Filter system & barrel stand
Back to main page




Now that the tank is together, time to install the filter system.

I wanted to use 2" PVC pipe because all the fittings and valves are really cheap, but I wasn't sure if I could gravity pull 3000 gallons per hour through a single 2" pipe. I know others use 2" pipe on their bottom drains, but those are usually hooked up directly to the pump like my pool pump is, so the pump actually sucks at the pipe and can keep it under pressure. So you have to understand that I am configuring my filter system to have a submersible pump just suspended in a barrel, so it is a real gravity feed situation, not forced with suction through the pipe.

The gravity flow pipe sizing charts that I could find are very confusing and I couldn't fully understand them, so I setup an experiment between 2 barrels with some 2" pipe. I discovered that at 1500 gph, the drop in level from one barrel to the other with about 3 feet of 2" pipe was around 4". At the time I was considering running my tank at 3000 gph, so it would make an even worse water level difference. So I decided to go with 3" pipe for the bottom drain. I considered going with 4", but the fittings and valves were double the price of the 3".



Here is my upside down bottom drain. Had a bunch of different ideas for how to make the bottom drain, so to narrow in on which solution to go with, I thought about the characteristics that I wanted to have.

1 - I wanted to have the water sucked from the very bottom of the tank. Since I have a current running in a round tank, I get what is called the "toilet bowl effect". This is a neat effect where debris are pushed towards the center naturally by the circular flow of the water. If you want to see what I am talking about, here is a neat experiment: take a 5 gallon bucket and fill it with water. Put some debris in there that will sink, but won't cloud the water. Then stir the water really fast with a stick so it spins. Now watch what happens with the debris -- it will all naturally collect in the center on the bottom of the bucket.

2 - Holes small enough so that a 6 month old koi can't go into the hole, also so an adult koi can't get it's mouth stuck in the hole and tear the side of their mouth.

3 - I wanted enough holes to have the same collective size as a cross section of the drain pipe, so they would not add resistance to the flow.

4 - Holes to be spread out enough so that if a small koi swam against the holes, it could free itself.

So after I did the numbers, this seemed like the best compromise from the materials I had available. I drilled 1" holes around the perimeter of the bucket, then cut the bucket off so the holes are cut in half. The horizontal pipe is high enough off the bottom so that a full grown 36" koi can swim under it without having to turn sideways.

But most importantly, I have a 1/4" hole drilled right at the point where the top of the pipe joins the elbow. This is a safety feature -- I know a number of people that each had some kind of failure like burst pipe in their filter system, and their pond was pumped entirely dry killing their entire collection. If I have a failure like that, the hole would break any siphon effect and the fish will still have 10" or so of water to survive in.

For the filter system, I decided to use 55 gallon barrels. Only problem is the barrels are only about 3' tall, and the tank is 4' tall. The first barrel is a vortex settling chamber and needs to be at the same level as the top lip of the tank, and the other 2 barrels are gravity return so they need to be above the lip of the tank.

I tried to join 2 barrels together, one on top of the other, but it didn't work very well, it leaked at the joint. So this was my other plan, make a cinder block stand.

I stacked the blocks and used a masonry blade in my circular saw to trim the blocks to get the desired heights.

After this photo I filled all the cells with rebar, then filled each cell with concrete. After that was cured, I filled the center with rocks and capped it off with concrete.

Here is the finished stand, it works just right. I should note that the 3rd barrel on the right is a 30 gallon barrel, not a 55.

Back to plumbing -- Here is how I attached the bottom drain to the first barrel.

The reason I continued the pipe straight up is so I can clean the main line with a flexible snake like people do with their toilets and sewer lines. While it is running, I can push the snake down and adjitate it back and forth loosening the debris, which will flow into the vortex barrel.

It really is amazing how much debris and muck can collect inside a horizontal pipe. It work just like a settling chamber.

This is the top view of the first barrel, my vortex seperator settling chamber. I know that it really is a lot smaller than most people would suggest installing. The most common rule of thumb is a dwell time of 5 minutes for heavy debris, and 20 minutes for the small light debris. That means if you have a 1000 gph flow rate, then you need a chamber that can hold atleast 83 gallons.

I settled on running 2 submersible pumps that have a combined actual flow rate of 2000 gph, and with the system running the vortex barrel only fills up to 52 gallons, which gives me a dwell time of just over 1.5 minutes. So you might be wondering why in the world I would waste the time and expense of setting up such a small settling chamber when I could simply add a submersible pump to the center of the tank and run a hose to the biofilter ? Well, the main purpose of that barrel isn't actually settlement, I am actually mostly interested in:

1 - Put the pump in a location that is really easy to access so I can maintain it. I have a couple of used pumps that I acquired from koi rescues, and they intermittently die on me, plus their intakes get clogged easily so it is nice to have it handy right there so I can maintain them.

2 - Shorten the discharge hose. I picked using a submersible pumps because they are really inexpensive to buy, and have a high flow rate per the amount of electricity used. But their one big fault is that they quickly loose flow rate as you add head pressure. So the shorter I can make the line that comes out of the pump, the less head pressure, the more flow rate, and a lower electricity bill.

Plus as it turns out, it does a bit of seperation.

Here are the two pumps I am using. The one on the right feeds the biofilter, and the one on the left feeds the UV sterilizer and the particle filter.

The reason I have them seperate is so that I can run the UV sterilizer at it's most effective flow rate, which is a much slower rate than I run the biofilter at.

Here is my packed column aerator on top of the biofilter. It aerates the water just before it goes into the biofilter.

This is a look across the top of the biofilter. You can just barely make out the horizontal pipe that makes my foam fractionator. The diagonal pipes that go out to the rim of the barrel are struts made from PVC pipe that hold the top of the tower pipe up. They simply have a piece of line running through them, and with the line tied tight, it does a good job of strengthening the aerator pipe.

This is the return intake from the biofilter. The pipe below the tee has a bunch of lines cut into so make a screen. The short pipe above the tee is an overflow, incase the screen gets clogged, the water will rise above the pipe then dump in there to return.

Well, this probably isn't up to NEC standards, but it works pretty good out here in the desert where it doesn't rain that much.

When it does rain, the door on the front keeps the rain from getting to the power strip.

Previosly I had this power strip mounted on a post, and put a bucket over top of it to keep the rain off it.

Here is my little UV sterilizer. It is only 9 watts, and rated for 1500 gallons max (half the size of my tank), but it seems to do the job.

Right now the UV returns directly to the tank, but I am going to reconfigure it to feed the fines filter (when I get around to making that filter)