How to size full siphon drains?

[scolley]

I'm planning a new tank, and am looking for guidance on the capacity of standard PVC sizes when used in a full siphon drain.

When draining at a full siphon (no air, all water, water level 1-2 inches higher than top of drain) how many gph would you expect from standard PVC pipe:

Pipe Size - ID
1/2" - 0.622"
3/4" - 0.824"
1" - 1.049"
1 1/4" - 1.380"
1 1/2" - 1.610"
2" - 2.067"
2 1/2" - 2.469"
3" - 3.068"

Thanks in advance!

[BeanAnimal]

Here are some rough calcs I threw together for another thread...

the "head height" is the total distance from the intake pool level to the discharge pool level or pipe exit (if the pipe exits above water).

The "bulkhead diameter" refers to the narrowest section in the drain.



No consideration has been given to pipe roughness, turbulance (reynolds number, smoothness of flow). Real world: multiply the numbers by maybe .85 or so.

More than you ever wanted to know about siphon overflows and how to make them fail-safe and silent:
http://www.reefcentral.com/forums/sh...readid=1310585
http://beananimal.com/projects/silen...ow-system.aspx

[scolley]

Hey Beananimal - Thanks!

I am a MAJOR fan of your "Silent and Failsafe" design. Have not read the entire thread yet (it's LONG). But that's because I am reading it. All of it. And that's an investment of quite a few hours. Great thread though. That was a major contribution to the hobby. Thank you.

As to your info posted here...

That's exactly the kind of data I was hoping for. Thank you!

I assume that the rise in flow rate as the drop increases is just like anything else falling - it picks up speed over distanced dropped up to a point were it levels out. If so, I'm planning on a drop of 4' or 5'. Do you have data that goes that far?

Beananimal I posted the thread's original question because I'd like to take the principles in your Silent and Failsafe method and apply it to a tank drilled through the bottom without traditional overflows. I mean three vertical pipes (hidden by rockwork) in the tank - plumbed through the bottom - that drain the tank, employing your three pipe types; Emergency, Full Siphon, and Open Channel. Would be awesome IMO.

Thanks for the help!

[uncleof6]

Originally Posted by scolley

Hey Beananimal - Thanks!

I am a MAJOR fan of your "Silent and Failsafe" design. Have not read the entire thread yet (it's LONG). But that's because I am reading it. All of it. And that's an investment of quite a few hours. Great thread though. That was a major contribution to the hobby. Thank you.

As to your info posted here...

That's exactly the kind of data I was hoping for. Thank you!

I assume that the rise in flow rate as the drop increases is just like anything else falling - it picks up speed over distanced dropped up to a point were it levels out. If so, I'm planning on a drop of 4' or 5'. Do you have data that goes that far?

Beananimal I posted the thread's original question because I'd like to take the principles in your Silent and Failsafe method and apply it to a tank drilled through the bottom without traditional overflows. I mean three vertical pipes (hidden by rockwork) in the tank - plumbed through the bottom - that drain the tank, employing your three pipe types; Emergency, Full Siphon, and Open Channel. Would be awesome IMO.

Thanks for the help!

Such a system would be interesting. However, surface skimming has to be taken into consideration. Siphon system intakes have to be sufficiently below the water surface, so they don't suck air. Trouble is getting it far enough below the surface, will negate surface skimming: you will be draining subsurface water, and the "slime" will stay on the surface. That is the whole idea behind "overflows" and "weirs." To facilitate getting the surface water out of the tank. Those that are using plain pipe drains (no overflow weir) are taking a huge step backward technology wise, and limiting the flow to "very low" if they wish to eliminate problems.

Jim

[scolley]

Holey Moley! I've hit the jackpot with this question.

Jim, before I reply to your post, I've go to say that you are one of the hand full of people on this forum that consistently knows exactly what you are talking about. It's painfully clear in all your posts. So it's a real treat to get your input. Thanks!

I HAVE considered issues with surface skimming. Your comment was right on target. And you were probably also thinking about the tremendous gas exchange function overflows provide - but were being too nice to hit me with too much at once. But to answer both concerns...

I'm only in to the exploratory phase of this, so I don't want to sound like I have "plans". I don't. Yet. But I've been thinking about this problem, and here's what I suspect...

