Why the rate has to be measured
Evaporation from an open surface is driven by the difference between the water and the air above it, and by how fast that air is replaced. Temperature, the humidity already present, and air movement all feed into it, and all three are properties of your particular room on a particular day. Any published rate is a rate for somebody else.
The measurement is trivial. Put a strip of tape on the inside of the tray at the water line, come back a day later, and measure from the tape down to the new line with a steel rule. At the defaults that drop is 0.20 inches over 24 hours, which on a 14 by 8 inch tray at 55 percent pebbles works out at 5.59 fluid ounces a day.
The alternative is to measure the refill instead of the level, which is more accurate on a shallow tray where a level drop of a tenth of an inch is hard to read. Both modes are on the page and both scale to a 24 hour figure from whatever span you actually used.
The pebble result, which surprises people
Pebbles in a tray cut two things at once. They take up space, so the tray holds less water — at the defaults, 55 percent of a 46.5 fluid ounce slab is stone, leaving 20.9 fluid ounces of actual water. And they break the free surface, so less of the tray is exposed water.
Those two reductions happen in roughly the same proportion, which means the level falls at about the same rate as it would in an open tray. So the refill interval barely changes — 3.75 days at the defaults either way — and what does change is how much water you carry to each refill. A pebble tray is a tray that runs dry just as often while holding less than half the water. That is not an argument against pebbles, which are there to keep the pot out of the water, but it is worth knowing that they are not extending anything.
The word "roughly" is doing real work in that paragraph. If the pebbles are large and rounded so that a lot of open water shows between them, the surface is cut less than the volume and the level falls faster than the open tray. Pack the tray with fine gravel and the reverse happens. The way to settle it for your own tray is to measure it both ways.
Surface, not depth
The rate follows the surface area. Doubling the depth of a tray doubles how long it lasts between refills and does nothing at all to the daily output; doubling the footprint doubles the output and leaves the interval where it was. That is the lever to reach for depending on which problem you have.
At the defaults, one tray delivers 5.59 fluid ounces a day. Reaching the 20 fluid ounces a day named in the target field takes four trays, which is 448 square inches of surface, or 3.11 square feet. The four of them together hold 83.8 fluid ounces and want refilling every 3.75 days.
What this does not calculate
It says nothing about humidity anywhere. The room the tray sits in has a volume, an air change rate, and other sources and sinks of moisture, and none of those are inputs here. A tray putting out six fluid ounces a day is a small contribution to a room and a large one to a sealed case, and only a balance over the actual space would tell you which.
It also has no view on whether any of this is worth doing for a plant. The target field takes a number you supplied from somewhere else. Nothing on this page recommends a humidity level, and no arithmetic can supply one.
Practical cautions
Water standing in an open tray indoors is a small stagnant pool. For most households that is unremarkable; for anyone whose immune system is compromised it is worth mentioning to a doctor. A tray that is not level will drain toward the low corner and over the edge before the arithmetic says it should be empty. And at the defaults a full tray is about 1.4 pounds of water, which is not much, but a deep tray on a large footprint is another matter — an open two inch fill on a 24 by 12 inch tray is around 21 pounds of water sitting on furniture.
Questions people ask
Do pebbles make a humidity tray last longer?
No, and this is the least intuitive result on the page. Pebbles reduce the water stored and reduce the exposed surface by roughly the same proportion, so the water level falls at about the same rate as in an open tray of the same footprint. The refill interval stays where it was and the volume you carry to each refill goes down. Pebbles are there so the pot sits above the water rather than in it, which is a structural job, not a longevity one. If your gravel is coarse enough to leave a lot of open water showing, the interval can even shorten slightly.
Should I measure the level drop or the refill volume?
Whichever you can read more accurately. On a shallow tray, a day of evaporation might be a tenth of an inch, and reading that against a tape mark is a coin flip; measuring the water it takes to refill to the same mark with a kitchen jug is far more precise. On a deep tray with a long span, the level is easy and the refill is fiddly. The page converts either into the same set of numbers, and running both once is a good way to check that your pebble fraction figure is right, because the two should agree.
How long should the measurement run?
Long enough that the change is easy to read, and short enough that the conditions have not changed underneath you. Twenty-four hours is a reasonable default because it covers a full heating cycle and a full day-night swing. Anything under about six hours risks the reading being dominated by measurement error, and anything over a week is averaging across weather you may not want averaged. Enter whatever span you actually used and the page scales it.
Can I use this to work out the humidity in the room?
No. That needs the volume of the room, how fast outside air replaces the air in it, and every other source and sink of moisture in the space, none of which are inputs here. This is a water balance for one tray: how much leaves it, how often it needs filling, and how many trays a given daily output would take. The humidifier sizing calculator does the whole-house version of that balance, and the humidor humidity drift calculator does it for a sealed case where the air itself turns out to matter less than the buffering material.
The tray dried out much faster than the calculation said. Why?
Usually one of three things. The heating came on, which raises both the temperature and the air movement over the tray and can easily double the rate. The tray is not level, so the water reached the low edge and went over rather than into the air. Or the pebble fraction is wrong — that figure is the one people estimate rather than measure, and being out by twenty points moves both the stored volume and the calculated rate. Measuring the fraction with a jug takes two minutes and fixes all the numbers at once.