Plumbing advice needed

[reef rider]

Getting ready to do plumbing on my 75 gal set up and had some questions. What's best to use, 80 or should I be using 40 schedule PVC? Potable or just industrial. Was going to go with Potable (safe for drinking water).Led to believe industrial could leach chemicals in time.I was going to go with 1" overflows and 1" return. Several friends said I should be using 1 1/2" for returns. Only shooting for 300-350 gph through sump, using a Mag 7 or possibly a Mag 9 return pump. Water volume will be around 110 gal total. I thought 1" would be fine. Advice, pros, cons? Thanks

[uncleof6]

PVC pipe and fittings are inert, and leach nothing, doesn't matter what flavor it is...sch 80 is a waste of money for our purposes. It is for high pressure, which aquariums no got. CPVC (also inert) is for high temperature.

Mag 9 pumps requires 1.5" pipe to get any flow out of it. It is even in the directions. Personally, I consider the whole line of pumps to be poorly designed/engineered, and not worth the money.

You should be headed towards 750 gph on this system, that aside, if using dursos, best make them 1.5" as well, as 1" won't deal with 350 gph, 50 gph maybe...

[reef rider]

Thanks for the info uncleof6. I will go with 1.5" overflow and return. Was wondering why PVC mfg. have a Potable PVC for drinking water. Good to know the info you provided on PVC. Always looking to save money! In your opinion, what would be a good reliable internal pump for the money, in the 750 gph range. Been away from the hobby for a few years and trying to come up to speed. Thanks again!

[McPuff]

Mag 9 pumps requires 1.5" pipe to get any flow out of it. It is even in the directions. Personally, I consider the whole line of pumps to be poorly designed/engineered, and not worth the money.

You should be headed towards 750 gph on this system, that aside, if using dursos, best make them 1.5" as well, as 1" won't deal with 350 gph, 50 gph maybe...

Hmm, I've got 2 x 1" drains and 2 x 3/4" returns... using a mag 18. It's working fine. Am I misinterpreting something here?

[reef rider]

Hey McPuff. I was going to go with 1" overflows and return. Many have said to go with 1.5". What's your setup like? Any pics? Mag 18...I'm assuming larger system than what I'm building.

[uncleof6]

Hmm, I've got 2 x 1" drains and 2 x 3/4" returns... using a mag 18. It's working fine. Am I misinterpreting something here?

Yes, you are misinterpreting your actual flow rate, which is far below what you think it is.

With only .88in² cross-sectional area for your two 3/4" pipes (it is actually a little lower) compared to 1.76in² crossectional area for a single 1.5" pipe, the friction loss is excessive. Friction loss is a function of the flow rate, cross-sectional area, and the length of pipe, and adds to the vertical lift.

Conventional wisdom for plumbing return systems is to upsize the pipe one size over the actual outlet size of the pump. This cuts the friction loss by ~ 1/2 - 2/3. This applies to all centrifugal pumps, save a subset within the class of pumps called positive discplacement pumps, where the flow curve is a straight vertical line (unaffected by vertical lift.) The Danner pumps are called "positive discplacement pumps" but due to engineering/design flaws, their flow curve is typical of common centrifugal pumps, and 1.5" outlet piping/tubing is required for the pump to operate on the flow curve, in the 9.5 and larger pumps. This puts them in the worst of class category.

Again, with durso/similar drains, physics tells us how they are going to behave, in a given situation. Once the pipe is 1/4 full of water, the flow becomes turbulent, the air and water mix, and you get the problems asscociated with all air assisted drains. It does not take a great deal of thought to realize that the flow rate required to fill a 1" pipe 1/4 full, is not very high at all. 50 gph is a ballpark figure, but 350 gph in 1.5" pipe is pretty common for the standpipes to not be "working fine." Siphons have different operating characteristics, and the above does not apply.

Some tweak their dursos/air assisted drains to near siphon, and get "decent" performance at the price of significant flood risk and instability. Some make claims that their air assisted drains are defying the laws of physics...simply put air assisted standpipes are low flow devices, the inventor of which, by his own admission, had no idea of the physics involved.

There is some variation from system to system, and no two are alike, however, fluid dynamics gives us the foundation to predict how something is going to behave. "Working fine" is ambiguous, and really does not say much of anything...

[reef rider]

Well explained uncleof6. Perhaps I should be re-considering my overflow design. I guess, like everyone I want something efficient, stable, and quiet. I was going to have an overflow box-36" long inside top of tank, 2 independent overflows,with stand pipes tee'd off the bulkheads outside the tank. Almost straight drop to the sump. Thought it would be simple, work well.

[SmoothSmoke]

Thanks uncleof6! I enjoy reading your posts on plumbing. Trying to take in as much as possible.

[reef rider]

I'm trying to understand this uncleof6. I don't want a high flow rate thru my sump. If it were to be around 300gph, could the stand pipe design work using 1.5" PVC? Is the threat of flooding due to the higher flow rate and the Pvc being too small to handle the volume of water? Bare with me. I'm not an engineer. Hope these questions aren't to dumb.

