Silent and Failsafe Overflow System

The probability of failure for me has been high. Some failures were electrical and some mechanical.
The probability of them both failing on the same day or week or while I'm on vacation is significantly reduced.
I have already had it save me from a prolonged outage in the year I've been running both of them at the same time.


What I mean by small is the height and thickness of the internal part of the overflow. I was not referring to the length.
 
Redundancy: a part in a machine, system, etc., that has the same function as another part and that exists so that the entire machine, system, etc., will not fail if the main part fails. E.G. the probability of the "redundant" part failing is at or near zero under normal conditions, and the probability only increases if the "main" part fails. In other words the "redundant part" is not in use.

In a system with both identical pumps running, the failure probability of the pumps is equal. That is not redundancy, as both pumps together are the "main part."

A second backup, or second failsafe is a redundant failsafe, used only if the primary backup/failsafe fails.

These pumps have a "rep" for failing, though it is not so widespread as claims state. Just the same, with both running it is just as likely that both will fail at the same time, as it is that only one will fail. You would need a third pump not running, to actually get to where you feel you are. The third pump would be the actual primary failsafe, not a redundant failsafe.

I would consider it 'partially redundant,' in the same way using 2 smaller heaters instead of one big one. The system may not be fully functional at the same level of flow as it would with both, but it would still have some flow, even if at a reduced level, which beats no flow at all.
 
in the computer world we would call it "hot/hot" or "active/active" redundancy as opposed to "master/slave" or "active/failover" redundancy.

both methods are valid approaches, however sometimes "staggering" components is smart with active/active, if you buy all the parts at once, they have a higher chance to all fail around the same time. in practice this is rarely done and usually any failures would be spaced out enough not to matter.
 
Staggering would make sense if it there were issues with a particular batch of devices since you could theoretically get devices from different batches and avoid having problems with all of them. If the failure is intrinsic to the device, though, the only way it would matter is if the MTBF is accurate to a time frame less than that needed to fix the broken device before the backup/secondary failed (meaning pump 2 would fail before pump 1 could be replaced in this case,) but we digress from the thread...

Relating redundancy to the Bean system, the reason it is so reliable is because it has that redundancy. With a return pump, partial redundancy is acceptable since the system should still survive without any catastrophes at half flow should one pump fail. With an overflow system that's not necessarily the case.

Take the 'wet Herbie' as an example - full siphon on one pipe with a trickle on the second and no dry emergency like the bean has. By design, the siphon cannot handle the full flow of the return pump and will flood if the trickle pipe gets blocked. In other words it has a single potential failure point. Many people argue that this is unlikely to happen and/or take precautions to make it less likely to happen, and are comfortable with the risks, which is fine. Safety is about reducing risks to an acceptable level and balancing the risks, the costs of mitigating the risks and the costs of a failure (i.e. will the flood be in your living room, or in an unfinished basement?) Ultimately you should understand the risks, though.

The Bean takes the Herbie system one or 2 steps further by adding redundancy - both the dry emergency is always ready if needed, and the airline tubing for the open channel can convert it to a full siphon, increasing it's capacity. In this case, both these systems provide full redundancy - the open channel and the dry emergency can potentially handle the entire flow of the system without a flood. I believe that both Uncle and Bean have had cases where both the siphon and the open channel have gotten clogged, meaning they were saved by the dry emergency, so although it may be an uncommon occurrence, it can happen.
 
Its always good to test as many scenarios as you can. Its better to happen while you have some control then when you just left out of town for several days to weeks.

I test by closing up the siphon. Then test closing up just the open. Turning of and back on the return pump.

Then close the siphon and the open. Again turning off and on the return pump.

See what happens.
 
Question For backsump overflow:
Tank 54 x24 x24"
Sump is 6.5" width at the back.
Return Pump eheim compact 2000

Howlong should the overflow surface skimmer should be? How big should the teeth be?


Sent from my GT-N7100 using Tapatalk
 
Question For backsump overflow:
Tank 54 x24 x24"
Sump is 6.5" width at the back.
Return Pump eheim compact 2000

Howlong should the overflow surface skimmer should be? How big should the teeth be?

haha that's almost become a loaded question around here.

my 120g tank is 48x24x24 and i went with 18" long. it's arbitrary at some point though. The best is "as big as you can go" meaning full-length, coast to coast style, or something even more ridiculous like multi-edge overflow, but no one reall ydoes that. I'm sure anything 16 - 24" would be fine in reality, just wouldn't be "quite" as good as full-length.
 
haha that's almost become a loaded question around here.

my 120g tank is 48x24x24 and i went with 18" long. it's arbitrary at some point though. The best is "as big as you can go" meaning full-length, coast to coast style, or something even more ridiculous like multi-edge overflow, but no one reall ydoes that. I'm sure anything 16 - 24" would be fine in reality, just wouldn't be "quite" as good as full-length.
Well mine is backsump, not sump at bottom. So no pipes etc. Can i go 12" wud it be enough??