1. The over flow surface skimming function is not that demanding - with most overflows providing far more than is actually needed. I can't prove that, but observations with my own tanks seem to bear that out. So if that assumption is true, a small overflow box on top of the lower-flow open channel stand pipe should be sufficient. Again, hiding this would require some creative rockwork. Protruding from the surface I'm sure. But I've created tanks before with interesting subject matter breaching the surface. Could look lovely. I'm thinking a reef crest...
2. As for the gas exchange, I'm assuming that I'll just offset the loss by a significant overflow of some form in the sump. Nothing says it has to happen at the DT.

Thanks for the feedback!

PS - Here's a couple of tanks I've done with subject matter going beyond the surface. My F/W planted discus tank a few days ago. And another of the same, sadly long since broken down. I'd LOVE to try the same thing with a reef.

[BeanAnimal]

Scolley,

Water flow...

Yes water is pulled by gravity. It (for our purposes) behaves like a solid when it is confined by a pipe, so the further it falls, the faster it falls. It pulls on the water above it as well. I will leave it to you to research if the "pull" is created by surface tension or displacement

In any case we use the basic Bernoulli equation to explain how much water falls how fast.

Lets take the 2" SCH80 Bulkhead with a 1.913 inch I.D. and 48" of head.

h = "height" and denotes the head (water height) above the hole.
g = "gravity" denotes the free fall of an object here on earth.. (32ft/sec)squared

So we find the velocity and then plug it into the basic formula:
Q = A * V

Lets do some math:
2" SCH#80 PVC pipe I.D. = 1.913 inches.

So:
h= 6" = 4'
A= 2.873" sq. = 0.01995' sq.
g= 32f/s sq.

Find the Velocity:
V = sqrt(2* 32f/s^2 * 4f) = 16f/s

Find the flow rate (Q):
Q= 0.01995f^2 * 16f/s = 0.3192 Cubic Feet per Second.

The Result:
0.3192 Cubic Feet per Second = 143.25 Gallons Per minute
143.25 Gallons per Minute = 8595 Gallons Per Hour.

Long drop, lots of pipe and fittings... lets say 7250 GPH

Now... the larger the drop, the more friction and turbulance affect the velocity and therefore, the overall flow. Also note that "terminal" velocity is reached much faster due to the friction and turbulance and that cavitation can occur if the pressure drop (caused by the water falling through the pipe) exceeds the surface tension of the water itself....


The bulk of the gas exchange can happen at the sump. The surface skimming is another story. While your idea of using the open channel as the surface skimmer is interesting, I don't thik it will provide enough surface skimming for a modest system. You may be able to increase the efficiency of the small surface skimmer by directing surface flow towards it with power heads or CL nozzles. Like anything else, there are certainly many ways to reach the same goal. Good luck with the project!

[kcress]

Hi Bean, yah got that chart in 36 to 60 inches? My tank is 36" deep and the sump is about 24" below that.

[BeanAnimal]

DOH! I posted the equation...

I wrote a FLASH calculator but have not had a chance to upload it yet... maybe I will get around to that soon.

[scolley]

Bean Animal – Thanks for the equations!

Being no math wiz myself, and without your FLASH calculator, I spent a little while wrapping my head around your explanation, and then putting it all in a spreadsheet. The two variables being pipe ID in inches, and vertical drop in feet. Works beautifully. Thanks!

For my application I’ll most likely need to know the flow rate of water in a 1” pipe. So I started playing with vertical values to increase the capacity of that size pipe. I found that with a vertical drop of 15,200,000,000,000,000 feet, I can move a whopping lot of water through that little pipe. Though the speed is so excessive that the pipe might vaporize from the friction (maybe I need a new thread to understand heat from water friction?). Or the water could turn into superheated steam at some point in its decent. That’s got to mess up the math. Or maybe the pressure of the gas will just burst the PVC.

I also noticed that at the above specified vertical drop, this water was also exceeding the speed of light. Only 2 tenths of 1% faster mind you. But it made me realize… this math calculates a constant, unlimited, acceleration of falling water in a pipe. Where does the damping factor come in?

Sorry, I couldn’t resist the humor. But is there reasonable way to calculate that? With the friction, turbulence, water tension, etc.? At what point does the acceleration cease?

Thanks again for the help!


PS - In case it seems like I'm being picky, my build-out path of least resistance is using a 1" pipe to achieve 1,800 gph flow. Anything less is not enough, and using a larger pipe means drilling I'd like to avoid. So I'm kinda hoping to get a pretty reliable number if possible. Thanks.

