Overload problem happened on boiler feed water pump 2

[SalvationTsai]

Greetings, and good day to every dear seniors, I found a overload problem occurred on a boiler feed water pump, but I am wondering why it will overload so much, and want to know whether my viewpoints are wrong or right, I will do my best to describe the every factors and situation in this system, and please ask any information I missed to mention, greatly thank you in advance.

Boiler: testing pressure 10kg/cm^2, normal operating pressure 5~6.5kg/cm^2,
hot water consumption rate 2.0m^3/hr

Working fluid is normal hot water, 80 Celcius.

Feed water pump: total head 120m, rated flow 4.8m^3/hr, NPSHr=5m

Motor: 7.5KW, 2 poles, 3600rpm
(but according to pump’s shaft power formula, I assume the efficiency of the pump is 0.7, motor efficiency is 0.7, then the overall rated power is approximately 3.2KW, so I don’t know why they use a over-rated motor to drive the pump, and this still makes the motor overload!)

The inlet pressure gauge reading and outlet pressure gauge reading is as follow---
Pressure-controlled valve open, inlet=1.7kg/cm^2, outlet=7.9kg/cm^2 <total head is 79m-17m=62m>

Pressure-controlled valve closed (shutoff head), inlet=1.7kg/cm^2, outlet=9.2kg/cm^2 <total head is 92m-17m=75m>

The pump will have a 10KW measured by power-gauge when operate under the pressure-controlled valve is closed; but according to the teach book, the power in the shutoff condition should by only higher than the pump operates in no-load condition, really don’t understand why it will overload (higher than the 7.5KW).

The pump will trigger the overload protection (higher than the 10KW), and stop the motor, when operate under the pressure-controlled valve is opened;

The calculated NPSHa is approximately 6.5m, is bigger than the NPSHr=5m (under rated flow and head), but since the actual head is almost half to the rated head, and then the actual flow should be a approximately double than the rated flow, thus makes the NPSHr goes higher, result in the cavitation happened.

I know the cavitation will make the inlet casing ring worn, thus let the pump efficiency goes lower, but could it exactly be the main factor to cause such extremely overload?

Thank you very much for reading my lengthy post, I sincerely thank you! And hope I could have a better direction to think on this problem!

 

[LittleInch]

Salvation - You again! I'm assuming this is a continuation of your previous posts. when you say Feed water total head 120m, what exactly do you mean? If this is the rated duy of the pump and you have a pump curve, please post it. If not please explina where this number comes from.

I suspect that what you have here is a pump running much more flow than it should as you are controlling on pressure instead of flow.

However your data is conflicting - 120m total head, but shut in flow head is only 75m?? There is something very wrong there.

Inlet pressure running of 1.7bar, but NPSHa of only 6.5m??

Th fact you seem to be cavitating is also leading me to the initial thoughts that you are actually flowing much faser than 4.8 m3/hr. Do you have any way of measuriung flow?

It looks like you need to control on flow to reduce power. Try going from a shut in condition then monitor power as you very slowly open the control valve. Measure flow if posisble and inlet and outlet presusres

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[SalvationTsai]

Dear good LittleInch,
it is so happy to have your teaching again! And yes this is indeed the continued studying on the boiler feed water pump.
The total head 120m is a rated head on the name plate of the feed water pump, and I don’t have the pump curve, because the person who responsible for this pump said the manufacturer didn’t provide it to him.

I am very agree on your good thinking that the flow is much more than it actually needed, but could this be the reason why it overloaded?

And I am very sorry and embarrassed for my bad data of the inlet pressure gauge reading, it should be 0.17Kg/cm^2 (I made a unit confusing between 0.17KG/cm^2 and the 1.7m)

Thus I have to correct on my total head derived by the reading of inlet gauge and outlet gauge.

