This article presents a novel method for the optimal integration of large-scale bio-SNG (biomass-derived synthetic natural gas) production system into the existing natural gas network. A thermodynamic model representative of a commercial bio-SNG process fed by forestry residues has been built and validated with experimental data. The variation of the capital and operational expenditures as functions of the power output is calculated, identifying the economies of scale of the process. A spatially-explicit model, based on a modified location-allocation algorithm, has been built to find the site and size of the plant(s) that minimise the levelised cost of the energy produced (LCOE). The combined effects of the capital, operational and transport costs resulted in an optimal LCOE of 86.3 (sic)/MWh for a single 71 MWSNG plant. On the other hand, the most profitable configuration has been found for a two-plant configuration, which has a net present value of 165.8 M(sic) and LCOE of 89.5 (sic)/MWh. The resulting optimal configurations are shown in maps, including the resource harvesting sites and the selected bio-SNG and grid injection points. The mean distance from harvesting sites to the one-plant injection point is approximately 95 km, which means transportation costs, while small are not negligible. A scenario analysis shows the impact of the resource cost and incentives for bio-SNG injection, showing that a low cost of the resource (45 (sic)/t(DB)) would reduce the one-plant LCOE to 59 (sic)/MWh, which is currently above the cost of natural gas for household consumers, 52 (sic)/MWh.