I favor Martin's suggestion. I will add to that a suggestion that you include a realistic X order stock/flow structure (third order ought to suffice) to model the acquisition of additional capacity, and to the degree it is relevant, a first order decay (stock/time to decay) to model the loss of capacity.