Pressure tank factors: In fact, tank weight as a percentage of the weight of pressurized fuel does not change at all when going from a very small to a very large pressure fed rocket. For composite pressure vessels made of carbon fiber and epoxy, one can roughly estimate the tank mass by using a "performance factor", which relates the mass of the tank to the enclosed volume and the enclosed pressure. To a first approximation, tank shape is irrelevant, as with composite construction the tank wall can be tailored to have only the required strength in each direction. Performance factors can take a couple of flavors, depending on the units used. In SI units, a common factor is expressed in kPa*m^3/kg at tank burst. For graphite/epoxy tanks, a rough number for estimation purposes would be about 300 kPa*m^3/kg. Using a safety factor of 1.5, you can derived a working performance factor of about 200 kPa*m^3/kg. This means that 1 kg of tank will enclose 0.1 m^3 at 2 MPa (about 300 psi) etc. Most designs for pressure fed launchers optimize at chamber pressures somewhere between about 1 MPa to 3 MPa. Assume a typical launcher has a tank pressure of 2 MPa (leading to a chamber pressure in the vicinity of 1.5 MPa). A 1 cubic meter tank will weigh about 10 kg and hold about 800 kg of kerosene, 1140 kg of liquid oxygen, or 1430 kg of hydrogen peroxide. This scales up linearly - a 100 cubic meter tank will hold 80 metric tons of kerosene etc. In all sizes therefore, tank mass is about 1 % of propellant mass if the pressure is at 2 MPa. At 1% of propellant mass, composite pressure fed tanks weight roughly the same as aluminum tanks for a pump fed vehicle. The composite tank for the pressure fed vehicle is running at perhaps 5 to 10 times the pressure of the aluminum tank for the pump fed vehicle (pump fed vehicles also require tank pressurization, to give adequate pressure at the suction side of the pumps). However, for tank construction purposes the composite tank wall material has something like 4 times the strength per unit mass as aluminum. In addition, the stiff, thick walls of the graphite/epoxy tank are self supporting, while the aluminum tank for the pump fed vehicle requires stiffening ribs of some form on the tank interior, and reinforcements whereever concentrated loads are applied to the tank. Using composite tanks, it is possible to produce pressure fed liquid propellant rocket stages with propellant fractions in the region of 0.93. As an existance proof, look at the current range of upper stage solid rockets with composite cases. These can have propellant fractions as high as 0.95. While the solid rockets have dense propellants (density about 1.75) they also have a substantial unfilled hole in the middle of the propellant. While pressure fed liquid stages have plumbing and valves that solids do not, solids have internal case insulation which is not present in liquids. Most important, solids typically operate at several times the operating pressure of pressure fed liquids, correspondingly increasing their pressure shell mass. Overall, we should expect newly designed pressure fed liquid rockets to have roughly the same propellant fraction as has been available with solids for years, to have a higher specific impulse, to have more environmentally friendly propellants (no more HCl in exhaust), and to have throttling, stop and restart capability.