Myth on increasing pipe diameter?

[Lavoisier]

Several folks (whom I hold in great respect) have told me that I should increase my pvc pipe diameter coming out of a high velocity pump. As I work on my new build (now over a year old!) I wondered if there was a practical limit in terms of distance, to the value one gains regarding friction loss by increasing pipe diameter from 1 1/2" to 2" before going back to 1 1/2"? In other words does it make sense to increase to 2" for 3 1/2 feet? My particular situation is as follows: My return pump is a Reeflo Barracuda 4' head--3600 gph) into two 4 way OMs. The distance from the 1 1/2" pump outlet (after ball valve) to the 1 1/2" inlet to OM (before ball valve) is about 3 1/2 feet. Likewise, my cl pump is a Reeflo Hammerhead (2' head--5750 gph) feeding two 1 1/2" Sea Swirls with the first distance 3 1/2' and the second about 6'.

Iin an attempt to answer this question I pulled out my copy of Aquatic Systems Engineering by Escobal but found the math in chapter 11 a bit daunting and turned to the Nomograph instead (http://flexpvc.com/graphics/head-los...raph-chart.gif). It appears that at 60 gps (my Barracuda) over a 3' span there is virtually no difference going from 1 1/2" to 2" diameter pipe regarding head loss in relation to psi, feet, or velocity. At 100 gps (Hammerhead) the differences at 3' or 6' appear to me to be neglible as well: psi .826 vs .225; feet 2.1 vs .525. Further, does Escobal points out that there is significant friction loss anytime the hydraulic flow moves into a larger diameter pipe. It seems to me that in a home aquarium we lose much more than we gain by increasing the pipe diameter in our plumbing. Please let me know if I am missing anything here.

 

[power boat jim]

I would bet using the largest radius elbows and as few as possible will do more to decrease back pressure then increasing pipe diameter. This observation just comes from personal experience.

 

[atarjanyi]

I have a BS in Chemical Engineering so I think I can handle this one.

1) Over that short of a distance the turbulence at the diameter change points will cause a greater pressure drop than you will save with a diameter increase.

2) If you have any elbows changing them from 90's to two 45's will cost more for the parts, but save on pressure drop and help maintain flow rate.

3) Cleaning build up from the inside of the pipe will help maintain flow rate much better in the long term than changing diameter that 1/2".

Alex

 

[Lavoisier]

I would bet using the largest radius elbows and as few as possible will do more to decrease back pressure then increasing pipe diameter. This observation just comes from personal experience.

Good advice.

I have a BS in Chemical Engineering so I think I can handle this one.

1) Over that short of a distance the turbulence at the diameter change points will cause a greater pressure drop than you will save with a diameter increase.

2) If you have any elbows changing them from 90's to two 45's will cost more for the parts, but save on pressure drop and help maintain flow rate.

3) Cleaning build up from the inside of the pipe will help maintain flow rate much better in the long term than changing diameter that 1/2".

Alex

This is helpful as well. Do you have a rule of thumb for cleaning out your pipe? Any special brush or tool?

 

[wolfblue]

1) Over that short of a distance the turbulence at the diameter change points will cause a greater pressure drop than you will save with a diameter increase.

Yep, not worth the trouble.

2) If you have any elbows changing them from 90's to two 45's will cost more for the parts, but save on pressure drop and help maintain flow rate.

I keep hearing that. But the truth is, two 45's add up to the same as one 90. Or it can even be slightly worse using two 45's. Two 45's are better for hammer, when you turn the pump on/off. But sweeps is what you really want for flow.

Make custom sweeps--- Glue a cap on a short length of PVC. Pack it tight as possible with sand. Glue another cap on that end. Carefully warm it up with a heat gun, and bend it around something with the proper radius. If it kinks, start over with a new piece. Its ok if you need to gradually change radius. But be careful about sudden changes or you'll end up worse than a 90. Cut the caps off and knock/wash the sand out.

And check out the valuable RC tool "Head Loss Calculator"

--John

 

[dkeller_nc]

I have a BS in Chemical Engineering so I think I can handle this one.

1) Over that short of a distance the turbulence at the diameter change points will cause a greater pressure drop than you will save with a diameter increase.

2) If you have any elbows changing them from 90's to two 45's will cost more for the parts, but save on pressure drop and help maintain flow rate.

3) Cleaning build up from the inside of the pipe will help maintain flow rate much better in the long term than changing diameter that 1/2".

Alex

This is correct, and I'll add that the friction loss from the smaller diameter pipe is largely confined to right at the output of the pump. In other words, it's helpful to use a larger diameter pipe if the pump's impeller housing will accommodate it, but not so much if the pump's impeller housing opening is smaller than the diameter of the pipe that it's feeding into.