Sent from my GT-N7100 using Tapatalk
 
Well mine is backsump, not sump at bottom. So no pipes etc. Can i go 12" wud it be enough??

Sent from my GT-N7100 using Tapatalk

it would probably handle the flow, assuming you keep it reasonable. However, the skimming effectiveness will be reduced at that size. You might want to go toothless if possible, to maximum use of the space.
 
it would probably handle the flow, assuming you keep it reasonable. However, the skimming effectiveness will be reduced at that size. You might want to go toothless if possible, to maximum use of the space.
That can cause fishes and inverts going into the sump

This is what i have. 30cm. Ordered from fish street
c18182974da8e4ca927aea2d6d5530c8.jpg


Sent from my GT-N7100 using Tapatalk
 
tried to find this answer as I saw it once before, but couldn't find again.

What sets the height of the water in your overflow? Is it the height of the "trickle" pipe?

Having issues where my overflow water level is too high and basically raises the tank water level rather than flowing through the weir. I'm guessing it's recommended to have the overflow level lower than the tank by 1/4", 1/2"?
 
tried to find this answer as I saw it once before, but couldn't find again.

What sets the height of the water in your overflow? Is it the height of the "trickle" pipe?

Having issues where my overflow water level is too high and basically raises the tank water level rather than flowing through the weir. I'm guessing it's recommended to have the overflow level lower than the tank by 1/4", 1/2"?

open your siphon's valve more?

technically you should be setting that water level with your siphon valve. opening and closing the valve on your full siphon will be what sets the water level in there. The trickle pipe will ultimately set the high point though.

too high is when it floods on your floor, imho. I don't think it matters whether the water level is higher or lower, as long as it's got margin that it's not close to spilling out, that's all that i care about.
 
Normally it sits like this.

5764f59001e2eb894fe1f4bd0979d7fe.jpg


Crack it a hair open, and I mean just a tiny tiny hair on a decent cepex calve and after about 5*+ minutes it eventually get to here and starts gurgling. I'm thinking my trickle tube on the left is too high?

3303093c91bfc8850e5acf12aa4f0869.jpg
 
Both your OC and siphon are way too high IMO

Siphon can be all the way at the bottom of your box. Literally, just enough pipe (black, in your case) to insert into the bulkhead and then into the first 90 and that's it.

OC can be just a hair above it or even at the same level. But I don't see an airline on your OC?? That is what keeps it from siphoning. You run the airline up to just below the emergency and that kicks the OC over to siphon when the level gets close to the emergency pipe
 
gotcha, thanks Floyd!
I had them higher because I didn't want my whole overflow to drain down on a power shutdown (it's a 10' long overflow so it holds a decent amount of water).
I'll get that anti-siphon tube done this weekend, just need to hit up bean's website for the parts.
 
not sure what you mean, i.e. why you are worrying about draining the overflow box?

BTW you have a seam separating in those pics. Vertical seam to the left of the OC.
 
yes, i didn't want to drain the entire overflow on shutdown.

What do you mean by vertical seam? an issue with my tank or how the pics appeared on the page?
 
yes, i didn't want to drain the entire overflow on shutdown.

What do you mean by vertical seam? an issue with my tank or how the pics appeared on the page?

i agree it's not worth draining that whole chamber on shut down, at least not on an internal overflow. on an external that's not as big of a deal.

again, i'd ditch the double-90s entirely on the siphon line. All you need is a strainer, and it can be 6" from the top, so it wont drain the whole overflow, just 6" worth.

Is there an airhole drilled into your open channel? It's hard to tell from the pics.

I think he's referring to that seam to the left of your open channel, whatever that acrylic thing there is.
 
The steady state level in the overflow is determined by 2 things: The level of the open channel and to a lesser extent the valve on the siphon.

Consider how the Bean is designed to operate - the siphon takes the majority of the flow with the open channel taking the flow above what the siphon can't handle.

Imagine that the siphon is established and the siphon flow and the return pump flow were perfectly matched, the water level could be anywhere between the inlet to the siphon and the ledge of the open channel (the point at the top of the inverted U where the water just makes it over the top to go down the standpipe.)

Now consider that the flow is slightly greater than the siphon can handle. The water level will rise until it starts making it over the ledge of the open channel. The more flow the open channel needs to handle, the higher the water will rise. This will continue until the flow in the open channel is 'maxed out,' at which point the water level rises to the level of the dry emergency.

For an external overflow like you describe, you don't necessarily need the inverted U, but if you omit it, the opening to the siphon pipe needs to be far enough below the surface that it doesn't entrain air. As Outdrsyguy1 noted, the further down in the box the siphon opening is, the more water will drain out when the power is cut. As long as your sump can handle the volume this isn't an issue; it should start back up again either way, but if you are interested in minimizing the amount that drains down, the inverted U will allow the opening to the siphon channel to be placed higher in the box than a straight open pipe.

I agree with Floyd - the pipes are too high in the box
 
thanks a ton everyone, i need to tweak it and see how things go. I'll drill the siphon line hole and install the fittings and move things down a bit.
 
Back
Top