Alan’s Book

Chapter 8 Filtration Systems

There is a well known saying – “many roads lead to Rome”; these words are very relevant to filtration systems! Although there are many ways of providing filtration systems for koi ponds this chapter will provide information on how to achieve good filtration by cost effective and simple means!!

The term “filter” conjures up images of an air or oil filter such as are used in cars; these filters are used to trap particles and remove them from the flow.

A filter in the context of a pond is vastly different; it is a living entity requiring the correct conditions in order to survive and operate. The ideal pond filter operates as a biological filter; living organisms ideally progressively process the various forms of Nitrogen (N) and organics which the fish produce, into less toxic forms.

The inhabitants of a pond filter are bacteria of many types and species, together with fungi and other minute protozoan type organisms; all of these creatures contribute to keeping water clean and pure in terms of fish keeping. They only exist and proliferate when suitable conditions are available to them; food, a suitable habitat and the correct environment in terms of pH, temperature and oxygen levels are some of the most important environmental factors which determine the success or otherwise of the diverse range of creatures which we need to assist us to provide good water conditions for our koi.

Our koi ponds only harness a small part of Mother Nature’s resources as they are usually devoid of plants. In the early days of keeping fish it became obvious that the ammonia produced by fish, which is extremely toxic, was a limiting factor. The quantity of fish which a pond could sustain was very small. The bigger the pond the more fish it could sustain; big ponds will hold more water and also have a higher surface area in contact with the pond water, and providing the depths are the same, it has a larger surface area in contact with the atmosphere.

The surface area of the pond in contact with the pond water provides a home for the bacteria, and the ammonia supplied by the fish provides a food source for these bacteria. The larger the surface area, the larger the colonies of bacteria which can be sustained and hence more ammonia may be removed.

The surface area in contact with the atmosphere (air) provides an interface for oxygen to diffuse into the water; in other words a large surface area will provide larger lung for the whole pond system.

Early Filtration Systems

Most pond keepers are limited by space so in order to increase the surface area available for the bacteria to live on, filter systems were added to ponds. These early “filters” were filled with materials which had high surface areas; typically gravel or similar materials such as Canterbury Spar were used. The addition of these materials effectively increased the area of the pond with which the water has contact. In order to expose the increased surface area to the pond water, it was necessary to pump the pond water through the filter.

Early pond filters acted as mechanical filters which trapped debris in the flow as well as allowing beneficial bacterias to colonise them and remove the ammonia from the water as it passed through it.

All filters which trap organic debris such as fish and food waste and algae, encourage the growth of bacteria and other organisms which feed upon this waste (heterotrophs); generally speaking these organisms form large and dense colonies which can, after a short time, cause blockages in the filter material. This can mean that the media which houses the ammonia>nitrite>nitrate removing bacteria (autotrophs) has limited amounts of water flowing through it due to “tracking” (uneven flows); the effective surface area of the filter is reduced and less ammonia can be processed.

Areas of the filter which become blocked do not have water flowing through them and are starved of the oxygen which is dissolved in the water. They become sour or “anaerobic” and these conditions are ideal for undesirable bacteria and to grow. High levels of these undesirable bacteria can allow toxins to arise which pollute the pond water. The undesirable bacteria which favour such conditions together with the toxins they produce, can adversely affect the health of the koi.

In order to control the amounts of organic processors (heterotrophs), their food source needs to be controlled. This is effectively done by reducing the quantity of organic matter which can gain access to the filter material which processes the inorganic wastes of ammonia, nitrite and nitrate.

One of the most important design features of a filter system is the removal of solid waste from the system. This vastly reduces the work required by the rest of the filter system.

The “vortex”, or an area where solid waste can settle, prevents much of the organic matter from entering the biological stage where ammonia is efficiently processed. Preventing the build up of organic processors (heterotrophs) in this inorganic stage is desirable as heterotrophs can compete for space and oxygen in the filter media as well as preying upon the autotrophs when their organic food supplies are low.