With regards to fitting friction loss, you may find this table helpful:

http://www.engineeringtoolbox.com/pvc-pipes-equivalent-length-fittings-d_801.html

 

[Lavoisier]

Great link, thank you.

 

[Lavoisier]

This is correct, and I'll add that the friction loss from the smaller diameter pipe is largely confined to right at the output of the pump. In other words, it's helpful to use a larger diameter pipe if the pump's impeller housing will accommodate it, but not so much if the pump's impeller housing opening is smaller than the diameter of the pipe that it's feeding into.

With regards to fitting friction loss, you may find this table helpful:

http://www.engineeringtoolbox.com/pvc-pipes-equivalent-length-fittings-d_801.html

So if I'm looking at this correctly the larger fittings greatly increase the friction loss. For example a 2" 90 sweep is roughly 70% more friction loss (5.7 feet of pipe) vs. a 1 1/2" 90 sweep (4.o feet of pipe)! That is quite a difference.

 

[dkeller_nc]

Yep, that's correct. But keep in mind that a friction loss in equivalent length of straight pipe is quite different than friction loss expressed in feet of head pressure. I'm not near enough to my books and tables to do this accurately, but a friction loss equivalent of 4 feet of straight pipe is about 6" equivalent friction loss expressed in head pressure.

 

[wakesetter]

Yep, not worth the trouble.


Make custom sweeps--- Glue a cap on a short length of PVC. Pack it tight as possible with sand. Glue another cap on that end. Carefully warm it up with a heat gun, and bend it around something with the proper radius. If it kinks, start over with a new piece. Its ok if you need to gradually change radius. But be careful about sudden changes or you'll end up worse than a 90. Cut the caps off and knock/wash the sand out.


--John

That is damn near genius. I have been doing custom installs for 15 years and have never heard of this. I just use spa flex instead, which still works well. I have heated pvc and bent it, but never used the sand idea.

 

[wakesetter]

My return pump is a Reeflo Barracuda 4' head--3600 gph) into two 4 way OMs. The distance from the 1 1/2" pump outlet (after ball valve) to the 1 1/2" inlet to OM (before ball valve) is about 3 1/2 feet. Likewise, my cl pump is a Reeflo Hammerhead (2' head--5750 gph) feeding two 1 1/2" Sea Swirls with the first distance 3 1/2' and the second about 6'.
.

I hope those pumps aren't in your living room. That has the capacity to be extremely loud, not to mention horrible on electricity. I just swapped out a similar pump situation on someones tank because it was sooo loud that they couldn't stand it. And it was costing them an arm and a leg in electricity.

 

[Lavoisier]

I hope those pumps aren't in your living room. That has the capacity to be extremely loud, not to mention horrible on electricity. I just swapped out a similar pump situation on someones tank because it was sooo loud that they couldn't stand it. And it was costing them an arm and a leg in electricity.

I have used Reeflo pumps for 3 years and have found them very quiet. They cost me (in Kansas City) about $20/month each to run. I'm assuming you swapped in a DC pump on the tank you mentioned because you would not get a meaningful watts/gal return on another AC pump. DC pumps would cost me about $10/month each to run so the savings is not significant enough to me to buy new pumps. I will say that if and when my Reeflo's give out (which should not be anytime soon!) I would certainly consider a DC pump. I haven't done much research yet so I would be curious to know which pumps would match Reeflo's flow (Hammerhead 6000g/hr.; Barracuda 4300g/hr).

 

[SGT_York]

Another point is that friction loss per foot is dependent on the pressure or velocity in the pipe.

The tables also compare the fitting to the pipe diameter which is great for calculating head loss but poor for comparing sizes. To expound, a 1/2" 90 fitting is equal to 3.6' of straight 1/2" pipe. a 2" 90 fitting is 8.6' of 2" straight pipe. However the friction loss of 1' of 1/2" pipe is more 35.5 psi than a 1' 2" pipe under the same pressure .09 psi. http://www.engineeringtoolbox.com/pvc-pipes-friction-loss-d_802.html

So for the 1/2" pipe at 3.6' * 35.5 / 100 = 1.278 decrease in psi at 600 GPH
also the 2" pipe at 8.6' * .09 / 100 = .00774 decrease in PSI at 600 GPH

The 2" pipe will score you 1.27026 psi less friction loss than the smaller pipe per 90 elbow.

3/4 vs 1.5
120 gph = .0018 vs .00135 very marginal difference
900 = .828 vs .0315 significant difference.

