Home › Guides › Measuring TDS in Water

Measuring TDS in Water: What Your Meter Actually Measures

Start with the fact that reframes everything else on this page: no handheld TDS meter measures total dissolved solids. Every one of them — the $8 pen and the $1,089 Ultrameter alike — measures electrical conductivity and multiplies by a conversion factor to produce the ppm number on the display. That's a legitimate and widely accepted method. But the conversion factor isn't a constant, most inexpensive meters don't disclose which one they use, and the difference between the common factors is large enough to change decisions. This guide covers what the measurement actually is, how much error the conversion introduces, how to calibrate and run a field protocol that produces defensible numbers, and how to interpret the result.

What TDS is, and the reference method

Total dissolved solids is the combined concentration of everything dissolved in a water sample — minerals, metals, salts, and dissolved organic compounds. The operational definition used across the extension literature is material smaller than 2 microns that cannot be removed by conventional filtration. Natural contributors include calcium, magnesium, sodium, potassium, carbonates, bicarbonates, chlorides, sulfates, nitrates, and silica, leached from soil and rock. Anthropogenic contributors include road salt, agricultural runoff, treatment chemicals, pipe corrosion products, and process water carryover.

The reference method is gravimetric and conceptually simple: take a measured sample, evaporate the water to dryness, and weigh what's left. Standard Methods specifies drying at 180 °C, and results are reported in mg/L. It's accurate, it's the number every other method is trying to approximate, and it's completely impractical in the field — it needs a laboratory balance, a drying oven, and hours per sample.

That impracticality is the entire reason conductivity meters exist. Everything that follows is about how good the substitute is.

What the meter actually measures

Dissolved minerals dissociate into ions, and ions carry charge. That makes conductivity a workable proxy for dissolved ion concentration: the more dissolved mineral content, the more readily water conducts. Distilled water conducts poorly precisely because there's almost nothing dissolved in it to carry current.

An EC meter applies a voltage across two electrodes at a known spacing and geometry — the cell constant — and measures the resulting current. That gives conductivity, in microsiemens or millisiemens per centimeter. A "TDS meter" is the same instrument with one extra step in firmware: it multiplies the conductivity reading by a conversion factor and displays the product as ppm.

This is not a criticism of the method. Conductivity-derived TDS is genuinely useful, it's what extension programs recommend for field screening, and for trend monitoring it's entirely sufficient. The problem is specifically the undisclosed factor, which is where a meaningful and avoidable error enters.

The EC-to-TDS conversion table

The conversion factor is not a constant. It increases with conductivity, because the relationship between ion concentration and conductivity in natural water is non-linear — as ionic strength rises, ion–ion interactions reduce the conductivity contributed per unit of dissolved mass. The table below is from NDSU Extension publication WQ1923, corroborated by SDSU Extension, and is the standard reference for natural waters.

The factor spans 0.50 to 0.82 across the range — a 64% spread from bottom to top. Any instrument that applies one number to that whole curve is making an approximation whose size depends on where in the range you're working.

How wrong a fixed factor gets

Inexpensive pen meters typically hard-code a single conversion factor. The two conventions in common use are the NaCl scale at roughly 0.50 and the 442 scale at roughly 0.70, named for a reference mixture of 40% sodium sulfate, 40% sodium bicarbonate, and 20% sodium chloride. Some meters use 0.65. Many don't document the choice anywhere.

It gets worse when you cross industries, because different fields standardized on different numbers. Agricultural extension guidance for irrigation water commonly uses EC × 640 — a factor of 0.64 — for water below 5 dS/m. Hydroponic nutrient charts are usually built on the 500 or 700 scale. Well-water guidance uses the sliding factor in the table above. The consequence is that a single water sample can be correctly reported as four different ppm values by four sources that are each following their own accepted standard:

Taking the NDSU table as the reference and applying each fixed factor across the range gives the error directly:

The practical shape of this: a 0.50-scale meter is accurate on very low-TDS water and reads progressively low as TDS rises, understating by nearly 40% on brackish water. A 0.70-scale meter is accurate in the 1,000–3,500 µS/cm band — which covers a lot of ordinary well and municipal water — and overstates badly at the clean end, which is exactly where RO permeate sits.

What conductivity can't see

There's a second and more fundamental gap between conductivity-derived TDS and gravimetric TDS, separate from the conversion factor: conductivity only detects things that carry charge.

Dissolved solids that are non-ionic contribute mass to a gravimetric TDS result but contribute little or nothing to conductivity. Sugars, alcohols, many dissolved organic compounds, and undissociated silica all fall into this category. Water carrying a meaningful non-ionic dissolved load can return a conductivity reading that implies far less dissolved material than the evaporation method would find.

