Plumbing 180g FW to 65g sump

[Jimothy]

Yes! Too powerful!
I think that point has drifted aside while I worked on other things. Back to it: whether or not I can return the magdrive24 (waiting to hear back from merchant) or recoup some of its value on craigslist I will need a less powerful pump. For the brand: I am willing to pay more for silence, though I don't know if I have 2x more flow than I need or just a couple hundred gph to lose...
A post in hardware hasn't yeilded any advice on make/model...

Siphoning from the bottom is just one of the funtions of the U pipe, the other is to get the water moving back into the sump because of the 2 inches (10 gallons) of rise in the display tank needed to get to the OF box so the sump doesnt run dry.

Adding a second U helped a little, so I am building a 3rd and 4th right now... pics in 30 min or so...

 

[jsoong]

Your baffles looked fine to me too. On the pic, it almost looked like there wasn't a gap at the bottom of your "return/filtration" compartment causing all that water to spill over the 1st baffle. Well, all the water going over top of that compartment would indicate that your pump flow is too much for your sump! You are going to have to use a smaller pump or set up that diverting T you had before. <BTW - if you can set up the diversion easily, it would be a good way to find out how much flow your system can handle before going out to buy another pump.>

As for the overflow on top, you can't have water spilling over the top since that would have too high an operating water level in your tank. At your first test, your pump was drawing air because you can't keep enough water in the sump at your current overflow level. My suggestion was to lower the operating water level so you don't drain down as much water into the sump at power off, allowing you to keep the sump fuller. If you want to take water from the top as well, you will need to make a skimmer box at the lower water level & connect that with a siphon. I am not sure if that is worthwhile doing. Your problem is the overflow box being too high compared to your platform, it might be simpler to re-do the overflow box & dispense with all the siphons....

But since you are already making more siphons, you may as well try that & see what happens.

 

[Jimothy]

took longer than 30 minutes

I will start the system at 9, giving the glued parts more time to dry...

 

[Jimothy]

OK, ready to go, everything set except pump running.

Here are photos of water levels when pump is off:
(BTW, getting both durso pipes to exactly the same level as minimum display was tricky, but testing without the U siphons shows that the 2 levels are equal)

I do not intend for the water to not spill over the OF, the U siphons are meant to pull from the bottom of the display and help to ease the transition...
Let's see what happens................

 

[Ziggy953]

How is it running with the modifications?

 

[Jimothy]

Internet went out last night before I could post...

The same basic problems occur, but slower...

The pump eventually runs dry, but it takes much longer than before.

The water runs over (instead of under) the first baffle, but it takes longer to happen and not as much as before.

The water goes over the overflow, and the IN flow is more than the 1" and 1.5" durso pipes can send back out. Since the pump can run longer before it runs dry, it continues to raise the display level, it got about 2" above the OF before it ran dry

So...
-If the durso pipes are at full siphon, how much flow is leaving the tank?
-This would be my maximum return flow?

The 2400 is too powerful, and noisy, I will replace it with a brand as close to silent as possible, what would you estimate the gph to be?

 

[jsoong]

I think it would be worthwhile to consider the parts seperately to get an idea of what each can handle. I will give you a theoretical approach for the estimating, hopefully saving you a bunch of experimentation.

I would use the laminar flow equation to estimate what flow you will get. <I know the flow won't be all laminar, but the turbulent equations are too hard to use, and we are estimating anyway!> This is the equation:

Flow = (Pi x r^4)/(8 x n x l) x dP

r = radius of your pipe, l = total length of pipe, n = viscosity of water (constant), dP is the driving pressure - or how high your water level is to drive the given flow.

Instead of trying to figure things out by plugging in the values directly, I would suggest taking an observation & using the equation as a proportion. This would account somewhat for the elbows, constrictions, & other imperfections in your piping that causes turbulence.

Since the Dursos appears to be the bottleneck, we will consider that first.

You need to measure what driving pressure is needed to handle your pump flow. To do that, you need to add more water to the sump so your pump won't run dry, and measure how high the water is in your tank once you get everything balanced. dP is the distance from the top of the water to the inlet. < Yes, you will have more water than is safe .... so make sure you do this outside if possible .... or be ready to plug both standpipes before you turn off the pump & drain out excess water in the tank. >

Once you know that height (we know that it is at least 2" above your overflow top), you can figure out what flow the system can handle at your planned water levels, like this:

Since r, n & l are unchanged, you can simpify the equation to:
Flow = dP/R ...... R = resisitance of the setup.

Using 2 sets of equations for the 2 flow conditions, combining the 2 equations ....

Flow1 / Flow2 = dP1 / dP2

Say Flow1 is your Mag 24 at 2000gph, and your driving pressure dP1 is 4", and you have only enough height in the setup to give a driving pressure of 1" (dP2) ....

