Basics: Filters
Filters perform 5 basic functions, biological filtration,
mechanical filtration, chemical filtration,
re-oxygenation and creating enough turbulence
such that the water temperature is kept uniform, and detritus is sucked into the
filter. While there are many types of filters, the ones most hobbyists will have
exposure to are the UGF (undergravel filter), powerfilter,
canister, and sponge filter, and how differently
these filters perform the 5 basic functions are what sets them apart. Some other
distinguishing characteristics are noise, ease of service and servicing interval, which
varies tremendously between filter types.
In the choices available to us, the canister filters are all good, but there are a few powerfilters which should be avoided. Watch for proprietary filter media (forces you into a recurring expense) and try to oversize the filter. A new filter sponge is biologically useless and mechanically inefficient (these sponges only start working when they are brown and slimy).
Sponges should be regularly cleaned but never over cleaned, and in most cases, very rarely replaced. Change your carbon every 2 months or don't use it at all (ie: planted tanks). Ammonia removing products are application specific and should not be regularly used outside their applications. Larger tanks (50g+) benefit from multiple filters, whether the same type (serviced alternately) or completely different types.
If noise is an issue (ie: bedroom aquariums), avoid air pump powered filters or accessories. The quietest pumps are powerheads and internal canisters (completely submerged and don't use the air injector), and external canister filters.
Details...
INTRODUCTION
With a multitude of different filter systems made by several manufacturers, and having such different characteristics and strengths, filters are easily the most debated topic, with opinions varying on what is the best and what is really needed. I'm going to try to be as objective as possible, relying on my experience from building my own filters, from selling various filter systems and the feedback from all my customers (who probably have close to every filter model known in the last 10 years). My own filter collection goes back to before the invention of the magnetically held impeller which has become the industry standard, so I've watched today's standards evolve to their current state. My own electro-mechanical engineering experience (in design and environmental stress) will hopefully also help keep my observations as factual as possible. For the purposes of simplicity, I'll only focus on the main filters mentioned above, and just touch on RUGFs, wet/dry filters, fluidized beds, diatom filters and algae scrubbers as they are related.
BIOLOGICAL FILTRATION
This is essentially the most important function a filter performs, and it begins with an understanding of the food chain in an aquarium. You feed your fish some fish food. Their solid wastes drop to the bottom and begins decaying releasing ammonia (NH3/4). The fish are also constantly releasing ammonia from their gills. Ammonia is toxic for fish, and cannot be left to accumulate in the water (would burn their gills). Excluding the diluting effects of water changes, or the effects of plants in the aquarium, NH3/4 would be left to build up if it wasn't for the bacteria which consume and convert it to nitrite (NO2). A different species of bacteria then converts the NO2 into nitrates (NO3).
In nature, the NO3 becomes nitrogen gas and returns to the food chain as the fish food you are dropping into the aquarium. In the small highly compressed environment of an aquarium, the food chain essentially ends with NO3 accumulating, which is then diluted through regular water changes. Plants will consume NH3/4, NO2 and NO3 (in that order of preference), and convert it to cellular growth, but it takes a significant amount of plants to match and keep up with a typical fish-load. It is easier to rely on your filter's biological filtration capability.
The bacteria of interest are nitrifying (the process of oxidizing ammonium salts to nitrite and then to nitrates) and aerobic (uses oxygen). They attach themselves to surfaces, making the surface feel slimy (bio-film). The combination of the moving water and the tremendous surface area of the filter sponge makes an aquarium filter an ideal home for bacteria. All filters have a biological capacity, depending on their flow rate (slow is better) and available surface area (bigger is better). Some filters keep the bacteria closer to the air (wet/dry filters and bio-wheels) resulting in a higher density of bacteria.
MECHANICAL FILTRATION
Mechanical filtration is simply removing detritus (particulates, feces, plant matter, scales, etc) from the water and trapping it where it can dissolve or later be removed. Mechanical filtration is a function of flow rate (high is good) multiplied by filter media efficiency (the smallest size particle which can be trapped). If the surface area is too small, then the filter will clog prematurely, making it ineffective. Some filters (ie: diatom filters) are designed for small particle filtration (water polishers), but their use is supposed to be occasional or they would need to be cleaned very frequently. All filters work better after the bio-film has coated its surfaces.
