Understanding Mechanical Filtration

Syd Mitchell's picture
Submitted by Syd Mitchell on Tue, 02/20/2018 - 13:41

By Syd Mitchell

A replica of a natural lake
Figure 1 shows the only essential elements for a koi pond - the bio-filter and a pump to circulate water through it. This is far short of what is normally regarded as a “proper” Koi pond but Koi would be quite content in such a simple system because, in the natural lakes where fish evolved, the only form of filtration they have is biological filtration due to naturally occurring populations of bacteria.  These bugs remove the main natural source of pollution in those lakes, the ammonia from fish, by converting it to nitrite, then nitrate which is used as a food by the plants in the lake.

Koi ponds sometimes have plants in them but there are rarely sufficient plants to remove all the nitrate the biofilter produces. Excess nitrate is usually removed by water changes but, with that exception, my earlier controversial statement, about the only essential element being the biofilter, should now make more sense. The above drawing is actually a replica of a natural lake, but on a much smaller scale.  To turn the replica natural lake into a Koi pond, some improvements are necessary.  As mentioned earlier in this series, water in Koi ponds contains suspended silt, small particles called fines and fish waste.  Any debris that is heavier than water will, not only settle on the floor of the biofilter where it can regularly be flushed away, but it will settle wherever the water flow is too gentle to keep it suspended. Much of it will settle on the pond floor.  This is undesirable because a build up of silt in the pond would encourage nasty bugs called heterotrophic bacteria which are not welcome in a Koi pond because of their habit of regarding Koi as another meal.

One way to prevent a pond floor silting up would be to vacuum it away every week but this would be a high maintenance solution and an ideal Koi pond should look after itself with as little work as possible on behalf of the pond owner.  What is needed is a simple method of removing suspended debris before it can settle in the pond.  This is called mechanical filtration.

Settlement chambers and electric brushes
Since suspended silt will tend to settle onto the floor whenever the rapid flow of water inside a pipe suddenly slows to a much more gentle flow in a larger space as described earlier, an early form of settlement chamber simply consisted of a large empty chamber. Water carrying silt entered near the bottom at one end, and as it slowed because it was no longer confined inside the pipe, it would drop nearly all suspended material that was heavier than water before leaving at the other end.  Anything that would float obviously wouldn’t settle and went straight through, so an improvement to the design was to put filter brushes into it them. The brushes, not only snagged floating fish waste and strands of blanket weed as they tried to pass, but they have an unexpected effect considering that they are under water.

The bristles are made from nylon or similar plastic materials and these hold, what could be described as an underwater electrostatic charge.  As tiny particles, such as fines, float past a bristle, the charge will attract and hold onto them until the brushes are taken out and rinsed clean.  This is only a weak effect but it explains why filter brushes, even when they are brand new and haven’t had time to build up a sticky bacterial biofilm, manage to collect so much floating debris. After the brushes have been in use for a few months, they will become home to a large colony of nitrifying bacteria.  This adds to the total amount of biofiltration in the pond, which is always a bonus, but the biofilm is also sticky which helps to trap and hold even more floating particles. 

The sticky biofilm is a benefit in that it makes the brushes more effective as a mechanical filter but it has a slight down side.  Brushes can be difficult to clean properly by the normally recommended method of swirling them around in a bucket of pond water. Cleaning in pond water is often recommended because it is well known that tap-water is chlorinated in order to kill bacteria that might otherwise breed in the supply pipes. It is therefore natural to assume that that chlorinated water also kills nitrifying bugs.  Whilst it is true that chlorine is toxic to them, it doesn’t kill them instantly.

Depending on the chlorine level in the tap-water, after an exposure time of several minutes, nitrifying bacteria begin to die. But cleaning bushes doesn’t take several minutes and, if a brush was cleaned with tap-water, only the bugs on the very top of the biofilm would be exposed to chlorinated water anyway.  Those deeper down in the biofilm wouldn’t even know that the ones on the top were being washed with chlorine, so I recommend cleaning brushes under a running tap or with a hose. As long as the flow of water isn’t so strong that it “jet-washes” the biofilm away, the bacterial colony on the brush will not be significantly harmed. A few unfortunate bugs at the top of the biofilm that were near to dying anyway might quickly succumb to the chlorine but removing old, weak bacteria allows room for more vigorous ones to take their place, so rinsing brushes with a hose is a good thing rather than a bad thing.

