Stone/rock filters have been the main filtration technique for most koi ponds for years. What is so good about stone? Most of the volume of the media is the bulky core of the stone which is biologically inert, resulting in a small amount of surface area. However, it does create an environment that is conducive to aerobic bacterial growth. My guess is the resistance created by the stone helps to equalize the pressure throughout the media, thereby preventing dead spots. Also water passing through the stone has an eddy current effect which creates an environment which helps bacterial growth. But how can we have the same effect in a more efficient, lighter media? The natural evolution is to consider mattress filters which have a larger surface area. Mattress, also called fiber matting, filters have been used successfully for many years, the most common being the air driven filters used for aquariums. All of the older types of mattress filters employ a solid piece of matting material with water flowing straight through it. All of these filters eventually channel and clog and become anaerobic if not cleaned regularly.
In Japan and England the checkerboard arrangement of matting material rather than a solid piece was found to be more efficient. This arrange ment seemed to create the eddy currents that are more conducive to bacterial growth. All the big koi breeders in Japan are now using this method. It is also the predominant method of filtration in England. In short, people with very expensive fish are using this system. Miyoshike, in Japan, and Peter Waddington of Infiltration Limited in England, have played a major role in developing this system. The King Koi, Pond America, and Nishikigoi Tancho filters employ this technique.
I have been experimenting with this material over the past six years with a combination of scientific and trial/error approach. I have had my share of frustration and dead fish to show for my efforts. The original use of this material was in containers designed for stone filters with heavy aeration under the matting. Although this works, it is energy inefficient. I like stone filters, upfow and down units can be used, but unlike stone filters, lateral flow can also be used. In general, the lateral flow units are easier to employ.
To build a lateral flow unit, the filter container should be a rectangular trench with a flat, or grooved bottom and one or two drain plugs. The container can be either fiberglass, plastic, or made from pressure treated wood with a liner on the inside. A waterfall entering and exiting the trench is preferable to aerate, or some other form of aeration can be used (i.e., a trickling filter or stream before water enters the trench filter). Air stones in the trench can also be used to increase the efficiency. The trench can be any length. The longen the better.
Since the mattress has very little resistance to water flow, the system should be designed to equalize the pressure throughout the media (i.e.. entry ports and exit ports should be equally distributed across the ends of the trench. A plate with holes can serve this purpose). Equalizing the water distribution entering and exiting these filters is the most difficult aspect of these units.
The media should be supported on a screen or grid slightly off the bottom. This allows sediment to be constantly purged. An even better design is grooves in the bottom of the container. The top of the media should be above the water level to prevent the water from channeling above the media. Mesh screens or bags can hold the media in place.
The media that I have been using is a polyester matting. The checkerboard design can be used but cutting this material into squares appears to be superior. Checkerboard units can be placed at either end to keep the squares in place, or egg crate grid could also be used for this purpose. Another design that is more effective is punching a series of holes in the matting in a large, small, large, small con-figuration in successive sheets, This increases the eddy current effect.
Mattress filters do best with a higher flow rate than stone. The optimal flow rate is probably about 10 gal./ minute/sq. ft., but I have seen considerably higher and slower flow rates be effective depending on the system. Lateral flow requires a more rapid flow than up or down flow. Often the filter is gravity fed before the pump, but it can also be after the pump. Up flow or down flow can he employed in rectangular or round containers, but round is prefeable.
To test the effectiveness of the unit, shake some of the media under water, especially near the bottom. If it turns cloudy, there is an inadequate flow rate in that area. If pepper flake-like particles shake off the media, and the water remains clean the flow rate in that area of the filter is adequate. Better the flow rate he too fast than too slow. The largest filter of this type I have built is 20 feet long, 4 feet wide, and 18 inches deep with a small waterfall in the center to aerate. The smallest I have built is 42 inches long.
These filters are the filters of the future and are more energy efficient and much more space efficient. Sooner or later we'll catch up with Engltnd and stop using gravel.