Drinking Water Standards and Testing: What to Check

Microbiological safety comes first because the health consequences are acute. The core test is E. coli, which must not be detectable in a 100 mL sample under the WHO reference and most national standards. Total coliforms are used as an indicator of treatment performance and distribution integrity; their presence does not prove faecal contamination, but it requires investigation.

Heterotrophic plate count is an operational indicator, not a health limit. A rising count at the tap points to biofilm, stagnation or a loss of residual, especially in a system fed by RO permeate with no chlorine. Turbidity is also a microbiological parameter in practice: above 1 NTU, particles shield organisms from UV and consume disinfectant, so most standards require clear water before the final barrier.

Chemical limits address chronic exposure. Nitrate at 50 mg/L as NO₃, arsenic and lead at 10 µg/L, and fluoride at 1.5 mg/L are among the values most often cited, and each has a treatment consequence. Nitrate is not removed by softening or carbon and requires RO or another demineralisation step. Nitrite is also commonly limited, with a short-term reference of 3 mg/L as NO₂. Arsenic and lead may require adsorption, coagulation or RO depending on the source. Fluoride may need reduction or, in some supplies, controlled addition. The local standard may set a lower value, so the design margin should be checked against it.

The table gives commonly cited WHO reference values for illustration, not a substitute for the local standard. The local limit may be lower, higher or absent, and the testing method may differ. Always design and report against the destination regulation, and state the reference used on the laboratory report.

Aesthetic parameters do not usually create acute health risk, but they decide whether people trust the water. Turbidity, colour, taste, odour, hardness and TDS are the common complaints. Iron above 0.3 mg/L or manganese above 0.05 mg/L stains fixtures even when the health limits are met. A supply can meet every health limit and still be rejected because it looks cloudy, stains fixtures or tastes flat. These parameters are also early warnings: a sudden change in colour or taste may indicate a treatment fault.

Typical targets are turbidity below 1 NTU, colour below 15 TCU, hardness 80–150 mg/L as CaCO₃ and TDS in the 150–500 mg/L range; chloride and sulphate are often kept below 250 mg/L. These are not universal limits. A very low TDS from RO can taste flat and corrode pipework, which is why remineralisation is part of the drinking-water train rather than an optional extra.

Sample at the source, at the treatment outlet and at the tap. Source samples tell you what the plant must remove; outlet samples prove the barrier is working; tap samples show what the consumer actually receives. For microbiological testing, use a sterile bottle, leave headspace, neutralise any disinfectant if the method requires it, keep the sample cool and deliver it to the laboratory within the holding time.

Field meters are useful for daily control of turbidity, pH, conductivity and chlorine, but they are not a substitute for laboratory analysis of microbiological and chemical parameters. Calibrate field meters against standard solutions, record the calibration, and send periodic split samples to an accredited laboratory. Keep microbiological samples at 2–8 °C and deliver them within 24 hours; chemical samples may have a different holding time. A result without the sampling point, date, method and detection limit is difficult to defend.

Frequency is set by the local standard and by the size and complexity of the supply. A small building plant may test microbiological parameters quarterly and full chemistry annually; a larger plant or a regulated supply may need monthly or weekly testing. Commissioning, a source change, a treatment change or a flood event should trigger additional sampling regardless of the calendar.

Between laboratory tests, run operational checks daily or weekly: turbidity, pH, conductivity, free chlorine or ORP, UV dose or ozone residual, and pressure differential across filters. These are not compliance tests, but they catch drift before it becomes a failed sample. Trend the results; a single reading inside the limit is less informative than a series moving toward it.

Compliance is shown by testing the finished water against the local standard, at a laboratory and frequency accepted by the local authority. Equipment certificates are not drinking-water approvals. The factory works under ISO 9001 and holds CE certification for exported equipment, but those cover quality management and product safety; they do not certify the water produced at your site.

Where a sanitary permit or material approval is required, obtain it for the exact configuration and destination market; a certificate for a similar model is not enough. Keep the validation report for any UV reactor or ozone contactor, the membrane specification, and the material certificates for wetted parts. Then repeat the finished-water test after commissioning and at the frequency the local rules require.

Keep the raw-water analysis, the design basis, the commissioning report, the as-built drawing and the operating log. The log should record daily flow, pressures, turbidity, conductivity, disinfectant residual or UV dose, every backwash and media change, and every calibration with the date and the person responsible. When a result is questioned, the log is often more persuasive than the equipment datasheet.

A treatment train is designed for a water quality, not for a source name. Rainfall, drought, agricultural runoff, a new borehole or a switch from municipal supply to tanker water can change turbidity, nitrate, hardness, iron and bacterial load within days. A plant that passed commissioning in the dry season may fail in the wet season without any equipment fault.

Sample the source in both wet and dry seasons before design, and repeat the analysis after any change. Adjust coagulant dose, backwash frequency, carbon replacement interval and disinfection dose to the new water. If nitrate or arsenic rises above the treatment capability, the train may need RO or a different adsorbent, not a higher chlorine dose.