Pressure-controlled valve open, inlet=0.17kg/cm^2, outlet=7.9kg/cm^2 <total head is 79m-1.7m=77.3m>

Pressure-controlled valve closed (shutoff head), inlet=0.17kg/cm^2, outlet=9.2kg/cm^2 <total head is 92m-1.7m=90.3m>

And I forget a big factor when under this situation, it is because I had installed a VFD, the frequency will be 44~47Hz when the pressure-controlled is valve closed, and doing 60Hz when the pressure-controlled valve is opened.

And I also have to correct that the pump will have a 10KW measured by power-gauge when operate under the pressure-controlled valve is closed (when frequency is 60Hz); 5.5KW measured by power-gauge when operate under the pressure-controlled valve is closed (when frequency is 44~47Hz);

And I have don’t have the tool to measure the flow, really sorry for this, but I will try to have one if I could and use the way you kindly provide to me, using the valve to control the flow and observing how the power will vary.

But in this case I am using VFD to adjust the head and flow, trying to match the actual head and flow, but still it keeps overloaded.

And wondering why it use a big motor(7.5kW) to drive a relevantly small pump(calculated rate power 3.2kW), won’t it doing bad(like wasting energy) to the whole system?

Thank you for your patient and gentle good teaching on me, really and vary gratitude to dear good you.

Best and Warmest regard.

 

[Artisi]

SalvationTsai;
Might be about time you stopped wasting your bosses time, admitted you are well out of your depths and get them to employ someone who knows what there are doing. A boiler-feed pump of this size shouldn't cause the slightest problem to any 'ENGINEER' with even the most basic understanding of pump hydraulics.
Are you writing a text book or a thesis, there has been nearly enough information forwarded to undertake a doctoral thesis on the subject.

One reason for the oversized motor could well be a based on the cooling needs of the motor running down to 44 Hz - which subsequently means that the cooling effect from the motor fan will be very low.

As littleInch has pointed out, too much conflicting and contradicted data most of which looks to be meaningless, you need to supply hard facts, inlet pressure, discharge pressure, flowrate, power input at the operating pump speed and a manufacturers pump curve, otherwise we are guessing just as you are as to what is really happening.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Salvation, before I answer your questions above, can you either photograph the name plate or list everything it says on it, manufacturer, serial no, type of pump, head, flow, speed, everything.

I think you are running this pump at about 2 to 3 times more flow than it was intended for. As you increase flow past the duty point the efficiency gradually goes down and hence power required climbs faster than the flow rate increase.

What you need to do is operate the pump at max speed but open control valve very slowly to keep a pressure of 12bar on the pump discharge then measure power.

You seem to have a lot of power losses when the pump is at no flow and min speed so something strange is going on. How is power worked out? What is voltage and current used?

A big motor is sometimes used just because it is the next size up from required duty, but motors only take whatever power they need, so a7.5kw motor with a3.2kW load will only use 3.2kW divided by it's efficiency, usually >90%.

Get the pump and motor plate details and we might be able to get somewhere.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[SalvationTsai]

To dear good Artisi:
Thank you for your kind suggestion and advice to me, and I am a voluntary helper and student on this case, trying to know more, but indeed I am very poor on the pump studying, but I do try my best to self-study on this domain, and sincerely apologize to you for my novice idea and poor data I collected.

And I am very happy to further know that I really forgot is the cooling problem when operating in a decreased frequency situation, I will try to focus on this issue!

And I am sorry for my negligence on some data I post at the beginning, really embarrassing to everyone.

Inlet and discharge pressure gauge reading when doing 60Hz, pressure-controlled valve opened, is 0.17kg/cm^2(inlet), 7.9kg/cm^2(discharge).

Inlet and discharge pressure gauge reading when doing 44~47Hz, pressure-controlled valve closed, is 0.17kg/cm^2(inlet), 9.2kg/cm^2(discharge).

Flow rate, I only know the rated flow rate is 4.8m^3/hr, don’t have the method to measure the actual flow rate.

Power input when doing 60Hz (pressure-controlled valve opened) is more than 10KW, and sometimes triggered the protection and force the motor to stop.

