I’m very happy with the 0.02% mixed indicator method of measuring low levels of alkalinity.  The method has been pivotal in robustly establishing suitable water for the low alkalinity Shrimphaus, and combined with deliberate copper removal, supports well-breeding populations of both Neocaridina and Caridina freshwater shrimp.

I also like using Oceamo‘s mass-spec based analytical service where their freshwater ICP-MS method for quantifying copper to parts per billion has been a real difference-maker.

That said, Oceamo seems to consistently report higher KH than what I measure.  In the latest go, I measured the Shrimphaus alkalinity at 15.2 ppm CaCO₃ equivalents, but the Oceamo test came back at 2.15 dKH which is 38.4 ppm CaCO₃ equivalents.  No idea why the discrepancy, but from extensive previous measurements I’m very comfortable that I easily get 1 ppm precision with the mixed-indicator method, so there is no way measurement variability can account for the difference.  Note that the fancy mass-spec method Oceamo uses only reports elemental concentrations – for measuring KH they use a chemical titration method (using a pH meter rather than a colourimetric indicator).

I have carefully calibrated the accuracy of the mixed-indicator method previously using defined solutions of NaHCO₃, but let’s roll that out one more time, just to be sure.  Full experimental write-up below.

TL;DR:  I’m happy with the accuracy of the mixed-indicator method. 

I have requested the raw titration data-traces from Oceamo and we’ll see what those show.

Gravimetric KH Titration — Method Validation

7 November 2026

Objective

Evaluate the accuracy of a gravimetric acid titration using bromocresol green–methyl red mixed indicator for measuring low alkalinity (10–20 ppm as CaCO₃) in aquarium water.

Reagents and equipment

• Titrant: APC Pure 50 mM HCl volumetric standard, diluted fivefold gravimetrically to nominal 10 mM.
• Primary reference: APC Pure sodium hydrogen carbonate (NaHCO₃), 99% minimum assay, product ECE6897-2X, expiry March 2031.
• Diluent: Halfords battery top-up water, conductivity 2 µS/cm.
• Indicator: 0.02% bromocresol green–methyl red mixed indicator.
• Equipment: Analytical milligram balance for titrations; separate 1 g resolution balance, checked against standard weights, for bulk water preparation.
• Temperature: Approximately 20°C.

All preparations and titrations were performed gravimetrically. Results are expressed in ppm alkalinity as CaCO₃ equivalent.

Standard preparation

Stock solution: 413 mg NaHCO₃ dissolved in 310 g Halfords water, giving approximately 793 ppm nominal alkalinity (assuming approximately 100% assay). Measured conductivity was 1,414 µS/cm, consistent with the expected concentration.

Working standard: 986 mg stock solution added to 20,970 mg Halfords water, producing 21,956 mg of solution at 35.61 ppm nominal alkalinity.

A second, lower-concentration standard was prepared by mixing 1,519 mg of the 35.61 ppm standard with 3,474 mg Halfords water, yielding 10.83 ppm nominal alkalinity.

Titration procedure

1. Weigh approximately 5 g sample directly into a plastic cuvette.
2. Add approximately 67–74 mg mixed indicator.
3. Add nominal 10 mM HCl incrementally, recording total cuvette mass and observing colour.
4. Approach the endpoint in small increments.
5. Define the endpoint as the last colourless state immediately preceding the first discernible pink, rather than the first disappearance of blue.
6. Calculate alkalinity from the mass of acid added, assuming 10 µmol H⁺ per gram of titrant.

No blank or indicator corrections were applied.

Experimental results

Blank and bicarbonate standards

All KH values in ppm CaCO₃ equivalent.

The Halfords blank required 8 mg acid to reach colourless and became pale pink after 12 mg. Its apparent alkalinity may include indicator and endpoint effects; no blank correction was applied.

The two standards demonstrated a consistent negative bias of approximately 3–4% at the operational colourless endpoint.

Endpoint colour comparison

The first pink endpoint gave improved recovery at the higher concentration but overestimated the lower standard. The colourless endpoint provided more consistent proportional recovery.

Shrimphaus aquarium-water matrix test

Two approximately 5 g samples were titrated:

Unspiked sample

• Shrimphaus water: 4,995 mg
• Indicator: 67 mg
• Acid to colourless: 154 mg
• Measured alkalinity: 15.43 ppm

Matrix-spiked sample

• Shrimphaus water: 4,030 mg
• 35.61 ppm standard: 975 mg
• Indicator: 68 mg
• Acid to colourless: 191 mg
• Expected combined alkalinity: 19.36 ppm
• Measured alkalinity: 19.09 ppm

Accounting for the reduced quantity of aquarium water in the spiked sample, incremental bicarbonate recovery was approximately 96.2%.

Summary of recovery

Conclusions

1. The gravimetric mixed-indicator method demonstrated consistent recovery of approximately 96–97% across the tested alkalinity range.
2. The customary colourless endpoint produces a modest negative systematic bias of approximately 3–4%. At 15.2 ppm measured alkalinity, this implies an estimated actual alkalinity of approximately 15.7–15.8 ppm.
3. The matrix-spike experiment showed essentially the same recovery as the reference bicarbonate solutions, providing no evidence of substantial interference from the aquarium-water matrix.
4. The indicator transition spans a narrow range of acid addition, with approximately 0.6–1.2 ppm separating the recorded colourless and first-pink observations. The actual colourless interval may be narrower.
5. No correction was made for indicator alkalinity or the apparent blank. Their contributions are incorporated into the observed overall method performance but have not been independently quantified.
6. Practical accuracy of approximately ±1 ppm at 10–20 ppm alkalinity appears reasonable, although formal uncertainty and repeatability have not been established through replicated measurements.

Overall conclusion: The existing gravimetric titration procedure, using the last colourless state before the first discernible pink as its endpoint, is suitable for low-alkalinity aquarium measurements. A calibration correction of approximately +4% is supported by the present experiments, subject to the limitations of the nominal reference standards and the small number of determinations.