The purging of stagnant or dead volumes in gas distribution systems is an important method for removing impurities and maintaining cleanliness. A combination of experimental investigation and computational process modeling is used to study the dynamics of impurity removal under variety of purge conditions. The controlled cycling of pressure during purge is found to enhance the cleaning process significantly, particularly in dead spaces. The process simulator was used to develop and analyze a pressure-cyclic purge (PCP) method and understand the conditions that would make PCP advantageous over steady-state purge (SSP). In particular, the effect of geometric factors, impurity surface interactions, flow rate, and cycle characteristics on PCP and its comparison with SSP was studied. The advantage of the PCP method, in terms of both purge time and gas usage, becomes more pronounced in systems with larger number and size of dead spaces and impurities that interact strongly with the surfaces. (C) 2015 American Institute of Chemical Engineers.