Early improvements in filter design

One of the first improvements in filter design was invented by Peter Waddington. The problems associated with blockages in early filter systems encouraged him to include a settlement area in his filter designs. This settlement area used the Vortex principle which encourages settlement at the centre of a circular chamber where water velocities are at their lowest. Take a bucket of water and swirl it in a clockwise direction (Northern hemisphere- anticlockwise if you are in the Southern hemisphere) add a handful of sand and you will find that it will all settle in the centre of the bucket. Using this fact of Physics allowed him to separate the heavy solids in the vortex chamber reducing the amount which could enter the biological stage of his filters and thereby vastly improve the filters of the day!

The separated organic matter, conveniently collected in a settlement area can be discharged to waste; if left within the pond system, the decomposition required and achieved by the various heterotrophs can utilise large quantities of the valuable oxygen available in the pond water.

Additional ammonia is also produced when organic waste is reduced by the heterotrophic bacteria; this adds an unnecessary burden to the autotrophs which must then process it through to NO3>N.

Modern Filtration Systems and how they perform efficiently

Modern filtration systems such as the “Nexus Easy” provide areas where heavy waste can be settled, together with pre filters which trap the lighter, finer particles with media such as K1 in the static mode. This means that the heterotrophic bacteria mainly colonise the static K1 where the organic debris is trapped, leaving the fluidised K1 relatively clean of organics such that they can process the dissolved ammonia without undue competition. Backwashing this pre filter eliminates the trapped organics and reduces the colonies of heterotrophic bacteria at the same time as the solids are flushed to waste when the chamber is emptied.

The commercially available systems as described above can provide an excellent, cost and space effective means of filtration for koi ponds however when these systems are used as stand alone filters they do have certain limitations. For example this type of filter system can only be used with an “in line” heat exchanger; “In line” heat exchangers tend to reduce the flow through the pond and when U/Vs are also added to the system substantial flow reductions can be experienced unless higher powered circulation pumps and larger bore pond return pipes are used.

When used in soft water areas they do not provide a separate chamber which allows the inclusion of buffering materials or an area of non turbulent water where a host of aquatic creatures ranging from protozoa to snails and water louse, can live. The silent and often unseen creatures which colonise such areas contribute in a natural way to the reduction of the detritus which settles out and the tiny particles of suspended detritus are removed by a variety of filter feeders who share this space. Since introducing snails and water louse to my final chambers which house coarse grade mattala matting, I have not found it necessary to clean these chambers whereas previously they needed cleaning at monthly intervals.

I use these commercial filters on some of my indoor ponds very successfully; although I have no need for additional buffering due to the high alkalinity of my water, I do provide a chamber following these units to house the aquatic creatures mentioned above.

Some clients who have installed these filters on outdoor ponds have experienced some management difficulties when blanket weed is present. Blanketweed can severely restrict the flow of water when it is trapped by the prefilter; the inclusion of brushes in the inlet section around the prefilter can trap the blanket weed and reduce these blockages.

My preference for a filtration system follows the same principle as the commercial unit which is combined in one plastic container, however for practical purposes such as trapping blanket weed before it blocks the prefilter, I prefer to use a settlement chamber which can intercept any blanket weed by way of coarse matting or filter brushes before it enters the relatively easily blocked prefilter. A further chamber following the biological fluidised K1 is a useful location for a radiator type stainless steel heat exchanger which does not reduces system flows. This further, last chamber is also a useful place to house buffering material such as lithaqua and oyster shells where water supplies are soft. It also allows an area where other medias which house the creatures mentioned above can be located.

Filter Media

K1 is a valuable media for housing the major ammonia reducing bacteria as it maintains a constant surface area for bacterial growth due to its fluidised state which prevents blockages and subsequent “tracking”. My preference for subsequent media is a coarse grade of matting known as mattala matting. This provides a place for the creatures mentioned above to live and consume the small amounts of organic debris and dead bacteria which leave the biological fluidised stage and allows us to utilise yet more of Mother Nature’s willing workforce from the aquatic environment. Other microscopic organisms can also colonise this area; these filter feeders can substantially improve pond clarity by removing tiny particles of suspended solids from the water flowing through them.

Air Pumps

The use of air pumps and air diffusers in pond systems helps to maintain oxygen levels throughout the whole water column by bringing the bottom water to the pond surface and continually exposing it to the atmosphere where it absorbs essential oxygen; trickle towers used in conjunction with this means of adding oxygen to the pond water means that large quantities of pond water are “de gassed” and any nitrogen gas from compressed air, along with carbon dioxide, are allowed to return to atmosphere. Air pumps are used to keep K1 fluidised providing an aerobic environment for the ammonia removing bacteria, a means of keeping the media fluidised and gassing off nitrous oxide during conversion from nitrite to nitrate thereby reducing nitrate levels.