Power input when doing 44~47Hz (pressure-controlled valve closed) is about 5.5KW.

But don’t have the manufacturer’s pump curve, really sorry for this.

Thank you for your good and kind teaching,
Best and Warmest Regards.

 

[SalvationTsai]

To dear good LittleInch:
Really thank you for your kind reply and good suggestion, I will try my best to obtain the information you mentioned.

And thank you for teaching me that the power will increase dramatically higher when past the duty curve (power required climbs faster than the flow rate increase), I do forgot the efficiency will decrease when out of the duty point.

According to your suggestion, pump at max speed, then if I slowly controlled the valve and makes it at 12kg/cm^2, then I should have the pump’s rated power consumption in theory, am I properly understand it?

And yes, I really feel the strange when pump is at no flow, but still consumes big power, and I will try to obtain the data you mentioned.

Really and very thank you for every good teaching you patiently told me, they are indeed and extremely helpful!

Sincerely thank you and with my Best Warmest Regards.

 

[LittleInch]

Salvation - I will await further information with interest.

In theory, at max speed you should have rated power at 12 bar with the flow that matches the pump, but your power at shut in head (5.5 kW) worries me that there is something else we haven't considered - this maybe due to the VSD - was the VFD a retro fit (put on later) or part of the original supply?

Also if you've been operating this pump for any time like this and having cavitation, your impellor will be damaged and the efficiency much reduced.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[ScottyUK]

VSDS on high head pumps are usually not worth the effort because so much Power goes into developing pressure rather than moving fluid. This sounds like such an application.

 

[1gibson]

Interesting discussion I guess, but high power at shutoff ("pressure control valve closed") means something is rubbing (bearings, wear rings) or packing is adjusted too tight (if applicable.)

 

[SalvationTsai]

To dear LittleInch and dear ScottyUK:

Thank you for your hospital help and generous teaching, It is my fortune to have it.

First photo is the pump motor, it was badly worn on the very left part, and I am hoping this could still provide any useful information.

Second photo is pump's name plate, kindly for your reference.

The power (measured by VFD) when pressure-controlled valve closed (shutoff head) is 5.45KW, current is 10.7A

The power (measured by VFD) when pressure-controlled valve opened is 11.49KW, current is 17.8A.

The VFD is a retro fit, which I want to decrease the frequency in order to match the actual head & flow for better performance and solve the overload problem.

Thank you for your kind instruction and help, sincerely thank you and with my best regards.
And please ask any information I missed to mention, thank you in advance.

http://i.imgur.com/7a4Pwd7.jpg[/IMG]]

http://i.imgur.com/AXRfNSE.jpg[/IMG]]

 

[LittleInch]

Salvation, I thought you had gone on holiday...

below link is your pump curve, but with only poor data. At 70 m diff head you are looking at a flow of nearly 15m3/hr. If you take an efficiency of 0.3 and a motor efficiency of 0.85 including vfd losses then you get 12kw.

You are flowing well beyond the end of the curve. See

http://www.shinkohir.co.jp/pump-svq/svq-ape.htm

What you need to do is run the pump at min speed and slowly open the valve MANUALLY watching the amps on the motor. Your pump has far too high a head for the duty so you need to control on either flow or discharge head of the pump (120m differential) NOT the pressure of the boiler. Your system is just totally wrong in the way it has been designed.

If you want to think of this as an electrical system designed for 12 V, but someone has put a 24V truck battery in the circuit. What happens if you don't put a variable resistor (control valve) in the circuit –the headlights burn really bright then blow up..... Your pump will cease to exist quite soon unless you follow the advice given here, but may already be too late...

