Karim's 1500gal dream reef

I modeled the vacuum fill rate and combined it with the flow on release = created a little dashboard for design options

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/0_zps0skzplop.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/0_zps0skzplop.jpg" border="0" alt=" photo 0_zps0skzplop.jpg"/></a>

based on that, I changed the connecting PVC structure to be 13 x 3" PVC
 
<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/1_zpsyeew4yzx.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/1_zpsyeew4yzx.jpg" border="0" alt=" photo 1_zpsyeew4yzx.jpg"/></a>

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/2_zpsbftliddf.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/2_zpsbftliddf.jpg" border="0" alt=" photo 2_zpsbftliddf.jpg"/></a>

basically, the best vacuum I can find has a hard time pulling more than 24" up. I could push it but after about 30 seconds, the flow rate drops so low, that having bigger more expensive tanks is not useful.

Same goes to for the PVC. With lower height and reduced starting velocity, the 4" PVC isn't critical any more. Straight 3" works just as well. Given the density of holes though, I'm going with a double plywood back with epoxy instead of just glass. There's still a back glass frame that bonds to the sides though.
 
I have a philosophical question now. I've designed to get a balance of both peak flow rate and highest average flow rate. The cycle time is very short (30sec up, 15 sec down) at 45 seconds because of this. The compromise is on the surge volume and displacement. Right now, the displacement distance is 2ft. That's not too bad.
With larger tanks and more time, I could get up to 10ft of displacement. That would take a fill time of 15 min and multiple vacuums. It would also change the peak flow substantially.

I'm not sure if more volume motion with higher peaks is better or if maximizing the average flow rate is better. It might be ok for the average flow to be lower as long as the peak flow is high enough. Not sure there's research on coral needs around this. It's not a critical design concern unless I choose to make the tanks larger to accommodate more flow volume.
 
Taricha. I went back to your derivation and mine. It lines up except for one part that I may have gotten wrong.

When I start my system, I have two tanks at two different heights. They're both connected to the display tank at a third height. In my practical case, we'll call the display tank height = zero (open air). The first tank has a vacuum so the water is at a height +h. The second tank is pressurized with air so that the height is -h. The delta between the two tanks is 2h and the settling point is zero. This is the "balancing scenario"

The "surge scenario" is just one tank at height +h that opens to the tank at zero. This settles to a height of zero also but with a delta of h.

In deriving Torricelli's equation for the two different systems, I get

v = sqrt (2g x delta) = sqrt (2gh) for the surge scenario

&

v = sqrt (g x delta) = sqrt (2gh) for the rebalancing scenario

Note that the rebalancing is missing the "2" in the sqrt and that shows up only because the height delta is doubled. This makes sense to me.

Then I go back to my differential equation starting conditions that I use to map the circuit model back into the fluid model. This is where I think I have an error. I'm using v= sqrt(2 g delta) in the rebalance scenario and delta = 2h. That means I'm off by a sqrt(2) in my calculation of v which then messes up all the rest of my coefficients.

That's not the same error we saw before (factor of 4), but I think it's related. A factor of two came from my doubling of the time constant, and I thought the other two was because I was using double the height (thinking I was right). But in light of the above, I think I was wrong there. But that error should have been a factor of sqrt(2), not 2.

What do you think?

Based on your derivation, what is your initial velocity v0 and the exponential form for the flow rate? I think I've fixed it but I would appreciate a second set of eyes again.
 
Oh... for a practical comparison: the tanks are 2' x 4' in area and a total height of 4' with 2' above the tank waterline and 2' below the tank waterline. So the delta at time zero is 4'

So I think v = sqrt(g x total tank height) = sqrt (32 x 4) = 11.3 seconds
 
karimwassef said:
Back to the "hairs" - do you agree that two elbows or more would stop the flow?
Click to expand...

90 degree elbows with this volume and speed of water would definitely make flow much less than theory.
My bigger concern would be significant sideways forces from deflecting that volume of fast moving water through an elbow, and the nature of the design is that the sideways forces would turn on and off constantly. Seems really hard on a structure.
 
Why am I so focused on this?

there's no way to move enough water down 12' of tank without a massive flow mechanism and doing it without pumps would be ideal for life. Doing it without splashing 50gals over the edge would be good too... ;)

I can't run powerheads of flow diverters down that massive of a tank without lots of ugly piping and DIY underwater electric cabling...

So this is a key part of this tank.
 
taricha said:
90 degree elbows with this volume and speed of water would definitely make flow much less than theory.
My bigger concern would be significant sideways forces from deflecting that volume of fast moving water through an elbow, and the nature of the design is that the sideways forces would turn on and off constantly. Seems really hard on a structure.
Click to expand...