CHEMICAL FILTRATION
This typically refers to using activated carbon (charcoal or chars in the UK) to remove inorganic toxins from the water. Carbon works by adsorption, where materials get trapped in the carbon's molecular structure. Carbon has a short life expectancy, basically hours to a few weeks, depending on the starting quality of the water. The industry recommendation is that carbon be replaced every 2 months, as it's bond with the contaminants it has trapped will begin to weaken, slowly releasing toxins back into the water.
Another chemical filter is zeolite (cat litter before additives) to absorb ammonia. This goes under various trade names (ie: Amrid) and is really application specific (ie: extended power failure causes biological component of filter to go anaerobic and high fish load will result in a critical ammonia spike).
Also included in chemical filtration are anything being used to modify the water's parameters, such as peat (moss or granules), terbang or oak leaves (which will all soften and acidify water), or dolomite, marble, limestone, aragonite, crushed coral etc (which will all make water more alkaline and hard). Click here to go to another page, for more information on products which change water parameters.
RE-OXYGENATION
This only requires a steady flow rate across the surface of a tank. If the flow down through the tank is uniform, then the re-oxygenated water will displace the O2 poorer water at the tank's bottom. Basically the tank's water will always try to achieve a ratio equilibrium with the atmosphere, so the greater the contact area, the better this equilibrium is maintained. Equilibrium is a ratio, not an exact match. For example water holds far less oxygen than air, especially as the water's temperature is increased. A fish's gills opened up are almost the size of their body.
TURBULENCE
A uniform flow through the tank will reduce any temperature stratification (layering). Directed turbulence is the 2nd main component of mechanical filtration. A good mechanical filter will trap small particles, but turbulence provides the pick-up power to lift and direct the detritus towards the filter's siphoning intake. In this regard, high flow rates are desirable, as is the ability to direct the returning water flow.
UNDERGRAVEL FILTER
The UGF pulls water into gravel and up its riser pipe(s). The gravel substrate is the filter media. It uses either air pump powered airstones in the riser(s), or powerhead(s). A variant available to some powerheads is to reverse the flow down the riser pipe and up through the gravel. This is then called a reverse UGF or RUGF, or RFUGF.
These filters pull the detritus into the gravel where it begins to be broken down by anaerobic bacteria. With a steady flow of water, the gravel is coated in aerobic nitrifying bacteria which removes the NH3/4 and NO2 from the water. Gravel vacuum periodically and you would seem to have a no-maintenance filter, however these filters have no practical chemical filtration capability, limited biological capability (a function of the gravel particulars and aggregate flow rate), and poor mechanical filtration (sucking the detritus into the gravel where it dissolves into organic compounds is not ideal). For heavily planted tanks, the roots choke the UGF vents and it's not possible to use food spikes (it pulls the fertilizer into the water column causing an algae bloom). It's operation can also be defeated by fish digging down to expose a portion of the plates. With all these limitations or characteristics, UGFs are still in common usage, paired up with other filters, or in fry tanks (where impellers would damage baby fish), or used as continuous gravel vacuums with their riser pipe leading to a powerfilter or a canister (which could get backwashed periodically). In lightly loaded tanks and/or with small fish, UGFs perform admirably, often keeping the water crystal clear.
Detritus will settle into the gravel, but not as deeply with an RUGF, making it
easier to gravel vacuum during water changes. A disadvantage is that the intake
of the powerhead in RUGF mode needs to be filtered, (ie: a sponge filter), so the sponge must be cleaned periodically.
POWERFILTER
Also called a HOB FILTER (hang on back), the models available today fall into 3 categories, BASIC, COMPARTMENTED or BIOWHEEL. All hang off your tank edge, are self-priming and pump water through filter media using a magnetically held impeller.
BASIC POWERFILTER
Typically uses a single vertical cartridge insert which is like a pouch of carbon, with some porous material on the outside of the pouch. Without naming any manufacturers or models, this filter design is seriously flawed. The carbon is the chemical filter, and it has a short working life. The porous material on the outside of the pouch (foam outer wrapping) holds your precious nitrifying bacteria, and should ideally never be thrown away, only gently cleaned. How do you keep the good bacteria and replace the carbon? To add insult to injury, the outside material cannot be cleaned many times, so you need to purchase a new insert which is only made by that manufacturer. To make matters worse, the overall foam surface area is small, forcing you to buy new inserts often. Avoid these filters.