[reef rider]

You stated I should be headed toward 750 gph. Did you mean actual gph through this system, or a pump rated at 750 gph knowing there will be head loss, bringing it down to around 300-350 gph range?

[yaksaredabomb]

Agreed on the return plumbing diameter (bigger is better, flow with 1.5" will probably be noticeably greater than with 1").

.... Once the pipe is 1/4 full of water, the flow becomes turbulent, the air and water mix, and you get the problems asscociated with all air assisted drains. It does not take a great deal of thought to realize that the flow rate required to fill a 1" pipe 1/4 full, is not very high at all. 50 gph is a ballpark figure, but 350 gph in 1.5" pipe is pretty common for the standpipes to not be "working fine." Siphons have different operating characteristics, and the above does not apply. ....

I recommend using a 1" siphon to meet nearly all of your drain's requirements, with a 3/4" secondary drain (1" at biggest). If it's possible for you to do this you can have a 100%, literally silent drain system.

I had an extremely noisy external overflow box (PF-800 gulping air) and came across a really neat idea. It was meant for internal overflow boxes but I adapted it to my situation. Wish I could give credit to the right person for it.

The mouth of the main drain line should be as low as possible in your overflow box. The mouth of the secondary drain line should be above it, near the top of your overflow box (more detail later).

Use a gate valve (ideally) or ball valve (if you must) to restrict your main drain line to the point that it just barely can't keep up with your return pump. At this point it should be working as a completely silent siphon, passing 99.5% of the drain water. The secondary drain will easily take care of the remaining trickle without a sound.

When you first turn on your return pump the water level in the overflow box will start down at the mouth of the main drain. The siphon will develop in the main drain (passing almost all the water) but won't be sufficient so the water level in the overflow box will rise very slowly until it reaches the mouth of the secondary drain. The mouth of the secondary drain therefore controls the water level in the overflow box.

If the primary return (the tuned siphon) ever has an issue and is partially or completely blocked, the secondary return will act as a fail safe. As the water rises, the secondary won't be able to keep up and will soon develop a full siphon itself. The new siphon in the secondary will easily gulp down the excess water in a flash, making a loud slurp when it finishes draining the backlog and the siphon breaks. The siphon in the secondary will form and break, making this noise over and over again (unless the main drain line clears) - thus alerting you to the obstruction in your main drain line!

Pipes are capable of carrying far more water once a siphon develops, especially if there is a large difference in water levels (like is typical from the water level in a display compared to the level in a sump).

The mouth of the secondary drain should be at the water level you want in your overflow box. I have mine about 2 inches below the main water level in the display tank, which is about 1 inch below the bottom of the interior box's "teeth". If it's too high the water level in the box may climb above the bottom of the teeth, thus slowing the drainage a bit and causing the water level to rise slightly in the display. That's not too big a deal but I don't like it since it adds to the system's volume and is more water that must fit in the sump during a power outage. If the secondary drain's mouth is too low then the water will fall too long a distance after it pours through the "teeth" causing a trickling or splashing sound.

If you position the mouth of your secondary drain correctly, and both drains terminate below the surface of the water in your sump, you won't be able to hear a trickle anywhere in your drain system. I'm telling you it's wonderful.

My 75G tank with ~300 gph return (after head loss) has been set up this way for about 2 years without a single problem. I've deliberately blocked my main return to test the 3/4" secondary many times and water accumulates until the secondary return is submerged (I had to attach it at right angles to the water's surface due to limited space in the overflow box). Once the mouth is submerged though it only takes seconds to suck the extra water down with a very loud slurp. You would probably even be fine with a 3/4" primary return, and I intend to change out my 1" for a 3/4" when I rearrange my plumbing (to make more space in the stand).

As far as flow rate is concerned - whatever "turnover" rate you've decided you want can be met by internal powerheads (Hydor, Tunze, etc), closed-loop pumps (external pumps that suck water from the display and return it directly to the display without exposing the water to air), and things like your sump and return.

If you want 10-20x turnover on a 75G tank you're looking at 750-1500GPH rushing through your sump and that's probably going to be noisy (as well as unnecessary). It could also make it hard to keep microbubbles from your skimmer out of the display tank since the water won't have enough "dwell" time to allow the bubbles to rise before they're sucked into the return pump intake. Much easier and calmer to push 300GPH through the sump and make up the remaining 450-1200GPH with internal powerheads and/or a closed-loop system.

In my tank I have 1800GPH of flow from two Hydors and even used to have another 1200GPH before another powerhead died. With 3000GPH of in-tank flow + 300GPH through the sump I thought I'd have a sandstorm and wondered if it would be too much for the fish, but it was just fine. Maybe the rockwork was breaking it up enough to not cause problems.

Good luck and hope this helps!

Edit: I just remembered why I did NOT use 3/4" for the main drain line and why it would probably be a bad idea. The 3/4" secondary has no fittings or obstructions, other than the bulkhead where it's attached to the overflow box. The main has the gate or ball valve though! Most valves tend to restrict the flow a little, even 100% "open". Some valves are designed to be the full diameter of the pipe or tube and don't have this "flaw". So while 3/4" may work perfectly fine for the secondary, before I (or you) decide to use it for the main it would be a very good idea to check out the valve used to dial in the siphon and make sure it's the non-restrictive type.

