(547e) Multiresolution Modeling and Optimization of a Natural Gas Liquefaction Process Using Detailed Spiral-Wound Heat Exchanger Models

Tsay, C. - Presenter, Imperial College London
Pattison, R., The University of Texas at Austin
Baldea, M., The University of Texas at Austin
Liquefaction processing makes up around 52% of the cost of liquefied natural gas, and many research efforts have centered around improving the energy efficiency of the process [1]. Spiral-wound multistream heat exchangers, allowing for thermal contact between a refrigerant evaporating on the shell side and one or more hot streams circulating upwards on the tube side [2], are commonly selected for this process and are essential for tight energy integration.

The mathematical modeling of multistream heat exchangers (MHEXs) requires accurately predicting phase transitions and the corresponding changes in physical properties, and the MHEX models available in commonly used process simulators are typically limited to solving a set of energy balance equations with the purpose of determining the outlet conditions of one streams (all other inlet and outlet parameters must be specified). The heat duty is then usually discretized into finite enthalpy intervals and a stream sorting algorithm is used to establish the structure of the energy balance equations for the exchanger. While these formulations are acceptable in sequential-modular flowsheet simulation software, they pose considerable difficulties for equation-oriented process optimization.

Various works have shown how careful process design and optimization can significantly improve the economic performance of natural gas liquefaction, with several proposed approaches for multistream heat exchanger modeling and optimization [2] [3] [4]. In our previous work, we demonstrated a pseudo-transient process modeling approach for seamlessly incorporating unconventional or detailed mathematical models into process flowsheets [6] [7]. Furthermore, we demonstrated that describing different unit operations at different resolutions and length scales can be employed to increase the design degrees of freedom and to simultaneously identify the optimal (detailed) design of the respective units.

In this work, we apply our multiresolution flowsheeting approach to optimizing processes with a detailed spiral-wound MHEX model, capturing both thermal performance and the detailed geometry of the device. We employ semi-empirical correlations for pressure losses and heat transfer within the MHEX and use pseudo-transient modeling concepts to develop a robust simulation and optimization approach for the resulting heat exchanger models. We demonstrate the methodology by optimizing, at the flowsheet level, the PRICO® natural gas liquefaction process and comparing the results to previous solutions in literature.


[1] MIT Energy Initiative. The Future of Natural Gas: An Interdisciplinary MIT Study. Cambridge, MA: Massachusetts Institute of Technology, 2011.

[2] G Stephenson and L Wang. Dynamic simulation of liquefied natural gas processes. Hydrocarbon Process Liquefied Nat Gas Dev, 37-44, 2010.

[3] RS Kamath, LT Biegler, and IE Grossmann. Modeling multistream heat exchangers with and without phase changes for simultaneous optimization and heat integration. AIChE Journal, 58(1):190-204, 2012.

[4] MA Duran and IE Grossmann. Simultaneous optimization and heat integration of chemical processes. AIChE Journal, 32(1):123-138, 1986.

[5] RC Pattison and M Baldea. Multistream heat exchangers: equation-oriented modeling and flowsheet optimization. AIChE Journal, 61(6):1856-1866, 2015.

[6] RC Pattison, C Tsay, and M Baldea. Pseudo-transient models for multiscale, multiresolution simulation and optimization of intensified reaction/separation/recycle processes: Framework and a dimethyl ether production case study. Comput. Chem. Eng., 2017. doi: 10.1016/j.compchemeng.2016.12.019

[7] C Tsay, RC Pattison, and M Baldea. Equation-oriented simulation and optimization of process flowsheets incorporating detailed spiral-wound multistream heat exchanger models. AIChE Journal, 2017. doi: 10.1002/aic.15705