate of the pump if the valve at discharge pipe fully open? Will it deliver 30m2/h and give my required head? He said need to install it at site first.

 

Yea, I don’t mind to deviate some from my required head and flow but at least I expecting vendor can provide a rigid value for the flowrate and the head. So that I can confirm that value will satisfy my process demand.

 

For info, my colleague did attend the FAT and they said that, they can’t do the FAT using the proper standard way. They have to make it manually as the pump is small as per your say. I don’t know how they do that but for me, as long as the pump delivers at my required flowrate and head, should be enough.

 

Sorry for my story =)

Edited by farid.k, 23 March 2015 - 09:02 AM.

[Zauberberg]

I believe what the vendor tries to say is that they cannot guarantee a flow point until they know the discharge conditions at site (elevation difference, suction and receiver pressure, piping and equipment pressure drops, etc.). Like Fallah says, the same pump will perform differently (= will develop different flows) in two systems which have different resistances to flow.

[samayaraj]

I think, the vendor is looking for pump curve from you for his pump. Looking funny.

 

#Samayaraj

Edited by samayaraj, 23 March 2015 - 01:21 PM.

[Zauberberg]

More likely for the system curve. If the pump has steep H-Q curve, flow variations will be significant with changes in pressure drop through the circuit.

[farid.k]

More likely for the system curve. If the pump has steep H-Q curve, flow variations will be significant with changes in pressure drop through the circuit.

 

I also have no idea for that. But, I expecting when I install the pump, and run it, it will deliver very close to my required flowrate and head. Because that flowrate is my basis for my pressure drop calculation.

Just want to ask one more thing, let say I want higher head, i have to close the discharge valve say 30%. So the flowrate will be reduced and head will be higher isn’t?

 

If my statement is right, I just can’t imagine how by reducing the flowrate by closing some percent of discharge valve will cause more head by that pump?

 

The thing that plays in my mind is, when I close say 30% of the discharge valve, upstream pressure (head) yes will be increased but the downstream of valve will remain low due to high pressure drop at that valve. Am I right? If I wrong, can anybody help me to imagine by reducing the flow by closing some percent of discharge valve will really increase the head?

 

I tried some simple experiment in the toilet. I open the tap water, connect to the hose, and pinch the hose just to reduce the flow and the water travel further distance compared to if I not pinch the hose. Is this the same situation with the pump curve?

 

Thanks so much for your time

[Zauberberg]

What you are asking is simply an immanent characteristic of centrifugal pumps - they develop more head with less flow, and vice versa. This applies for almost every type of centrifugal pump, except for some pumps which develop less head at shut-off condition than at a certain non-zero flow.

 

Look at the attached image of a pump curve, and the system curve which depicts how pressure drop changes with the flow. It starts at zero coordinates (if static pressure difference is not shown) and it rises with flow. At the operating point, the pump curve and the system curve intersect each other.

 

If pressure drop of your system in reality is higher than calculated, the system curve will rotate towards the Y (head) axis. This means that the pump now has to develop more head OR the system pressure drop needs to be reduced. How? By opening the control valve more. But, if the hydraulic system is incorrectly calculated and you cannot achieve the required flow at fully open control valve, the pump will back off on its curve to the required head conditions, and the resultant flow will be less. In some cases, when the pump curve is flat, it might not be even able to produce any discharge flow. This is what in reality means system curve - the vendor cannot guarantee performance of the pump if you have miscalculated the system curve and the available pressure drop across the control valve.

 

Attached Files

•  intersection pump and system.jpg   18.5KB   4 downloads

[Zauberberg]

Have a look at the attached paper. Another good article from Joe Evans. You may wish to visit his website: http://www.pumped101.com/

 

Attached Files

•  How Curve Shape Can Influence Pump Selection.pdf   94.78KB   56 downloads

[breizh]

Farid ,

another great resource to clear your mind.

 

Instead of using valve you may also consider a variable speed drive (VSD ) .

 

Good luck.

 

Breizh

[Zauberberg]

Variable speed pump is probably the most effective solution if you expect significant variations in Head or Flow requirements. VSD also reduces the pump startup torque which is very significant design parameter, especially if we are talking about high power pumps. I suppose someone from the Electrical Department is looking into power demand for the electrical substation, particularly for the fixed-speed pump design scenario.

 

Let's imagine that you have calculated head loss in the circuit (without control valve) for 50 m³/h flow, and this head loss equals 55m. If you choose a 70m head for the pump (at 50 m³/h flow point), this gives you 15m available differential head across the control valve. I believe most design standards would agree with this approach, perhaps they would even recommend to reduce the available differential head across the control valve for a few meters. Some guidelines have been discussed recently at http://www.cheresour...c-calculations/

 

Now let's imagine that your hydraulic calculations have been done incorrectly, and that the actual head loss in the circuit is 68m (at 50 m³/h flow point). This leaves you with only 2m differential head across the control valve. It is very likely that the valve will not be able to pass 50 m³/h at 2m differential head, so the pump will be undersized and would be able to pass less than 50 m3/h. In extreme situations, if system pressure drop exceeds 70m, the pump which has the shut-off head of 70m will not be able to pump anything.

 

Read carefully the article at the referenced thread (it deals with control valve specification with regards to hydraulic resistances in the loop), as well as Joe Evans' article I posted in this thread. It gives excellent examples on what type of pump you should be looking for, depending on system (piping and equipment) head loss behavior.

[Zauberberg]

I believe this is another great thing to read for your particular subject.

Attached Files

•  Centrifugal pump specification and selection.zip   5.54MB   58 downloads

[farid.k]

Thanks so much for the technical support given.

 

I don’t think the using of VSD to effluent pump would be economical. This is just for effluent pumping. Not for any process that need varies of flow. I think using VSD would cost more money. Might be more expensive to the effluent pump it self