Good luck and let us know how you get on.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Artisi]

Five things,
1. speed correct the curve - the curve is drawn at 3600rpm and the motor name plate gives a motor speed of 3490rpm.
2. The flow rate shown on the pump name plate of 4.8 m3/h is not the flowrate for your system, it is probably the flowrate at the pumps best efficiency point.
3. now that LittleInch has done your job and went to the internet to find the curve that should have been the first thing that you did things are now a little clearer.
4. The pump is probably operating just as it should, it delivers the flow at the required input power to meet the head imposed on it.

and finally - you are lucky be have been given all this advise, as a student, which you advised a couple of posts back you shouldn't be posting in Eng-Tips - please read the conditions for posting in Eng-Tips - no student postings. But what is done is done, so get on with sorting out your problem now you are aware of what is going on.


It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[LittleInch]

Salvation,

Also consider that a retrofit VFD needs to ensure that the motor is Ok for it. Fixed speed motors may either overheat due to lack of cooling flow as noted by artisi or suffer from current leakage and issues with bearings. My guess is that the VFD was just added without checking with the motor vendor if this was Ok - You could easily burn out the motor quite quickly.

For your info I attach a pretty basic flow diagram and information that you need to find out. Flow diagrams are normally drawn left to right - don't ask me why that's just the way it is...

What I'm a bit puzzled about is how this thing operates. You quote a boiler feed supply required of 2 m3/hr. From the pump curve I think its doing about 14 m3/hr. Does the pump start and stop a lot?


My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[ScottyUK]

Worthy of note: it's a Class B motor - Class B windings are not typically wound with materials which are suited to VFD use. The materials used for Class H machines and some Class F machines are more tolerant of the output waveform of a VFD although many manufacturers have types which are specifically designed to withstand the fast risetime switching transitions found in VFD outputs.

 

[SalvationTsai]

To dear good LittleInch:

Sorry for my late reply (was trying to organize it more clearly), and many thank you for your every teachings.

First I assume I am applying 46Hz (by VFD), and thus I should have a new lowered pump curve.
Using the pump affinity law, H is proportional to the square of Hz, then applying rated 60Hz has a rated head=120m, I could derived a new head=70m when Hz=46.

The below is my concept drawing on the new curve, and with some questions.

In my mind, I should have a better pump efficiency and a much lowered power consumption (due to P is proportion to cubic of Hz); but I don’t really clear how to find the pump efficiency when at H=70m, Hz=46, is it because the manufacturer didn't provide the efficiency curve on the H-Q chart?
And I am not quite understand why dear good you could know the pump efficiency is 0.3, when pump is 60Hz, actual head=70m? Is it a assumption value? I know it is because the pump efficiency will be dramatically low when fall out to the end of the curve, but I don’t know how to find the good proper value of it.

And I will have a Q=5m^3/hr actual flow when I am operated in 46Hz, actual H=70m, and thus will I have a bit increased NPSHr compared to the NPSHr which is operated in 60Hz, rated Q=4.8m^3/hr? From my poor knowledge I know the NPSHr is proportion (but not linear) to Q. And another question is how could I know the rated NPSHr in this case (I am originally be told that it is 5m)? I know it should be provided within the H-Q chart, but I don’t see it from the chart or data from the web side, did I miss it?

I will try to run the pump at minimum speed(but in safety the lowest will be 30Hz), and monitoring the current vary when I slowly manually open the valve, but truth to speak I don’t have the chances or rights to operate it, because I am only a voluntary helper, very novice and naive on this case, but I think if I propose a decent concept to the person in charge of it, then I think I will probably could have a chance to try on this.

And your clear and brilliant explanation (using the 12V and 24V) makes the whole problem better to understand and the resulting consequences, when the pump is ill designed to the system.

And yes, very thanks to ScottyUK’s vivid explanation for the effect when VFD is installed on Class B, H or some Class F.

I am very thank you for your kind teaching (attachment), I will read it over and over again.
And the application after we retro fit the VFD, will be operate the pump in 44~47Hz when boiler don’t need the water (boiler pressure higher than 6.5kg/cm^2, makes the valve closed), and operate the pump in 60Hz when boiler needs the water (boiler pressure lower than 6kg/cm^2, makes the valve opened), making the pump is standby in 40~47Hz, and start to provide the hot water in 60Hz, so basically it is standby-and-start, a bit different from yours stop-and-start, but I think yours concept is much more better (save more energy when stop it, compared to standby).


