Coil Filters - Introduction:
 

Maintaining excellent water quality in my reef tank is of course on of my prime concerns. My canister and fluidised bed filters provide excellent conversion of Ammonia ions through to Nitrite then Nitrate ions. Given the reasonably heavy bio-load in my tank and the number of sensitive invertebrates I also need an efficient method of removing Nitrate ions. There is clearly some level of denitrification taking place in the tank - which I assume is taking place in anaerobic regions of the rock and sand, however this is not sufficient to maintain a balanced environment. Nitrate reducing resins and sponges placed in canister filters don't appear to work well so I have spent some time researching and developing more experimental approaches.
 

Nitrate Reduction Basics:
 
The Nitrogen cycle contains the following phases:

Ammonia ions produced by fish waste and rotting material.
Ammonia ions converted to Nitrite ions by aerobic bacteria.
Nitrite ions converted to Nitrate ions by aerobic bacteria.
Nitrate ions converted to free Nitrogen and Oxygen gas by anaerobic bacteria.

Producing the aerobic conditions for the Ammonia -> Nitrite -> Nitrate process is fairly easy in the aquarium environment - after all, the water is heavily oxygenated through the water surface, vigorous water movement and externally introduced aeration. However, to produce the conditions necessary for the anaerobic Nitrate -> Nitrogen/Oxygen process is not so easy. A number of methods exist to try to promote these conditions, most of which involve creating a region where water is allowed to move very slowing, thereby allowing the dissolved oxygen to be consumed and so rendering the water anaerobic. A very fine balance has to be achieved between ensuring the water flows slowly enough to go anaerobic and ensuring the water does not flow so slowly as to start to produce Sulphur Dioxide (which not only smells bad but is poisonous to the tank inhabitants. The following methods exist:

Porous rocks and sand substrate (denitrification occurs naturally to some extent in mature systems).
The "wet" phase of a wet/dry trickle filter.
Under-gravel box filters.
Coil filters.

Coil Filter Design:
 

The basic design of a coil filter is quite straightforward. Water to be filtered is allowed to pass VERY slowly through a long narrow tube. The tube is coiled up to fit it into a reasonable space, hence the name coil filter. As the water enters the top it will be oxygenated. As the water travels along the tube, aerobic bacteria in the water will slowly consume the dissolved oxygen. If the tube is long enough and the water flow slow enough, sometime before the end of the tube the water will be completely anaerobic. At this point, denitrifying bacteria can begin to work. The water is passed through a suitable medium which promotes anaerobic bacteria growth. Once the water has passed through the denitrifying region it can be re oxygenated and returned to the tank.
 

My Experimental Coil Filter:
 

I have constructed a coil filter which works as follows:

Water from the canister filter system is pumped up to the reservoir in the coil filter. As the flow rate into the coil filter needs to be very low, most of the water entering the reservoir flows straight through and back into the aquarium. I have included two parallel coils of tube to double the capacity of the filter. Two taps have been used to regulate the water flow into the filter to about 1 drop per second each. Water exits the bottom of the tubes and rises up through the inside of the container which has been packed with Kent Nitrate Sponge media. I have "seeded" the Nitrate sponge media with a denitrifying bacteria product from Aquamarine. As the water reaches the top of the inner container it overflows into the outer container where a large wood air stone connected to an air pump re-oxygenates the water before it flows back to the aquarium.

The outer filer housing and reservoir was constructed from 5mm clear acrylic sheet, bonded with Araldite and sealer with aquarium sealer. The inner container is a small plastic waste paper bin. The two tubes are each about 18m of 9mm internal bore plastic tube which has been carefully coiled around the inside of the inner container and held in place with Araldite. All of the pipe fixtures were made from standard fixings available from my local D.I.Y. store.

There are a number of mechanisms in the filter which should encourage denitrification:

The long journey through the tube.
The Kent Nitrate sponge media.
The denitrifying bacteria.

As a result it will be difficult to tell exactly which, if any of these are affecting nitrate levels. As I have only recently set up this system I do not yet have any solid results to share on the effectiveness of this kind of filter.