Fluid Mechanics-head loss/GPM change adding 2nd line for pumping

[saxman2u]

I have a rainwater collection system in Austin, TX, where we have to pump water to our main 30,000 gallon tank. The system works great. My question is, our 3 HP pump pumps about 85 gallons per minute. The length of the 2"pump line is about 300 ft, elevation change is about 50 feet higher. Throw in about 12 elbows and you have your head loss, it is around 90 feet of head according the pump curve sheet.

My question is, when i installed the system years ago, i stubbed in another 2" line that is currently not being used. Would this decrease my current feet of head? If so, by how much? Just curious, I am wondering if it's worth the time and energy to plumb in this line as well. Occasionally, the pump can't keep up and the small 2500 gallon "stage" tank overflows since the pump can't keep up with a rain rate of 8" per hour rate rain event that lasts a few minutes.

pump curve here for 3HP pump:
https://www.google.com/search?q=pen...BAgHEA0&biw=1280&bih=846#imgrc=r7OqLyuhGhA1LM:

Thanks,
-S

 

[jrmichler]

Let's see if we can solve this without detailed calculations. The hard core physics types will hate me, but we engineers prefer simple approaches that are "good enough".

If I understand correctly:
1) You have 50 feet of static head.
2) You have one 2" line, and the total flow is 85 GPM.
3) The pump curve shows 90 feet of head at 85 GPM. That's 50 feet static plus 35 feet dynamic flow loss.
4) You have a second 2" line parallel to, and the same length and number of elbows as the existing line.

Now for the simplified approach:
1) A second line in parallel cuts the total flow resistance in half at the same flow.
2) Flow loss is roughly proportional to flow velocity squared. Look up Moody diagram if you want better numbers.
3) With a single line, flow rate is zero at 50 feet, 85 GPM at 90 feet, and the flow curve is parabolic.
4) With two line, the flow rate is zero at 50 feet, 170 GPM at 90 feet, and the flow curve is parabolic.
5) Sketch both curves on the pump curve. The estimated flow with two line is where the second curve intersects the pump curve. My WAG (Wild Ass Guess - no calculations) estimates about 120 GPM.

 

[saxman2u]

Thanks, that is what I was thinking. Since this is for rainwater collection, that is a substantial amount for our setup. I might add this line into gain 30 + GPM. Will make it easier on the pump too.

Since I pump my water, I also send it through a sand filter like you would see at a residential pool. The pressure right now is 35 PSI. if I add a second line, should'd the pressure go down too?

-S

 

[jrmichler]

Will make it easier on the pump too.

Well, no. The shaft horsepower to a pump is very roughly proportional to the flow. Low flow is less power, higher flow is more power. If the flow gets beyond the end of the curve (150 GPM on your pump), then many pumps will overload the motor.

The radial load on the shaft is zero at the best efficiency point, and high at both low and high flow. High shaft loads show up as broken shafts, overloaded bearings, and failed seals (due to shaft deflection). Your pump has its best efficiency somewhere near 80 to 100 GPM, and is best run at flows between 60 and 130 GPM.

In practice, all of this is almost meaningless for small pumps like yours. Just don't stop the flow completely while the pump is running, and you will be all right.

 

[JBA]

With your 50 ft of static head, even with zero flowing head loss the pump curve shows that the pump would deliver approx 150 GPM so assuming you cut the flowing head in half, which is what can be expected by adding a second line with the same fittings then your total head will be 50+ 17.5 = 67.5 ft then based upon the pump curve the expected flow will be approximately 125 GPM because the pump is capable of delivering at that rate with that combined head so there will be no reduction of flow in the two lines that would result in a reduction of flow head loss.

Edit: If you were increasing the size of the single line then it would require the other approach.

 

[saxman2u]

Thanks for all of the information. I will report back after i install the parallel line within the next few weeks.

 

[miltos]

The shaft horsepower to a pump is very roughly proportional to the flow

In fact, pump motor output power P2 is proportional to the cube of flow, for a given network, without height difference, if you vary the motor's rpm.

..so assuming you cut the flowing head in half, which is what can be expected by adding a second line with the same fittings then your total head will be

There is no reason to suppose that the flow head will be divided by 2. A more logical very first approach assumption to start with would be that the flow would be divided by 2, so the flow head would be divided by 4.

Anyway, you can calculate the resulting flow easily.
I tried to make some calculations for you, but I roughly estimated that your actual flow is about 75gpm, instead of 85gpm that you said.
So, I made just a very rough approximation of the flow, if you use the second pipe in parallel, which is 100gpm.