A searchable database for prediction of emission compositions from upstream oil and gas sources

Abstract

Atmospheric emissions from oil and gas production operations are composed of multiple hydrocarbons and can have large variations in composition. Accurate estimates of emission compositions are needed to estimate the fate and impacts of emissions and to attribute emissions to sources. This work presents a database, constructed with empirical data and thermodynamic models, that can be queried to estimate hydrocarbon compositions from emission sources present at oil and gas production sites. The database can be searched for matches using between two and seven well parameters as query inputs (gas-To-oil ratio, API gravity, separator pressure, separator temperature, methane molar fraction in produced gas, ethane molar fraction of produced gas, and propane molar fraction in produced gas). Database query performance was characterized by comparing returns from database queries to a test data set. Application of the database to well parameters for tens of thousands of wells in the Barnett, Eagle Ford, and Fayetteville production regions demonstrates variations in emission compositions. Ethane to methane ratio varies by more than an order of magnitude from well to well and source to source. VOC to methane ratios are comparable in variability to ethane to methane ratios for most emission sources, but have a higher variability for emissions from flashing of liquid hydrocarbon tanks.