Why an 8 Ounce Bowl Keeps Water Fresher Than a 2 Litre Fountain

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Because the tap is the only water source in the house with a residence time measured in seconds. A pump makes water move. It does not make water new, and those are different properties with different consequences.

Here is the whole argument in one line. For any bowl held at a constant level, where what goes in equals what the animal drinks, the mean age of the water in it is the volume divided by the drinking rate. The pump does not appear in that expression. So a deliberately small bowl holds fresher water than a large circulating one, every time, and the gap is not small.

Volume over drinking rate, and nothing else

A CATnK whisker-friendly elevated cat water bowl publishes a wide, shallow, open shape with lowered sides so whiskers stay clear of the rim, an elevated base to ease neck and back strain, a curved surface said to help the cat see the water, food-safe melamine that is shatterproof and dishwasher safe, and a capacity of 8 ounces, which is one cup.

And it publishes the reasoning, which is the part worth quoting: "It holds a stay-fresh 8 oz, one cup, deliberately modest to encourage daily refills so the water stays fresh rather than sitting stagnant."

That sentence is correct and this is the number behind it. Eight ounces is 236.6 millilitres. A cat drinks a representative 200 millilitres a day. So the mean age of the water in that bowl is 236.6 divided by 200, which is 1.183 days.

A set of 12-set triple filtration cat water fountain filter replacements publishes the other end of the scale, and usefully publishes the fountain it belongs to: square pet fountains of 2.2 litres, or 73 ounces. Each set pairs a cartridge of cotton, activated carbon and ion exchange resin with a pre-filter sponge for hair and a high density 8D sponge to protect the pump, and the maker advises replacing every two to three weeks, soaking each new filter for five minutes and rinsing for 30 seconds to clear carbon dust.

Two point two litres against the same 200 millilitres a day is a mean age of 11.00 days. Nine and a third times the small bowl.

And the distribution matters more than the mean. A pumped fountain is a well mixed vessel, so the age of the water in it is exponentially distributed, which means the fraction older than any given age is the exponential of minus that age divided by the mean. Pick three days as a threshold:

In the 8 ounce bowl, the fraction older than three days is the exponential of minus 3 over 1.183, which is 7.9 percent. In the 2.2 litre fountain it is the exponential of minus 3 over 11, which is 76.1 percent.

So 92 percent of the small bowl is under three days old and 76 percent of the fountain is over it. A ratio of 9.63 on the stale fraction, decided entirely by reservoir size and not at all by filtration or circulation. Full record in our review of the CATnK elevated bowl.

Where our own guide gets it half right

Our page comparing automatic pet water dispensers across gravity, plumbed and fountain types describes a fountain like this: "An electric pump circulates and filters water continuously, so it stays fresh, oxygenated and moving."

Two of those three words are right. Moving is obviously right. Oxygenated is right and measurable, because a falling stream entrains air and a still bowl does not. Fresh is the one that does not follow, and the arithmetic above says why: circulation changes nothing about age, and age is what fresh means. The same page is exactly right about the gravity type, that "the standing water is neither filtered nor moving, so it needs regular cleaning", and the residence time is the quantity that tells you how regular.

It is also right, in the sentence that matters most for a cat, that "many cats and some dogs drink far more from flowing water". That is a real effect and it is about perception rather than purity, which is the honest division of the subject: a fountain wins on getting the animal to drink and loses on how old the water is, and a listing that claimed both would be claiming one thing twice.

Evaporation takes the water and leaves everything in it

Now the second mechanism, which has a surprisingly clean form. Evaporation removes pure water. The top-up that replaces it brings dissolved solids with it. Nothing removes solids except the animal drinking them.

Write the balance at steady state. Solids coming in equals solids going out, so the tap concentration times the sum of drinking and evaporation equals the bowl concentration times the drinking alone. Which rearranges to something you can hold in your head:

The bowl settles at the tap concentration multiplied by one plus the ratio of evaporation to drinking.

Put numbers on it. In a 22 degree room at 50 percent relative humidity, a water surface sitting near 18 degrees evaporates at about 74.5 grams per square metre per hour, by the same psychrometric route our page on why merino and polyester dry at the same rate works out for a wet shirt.

A 2.2 litre square fountain is roughly 18 centimetres on a side, which is 0.0324 square metres of still surface, plus a wetted ramp and falling stream of perhaps another 0.012, so about 0.0444 square metres. That loses 3.31 grams an hour, which is 79.4 millilitres a day. Against 200 millilitres of drinking, the ratio is 0.397, so the fountain settles 39.7 percent above the tap.

The 8 ounce bowl is wide and shallow by design, so call it 0.013 square metres. It loses 23.3 millilitres a day and settles 11.7 percent above.