[reef rider]

Thanks for the reply Yaksaredabomb. That's what I'm going to do, low flow thru the sump and higher flow in-tank via powerheads. I just saw a video on Youtube, by "JoeK". Interesting and non intrusive overflow on a 65 gal reef. My design was similar to his off the back of the tank with the stand pipes. However he just used a piece of PVC running the length of the tank to maximize surface skimming. Looks interesting, in-expensive, and sounds quiet. Claims he ran it over a yr. with no problems. Anybody seen it? May try this.

[uncleof6]

I'm trying to understand this uncleof6. I don't want a high flow rate thru my sump. If it were to be around 300gph, could the stand pipe design work using 1.5" PVC? Is the threat of flooding due to the higher flow rate and the Pvc being too small to handle the volume of water? Bare with me. I'm not an engineer. Hope these questions aren't to dumb.

Why not? There is no valid information to support the notion that that higher flow rates are bad...the topic is a complex interdisciplinary course in physics, chemistry, and biology, surface skimming/renewal (rather than mixing organics back down with low through rates and high mixing rates,) what actually influences skimmer efficiency, and last but certainly the main reason: higher flow rates, mean bigger pumps, and higher electrical costs. Finally, there is the ambiguous "no issues," "works fine" anecdote that says absolutely nothing. Suffice to say that the volume of common advice and reasoning is mostly hogwash. Reaching at straws such as the "turnovers" are created by powerheads in the tank, (a turnover is for our purposes defined as water in/water out, and always has been) water "rushing" through the sump, tons of flow, the sump will be loud (a design issue not a flow rate issue) is intended to make one feel foolish, if they jump the flow rate up above what was recommended for undergravel filters plate systems from 30 - 40 years ago. The open systems in use today, do not work the way these old systems work, and the thinking needs a major overhaul. (actually a return to logic.)

You stated I should be headed toward 750 gph. Did you mean actual gph through this system, or a pump rated at 750 gph knowing there will be head loss, bringing it down to around 300-350 gph range?

Yes the flow through the system as a whole. Powerheads provide vertical mixing, something that is lacking in top out/in type systems. This is the reason powersheads were employed, not to add to the "turnovers" which no one is even getting close to anyway...with some of these "rules of thumbs" heading upwards to 20000 gph. (depending on tank size.)



Yaksaredabomb: Impressive post, unfortunately some of it is inaccurate, (too many different sources of information most likely.) I would go through it a piece at a time, but I don't have the time.

It is not possible to credit one person, with developing what has come to be known as the "herbie" type overflow (it is actually a drain system.) Herbie modified a design (siphon and dry emergency) that was already in use, to fit in a corner overflow. The concept is lost to antiquity, in the quest to circular file the durso standpipe, which proved to be more trouble that it was worth, in marine systems. With low flow freshwater systems it excelled in stopping the long water fall in corner overflows, which was its intended purpose.

Who the first was to employ a siphon system, who knows. It wasn't me, more likely it was many in different locations with no contact, saw it as the logical solution. A couple years ago, I suggested a siphon system in a badly overrated eshopps overflow, and over protests and heckling by "heavy hitters" it cured the problem, but since the failure point is the over the back siphon tubes, there is not much use for a dry emergency...

Couple things that are brief: Siphon bulkhead should be the smaller of the two if they are different sizes. You always want the back up to have a higher capacity (irregardless of valve settings on the siphon.) It is just good common sense, if nothing else.

Dry emergency: means DRY no water in it. Therefore the dry emergency inlet needs to be ABOVE the normal running water level in the overflow. Water flowing in the dry emergency, negates the fail safety feature, as the dry emergency is now a plug risk, just as the siphon is...

Pipe size vs flow rate: Flow rate in a drain system is limited by the bulkhead size. (actually the smallest diameter opening in the drain system.) There is an approximate max theoretical limit for flow through the bulkhead based on Bernoulli's equation, according to the size of the bulkhead, and the length of the drop. The pipe size influences how close to that max limit the drain system will get. Larger pipe size, on the same size bulkhead, will flow *more* than the smaller size pipe, due to less friction loss. 1" pipe, is the bottom end size for a drain line. Smaller is asking for trouble. It is far to easy to plug up 3/4" pipe. Larger is optional, depending on your flow rate however, 1" bulkheads and 1.5" pipe is the design for the best siphon system in use today. Granted, a bit much for a pico or nano...

[McPuff]

Yes, you are misinterpreting your actual flow rate, which is far below what you think it is.


Wow, really great information and I do appreciate it! I have long thought that my return pump was too powerful for what I really need and it turns out most of that power is actually going to waste anyway. And 150W running constantly is quite expensive when accrued over an entire year (even month). This reaffirms my plans to go with a smaller return pump (perhaps DC) to save a TON of energy and increase efficiency in many aspects of my reef.