The loss in head pressure was so massive, it wouldn't flow at all... at least that's what I get with just two elbows!
 
karimwassef said:
Based on your derivation, what is your initial velocity v0 and the exponential form for the flow rate? I think I've fixed it but I would appreciate a second set of eyes again.
Click to expand...

I get the same starting velocity for both conditions because the difference in water level is the same in both.

The only difference I get when setting up my diff eqn. is that in the "surge" case every bit of volume out the bottom pipe is equal to the volume drop in the container...
Apipe*dx = Acontainer*dh
But in the connected containers, the volume out the bottom both subtracts from one and adds to the other connected containers. so it must be doubled when accounting for the change in the water height difference.
Apipe*dx * 2 = Acontainer*dh


Aside:
I don't think your exponential solution is a bad model, and I think you can make it work, but (if it matters to you) I still can't get an exponential as the solution to the diffeq.

I still don't get an exponential solution to the diff equation, because it comes from the water velocity out being proportional to sqrt(h) and not proportional to h.

my equation is not dVolume/dt = - (constants)*Volume

my equation is dVolume/dt = - (constants) * Sqrt(Volume)

https://www.wolframalpha.com/input/?i=V%27(x)+%3D+-+C+sqrt(V),+V(0)+%3D+V0
 
Oh, and crazy idea you might consider.
Maybe instead of a big vacuum pump pulling many gallons of water up by sucking air.....
Why not just pump air into one underwater tank and let the other have water, then let that imbalance create the flow.

Guessing the pump would be way cheaper.

Sent from my SAMSUNG-SGH-I337 using Tapatalk
 
I'm planning on having the vacuum suck air out of one container and pump it into the other container. The idea being that if the volume of air is the same, then they're guaranteed to level back to the tank waterline. This is important because I'd like to create this massive flow flow changing the waterline significantly (minus the wave effect due to propagation speed).

My current surge on my tank is awesome and truly bubble free... except for the secondary surge into the overflow and down to the sump. That creates a massive bubble effect.

If I can get the two tanks to passively balance, it simplifies the flow immensely. It really should emulate a circulating loop of flow.

Danny (jcca) actually suggested pressurizing the containers to drive flow instead of a vacuum. I'm incorporating both ideas.. negative pressure in one and positive pressure in the other and then letting their natural settling drive the flow.

To be honest, I haven't confirmed that the vacuum flow and pressure rating will support both the sucking and blowing... it needs to be verified.
 
Karim, why not just get two hydrowizard ECM63 pumps and let them do all the work. It would probably be cheaper and I'm sure you'd get more than enough flow! Heck, you could probably even use 4 of the ECM42 and be happy. Definitely cheaper than the vacuum approach.
 
actually, the peak flow isn't enough with that pump. I need to reach a peak of 50,000 gph and they peak at 13,500. So I would need 4 per channels... multiplied by 4 channels = 16 powerheads. At $2000 each = $32,000. The vacuums each cost $100 and I may need two per channel = $800. The vacuum channels and plumbing is maybe $2000 (that includes 16 valves for the air)... so ~$3000 to achieve up to an instantaneous peak flow of 150,000 gph.

the vacuum chambers push and pull equally in a circular loop. They do this passively (without active control) just by ensuring the same levels in the storage tanks. I may be able to get the powerheads to do the same, but it would need active control.

also, I don't want a single round outlet flow (pipe outlet). I want a flow pattern that is uniformly pushing in the same direction through the whole aperture.

I also want to avoid directly pushing water through high shear. The theory being that pumps rip plankton apart and that kills the lifecycle in the captive reef. I would like to do it all without any mechanical force directly in the water.

My original approach was to use a matrix of 10 powerheads (Jebao) per channel (40 total at $80 each = $3200). That's more affordable, but also unproven. The wire bundles for power and control were becoming a nightmare ... not to mention having to get in there for maintenance. Not having any mechanical devices in the water creates a significant relief there. There's no potential for fouling or impacted flow due to impellers jamming or being invaded by life.
 
taricha - ok... I've reformulated by result. What is your time constant for two tanks at 2' x 4' x 4' centered around the display tank waterline. One is full to the top (2' above waterline) and one is full of air (2' below waterline). They're connected to the display with 12 pipes of 3" PVC.

I get a time constant of 2.5 seconds which means that most of the flow occurs in about 5 seconds (twice the time constant).

time constant = 4 A h / ( n pi D^2 sqrt(2 g h) )

A = 8 ft2
h = 2 ft
n = 12
D = 3 in

time constant = 2.4 sec (volume delta drops from 100% to 1/e = 37%)
time to mostly purge = 2 x time constant = 4.8 sec (volume delta drops from 100% to 1/e^2 = 14%)
time to near total purge = 3 x time constant = 7.2 sec (volume delta drops from 100% to 1/e^3 = 5%)

This is without taking into account the hairy parts of the equations.
 
dave.m said:
This has been repeatedly disproven.

