Previous works on MDA [1] [11] have shown that most promising catalysts results from the combination of a transition metal with a zeolite. Particularly, molybdenum impregnated on a HZSM-5 zeolite support is one of the catalysts most studied. The unusual resistance to deactivation by carbon deposition of that zeolite [12] could probably be one of the key factors for functioning in MDA. Furthermore, making-up methane feed with a small concentration of hydrogen plus steam, is reported to stabilize MDA suppressing deactivation effectively [13] . Nevertheless, previous reports [1] [11] [13] [14] suggest low methane conversions (~10%) for steady state MDA.

For MDA, a trickled bed reactor is recommended where both natural gas and the liquid products flow down across a bed of pelletized Mo/HZSM-5 catalyst. It should be anticipated the necessity for catalyst regeneration in the case that deactivation by carbon deposition is not stabilized. One alternative is to stop MDA stage by- passing methane directly to MRC stage while catalyst regenerating. Another alternative is to employ two parallel reactors to allow catalyst regeneration in one reactor while carrying MDA in the other.

Two alternatives are possible for connecting MDA to MRC: one is to separate the liquid MDA products leaving non-reacted CH₄ plus H₂ product to be directed as the gaseous reactants for MRC. The other is to direct all MDA output to MRC. This latter is backed by recent report suggesting a synergistic effect between coal pyrolysis and methane aromatization [15] . Certainly, the aromatics produced by MDA may improve diffusion of the gas reactants to the carbonaceous surface enhancing liquefaction.

Catalytic agents and/or solvents should be introduced for MRC. It is well known that the use of solvents improves coal liquefaction. Previous experiments on the liquefaction with H₂ of a Venezuelan coal [16] demonstrated the advantage of using H-donor solvents. Therefore, in addition to the aromatics coming from MDA, certain fraction of liquids produced in MRC stage should be pre-treated with H₂ and recycled to function as H-donor solvent for the liquefaction.

Early experiments using batch reactor yield high conversions (~50%) when reacting CH₄ and H₂ mixtures with coal [3] [5] , as well as in the case of Venezuelan heavy oil [17] . The possibility of adapting two parallel batch reactors operating similarly as in delayed coking processes may be one alternative to consider for MRC.

Several other alternatives may be investigated for continues flow MRC process reactor: A fluidized-bed or ebullated-bed as employed in the H-coal direct-liquefaction process [8] should be taken into account. The use of small coke particle size is recommended in order to ensure more solid surface area that should improve reactant gas diffusion phenomena to favor coke reaction kinetics. Tubular reactors where a slurry of the carbon powder plus a solvent containing dissolved gases (CH₄ + H₂) flowing continuously have also been experimented [4] , as well as microwave plasma reactor [6] and flash pyrolysis [2] both under CH₄ supply. A novel reactor experimented using Venezuelan petroleum coke under steam gasification employing solar energy [18] could also be considered introducing the variance of employing CH₄ instead of steam.

Coke liquefaction residue extracted from MRC may be an appropriate raw material for activated carbon manufacture as proposed else where [19] . In addition, it could have properties similar to calcined coke, a commercial material useful for application in metallurgy and other carbon consuming industries. The possible application of the residual coke for agriculture and desert greening as proposed earlier [20] should also be investigated further.

Figure 2 shows a basic flow diagram for the proposed two-stage process for co-upgrading natural gas and petroleum coke into liquid hydrocarbons. Notice that non-reacted gases are recycled to MRC stage after stripping of H₂. In addition, certain distillate fraction is recycled to MRC reactor to function as solvent to enhance reactant gas diffusion to solid. Therefore, H₂ stripping from MRC output gases may also be available for the hydrotreating reactor as show in Figure 2, an operation intended to produce required hydrogen donor solvent.

Basic engineering analysis employing reactions I and II (n = 8), assuming coke is the limiting reactant employing excess methane, and discrete values of MDA conversion (10%) and coke conversion (30%) to light petroleum, derived from the literature investigated; therefore, it is estimated that 1.500 m³ (1 atm, 25˚C) of natural gas is required to produce 8 bbl of light petroleum for each ton of petroleum coke processed. Venezuelan petroleum coke produced today (20,000 ton/day) would be equivalent to 160.000 bbl/day of light petroleum using this process.