Chemical Filtration and Nutrient Management

Mechanical filtration catches particles (Module 15). Chemical filtration uses media that hold on to substances dissolved in the water: organics that yellow it, medications, metals and other contaminants. The skimmer removes dissolved organics too (Module 16). Chemical media sits after the mechanical filter, so particles don't clog it.

Nutrient management is the wider job of keeping nitrate and phosphate at the right level: not zero and not too high (Module 6). Some of it's done with media, but most of it's done other ways: water changes, algae that are grown and harvested, and bacteria. This module covers all of them and how to tell which one is working.

Activated carbon adsorbs dissolved organic compounds, including the yellow substances that tint water and cut light, along with odor, dyes, medications and the allelopathic chemicals corals release (Michael Maddox, President of IMARCS). It doesn't remove ammonia, nitrite, nitrate or phosphate. Because it removes medications, take the carbon out before medicating, and put fresh carbon in after the treatment to remove what's left.

Rinse new carbon in RO water to wash out the dust. Use it in a mesh bag in a high-flow spot in the sump, or in a reactor (section 4), where all the water passes through it several times a day. Follow the maker's dose. Change it every 4 to 6 weeks, or sooner if the water starts to yellow. A quick check: water in a white bucket shouldn't look yellow. Use good-quality carbon, the recommended amount, and no more flow than it needs, so it doesn't shed fine dust into the tank.

Poly-Filter is a branded product, a chemical filter pad made by Poly-Bio-Marine. It removes contaminants and toxic materials, organic wastes, medications after a treatment, and excess ammonia while the biological filter adjusts. It doesn't remove calcium, magnesium, strontium, barium or alkalinity supplements, so it doesn't strip what you dose.

The pad changes color as it catches things, and the color shows what it has caught; the meanings are printed on the package. A color you don't expect is a clue to find the source. Replace it when it's exhausted and has changed color, and use a new pad for each medication removal; rinsing it again and again makes it clog faster. IMARCS runs Poly-Filter during allelopathy events and after a crash (section 14) (Michael Maddox, President of IMARCS).

In a media reactor, water goes down a center pipe to a plate at the bottom, then rises up through the media and out the top. Every drop that enters has to pass through the media. Slower flow means longer contact with the media. Water can also cut channels through a bed of media and skip most of it.

Each media wants its own flow. Carbon is soft, so it runs very slowly with no tumbling, or it breaks into fine dust. GFO runs at a moderate flow, just enough for a slow, slight tumble that prevents channeling and keeps the grains from fusing into a block; vigorous tumbling grinds it to dust. Never tumble carbon and GFO together; the hard GFO grinds the carbon. Biopellets need enough flow to tumble, and no sponges inside the reactor. Set the flow with a ball valve, starting from the media maker's suggested rate, then watch the bed.

IMARCS rule (Michael Maddox, President of IMARCS): as a general rule, IMARCS doesn't use phosphate binding material. IMARCS manages nutrients with bacteria dosing (Module 8) and water changes (section 7). You still need to know binders, because many systems use them.

Granular ferric oxide (GFO) is iron-based media that binds phosphate. Rinse it until the water runs clear, start with a small amount, and run it in a reactor with a slight tumble (section 4). Replace it when phosphate starts to rise again, often every 4 to 8 weeks. Aluminum oxide binds phosphate and silicate. It isn't completely insoluble in seawater: aluminum can leach out and irritate corals. In IMARCS experience, aluminum oxide irritates anemones (Michael Maddox, President of IMARCS). Some kinds warm up when they first touch water, so pre-wet and rinse them. Lanthanum chloride is dosed as a liquid; it joins with phosphate into solid flakes, so it's dosed in very small amounts just before a filter sock or other mechanical filter that catches them. The flakes can harm fish through their gills.

Whatever the binder, the risk is the same: phosphate dropping too fast stresses animals, and phosphate at zero starves the system (section 6).

Ammonia, nitrate and phosphate aren't simply waste. What matters is how much: at the right concentration they're food, and at too high a concentration they're toxic (Module 6). Zero is harmful, because the organisms that use them go hungry. Nitrate should ideally be no more than 30 ppm (Michael Maddox, President of IMARCS). Phosphate should stay detectable; too much fuels nuisance algae and slows calcifying corals (Module 6).

Stable but non-ideal numbers do far less harm than constantly chasing numbers, which causes swings (Michael Maddox, President of IMARCS). Change any nutrient level slowly. Nutrient levels and alkalinity are linked: generally, lower nutrient levels need lower alkalinity (Michael Maddox, President of IMARCS).

New salt water has no nitrate or phosphate, so a water change dilutes them: a 10% water change takes out about 10% of what's dissolved. Four weekly 10% changes take out about 34%, roughly the same as one 40% change a month. If the inputs stay steady, regular changes make nutrients level off instead of climbing forever: with weekly 10% changes, they never rise to more than double where they started.

