Why a low conductivity reading does not prove silica-free blending water

Key takeaways

• Conductivity is blind to silica. Silica is a weakly ionised species, so it “cannot be easily detected using conductivity” (Merck Millipore).
• It is an early-warning ion. The same source reports that silica is released from ion-exchange resin “before the drop in resistivity is detected,” which makes it a better indicator of resin exhaustion than the conductivity meter on the skid.
• Natural waters carry a lot of it. Australian utilities report mean silica from 0.6 to 104 mg/L across regions, and the national guideline sets an aesthetic ceiling of 80 mg SiO₂/L to limit scale (NHMRC).
• There is no spirits-specific silica limit we can cite. We found no published sensory threshold for silica in diluted spirit, so the case for measuring it rests on process control and scale, not on a proven flavour effect.

A low conductivity reading tells you that very little ionic material is present in a water. It does not tell you that silica is absent, because silica barely contributes to conductivity in the first place. For blending water, the practical consequence is that a certificate listing only conductivity or TDS leaves one of the earliest-failing contaminants unchecked, and a buyer who cares about consistency should ask for reactive silica as a separate line.

What is silica doing in water at all?

Silica in drinking and process water is overwhelmingly dissolved silicon dioxide in the form of orthosilicic acid, Si(OH)₄. The Australian guideline describes this as the soluble form and notes that at higher concentrations it polymerises, “forming polysilicic acids, followed by the formation of colloidal silica” (NHMRC).

It comes from the rocks the water has passed through, which is why it varies by geography. Utilities typically report it as SiO₂, though the figure can be converted to elemental silicon using the ratio of the two atomic masses. The same NHMRC page gives the working numbers: crystalline silica dissolves only to about 6 mg SiO₂/L, while amorphous silica dissolves to 100–140 mg/L at 25 °C, and that second figure is the basis for the 80 mg/L aesthetic value.

Why can’t a conductivity meter see it?

Conductivity measures how readily a solution carries current, and that depends on free ions. Silicic acid is a very weak acid, so at neutral pH most of it sits in water as an uncharged molecule rather than as ions. Vendor technical literature states the consequence directly: silica “is weakly ionized in water, consequently it cannot be easily detected using conductivity” (Merck Millipore).

This is the same structural limit we described for total dissolved solids and conductivity conversions: a single electrical number is a proxy for ionic load, and some contaminants simply do not move it. Dissolved gases and non-ionic organics are others, and reverse osmosis has its own list of things it does not remove.

Why does silica leak first from ion-exchange resin?

Strong-base anion resins capture silica as the silicate anion, but silica is held less tightly than stronger anions such as chloride and sulphate. As a resin bed nears exhaustion, the more weakly held species are displaced first. Merck’s technical article puts it plainly: “silica is also known to be the first ion to breakthrough when the ion-exchange resin approaches depletion,” and reports that silica is released before the resistivity drop becomes visible.

For a plant operator that is a monitoring point. A demineralisation train alarmed only on conductivity may deliver silica-bearing water for some time before the alarm trips. Plants that serve silica-sensitive industries therefore monitor silica directly, and the same logic transfers to a blending-water line, even though spirits are less demanding than semiconductor fabrication.

Does reverse osmosis remove silica?

Largely, yes, but not perfectly, and not without cost. The vendor sources we reviewed describe thin-film composite membranes as rejecting the large majority of dissolved silica, and describe RO plus electrodeionisation as an effective pretreatment that leaves a low ionic load for the polishing resins. The same literature warns that silica at high concentration can itself foul the membrane by forming a silicate layer on its surface.

We did not find independent, peer-reviewed rejection data that we are willing to quote as a number, so we do not give one. Rejection depends on membrane, recovery, temperature and feed chemistry; the figure that matters is the one measured on the water actually being supplied.

Does silica change how whisky tastes?

We could not find evidence that it does. The NHMRC page, our most authoritative source here, contains no statement about taste, and its aesthetic value is explicitly a scale-control limit. We also found no sensory study of silica in diluted spirit. Anyone claiming a flavour effect at the low levels expected in treated water is, as far as we can verify, going beyond the evidence.

The reasons to care are narrower and more practical:

• Deposit and haze risk. NHMRC notes that dissolved silica “will react with the surfaces of glass and metals and begin to form a white precipitate.” Whether that is relevant at the single-digit mg/L levels of treated water is a question for the producer’s own stability testing, but it is a reason to know the number. The haze mechanisms that matter most in reduced spirit are driven by the spirit’s own congeners.
• Consistency. A water that is silica-free in March and silica-bearing in September because a resin bed was run past its capacity is a specification drift problem whether or not anyone can taste it.
• Diagnostic value. Because silica leaves first, it tells you early that the treatment train needs attention, before the parameters that are certain to matter are affected.

What should a buyer ask for?

1. Reactive silica (as SiO₂) on the certificate of analysis, with the method and detection limit stated, alongside conductivity at 25 °C. Our overview of how blending water is tested lists the parameters we would expect to see.
2. The regeneration or replacement trigger. Ask whether the supplier changes resin on conductivity alone or on silica as well.
3. Feedwater silica. A source with high natural silica loads the polishing stage harder. The seasonal variation of that source is worth asking about.
4. A number, not an adjective. “Low silica” is not a specification. Agree a limit and a test method, as you would for any other line in a demineralised water specification.

Glossary

• Silica (SiO₂) — silicon dioxide; in water, usually reported as the dissolved, reactive fraction.
• Orthosilicic acid — Si(OH)₄, the soluble monomeric form of dissolved silica.
• Colloidal silica — polymerised silica particles suspended rather than truly dissolved.
• Breakthrough — the point at which an ion-exchange bed stops retaining an ion and it begins to appear in the product water.
• Conductivity — the ability of water to carry an electric current, in µS/cm, used as a proxy for ionic content.
• Reactive silica — the fraction of silica that reacts in standard colorimetric tests, usually the figure reported on a certificate.

Frequently asked questions

Does low TDS mean low silica? Not necessarily. TDS is usually estimated from conductivity, and silica contributes very little to conductivity (Merck Millipore). A water can read low on both while still carrying measurable silica.

Is silica in water harmful? At the levels in treated water, we have no evidence of harm. Australia’s health-based guideline is 100 mg Si/L (about 210 mg SiO₂/L), far above levels expected in purified water (NHMRC).

Why do process engineers watch silica instead of conductivity? Because it leaks from ion-exchange resin before conductivity changes, so it gives earlier warning of exhaustion (Merck Millipore).

Is there a silica limit for water used to reduce whisky? We could not find one in the spirits-specific standards we reviewed. Buyers should agree a figure contractually with the supplier. For Indian regulatory context, see water quality standards for spirits and FSSAI.

Can silica cause scale in a distillery? Yes, in boilers and heat exchangers at high concentrations; the NHMRC aesthetic value of 80 mg SiO₂/L exists specifically to limit silica scaling (NHMRC).

Blenders International · Water Science series