If you already own one of these, some people have managed to replace the carbon
while keeping the foam outer wrapping for a few cycles. For the price of a filter,
set this type aside as a spare, and purchase a proper filter.
COMPARTMENTED POWERFILTER
There are several compartmented filters on the market, but the industry standard
is probably the AquaClear model from Hagen. This is a workhorse which many hobbyists
are familiar with. Compartmented filters keep their filter media separate, stacked
in a vertical or horizontal configuration. This allows each media to get the attention
needed, without significantly affecting the others, for example:
- Carbon, replaced every 2 months or more often for poor water conditions
- Sponge media, rinsed and unclogged as needed
- Sintered glass, rinsed every 6 months, partial replacement at 12 months
- Peat, replaced according to effects
- Coral or other calcium leeching products, rinsed to prevent clogging
The AquaClears stack their media horizontally, in a removable filter basket. The following table provides a quick comparison between models. The information came from Hagen's site here.
HAGEN
| Model | Tank Size in | Flow Rate |
| Number | US Gallons | in GPH |
| AquaClear Series |
|---|
| Mini | 20 | 33-100 |
| 150 | 30 | 50-150 |
| 200 | 50 | 66-200 |
| 300 | 100 | 100-300 |
| 500 | 150 | 166-500 |
Specific to AquaClears:
If the filter is level after installation, and the filter media were to get clogged from neglect, the water would overflow back into the tank from the impeller cavity on the right side of the filter. However, due to the various thicknesses of trim pieces used by different aquarium manufacturers, the filter is rarely level and is usually angled forwards or backwards. If angled backwards, and the media were to clog, then your overflow would run down the back (and out) of the filter, on to your floor. Hagen compensates for this by adding a plastic tab which is part of the motor assembly, or snaps into the box. This tab forces the filter to angle forwards. It's my experience that this potential hazard only exists with the smaller filters (Mini and 150), and is most likely to occur with a used filter (ie: garage sale), where the tab has been broken or gotten lost. It's a good thing to be aware of.
Many experienced hobbyists run their AquaClears in a dual sponge configuration. If using carbon, then the carbon sits between the 2 sponges. If there is no carbon (or other non-sponge media), then the sponges are installed with a gap between them. This achieves a very simple and economical bacterial redundancy. When the filter needs to be serviced (ie: flow decreasing), then the filter basket is lifted out, the bottom sponge removed to be cleaned, the top sponge is pushed down to the lower position without being cleaned, and the cleaned sponge now goes into the top position (in other words, rotating their order, and only ever cleaning the sponge which came from the bottom position). As nitrifying bacteria double their population in less than 24 hours, the total destruction of all the bacteria from one sponge would be mostly compensated for in a short period of time. This is also an inexpensive upgrade when you need to increase your biological filtration capability, however as you put more capacity in a single filter, you reduce your tank operation's safety margin. A brief power failure can have disastrous consequences on an over-stocked tank, so this filter upgrade trick is really a short term fix while you plan a more permanent upgrade (larger tank, more or bigger filters, reducing the fish-load etc.
If not using dual sponges, then sponge rinsing should always be done in water which came from the tank (to avoid shock, chorines or chloramines causing excessive die-off in the good nitrifying bacteria).
Like most powerfilters, AquaClears will give many years of service if the impeller is never run dry. These filters need to have water in their impeller well to self-prime.
There is a plumbing size difference at the 300 (300 and 500 share the same diameter pipes, while the smaller models share the same smaller pipe). Extensions are available for both models.
If I was to pick a favorite, it would be the 300. The 500 is just too large for most of my applications. When I get over 60g, I like to use multiple filters. I would use 500s if I was over-wintering Koi or keeping large Oscars.