Plugging everything in gives you a Flow2 of 500gph .... this is what gph your setup can handle. <I would not run a pump right at the edge of system capability .... cut back 25-50% to handle the invariable glitches during operation.>

Of course, I am only guessing at the numbers by looking at the pics. Try with real measurements & see what you get. The Dursos should be able to handle more than 500 gph, even at 1" of head height .... so you might be able to tune the system for better performance <eg - changing the air hole size>. If you can't get the Dursos to handle an acceptable flow rate, you'll have to consider using an open channel design <in an open channel or full siphon setup, dP is the height between your overflow & your sump .... so can handle much higher flows>, or re-designing the overflow. The other option - dropping the height of the Dursos to gain more head height is not safe with the siphons taking water from the bottom - since the minimum water level would be too low & you'll flood the sump with a power out.

 

[jsoong]

One other thought came to me overnight ....

Your pump is in a small compartment, giving you very little water to work with. For testing, why not move the pump into the main "fuge" compartment. This might allow you to run the system & get your levels without the pump going dry.

In real operation, you would want the pump in the last compartment as those baffles acts as air & dirt trap to reduce clogging of the pump (& bubbles in tank).

 

[Jimothy]

I am installing a ball valve in the return line to get a rough idea of how many gph I want in a new pump, thanks for the formulas, that will help to confirm my finding or maybe narrow it down further.

The problem with the open pipe is that the water level fluctuates between a full siphon and a gurgling drain until water builds back up to create a siphon which drops it again bringing air into the pipe...

Though if it offers much more flow maybe I should have 1 open and 1 durso?

--->The hole in the top of the durso, I thought this was to vent air as the water levels moved, would a larger hole lead to faster flow?


...I will see if there is enough room to move the pump to the fuge compartment as well

 

[Jimothy]

I can't move the pump's position within the sump but...

With ball valve half closed the systems runs!

The display level gets to 1/4 inch above the OF...
The teeth I am adding are about 1/2 inch tall so there should be enough flow through the U siphons with some good skimming of the surface...
The durso pipes, they don't flow much with the holes open, so after plugging the big one with the tip of my finger and seeing the OF level instantly drop I realize my holes are too big...
Little bit of masking tape and a pin to slowly widen a hole in it led to a alternation between full siphon and a slow fill up of the OF box, full siphon and then slow fill...
I am sure a balance needs to be found but I am happy to just see things finally doing what they should.

Water is still coming over the first baffle, I think the only solution is to add height to it? Silicone sealant and some plexiglass? Silicone sealant is also how I am gluing the teeth on?

GPH... Hard to get more detailed than 'ball valve half full'
I can't really adjust anything once it is running, or make any slight adjustments for that matter, I can only be pretty sure that it is closed exactly halfway...

So, I plan for a gph that is slightly overpowered, and get the T off valve set up again to fine tune the flow.
Do you think it should be more like 1200 or 1600?
Do you have any advice on tweaking the dursos?

 

[jsoong]

I found the same thing as you with the Durso .... It is rather difficult to tune. If you plot flow vs hole size, I would think that it would be a "U". Flow will be highest with no hole (full siphon) and max size hole (full open). Unfortunately, noise will be lowest only at the min flow. The trick is to find the right size Durso so that the min flow is greater than your pump flow. <Note that you can use bigger size pipes for the Durso than your bulkhead ... you connect it with a reducer .... It doesn't matter so much with the bulkhead size as the bottle neck is the Durso mechanism.>

I have not been able to get a Durso working as quiet as I like. I get the same problem as you. With a small hole (and low noise), the flow is not high enough so water backs up and convert the Durso into a full siphon. At this point, the flow is too high, and it sucks the water down & make a big slupping sound. The level drops, converting back into low flow Durso ... the cycle repeats .... The slupping noise is worse than running the system open. That is the reason I will be trying the full siphon with an emergency open setup. <People with sucessful Durso I think have the water fluctuating between the middle of the T & the middle of the downturned elbow .... I can't seem to find the right size hole to do that ....>

Estimating pump flow with a ball valve is tricky. The flow vs reisitance curve is not linear, so 1/2 open valve does not equal 1/2 the flow. One way to find out is to do the t-divertor setup & measure the flow in the diverting limb. Another way to estimate is to look back at the flow vs head height table. Take the gph rating at 1/2 the max head height (minus current head height) and that might be where you are at. [Looked at the table & it is around 800-1000 ghp for the Mag 24.]

I wouldn't do anything to the sump baffle at this point. Do that once you have all your flows settled (ie pump, overflow, etc.) or you end up having to redo things. You can raise the 1st baffle height - as long as you have enough room on top for water to spill over (safety in case of blockage). <Otherwise you will need to increase the gap at the bottom & maybe increase the space between the 1st & 2nd baffle ... that means taking the thing apart ..... >

 

[jsoong]

Thought of a better way to figure out your pump flow.

