Production planning in flexible manufacturing requires the updating of work schedules each time a new product order is acquired. To this end, a Production Planning (PP) strategy should be devised by solving two coupled problems, embedded into an iteration: (i) Product Orders Scheduling, (ii) Production Capacity Planning. This paper presents a new algorithm to solve the Production Capacity Planning (PCP) Problem for a complex manufacturing system composed of a network of bufferized time-sharing workshops (i.e. shops able to share their service rates among different part types). The solution procedure developed allows one to minimize the Global Flow Time (GFT), while verifying service and storage capacity constraints, in association with a given future pattern of product orders. An example derived from an existing production process is outlined, to show the effective usefulness of the proposed PCP procedure, and its easy integration with existing scheduling tools