BIO-WHEEL POWERFILTER
The Marineland bio-wheel filters (Marineland holds the bio-wheel patent) have models which are comparable to both the BASIC and the COMPARTMENTED types. The main difference is the bio-wheel. On these filters, the vertical cartridge(s) can be replaced without significantly affecting the nitrifying bacteria as most of the bacteria are in the bio-wheel. Nitrifying bacteria are aerobic (need oxygen), so the bio-wheel provides a more O2 rich environment for them.
Some models use proprietary filter media which fills quickly, so it's forte is not mechanical filtration, and it's chemical filtration is good to average (depending on the model) typical for powerfilters.
There is a large retailer which sells a stripped down Marineland filter which
has no bio-wheel. This makes it into a BASIC filter, (see comments above). It
is possible to install a separately purchased bio-wheel and change the filter
cover to bring it back to stock condition, but the expense of doing it this way
is usually prohibitive.
A brief overview of the Marineland bio-wheel filters is below and more information can be found at their site
Penguins here,
Emperors here and as an add on for canister filters or powerheads
here.
MARINELAND
| Model | Tank Size in | Flow Rate |
| Number | US Gallons | in GPH |
| Penguin Series |
|---|
| Mini | 20 | 100 |
| 125 | 30 | 125 |
| 170 | 50 | 170 |
| 330 | 70 | 330 |
| Emperor Series |
|---|
| 280 | 50 | 280 |
| 440 | 80 | 400 | >
| Pro Series (not a powerfilter) |
|---|
| 30 | n/a | n/a |
| 60 | n/a | n/a |
CANISTER FILTER
The canister filter comes in 2 basic models whose only difference is in location. The typical canister sits on the floor, and is connected to your aquarium by 2 hoses. The HOT (hang on top) canister hangs from your tank edge, but is otherwise identical to the regular canister filter.
Canister filters are probably the best overall filter (comparing the basic four types), but also the most expensive. They provide excellent biological filtration (large surface area and a slow flow, about x2 to x4 the tank volume per hour). They provide excellent mechanical filtration (large surface area allows tighter weave trapping smaller particles) and the directed flow allows better pickup power. They typically have extra chambers for holding a variety of chemical filters. There are several manufacturers, but the main players are Eheim and Fluval (Hagen) who have been with the industry for decades, and a more recent arrival, FilStar (Rena or Aquarium Pharmaceutical in the US).
Eheim has the distinction of having the most models, most features, most difficult to pronounce filter media names and they are expensive. Some filter media examples: Ehfifix, Ehfikarbon, Ehfisynth, Ehfimech, Ehfisubstrat, Ehfitorf, and Ehfilava. What they lack in Marketing skills, they make up for in filtration. Their Pro series are their top of the line. The Classic series were very good filters in their time, but now lack modern features. The ECCO series were introduced to prove they can make an inexpensive filter, to compete with Fluval and now Filstar, but not compete with their other more-expensive models (a difficult thing to do).
While Eheim may have run out of places to go, Fluval has probably reached their prime, making the best filter they have ever designed. Earlier Fluval series -01 through the -03 and even early -04 suffered from a variety of quirks. The current -04 series has undergone progressively smaller design changes, with the latest models being solid performers.
I've also included Filstar in the data tables below, but I have no personal experience with them. I'll gladly revise this if Rena sends me one to test. Based on customer and newsgroup feedback, besides the occasional production defect from any of these companies, there aren't any current production canister filters which should be avoided. As for favorites, a few years ago, my choice was Eheim, and in a few years, it might be someone else. Based on my limited experience with Eheim & Fluval experience, if I was buying a filter today with my money, it would be a Fluval 304 (power, performance, ease of maintenance, running cost).
EHEIM
| Model | Tank Size in | Flow Rate | Heater |
| Number | US Gallons | in GPH | ? |
| Classic Series |
|---|
| 2211 | 40 | 63 | |
| 2213 | 66 | 103 | |
| 2215 | 93 | 135 | |
| 2217 | 160 | 209 | |
| Professional Series |
|---|
| 2222 | 40 | 63 | |
| 2224 | 66 | 132 | |
| 2226 | 93 | 172 | |
| 2228 | 160 | 198 | |
| 2322 | 40 | 63 | 150 W |
| 2324 | 66 | 132 | 150 W |
| 2326 | 93 | 172 | 180 W |
| 2328 | 160 | 198 | 210 W |
| Wet/Dry Series |
|---|
| 2227 | 93 | 145 | |
| 2327 | 93 | 145 | 180 W |
| 2229 | 160 | 145 | |
| 2329 | 160 | 145 | 210 W |
The table above was made using information from Eheim's web site here.
