6' Recirculating skimmer

[BeanAnimal]

I am about 75% done with my skimmer.

Background:
This project started out as a classic airstone Counter Current design for my inwall 75 gallon setup. The total system volume is around 150 gallons and a mixed reef with Soft, Hard, and LPS. The current skimmer on the the system is an Excalibur and just can't keep up with my overfeeding. My xenia and zoanthids are out of control and the algea is starting to become a problem.

Even though I built a HUGE sump, I don't have the clearance for an in sump skimmer (and had always planned on an external model).

After doing some research and finding nothing appealing in external skimmers for price and performance I decided to DIY.

As I studied threads here and other places I became aware that the NW designs introduce much finer bubbles than any reasonable airstone can provide with anything less than a full blown compressor. Needless to say the the project evolved from airstone to recirculating.

I don't plan on going into the details of the reasoning behind my decisions other than to say that I used every inch of vertical height in my fish room and 6" is about the widest body I could live with.

That said, some photos will follow. As the next week or so pass, I hope to get the finished photos up and then update the thread in a few weeks with the observed results in tank.

Enjoy...

PS
The higher res images will be here for a while. At some point I will replace them with low res thumbnails and you can then click on the image to be taken to the high res (don't want to kill my server with RC images).

(Title-changed by request)

[BeanAnimal]

The basic skimmer base without it's top. Notice the cam-loc fitting on the drain. This is to facilitate easy draining to the floor drain or connection of a pump to store the water if I need to drain the unit for maintenance.


A side view showing the basic plumbing. The feed pump water is piped into the OR3700s intake plumbing.


Another view of the plumbing from the top this time. The flange holes are 3/8 16 TPI and tapped.
[url=http://www.reeflogix.com/images/rc/skimmer/topview_b.jpg]a

[BeanAnimal]

Here is how I made the flange. It is 1" PVC plate routed to fit the outer tube and inner baffle tube. Waterproof epoxy was used to secure the outer base, as the fit was not tight enough to use PVC cement. The outer tube is 6" OD harvel clear PVC


Here is a finished view of the flange. The outer ring of holes are clearance holes for the nylong bolts that attach the flange to the skimemr base. The inner set of holes are the water returns for the recirculation pump.


Here is the gasket that will seal the flange to the body. I cut it out of EPDM rubber (neoprene) using a very sharp knife and the flange as a pattern. I used a hollow punch to put the clearance holes in it. If you look close you will see the feed pump plumbing does not fully connect with the recirculation pump plumbing. The DIY input bell was made by heating and stretching the 1" thinwall PVC over another PVC pipe. It is off center here as it is not pushed in.

[BeanAnimal]

Here you see the baffle plate partially cut. I used a drill press and indexed it around the center pin. The bit was a single flute countersink. I finished the holes off with a round bottom 3/16" router bit in the drill press (to prevent cracking)


Here is the finished baffle. It is tapped in the center for 3/8" 16 threaded nylon rod. I did this in case I needed to remove or reconfigure the plate. The o-ring was used for testing purposes to determine how much the OR3700 pushed through the holes in the bathtub. Not scientific but it gave me an idea of I was on the right track. The Oring is not needed in the finished skimmer.


Here is the baffle chamber with the uniseal that will connect it to the outside tube and the pump. Not much more to say. It is about 11" tall.


The inner chamber and the plate are held to the flange with the 3/8 threaded rod and this DIY locknut. It is just a piece of tapped acrylic.

[BeanAnimal]

This is a view of the baffle chamber inside the main tube... not much to say here.


This is a basic full shot of the skimmer. It is too tall to get into the picture but it will stand about 7' with collection cup. The reaction chamber is around 5.5' from flange to cup.


Hre is a look down the throat of the skimmer. You can see the bubble plate, RC feed pipe and faintly see the return holes. The inner chamber is 5" acrylic and again the body is 6" PVC (about 6.25 ID)

[BeanAnimal]

Here is a bracket that will be velcored to the main body. It will fit higher than this, but the photo gives you an idea of what it is for. The FEED and RETURN pipes will be strapped to this to keep them rigid and straight over time.


Here is a better view of the feed water mix. The idea is that all of the feed water (about 150 GPH) gets sucked into the RC pump right away instead of possibly bypassing to the return. The wide bell is designed to allow the 900 GPH RC pump to also do it's job with the rest of the water in the base.
.jpg

[Roland Jacques]

Nice job BEAN, i really like your tank water feed , the way you tied it into the pump is good. i used a tee but i like yours better.

how many holes and what size hole did you end up using? it looks like 60 at 3/16"id.

[BeanAnimal]

This is not the final plate. Honestly... I must have made 10 of the things. I ended up with right at 100 holes on the final plate. I thik someplace around 4 sq inches of area... I don't have my notebook handy.

[BeanAnimal]

Other notes:

The base is 1/2" cast acrylic. The joints were OK but I gummed em up with WO 40 (just had to try the stuff even though I didn't need it)

I built a DIY venturi for the intake by thermoforming 1" PVC pipe. At 6" of head it pulled well over 20 SCFH. The unit had a very refined shape (30 degree input taper and 5 degree exit taper). I did my homework on it and built it according to sound fluids engineering principle.

However in my haste I made it the wrong length and did not realize until I glued it up and test fitted it. (I had already cut the hole in the body and base) and since scrapped it for a stright pipe and tapped hole for a john guest fitting.