Turn your pump off & let water drain down to the sump. At this point, the water level should be higher than all the baffles. Turn pump back on & note the speed the water level goes down. <Do the timing before the water starts coming back from display and before the water level drops below the baffles.> From this, you should be able to calculate the gph:

GPH = Sump surface area (sq inches) x Speed water level drops (in per minute) x (60/231)

 

[Jimothy]

Speed water level drops... This means if it drops 1 inch in 3 seconds the number I use is 20/1?


Thought of something... Because my 2 standpipes come together before dumping into sump, wouldn't they both have to be full siphon for one of them to be?

 

[jsoong]

If it drops 1" in 3 sec, then yes, the speed of drop is 20" per minutes.

Depends on how you connects the 2 standpipes .... Generally, it is better to have the 2 completely separate, since any constriction in the connection will degrade the performance of the standpipes. The least resistance is a connection using a Y. A T connector adds a fair bit of turbulence & you might get into airlock & other problems, but I would think that with your pipe size, it is mostly an issue of noise.

I think you can get a full siphon on one of the standpipe as long as the connection is well below the display. You only need a siphon to draw water pass the inverted U. Once it get pass that, the water just falls. An opening below that level should make no difference.

 

[Jimothy]

Yes, the undertank plumbing is not ideal, though it works and was difficult to build and connect, the only reason I would change it is if I can't find a suitable silent submerged pump and need to change to an inline.

Did two 1" water drops in the sump, one was 10sec and the other was 11

10sec ---> GPH = surface area (18x48 x water drop (6/1) x (60/231) = 1346.5

11sec ---> GPH = surface area (18x48 x water drop (5.455/1) x (60/231) = 1224.2

So I want to find a pump that, after head loss, pumps 1300gph?


BTW... The U's I created for the OF work great in pumping all sump water into display to be siphoned out.

I have help today moving the tank aside and stand outside to be sanded and finished, then back in on wednesday.

Going to try again in hardware for recommendations on silent 1600(?)gph pump...

 

[Jimothy]

Looks like Quiet One pumps are quieter and more energy efficient than MagDrive, and cheaper at foster & smith. Am I missing something?

If not then I will get the 6000 or the 9000...
The 6 has 1500gph
The 9 has 2300gph

Do you think that 1500gph will work out after head loss?
I know that 2300 is nearly as powerful as my current MagDrive, but it appears to use less power and if it is silent I can T off the extra flow, I just don't want to make the mistake of ending up with not enough...

 

[jsoong]

A lot depends on how much water flow you need.

This is the flow table:
Quiet One Pond Pumps

Your head loss will be the height difference between the tank water level & the sump water level. There will be some loss from the pipe/ spray bars, etc. but that will not be too significant .... may be 10% less.

Looking at the 6000 table, it shows flow of around 1000-1200 gph at 3-4'. Which I would guess to be close to your head loss. Is that kind of flow rate enough for the turtles?

A smaller pump will be quieter, but it will be less flexible if you are running it at its limit. Eg. you need a new pump if you want to redo the plumbing & go for higher flow. OTOH, the 9000 will still be running over 2000 gph with head loss ... That's a lot of flow. Too bad there isn't an in between size.

 

[Jimothy]

Yes... too bad...

Again, my #1 concern about a new pump is silence, not many advertise the level of noise produced. I read that some of the external ones, like Reeflo, can only be heard a foot or so away. I am willing to go in-line if the right pump for my needs (silence, 1300gph, magnetic drive) turns out to be in-line and not submerged.


Not really sure how much flow the turtles 'need'
Their top needs are diet, basking temp, and UV light, all taken care of.
I don't think they would like living in a rushing river but I am not too concerned, they can find places to sleep like on or under the ramp just fine.


The 'right' flow is one that keeps water moving toward the overflow and U siphons and doesn’t let waste build up. I would place the height at 5', and a T off with spraybar might increase this to 6' (?), making my flow 900gph, do you think this is enough to keep water moving across a 6' long, 2' wide tank?

 

[jsoong]

900 might be a smidge low .... You might get some dead spots in a big tank. I know reefers aim for 12-15x water turn over, and since your tank is 1/2 full, I am guessing 1000-1500 gph to start.

Since you have the Mag24 set up, why don't you try chocking the flow down to 900 & see if that is acceptable? <May be put a pinch of fish food or some other small particles here & there & see how fast it gets disperse into the water column.>

 

[ThatGuy619]

well with the amount of bends you have its definitely not 2400. for my reef setup i have a mag 9.5 on my return for my 100g tank. for my 200g cube im building im going to use a mag18. my mag 9.5 is dead silent. a mag 9 would be perfect for turtles. i have a 210 turtle tank with 2 fx5's running it.