That said, I'm hoping you might consider addressing a few questions:
1) For my 120 (60 gal sump ~40% full) what flow rate should I try to achieve through the sump? Head is about 5'. And what type/brand of pump would you recommend given your dislike for Danner mag drives?

2) I already have 3/4" bulkheads (i.e., drilled holes cannot be changed) for my returns going into the display... in fact, I actually have a 1" SCWD in line as well... would it makes sense to increase the return pipe diameter up to/after the SCWD and until the bulkheads? If I were to install a DC pump, the outlet is already 1 1/4" I believe so I would still need to increase pipe diameter a bit. But then flow would be restricted at the SCWD and again at the bulkhead. Will this still lead to improved flow rate from the return pump or negate the increase in pipe diameter. Of course I could remove the SCWD because at lower flows, the switching is going to be quite slow and it's effects not really worth the effort. Besides, random current is already being created within the tank by pumps.

Thanks again for the previous comments and clarification. I am always striving for efficiency and it looks like I've got a great opportunity to improve the mechanical and biological (i.e., skimmer) efficiency in my system.

[reef rider]

I will go with a higher flow rate through my system. Thanks again for the great info uncleof6. Still would like a good internal pump recommendation.

[uncleof6]

RLSS Waveline DC series of pumps...

[yaksaredabomb]

Why not? There is no valid information to support the notion that that higher flow rates are bad...

I don’t see why they are bad either, at least for the water quality, but you mentioned a few good reasons to keep flow rates through the sump “reasonable”. More flow means bigger pumps, more electricity costs, and more noise and turbulence in the sump to start with. More heat input to the system as well if using a submersed return pump. And greater difficulty keeping microbubbles from the skimmer out of the return pump intake.

…water "rushing" through the sump, tons of flow, the sump will be loud (a design issue not a flow rate issue) ...

It’s simple physics and geometry to see 750GPH dispersed in a 75G tank will produce far less “rushing” and noise than 750GPH through a sump that is likely 30G at best and perhaps only half full to set proper skimmer depth and have reserve capacity for a few gallons to backfill in a power outage before the siphon breaks in the return line (typically located below the display tank’s operating water level) kick in. Maybe the skimmer can be raised up a bit to get the water level higher in the sump but usually vertical space is limited for doing so – I know I have next to no clearance for raising my skimmer, unless I were to buy a smaller skimmer I suppose.

Say there are maybe 15 gallons in the sump, perhaps a reasonable assumption. So 750/75 is 10, while 750/15 is 50. That’s 5 times the flow per volume! So sure, maybe one can keep that quiet and avoid trickling sounds, but it’s not necessarily a simple challenge to meet.

It could also make it hard to keep microbubbles from your skimmer out of the display tank since the water won't have enough "dwell" time to allow the bubbles to rise before they're sucked into the return pump intake.

About the microbubbles – it seems like a common design to have baffles as an attempt to prevent bubbles from the skimmer from reaching the return pump chamber. By forcing all the water to move down for a few inches, while the bubbles naturally want to move up, many or all of the bubbles can be eliminated from the water stream. A common sump width is 12”, and baffle width may be 1”. So about 12 square inches in cross section, where water is moving downwards. If 750GPH was passing uniformly through this cross section, it would have to be flowing at:

1 cubic inch = 0.00433 gallons. 750 gallons/hour = 173,210 cubic inches/hour.
12 cubic inch wide path means 14,434 “chunks” of 12 cubic inches/hour or 4.01 “chunks” per second.
So the water will be traveling downwards at about 4” per second.

In reality, the flow is not uniformly passed due to corners and drag and fluid dynamics I’m not smart enough to calculate, but it doesn’t take a degree to know the water in the center of a stream flows faster than the water at the edges. To take a blind stab at accounting for this effect, I’m going to say perhaps the water in the middle of the cross section will be traveling at 6” per second. This might not be the way you would have calculated it, and my assumption about the fluid dynamics may not be very good - but I think it’s accurate enough for our purposes.

This means any bubble large enough to rise at a rate faster than 6” per second will be eliminated, while those too small will pass right through the baffle. In my opinion that flow rate is far too fast. Maybe the sump can be designed with 2” baffles to slow the linear flow rate, but space in a sump is usually pretty limited and the extra space may not be in the budget (I couldn’t fit it into mine without having to downsize my skimmer or eliminate my fuge section).

So that’s a really long way to say flow rate through a sump is not free and there are real downsides to going overboard with flow. I’m NOT saying I know how much flow is “reasonable” or how much flow is “going overboard” - I certainly am not qualified to make up those numbers myself. Perhaps it’s best to go with uncleof6’s recommendation of 750GPH – just be aware of some of the above so you can try to minimize the consequences of a higher flow rate.

Yes the flow through the system as a whole. Powerheads provide vertical mixing, something that is lacking in top out/in type systems. This is the reason powersheads were employed, not to add to the "turnovers" which no one is even getting close to anyway...with some of these "rules of thumbs" heading upwards to 20000 gph. (depending on tank size.)
Thanks for distinguishing between “flow through” versus “vertical mixing”. That’s interesting about powerheads not being meant to add “turnovers”, and I think I see where you’re coming from, but it goes against many recommendations I’ve heard and even some logic.

