The evolution of aquarium filtration technology has always intrigued me. In essence, its our attempt to imitate, as closely as possible, natures own method for processing aquatic pollutants in the form of particulates, dissolved organic compounds, ammonia, nitrite and nitrate.
To achieve this goal, filtration technology has evolved greatly over the years, but still falls short of achieving the ultimate goal, which, in my opinion, is a truly closed system in which water changes become obsolete and only the replacement of evaporated water is required.
Now, Im not naive enough to believe that the home aquarist in 2008 can achieve a closed system with no water changes, but I do believe that its possible to come far closer toward that goal than most people believe is attainablea system that would minimize water changes with respect to frequency and overall percentage of water volume, without compromising water quality.
One thing that some people seem to forget is that nature doesnt do water changes. You may say that rain amounts to a water change, but in areas with little or no rain, lakes, streams and ponds all thrive without the benefit of water changes.
So, how does nature achieve this? Well, a bio-load far lower than the average aquarium is certainly a heavily-contributing factor. However, these natural systems also incorporate a multi-tiered layering of filtration elements that the home aquarist rarely does.
The filter system Ill outline here will attempt to incorporate every component of the natural filtration process in order to come as close as possible to achieving the ultimate goal of a closed system. I am not a true expert in the field of filtration, wastewater management or environmental engineering, and I may have overlooked certain essential components that would compromise the effectiveness of my ideas. I have not yet built or attempted to use this filtration system, so its still in the theoretical stage. However, no concept utilized in this design is new or attributable to me, and the ideas Ive borrowed have been proven effective.
One thing Ive noticed in studying aquarium filtration is that there seem to be three separate core groups of peoplemarine aquarium hobbyists, freshwater aquarium hobbyists and koi/fancy goldfish hobbyists, many of whom have experience with large filtration systems designed for ponds with heavy bioloads. Ive also noticed that hobbyists in one of those three core groups can sometimes be prone to falling behind the hobbyists in another, and that theres been a failure, to an extent, to cross-pollinate other groups with new information and technology.
Of course, there are also cutting-edge technologies being employed in the areas of wastewater treatment and environmental engineering. This filter design attempts to borrow ideas from all of the aforementioned hobbyist groups and industries.
Before I continue, Ill state that this design is not cheap to implement, even if done in DIY fashion. The main reason for this is the usage of premium filtration media, which is typically very expensive (around $35-55 per KG). A 1,000-gallon tank heavily stocked with predatory fish would likely require somewhere in the neighborhood of 30-40 KGs of filtration media for optimal results.
The system Ill outline is configured to handle approximately 1,000 2,000 gallons, depending on how heavily the aquarium or pond is stocked. However, the design can be scaled down to very small levels and still incorporate the same technology and general principles. As its scaled down, its size (footprint) and overall cost also decreases.
What Im looking for is feedback from the community here about my ideas. Negative feedback is just as good as positive feedback, and probably even better. Id like people to try to find flaws with these ideas so that they can be refined. This is very much a rough draft in its current state. The design is broken down into five filtration stages, each of which will be described in detail.
Stage #1 Water exits the tank and is mechanically filtered.
This system uses an overflow to remove water from the surface or upper region of the tank. In the event that the next stage of the filter is place higher than the overflow, a water pump will need to be used to pump water out, instead of an overflow, which relies on gravity. In this initial stage, water exiting the tank flows through a mechanical filter comprised of a fine, fibrous filtering agent like a filter mat, wool floss, etc. This is intended to remove large particulate waste before it reaches the second stage of the filter. I wont spend much time on this section, as overflow devices with mechanical filtration incorporated are readily available and in widespread use already.
Stage #2 Water is dispersed evenly over a Bakki Shower for the first stage of the biological filtration process, at a rate of 1.5 times tank volume per hour.
