Enabling cost effective mean of reducing pollutant emissions from gas turbines

碳氢化合物燃料是发电的主要能源。燃烧过程排放的污染物对环境和人类健康都有影响。为了限制这些影响，世界范围内实施了严格的立法，对燃烧系统的设计提出了相互矛盾的要求。本提案的目的是开发一种新的方法，用于湍流燃烧环境下的排放预测，并应用于工业燃气轮机。

Hydrocarbon fuels are the main energy source for power generation. Pollutant emissions from combustion processes have environmental and human health impacts. Stringent legislation enforced worldwide to limit these impacts imposes conflicting requirements on the design of combustion system. The aim of this proposal is to develop a novel methodology for emissions predictions in turbulent combustion environment with application to industrial gas turbines. Detailed simulation of the combustion process still exceeds computational capabilities. The influence of the uncertainties in the model assumptions, in their parameters, and in the definitions of the boundary conditions further complicates the problem. As a consequence, reliance on costly experimental testing is the only alternative for the development of gas turbine combustion systems. The theory of uncertainty quantification (UQ) has been in intense development in the last few decades and has found applications in many scientific and engineering disciplines. UQ provides a rigorous methodology to drive improvements in all aspect of numerical modeling analysis. This paradigm shift provides a new framework to systematically study otherwise undetermined problems. The objective of this Discovery Grant is to build an UQ framework that will help to address the limitations of current emissions prediction methods. The long term goal of our research is to develop predictive capability, with low level of uncertainty, for the all combustion processes of importance for gas turbines. Three integrated projects form the core of the proposal; reduction uncertainty of natural gas chemical kinetics mechanism including NOx chemistry, development of novel turbulence chemistry interaction model for use in chemical reactor networks (CRN) simulations, and development of a consistent CRN methodology for emissions predictions with quantified uncertainty. UQ is a step change in computational modeling that brings together scientific investigation, model validation, and engineering design into a common framework that will enable better decision making. HQP trained with this skill will be in high demand in all branches of engineering and science. This research program will benefit gas turbines manufacturer present in Canada and therefore to the Canadian economy.