From a pure performance standpoint, if you multiply the overall density (including the rocket motor casing, nozzle hydraulics packages, etc.) times the Isp (specific impulse, or lbs of thrust per lb of propellant used per second), you get the Density-Specific Impulse. Solids in the class of the Shuttle SRBS have a relatively higher Density-Specific Impulse than similar-performing liquids would have in the configuration of the Shuttle stack.<br /><br />The original shuttle concept called for a re-usable, manned fly-back booster, with jet as well as rocket engines. I don't recall if they intended to launch horizontally or not, but it is possible.<br /><br />When it became obvious that you could not build the whole thing for an amount of money that could be obtained from Congress. The technology developments required for the TPS (tile system), and the SSME's ate up too much money. The Shuttle Solid Rocket Boosters could be recovered by parachute. The amount of money saved on the first three or four flights (NOT including the one where the ringing frequency of the separation ring firing triggered the Main Deck Fitting separation explosive bolt firings and we lost both boosters...STS-4) paid for the development of the parachute subsystem (SRB Decelerator SubSystem), about $24M. <br /><br />Solid rocket boosters have been used very successfully (though NOT without spectacular failures) on the Titan III and IV. <br /><br />The rationale behind using a single SRB for launching a CEV is that the boosters are already "man-rated"! This saves a long and expensive series of static and flight tests that would be required for any other vehicle, including the Atlas V and Delta-IV.<br /><br />If this sounds great, I've got a launch pad on Merritt Island I'll sell you for 5M laudinium credits!<br /><br />Yeah, I've worked liquids and I'ves worked solids, and for manned and/or high-value payloads, I wouldn't want anything to do with a large firecracker, an escape tower built into the forwar