Many of you may be asking, Whats a Bakki Shower? Although this filtration device has been in use primarily in Asia and Europe for more than five years now, its not in widespread use in North America, and has not become very popular outside the realm of serious koi and fancy goldfish hobbyists, most of whom use them for ponds.
In essence, a Bakki Shower, invented in Japan, is a tall, multi-tiered trickle filter device that showers water over a series of vertically-stacked bins, each of which contains a premium high-surface area filter medium like BioHome or Bacteria House. Without using one of these filtration media, a critical component of this filtration system will be lacking.
The reason for this is that both products claim (and their claims have been verified by countless users of these products), that the design of their filtration media employs a capillary system in which primary capillaries receive generous supplies of oxygen and therefore promote the colonization of aerobic bacteria, while secondary capillaries receive very little oxygen, and therefore promote the colonization of nitrate-consuming anaerobic bacteria in these dead zones.
Other filtration media, like bioballs, sponges, etc., do not achieve this effect and only promote the colonization of aerobic bacteria used to process ammonia and nitrite. The photograph below on the left shows a Bakki Shower design, while the photo on the right shows the capillary system of media like BioHome and Bacteria House.
An article with more information on Bakki Showers can be found here:
http://www.makc.com/bakki.pdf
Both BioHome, which is made from sintered glass, and Bacteria House, which is made from kiln-fired ceramic, are inert and will not affect water chemistry.
In Stage #1, heavy, extremely efficient removal of ammonia, nitrite and nitrate takes place. Additionally, due to the open design of the Bakki Shower, some nitrogenous gasses are displaced, and oxygen is infused into the water during the shower effect. This massive infusion of oxygen allows the aerobic bacteria colonies in the filter media to work at optimal efficiency.
The water then exits the Bakki Shower via gravity into the sump, which, for a 1,000-gallon tank, could be something like a 300-gallon Rubbermaid tub, available at places like tractor supply stores for around $200. Once the water enters the sump, filtration Stage #3 begins.
Stage #3 Detoxification and Additional Biological Filtration
Stage #3 filtration is achieved through a two-tiered process. The first tier is the placement of a thick bottom layer of sand in the sump, on top of which a thinner layer of fine gravel is placed. The bottom sandy layer should be at least 5-6 inches deep. Its purpose is to create another dead zone in which little oxygen exists, thereby promoting the colonization of additional anaerobic bacteria to break down organic waste and nitrates.
The second tier of Stage #3 filtration is what some refer to as a vegetable filter or refugium. In this case, our vegetable filter is a reed bed consisting of the aggressive, highly-efficient waste removing Common Reed
(Phragmites australis). Although my filter design is based on the use of a reed bed populated by
Phragmites australis, other plants can be substituted. According to Wikipedia, the following would all be acceptable for use in a filter like this:
- Phragmites australis, for temperate climates, One of the best plants for water purification, will escape cultivation and become invasive, is now invasive in many areas
- Nymphaea alba; for temperate climates, depth 60-120cm, One of the best plants for water purification
- Sparganium erectum, for temperate climates, depth 60-120cm, One of the best plants for water purification
- Iris pseudacorus, for temperate climates, depth 0-20cm, One of the best plants for water purification
- Schoenoplectus lacustris, for temperate climates
- carex acutiformis, for temperate climates
The reeds growing in the sump would be planted approximately 5 apart, throughout most of the surface area of the sump. Only a small section would be walled off for placement of heaters, water pumps and other equipment. This area would have to be designed so that sand and/or gravel would not enter it. I wont spend much time discussing this process, as successful sump design is a well-established topic.
Above the reed bed, a strong light source on a timer would give the reeds the energy they need to thrive. Something along the lines of a 75W or 175W metal halide light system would be advisable, as the light intensity of fluorescent bulbs decays greatly with distance. Ideally, some type of opaque screen would be placed between the Bakki Shower and the light source in order to minimize light exposure to the filtration media. Nitrifying bacteria require darkness, or at least minimal light, in order to work at optimal efficiency.
