Dual Return Pumps

[reely989]

I like running two returns. I’ve had pumps act up or whatever and the system keeps chugging. Hell I also keep a third one as a backup in case one fails and I need to swap one out.

 

[xxjt2oxx]

So here is a rough idea of my plumbing. Yellow (3/4") is the front side and pink (3/4") is the back, and they're tied together on the green (1") and down to the pump.

I have an Eshopps Prodigy L on each end of the tank, so the drain pipes from that go straight down and I lose the entire center ~12" of the 24" end. I keep the return lines far enough away from the tank to fit in my Vortechs, which is where space gets tight. The bulkheads for the returns are all 3/4"

That all being said, I never thought about the 1" barb fitting being the current bottleneck. I have the 4 ball valves, and interestingly I have noticed that I don't get much of an increase in flow when I have all 4 wide open vs. 2 open/2 closed, so there may be enough untapped potential in changing that fitting to at least get me to a happy place for now.

 

[UncommonSense]

So here is a rough idea of my plumbing. Yellow (3/4") is the front side and pink (3/4") is the back, and they're tied together on the green (1") and down to the pump.

I see room for even further improvement here, just messing with the three return tees under the tank:

if you can step up to 1.25” Sch.40 PVC (using a 1.25” BSP to 1.25” NPT adapter fitting at the pump, or a 1.25” BSP to 1.5” barbed fitting), that can be reducing tee-d to two 1” sch.40 pipes, which are then themselves reducing teed to your four 3/4” sch.40 pipes…

This would be a perfect-world plumbing scenario which should get you into the realm of realistically moving 10X turnovers that you’re targeting!


However,

have an Eshopps Prodigy L on each end of the tank

These are only rated for 900GPH each!

So, if you use the above plumbing arrangement, reducing 1/4” in pipe size at the outlets of each tee, you will be physically capable of pumping more water into the tank than the drains can handle!

(not with your current pump [<1,800GPH observed after just head loss], but with a bigger pump, this would need to be a consideration!)

 

[xxjt2oxx]

Are the ends of the tank the parts that are "in wall"? No way to go over the overflows?

Eliminating the 3/4" run will help, but 1" is just a lesser problem.

Revising my calculator estimate from earlier to use 1"still gives +5 feet of head pressure.

May not get the target flow with 1", but seems like it should still be enough.

It would also be interesting to me if you were to measure your actual flow rate now before making any changes. Calculations and estimates are nice, but...it's nice to know for sure sometimes. How much flow are you really currently getting?

I see room for even further improvement here, just messing with the three return tees under the tank:

if you can step up to 1.25” Sch.40 PVC (using a 1.25” BSP to 1.25” NPT adapter fitting at the pump, or a 1.25” BSP to 1.5” barbed fitting), that can be reducing tee-d to two 1” sch.40 pipes, which are then themselves reducing teed to your four 3/4” sch.40 pipes…

This would be a perfect-world plumbing scenario which should get you into the realm of realistically moving 10X turnovers that you’re targeting!


However,


These are only rated for 900GPH each!

So, if you use the above plumbing arrangement, reducing 1/4” in pipe size at the outlets of each tee, you will be physically capable of pumping more water into the tank than the drains can handle!

(not with your current pump [<1,800GPH observed after just head loss], but with a bigger pump, this would need to be a consideration!)

2000gph was kind of a rough goal - I feel like the Prodigy would be very difficult to keep silent at full capacity, but that's just my opinion. Realistically I would probably be coming in around 1400gph as my regular use case top end flow.

Long term, I think I'm going to run 2 smaller pumps - one to the front returns and one to the back. I need to look at some of the fittings and really size up the barb section. I like the idea of being able to survive if I'm away or on vacation and not able to get the pump swapped in a timely manner.

 

[UncommonSense]

Long term, I think I'm going to run 2 smaller pumps - one to the front returns and one to the back. I need to look at some of the fittings and really size up the barb section. I like the idea of being able to survive if I'm away or on vacation and not able to get the pump swapped in a timely manner.

In this case, I’d say stick to two 1” pipes, one for each pump… this way, you can omit the check valves completely by having each return pump on fully independent plumbing!

If vibration mitigation is required, you would 1.25” tubing/barbed fittings for each pump to maintain the usefulness of 1” pipe; moving 700GPH+ with minimal frictional flow losses! (Basically, maximum observed GPH per-watt of power consumed)

 

[xxjt2oxx]

In this case, I’d say stick to two 1” pipes, one for each pump… this way, you can omit the check valves completely by having each return pump on fully independent plumbing!