Water changes take out nitrate and phosphate evenly. IMARCS manages nutrients with water changes and bacteria dosing (Module 8), and a water change is the first thing to reach for when something's wrong: when in doubt, do a water change (Michael Maddox, President of IMARCS). Match the new water to the system (Module 5).

A refugium is a separate space connected to the system where macroalgae and tiny animals grow. Chaetomorpha (chaeto) is common: it grows fast, is easy to harvest, and gives copepods and amphipods a place to breed. As the algae grow, they take up nitrate and phosphate.

The refugium light usually runs on a reverse photoperiod: on while the display is dark, often for 8 to 12 hours overnight. The algae take up CO2 while the display's corals aren't, which helps keep pH from swinging at night (Module 6). Harvest macroalgae to export the nutrients they've utilized: the nutrients only leave when the algae leave. Algae left to die back release them again. Before you throw harvested chaeto away, shake it gently in tank water to return the pods.

An algae turf scrubber runs water quickly over a rough, brightly lit surface. Algae grow on the screen and take nitrate, phosphate, ammonia and some metals, such as copper, out of the water. Like a refugium, it only exports what you harvest.

Harvest it every 7 to 21 days, and sooner when the growth turns black, fills the scrubber, starts coming loose, or nutrients rise. Keep the growth under 20 mm thick so light still reaches the screen.

Carbon dosing adds a liquid carbon source, such as vodka, vinegar or a product, so bacteria grow and take up nitrate and phosphate (Module 8). It starts low and rises slowly, and the skimmer removes the extra bacteria. Carbon dosing doesn't require biopellets: a liquid carbon source does the job on its own. IMARCS manages nutrients with bacteria dosing (Module 8) and water changes (Michael Maddox, President of IMARCS).

Biopellets are one alternative, a solid carbon source: small pellets of biodegradable plastic, usually PCL or PHA, in a reactor. Bacteria grow on the pellets and feed on them. The pellets must tumble, which sheds the bacteria into the water; aim the reactor outlet at the skimmer intake so they're exported. A typical full dose is 236 mL (1 cup) per 189 L (50 gallons). Soak the pellets in tank water for 24 hours, start with about 25% of the full dose, and add 25% a week. Results take about 4 to 8 weeks; don't change anything else meanwhile.

Watch for: cloudy water, a bacterial bloom that can pull oxygen down (point powerheads at the surface); cyanobacteria if phosphate falls more slowly than nitrate; pH and alkalinity drifting down; and pellets clumping if flow is too low.

In a deep sand bed, at least 10 to 15 cm (4 to 6 inches) of fine sand, the bottom layers get almost no oxygen. Bacteria there turn nitrate into nitrogen gas (Module 6, Module 8). Worms and snails stir the top few centimeters, drawing water deeper. Coarse sand traps detritus and needs to be deeper to form those zones.

Many IMARCS systems have deep sand beds (Michael Maddox, President of IMARCS). They pull nitrate down, but phosphate keeps arriving with food and builds up, so the balance between them drifts. That isn't a big problem unless the system has calcifying animals (Module 6). Phosphate comes out with water changes.

Phosphate can be brought down with water changes, with chemical media and dosing, biologically with bacteria (carbon dosing, biopellets) or algae (refugia, scrubbers), or a combination. They don't all remove the same things. Water changes take everything out evenly. Algae and carbon-fed bacteria take up both nitrate and phosphate. Binders take only phosphate. Sand beds take only nitrate.

That last point matters: a method that removes only one of them pushes the two apart. A deep sand bed can leave phosphate building up; a binder can leave nitrate building up. Biopellets can bring nitrate down faster than phosphate, which can bring cyanobacteria. IMARCS manages nutrients with bacteria dosing (Module 8) and water changes, and doesn't use binders as a general rule (Michael Maddox, President of IMARCS). Whatever is used, change one thing at a time and slowly, so you can see which change did what (Module 1).

Test nitrate and phosphate at least weekly and log every result, so you can see a trend before it becomes a problem. A level rising for weeks means the method isn't keeping up with what goes in; a level falling toward zero means it's removing too much. Retest any reading far from the last one before recording it (Module 7).

Watch for these traps. Algae can hide phosphate: they take it up as fast as it arrives, so a low reading with algae growing can still mean too much phosphate is going in. Units: some nitrate kits report nitrate-nitrogen (N-NO3); multiply by 4.4 to get nitrate. Media: GFO is used up when phosphate starts rising again, and testing the water leaving a reactor shows whether its media is still working.

Allelopathy is corals releasing chemicals that harm their neighbors. Spawning events fill the water with eggs and sperm. During allelopathy events, IMARCS runs Poly-Filter and fresh carbon; carbon also takes up the chemicals corals release. Spawning events get fresh carbon and internal canister filters with micron cartridges (Module 15), not Poly-Filter (Michael Maddox, President of IMARCS).

After a crash (Module 9), put in fresh carbon, add a Poly-Filter, and do large water changes, repeated as often as needed. Keep spare carbon and Poly-Filter pads on hand so they're there when needed. And when in doubt, do a water change (Michael Maddox, President of IMARCS).