For most natural waters this hardly matters — the dissolved load is overwhelmingly ionic and the proxy holds well. It matters in specific cases:

• Food and beverage process water carrying sugars or organics — conductivity substantially understates total dissolved load.
• Silica-bearing source water, which is a serious RO membrane scaling risk that a conductivity reading will not warn you about. Silica requires its own analysis.
• Wastewater-influenced sources with elevated dissolved organic carbon.
• Ultrapure water systems, where the resistivity measurement is genuinely sensitive to ionic content but blind to non-ionic TOC — which is why semiconductor and pharmaceutical water specifications control TOC separately rather than inferring it. See the deionized water guide for the ASTM purity type framework.

EC units, decoded

Conductivity units are a persistent source of confusion, largely because two naming systems coexist — the older mho (ohm reversed) and the SI siemens — and they're numerically identical.

"Micro" is one-millionth; "milli" is one-thousandth. Older extension and agricultural literature tends to use µmhos/cm; modern instruments display µS/cm or mS/cm. When you're comparing a reading against a published table, confirm which unit the table uses — the difference between mS/cm and µS/cm is a factor of 1,000, and it's the single most common error in reading these tables.

TDS units are simpler: mg/L and ppm are equivalent for dilute aqueous solutions, and ppt means parts per thousand, so 1 ppt = 1,000 ppm.

Meter selection — the specs that matter

SDSU Extension's guidance on meter purchasing is blunt and correct: don't buy the cheapest one. But "buy a better one" needs specifics, because the marketing language on inexpensive meters is designed to sound like specification without being any.

Established instrument manufacturers in this space include Myron L, Hanna Instruments, YSI, Apera, Milwaukee, and Extech. That doesn't make every product from them right for every job, but it does mean published specifications, documented cell constants, available calibration standards, and service support.

Calibration, including the field method

An uncalibrated meter produces numbers that look every bit as authoritative as calibrated ones. Conductivity cells age, electrode surfaces foul, and cell geometry shifts with physical shock — all of which move the reading without any indication on the display. NDSU Extension is explicit that meter accuracy should be verified against a calibration solution before sampling.

When to calibrate

• Before each sampling session, as a verification step even if no adjustment proves necessary
• After a battery change — power interruption can disturb the stored reference
• After dropping the meter on a hard surface, which can shift electrode geometry
• After prolonged storage, where electrode drift is common
• Any time a reading is unexpectedly high or low for a source with known history

Commercial calibration standards

The reliable approach is a purchased standard of certified conductivity. Common values are 84, 1,413, and 12,880 µS/cm — choose the one nearest your working range, since calibrating at 84 µS/cm and then measuring brackish water verifies almost nothing about performance at the top of the range. For drinking-water and food/beverage work, 1,413 µS/cm is the usual choice because it sits near the EPA secondary standard range.

The NDSU field method using table salt

When a commercial standard isn't available, NDSU WQ1923 documents a validated field procedure producing a solution near 2,600 mg/L TDS. That target wasn't arbitrary — it's approximately the level at which livestock begin showing adverse effects, and NDSU notes the conversion factor changes little above 2,700 mg/L, so a meter verified there reads reliably across the whole high range.

The same discipline applies to any field instrument — the pH meter calibration guide covers the equivalent protocol for electrometric pH, including slope acceptance criteria and compliance documentation.

Field testing protocol

This protocol combines NDSU WQ1923 and SDSU Extension guidance with standard industrial practice. It applies to wells, ponds, streams, process water, cooling tower basins, and boiler feedwater alike.

1. Calibrate first, and log it. Verify against a standard before collecting any data, and record the result. A calibration record is what makes the day's readings defensible later.
2. Fix and document sampling locations. For any monitoring program, establish permanent sampling points with GPS coordinates or unambiguous physical descriptions. Trend data from inconsistent locations isn't trend data.
3. Rinse the container three times with the water being sampled before taking the final sample. Residue from a previous sample is a real and common source of error.
4. Collect representatively. Surface water: scoop through the water column rather than skimming the surface. Wells: sample from active flow, not water standing in the pipe. Process water: mid-stream, never from a dead leg.
5. Let the reading stabilize — typically 15–30 seconds while the probe equilibrates to sample temperature. Reading before stabilization is reading the probe's thermal transient, not the water.
6. Immerse the probe fully, off all surfaces. Every sensing electrode submerged, and the probe not touching the sides or bottom of the container. Container walls distort the electrical field and produce false readings.
7. Swirl gently while reading to displace the stagnant boundary layer at the electrode surface so the reading reflects bulk water composition.
8. Record with full context — date, time, location, water temperature, EC in µS/cm, TDS in mg/L with the conversion factor used, and any observation about appearance or odor. A ppm value recorded without its factor is not reproducible.
9. Rinse the probe with distilled water between samples to prevent carryover, particularly when moving from high-TDS to low-TDS samples.
10. Apply the action threshold. NDSU's field trigger for laboratory submission is EC ≥ 6,000 µmhos/cm or TDS ≥ 4,500 ppm. Above that, field screening has done its job and speciated laboratory analysis is warranted.