If vibration mitigation is required, you would 1.25” tubing/barbed fittings for each pump to maintain the usefulness of 1” pipe; moving 700GPH+ with minimal frictional flow losses! (Basically, maximum observed GPH per-watt of power consumed)

We have it in between the main entryway of the house, and a playroom that is eventually going to be a "wellness" room so my wife can practice out of the house - so sound level/vibrations to a minimum is important.

Will probably try to get this done next week and will come back with an update :)

 

[mcarroll]

So here is a rough idea of my plumbing. Yellow (3/4") is the front side and pink (3/4") is the back, and they're tied together on the green (1") and down to the pump.

I have an Eshopps Prodigy L on each end of the tank

As noted elssewhere, this is rated for up to 900 GPH. I don't know how they calculate that, but it looks like three 1" gravity drains. 1" gravity drains can handle up top 600 GPH each, but are only silent up to about 50% of the rating....so ≤300 GPH is the usual target flow for a single 1" drain.

You might have to experiment with yours, but I'm guessing there will be a limit well south of 900 GPH where performance suffers due to noise or other factors.

(You don't run drains at 100% of their rating for other reasons too.)

so the drain pipes from that go straight down and I lose the entire center ~12" of the 24" end. I keep the return lines far enough away from the tank to fit in my Vortechs, which is where space gets tight. The bulkheads for the returns are all 3/4"

As long as you limit the 3/4" to the ends of the plumbing run (where the water enters the tank), they actually aren't much of a restriction. A single 3/4" outlet with 1' of 3/4" plumbing would be able to carry your 20 GPM (1200 GPH) without much added friction loss. Less than +1 foot of head pressure.

So re-plumb the pump with its ideal plumbing size (1.25 inch?) and only adapt down to 3/4" at the last possible moment. Eliminate as much of the 3/4" (and even 1") as possible.

Further mods would be unnecessary or optional.

2000gph was kind of a rough goal - I feel like the Prodigy would be very difficult to keep silent at full capacity, but that's just my opinion. Realistically I would probably be coming in around 1400gph as my regular use case top end flow.

As I said you may have to experiment to find the actual limit of your drains where they will run silently/in a generally acceptable fashion, but I would not expect to get much more than 1000-1200 GPH of actual flow through both drains. (That way you'll be happy if you do get more, but not bummed if the limit is <2000.)

Long term, I think I'm going to run 2 smaller pumps - one to the front returns and one to the back. I need to look at some of the fittings and really size up the barb section. I like the idea of being able to survive if I'm away or on vacation and not able to get the pump swapped in a timely manner.

Not a bad idea, but IMO don't skimp on the pumps....use same-or-better quality as what you have now.

After all, you aren't trying to ask for a higher failure rate just because you have redundancy. :)

For what it's worth, redundancy is not a perfect strategy....twice as many devices means, among other things, doubling the chances of failure, whatever those chances are. Often one high-quality pump really is ideal, and it can be for more significant/practical reasons than redundancy. Check out this link:
https://en.wikipedia.org/wiki/Redundancy_(engineering)

 

[UncommonSense]

1" gravity drains can handle up top 600 GPH each, but are only silent up to about 50% of the rating

I think these numbers may come from the drainage industry, wherein the max flow rate is measured with the pipe ~97% full of water, and a small residual air gap…

In the case of a multi-drain aquarium plumbing configuration, the main drain is typically run at a metered full-siphon, dramatically increasing maximum flow rates!

(My hexagon tank, for example, is getting in excess of 600GPH of observed return flow, and its 1” main drain has its associated gate valve opened maybe 50-60%!)


In the case of the prodigy L overflow, and its 1” drains, I suspect that 900GPH is actually a safe/silent operating limit… I’m unsure if the bulkhead holes through the tank wall are the limiting factor for flow, but a 1” sch.40 PVC drain at full siphon should be capable of a bit more than the 900GPH the manufacturer rates them for!

 

[mcarroll]

I hear you, and as I said I don't know how they calculate it, but Eshopps says "up to 900 GPH"....they don't say it how you said it.

Further, it's only rated up to 120 gallons suggesting a target "normal" flow rate somewhere around 400-600GPH, IMO. (120 gallons is only going to need 120-480 GPH return flow in most cases....Eshopps doesn't seem to suggest a target flow unfortunately.)

Even if you wanted to suggest running a drain system at 100% of capacity, that's only 1800 GPH for both units combined – still not quite the target flow.

