ICP testing for reef tanks: ATI, Triton, Oceamo, ICP-Analysis

What ICP measures

Inductively coupled plasma spectrometry vaporises a sample in an argon plasma at around 8,000 K, atomises every element in it, and measures the wavelengths of light each atom emits. The result is a list of element concentrations: calcium, magnesium, strontium, potassium, lithium, boron, plus iodine, bromine, manganese, iron, zinc, copper, and a dozen heavier metals down to lead and uranium.

For a reef tank that means you find out, in one report:

• Macros: Ca, Mg, K, S, Sr, B, Na, Cl. The big-picture salt composition.
• Salinity (by total ionic strength): ICP can derive salinity from the sum of major ion concentrations, more accurate than a refractometer.
• Traces: I, F, Br, Si, plus the depleting elements like Sr that NSW carries at 8 ppm but most reef tanks run down to 5 ppm.
• Heavy metals: Cu, Zn, Ni, Pb, Cd, Sn, Hg. Tested at parts-per-billion. Useful for catching equipment contamination.
• Non-ICP add-ons (varies by lab): Nitrate, phosphate, dissolved organic carbon (DOC), ammonia — measured by separate methods on the same sample, included with premium ICP packages.

When to run an ICP

The three situations where ICP earns its money:

1. New-tank baseline. Run an ICP about 4–6 weeks into a new build, after salt mix has stabilised and you have rock, sand, and circulation but minimal livestock. This is your reference point — you'll compare every future test against it. Catches early problems like an out-of-spec salt mix batch or a heavy-metal-leaching piece of equipment.

2. Quarterly check on a stable system. Once the tank is running and growing coral, an ICP every three months catches slow drift — strontium depletion, potassium depletion, trace contamination from new equipment, salinity creep from evaporation top-off failures. Cheap insurance.

3. Problem diagnosis. Coral is paling, pixelating, or dying in a pattern that doesn't match any obvious chemistry kit reading. The standard tests (Alk, Ca, Mg, pH, NO₃, PO₄) all look fine but something is wrong. ICP often surfaces the missing piece — depleted strontium, low iodine, high copper from a corroding screw, accumulating ammonium from a non-cycling biopellet reactor.

ATI vs Triton vs ICP-Analysis

Three labs dominate the hobby. Picking between them is mostly about geography and what add-ons you want.

ATI Labs (USA, Atlas portal): The North American standard. Tests 33 elements, plus NO₃ and PO₄ on the premium tier. Reports come back with an aquarium-keeper's commentary explaining each element's consensus target. Atlas portal is the cleanest results UI in the hobby. Costs about $50, ships from US. Turnaround 7–10 business days.

Triton Lab (Germany, EU): The original. Triton ICP-OES plus N-DOC includes dissolved organic carbon, which is the parameter Triton's no-water-change method monitors above all else. About €30. Triton's reports are tied into their method recommendations, which is great if you run Triton and somewhat opinionated if you don't. Turnaround 4–7 days in the EU.

ICP-Analysis (Fauna Marin): Cheapest and fastest in Europe at about €30. ICP only — no separate nitrate, phosphate, or DOC test. Reports are number-only with minimal commentary. Good for routine quarterly testing once you know how to read the numbers yourself.

Oceamo (Triton method partner): Uses ICP-MS (mass spectrometry, not OES) for higher precision at the trace end. Used by serious Triton-method tanks where you actually want sub-ppb iodine readings. About €60. oceamo.com.

Mixing labs over time produces apparent drift that isn't real — different labs report slightly different numbers for the same sample. Pick one and stay with it.

Reading the results

Read in this order:

1. Salinity first. If the ICP-derived salinity doesn't match your refractometer, recalibrate. A 35.5 ppt refractometer reading against 33.0 ppt ICP usually means a refractometer drift, not a tank problem.
2. Macros (Ca, Mg, K, S, B, Sr). Calcium and magnesium against your test-kit values — they should match. Potassium often comes in low because most reef hobbyists don't test for it. Strontium tells you whether your salt mix is reef-suitable.
3. Halides (I, Br, F). Iodine depletes fast — many tanks run at 0.02 ppm vs. NSW 0.06 ppm. Iodine deficiency stresses ornamental shrimp, doesn't generally affect coral. Bromine and fluoride are inert; ignore unless they're dramatically off.
4. Heavy metals. Don't panic. Copper at 0.005 ppm is detectable but not toxic (toxicity threshold is ~0.02 ppm). Lead at 0.001 ppm is detectable but not actionable. Randy's overview of trace elements in reef tanks is the right reference for what matters and what is noise.
5. Compare to last quarter. A single ICP gives you a snapshot. Two ICPs three months apart give you a trend, which is what actually matters.

Common flags and fixes

• Low strontium (< 4 ppm). Common after several months of water changes alone. Coral consumes Sr. Top up with a strontium supplement or switch to a salt mix richer in Sr (Tropic Marin Pro Reef, Red Sea Coral Pro).
• Low potassium (< 350 ppm). Most home tests don't measure K, so it depletes invisibly. Acros take K out faster than NSW supplies. Dose K-balance or switch salt brands.
• Elevated copper or zinc. Almost always a corroding piece of equipment. Likely culprits: stainless steel screws or impellers in metal-halide reflectors, non-marine-rated return pump bearings, brass plumbing fittings someone snuck in. Hunt and remove. Carbon-filter for two weeks. Retest.
• Elevated nickel. Hint of stainless contamination. Same fix.
• Low iodine. Normal in skimmed/carbon-filtered tanks. Supplement with Tropic Marin Iodide or Lugol's if you have ornamental shrimp; ignore otherwise.
• Elevated boron (> 8 ppm). Usually a salt mix with elevated boron (some Red Sea mixes). Coral doesn't care. Ignore.

What ICP doesn't catch

ICP measures elements, not molecules. Several things matter to a reef tank that ICP cannot see:

• Alkalinity. ICP measures total carbon if the sample has been acidified, but the dKH reading you care about is the buffer state, not the carbon content. Use a Hanna or Salifert titration kit.
• pH. ICP destroys the sample. Use a probe or test kit at the tank.
• Dissolved gases. O₂ and CO₂ outgas during transit.
• Bacterial biomass. The skimmer-quality, cyano-presence, biofilm-on-rocks questions are biological, not chemical.
• Coral allelopathy. Soft corals release sesquiterpene compounds that stress SPS. Not on any ICP report.
• Ammonia (NH₃/NH₄⁺). Short-lived; some labs include it on the side as a separate non-ICP test.

Read it alongside your normal test-kit log — see the parameter index for the kit-and-target reference for the day-to-day numbers. The ideal reef parameters guide covers what target ranges to compare against.