An argument with a colleague on the subject of pump head measurement in the open-loop (condenser pump) that suction side friction loss from source will not be calculated in the condition where the NPSH is positive or NPSHA > NPSHR. My opinion is that the all side friction loss put in the calculation. what is the technical or fundamental point of view?? Any help will be much appreciated!
Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
suction side friction loss 5
- Thread starter moideen
- Start date
You mean if suction friction loss is included in the calculation of NPSHA?
Can you elaborate on why your colleague would omit it? I've never heard of that and I don't omit the term. One _could_ argue, that omitting it would yield a more conservative result but honestly I don't see the point in that.
Can you elaborate on why your colleague would omit it? I've never heard of that and I don't omit the term. One _could_ argue, that omitting it would yield a more conservative result but honestly I don't see the point in that.
- Thread starter
- #3
-
1
- #4
LittleInch
Petroleum
Well he might have a point, but only if the inlet side friction loss is say 5% or less of the discharge head.
but you still need to work out what the inlet head is, add the pump differential to work out the discharge head.
If the pipe is very short, then the majority of the inlet head will be the static head. The issue is how much higher than NPSHR do you have?
It's not good practice to ignore losses when the basis for ignoring them can change.
A diagram and some data would help us here.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
but you still need to work out what the inlet head is, add the pump differential to work out the discharge head.
If the pipe is very short, then the majority of the inlet head will be the static head. The issue is how much higher than NPSHR do you have?
It's not good practice to ignore losses when the basis for ignoring them can change.
A diagram and some data would help us here.
Remember - More details = better answers
Also: If you get a response it's polite to respond to it.
The pump total head csn only be established with calculating the head from the inlet to the discharge point, that includes all friction losses added to the static head. NPSHa / NPSHr is a seperate calculation.
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.)
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.)
It sounds like he is simply ignoring any contribution to total head/pressure from NPSHa. Additionally he is somehow sure that NPSHa is greater than NPSHr. In which case it is a matter of if including NPSHa makes any significant difference to discharge head/pressure or has any significant effect on flow. If it doesn't ammount to any significant difference in pressure or flow, then its a judgement call to not waste time in actually calculating suction pressure at the pump inlet.
-
2
- #7
The statement that friction losses can be ignored on the suction side of NPSHa>NPSHr is false. Imagine a pressurized tank at 50 psig (static head + tank pressure), and line losses of 40 psig to the pump suction. The pump discharge head is 100 psig. NPSHr may be, say, 25 feet of water absolute. Though there is adequate NPSHa, it would be foolish to ignore the 40 psig in line losses to the pump because it accounts for a large percentage of the overall system pressure.
That is a bit of an extreme example, but it shows that his assumption can lead to serious oversights.
Many pumps and suction lines are designed to minimize suction losses, so he may be right, sometimes, within an acceptable error tolerance. However, he may also be waaay off, depending on conditions. Thus, ignoring line losses if NPSHa > NPSHr is bad practice. Lazy, too.
That is a bit of an extreme example, but it shows that his assumption can lead to serious oversights.
Many pumps and suction lines are designed to minimize suction losses, so he may be right, sometimes, within an acceptable error tolerance. However, he may also be waaay off, depending on conditions. Thus, ignoring line losses if NPSHa > NPSHr is bad practice. Lazy, too.
No, NPSHa is that head available at pump inlet after subtracting all suction line losses (your 40psi) , thus if you are sure somehow that NPSHa > NPSHr, it is so. NPSHa = 50+14.73-40-1 psi vapor pressure = 23.73 psia => 56.7 ft (of water), which is probably going to be more than enough suction head for your extreme example pump.
