Probably the most important factor affecting space mission cost is the cost of launch vehicles. Many of the major launch vehicles are designed to place payloads of 1500-6000 kg into geostationary transfer orbit, at a typical cost of $50M - 150M. Such launch costs are far to high for a low-cost mission and in any case a launch on a major launch vehicle could not be arranged at short notice. Although it may be possible to share a launch as an auxiliary payload and wait in orbit for a suitable target, as explained in Section 3 this is not a preferable option. Instead it is preferable to have a dedicated launch as required into a suitable parking orbit for the target in question.

Small Expendable Launchers. Many of the early Western space launchers had, by modern standards, very small payloads. The US Vanguard rocket could place 20 kg into LEO, whilst the French Diamant and British Black Arrow both had LEO capabilities of approx 50 kg [Baker78]. From the mid-1960s onwards launcher development concentrated on increasing payload weight, but during the 1980s the growing interest in small satellite development lead to a number of small and (relatively) low-cost launchers being produced. A further source of small launchers opened up in the late 1980s with the decommissioning of many US and former Soviet nuclear missiles. Russia in particular has been keen to convert former military missiles into launchers which can be sold for foreign currency.

Table 4.1 summarizes several of the small satellite launchers either available or the subject of credible proposals [Wertz96, Jane's, Bauer96]. Note that the 'Scorpius' programme is still in the early stages of development, whilst the prices quoted for Russian small launchers (both converted ICBMs) are extremely approximate. Taurus, LMLV* and Rokot all have sufficient capacity to carry the final configuration of probe and upper stage into any desired inclination LEO. The availability of such a launcher at relatively short notice (possibly by a flexible pre-purchased launch slot) would be a matter for a more detailed commercial study of this proposal to confirm.

[*Lockeed Martin Launch Vehicle - first successful flight 23 August 1997.]

Novel Launch Systems: Reusable Launchers. A frequently-proposed solution to the high cost of access to space is the development of reusable launch vehicles. It has often been pointed out that if aircraft were built to purpose for every flight and scrapped thereafter then air travel would be as expensive as space launch. Unfortunately the technical challenges involved in building a reusable launch vehicle (RLV) are considerable, as is evident from the continued use of expendable launchers some 40 years after the launch of the first artificial satellite. The one operational launch vehicle with any degree of reusability, the US Space Shuttle, is as expensive to operate in terms of $/kg to LEO as most expendable launchers and requires a large support operation to refurbish each Orbiter after flight.

The main difficulty with building an RLV is that of achieving orbital velocity (9,500 m/s including drag and gravity losses) without using disposable stages or tanks. The final velocity attained by a rocket is given by the 'rocket equation' [Larson92]:

(4-1)

where V_e is the effective rocket exhaust velocity, M the initial mass and M₀ the final mass. V_e is often expressed as g_oI_sp, where g_o is the acceleration due to terrestrial gravity (i.e. 9.81 m/s²) and I_sp is a quantity known as the specific impulse of the particular propellant combination and rocket engine being used. Specific impulse is measured in units of s, and can be thought of as the thrust per unit weight of propellant. Commonly used propellants have values of I_sp between 260s and 450s; to achieve a final velocity of 9,500 m/s with them would require a mass ratio of 9 to 41. Mass ratios much above 15 are very difficult to achieve, especially with any significant payload, whilst those propellants with sufficiently high I_sp to allow lower mass ratios, such as liquid hydrogen/liquid oxygen, are low-density, leading to difficulty in building large, structurally strong tankage. Also, to maintain a reasonable level of acceleration, a launcher must significantly vary its thrust during its burn. Expendable launchers drop off structure such as empty tanks and the large engines needed for takeoff, thus maintaining their effective mass ratio and lowering thrust throughout flight. Building an RLV without doing this is thus a difficult proposition.