Other Filtration Methods

Other filtration systems are employed for fish keeping including bubble bead and sand filters. I tend to avoid the use of these units preferring not to pass pond water through the waste matter which is inevitably trapped by the mechanical process of filtration upon which these units rely. Any trapped organic matter deteriorates very quickly and water flowing through it can dissolve undesirable products from it.

Nitrate – The end product of filtration in koi ponds.

The effect of nitrate levels in the water on koi health and growth rates is a controversial subject in the koi hobby. Many believe that nitrate levels above 10 ppm compromises the immune system of the koi, and makes the koi more susceptible to diseases. Others seem to believe the koi can tolerate higher nitrate levels acceptably. But no one in the hobby claims higher nitrate levels are good for the koi. Control of nitrate levels can be accomplished in several ways.

Large trickle towers or showers can be incorporated into the filter system designs which can prevent the build up of nitrate by directly de-gassing nitrogen oxides from the filter into the air and out of the water. The major problem with this type of system is heat loss as the large surfaces areas of water can be cooled significantly during cold weather periods. Enclosing the trickle tower to reduce heat loss may prevent the correct exposure to the atmosphere and inhibit nitrogen release.

Another way to control nitrates is to use enough pond plant life to consume some of the ammonia directly from the pond water which will reduce the quantity of nitrate being produced by the biological filter system.

With careful initial planning of the pond and filtration system this method of nitrate removal can soften the impact that a formal koi pond sometimes imposes in a garden situation and can work acceptably to control nitrate levels at high plant life levels, low fish stocking densities, and low feed rates. The limitations here are that plants will only consume ammonia/nitrates during their growing season so this method alone will not accomplish the desired reductions during the plant’s dormant period.

Nitrates can also be reduced by arranging for some of the water which is returned to the pond after filtration to pass slowly through a gravel bed or other such media to encourage anaerobic areas where the nitrate may be reduced to atmospheric nitrogen and oxygen.

The term “anaerobic” sometimes causes confusion with respect to nitrate reduction. Blocked areas in a filter bed become anaerobic (unable to obtain oxygen) because they are blocked, usually by fish/food waste and cannot access the oxygen within the water column These conditions are undesirable as many of the pathogenic (harmful) bacteria use this waste as a food source and produce toxins such as hydrogen sulphide (rotten egg smell) into the pond water.

In a nitrate reducing filter bed the aim is to circulate relative clean water very slowly through a filter bed, minimizing the oxygen available from the water column for the biological process of nitrate reduction. Under these conditions some of the oxygen (O2) is taken from the nitrate (NO3) and the end process is nitrogen (N) and oxygen (O2). So it could be argued that this method of nitrate reduction is not truly anaerobic as oxygen is using in the process.

Yet another approach for controlling nitrate levels is simple water exchange. Water exchange works for nitrate control if you have source water low in nitrates, and the source water is cheap enough to afford the water exchange.

Some koi ponds use a selection of the above techniques as well as aerated and fluidised K1 media in the main biological filter to control nitrate levels. Good filter design, reasonable water exchange rates, low stocking densities, and some level of desirable pond plants should ensure low nitrate levels.

Filter Activity

The temperature of the water largely governs the speed at which the bacteria multiply, so time to filter maturity is usually governed by the pond temperature. The higher the temperature within the normal range, the faster the filter bacteria will both multiply and process their respective food sources.

The nitrogen cycle describes the process of conversion of ammonia through nitrite to nitrate and in an ideal situation the nitrates would be used as plant food; the fish eat the plants and the cycle is complete.

The bacteria which deal with ammonia are relatively robust and are not usually affected by the medicated pond treatments which we sometimes find necessary for the welfare and health of our koi. The bacterial colonies which process the nitrite however are quite sensitive to changes in the environment and can be killed off or seriously reduced by some treatments. Being a much slower growing bacterium, in newly established ponds it invariably takes much longer to reach adequate levels where it can process the nitrite as quickly as the ammonia. Due to this nitrite levels can persist for some time. Test kits are available similar to the one for ammonia described below.