Dave.M
Click to expand...

where? when I researched algae farms and fisheries breeding programs, the consensus was that pumps kill life at that scale.

It's why they use slow motion paddle wheels for aeration for example.

I spoke to the breeding expert at the LI museum aquarium and he confirmed that pumps cannot be used because of the forces they exert on very small scale life.

The organisms that do survive have hard surfaces (like dinos, etc...) but the embryonic stages of shrimps and fish don't.

So I guess the answer is that it depends - but the sensitive stuff gets mashed up. :)

The other reasons I prefer non-contact are important too.
 
It has come up often in articles over the years. I did a quick Google on "copepod amphipod killed pump aquarium" and got a variety of answers. Here's one article that clarifies Harpacticoid copepods are more likely to survive pumps than Calanoid copepods. If you wade through the flotsam and jetsam there is also mention of baby Banggai cardinals safely passing through circulation pumps. It's a common enough question so please don't get the impression that I am dismissing you concerns.

Dave.M
 
No worries. It also depends on the age of the babies. Some are more robust. I can't find a zero pump solution yet, but the circulation volume will be minimal compared to the surge flow, so the majority of motion should be "soft".

The idea is to increase the odds of the more gentle babies surviving and potentially becoming plentiful enough to get to the next stage.

There's no doubt that this tank is experimental. I'm doing many things that are unconventional because I believe they can be better... and maybe so much better that they could improve the hobby for all of us.

I also have very finite resources and a very big tank to support so I need to think out of the box to get the cost back into the box :)

It's why the sump is now going to be cinderblock and concrete.. and the surge tanks are likely to go plywood.

The biggest ongoing expense right now is still cooling and the increased power consumption of the vacuums... and the cost of RODI water has crept up to becoming silly (monthly change).

It's still evolving but I always always always appreciate the input here. Sometimes, it's easy to get excited in a vacuum (pun) but the smart questions and feedback here keep it grounded.
 
Ok. Big news. We signed the docs on the new house. Location is Prosper, TX.

It's not a southern facing back yard, but it's a big western facing back yard with enough clearance that I can get sunlight from 8:30am to 3:30pm every day of the year!

It will require a separate conservatory/sunroom though- not attached to the house. I'm working on a glass walkway to bridge the space to the house.

Quotes for the sunroom are coming it very high, but it's progressing.
 
ok. with the site finalized, I decided to map out the neighborhood with existing and future structures to find the best location the tank. There are some tall fences (11'), a few big trees (16'), the house itself and even a water tower... all casting shadows at different times of day and year.

A little sketchup ray tracing shows the times when the tank will be in full sunlight (Dallas). I ran the same for Boulder just to see how much difference it makes:

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/1_zpsuycknqur.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/1_zpsuycknqur.jpg" border="0" alt=" photo 1_zpsuycknqur.jpg"/></a>

Average is about the same, but northern locations will have more swing.

A little more calculation on angle of incidence throughout the year... and I converted the actual hours of direct sunlight to effective hours... this should be equivalent to full real reef sunlight (in the tropics where the sun is directly overhead most of the time):

<a href="http://s1062.photobucket.com/user/karimwassef/media/Designs/2_zpsovdxomeb.jpg.html" target="_blank"><img src="http://i1062.photobucket.com/albums/t496/karimwassef/Designs/2_zpsovdxomeb.jpg" border="0" alt=" photo 2_zpsovdxomeb.jpg"/></a>

So I was kind of happy that I always had at least 8 hours of sunlight every day, but the winter is rough. Even with direct sunlight, the angle cuts the light by 50% to 4 hours of equivalent direct sunlight.

So winter ~4 hours, summer ~10 hours... average of ~7 hours throughout the year. It's not bad, but the winter seems very short.

here are some links that really helped along the way:

http://www.itacanet.org/the-sun-as-a-source-of-energy/part-1-solar-astronomy/
http://www.itacanet.org/the-sun-as-a-source-of-energy/part-3-calculating-solar-angles/
http://www.itacanet.org/the-sun-as-...t-2-solar-energy-reaching-the-earths-surface/
 
You must log in or register to reply here.

Similar threads

joekidwell
JBJ Nanocube MP10
Replies
4
Views
3K
jeremy007
J
joekidwell
How do I stop light from flooding the room
Replies
6
Views
2K
Sk8r
Sk8r
Reefers4U
Citron Goby Death
Replies
2
Views
551
Reefers4U
Reefers4U
R
coral for sale in Akron/Cleveland
Replies
1
Views
641
Ron448
R
E
FS: 45 SC Aquariums, sump, stand, reactors, UV sterilizer and more
Replies
3
Views
825
Zoahoarder
Zoahoarder
Back
Top