How does a coral or a fish know where the flow is coming from, and why would they care? More vertical mixing will keep food suspended longer than more flow through. Vertical mixing will remove waste products from coral just the same as flow through. Same goes for gas exchange – it doesn’t matter what mechanism is used to ensure new water is constantly circulated to the surface of the display. Most of the biological filtration is happening in the rock and sand in the display tank and doesn’t depend on flow through the sump, unless one has a large fuge I suppose.

Sure vertical mixing won’t get mechanical filtration done, won’t get nutrients to the skimmer, won’t get phosphates to a GFO reactor, etc etc, but so long as these things are still getting done *well enough* then what’s the problem with using powerheads to give the fish exercise and keep the corals happy?

Maybe that’s the point – without a higher “flow through” rate these other tasks are *not* getting done satisfactorily by most fish keepers.

Yaksaredabomb: Impressive post, unfortunately some of it is inaccurate, (too many different sources of information most likely.) I would go through it a piece at a time, but I don't have the time.
I appreciate your comments, especially given that I’m sure you have much more experience than I do, and if you ever find the time I would be interested to hear more about the inaccuracies. I tend to be long-winded, but to some extent I think it's a little like "showing one's work" on a homework assignment - it makes it easier for the teacher to point out where you went wrong haha.

It is not possible to credit one person, with developing what has come to be known as the "herbie" type overflow (it is actually a drain system.) Herbie modified a design (siphon and dry emergency) that was already in use, to fit in a corner overflow. …. Dry emergency: means DRY no water in it. Therefore the dry emergency inlet needs to be ABOVE the normal running water level in the overflow. Water flowing in the dry emergency, negates the fail safety feature, as the dry emergency is now a plug risk, just as the siphon is...
Apparently the design I implemented is not a true “herbie” in that case, because the secondary drain I mentioned was never intended to be a “dry emergency” and the design won’t work (or would be extremely finicky to design or tune) without the second drain being wet.

In the design I’m using, the main drain (the one with a valve) is tuned such that a siphon develops that is not quite capable of handling all the flow. The remaining tiny amount of flow fills the overflow and drains silently down the secondary drain as the slightest trickle. To maintain a silent siphon, keep the water level at a steady level, and keep the secondary dry, the main drain would have to be tuned to carry 100.000% of the flow, exactly.

100.001% and the siphon will occasionally drain the overflow and make the obnoxious slurping sounds. Because the siphon has broken, the main drain will no longer be capable of carrying the entire flow and so the water level in the overflow will rise. Once the water level has risen sufficiently far above the mouth of the main drain so air is no longer sucked in, the siphon will reform and gurgling will cease….for a short time until the siphon empties the overflow again. The cycle will repeat endlessly, driving you and/or your SO insane.

99.999% and the siphon will not be able to keep up, causing the water level to rise until it reaches the mouth of the secondary. That’s how mine is configured. The output of the secondary is sometimes literally only dripping because it is passing so little water, but I can’t stop the dripping entirely because it is impossible to hit that 100.000% exactly and I am not concerned enough about maintaining a “dry emergency” drain to put up with the trickling and slurping that tuning to 100.001% causes.

Regarding the need for a “dry emergency” with my setup:

Yes, a dry emergency would offer lower risk. However, both the main AND secondary drain lines would need to be significantly obstructed before causing a problem. That’s more redundancy than many people with a single drain have, and it would be enough to satisfy me.

However, I’ve also made it impossible for my display tank to overflow EVEN IF both drains are COMPLETELY blocked and it’s not that hard. There simply isn’t enough water in the return chamber of my sump to allow the display to overflow. The return pump would run dry before the display overflows.

So if the redundancy of having two drains isn’t enough (even if they *are* both “wet”), I’m not risking a spill. I’m risking burning out my return pump. I’m quite comfortable with that risk – it’s less than in a standard 1-drain setup after all - but I realize not everyone would be. I suppose I could add a THIRD drain as that “dry emergency”, but that sounds obnoxious and I’m already satisfied with my current level of risk.

….A couple years ago, I suggested a siphon system in a badly overrated eshopps overflow, and over protests and heckling by "heavy hitters" it cured the problem, but since the failure point is the over the back siphon tubes, there is not much use for a dry emergency...
Sounds exactly like my overflow! I actually bought a second over the back siphon tube, thinking that would add redundancy and help prevent failures. However, what really happened is that the flow through each tube was cut in half and it was much easier for bubbles to accumulate at the top of the u-tubes. Instead of reducing risk I had increased it!

That’s part of what makes me advocate for smaller drains when using siphons (within reason of course). A constant flow rate (set by the return pump) through a 1” drain will have more than twice the linear velocity than the same flow rate through a 1.5” drain. Kinetic energy is 0.5 times the mass times the velocity squared (1/2*m*v^2). With the water in the 1” drain moving more than twice as fast, it has more energy and isn’t so easily plugged as one might think! Of course, reason has to apply here too - a 1/4” drain will be more easily clogged than a 4” drain. It would take a lot to clog a 3/4" drain, though, like perhaps a fish (and two fish suicides at once?). Algae won’t do it and neither will snails at those water speeds.