To dear good Artisi:

1.So I think the curve in 3490rpm should be drag lower than the curve in 3600rpm, could my assumption correct?
2.Yes I know the 4.8 m^3/hr is the rated Q when in rated H, and not the Q in this case.
3.I am really thankful for knowing that I could find the H-Q curve in Internet, without dear good LittleInch, I won’t know this way to obtain this vital information.
4.Just like dear good LittleInch said, I am now further known (clearly) how and why the overload really happened, the bad efficiency is the critical factor for it.
5.I really feel bad for thinking me like I am a cheater, I am indeed a student, but this is not any report or homework; the professor happens to have this pump overloading case, and ordered a PhD student to deal with this, and want me to help carrying the tools and running errands in order to retro fit it, I don’t responsible for this overloading case, but I am interesting on it, and want to know more on pump engineering. My school doesn't have any class teaching the pump knowledge, so I could only self studying and make question on this great forum, and many thanks to everyone who helped me, I indeed received a lot of correct concepts in here, this kind of joy is beyond words.

 

[LittleInch]

Salvation,

As with a few of these posts we gradually get more information, but you still need to get more to determine what exactly is happening. What you need to do is observe the operation of this pump for at least one hour and much more if possible noting every 5 minutes or when something changes and note the following minimum.
time, motor speed or htz, motor temperature (if known), electrical power in, inlet pressure, outlet pressure, %open of the CV (from control system or slider on the pump), inlet pressure to boiler, change of operation or operator action.

Only then will you be able to show your tutor what is happening and why.

Back to your questions:

Pump efficiency I estimated based on experience of what it is at extreme end of curve. It should be supplied by the pump manufacturer, but in this case the pump curve is very basic and doesn't have NPSH, power, efficiency etc which it should have. Try e-mailing the company with the pump serial number requesting a full pump curve with this data. They may send one back, but if you don't ask you won't know. It should have been supplied with the pump but maybe has become lost.

NPSHR will be higher due to lower speed, but difficult to calculate - maybe 2 to 3 more m, but nothing much more than that.

You will need to run the pump at about 46 to 50 htz to get enough head to get some flow - 30 is much too low.

Without the data listed above it is difficult to say exactly what is going on, but it sounds as if the operation is being very bad for the pump - if the pump is "turned down" when the boiler pressure is >6.5 bar and the control valve is closed, then this is very bad for a pump to have no flow for any length of time and will damage the pump over time. If the pump is then "turned up" when the pressure is below 6 bar and the control valve opened then it will flow water in 3 times more than the pump capacity - again very bad for the pump. What you should be aiming for is a more or less constant flow or turn the pump off when you don't need it or at least have 1-2 m3/hr going through the pump as a re-cycle back to the inlet tank. Also for this pump you probably shouldn't run it much more than about 50 htz or you will flow far too much through the pump and make it overload - Always remember your pump at 60 htz is like the 24 V battery when you only need 12...

Think of the no flow operation like a car without a cooling fan for the radiator. What you are doing is driving it at 100 MPH, then stopping dead, but still with the engine running, what happens - the engine overheats very quickly and the radiator boils. What the car (and the pump) needs is to keep moving so that cooling air come into the radiator - same thing with a pump. Either that or turn off the engine if you're stuck in a traffic jam (turn off the pump if no flow). Or fit a cooling fan (minimum flow re-cycle).

See if you can observe the operation for long enough to make sense of what happens then you can show what happens if you run it differently and hopefully solve your problem.

I hope this helps you understand the basic principles.

Good luck and let us know how you get on.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[SalvationTsai]

To Dear good LittleInch:

I am thankful for your continuing hospitable teaching, and I am sorry for my very late reply, was trying to sum up the concepts and the solutions, and also wait for the more detailed pump data from the manufacturer; the manufacturer sent me the data promptly, but the e-mail goes to the junk box thus makes me hard to find it until the yesterday.