Eheim also has an ECCO series 2231, 2233 and 2235 (for 35, 60 & 80g according
to their web site, 22, 44 and 66g according to a distributor), but the latest
distributor listings have them replaced by the 2232, 2234 and 2236. In any case,
I was unable to find their specification for filter flow, to include in the table
above.
FLUVAL
| Model | Tank Size in | Flow Rate |
| Number | US Gallons | in GPH |
| -04 Series |
|---|
| 104 | 25 | 125 |
| 204 | 40 | 180 |
| 304 | 70 | 260 |
| 404 | 100 | 340 |
The table above was made using information from Hagen's web site here.
FILSTAR
| Model | Tank Size in | Flow Rate |
| Number | US Gallons | in GPH |
| XP Series |
|---|
| XP1 | 45 | 250 |
| XP2 | 75 | 300 |
| XP3 | 175 | 350 |
The table above was made using information from Aquarium Pharmaceutical's web site here.
Besides the maximum tank size and flow rates data in the tables above, other features
of interest are shut-off and quick-disconnect valves, filter media extra or included,
the ability to re-use filter media, hose opacity (opaque hoses reduce clogging
due to algae), accessories (pre-filters, spray bar, open pipe, air injector, skimmer
and primer). Note that the value of any of these features or accessories depends
on the individual's application. The size of the filter-box will also influence
your cleaning interval.
Interesting observations can be made when comparing tank sizes and flow rates from different manufacturers. Lets look at our options with a 40 gallon tank.
40G COMPARISON
| Model | Tank Size in | Flow Rate | Changes |
| Number | US Gallons | in GPH | per Hour |
| 2222 | 40 | 63 | 1.6 |
| 204 | 40 | 180 | 4.5 |
| XP1 | 45 | 250 | 6.3 |
There isn't a straight-forward recipe for designing a good filter. For example: the lower your flow rate (within limits), the better your biological and chemical filtration. The higher your flow rate, the better the pick-up power will be inside the aquarium, and if your filtration screen size is small enough, you will have much better mechanical filtration. The higher the flow rate, the quicker (potentially) your sponge will clog, reducing your servicing interval time, but doing a more effective job of removing detritus. However, a sponge which clogs too quickly, may drop your flow rates too fast, and without fresh oxygen from the water flow, your nitrifying bacteria will have some die off. All filters have some bypass (ability of water to bypass clogged filter media) and this is essential to keep the motor from overheating. Too much bypass removes some of the filtering functionality. There are different strategies being used by the 3 manufacturers, in terms of filter screen size, flow rates and bypass.
From a servicing perspective, your filter's media should be stacked and of a composition that causes them to all become clogged at the same time. This gives you the longest interval between servicing. However with the requirement to keep our nitrifying bacteria alive, this changes everything. The ideal filter should just never get clogged. This is not as silly as it sounds. Detritus caught in the filter slowly dissolves, so with a large enough ratio of filter media to fish-load, your filters would clean themselves as quickly as they collected debris. This is evident on overfiltered tanks. If you can overfilter (while reducing the extra turbulence of the oversized filter), then you would find very little reason to have to clean your filters (always finding them clean, every time you checked).
It would be interesting if someone made transparent filter housings. I'd like to see what the flow patterns looked like under different load conditions. Fluval appears to have approached this issue (dealing with clogging and keeping bacteria alive) by using a very large front end sponge filter. During normal operation, the filter media compartments behind the sponge slowly clog up, and although undesirable, this does not cause any disaster. These compartments hold the chemical filtration (ie: carbon and ammonia removers) which work best at low flow, and sintered glass (Bio-max) which has mostly anaerobic bacteria, which by their nature, do not require oxygen. The filter then begins bypassing the chemical filtration compartments, but still pushing water through the sponge, which we are all led to believe is the main housing for bacteria, because of the tremendous surface area it has.