3/4" cross section is roughly 1.75 square inches.
300GPH in my system = 69,300 cubic inches.
About 11 feet per second – good luck snail :P

Pipe size vs flow rate: Flow rate in a drain system is limited by the bulkhead size. (actually the smallest diameter opening in the drain system.) There is an approximate max theoretical limit for flow through the bulkhead based on Bernoulli's equation, according to the size of the bulkhead, and the length of the drop. The pipe size influences how close to that max limit the drain system will get. Larger pipe size, on the same size bulkhead, will flow *more* than the smaller size pipe, due to less friction loss. 1" pipe, is the bottom end size for a drain line. Smaller is asking for trouble. It is far to easy to plug up 3/4" pipe. Larger is optional, depending on your flow rate however, 1" bulkheads and 1.5" pipe is the design for the best siphon system in use today. Granted, a bit much for a pico or nano...
In general agreed – especially on the return side. On the drain side though, if using a siphon, larger pipe may just take up more space, be more difficult to work with, and make it more difficult for the siphon to form. YMMV but I can say with much confidence that a siphon in a 3/4" ID hose with less than 3 feet of drop is significantly *more* than capable of handling my return pump’s output of about 300GPH. I tested this many times before I was comfortable sleeping at night haha.

My return pump:
Rio 10HF HyperFlow Water Pump
http://www.riopump.net/products_pumps/hyperflow_desc.html
http://www.riopump.net/images/hyperflow/hyperflow_graph.jpg
3/4" output, 1” thin-wall PVC return, non-restrictive outlet, 3 elbows, with output about 3 feet above the water level in the return chamber.

1” ID has about 78% more area than 3/4", so I really can’t see needing more than 1” diameter unless you want more than 600GPH or have less than 3 feet of drop. The physics say 2 feet of drop would only pass 2/3 the flow (makes sense because the water in the drain would only weigh 2/3 as much, so the force would be 2/3 as much, so the pressure would be 2/3 as much, etc).

[FrostyChicken]

Conventional wisdom for plumbing return systems is to upsize the pipe one size over the actual outlet size of the pump. This cuts the friction loss by ~ 1/2 - 2/3.

Uncleof6 - Am I interpreting this correctly that a pump with a 1.5" outlet, for example, should be stepped up to a 2" return line? If this is the case should I also run the 1.5" PVC at least 15" (10x the pipe diameter) before making the transition?

[yaksaredabomb]

Conventional wisdom for plumbing return systems is to upsize the pipe one size over the actual outlet size of the pump. This cuts the friction loss by ~ 1/2 - 2/3.Uncleof6 - Am I interpreting this correctly that a pump with a 1.5" outlet, for example, should be stepped up to a 2" return line? If this is the case should I also run the 1.5" PVC at least 15" (10x the pipe diameter) before making the transition?
I tested it both ways with my saltwater mixing station pump and found it best to make the conversion as early as possible. In this case the pump outlet was 1/2" male thread so I screwed on a 1/2" female thread to 1" slip adapter. Flow was substantially better that way than when using any length of 1/2" pipe before the adapter.

My testing wasn't very scientific and it's possible my finding was wrong or at least not typical. Maybe Uncleof6 knows something more or has a different suggestion.

Link to the pump model I tested:
MODEL 7 MAGNETIC DRIVE PUMP 700GPH
http://www.dannermfg.com/Store/Products/Danner/PID-02517.aspx

[Jay Fortay]

This thread is awesome! Great dialogue in here. Thanks to those who are contributing.

[uncleof6]

Uncleof6 - Am I interpreting this correctly that a pump with a 1.5" outlet, for example, should be stepped up to a 2" return line? If this is the case should I also run the 1.5" PVC at least 15" (10x the pipe diameter) before making the transition?

Yes, but with conditions. Submersible pumps behave differently than external pumps. This applies to the pump INTAKE.

Generally an external pump, with a 1.5" outlet should be upsized to 2". Absolutely. However, if the pump inlet is also 1.5", the inlet pipe should be upsized to 2" as well. If not the pump will cavitate, and tear itself apart. It is a rather complex topic. You can research it if you want.

A submersible always has a flooded intake, (no inlet piping) so only the outlet need be upsized.

The pipe should be upsized at the outlet of the pump. Yes you will need that bigger valve...no skimping to save $ on the valve.

The 10x rule applies to the pump intake (only for external pumps) and the size applies to the entire length of the intake plumbing. 10x the diameter straight run after an elbow/change of direction. Again a rather complex subject.

[reef rider]

I am going to run a gravity fed display 20 gal cube refugium off one of my overflows with dual return to the sump. This is how I was going to achieve 100 gal total water volume in this system. Do you think Lowering the flow to the refugium is a good idea or not necessary? My sump is a 30gal long and was going to have 3 baffles after skimmer, and 2 before the return pump. Thought this would get rid of all air bubbles.I too love the dialogue. Very educational. This has been most helpful and thought provoking. Thanks, ,to all, for your contributions, and your time.

