Enhancing refrigeration capacity of turbo expander by means of multipoint design optimization

Abstract

Turbo expander produces cold energy in cryogenic Air Separation Unit (ASU) to meet the requirement of liquefied air fractional distillation. The study is initiated by recent commissioning operation of a single-point designed turbo expander in the 1600 Nm3/h oxygen yield ASU, where the online measured cold energy at an off-design operation point is under the requirement. To enhance the expander refrigeration capacity over a wide-spread condition and support the ASU flexible operations, a novel multipoint design optimization method is proposed. Characterization of the ASU expander condition point is done by allowing for the expander real status in the operational ASU; a 3-factors and 5-levels’ uniform design experimentation scheme is proposed and used to identify the multiple design points in the 3D space (formed by the expander inlet pressure and temperature, and shaft speed) and they are used to establish the multipoint objective function; adaptive surrogate model-based optimization algorithms are incorporated to enable the aim-oriented global optimal searching with considerably reduced CFD computations. The multipoint optimized expander impeller is manufactured and used to replace the original. In the restarted ASU, the expander operates at a new condition point but produces high-level overall performance and sufficient cold energy for liquefied air fractional distillation.