It is far less evident to understand how Eheim approaches this problem. They appear
to allow less bypass, and use a lower flow but perhaps with higher torque (not
as influenced by initial clogging). The media used also appears to spread the
nitrifying bacteria more uniformly throughout the canister.
The next tank size where filters could be compared is the 66 to 75 gallon range,
shown in the table below:
66-75G COMPARISON
| Model | Tank Size in | Flow Rate | Changes |
| Number | US Gallons | in GPH | per Hour |
| 2213 | 66 | 103 | 1.5 |
| 2224 | 66 | 132 | 2.0 |
| 304 | 70 | 260 | 3.7 |
| XP2 | 75 | 300 | 4.0 |
The flow range between manufacturers has now decreased from a 4 fold difference to under 3 fold. As tank sizes get larger, flow rates (as compared to tank sizes) decrease. This is valid for all filter designs. Powerfilters typically start much higher (x5) before dropping. Canister filters start lower (x2 to x4) and then drop further as the tank is larger. Pond filters (as a comparison) are typically rated at x 0.5 the volume of water per hour.
SPONGE FILTER
Either air powered or using a powerhead, the sponge filter has several advantages, such as the ability to be installed at any depth and they provide complete flexibility to directing the water flow (only canisters are slightly better due to their smaller hose getting into tighter locations). The sponge material used is a very tight weave, operating as a cheap water polisher (only the diatom filter is better at polishing, but crystal clear water is achievable with any filter). These filters are safe to use in a fry tank (probably the best filter for this), and quickly set-up, moved and cleaned (ideal for any temporary tanks, like quarantines). With their small size, they can be left operating in a main tank and quickly moved with a full biological capacity to another location. Also useful for augmenting existing filtration or adding a little extra turbulence where needed. A disadvantage is that they have no chemical capability and limited mechanical capability. A variant of the sponge filter (or a variant of the canister) is the internal canister filter. These have the same characteristics of the sponge filter, but use a more porous sponge, so they have better mechanical filtration but poorer water polishing capability.
SERVICING
Servicing varies according to the fish-load and the surface area of the filtration media. Large bodied fish are a heavier fish-load than smaller fish, even if the total linear inches are the same. The food chain starts with the fish food you add into the tank (and vegetation consumed in the tank). The more food consumed, the quicker the filter fills up. A certain amount of detritus dissolves back into the water, so larger filters with a greater contact area, will allow more to dissolve back into the water, to be removed by water changes. If you consider gravel vacuuming a normal part of any aquarium's maintenance, then UGFs are almost maintenance-free, requiring only the once or twice a year vacuuming through the riser pipes. UGFs use the entire surface area above their plates as their filtration media. Canister filters are the next in terms of requiring little service, primarily because they have a large filtration media surface area. Powerfilters due to their high flow rates and limited surface area will need to be serviced the most frequently, but they are also the quickest to service. Sponge filters by design are usually light duty filters so servicing is not comparable with the other systems mentioned. They cleaner the water, the longer they can go without servicing, performing their biological filtration and water polishing.
CLOSING REMARKS
Large tanks benefit from running multiple filters, each typically sized for 2/3 of the tank's capacity. The difference between a filter breakdown being fatal or being a nuisance depends on whether it's the only filter, or running in tandem with other filters. A stopped filter will oxygen-starve the nitrifying bacteria inside, and lose it's biological capability. The delay in buying parts to repair a filter can be fatal to an aquarium's inhabitants.
Stagnant filter water should never be poured into an aquarium (ie: after a prolonged power failure). During a power failure, float the filter sponges in the tank to preserve some of the bacteria. How critical the filter is depends on the tank's fish-load. The higher the fish-load, the less operating margin you have, and the more dependant you are on the filters. As fish grow larger, it is sometimes necessary to upgrade your filter to keep up with the fish-load. The amount of filtration needed is directly proportional to the amount of food dropped into a tank (the start and the end of the food chain), so any filter changes or problems should be accompanied by a reduction in feeding. Most filters do nothing or very little about the accumulation of dissolved organic compounds, so regular water changes are a must to keep these diluted. The industry standard is 20% a week or 10% a week on planted tanks, removed from the gravel with a gravel-vacuum.
|
