I'm afraid I must reverse myself on the idea of using solar power to electrolyse water into oxygen for life support. While water is the most convenient package for bringing oxygen and hydrogen to the platform, some examination of how to fit the processing system into the hab reveals that it is not the most efficient way to deliver oxygen and hydrogen once you get it there.

Between breathing and power generation, the crew requires about 2kg of oxygen per man day. Producing this from water also produces three other products: hydrogen, half of which (the same number of moles as oxygen produced) is stored for power generation and the other half is sent directly to the lift cells to replace daily losses due to leaks and diffusion; water vapor, which must be condensed out and can be returned to the system; and ozone (a poison), which must be removed by converting it to usable oxygen with a catalyst, returning it to the system. In addition to a heat exchanger and catalyst bed, this also requires electrolysers, solar panels to run the electrolysers, tanks and compressors to store the gas, and regulators & tubing to distribute the gas to the crew and back to the fuel cell.

We would use two 10kW reversible fuel cells in this scenario. With two, we can alternate them between producing gas and producing power. A single reversible fuel cell doesn't have the power output of other types of fuel cell of the same mass, but does have the same or better gas output ratio as a dedicated electrolyser of the same size, so it's better to just use two (or one reversible and one high power fuel cell) and have the reserve power handy.

We require at least a 24 hour reserve of gas, because we will be unable to obtain more overnight and must have sufficient reserves to repair a non-functioning system during the day. Unfortunately, this requires a large amount of tankage, for at least 1000 moles. Considering the low rate at which we can expect gas to come out of the electrloyser and feed through both the catalyst bed and heat exchanger, I don't think we can expect to achieve high pressure in the tanks if we're using half their contents or more on a daily basis. Rather than the hundreds of atmospheres found in industrially bottled oxygen cylinders, I think the most the system will be capable of is the ten or twenty atmospheres of a garage air compressor. Unfortunately, it means that the number and volume of all the tanks required to store all of this gas will be quite large. Though the walls are thinner on low pressure tanks, there will be more of them and so they will weigh more than for gas bottled at higher pressure. Chilling the tanks by storing them outside will reduce their required volume, but also reduces compressor efficiency and still does not come near the reduction possible by storing the gas at higher pressure. And storing at high pressure can only reduce their required mass by about a third. The tanks will still be heavy because their walls must be thicker.

It is a testimony to the VBP's versatility that 1500kg is still within our mass budget. However, shipping up enough oxygen on the elevator for 50 man-days requires just 45kg of oxygen in a 40L dewar for the crew. One crewman could probably lift each full dewar by himself. The distribution system for liquid oxygen is relatively simple and light. Furthermore, it isn't necessary to keep feeding a fuel cell because a fuel cell isn't necessary. So long as it is recharged daily by the solar array, a bank of nickel/metal-hydride batteries can do the same work as a fuel cell at the same weight without requiring a drop of fuel. (We will not need two, although it is desirable to break a single battery bank up into parallel.) And halve the size of the solar array, since it no longer needs to crack water for the crew's air. The entire system could easily be accomplished with less than 600kg, including the dewar.

So, rather than introducing an equipment problem to solve a grocery problem, I recommend we get rid of both problems. Keep shipping up the oxygen, which has to go up no matter what, only ship it in LOX form which is immediately usable upon delivery. This way, we will not be limited to a day's reserve of breathable air. Abandon multiple reversible fuel cells, and the mass of systems and added fuel that go with them, in favor of compact banks of metal-hydride batteries that do not require fuel or extensive maintenance.