This study aims at high-fidelity modeling and mechanistic optimization of gas dehydration and NGL (Natural Gas Liquids) systems of a commercial natural gas plant based in Abu Dhabi, UAE. Scope of the work includes development of models, validation of models with plant data, optimization analysis and real-time validation at the plant site.
In this work, we developed a dynamic model for the gas dehydration system and a steady state model for the natural gas liquids recovery unit. An advanced process simulator that follows equation-oriented approach is employed as the modelling and optimization platform. We first show the comprehensive plant data reconciliation followed by the model validation using the operating data of the years 2016 and 2018, to ensure that the model predictions match the real plant operation. We then present how the mechanistic optimization entity result in the best operating conditions for the natural gas liquids recovery system. We also show the optimization analysis that aims at maximizing the adsorption cycle time for the dehydration unit while minimizing the total heating duty required for the regeneration of the molecular sieve beds.
Optimization analysis reveals a significant increase in the annual net revenue of natural gas liquid recovery unit as a result of modifying various process operating conditions that lead to higher liquid hydrocarbon production and lower operating costs related to steam and refrigeration. Similarly, optimization analysis of the dehydration system indicates that adsorption-step time can be increased to a higher value, which results in significant reduction of regeneration costs.
As a next step, we aim to carry out the validation tests on the plant site to verify and implement the model recommendations in the real plant to verify the model recommendations. We also plan to derive the set of operating guidelines that allow the operators to drive the plant towards optimal operation.
To the best knowledge of authors, this study is the first effort to build a holistic model comprising of the dynamic dehydration model and steady state NGL model on a common platform.
Rigorous validation of the model is performed using real plant data of two calendar years.
Scope of this study also includes real time validation of the model recommendations through tests performed at plant site.