The pipe is most of it, and it is easy to get wrong
A foot of half inch type L copper holds 0.01212 gallons. That sounds like nothing until you count the run properly: a collector on a roof and a tank in a basement is not a 35 foot run, it is a 70 foot loop, and 70 feet at that bore is 0.85 gallons. Go up to three quarter inch type L at 0.785 in and the same 70 feet holds 1.76 gallons, because volume scales with the square of the bore.
The trap is the nominal size. Half inch is not a bore, it is a label, and the actual bore behind it runs from 0.475 for PEX to 0.602 for schedule 40 PVC. That is a sixty per cent spread in gallons per foot for pipe everyone calls the same thing. The bore table on the page shows what the same run holds across six of them.
Adding up the fill
With the defaults: two panels at 0.35 gallons is 0.70; 70 feet at 0.545 in bore is 0.848; the second size is switched off; components add 0.60. That is 2.15 gallons in the loop. A 40 per cent fill from neat concentrate takes 0.86 gallons of product and 1.29 of water, and since concentrate comes in five gallon containers, one container does it with plenty left.
The arithmetic that matters is the last part of the fill. Once the loop is full and the pump has purged the air, the temptation is to top up whatever settles out with whatever is to hand. Water is what is to hand. Topping up two litres of a two gallon charge with water moves the concentration by several points, which is a real change in the freeze point and not a rounding error.
Growth is not the same question as the vessel
The page reports how many gallons the charge grows by between the temperature you filled at and the temperature the loop runs at, because that number is a property of the fluid and the volume and belongs here. It does not size the vessel that absorbs it. That is a Boyle law calculation involving the precharge, the fill pressure and the ceiling you are working to, and it lives on the expansion tank page.
Worth saying plainly: the growth figure on this page is for normal running. The volume a loop displaces when a collector stagnates and its contents flash is a different and much larger question, and it is not arithmetic this page or any other page here will do for you.
Filling an empty loop and changing a full one
These are two different sums and confusing them is the usual mistake. Filling empty: the product and the water go in at whatever ratio gives the target, straightforwardly. Changing a full loop: whatever is already in there dilutes whatever you add, so you have to drain a calculated fraction first, and if the loop will not drain completely you have to do it more than once. The second problem is on the coolant mix page, which handles the repeated-exchange case.
What is not decided here
Nothing on this page says what concentration a loop should carry, what freeze point is adequate for a location, which fluid product is appropriate, or how often it should be tested. Those come from the fluid data sheet, the equipment manufacturer and whoever designs the system. The freeze point field exists so the figure you took off the data sheet travels with the rest of the numbers, not because this page has an opinion about it.
Questions people ask
How much glycol does a solar hot water loop hold?
It depends almost entirely on the pipe. With two panels at 0.35 gallons each, 70 feet of half inch type L copper counted for both legs, and 0.6 gallons of exchanger, separator and pump, the loop holds 2.15 gallons. Move that pipe up to three quarter inch and the same loop holds 3.06. Panels are usually the small part; the run to and from them is the big one.
How many gallons of concentrate do I buy for a 40 per cent fill?
Loop volume times the target divided by the strength of what you are buying. A 2.15 gallon loop at 40 per cent from neat concentrate takes 0.86 gallons of product and 1.29 of water. From a 50 per cent pre-diluted product it would take 1.72 gallons of product and 0.43 of water. If the target is above the product strength it cannot be reached at all, and the page says so.
Why does the calculator ask for the bore rather than the pipe size?
Because the nominal size is a label and volume goes as the square of the actual bore. Half inch copper, half inch PEX and half inch PVC hold 0.0121, 0.0092 and 0.0148 gallons per foot, a sixty per cent spread for three pipes everybody calls half inch. Over a 70 foot loop that is the difference between 0.64 and 1.04 gallons, which changes how much concentrate you buy.
Can I just top up the loop with water?
You can, and it dilutes the charge. On a small loop it does not take much: adding half a gallon of water to a 2.15 gallon charge at 40 per cent leaves it near 30 per cent, which is a real change in the freeze point rather than a rounding error. Whether that matters, and what the concentration ought to be, is a question for the fluid data sheet and the system designer, not for this page.
Does this size my expansion tank?
No. It tells you how many gallons the charge grows by between the fill temperature and the running temperature, which is a property of the fluid and the volume. Turning that into a vessel size needs the precharge, the fill pressure and the ceiling you are working to, and that Boyle law calculation is on the thermal expansion tank page. What a stagnating collector displaces is a larger question again and is not arithmetic offered anywhere here.