In recent years though, advances in materials technology have given rise to the possibility that it may be possible to build a single-stage-to-orbit (SSTO) RLV. By using strong but light materials such as composites and advanced alloys it may be possible to build a vehicle light enough in relation to its size that it has a high enough mass ratio to achieve orbit with a useful payload. The NASA/Lockheed-Martin X-33 is being built as a technology demonstrator for such a vehicle, and should carry out suborbital flights in 1999. It is hoped that such work will lead to the development of operational SSTO-RLVs within the next 10 years. One of the main objectives of such research is to develop an RLV that can be refurbished between flights with little more than the routine servicing required by conventional aircraft. Considerable progress was made in this area by the McDonnell-Douglas DC-X/XA before it was lost in a flight accident, with rapid turnaround by a small flight crew being demonstrated. If successful this research should lead to an RLV that can fly at short notice (under a week) to deliver to orbit a payload at relatively low cost (under $1000/kg to LEO). Such a vehicle would be ideal as a launcher for a rapid-response probe, as it would make it feasible to procure a dedicated launch into a specific parking orbit at short notice and reasonable cost. However, as mentioned, it is unlikely that such a vehicle will enter operational service before 2007, with the 2010-2015 period being more likely. As such, whilst the launch of rapid-response missions is likely to become a much easier proposition with the development of SSTO-RLVs, this will not happen in the near future.

The SSTO-RLV is not the only approach to low-cost space access though. Making any element of an ELV reusable should reduce costs**; such savings are likely to be most pronounced if the part reused is the largest component, i.e. the first stage. The first stage is a particularly promising candidate for replacement, as not only does it have to carry the payload and upper stages, but it must overcome the most significant delta-V overheads (atmospheric and gravity drag) associated with launch. These overheads typically add 1,500-2,000 m/s to the delta-V required to reach normal LEO orbital velocity of 7,800 m/s. Even small improvements in this area can bring dramatic benefits. A good example is the Orbital Sciences Corporation (OSC) Pegasus air-launched rocket, which uses a converted L-1011 airliner as a reusable 'zeroth stage'. Even by taking the remaining stages to just 15,000m altitude and 250 m/s velocity, the Pegasus rocket is approximately halved in takeoff weight from a ground-launched version of equivalent performance. An additional factor in making just the first stage reusable is that its recovery is much easier than for an SSTO. An SSTO returning from orbit must cope with re-entry at approximately Mach 25, whereas a reusable first stage would reach a peak velocity of Mach 9 to Mach 12.

[**This is true so long as the element in question actually is easy to refurbish, and the cost of developing such a capability is not excessive. The US Space Shuttle, for instance, reuses all elements except its external fuel tank. However, the high development cost and the effort involved in refurbishment between flights result in its cost to orbit being as high as for most expendable launchers.]

A number of partly reusable launch vehicles (PRLV) have been proposed around the concept of a reusable first stage - often air-launched in some way - being used to boost the payload and upper stages onto a suborbital trajectory. During 1995 OSC carried out development work on the X-34, effectively a reusable vehicle that replaced the first stage of Pegasus. Although the X-34 has been redesigned as a suborbital test vehicle, other companies have made similar proposals. Kelly Space have started development work on the Eclipse winged launcher that would be towed behind an airliner before climbing on rocket power into a trajectory where solid-propellant upper stages and the payload would be released from a cargo bay. An alternative proposal is being developed by Pioneer Rocketplane; as the author has a more detailed knowledge of this concept (through acquaintance with one of its principal designers) it will be described in more detail.

Pioneer Rocketplane's 'Pathfinder' (originally designed as the US Air Force 'Black Horse' spaceplane study) is a small spaceplane that enhances its performance via aerial propellant transfer. It takes off using two conventional augmented turbofan engines, carrying a full load of kerosene fuel but only sufficient liquid oxygen (LOX) oxidiser to pre-cool its LOX tanks. It then rendezvouses with a tanker aircraft to take on a full load of LOX before using a LOX/kerosene rocket engine to climb to 100 km altitude at a speed of Mach 12. Here it releases a solid-propellant upper stage (typically a STAR-48 or -63 boost motor) which accelerates the attached payload into orbit. The Pathfinder then re-enters and relights its turbofan engines to return to base. Aerial propellant transfer allows the Pathfinder to reduce structural weight as its wings and landing gear are sized to support a much lower weight than if it took off with a full propellant load [Burnside Clapp94]. Calculations indicate that Pathfinder should be able to place 1,000 kg into polar LEO, or more into orbits of lower inclination, for a proposed cost of $5M. Pioneer are currently carrying out a design study on the Pathfinder for NASA as a contender for the Bantam-X RLV proposal; it is also a contender for the proposed USAF space sortie vehicle. If development is proceeded with, Pioneer propose to make operational flights early in the next decade.

Summary. The availability of an RLV or PRLV would enormously increase the feasibility of a rapid-response space probe as it would allow relatively economical launch at short notice. The mission could be carried out using one of several existing small satellite launchers but the costs would be higher and the question of launch availability at short notice remains open.