High performance oxidizers that are similar to LOX performance (of which there really aren't many reasonable ones), are generally very difficult to obtain and handle. LOX is currently the best (and in our opinion, probably the safest) high performance oxidizer available for both booster stages and upper stages. H2O2 (hydrogen peroxide) and WFNA (white fuming nitric acid) are about the only other realistic choices. N2O is possible for some applications, but it's a relatively lower-performance oxidizer.

In order to be a reasonable oxidizer, H2O2 must be at very high concentration. Although it is more dense than LOX (an advantage), H2O2 has the unfortunate quality of self-decomposition, and a greater ease of self-decomposing at slightly lower concentrations. In addition, obtaining the required high concentration H2O2 may be almost as difficult as the N-Prize itself (that's a joke of course, but it's quite tough to get) – not to mention that it's prohibitively expensive. In our estimation, this is a non-starter from a practical/testing perspective, and storing reasonable quantities is simply not a good option. And, distilling lower concentrations to high purity is again problematic. So, H2O2 is somewhat easily ruled out for this attempt, if just for practical reasons alone.

WFNA is a somewhat similar story, but with somewhat different problems. Like H2O2, it appears as a reasonable choice for an oxidizer. But, it's a very corrosive chemical; long term storage of uninhibited WFNA can pose a problem. It does not pose as great a decomposition risk as H2O2 (there is still some risk), and although it's not as high-performance as LOX, it's also more dense and thus can save tank weight. WFNA has the advantage that it can be hypergolic with many fuels. Preparing quantities was not a problem and such is reasonably safe (well... as far as very hazardous procedures go). Unfortunately, our experience is that the general availability of bulk WFNA is again lacking when compared to LOX, and that's somewhat problematic (especially considering that WFNA does not reach the performance of LOX).

In a pressure-fed system the oxidizer density can become quite important (namely due to heavier tanks in the higher-pressure systems); these higher density oxidizers can provide about a 25% reduction in tank size over LOX which allows for a smaller, lighter dry mass per stored propellant mass; this compensates somewhat for the overall decreases in performance in some systems. However, in lower-pressure and/or pump-fed systems the tanks comprise a relatively smaller fraction of the mass, and the LOX tends to win out.

Given the typically higher performance of LOX and our experiences with the far greater ease of obtaining it (plus several other aspects that are specific to our particular designs), LOX is (and has been) our oxidizer of choice. However, we are not exclusively committed to LOX for the whole system and remain open to considering other oxidizers (including WFNA) in some of our approaches.

Our simulations indicate that LOX/LH2 performs worse (overall) for small boosters and perhaps marginally better for stages to LEO, depending on the feed system. Combining the worse booster performance with the added size and weight of the upper stage tanks only makes matters worse at these smaller scale sizes, and the overall performance of the whole system is reduced. In addition, LH2 is highly cryogenic and far more difficult to handle than LOX.

Note that if the overall mass fractions remained identical from system to system, LOX/LH2 would significantly outperform the LOX/Hydrocarbon. The problem stems from the very low density of LH2 and the size/weight of the tanks required to store it; such would necessitate more advanced technologies for LOX/LH2 to be competitive (mass wise) with the LOX/Hydrocarbon system, especially when pressure-fed.

At the same time, LOX/LH2 in upper stages could work advantageously in a pressure-fed system due to the relatively lower pressures required in a vacuum engine (and thus the ability to retain reasonable mass fractions). Although this would increase delta V it would also increase the complexity considerably since LH2 (as mentioned) is very difficult to work with and store.

We are currently looking at a few different systems, but are not ready to disclose our approaches yet.

Wow... It's amazing how fast that spread and how many questions we got on this. Yes, we are experimenting with a unique expander cycle engine but there is no guarantee we are actually going to use it for the N-Prize.