<i>Theoretically, from an economics perspective, this means keeping an extra two engine production lines open. For the SRBs, the refurbishment and assembly infrastructure alone costs $400m per annum before you build the first unit. Then you have maintaining the VAB, crawlers, pads, and a J-2X and Ares-I production line that only pushes out six units a year.<br /><br />Exactly how many Atlas and Delta CBCs do you get for that sort of money?</i><br /><br />I don't know what the overheads for the EELVs are. But they are about $180~240 million a flight at the current (relatively low) flight rates. I don't think that the J-2X is more expensive to make than say the RD-180. I am not sure you'll get more engines for the money if you go with the latter. The SRB is also the simplest and cheapest booster type that can be made. There is no turbomachinery, no pressure tanks, no insulations, none of that. Basically, its a tube with solid propellant poured into it. The only "complicated" part being the thrust vectoring nozzle. And this is still no more complicated than the typical gimbal for liquid engines.<br /><br />In short, given the same flight rate and the same production volume, solids are ALWAYS cheaper. They are also storable fully fueled and practically indefinitely.<br /><br />The J-2X is actually a less complicated tap-off cycle engine than the Staged Combustion RD-180 which is about as complicated and as highly stressed as the SSME (RS-24).<br /><br />I am pretty sure that by employing the same quality standards and at production volume, an ARES based LV system is cheaper than either the Delta IV or the Atlas V. The reason being that it uses a relatively simple upper stage engine and a solid booster instead of two liquid engines and stages.