Leaking toilet water loss
A leaking toilet is the most common continuous water loss in a UK home, and most of them are silent. The measured residential average is 400 litres a day, which is 16.7 litres per hour, and around 8 per cent of toilets are leaking at any given time. This calculator converts what yours is visibly doing into a band, because a visible classification is not a measurement and reporting it as one would be false precision.
Last reviewed 14 August 2026
What this calculator tells you
It converts what a toilet is visibly doing into a range of litres per day, and then into the volume and cost that range produces over a year.
- A daily volume, as a band in litres per day, not a point value
- An annual volume in cubic metres per year, as a band
- An annual cost, at your own supplier rate, as a band
- A severity classification and what to check next
Every output here is a range, and that is deliberate rather than a limitation. The input is a category you chose by looking, so a single number out of it would be precision the input never contained.
How to take the measurement
A leaking toilet is usually silent and leaves no mark, so the useful test is a dye test rather than looking or listening. It takes half an hour and needs nothing you would have to buy specially.
- Put dye in the cistern. Lift the cistern lid and add enough food colouring to the water to colour it clearly. A dye tablet does the same job and is less likely to be spilled.
- Do not flush. Leave the toilet completely alone for at least 30 minutes. One flush restarts the test, so if the toilet is in use, do this somewhere it will not be.
- Look in the pan. Come back and look at the water in the pan without flushing. Colour in the pan means water has passed from the cistern into it, which is a leaking flush valve.
- Check the overflow as well. While the dye is in, look at the external overflow pipe if the toilet has one. Coloured water there means the fill valve is passing and the cistern is overfilling, which is a different fault with a different repair.
Colour in the pan means the flush valve is passing. Colour in the external overflow means the fill valve is passing and the cistern is overfilling. Those are two different faults with two different repairs, and the dye test tells them apart in the same half hour.
If the dye test is negative and you still suspect a loss, the mechanism itself may be the answer. which mechanism your cistern uses explains why some UK cisterns cannot leak silently at all.
Using the calculator
Rate of loss
What your result means
The band you selected maps to a range of daily volumes, and the four bands describe what a toilet doing that is usually losing.
| What you can see | Daily volume | What is usually happening |
|---|---|---|
| Weeping into the pan, visible only as a wet streak or with dye | 30 to 200 litres a day | A flush valve seal held slightly open by debris or limescale. |
| A visible trickle running down the back of the pan | 200 to 400 litres a day | The same fault, further advanced. The measured residential average sits at the top of this band. |
| A constant flow, audible if you listen | 400 to 750 litres a day | A flush valve failing to seat at all, or a fill valve running continuously. |
| Water discharging from an external overflow pipe | 750 litres a day and above | The fill valve is passing and the cistern is overfilling, so water runs to waste continuously. |
Two things about that table are worth knowing before you rely on it. The boundaries are a practical classification and no published source was found for any of them, which is why they are adjustable inputs on this page rather than fixed constants. And the one measured figure in the whole set, the residential average of 400 litres a day, falls exactly on the boundary between the third and fourth rows. So the scale is not arbitrary. It is also not measured, and this site will not describe it as validated.
One way to picture the measured average before comparing it to anything: across the UK a single flush is capped at 6 litres, so a toilet losing 400 litres a day is losing more than 66 full flushes a day, every day, silently. That converts a volume most people cannot picture into one they can, and the whole band table does the same.
| Band | Litres a day | Full flushes a day |
|---|---|---|
| Weeping, at the bottom | 30 | 5 |
| Trickle, at the bottom | 200 | 33 |
| The measured average | 400 | 67 |
| Overflow, at the bottom | 750 | 125 |
The two halves of that comparison are not equally strong, and saying so is the point. The 6 litre cap is read directly from the instruments and is exact. The 400 litres a day average is measured but reached through an intermediary rather than from the original research, and the band boundaries have no source at all. A ratio is only as good as its weaker half.
Now put those bands against what a water meter can actually measure, because that comparison explains something people report constantly and rarely have a mechanism for.
A domestic meter is specified for accuracy only above its minimum flow rate. On a typical DN15 meter that is 15.6 litres per hour, which is 374.4 litres a day. Below it, the standard defines no permitted error at all. Dividing each band by 24 puts the two scales side by side.
| Band | As a rate | Against the meter |
|---|---|---|
| Weeping | 1.3 to 8.3 L/hr | Entirely below the specified range. The standard says nothing about what the meter does here. |
| Trickle | 8.3 to 16.7 L/hr | 87 per cent of the band is below the specified range. |
| Constant | 16.7 to 31.3 L/hr | Entirely above it. |
| Overflow | 31.3 L/hr and above | Entirely above it. |
So it is not simply that the average sits near the boundary. Two of the four bands fall entirely or almost entirely inside the range where a domestic meter has no specified accuracy whatsoever. A weeping toilet at 100 litres a day is running at about a quarter of that minimum flow, and nothing in the standard describes what the meter registers there.
That is the mechanism behind a leaking toilet running for years without the bill moving noticeably. It is not that nobody looked. It is that for two of the four bands the instrument is being asked to measure below the range it is specified for.
One caution about that comparison, because the two halves of it are not equally strong. The meter figures are read directly from the international standard and from a manufacturer datasheet. The toilet bands have no published source at all. So the arithmetic is exact and checkable, and one side of it is a convention this site chose. It shows where our own categories sit against a measured instrument. It does not turn those categories into measurements.
How the calculation works
The band gives a low and a high figure in litres per day, and both are carried through the same arithmetic, so every output is a range with the same shape as the input.