However I don't like to suggest that since running drains at 100% leaves no margin for error. If something happens to one of the drains, the other has insufficient (maybe zero) excess capacity to keep the system running. (ie we create a flood potential that doesn't need to exist)

https://www.eshopps.com/product/prodigy-l/

SPECIFICATIONS​
Item#​	95858
Dimensions​	Internal: 12×5/8×6″ (305x16x153 mm) External: 12x3x6″ (305x76x153 mm)
Flow Rate​	Up to 900 GPH (3,407 LPH)
Intake​	70-120 gal (265-454 L)

 

[UncommonSense]

However I don't like to suggest that since running drains at 100% leaves no margin for error. If something happens to one of the drains, the other has insufficient (maybe zero) excess capacity to keep the system running. (ie we create a flood potential that doesn't need to exist)

Ah, this is why each of these two overflow boxes has double redundancies; each box will have a 1” secondary drain standpipe higher than the primary drain, and an additional 1” emergency drain standpipe higher still!

— this would be a typical three-drain BeanAnimal overflow (just two of them, so 6x 1” drains in total)!

The 1” primary drain (AKA siphon channel) is, as the name suggests, under full siphon…

The 1” secondary drain (AKA open channel) is ready to accept the full return flow rate if primary drain ever clogs…

then, the third 1” emergency drain (AKA emergency channel) kicks in if both the primary, and secondary drains manage to clog simultaneously!

(And, there’s two of these triple drain overflows on this tank, both operating in the same way!)

 

[mcarroll]

Ah, this is why each of these two overflow boxes has double redundancies; each box will have a 1” secondary drain standpipe higher than the primary drain, and an additional 1” emergency drain standpipe higher still!

— this would be a typical three-drain BeanAnimal overflow (just two of them, so 6x 1” drains in total)!

The 1” primary drain (AKA siphon channel) is, as the name suggests, under full siphon…

The 1” secondary drain (AKA open channel) is ready to accept the full return flow rate if primary drain ever clogs…

then, the third 1” emergency drain (AKA emergency channel) kicks in if both the primary, and secondary drains manage to clog simultaneously!

(And, there’s two of these triple drain overflows on this tank, both operating in the same way!)

I'm not a believer in the necessity for these kinds of drain setups. IMO they do not offer any real benefit to regular reef tanks.

In the days of closed loops and using return pumps for tank flow (around 2000) they made some sense – folks were driving sump flow rates to 100+% on many cases – there was no margin for error, especially if you only had one drain box, as was somewhat common back then. So various improved drain systems were invented (BeanAnimal was one) to contend with these non-ideal circumstances. Thankfully powerheads have more or less eliminated this usage.

Tanks with two drain boxes (ie Reef Ready tanks, et al) already have failure redundancy. As long as the "rule" of using 50% of the drain flow rate (which is also conveniently the rate where the drain stays silent) is adhered to, the remaining drain can always take the failed drain's flow – plus that drain gets really noisy so you have an audible alert of the failure state.

But if each drain is maxed out to 100% (BeanAnimal or otherwise), you get a flood if one of them becomes even partially blocked.

(Pretty sure you know this; stating it for the OP)

Hopefully the OP will experiment as suggested.

 

[UncommonSense]

But if each drain is maxed out to 100% (BeanAnimal or otherwise), you get a flood if one of them becomes even partially blocked.

I’m not sure you fully understand my above reply!

There would be no situation where normal operation would max out all three drains in a BeanAnimal configuration, outside of trying to out-flow all three drains with massive return pump(s)…

I’m saying that ONE 1 inch schedule 40 PVC drain (out of the three total 1” drains in one overflow box) in full siphon is capable of silently flowing the 900GPH the overflow box is rated for.

The secondary or emergency drain only ever accept full flow if the lower standpipe(s) are completely clogged!

— The confusion here, as stated above, seems to come from the use of drainage calculations for plumbing open to atmosphere, vs at full siphon with no entrapped air! (You’re getting a much lower max GPH number than the real-world shows is possible, and very reliable!)


(And keep in mind that even an open pipe with no metered restriction in the form of a gate valve will form a full siphon once water packs up enough head pressure above its inlet… this siphon isn’t continuous, it starts and stops repeatedly due to the lack of the metered restriction, but it will dramatically increase the open pipe’s drain rate during each siphon even, inherently preventing a room flood!)

 

[Privateye]

Just a note about the friction loss calculations: because the line is split 4 ways, after that point you would divide the total flow by 4 (if we assume all lines are equal) and calculate the head loss based on that flow rate.

Prior to that point, you'd use the total flow rate.

 

[xxjt2oxx]

There’s a lot of information between here and my last post, but based on first hand experience with the Prodigy L, the limiting factor is not the 1” pipe, but how fast the water can flow through the weirs and into the back box and how quiet you want it to be.