Yes, that's true. It would be rather (shall we say extremely) uneconomic (shall we say foolish) to design a 40psi pressure drop in suction to feed only a 100 psi pump discharge, but if discharge was 1000 psig, you're talking less than 4%, which might be an extreme condition caused by a small probability malfunctioning control valve snapping to minimum open limit, or simply not a bad safety factor, if you think you need one, which maybe you might want to have if you didn't actually do the calc. Its a thing of relative value and the OP did not give us anything we can relate to in that manner. So if you're sure that NPSHa > NPSHr ... you don't have to prove anything else. Of course you might not have an optimum design there, but maybe all you need is a pump cost estimate. There is little point to calculate things, or to a level of accuracy, more than you need to solve the immediate problem at hand.
Yes, that's true. It would be rather (shall we say extremely) uneconomic (shall we say foolish) to design a 40psi pressure drop in suction to feed only a 100 psi pump discharge, but if discharge was 1000 psig, you're talking less than 4%, which might be an extreme condition caused by a small probability malfunctioning control valve snapping to minimum open limit, or simply not a bad safety factor, if you think you need one, which maybe you might want to have if you didn't actually do the calc. Its a thing of relative value and the OP did not give us anything we can relate to in that manner. So if you're sure that NPSHa > NPSHr ... you don't have to prove anything else. Of course you might not have an optimum design there, but maybe all you need is a pump cost estimate. There is little point to calculate things, or to a level of accuracy, more than you need to solve the immediate problem at hand.
- Thread starter
- #9
Thanks for all the timely responses. Our dispute began in the face of a new two retrofit project. In one project, the cooling tower and pump very close, and in the 2nd project, the cooling tower placed rooftop and plant room at the ground level. 30-floor building and vertical height is around 100 meter.in the 2nd project, the cooling tower leaving water direct go to the condenser then come to the pump inlet (cover the condenser drop also) in both case, NPSHA is strong, higher than NPSHR. But, his argument is that head calculation to be started from the pump outlet only. He doesn’t look at the pump suction side especially in the 2nd project 100-meter suction pipe frictional loss. He simply ignores the suction side loss being the NPSH is sufficient. my experience totally disagrees with him.
as Artisi said, NPSH is a separate calculation and it is subjected to the only suction side design
thanks
Moideen
as Artisi said, NPSH is a separate calculation and it is subjected to the only suction side design
thanks
Moideen
Moideen: correct, the pump performance in terms of flow and head is only known by calculating all the friction losses plus the static head from the inlet to the discharge point - once this is known the pump duty point (from the pump performance curve)gives the NPSHr for that particular condition - otherwise it's guess work.
In your case the friction loss for 100m of inlet pipe should be established.
Calculate the total head across the pump from inlet to outlet, establish the flowrate at this total head, see what the NPSHr is at this duty and then establish what the NPSHa will be. It's not rocket science and usually straight forward unless you have a very convoluted pipe system, elevated temperature or other points that might influence the calculation.
Forget all the academic round and round, just concentrate on the practical aspect of the installation.
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.)
In your case the friction loss for 100m of inlet pipe should be established.
Calculate the total head across the pump from inlet to outlet, establish the flowrate at this total head, see what the NPSHr is at this duty and then establish what the NPSHa will be. It's not rocket science and usually straight forward unless you have a very convoluted pipe system, elevated temperature or other points that might influence the calculation.
Forget all the academic round and round, just concentrate on the practical aspect of the installation.
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.)
Again I will disagree a little bit. In a closed loop system where suction is a vertical downpipe from the 30th floor, it is very unlikely to be a problem. A flat closed loop system is more problematic, so doing the calc is probably more necessary to acctually insure NPSHr is satisfied. In my experience a 30 story building with a suction pipe size selected by an engineer with experience is not going make trouble. If you don't know, do the calc. If you trust the other guy's experience, don't waste time arguing about 30th floor suction line. Ask how the loop will be filled.
-
2
- #12
Compositepro
Chemical
This is an open loop not closed. Not that it matters except to show a lack of attention to detail.
The operating point of the system is determined by where the pump curve intersects the system curve. The fact that NPSHA>NPSHR means that the suction losses do not affect the pump curve. But the suction losses are absolutely a part of the system curve just like the friction losses in the discharge piping. You cannot ignore the suction losses for any flow calculations. It makes no difference if the friction loss is on the suction side or the discharge side.