[mikespeed6]

Personally never had a problem with my mag drive pumps I know others disagree but I've never had a failure that wasn't self inflicted......

[uncleof6]

1” ID has about 78% more area than 3/4", so I really can’t see needing more than 1” diameter unless you want more than 600GPH or have less than 3 feet of drop. The physics say 2 feet of drop would only pass 2/3 the flow (makes sense because the water in the drain would only weigh 2/3 as much, so the force would be 2/3 as much, so the pressure would be 2/3 as much, etc).

Small bites....

Weight of the water has nothing to do with it. That would be the same as saying that a 10' high column of water 1" diameter, will have less pressure at the bottom, than there is in the ocean at the same depth (10').

From Newton, we know that in a vacuum, a pea will fall at the same rate as a 600 lbs ball of thread....acceleration due to gravity, no lift, no drag, no viscosity, no density. Blah blah...

Bernoulli's equation for inviscid flow, uses velocity, acceleration due to gravity, elevation, pressure, and density. Neither pressure or density involve weight. As pressure increases, the flow decreases, and vice versa. Weight involves mass x acceleraton due to gravity...it is a static property.

Maximum possible drain rate for a tank with a hole or tap can be calculated directly using Bernoulli's equation, and is found to be proportional to the square root of the height of the fluid in the tank (above the opening). (Toricelli's Law.)

http://www.beananimal.com/media/6385/bernoulli.gif where v is velocity, g is acceleration due to gravity, and h is the height. The water has velocity because gravity is pulling it.

With the velocity in hand, we can easily determine the flow rate through the opening (bulkhead):

http://www.beananimal.com/media/6859/flow-rate.gif where Q is the flow rate, A is the crossectional area of the opening, and v of course is the velocity.

Once pipe is added to the bulkhead, it all goes south. Although a 24" drop (1" open bulkhead) can "theoretically" flow ~1660 gph, because the pressure increases in the pipe, the flow slows (friction loss.) What started as 1660 gph is now ~ 1200 gph. Increasing the pipe size, to 1.5" reduces the pressure in the pipe. Lower pressure = less drag, less friction loss, = higher flow rate, so the value jumps up to ~1500 gph.

What size pipe should you use? Heck don't ask me... I will say that because friction loss in 3/4" pipe is extreme, it is not a good idea, along with that it is too darn easy to plug up. Tank manufacturers pop 3/4" holes in 240 gallon tanks, heck they build tanks, they don't know, and don't care, how the rest of it works... 1" or 1.5" pipe? Depends on what it is you are trying to do.

[yaksaredabomb]

1” ID has about 78% more area than 3/4", so I really can’t see needing more than 1” diameter unless you want more than 600GPH or have less than 3 feet of drop. The physics say 2 feet of drop would only pass 2/3 the flow (makes sense because the water in the drain would only weigh 2/3 as much, so the force would be 2/3 as much, so the pressure would be 2/3 as much, etc).Small bites....

Weight of the water has nothing to do with it. That would be the same as saying that a 10' high column of water 1" diameter, will have less pressure at the bottom, than there is in the ocean at the same depth (10').

From Newton, we know that in a vacuum, a pea will fall at the same rate as a 600 lbs ball of thread....acceleration due to gravity, no lift, no drag, no viscosity, no density. Blah blah...

Bernoulli's equation for inviscid flow, uses velocity, acceleration due to gravity, elevation, pressure, and density. Neither pressure or density involve weight. As pressure increases, the flow decreases, and vice versa. Weight involves mass x acceleraton due to gravity...it is a static property.
That’s really interesting! Thanks for taking the time to explain it properly versus my somewhat flimsy use of terms like “weight” and lack of equations.

At first I didn’t think I was guilty of what you spoke of.

I know the pressure in the bottom of a plugged 3-foot tall, 3/4" diameter pipe is the same as the pressure in the bottom of a 3-foot tall, 1” diameter pipe. This is true even though there will be about 78% more water in the 1” pipe, so the water in it will weigh about 78% more. Yet the pressure is the same. The difference in that case is the force. Since Force = Pressure times Area, the downwards force on the plug at the bottom of the 1” pipe will be about 78% more. And because Force = Mass times Acceleration, 78% more force ought to mean you can move 78% more mass (since acceleration due to gravity will remain constant). Because water is relatively incompressible, 78% more mass means about 78% more volume – so 78% more flow.

Further, when I said a 2-foot drop would move 2/3 the water of a 3-foot drop (pipe size being equal) I was considering the same equations. My logic went: Pressure = Force/Area, Force = Mass * Acceleration, and the mass in the 2-foot pipe will be 2/3 the mass in the 3-foot pipe. Therefore, pressure at the bottom of a 2-foot pipe will be 2/3 the pressure at the bottom of a 3-foot pipe. As in the above paragraph, I reasoned 67% of the pressure would only move 67% of the mass (under the same acceleration due to gravity) and therefore the flow in a 2-foot pipe would be 2/3 the flow in a 3-foot pipe of the same diameter.