A factor of 3.39 between them, and here is the uncomfortable part: the cause is the wetted area a fountain must have in order to be a fountain. The falling stream that makes the water appealing is also the surface that concentrates it. You cannot have one without the other, and no listing in this category publishes a wetted area, so the figure cannot be computed for any specific product.

The time constant on that approach is the same volume over drinking rate as before, 11 days, which means a fountain is most of the way to its steady state inside a single filter interval. And a running tap sits at exactly the mains concentration, forever, because nothing has had time to evaporate out of it. Three variables separate a tap from a fountain, and this is one of them.

Why a recirculating filter is rated in weeks rather than litres

The two to three week interval on those cartridges is the right kind of rating, and the reason is worth working out because it looks wrong at first.

Over seventeen and a half days, a 2.2 litre fountain serving one cat takes in 3.5 litres that get drunk and 1.39 litres that evaporate, so 4.89 litres of genuinely new water, which is 2.22 reservoir volumes. That is all the contaminant the carbon ever has to deal with.

Meanwhile the pump, at a representative 80 litres an hour, has pushed 33,600 litres through that cartridge in the same period. Which is 6,871 times the new water, and essentially all of it is water the carbon has already cleaned, going round again.

So a volume rating would be meaningless here: the cartridge's throughput has almost nothing to do with its load. An interval is the correct unit, because what the carbon has to adsorb is proportional to time rather than to litres pumped. Our page on why aquarium carbon lasts about a month works the adsorption capacity side of this properly, and the same logic applies with a much smaller loop.

It also explains something in the kit that otherwise looks arbitrary. The set contains a carbon cartridge and two sponges, and the sponges are there for hair and debris. Hair is generated continuously, inside the loop, by the animal. Dissolved load only enters with new water. So the two components have genuinely different clocks, and our guide to keeping a pet water fountain clean and filtered gets this right in practice when it says to keep both sponges clean rather than replacing them on the cartridge's schedule.

What the listing does not publish is the resin volume, and the ion exchange resin is the one component that could genuinely address concentration. Work backwards from the interval: 4.89 litres of new water at a representative 150 milligrams per litre of hardness is about 730 milligrams to capture, which needs around 7 millilitres of resin at two equivalents per litre. That is a small and entirely plausible amount. It is simply not stated, and neither is the carbon mass, so the interval has to be taken on trust rather than checked.

Eight litres, or 1.1 gallons, and they are not the same reservoir

A 1.1 gallon gravity pet water dispenser with a steel bowl publishes a removable stainless bowl, an adjustable base lifting it from 3.15 to 6.3 inches, a replaceable natural stone filter, a reusable ice pack, and the opening line of its own description: "This 8-liter gravity water dispenser holds about 1.1 gallons".

Eight litres is 2.1134 US gallons, or 1.7598 imperial. One point one US gallons is 4.1640 litres, and 1.1 imperial gallons is 5.0005. The two published figures are out by a factor of 1.9213, and unlike some pairs of this kind, no gallon convention reconciles them. The product's own URL carries the 1.1 gallon figure, so whichever half is wrong, the wrong one is the one in the address.

Its own duration estimates arbitrate, and they come down on the litres. The listing estimates "5 to 7 days for large pets", which at 8 litres implies 1,143 to 1,600 millilitres a day and at 1.1 gallons implies 595 to 833. A large dog drinks somewhere between one and a half and three litres a day, so only the 8 litre reading produces a sensible number. The small and medium figure, 8 to 12 days, implies 667 to 1,000 millilitres at 8 litres and 347 to 521 at 1.1 gallons, and both of those are plausible, so it cannot settle anything.

Take the 8 litres, then, and apply the residence time. For a cat at 200 millilitres a day, that reservoir takes 40 days to empty. And a sealed gravity column does not mix: water goes in all at once and comes out from the bottom, so the water drunk on day twenty is twenty days old and the last of it is forty. The mean age over an emptying cycle is 20 days, which is nearly twice the fountain's and seventeen times the 8 ounce bowl's.

That is not a defect of the design, it is the definition of the design, and the listing is honest about the things it can be honest about: the ice pack "is not a fridge" and the stone filter "needs replacing over time". The figure that would complete the picture is the one above, and a natural stone filter removes essentially nothing dissolved, so it does not touch the concentration term either.

The feature that makes it hands-off is the feature that makes it old

A Ficuswin 67oz automatic dog water dispenser publishes a wall mount, a 67 ounce capacity, a patented mechanical valve needing no electricity that refills the bowl when the level drops below about 50 ounces, a copper connector and hose for a three quarter inch garden thread, and solid metal mounting.

Those two numbers are the interesting pair. Sixty-seven ounces is 1.9811 litres and fifty is 1.4787, so the valve's operating band is 17 ounces, which is 0.5027 litres, or 25.4 percent of the capacity. The other 74.6 percent sits below the trigger level and never cycles until the animals drink it down there.