Interpreting the number

There is no universally good TDS value — only values appropriate to a use. The EPA's 500 mg/L figure is a secondary standard addressing taste, odor, and appearance, not a health-based enforceable limit, and it gets cited far outside the context it was written for.

Two application-specific notes worth pulling out. In hydroponics, TDS isn't a contaminant measure at all — it's the nutrient dosing parameter, deliberately managed from around 400 ppm at seedling stage to 3,500+ ppm for heavy feeders. The 500 mg/L drinking water figure is irrelevant there, and the conversion-scale problem is acute because published crop targets rarely state which scale they assume. That's covered in detail in the hydroponics EC, TDS & pH guide. In RO systems, TDS is the performance metric: measure feed and permeate, and rejection percentage tells you membrane condition directly. Rising permeate TDS is the earliest available warning of membrane damage or O-ring bypass.

Livestock and agricultural thresholds

Agricultural operations are among the heaviest users of field TDS testing, and the thresholds are well established across the extension system. Elevated TDS reduces water intake, which reduces feed intake, which shows up as production loss before it shows up as an obvious health problem.

Sources of high TDS and treatment

TDS doesn't rise randomly. Each source has a characteristic signature, which is why the correct response to a high field reading is speciated laboratory analysis rather than immediate equipment purchase — the treatment depends on what is dissolved, not how much.

FAQ

What does a TDS meter actually measure?

Electrical conductivity, converted to a ppm estimate in firmware. The reference method for true TDS is gravimetric — evaporate a measured sample and weigh the residue, dried at 180 °C. Conductivity is a well-accepted proxy for that, not a direct measurement of it.

What is the EC to TDS conversion factor?

It varies with conductivity, from 0.50 at 100 µS/cm up to 0.82 at 10,000 µS/cm on the NDSU reference table. Meters that hard-code a single factor — commonly 0.50 for the NaCl scale or 0.70 for the 442 scale — are exact at one point and diverge from there, by as much as 39% at the extremes.

Why do two TDS meters give different readings on the same water?

Most often because they use different conversion scales, which alone accounts for up to a 40% spread between correctly functioning instruments. Other contributors are calibration state, temperature compensation differences, and cell constant mismatch with the sample range. Compare EC readings rather than ppm readings and most of the disagreement disappears.

What is a good TDS level in water?

Application-dependent. EPA's secondary standard is 500 mg/L for drinking water, an aesthetic rather than health-based limit, with under 300 mg/L generally preferred for taste. Livestock: under 3,000 ppm broadly safe. Hydroponics: 1,200–3,500 ppm by design. Ultrapure process water: under 10 ppm. A "good" number outside a stated application isn't meaningful.

Does a TDS meter measure water hardness?

No. Hardness is calcium and magnesium; TDS is all dissolved solids. High TDS with low hardness is common in sodium-dominated water. A properly working softener will barely change a TDS reading, because it exchanges hardness ions for sodium rather than removing dissolved mass.

Does "lab grade accuracy" mean anything on a TDS meter?

No. It's unregulated marketing language with no defining standard. The specifications that determine whether a reading is defensible are accuracy as a percentage of reading, cell constant, temperature compensation coefficient, resolution, and traceable calibration certification. If those aren't published, the adjective isn't a substitute.

How do I reduce TDS in water?

Dissolved solids can't be removed by conventional filtration, which only captures suspended particles. The options are reverse osmosis (most cost-effective in most commercial cases), distillation, electrodialysis, deionization, or blending with a lower-TDS supply. Which one is appropriate depends on what the dissolved load actually consists of, which is a laboratory question rather than a meter question.

When should I send a sample to a laboratory?

NDSU's field trigger is EC ≥ 6,000 µmhos/cm or TDS ≥ 4,500 ppm. Beyond that threshold, also send for compliance reporting, when establishing a baseline on a new source, when TDS has shifted more than about 20% from an established baseline, or whenever the number will be used by anyone outside your organization.

Sources: Prasad, Laxmi, and Miranda Meehan. "Using Electrical Conductivity and Total Dissolved Solids Meters to Field Test Water Quality." NDSU Extension Publication WQ1923, North Dakota State University, revised January 2025. · Salverson, Robin. "Performing a Field Test for Livestock Water Quality." SDSU Extension, South Dakota State University, updated November 2023. · MSU Extension Water Quality Program / Montana Well Educated Program, "Alkalinity, pH, and Total Dissolved Solids," Montana State University. · US EPA National Secondary Drinking Water Regulations. · Standard Methods for the Examination of Water and Wastewater, Method 2540 C (Total Dissolved Solids Dried at 180 °C). Fixed-conversion-factor error figures in this guide were calculated from the NDSU WQ1923 reference table. This guide is informational and does not constitute engineering, regulatory, or veterinary advice.