I have a spreadsheet I use for estimating the mass of fuel tanks. It has been useful for indulging my obsessions, so I'm posting it here in case someone else finds it useful. Or wrong. I'd love to hear about either.
It has two tabs, one for spherical tanks and one for cylinders with spherical endcaps. I'll consider doing a 2:1 endcaps tab if someone actually wants such a thing. The spherical tank tab assumes you only want one fluid in the tank, while the cylindrical tab assumes you want two fluids separated by a flat circular bulkhead.
First is a set of user-entered values that define the outside dimensions and material properties of the tank plus the properties of the fuel and oxidizer. Enter these (through row 20) and set a desired safety factor. Row 31-34 will provide minimum thickness values; you can enter these or your own values in rows 22-28. The sheet will calculate where to place the bulkhead based on the fuel ratio.
Results are shown in three groups: blue for tanks without ullage, purple for tanks with pressurized helium gas and green for tanks with liquid nitrogen (not present for spherical tanks). Note that a tank without ullage is likely to have a higher liquid retention value, so these are not a good apples to apples comparison. Given are the dry mass, fully-fueled takeoff mass, nominal burnout mass and the net amount of 'burned' mass, plus the mass fraction. Mass fraction is calculated as burnout mass divided by takeoff mass, so it includes the mass penalties of unburned fuel, ullage gas and tankage for the ullage gas.
Also provided is a first cut estimate of the mass required to equip this tank with a PV powered cryocooler, excluding batteries and power conditioning hardware. The system is sized to provide zero boiloff in Earth orbit for a full tank. It will be oversized since some of the tank's surface area will have a view of deep space at 4 °K instead of Earth. It also assumes a sunshade.
Next are a series of calculation steps. If you want to know the mass of helium required or the length of the liquid N2 cylinder these are available here. After that are geometric calculations for three types of clustering: hexagonal, square and triangular. Lengths of support members and available clearance for ullage tanks and fuel lines are provided. Spherical tanks are only assumed to have a 2d packing; more efficient 3d packing geometries are possible.