Calculating system curve with different pipe sizes 1

[Nutzman]

Good day all,
I'm sitting with a problem, in that I have a piping system in which the pipe system has different pipe diameters for different "parts" of the system. I have an 80mm line, branching off into a 25mm line, as well as a 32mm line. Both (branches)have different flow and head requirements. How do I "compile" then into a system curve?
All input will be welcome.

 

[LittleInch]

You just need to do it in sections.

If the 25mm line is a branch off the 80mm and the pressure at the branch point is dependant on other users then you have a problem.

You can simply bound the problem by assuming that at best, you are the only user of any common system or assume that everyone is using the common parts and hence pressure available at your branch will be lower.

So you then have two curves for your branch flow and reality will be somewhere in the middle.

Do you have a schematic of the whole system? If not you will need to draw one to work anything out.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Maybe you need some detail as to how to determine your system curve?

You will have to know, or assume the flow to each branch end point and the end pressure of each each branch. Then you can determine the pressure drop in each segment of pipeline, given the flow rate in the source pipeline and the flow going to each branch. You will have two overall totals, one from source to end of the 25mm branch and the other, the sum of pressure drops from source to the end of the 32mm branch. Select the greater of the two as your required design head for your source flow rate.

Do that same procedure for each possible flow rate going to each end of branch. Plot each of the equivalent heads of each pressure drop against each corresponding source flow rate. The resulting plot will be your system curve.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[Artisi]

1.0 petroleum:good answer, was just trying to get my head around the best way to present what you have said, now saves me the trouble.
Although I think each branch line will require some sort of regulation at entry to control flow and required pressure.

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.)

 

[Nutzman]

Good day LittleInch,
Lets see if my hand sketch makes sense.
Essentially 3 pumps in parallel.
Each pump can do 9.5 m^3/h @ 49.8m
Tyre wash requires 4.1 m^3/h
Sprinkler gate requires 2.7 m^3/h @ 40m
Hand hose use is intermittent flow = ??

Tyre wash flow goes into an open tank, so only piping lose's and elevation.

Hoping you can decipher my left handed sketch.

Thanks for your help.

 

[1503-44]

Sell 2 pumps on ebay. Buy a vfd and run at 1/2 +/_ speed.
Sell 3 pumps on ebay. Buy one with half the head at 5m3 capacity.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

I was going to say don't worry about the left hand thing - all good engineers are that way, but that is a bit of a scribble....

what is the strange labyrinth thing on the end of the 66mm line? A tank and then another pump? What controls the flow into the tank?

If you're pretty close to the pump discharge, the pressure won't vary much so you could take that as a constant or even negligible in the system curve calculation.

But are you trying to get a total system curve or what?

you can only get a worst case with everything on curve or a best case with only the hose on. Anything else you're somewhere in the middle.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Sorry LI, no good engineers write like that. Doctors yes.
Not much there that is actually readable.

Nuts, take my advice and ... call a plumber.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

What follows 0.00?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Nutzman]

Hi LI,

The "strange labyrinth thing", is a settling tank (6m container), that feeds the pump for the tyre wash. The (3X) pump that feed the settling tank, is 30m away and an open ended pipe. Thereby creating no resistance against the pump. Whereby the pipeline along the bottom of the page is 80+175m long and makes up the majority of the back pressure. But being a centrifugal pump, the water will seek the path of least resistance. So the pump falls off the curve feeding the settling tank and a fraction goes to the sprinkler at the gate. Other than putting in an orifice before the settling tank to balance the system, I don't see how I can "develop" a system curve showing all the variables?

 

[1503-44]

OK LI. Tell him again how I told him to make the system curve. I think Artist gave me a star for that, so it must have been a pretty good instruction. He doesn't give stars for nothing.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[Artisi]

Nutzman, you're correct, as the system is open ended with out any flow or pressure control, the calculations /permeations are virtually unknown, certainly beyond what my old brain can handle now-a-days.

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.)

 

[1503-44]

You have to control outflow, or try to operate a run-away pump.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[Nutzman]

Perhaps a couple of variable orifices (with there handles removed), would assist in balancing the system?

 

[1503-44]

Perhaps yes, but orifices have different pressure drop effects at different flows, so leave the handles on.

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[Artisi]

Or simply fit an el-cheapo gate valve at each outlet and enjoy the fun balancing the system.

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.)

 

[1503-44]

Remember that Dutch boy method?

Reality used to affect the way we thought. Now we somehow believe that what we think affects reality.

 

[LittleInch]

OK,

First draw a PROPER diagram listing all the flows and pressures, lengths etc

Then divide your system into sections between nodes where you have branches.

All you know for sure is that flow into anode = flow out of its branches and the pressure at the node is the same for all branches.

Then work out the combinations you want to generate a curve for.

Then work out for the different sections what the flow and pressure are for that section based on elevation change and friction.

Then make up a spread sheet and start altering numbers so that for regular flow changes, you can work out the head.

For your system because for some strange reason you have an open ended pipe, once the head at the first node gets to 30m, then most of the flow will go that way.

Looking at it from on high, your pumps will run off the end of the curve and you'll get a lot more flow than 4 m3/hr and your tank will overflow.

Why are you doing this?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Nutzman]

HI LI,

It's a project I "inherited". I suspect it's original design has no resemblance to what it is today? I think it is a add on, to an add on? I'm fully aware of the fact that their is "no" system resistance to the settling pond and that the pump see's that as path of least resistance and will run off the curve and besides tripping the 2.2 kW motor, will eventually lead to mechanical failure.
Thank you for the pointers on how to balance the system. But I am under time constraints to come up with a solution, combined with a shoe string budget.

 

[LittleInch]

You could try this little device which is reported as working well for a fixed flow within 10 to 15% regardless of pressure drop

https://maric.com/

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.