A new diagnostics tool for water injected gas turbines - emissions monitoring and modeling

Natural gas-fired cogeneration systems are commonly used for large-scale industrial energy production – both electricity generation and heat recovery. Industrial cogeneration currently represents about 8% of the U.S. total electricity generation capacity. Plans call for cogeneration to increase to 20% of the generation capacity by the end of 2030 [1, 2]. Industrial cogeneration systems attain both high thermal efficiency and low emissions. The attainment of low emissions from natural gas fired turbines, in particular low NOx emissions, is of considerable environmental importance especially as coal becomes a less favorable fuel source. Our current project addresses emissions and performance modeling of the 20 MW natural gas-fired cogeneration system located at Louisiana State University. Water injection is used to help lower emissions. Data reconciliation and gross error detection are performed to adjust measured variables and determine efficiency. A continuous emission monitoring system (CEMS) has been recently installed to measure both the NOx and O2 concentrations in the exhaust; CO is also measured. These concentrations have been used to validate an emissions kinetics model, based on GRI-Mech 3.0, in order to predict NOx, CO and O2 concentrations leaving the system. The kinetics model is used within a chemical reactor network consisting of perfectly stirred reactors and plug flow reactors to represent the turbine combustion in both the primary and dilution zones. Changes in the measured emissions of certain species combined with a detailed kinetics model are used to indicate the onset of problems in the cogeneration system.

Name thesis:degree_name

Doctor of Philosophy (PhD)

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Chemical Engineering

Grantor

Chemical Engineering

Year dc:date.available

2013

Author dc:creator

• Syed, Mohammed Shafi

• gas turbines
• data reconciliation
• emissions
• modeling

Statement dc:rights

• unrestricted
• Release the entire work immediately for access worldwide.

Identifier

etd-06062013-010835
https://repository.lsu.edu/gradschool_dissertations/994

OAI identifier oai:identifier

oai:repository.lsu.edu:gradschool_dissertations-1993