Related Links Site Map About 

Project: DIY Jumbo Sized Fluidized Bio-Pellet Reactor

I recently decided to try out the latest rage in nitrate/phosphate reduction - bio-pellets ( for info see my Reef Blog from 7/7/2010). The vendors of the pellets (I purchased mine from Warner Marine) all recommend approx. 1/2 liter of pellets per 100 gallons of water to be filtered. I usually stick with recommended amounts - I always like to err on the side of caution - so, for my system, I need around 2-1/2 to 3 liters of pellets. Current general opinion seems to be that the pellets require a fluidized media reactor for maximum effectiveness. That's a bit of a problem as there are no (affordable) fluidized media reactors on the market that are large enough to hold four liters.

So what to do? Well, I happened to have a 24" long piece of 6" cast acrylic tubing that had originally been intended as the neck for a DIY skimmer that never materialized. It was perfect for a jumbo sized DIY reactor, instead. Add to that some 3/8" black and 1/4" clear acrylic sheet, a few 1/4"-20 nylon thumb screws, a length of 1/4" silicone o-ring stock (all of which I had laying around as extras from previous projects), and just about all the material requirements, for the reactor, were taken care of. The only parts I had to purchase were a few 3/4" Sched-40 PVC fittings and two 6" circular sewing grids from Jo-Ann's sewing supplies.

How does a fluidized reactor work?

Before I start on the description of the build, it might be good to describe how a fluidized media reactor functions - skip this paragraph if this is not new to you. The term fluidized media comes from the fact that the media - such as sand, carbon, GFO, or in this case bio-pellets - is forced into a fluid like (ergo fluidized) tumbling motion by the flow of water through the reactor. This can only take place if the water entering the reactor (in sufficient volume and force) is first sent to the bottom of the reactor and then flows up through the media. The drawing illustrates this principle:

Of course, in addition to the reactor, a pump is required to move the water from the system, through the reactor, and back to the system.

A little about the design

This reactor would be holding a relatively large amount of media. Add that to the fact that bio-pellets are fairly heavy, and one can assume that there will have to be a strong current through the reactor in order to cause the pellets to enter into a fluidized motion. Taking this need for more water flow into consideration, I dimensioned all the fittings a size larger than the 1/2" I usually use. I assumed that 3/4" water lines would suffice to get the 2-1/2 to 3 liters of pellets moving sufficiently.

I've been building fluidized reactors for years using the same basic and very practical design. It's seen on many commercial reactors and with good reason. The design is straight forward, comparatively simple to build, and functions very well.

And the component parts

The top consists of a keyhole flange made of 3/8" cast acrylic sheet. I like to use black, but any color will do. I use a template whenever I need to make a keyhole flange. Years ago, I made a set of flange templates - one for every standard size of acrylic tubing, from 2" all the way to 12". The templates are made from 1/2" MDF and minimize the effort necessary to complete a flange. Using a template and my router table, I can complete one in less than half an hour. I have a page on how to make the templates - check it out.

I used 3/4" female threaded couplings for both IN and OUT through-leads. That way, later, I can change out, or realign, the 90deg elbows for straight couplings if that better suits the installation conditions.

Anytime I have a need to glue acrylic to PVC, I use the two part epoxy glue Weld-On 40. It is extremely strong and fairly easy to work with. I say fairly easy because you do have to put a little effort into mixing it up accurately, as the instructions call for a 20 to 1 mix - resin to hardner.

Viewed from the bottom, you can see the fittings used below the top flange. The longer of the two (water in) is a 3/4" slip coupling. When the top is placed on the reactor, it engages (slips over) the central water feed tubing that leads to the bottom of the reactor. The "water out" lead just ends in a short stub (just a slice of a normal 3/4" coupling added for strength) beneath the top flange.

The bottom grate consists of a 6" cast acrylic disk mounted on a short piece of 3/4" PVC tube. A large number of 1/4" holes were drilled in the disk to handle the water flowing up from the bottom of the reactor. Above the disk, I added a 3/4" PVC coupling. When assembled, the open end of the coupling receives the central tube carrying the incoming water to the bottom of the reactor. Below the disk, I drilled a number of 3/8" holes in the short piece of tubing to function as an exit for the incoming water. I made sure to drill an equal number of holes all the way around the tube to alleviate any chance of dead spots, in the flow, as the water travels up through the pellets.

I then placed a flexible plastic mesh over the top over the bottom grate. It has very small openings and keeps the media from falling through the 1/4" holes in the grate.

You can get the plastic mesh circles at JoAnn's or Michael's. The material is called plastic canvas and is actually available wherever sewing supplies are sold. The number 7 mesh seems to be the only mesh size available. It seems to be about the optimal size for bio-pellets, at least for the EcoBak pellets I have (from Warner Marine). The material is very easy to cut to any desired shape. I got the idea from a thread on the Reef Chemistry forum on Reef Central. For those addicted to shopping on-line, here's a link to a source on the web: Plastic Canvas

The upper grate ensures that media cannot exit the top of the reactor with the outflowing water. It is made up of a perforated acrylic disk, just like the bottom grate, but the 3/4" PVC coupling is on the bottom of the disk, not the top. The couplings - one on the top of the bottom disk and one on the bottom of the upper disk - serve to connect the two disks with the central tubing and allow the whole assembly to be taken apart for cleaning, etc.

The short length of tube extending from the top, fits into the coupling on the bottom of the top flange. There is a small o-ring around the tube which serves as a seal to stop the incoming water from leaking into the upper portion of the reactor.

Here is the internal assembly. It consists of the upper and lower grates and the connecting 3/4" PVC tube. As described earlier, the tubing is not glued in place. It merely slip fits into the couplings so that the whole thing can be disassembled if need be. The plastic canvas is hanging down on one side because I hadn't yet fastened it to the grate. For that I use very thin nylon fishing line:

The only parts left to show, other than the main reactor body is the 1/4" silicone o-ring and the eight 1/4"-20 thumb screws. I used to use 1/8" o-rings but, recently, I experienced some trouble with leaking reactors. Since I changed to 1/4" I've had no further troubles. I make the o-rings out of stock I get at McMaster-Carr. I cut it to the required length and then glue the ends together with clear silicone to make the ring.

A 1/4" wide and 1/8" deep circular groove is routed in the lower flange for the o-ring to fit in to. I do this using a Jasper circle cutting jig. I describe the jig in the page on making keyhole flanges.

At this point, I should probably note that some feel that Sched-80 PVC  fittings should be used on this type of equipment, instead of the much cheaper Sched-40 fittings. I can only say that I have always used the less expensive material and have never had a problem of any kind.

And... the completed reactor

When you put it all together this is what the reactor looks like. It has a footprint of 8-1/2" and stands 28-3/4" tall. It doesn't look especially large in the pic, but remember that those hose barbs at the top of the reactor are 3/4":

In order to keep all the bio-pellets in this reactor fluidized, I'll have to utilize a strong pump. I have an extra Eheim 1262, so I may try that for a while - or, I just may buy a Mag 9.5 (a 900gph pump, just like the Eheim). The Mags are quite reasonably priced and I'd be able to use the Eheim somewhere where its extreme reliability is more needed.

08/09/2010 - Here's a short video showing the reactor in action. The reactor contains 2 liters of Warner Marine EcoBak pellets and is being run using a Mag 9.5 pump throttled back to approx. 500gph.

I'll be writing up a page about my adventure with bio-pellets - I'll report on how the pump, reactor and the pellets themselves, worked out.



Currently under general maintenance.

Please check back shortly.