Gallons out are not gallons stored
The multiplier is the part people find surprising. If the store is at 155 °F, the cold is at 55 and the water is used at 110, then every gallon of store carries enough heat above the cold supply to make 1.82 gallons at the delivery temperature: (155 − 55) ÷ (110 − 55). A 44 gallon hot layer therefore delivers 80 gallons. That is not a trick; it is where the heat actually is, and it is why a small hot store beats a large tepid one.
Run the same arithmetic on the stirred version of the same tank. 44 gallons at 155 and 36 at 80 average out at 121.3 °F, and the whole 80 gallons then yields 80 × (121.3 − 55) ÷ 55, which is 96.4 gallons. The stratified reading gives 80.0 from the top layer alone plus nothing from the bottom, because 80 °F is below the delivery temperature and cannot contribute.
So mixing sometimes wins
That is the honest result and the page reports it either way round rather than assuming stratification is always better. When the bottom layer is below the delivery temperature, stirring promotes some of that dead volume above the line and the total usable figure can go up. When the bottom layer is already above the delivery temperature, stratification and mixing deliver exactly the same total, because the heat is the heat and the blending multiplier is linear in temperature.
Where stratification genuinely wins is not in this snapshot at all. It wins in what it does to the collector: a cold bottom layer sends cold fluid back to the panel, the panel sits higher on its efficiency curve, and it collects more tomorrow. It also wins on the outlet temperature during the draw, which the layer-by-layer table shows — a stratified store holds its temperature until the hot layer runs out, and a mixed one slides all the way down from the start.
Reading the break
Two thermometers and a tape measure will do it. Feel down the side of an uninsulated tank and the band is often findable by hand; on an insulated one, a multi-point probe or a strap-on sensor moved down in steps works. The band is thicker than people expect, usually several inches, and it thickens every time the tank is stirred — by a high inlet velocity, by a recirculation return dumped into the middle, by a heat exchanger loop running fast.
The sensitivity block at the bottom of the page shows what a fifteen point error in that reading costs. It is usually the largest uncertainty in the answer, larger than either probe temperature, and it is why the page asks for it rather than quietly assuming a usable percentage the way most tank calculators do.
What this is not
It is a snapshot of a store as the probes read it at one moment. It does not model the collector putting heat in while the draw is going out, it does not model the backup heater, and it does not know the tank is losing heat through its jacket the whole time. It states no temperature and issues no verdict about whether anything is correctly sized or safe. Water stored at solar temperatures scalds, and how a system keeps that away from a tap is a design question for the people who design it.
Questions people ask
How much usable hot water is in a stratified tank?
More than the tank holds, if the store is hotter than the delivery temperature and cold is blended in. With the defaults, a 44 gallon layer at 155 degrees blended to 110 with 55 degree cold yields 80 delivered gallons, because each stored gallon carries the heat for 1.82 delivered ones. The layer below the break at 80 degrees contributes nothing, since it is under the delivery temperature.
Is a stratified tank always better than a mixed one?
No, and the page will tell you when it is not. If the bottom layer is below the delivery temperature, stirring lifts some of that dead volume above the line and the usable total can rise. If both layers are already above it, the two answers are identical. What stratification reliably wins is a colder return to the collector, which raises tomorrow collection, and a steady outlet temperature during the draw rather than a slide.
How do I find the thermocline in my tank?
A strap-on thermometer moved down the side in steps, or a multi-point probe if the tank has one. On an uninsulated vessel you can often feel it. It is a band rather than a line, commonly several inches thick, and it gets thicker every time the tank is disturbed. Enter it as a share of tank height; the sensitivity block at the bottom of the page shows what a fifteen point error costs you.
Why does my solar tank lose its layers?
Something is stirring it. A high inlet velocity, a heat exchanger loop circulating fast, a recirculation return discharging into the middle of the tank, or a draw that pulls hard enough to entrain the layer below. Low flow through diffusers, side-facing inlets and taking the draw from the very top are the arrangements that preserve stratification, and they are decisions made when the system is built rather than afterwards.
Can the usable volume be more than the tank size?
Yes, whenever the store is hotter than the delivery temperature and there is cold available to blend with. The tank is storing heat, not water, and the delivered gallons are the tank gallons multiplied by how far the store sits above the cold supply relative to how far the delivery temperature does. It also means the same tank delivers far fewer gallons in winter, when the incoming cold is colder and the multiplier shrinks.