Which is the inversion worth noticing. A narrow top-up band is what makes a float valve convenient: the bowl is always nearly full and nobody has to think about it. It is also what maximises the mean age, because topping up a vessel is not the same operation as replacing its contents. Only replacement resets the clock. A valve with a wider band would be less tidy and the water would be younger.

And this one is outdoors by design, aimed at kennels and poultry runs, where the water is warmer and the evaporation term is larger than any of the indoor figures above. The right use is the one the record describes, a high consumption setting where a 1.98 litre bowl turns over in a day or two because six birds or a working dog are drinking from it. In that case the residence time looks after itself, which is the honest version of the recommendation.

At the other extreme sit a pair of PetBonus collapsible silicone dog bowls, which publish 34 ounces each, food grade BPA-free silicone, a fold-flat height of 0.8 inches, an expanded size of 7 by 5 by 2.65 inches, dishwasher safety on the top rack, and the honest note that soft silicone can flex and tip when full.

Check the geometry, because it is a good one. An elliptical bowl inside a 7 by 5 inch footprint has a mouth of 27.49 square inches, and at 2.65 inches deep that is 72.84 cubic inches brim full, which is 1,194 millilitres. The stated 1,000 millilitres is 83.8 percent of brim full, which is exactly where a usable fill line should sit. Folded, 7 by 5 by 0.8 inches is 459 millilitres of envelope, so the bowl packs down to 30.2 percent of its expanded footprint volume.

And its residence time is minutes, because a travel bowl is filled, drunk from and emptied in one go. That is the freshest water in this entire article, and it is the one product here with no reservoir at all.

What this category does not publish

No fountain or dispenser in this catalogue publishes a flow rate. That is the quantity the whole appeal of a fountain rests on, since the argument for one is that moving water encourages drinking, and how fast it moves is the argument. None publishes a wetted area, which is what sets the evaporation and therefore the concentration. None publishes a pump wattage, which would also be a heat input into a small volume of water. And none publishes a recommended replacement interval for the water itself, as distinct from the filter, which is the only intervention that actually changes the age.

What the category does publish, and gets right, is one capacity claim with its own reasoning attached. The 8 ounce bowl explains why it is small. That is the only product here whose headline number comes with the mechanism behind it, and the mechanism is correct.

Our roundup of cat water fountains covers the category. The figure to carry into it is the reservoir volume divided by what your animals actually drink, and the useful threshold is that anything over about three days of supply is a bowl of water with a pump in it rather than a source of fresh water.

The honest limit, and the three numbers to carry

What is exact here is both results. The mean age of the contents of a constant volume vessel is the volume divided by the outflow, and for a well mixed vessel the age distribution is exponential, so the stale fraction is the exponential of minus the threshold over the mean. The concentration result comes from a solids balance with two terms and reduces to one plus the ratio of evaporation to consumption with nothing else in it. The gallon arithmetic on the gravity dispenser is multiplication. The bowl geometry on the collapsible pair is multiplication.

What is representative is every quantity of water. The 200 millilitres a day a cat drinks is the figure the whole page scales on: a cat drinking 300 shortens every mean age by a third and lowers the fountain's steady state excess from 39.7 percent to 26.5, while the ratio between the small bowl and the fountain is untouched because both divide by the same number. The 0.0444 and 0.013 square metre wetted areas, the 80 litres an hour pump rate, the 150 milligrams a litre of tap hardness, the two equivalents a litre of resin capacity, the 22 degree room at 50 percent humidity and the 18 degree water surface are all in the same category. The exponential age distribution assumes a well mixed vessel, which a pumped fountain is and a sealed gravity column is not, which is why the gravity case is treated separately as plug flow.

What this page deliberately does not compute is three things. It does not say what water age or concentration a cat will accept, because that is an animal preference and there is no measurable in any of these listings that bears on it; all that is computed is the physical difference between the options, and the preference remains the cat's. It does not compute biofilm, which is what the companion question about an unreachable pump housing is really about, because biofilm growth needs a surface area, a nutrient concentration and a residence time at a temperature, and while this page supplies the residence time, no listing supplies the first two and nothing in a product description could. And it does not compute the flow or the sound of a fountain against a tap, because no fountain here publishes a flow rate and the acoustic argument would need one; what can be said is that a kitchen tap at an ordinary domestic flow delivers several times the volume per second of any small pump, and that the water arriving is at the mains concentration by definition.

Three numbers to carry. Mean water age is the reservoir divided by the daily drinking, which makes an 8 ounce bowl 1.18 days and a 2.2 litre fountain 11.0, and a pump changes neither. The steady state concentration is the tap's multiplied by one plus evaporation over drinking, which is 39.7 percent for a fountain and 11.7 for a small bowl, and the difference is the wetted area the fountain needs. And topping up is not replacing: a valve that refills a bowl from 50 ounces back to 67 is cycling a quarter of its contents and leaving three quarters where they are.