In most of the missile systems , liquid bi-propellant systems are used and there we have separate tanks for the oxidizer and fuel. There is a separate tank with an inert gas maintained at high pressures. This is connected to the fuel and oxidizer tanks. Each of the fuel and oxidizer tanks have a partition inside using a diaphragm and the pressurized inert air from the third tank is pumped into the fuel and oxidizer tank to maintain their pressure levels.

This is done to reduce the weight of the system and also space optimization. Places where weight and space are not critical there a separate pressurization system may be used to reduce the complexity.

What you described is separate pressurization systems sharing only inert gas tank.

What i am asking why people bother with separate valves and regulators for fuel and oxygen if it is enough to put connecting pipe between pressurization volumes in oxidizer and fuel tank to equalize pressure? Or even flexible diaphragm between fuel and oxidizer in shared tank? Is something wrong with these ideas?

Usually you have high pressure differences between the liquidising pressure of the oxidizer and the pressure you neet to maintain for the liquid propellant. This alone makes ist impossible to pressurize the tanks with the same pressure.

The diapragma solution is complex due to the mass and volume loss in the tanks on both sides during operation - how do you want to maintain the RIGHT pressure for both tank sides?

Oleg: You must specify if your question applies to storable or cryogenic propellants. I assume you deal with storable propellants. You must specify if your application is a missile or an upper stage or a satellite. In a missile, pressurant and propellants are usually separated by burst discs for one-time operation. You cannot open two burst discs from the same pressure source since the discs release at slightly different pressures, even if they are designed and made identical. For a satellite or a space probe, having only one common pressurant tank for both fuel and oxidizer proved disastrous in the case of Mars Orbiter on August 21, 1993. The most likely failure mechanism was that N2O4 oxidizer had diffused upstream through the Teflon check valve and when the system was activated after a long interplanetary cruise, N2O4 mixed with MMH and exploded in the feed lines. Subsequently, the design was changed to two pressurant tanks in Cassini and it worked fine. Propellants permeate through diaphragms and leak past pistons in positive expulsion tanks.

For Eckart Schmidt:

I thought about propane/LOX combination for launch vehicle akin cancelled FALCON project. They give a clue about "alternative tank configurations" possible with these fuels, so i am investigating.

Yes, i know about upstream drift of vapors. Similar problem has crippled Akatsuki probe. But for launch vehicle, given smaller time to diffuse, problem may be manageable.

For Stefan Kramer:

The diaphragm solution i imagine is flexible polyimide diaphragm between propane and LOX operating at near zero differential pressure (see attached image). Therefore, only LOX need to be pressurized (by VAPAK system or conventional He pressurant)

Density of LOX and chilled propane is similar, and static pressure differential is essentially zero, therefore the diaphragm can be very thin (we test 75um membranes currently)

Added file depicting the tentative diaphragm shape

I see some serious challenges here. I have never worked with propane, so hopefully my numbers are good. First, you must sub-cool the propane from its boiling point of ~ 230K to the boiling point of oxygen (~90K) prior to loading it. If you don't, with the large heat transfer area in the design, you will get massive boil-off of LO2 during tanking and countdown. Sub-cooling propellant will come at substantial launch facility expense, adding to processing, storage, and transfer system costs. Second, the tanking process must load the propellants such that no significant head difference occurs which would lead to membrane system damage. This would be a significant challenge to loading control systems. Third, the membrane system must reduce the risk of mixing of propellants to an acceptable level. That means it must provide a minimum protection of surviving a single failure (2 membrane systems?). Mixing of propellants is an unacceptable risk to personnel and launch pad facilities. Fourth, you must show that this membrane design provides a performance (and/or cost) gain that is worth all of this risk. That looks to me to be a trade between pressurization system hardware and separate tank weights vs membrane system weights (as a minimum). There may be other weight penalties and launch pad costs if experienced launch service and rocket engineering minds were put to work on it for more than the few minutes that I have devoted to it.

Well, i tried sub-cooling propane in tank. Load first full tank of LOX, open oxygen bleed valve and start bubbling warm propane to inner bladder. Of course, it works for small (