- Annual volume in cubic metres, the daily volume × 365 then ÷ 1,000
- Annual cost, the annual volume × your volumetric rate
- Multiplied by the number of toilets, if more than one is doing it
Nothing is averaged into a single figure at any point. Two toilets in the trickle band produce a band twice as wide, not a point value twice as large, because combining two uncertain quantities does not make them certain.
The sewerage rate applies here in a way it does not for every leak on this site. A toilet leak drains to the sewer by definition, so unlike a loss on an underground supply pipe, your supplier rate per cubic metre is needed for both halves of the bill rather than only the water half.
Assumptions and sources
Aquadis+ DN15 and DN20 specification sheet
Itron · manufacturer source
- Q1 minimum flow rate, DN15, Q3 2.5, R160: 15.6 l/h
- Q2 transitional flow rate, DN15, Q3 2.5, R160: 25 l/h
- Q3 permanent flow rate, DN15: 2500 l/h
- Typical starting flow rate, DN15: 0.4 l/h
- Typical starting flow rate, DN20: 2 l/h
- Q1 minimum flow rate, DN20, Q3 4.0, R160: 25 l/h
- Flow rate at which accuracy is within +/- 5%, DN15: 3 l/h
- Flow rate at which accuracy is within +/- 2%, DN15: 5 l/h
- Minimum scale interval, DN15 and DN20: 0.02 l
Leaky loos, why it's not as simple as faulty flush valves
ech2o, reporting research attributed to Thames Water · secondary source
- Toilets inspected: 58551 toilets
- Toilets found leaking: 4854 toilets
- Prevalence: 8.3 %
- Average residential leak rate: 400 l/day
- Average commercial leak rate: 2100 l/day
- UK-wide average across all WCs, leaking and sound: 20 l/day
- Share of leaking toilets with flush valves rather than siphons: 81 %
Litres in a cubic metre
Exact by definition
SI definition. The litre is defined as one cubic decimetre, so one cubic metre is exactly 1,000 litres.
Hours in a day
Exact by definition
Definition of the day as 24 hours.
Days used for an annual figure
Editorial convention
Annual figures on this site are a daily rate multiplied by 365. Leap years are not modelled, so a leap year is understated by one day, which is 0.27 per cent.
Toilet leak daily volume bands
Illustrative assumption
Adjustable assumption. The four visible categories are a practical classification and no source was found for any of their boundaries. What is measured is the average: a leaking toilet loses about 400 litres a day. Two things can be said for the scale without overstating it. The measured average falls exactly on the boundary between trickle and constant flow, and the separately published 215 to 400 litres a day range sits inside the trickle band. So the scale is not arbitrary. It is just not measured.
The evidence behind this page splits in two, and it is worth being exact about which half is which.
Measured, but second hand. The prevalence figure and the 400 litres a day average come from research attributed to Thames Water and reported by ech2o in 2021. The underlying Thames Water publication was searched for and not located, so this site cites the intermediary that was actually read. Nothing here should be taken as us having read the original, and if it turns up it replaces this.
Not measured. The four band boundaries have no published source at all. They are a practical classification, they render as editable inputs so you can replace them, and the two things that can honestly be said for them are that the measured average lands on one boundary and the most widely quoted published range sits inside another band. That makes the scale not arbitrary. It does not make it measured.
The dye test itself needs no citation. It is a procedure rather than a figure, and it either shows colour in the pan or it does not.
What to do next
Four things, in this order, and the first two cost nothing.
- Dye-test every toilet in the property, not just the suspected one. Prevalence is high enough that a second leaking toilet is a realistic possibility rather than bad luck.
- Look at the external overflow pipe from outside. A wet patch below it, or staining on the wall, means the fill valve has been passing for a while even if nothing is dripping now.
- Lift the cistern lid and see which mechanism is fitted. A siphon and a drop valve fail in different ways and only one of them can leak silently.
- Confirm the rate at the meter once you know which toilet it is. A dye test proves a toilet is passing; only the meter puts a number on how much.
That last step is worth doing before any repair, because it turns a category into a measurement. confirm the rate of loss at the meter isolates the property and reads the difference over a measured interval.
Common questions
- How much water does a leaking toilet actually waste?
- The measured residential average is 400 litres a day, from research attributed to Thames Water and reported by a third party. That is an average across leaking toilets rather than a figure for yours: a weeping valve loses far less and an overflowing cistern loses far more.
- Would I not hear it?
- Usually not. The most common fault is a flush valve passing slowly into the pan, which makes no noise at all and leaves no mark. That is why the dye test exists: it makes a silent loss visible in half an hour without any tools.
- How common is it?
- Around 8 per cent of toilets are leaking at any given time. The larger sample behind that inspected 58,551 toilets and found 4,854 leaking, which is 8.3 per cent. Of the ones found leaking, 81 per cent had a flush valve mechanism rather than a siphon.
- Why does this calculator give a range instead of a number?
- Because the input is a classification of what you can see, not a measurement. You are telling it whether the water is weeping, trickling, running or overflowing, and no amount of arithmetic turns a visual category into a precise litres per day figure. A range is the honest output of a range of an input.
- Where do the four bands come from?
- They are a practical classification rather than a measured set of boundaries, and no published source was found for any of them. They ship as adjustable inputs so you can change them, and the only measured figure behind them is the 400 litres a day average, which falls on one of the boundaries.
- Will the meter show it?
- An average one, just about. 400 litres a day is 16.7 litres per hour, and the lowest flow a typical UK domestic meter is specified to measure accurately is 15.6 litres per hour. So the average leaking toilet in Britain runs at roughly the slowest rate a domestic meter is built to measure properly, and a less severe one sits below that.