To address the Newton example – yes, a pea may fall at the same rate as a 600 lb ball of thread….but in terms of mass falling per foot, per second, the mass of the ball of thread will be falling at a greater rate. A 1-gallon pitcher of water will fall at the same rate as a 55-gallon drum of water, but the “flow rate” if you will is much greater for the water in the drum. More water is moving the same distance in the same time, and the difference in flow rate is 1:1 directly proportional to the size of the containers being dropped (the water in the 55-gallon drum is “flowing” 55 times faster). So this supports the idea that a 1” pipe carrying 78% more water than a 3/4" pipe will have a flow rate 78% greater.

HOWEVER, I then realized I was using entirely static equations. Maybe that’s okay – I think P=F/A and F=MA can’t be thrown out the window entirely – but maybe there are other considerations and equations that come into play in a dynamic situation? And of course, the Newton example relies on a vacuum. No, a feather will not fall as fast as a bowling ball when dropped out a normal window in real life. The feather has more drag and will go slower, just as the water in a 3/4" tube will have more drag and will go slower. The circumference of the 1” tube is only 33% longer but its cross sectional area is 78% larger! So it is easy to see the surface area to volume ratio is much smaller in the 1” pipe and therefore friction from the wall will have a much smaller effect in a 1” pipe compared to a 3/4" pipe.

Maximum possible drain rate for a tank with a hole or tap can be calculated directly using Bernoulli's equation, and is found to be proportional to the square root of the height of the fluid in the tank (above the opening). (Toricelli's Law.)
I’m not sure I can even picture that relationship. Here’s a unitless table to try to grasp the proportions:

Max drain rate | Height of fluid above opening (directly proportional to pressure at the opening)
1.00 | 1
1.41 | 2
1.73 | 3
2.00 | 4
2.24 | 5
2.45 | 6

So with the same size hole, doubling the height of fluid above the opening (thus doubling the pressure at the opening) does NOT double the drain rate? In fact, it only increases the drain rate by about 40%? That’s crazy and nowhere near what I would have guessed before! Pretty neat to learn about that – thanks.

Once pipe is added to the bulkhead, it all goes south. Although a 24" drop (1" open bulkhead) can "theoretically" flow ~1660 gph, because the pressure increases in the pipe, the flow slows (friction loss.) What started as 1660 gph is now ~ 1200 gph. Increasing the pipe size, to 1.5" reduces the pressure in the pipe. Lower pressure = less drag, less friction loss, = higher flow rate, so the value jumps up to ~1500 gph.
So when I said "I really can’t see needing more than 1” diameter unless you want more than 600GPH or have less than 3 feet of drop”, it turns out that is only half the ~1200 figure you came up with for only 2 feet of drop, and 3 feet should pass even more! So my 600GPH rough/anecdotal figure may be quite conservative after all. I’ll chalk that up to dumb luck though for sure – it seems you definitely know a lot more about the complexities of the fluid dynamics than I do. It’s just comforting to see the proper calculations based on Bernoulli's equation aren’t completely at odds with what I’ve seen about how effective siphons can be.

-----------------------------------------------------

So the bottom line so far seems to be: siphons are a very, very effective means of draining that can reduce your minimum pipe size requirements! You might be able to use a much smaller pipe than you would have thought. But, the exact figuring of how much flow you’ll be able to achieve can get complicated. Further, drop height and bulkhead size can be just as important as pipe size.

For instance, though 1200 GPH is a decent estimate of the flow in a 2-foot vertical, straight section of 1” ID pipe, the capacity may only increase ~23% to ~1500 GPH with a 3-foot drop (a more typical height difference between display and sump) if uncleof6 is to be trusted and I've understood him correctly. In reality almost no one will have a simple setup like that and any fittings, valves, bulkheads, etc may reduce the capacity, perhaps substantially, especially if they reduce the minimum ID found throughout the drain.

So, as uncleof6 said, 1" or 1.5" pipe? Depends on what it is you are trying to do.

In any case, it is probably wise to go one pipe size larger than what you think you'll really need (just as you would for a pump). If you're looking at 500GPH and a siphon in 3/4" pipe really ought to handle it just fine, maybe go with 1" anyways. 1000GPH should be comfortable in 1", but 1.5" will absolutely be okay. More than one size greater though and you're just making things more difficult, bulky, and expensive for little or no reward.

[FrostyChicken]

Uncleof6 - You seem well versed on the intricacies of plumbing and the fluid dynamics behind it. The pump I'm looking at using has a 1.5" outlet but the 180 reef ready tank I'll be setting up in about 2 months has 3/4" return lines. In your opinion what is the best way to make that transition with the least amount of flow loss possible?

[uncleof6]

Pretty simple. 1.5" - 2" over the back of the tank. Reducing the line at any point to 3/4" will kill the flow...perfect example of tank manufacuturers not knowing/not caring how something actually works. 3/4" is useless on a tank that size.

[FrostyChicken]

In that case would it be beneficial to use the holes already drilled for the 3/4" return lines as supplementary or backup drains instead of just capping them off?

[uncleof6]

No, because the flow capacity for a 3/4" bulkhead is insufficient for a tank this size.