A turbidity tube reading is roughly inverse to cloudiness, so a depth of 10 centimeters is about three times cloudier than 30 centimeters, and the carrier’s 30 centimeter “good” line is a screening threshold, not a guarantee that water is clear enough for chlorine or sunlight to work. A slow sand filter is unreliable above about 10 NTU and chlorine disinfection assumes under 1 NTU, so a reading that passes the tube can still fail treatment, and the pattern of readings across the system says more than any single value.
Part of Community Water Testing
Interpreting turbidity is the reasoning that turns a clarity reading into a decision: what the number implies about particles, germs and treatment, where in the system it points, and what to do about it.
A Clarity Reading Is Strong Evidence About Particles And Weak Evidence About Germs
Turbidity measures particles that scatter light. It is a good measure of the condition of the water for treatment and a poor measure of the water’s safety. Three things follow.
The scale. The black mark disappears when the light lost along the path down to it and back up equals what the eye can detect. For a single kind of particle that loss is roughly proportional to the depth multiplied by the amount of particles, so depth × turbidity is about constant. This is an idealisation: it does not hold exactly for mixed suspensions, but it is a useful guide. Halving the reading roughly doubles the turbidity, so a reading of 10 cm is about 30 ÷ 10 = 3 times cloudier than one of 30 cm.
The thresholds. Different processes need different clarity. The sand-filters page says a slow sand filter does not remove turbidity above 10 NTU reliably and needs pre-treatment (Sand Filters). The WHO guideline I am recalling for effective chlorination assumes turbidity below 1 NTU. The SODIS pages use a read-through-the-bottle test (SODIS Method). The carrier’s tube bands, over 30 cm good, 10-30 marginal and under 10 bad, are screening bands for distribution. Without a calibration against a measured standard, the tube cannot say which of those thresholds a reading corresponds to. Use it for change and for ranking, not as proof that chlorine will work.
The causes. A rise has a cause, and the place in the system where it appears points to it:
| Pattern | Likely cause | Response |
|---|---|---|
| Rises at the source after rain | Runoff entering; catchment or intake exposed | Divert surface water, repair the intake (Source Protection) |
| Source clear, tank outlet cloudy | Sludge stirred up in the tank, outlet too low | Clean the tank; set the outlet above the floor |
| Tank clear, one tap cloudy | Sediment or slime in that branch pipe | Flush the branch; check for a leak |
| Whole system rising slowly | Basin or filter failing | Desilt, scrape, check the rate |
| Brown water after a repair | Disturbed sediment | Flush until clear, then test |
What You Need
- Tube readings from several points, taken the same way each time
- A log with dates and weather so rain and turbidity can be matched
- A map of the system to place each reading
Method: Read The Pattern, Not Just The Number
Step 1: Take readings at the source, after each treatment step, at the tank outlet and at a near and a far tap.
Step 2: Compare each with the one before it. A step that does not improve the water is failing.
Step 3: Compare with its own history. A spring that reads 45 cm and then 25 cm has changed, even though 25 cm is still not dreadful.
Step 4: Link changes to causes from the log: rain, repairs, visitors, animals.
Step 5: Decide: adjust treatment, find a cause, or stop distribution.
Worked Example: How Much Does Each Step Remove?
On a storm day the tube reads 8 cm at the source, 14 cm after the basin and 40 cm after the filter. With turbidity taken as proportional to 1 ÷ depth:
- After the basin: 8 ÷ 14 = 0.57, so about 43% of the turbidity has gone.
- After the filter: 14 ÷ 40 = 0.35 of what entered, so 65% of the remainder has gone.
- Overall: 8 ÷ 40 = 0.20 of the raw turbidity is left, so about 80% is gone.
These are approximate, because the idealisation does not hold exactly, but they show which step does the work and which one is being asked too much. The filter’s outflow reads 40 cm, which passes the carrier’s good line, yet the raw water was so cloudy that the filter may be near the limit of what the sand filters page says it can handle, and the water should not be chlorinated on the assumption that it is clear.
What the treatment page’s advice to double the dose for turbid water means. It is a stopgap; the remedy for cloudy water is to settle and filter it first, since particles shield germs and consume chlorine (Chlorination).
What A Clear Result Does And Does Not License
A clear reading at one tap at one moment says the water was low in particles then. It does not show the water free of germs, dissolved chemicals or a sudden event, and a safe-looking result the day before an outbreak is normal.
Common Mistakes
- Treating “good” as ready for chlorination. Tube bands are not NTU.
- Looking only at the final water. The pattern between steps finds the fault.
- Ignoring slow drift. A slow rise is a failing step.
- Not matching readings with rain. The cause then stays unknown.
Safety
Cloudy water defeats disinfection, and clear water can still be unsafe
Chlorine and sunlight both fail in cloudy water, because particles shield germs and use up the disinfectant; clearing the water first is a requirement, not an optional step. Never judge safety by clarity: bacteria, viruses, parasites and dissolved chemicals leave water clear. After a flood, a repair or an illness in the community, treat the water as suspect whatever the tube says.
What’s Next
- Learn next: Biological Indicators — evidence about germs, which turbidity cannot give
- Combine with: Sedimentation — the step that fixes cloudiness