[BeanAnimal]

Short answer:

Yes there are terms that can be added to the equation to compensate for friction and the resulting turbulance and resistance to flow. The problem is that determining the variables to plug into those terms is too complicated to make it worth the trouble (for our purposes at least). For any reasonable pipe size and drop we encounter, these numbers should be in the ballpark.

There are also many different equations that attempt to answer the complex question of single phase flow in a closed pipe. The Darcy-Weisbach equation is one of the most common. The Hazen-Williams equation is also common. Both equations take into account the friction loss in the piping system. The Manning formula is good for open channel flow that is driven by gravity. If you are bored you can build another spreadsheet

[scolley]

Originally Posted by BeanAnimal

If you are bored you can build another spreadsheet

Thanks. Bored I'm not. But I'd be happy to do it, but unfortunately I don't know how. Maybe someone else with a better math head will look up those equations. I'm personally sure the math would beyond me.

I guess I was hoping that there might either be rules of thumb, or pre-made tables (like books of logarithms) that had that information.

Originally Posted by BeanAnimal

For any reasonable pipe size and drop we encounter, these numbers should be in the ballpark.

If I apply these equations to a 1.033 ID piece of PVC with a 3.5' drop (which is my particular application), I'm calculating 2,346 gph. And I'm having trouble wrapping my head around that being remotely possible. If I take your .85 fudge factor above that's still right at 2,000 GPH.

I have no experience with this kind of flow (my experience is in F/W tanks), but I've got a Reeflo Dart for a return for this tank. In testing it at 0' head through a 1.5" discharge I've found it BLOWS water at 3,600 gph or 1 gallon per second.

It just seems really tough to imagine that I can get MORE than half that rate of discharge through a tube with an ID that's LESS than half the area - moved by gravity alone. Defies logic.

Either my math is wrong, or the equations cannot be taken at face value unless it's short runs of wide diameter pipe. Bummer. I think I need to look up that Manning formula. I guess my problem was more difficult than I had hoped.

Thanks again for the info and assistance.

[uncleof6]

Most of us running darts, are using 1.5" drain pipe. Use 1.5" pipe on the 1" bulkhead, and forget the math. I hate math.

Jim

[scolley]

Originally Posted by uncleof6

Most of us running darts, are using 1.5" drain pipe.

Aaahh... are you suggesting 1.5" full siphons, expecting the loss from the 1" bulkhead to be an insignificant "momentary restriction"? I LIKE that thinking! Then with a 2nd one as my open channel, I should be fine?

I guess the only place I get uncomfortable is the emergency drain, which would have the same restriction. So if the emergency was ever fully needed (blockage to the full siphon) I might be choking the Dart back a bit. But it does not sound like too much. Cool!

Thanks for that suggestion. Can you please confirm that I've correctly interpreted your suggestion?

Originally Posted by uncleof6

I hate math.

No kidding. I'm a card carrying member of that club.


BeanAnimal - I've been messing with the calculators you suggested. And unless I'm using it incorrectly (a REAL possiblity), an online manning calculator is putting my 1.033" diameter 3.5' pipe at a discharge of right at 0.067 cubic feet per second. Which - BTW - is EXACTLY what your formulas above project.

So unless I'm using this calculator incorrectly, all my whining above was without cause, and your statement about the numbers being "in the ballpark" was right on target. Sorry for being a pain. I guess sometimes "common sense" is just wrong. It's looking like the numbers don't lie.

AGAIN - thanks for the help!

[uncleof6]

Bean used 1" bulkheads on his system (because they were already there) seems to work well for him I use 1.5" all the way, and most of the time I HAVE to pushing 2800 plus with the darts, and not taxing the system. Am certain the 1.5" with 1.5" bulkheads will top 3000 gph and beyond. But the variables vary......... The systems I have created, fall in the 240 to 300 gallon range, so I don't have a need, really, to go above 3K. I seem to recall that bean's tank would handle close to 3000 gph (for proof of concept)..........with the 1" bulkheads and 1.25" street ells in the box.

Ever heard the phrase: "Going after a mouse with a bazooka?" I think with drain systems, that is a good way to go about it. Have more than you need.

If you do wind up using a 1" siphon, use a 1.25" open channel-- 1" dursos @###@!!&*^@@. And then go to 1.5" for your dry emergecny--- alrighty then.....