Obviously these calcs don't consider the turbulence with size differences. Turbulance will be significantly affected depending on how much head pressure restrictions you have or by how much velocity you reduce. I.E. it will significantly reduce the friction loss if you use one small lockline at the end due to the significant reduction in velocity. Another point is the lifting capacity of your pump in terms of headloss but that is for a different post. The main point is that fluid mechanics is very complex and there are always multiple variables. Guess that's why it is it's own engineering profession.

I'd also like to point out that two 45's do not equal a 90.

 

[BeanAnimal]

of Aquatic Systems Engineering by Escobal but found the math in chapter 11 a bit daunting and turned to the Nomograph....

....Please let me know if I am missing anything here.

FWIW... I have found the Aquatic Systems Engineering text to be rather bug riddled and therefore don't really trust much of what is in it.

Without going into a lot of detail:

I was doing some work with the thermodynamics chapter (in preparation for my thermodynamics article on my site) and found the heater sizing integration to have huge error rates outside of tanks in the 50-150 gallon range.

Assuming that I was in error, I consulted several physicists and thermodynamic engineers in hopes of gaining a better understanding of what I may be missing. In one case I posted to a well know physics forum, posing as a student looking for information, as I was a bit embarrassed to use the BeanAnimal username I am associated with. To my shock, not only did EVERY source (both private contacts and on the public forum) point out that Escobal had made a HUGE error in building the equation, but TWO of the consulted experts (one from the forum and one private referral) pointed toward the thermodynamics article at BeanAnimal.com as a "proper" way to calculate conduction!!

So the error? Escobal neglected to convert units from (if I remember) metric to SI for part of the his base thermodynamics equation. That base equation is a HUGE part of the integration he used as a basis for the entire chapter and resulting heater sizing formula. The formula appears to work for a small range of reasonable tank sizes, but anything bigger or smaller, the numbers get crazy. The error likely went unnoticed because the author built the equation and checked it against a few common tank sizes and the editor is never as smart as the author. The books are now out of print, the publisher out of business, and there is nobody to contact to make the corrections...

I have tried to contact Mr. Escobal but he has been MIA for many years, and I am sure he could care less anyway.

Without going into more detail... there appears to be other similar errors in the book.

So is the fluid dynamics chapter correct? I will leave that to you

 

[Lavoisier]

Another point is that friction loss per foot is dependent on the pressure or velocity in the pipe...

Thank you for your response. I will go back and reconsider going to 2" pipe out of my 1 1/2" Reflo pumps!

FWIW... I have found the Aquatic Systems Engineering text to be rather bug riddled and therefore don't really trust much of what is in it...Without going into more detail... there appears to be other similar errors in the book.

So is the fluid dynamics chapter correct? I will leave that to you

I appreciate your response. The book was not cheap and with your research I wish I had not spent the money!! As with any iconic authority we still have to check particulars and then may be forced to reconsider...

 

[mr.maroonsalty]

I spent quite a bit of time mulling over friction loss charts and equivalent lengths; I chose the step up to 2" out of my hybrid dart with the snapper wheel. After I swapped out the 1.5" I need to restrict my syphon a wee bit instead of opening it a tad as I expected. I'm happy for the increase in diameter simply because of fri tion loses that are starting to show up from growth inside the pipe, plus it gives me the option to install the dart wheel. With the lower velocity pump I was curious if the additional weight of water in the 2" pipe added more head preasure thus causing the slight decrease in flow as seen through having to restrict the syphon a tad?

 

[Aqualund]

no, the only weight to consider is the weight of the water directly over the diameter of the opening.

 

[BeanAnimal]

I spent quite a bit of time mulling over friction loss charts and equivalent lengths; I chose the step up to 2" out of my hybrid dart with the snapper wheel. After I swapped out the 1.5" I need to restrict my syphon a wee bit instead of opening it a tad as I expected. I'm happy for the increase in diameter simply because of fri tion loses that are starting to show up from growth inside the pipe, plus it gives me the option to install the dart wheel. With the lower velocity pump I was curious if the additional weight of water in the 2" pipe added more head preasure thus causing the slight decrease in flow as seen through having to restrict the syphon a tad?

The "weight" of the water has nothing to do with the head...

If you stand at the bottom of a 10' deep swimming pool, the weight of the column of water pushing down on your head is the same as if you were standing at the bottom of a 10' deep ocean.

 

[mr.maroonsalty]

In other words if one could somehow focus the water weight of x(10•1000) on top of a pump, it would displace the same amount (more in fact with less friction loss) as an 1 1/2" pipe with same head height?

 

[Aqualund]

dunno what you mean by this x(10•1000)