In Stage #3, ammonia and nitrite are processed by the gravel substrate in the sump, as well as by the reeds, which use both as fertilizer. Nitrate is also processed by the lower sandy anaerobic layer, as well as by the reeds as fertilizer. Additionally, the reed bed removes other pollutants from the water, including chemicals like lead, arsenic and other heavy metals.
Reeds and similar highly-evolved plants are extremely aggressive in competing against other plants for nutrients, to the extent that the reed bed would likely starve algae, thereby inhibiting its growth in the tank dramatically or entirely. This may be a problem for people with algae eaters like certain plecos in their tank, which would need to be fed an algae supplement.
The reed bed could also potentially starve out desirable aquatic plants in the tank. It may therefore be necessary to add aquatic plant fertilizer or CO2 to the system, depending on the size of the reed bed and how many desirable aquatic plants are in the tank.
Those whove read this far and have been paying attention will note that the filter has theoretically, through a combination of processes, addressed the removal of ammonia, nitrite and nitrate, rendering the first two immeasurable, and the third either immeasurable or extremely low (10PPM or lower). Users of Bakki Showers typically report undetectable nitrate levels in their ponds.
However, one component is still missingthe removal of dissolved organic compounds from the water (DOCs), and thats where we get to Stage #4.
Stage #4 Removal of Dissolved Organic Compounds
In a saltwater system, DOCs are typically removed in large part by the implementation of a protein skimmer (foam fractionator), which uses the high specific gravity of saltwater infused with oxygen to remove surficants from the waterwhich represent raw DOCsand collects them in a device that can be emptied and cleaned.
The removal of these DOCs improves water quality and clarity dramatically, and lowers the overall bioload of the system by removing a large percentage of the overall organic waste produced by the tank before it even has a chance to be broken down into ammonia.
Many may note that protein skimmers are a saltwater device not applicable to freshwater. However, some designers do manufacture skimmers for use with freshwater. Theyre generally more expensive, and cannot rely on merely employing a passive process like the Venturi valve skimmers used in saltwater set-ups. The reason is that the far lower surface tension of freshwater requires a great deal more energy to create enough agitation to separate these surficants from the water such that the effluent can be collected in substantial quantity.
Below is a link to one such skimmer product designed for use with freshwater tanks:
http://www.tmc-ltd.co.uk/pond/sander-freshskim.asp
These products are typically paired with ozonators to maximize their efficiency in removing DOCs from the water. In my set-up, both the skimmer and the ozonator would reside in the area of the sump walled off for such devices.
Also in the sump is a water pump with a check valve to return filtered water back to the tank. The vertical placement of the bulkhead, perhaps used in conjunction with a reverse undergravel filter, could be implemented to prevent solid waste from settling at the bottom of the tank, instead forcing it to the surface. Others here have a great deal more experience with this than I do, so I welcome ideas about how to build a system that would allow for the use of substrate in a big tank without the need to vacuum gravela task that becomes tedious and impractical in such a large tankespecially one with a depth greater than a couple of feet.
Stage #5 In-Tank Biological Filtration
Stage #5 doesnt contribute greatly to the overall filtration design, but merely consists of the natural aerobic bacteria that will grow on the substrate and other aquascaping in the tank, like pebbles, rocks, driftwood, etc. This stage will likely contribute minimally to the overall filtration process, but is worth mentioning nonetheless.
Conclusion
I firmly believe that this filter design is superior to just about any other system out there, and that if one has the space to house it, this system comes as close as possible to achieving a closed system with the technology currently available to the hobbyist.
The design also infuses a great deal of oxygen into the water, which can be enhanced by placing additional aerators in the sump.
The system is designed to address all of the polluting agents found in aquarium waterammonia, nitrite, nitrate and dissolved organic compounds.
I welcome feedback on my design, whether positive or negative. Did I miss anything? What are the potential problems with what Ive proposed?
I don't know if it was already done because I only read the original post so far 
