No abstract

Purifiers are widely used in refinery hydrogen networks. The placement of purifiers is critical in reducing the utility consumption. Various conceptual methods have been employed to find the optimal placement of the purifier inside the hydrogen network. However, previous research only adopts the performance based purifier model, leaving the mechanism based purifier model behind. More practical and simplified approaches are expected to deal with the complicated mechanism models. This paper improves the existing graphical approach in two aspects: incorporate algebraic equations into graphical methods to find the initial location of purifiers with complicated models; develop a novel pinch sliding approach to simplify the shifting procedure of composite curves. In this method, both performance based and mechanism based models of purifier can be solved. Another advantage of this targeting approach is that the flow rate of streams to the purifier also can be optimized simultaneously.Case studies indicate that this method is effective for both hydrogen and water networks.

During the last decade, the design methods of hydrogen and water networks have been improved greatly. Since the resulting network structure featuring minimum utility consumption is not unique, other properties such as disturbance resistance have drawn more and more attention. In this article, a novel concept, Mixing Potential, is proposed to improve the disturbance resistance ability of the networks in the design stage. This concept originates from measuring the concentration fluctuation of a single sink, and could be calculated by its graphical and algorithmic definition, respectively. In addition, a sufficient condition for minimizing the Mixing Potential of a single sink has been proved. Based on this sufficient condition, a graphical and its corresponding algorithmic method are proposed to design the hydrogen and water networks with minimum utility consumption. Literature examples illustrate that the disturbance resistance ability of the network can be improved by adjusting the satisfying order of sinks. © 2014 American Institute of Chemical Engineers AIChE J, 60: 3762–3772, 2014