THERMODYNAMIC ANALYSIS OF AMMONIA-WATER-CARBON DIOXIDE MIXTURES FOR DESIGNING NEW POWER GENERATION CYCLES

Gupta, Ashish

University of Buffalo (United States)

关键词: Power Generation;    Carbon Dioxide;    37 Inorganic, Organic, Physical And Analytical Chemistry;    Enthalpy;    Ammonia;   

DOI  :  10.2172/836708
RP-ID  :  NONE
RP-ID  :  FG26-98FT40109
RP-ID  :  836708

美国|英语

来源: UNT Digital Library

【 摘 要 】

This project was undertaken with the goal of developing a computational package for the thermodynamic properties of ammonia-water-carbon dioxide mixtures at elevated temperature and pressure conditions. This objective was accomplished by modifying an existing set of empirical equations of state for ammonia-water mixtures. This involved using the Wagner equation of state for the gas phase properties of carbon dioxide. In the liquid phase, Pitzer's ionic model was used. The implementation of this approach in the form of a computation package that can be used for the optimization of power cycles required additional code development. In particular, this thermodynamic model consisted of a large set of non-linear equations. Consequently, in the interest of computational speed and robustness that is required when applied to optimization problems, analytic gradients were incorporated in the Newton solver routines. The equations were then implemented using a stream property predictor to make initial guesses of the composition, temperature, pressure, enthalpy, entropy, etc. near a known state. The predictor's validity is then tested upon the convergence of an iteration. It proved difficult to obtain experimental data from the literature that could be used to test the accuracy of the new thermodynamic property package, and this remains a critical need for future efforts in the area. It was possible, however, to assess the feasibility of using this complicated property prediction package for power cycle design and optimization. Such feasibility was first demonstrated by modification of our Kalina cycle optimization code to use the package with either a deterministic optimizer, MINOS, or a stochastic optimizer using differential evolution, a genetic-algorithm-based technique. Beyond this feasibility demonstration, a new approach to the design and optimization of power cycles was developed using a graph theoretic approach.