The unit will be fed via an airpump (20 liter HiBlow or Alita). I may one day go back and do another venturi... but for now this will do.

[AcroSteve]

Nice. How much air do you figure you can pump into it?

[BeanAnimal]

No idea yet... From most accounts the pump will handle 20 SCFH at full flow. At this depth I am nto sure how it will react with that much air shoved into it.

Dale (tinygiants) was pushing somewhere close to that (If I remember correctly) into a 6" PVC airstone skimmer. He might have even been pushing 40.

The pressure at that depth will be just under 1.75 PSI so a 20 liter Hiblow will do about 17 (if the pump can handle it in this setup). The 40L Alita will be too much and a wast of energy, not to mention the noise. I am not quite sure what airpump I am going to end up with. The Alitas are to expensive to buy two 10 Liter units, so the HiBlow (same as alita) 20L may be just the ticket.

[smjtkj]

Looks good Bean!

[Pbrown3701]

Looks great Bean! Do you really think you need the diffuser plate on a 6' tall skimmer though?!? I'll be interested in hearing your answer on that question after everything is up and running.

[BeanAnimal]

Yes, I think the plate will make a difference. Running fresh water with and without the plate at 5 SCFH even showed a difference.

Bean

[BeanAnimal]

I have given a lot of consideration to skimmer floods and have a few ideas on how to avoid them with this skimmer. I have not yet decided on the style of collection cup (sealed with a remote holding pot, or standard vented cup).

I am not sure how well the "auto shut off" pots work. It will be a week or two until I get any testing of that nature done.

The skimmer will be fed via a MJ 606 out of the overflow and returned over the top of the tank. The MJ should be enough to feed the skimmer due to there being only a few inches of head.

This also means that when the power goes out, the skimemr will not drain. The ATO should work no matter if I have the skimmer online or offline.

[ChemE]

Bean, I've also given a lot of thought to designing out skimmer overflows. My best idea so far is to put a removable waste collector in chamber 0 of the sump. Chamber 0 is higher than chamber 1 and water in from the display enters in chamber 1. Chamber 0 is able to overflow into chamber 1. With this design if the skimmer goes nuts and starts pumping tank water into the collection cup into the waste chamber the tank water will just cycle back into the sump. Worst case scenario here is that the waste collector is nearly full of espresso when the skimmer starts going nuts and all that waste goes back into the tank. Is this better than a flood? I'm not sure I'm convinced yet; I'll be curious to see what you all think.

[BeanAnimal]

Yeah, my first basic plan was to simply plumb an overflow into the cup that drains back into the tank. If it goes nuts... well you get the idea.

Now you could do this with a simple quick connect fitting (like the camlock I have on the drain)

-=OR=-

You could but an overflow tube and elbow on the side of the cup. The neck flange would have a gutter in it that this drained into. The gutter would be connected to the sump. That way when you remove the cup there is nothing to disconnect.

Second form of safety:
I simple float switch in the breather portion of the Hartford TEE would act as an air/recirc pump cutoff in the event of a clogged drain.

I have also considered a tall cup with a float switch that normaly stays dry. If the skimmer went nuts this switch would cut the air or recirc off.

I have other ideas, but none tested yet.

[Pbrown3701]

i run my collection cup drain to an empty icetea jug. In the screw on cap, there is the tube that goes to the collection cup and then another that goes back to the tank. Because the container is sealed, if the skimmer goes nuts, then it flows back to the sump.

I have had about 3-4 inches of scum and then have it go nuts and add it to the sump. My anemony shrinks up for a couple of hours and then everything is back to normal (except the skimmer works overtime for a day or so).

[RichConley]

Hey Bean, nice work.

Are you planning on running the OR on an airpump, or just venturi? If you're running venturi, theres always the autoshutoff type collection cups, but I dont think that'll work on an air pump.

[BeanAnimal]

Air pump... No way I will get as much air as I want with 5' of head on this thing.

The "shut off " type cup would work, but it would put abput 4 PSIG on the system and pump if it dead heads. The way around that is to leave a bleed valve open on the air feed. That way when the cup backs up the bleed valve keeps the pressure down.

Thanks for the complements Rich (and everybody else who has tipped their hat).

[tinygiants]

Nice skimmer.

The foam head will actually collapse with much less backpressure than 4 psi.

Try this: On a skimmer drain line, run it to a cylinder of water. The foam head will collapse when the drain line gets about 2 - 3 inches deep. This of course assumes a sealed collection cup. This method will not create the large backpressure damage to the pump. The air will still exit the skimmer through the drain line, but no new foam will enter the cup.

Dale

[BeanAnimal]

Yup did not think about the drain line. But then again the water would have to be pushed out of the skimmer first of it was all sealed (at least in this setup).... now thats a scary thought!

[BeanAnimal]

Before I forget, I neglected to credit MarkS from here at RC for writing the GhostScript to create the flange patterns (I even easily adapted it for use on the baffle plate patterns)

Here is the Thread

I spent some time in autocad and some other tools creating the flange templates, then ran accross his little script by accident. I have since used it as it is less cumbersome than the big powerful design tools. I can change a flange or baffle design in seconds.

I did have to make some modification to the script to allow for larger paper sizes and centering.

Thanks Mark!

[AcroSteve]

I use a float switch in my collection container. Works pretty well, but adds some electrical devices that might be avoided with another method.

[elfa]

I do the same as Steve, but I have in addition added a second float switch that activates an air pump when/if the skimmer shuts off.

Vidar