I’m pretty sure the single 1” siphon unrestricted will be able out perform the how fast it can fill, and if you drain the box too fast, the water flowing in from the tank drops in and is noisy.

Once I get the plumbing/pumps situated I’ll test it out!

 

[mcarroll]

I’m not sure you fully understand my above reply!

There would be no situation where normal operation would max out all three drains in a BeanAnimal configuration, outside of trying to out-flow all three drains with massive return pump(s)…

I’m saying that ONE 1 inch schedule 40 PVC drain (out of the three total 1” drains in one overflow box) in full siphon is capable of silently flowing the 900GPH the overflow box is rated for.

The secondary or emergency drain only ever accept full flow if the lower standpipe(s) are completely clogged!

— The confusion here, as stated above, seems to come from the use of drainage calculations for plumbing open to atmosphere, vs at full siphon with no entrapped air! (You’re getting a much lower max GPH number than the real-world shows is possible, and very reliable!)


(And keep in mind that even an open pipe with no metered restriction in the form of a gate valve will form a full siphon once water packs up enough head pressure above its inlet… this siphon isn’t continuous, it starts and stops repeatedly due to the lack of the metered restriction, but it will dramatically increase the open pipe’s drain rate during each siphon even, inherently preventing a room flood!)

I think I do understand, and same for you; I suspect we're just talking about two different aspects.

Maybe this is still interesting for the OP (I hope!) so I'll keep going. LOL

(Sorry for the length....)

It seems like you're imagining one of the drain pipes inside a Prodigy L becoming blocked.

I'm not sure how that would happen in a system like the Prodigy L, but I'm sure it's possible...just not likely.

We're talking about the Prodigy L drain system as a whole (or anything similar) and asking what is the side effect of slowing (or blocking) half of the drains in a system that uses two operating NORMALLY at 100% of their maximum capacity? (Ie. to reach as close to 2000 GPH as possible)

First realize that the pipes themselves don't clog 99.999% of the time because the drains are designed with the teeth as a pre-filter. Anything that can get through the teeth will easily fit down the pipes without clogging.

But just like a trash cage/rack in a lake or pond overflow that catches larger debris...

...the teeth of the drain get clogged (not the pipes) which then causes the drainage rate to slow which causes water to back up.

That's flood potential.

Take a look at a practical example of the filter teeth getting blocked in this video. (I can't seem to find any aquarium videos of this scenario....I think it must virtually never happen, but I dunno maybe it's Google bias):
fast forward to about 7:30 to see the drama unless you want to hear all the explanation, which is relevant but lo-o-o-ong


On the Prodigy L and most other similar drain setups (from aquariums up to lakes) the weir teeth (or trash cages) are designed to catch blockages *before* they reach the drain pipes.

So normally the teeth will be the location of any actual slowdown or blockage. (If that's not true then we should have been designing drains differently all along.....but this is an almost-universally common drainage design for anything from lakes to tanks.)

So, we take the location of a likely blockage in mind (the teeth), and Eshopps "max 900 GPH" rating as given.

We're running two of Prodigy L's at 900 GPH for a total potential flow of 1800 GPH.

When a blockage happens (partial or total), it reduces the available flow to *all three pipes* in Prodigy L.

If the second Prodigy L was already running at 100%, where is the "leftover" flow from the slow/blocked drain gonna go?

At this point the problem is mathematical – how about an analogy?

We now have too many apples to fit into the baskets we have available (one basket is too small). Now some of the apples are going to be outside the baskets. We want the apples *in the baskets* at all times. ;)

Translate apples to water.

We don't run aquarium drains at 100% (anymore) simply because it eliminates our margin for error creating potential for bad things to happen where there need not even be a possibility. (Pretty much regardless of design or the number of pipes involved.)

"Lower" flow through a sump is better in all ways than "maximum". Keeping drain flow under 50% of maximum drain capacity is normally not an issue for anything....and it gives you exactly the margin for error you want in case "something" happens to your other drain.

Having two drains is the redundancy that matters in this case, not really how many pipes in each drain.

Each of the two Prodigy L's should be able to carry the total sump flow by itself. (Meaning ≤50% drain utilization by design.) You can't get 1800 GPH *and* have a margin for error.

Again, drain systems like Herbie or Bean were invented for applications in the 1990's like closed loops or return pumps that were also providing flow for a reef.....applications that have been replaced by a much better alternative.

There's generally no practical reason to have more than about 1-4x flow in a sump anyway....which I'd guess is the reason Eshopps spec'd the Prodigy L for up to 120 gallon tanks and not larger.