I humbly propose the following two solutions, and hope you could give me any advice.

Solution 1:
According to the condition for the boiler valve is opened (pressure lower than 5.0kg/cm^2) or closed (pressure higher than 6.5kg/cm^2), I assumed a commanding sequence into the VFD, in order to control the boiler hot water feed pump.

My pump will turn-up(power-on), running at approximately 48~50Hz, when the boiler pressure is at 5.2kg/cm^2, before the valve is going to open; the reason why I turn-up the pump prior to the valve’s opening, it is because I think this could make some kind of warming up to the pump, and prevent the water from the boiler flow back through the pump to the hot water storage tank.
And after the boiler pressure is at 6.5kg/cm^2, I will let the pump to run 15~30 more seconds, in order to make sure the valve is indeed closed, then shut-down(power-off) the pump (like the dear good you said “Either that or turn off the engine if you're stuck in a traffic jam”).

Solution 2:
Will it help if I add one recirculation loop(hot water tank--->pump--->hot water tank), in order to have the water flow to cool down the pump(like dear good you said “to keep moving so that cooling air come into the radiator”)? But later I think it is quite wasting the energy, thus make solution 2 a bad idea but could still poorly solve the problem?

The following is the more detailed data from the manufacturer, but the pump efficiency is very poor, even in the best operating point, is it normal? And still no data of the NPSHr.

And I will first let dear good you know if I got all the testing data which you taught me to collect, but this case is in a tanker ship, and it is scheduled to return in months, and also hoping they could let me board to the ship and begin my collecting the relevant data!

Thank you truly and with my best and warmest regards to dear good you.

 

[LittleInch]

Salvation,

Congratulations on getting the data - I hope you see how much easier it is when you get all the required information. At least this answers a few questions we had earlier, which is:
Why do you need a 7.5kW motor - answer, because the pump is very inefficient ( the motor isn't great either). To have an efficieny of 25% at duty point is very low. I would normally assume 70% for a decent sized pum or maybe 60% for smaller units so 25% is quite low.

Why do you still use 5.5 kW at no flow - answer, because the pump is very very inefficient at low / no flows and even at a lower speed, you're loosing a lot of power going nowhere.

I think I have mis understood one of your comments when you said "pressure controlled valve", I read this as "pressure control" valve. I think now that the valve is basically an on / off valve, not a control valve. You really need a control valve on this line somewhere to provide flow and presusre control, but don't seem to have one??

Your solution 1 looks good to me - You might need to adjust the speed control until the pump produces enough flow, but not enough to trip the over load protector, i.e. monitor the power going in as well. This is still a poor design, but would solve your original problem.

Your solution 2 you are correct this would waste energy, but probably no more than it is doing at the moment if the pump stays on all the time, but at low speed. The issue will be how many times solution 1 would need to start and stop in an hour. More than about 4-5 starts/ hour will proabably overheat either the VFD or the motor as there are some effects on start which mean the motor needs to run for a set period to cool down. Therefore if there are lots of starts you need to keep the pump running on minimum flow (about 1.5 m3/hr) to avoid overheating of the motor or the VFD unit.

For your information, centrifugal pumps are essentially fixed pressure units, give or take 10%, but variable flow. If the system curve or friction losses / pressure it sees at the discharge changes a lot, you either need to add a control valve to provide additional frictional resistance to maintain pressure and restrict flow to the pump maximum (which your system doesn't seem to have) or reduce the speed - which you do have, but are not controlling it on any process inputs.

Positive displacement pumps (piston, diaphragm, screw etc) are constant volume pumps which can provide variable pressure. Varying the flow requires either bypass back to the inlet or variable speed pumps. Your duty is more suited to a PD pump, but I guess you're stuck with what you have...

I just hope by the time the tanker comes back there is enough of a pump left to do some tests or modifications on.

I hope this has answered your questions, but come back if you need more. Good luck.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[Artisi]

For the size of pump etc, the efficiency is probably reasonable.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)