[scolley]

Thanks for the help Jim. Will reply in more detail after work.

Thanks!

[scolley]

Thanks Jim. Sounds like I'm good to go! Based on what you are saying, coupled with the fact that I was planning on choking the return back to something a little north of 1800 gph anyway (10x DT volume), it should be a non-issue.

I'll accept that I'll have a momentary choke point on my outflow at the bulkhead, but crank the the siphon up to 1.25 - just to enhance the capacity beyond a consistent 1" pipe. And the emergency up to 1.5" to ensure its capacity exceeds that of a prospectively failed siphon. Cool!

Thanks a mil' guys. I'm good to go! :-)

[uncleof6]

Scolley, I was kidding about the 1.5" dry emergency. But that certainly would negate ANY doubts that the emergency would handle the full flow of a plugged siphon, and the open channel and some more to spare. I would, however, go for 1.25" across the board. Bigger is always better.

Jim

[scolley]

Originally Posted by uncleof6

I would, however, go for 1.25" across the board. Bigger is always better.

Understood. But following that same thinking, I'll go 1.25" siphon and open channel, but 1.5" drain.

Originally Posted by uncleof6

Scolley, I was kidding about the 1.5" dry emergency. But that certainly would negate ANY doubts that the emergency would handle the full flow of a plugged siphon, and the open channel and some more to spare.

Jim, you may have come up with another improvement to BeanAnimal's Silent and Safe design. I would not have though another was possible given the length of it's thread. You just posted it in the wrong place.



BACK TO THE POINT OF THE THREAD
I'm thrilled that I've got a solution to my particular problem. Thanks guys! I phrased my original problem - posing a generic question rather than referencing my specific problem - because I think an answer to the question has value to the community. So I'd like to pursue it further if possible.

BeanAnimal has posted a fantastic great chart for drops up to 12". But it's understood that other factor - like drag and turbulence - will begin to slow down the flow at some point. So we need two thing IMO...

1. Data that shows the flow rates for drops longer than 12". I would think at least 4' for common "discharge under the tank" applications, and up to 20' for "I've got my sump in my basement". 20' is also approaching the practical limit of many pressure rated return pumps.
2. An understanding of what depth the constant acceleration begins to be significantly influenced by the various factors affecting water velocity in the pipe.

I used this Manning Calculator and came up with VERY similar numbers to what BeanAnimal's equations from his second post predicted. In that calculator I used the following values for the variables.

Select Units: Use feet and seconds
Selection Calculation: Use Velocity (V) and Discharge (Q)
Area, A (ft^2): 0.00582
Wetted Perimeter, P (ft): 0.2704394
Channel Slope, S (ft/ft): 1
Manning n: 0.01

Was I using the correct values? And is this the correct calculator for the problem? I question my 1 for the Channel Slope (don't know actually).

And I'm wondering if this the correct tool anyway. There's nothing here to specify the length of the pipe. And it seems to me that without that, we are still dealing with equations that yield constant acceleration, which cannot be right.


Thoughts?

[BeanAnimal]

Short answers:

The chart only goes to 12" of head because that is all that threw in the spread sheet. I posted the formula I used (a simple example of the Bernoulli equation). It works for any reasonable head and pipe diameter we will encounter. It ignores friction, so we fudge the results a few percent low and we are close enough for anything we do.

No, you can not use the manning formula for a vertical flooded pipe. The formula predicts open channel flow and is used in non vertical pipes.

If you want "better" data then you would use the Bernoulli formula to get a rough idea and the use the Hazen-Williams formula to correct the resust. Again, for our vertical siphons that are VERY short and have NO SLOPE there is really no need to bother with Hazen-Williams. It is much easier to just fudge our ideal Bernoulli number down a bit.

Darcy-Weisbach is even more complicated, though it gives a somehwat more accurate result. Again, for the short distances we are working with, it is just not worth the trouble.

The 1.5" pipe connected to the 1" bulkhead only ensures that friction is not a problem. Organic material builds up over time and a 1" pipe is much more prone to fouling and affecting the balance of the system.

Rest assured, you are putting too much thought into this

[scolley]

Originally Posted by BeanAnimal

Rest assured, you are putting too much thought into this

Thanks! It certainly would not be the first time.

But generally speaking, that's why my tanks all tend to work the 1st time Research, planning and preparation is the key. But for application, this case seems to be closed. Thanks!

[JMorris271]

...