During periods when nitrite levels persist in our ponds the toxicity of nitrite may be reduced by the addition of salt (sodium chloride). This should be added at the rate of 1 Kg/tonne (1000 L or 220 imperial gallons) giving a sodium chloride concentration of 0.1%. It is important to do this if nitrite levels persist in a pond as nitrite interferes with the ability of the blood to carry oxygen around the body.

The amount of ammonia in a water sample can be simply measured by liquid test kits where drops of a reagent are added to a measured amount of water to be tested and the amount of ammonia is determined by the colour of the sample when compared with a colour card provided with the test kit.

Preparing the filter before adding the koi

As ammonia and nitrite are toxic to fish it would be wise to have colonies of bacteria operating to consume these toxins before our koi are introduced to our ponds.

This can be done by adding ammonia and some pellets to the system which will provide a food source to initiate the nitrogen cycle. The necessary bacteria are always present in the pond environment and will respond to the suitable food supply of ammonia by enlarging their colonies in proportion to the food source supply. Bacterial cultures are also available which can reduce the time for the filter to become effective. The addition of mature filter media of the same type as is being utilised in the filter being matured, will help to ensure that the specific species will be cultured on the media used. This can be important as the nitrite processing bacteria which colonises K1, seems to differ from that which colonises Jap matting.

Detail – Once a suitable food source (ammonia) is available in a pond, the bacteria which process it to less toxic compounds will multiply in response to it.

It is important to keep the levels of ammonia fairly low (circa 0.5 to 1 ppm) during the period when bacterial colonies are multiplying as the bacteria which process nitrite (produced from ammonia) do not operate well in an environment with a high level of ammonia.

The ammonia reducing bacteria nitrosomonas produce nitrite which is a food source for other bacteria nitrobacter which converts nitrite to the relatively low toxicity nitrate. Nitrate is a plant food; the koi eat the plants and the cycle known as the Nitrogen Cycle is complete. In some filters with strong aeration the production of nitrate is reduced when the gases nitric oxide and nitrous oxide are liberated; it is a preferable to keep nitrate levels as low as possible.

Ammonium sulphate and household ammonia are a cheap source of ammonia and may be added to a pond to produce an ammonia reading of about 0.5>1.0 ppm. Add small quantities until the desired levels are achieved.

Koi pellets contain organic material so it is wise to add a handful of pellets to the system being conditioned for koi keeping and these pellets will provide a carbon source for the bacteria to use as they multiply and other sources of food which will help other bacterial colonies multiply in readiness for when the koi are introduced and fed with pellets

Temperature plays an important role in determining how long the nitrogen cycle takes to complete and this makes it difficult to be precise about how long it will take. At higher temperatures the time is reduced considerably.

The order in which things happen can be precisely predicted providing the environment is aerobic (rich in oxygen).

As stated in a previous chapter bacterial action performs best at a pH of 8.1 and an alkalinity value of between 100 and 150 ppm. An ideal temperature would be about 30C but temperatures equating to normal pond operating temperatures may be best to replicate.

Great care should be taken when filter systems are being established such that the KH is maintained at a suitable level to support the pH. Failure to monitor KH and pH levels may result in a “pH crash”. If this is allowed to happen, the filter bacterial action will cease.

Bicarbonate of soda may be used to maintain the alkalinity and pH at the required levels”.

Requirements for optimal biofiltration

? Enough alkalinity to absorb acid from biofiltration, pH range 7.5 to 8.5

? Carbon source for biofilm development

? Phosphate source for biofilm development

? Variety of trace minerals, particularly important for nitrite conversion

? High oxygen content at biofilm, preferably high air flow too for low nitrates

? Enough surface area to grow sufficient biofilm

? Regular disposal of dead biofilm (finite lifetime) and other solids

? Do not allow the filter to freeze during winter, earliest possible start in spring, try to keep media active through winter if possible.

Filtration is a big and fascinating subject; many of us hold varying views on how it should be done. My views are that we should keep it as simple as possible and where possible harness the essential requirements laid down by Mother Nature; these include good exposure to the atmosphere and the provision of a habitat for diverse species of organisms.

This entry was posted in Miscellaneous. Bookmark the permalink.

Comments are closed.