The logical error here appears to be that someone thinks that pump head is how high the pump will pump liquid, which only true under specific conditions. What is true under any conditions (where NPSHR is met) is that the pump head is the difference in pressure between the suction and the discharge of the pump. Any pressure losses on the suction side of the pump will result in exactly the same pressure reduction on the discharge side.
The operating point of the system is determined by where the pump curve intersects the system curve. The fact that NPSHA>NPSHR means that the suction losses do not affect the pump curve. But the suction losses are absolutely a part of the system curve just like the friction losses in the discharge piping. You cannot ignore the suction losses for any flow calculations. It makes no difference if the friction loss is on the suction side or the discharge side.
The logical error here appears to be that someone thinks that pump head is how high the pump will pump liquid, which only true under specific conditions. What is true under any conditions (where NPSHR is met) is that the pump head is the difference in pressure between the suction and the discharge of the pump. Any pressure losses on the suction side of the pump will result in exactly the same pressure reduction on the discharge side.
No arguments from me about that. Entirely true. My only contention is, if something is small enough to ignore it, no matter what that might be, then ignore it and get on with next business. The longer story behind that, if you're interested is...During my very first performance review as an EIT, I got rung up for wasting too much time running numbers to see if I needed to use a W6X15 beam, or if I needed to use a W8X17. My bosses answer to which was, I spent more money running the stupid calculation than I saved in steel cost and I should have selected a W10X21, even though I needed 1000 of them, because that was an unwritten plant standard to allow quick and easy fabrications using the unofficially standard jigs, clip angle cutters and bolt hole locations all similar and get the refinery up and running one day earlier. I had no real argument to that then, or now. That is my one and only point. If you want to be a bean counter, than count beans. Engineers should also consider multiple constraints in their solutions, one very important real life constraint is "the quicker the better". But obviously not in the case of 737MAX.
Yes it is open, ... not that that matters.
Yes it is open, ... not that that matters.
-Fortyseven
You are misunderstanding OP’s question:
his point of view that the suction friction head loss will not add in the total head of calculation if the suction side meets above NPSHR. So the calculation will start from pumps discharge to the cooling tower nozzle drop
This is OP’s clarification on his colleague’s statement. It also happens to be 100% wrong. Suction friction ALWAYS is considered in the total head calculation.
There may be cases in which the error margin from his colleague’s method, though incorrect, had an acceptable margin of error. That, however, is not OP’s question. OP is asking for the technically correct approach.
Also, knowing that NPSHa > NPSHr means that he has already calculated suction side losses. I don’t understand why he wouldn’t include that in the final head loss calc if he already has the suction side losses calculated. Hence my “Lazy” description.
You are misunderstanding OP’s question:
his point of view that the suction friction head loss will not add in the total head of calculation if the suction side meets above NPSHR. So the calculation will start from pumps discharge to the cooling tower nozzle drop
This is OP’s clarification on his colleague’s statement. It also happens to be 100% wrong. Suction friction ALWAYS is considered in the total head calculation.
There may be cases in which the error margin from his colleague’s method, though incorrect, had an acceptable margin of error. That, however, is not OP’s question. OP is asking for the technically correct approach.
Also, knowing that NPSHa > NPSHr means that he has already calculated suction side losses. I don’t understand why he wouldn’t include that in the final head loss calc if he already has the suction side losses calculated. Hence my “Lazy” description.
Yes, I believe you are right. Thank you. I did misunderstand the question. In which case (30 floor downcomer) I find it convenient to use a reference point of the lowest posible liquid level in the suction downcomer when the pump is stopped. You can check all suction loses from that point as the pump starts and goes from 0 flow to 100% flow, noting any change of liquid level at the ref point as might occur. Sorry for going into the weeds.
- Thread starter
- #16
- Status
Similar threads
- Locked
- Question