A practical design
The drawing above shows the design of a typical settlement chamber.  Silt laden water enters at the bottom at one end and leaves at the top at the other end.  For clarity, only three brushes are shown.  In practise, the more brushes, the more floating debris will be collected.

The chamber should be as long as possible to allow more time for silt to settle to the bottom.  If the floor is sloped and has a drain as shown, cleaning can be done in a similar fashion to flushing the biofilter chamber.  To complete the design, two additional valves have been added, one on the inlet and one on the outlet. These can be closed during cleaning to isolate the chamber and prevent unnecessary loss of water from other chambers in the filter system.

Size matters?
Flushing a biofilter chamber is simple because the silt collects in a relatively small area which is within easy reach of the “suction” from the drain when the waste valve is opened. For a settlement chamber to be effective, the path taken by the water flowing through it must be as long as possible in order to allow the maximum amount of time for silt to settle. So a settlement chamber has to be large.  This also obviously means that the floor area is large too.

Whilst a settlement chamber is a very effective way to remove fines, strands of algae and suspended silt, it does have the disadvantage of its physical size. Also, the large floor area of a settlement chamber means that, wherever the drain is positioned, most of the accumulated silt will not be quickly “sucked” down the waste pipe, it may mean that the entire chamber has to be emptied so that any remaining silt can be hosed towards the drain.   So, although being effective at doing its job, its disadvantages are that, it is not only large, but that flushing away the silt may use a lot of water and it may need to be hosed to completely wash the silt away. Wouldn’t it be nice if a settlement chamber could be made smaller and as easy to clean as the biofilter chamber in figure 1?  Well it can, we call it a vortex.

The vortex
The water flow pattern through a vortex is as follows:  Water enters at the bottom of the chamber through a pipe that is offset from the centre so that the water is made to follow a slow helical path from the inlet at the bottom towards the outlet at the top.  (The path isn’t a spiral; spirals are flat and can be drawn on a sheet of paper.  A helical path looks more like a spring).  See drawing below.

Having to complete several circuits before it can leave the vortex, the total length of the path taken by the water is very long so two things happen to silt laden water while it is in a vortex. Firstly, due to the long distance that it has to travel between the inlet and outlet and the fact that it is now travelling very much slower than it did when it was speeding along the pipe leading to the vortex, it has plenty of time for silt to settle out.  In this respect, a vortex is simply a settlement chamber where, instead of a long straight water path, the path is coiled back around itself.  Secondly, friction loss between the sides of the chamber and the water slows the boundary layer even more, causing even more silt to settle out. As before, the accumulated silt can be flushed away by opening the waste valve for a few seconds. For clarity in the diagram, I have not included the brushes that made such an improvement in the original settlement chamber.  There are arguments for and against putting brushes in a vortex.  Much depends on the flow rate so it is impossible to have hard and fast rules. All systems are different, so a “suck it and see” approach is best.  If you don’t have any, try some, if you think they improve the situation, keep them, if not take them out again.

If the vortex, with or without brushes, is now added to the evolving Koi pond design, as in the drawing above, most of the silt will settle out in it because suspended silt will settle at the first opportunity as the water slows from a high speed in the pipe to a very much slower speed in the first chamber it enters – the vortex. It won’t totally solve the silt problem in the biofilter, these will always need an occasional clean but the biofilter floor and the media will stay clean for much longer.  Also, with so much of the silt and fines being regularly flushed to waste, there will be very little left to settle onto the pool floor.  Isolation valves have also been added to allow the individual chambers to be quickly flushed to waste or even drained if necessary, without unnecessary loss of water from other chambers.

The simple mechanical filtration action of a vortex will make a big improvement in the pond water quality. There are other types of mechanical filters and, as with different types of biological filters, these will be covered later in the series when these individual elements are discussed in greater detail but for now, a simple vortex will suffice as the only means of mechanical filtration.

Well not quite.  Since this part of the series is about mechanical filtration, it makes sense to put a simple dome over the 4” pipe leading from the pond to the filter system shown above.  The cover on a bottom drain, not only acts as a mechanical filter to stop twigs, blanket weed and fish from being drawn into the bottom drain, but helps to slightly improve the water circulation too. 


!If you're not havin' FUN, you're not doin' it right