ONLINE FIXED TANK VOLATILE ORGANIC CONTENT QUANTIFICATION USING MACHINE LEARNING

Abstract

Tank emission calculation is typically performed offline, and manual data gathering from various tanks could be a timely and inefficient task. This paper presents iCON’s first principal process simulation software to perform tank emission quantification by leveraging its extensive database of all process simulation models and chemical components’ physical properties. Hence, a lot of operation cost is saved by eliminating the need for inventory sampling activity for each tank. This paper also will discuss online tank emission work processes to make them sustainable. iCON Tank Emission complies with the United States Environmental Protection Agency (USEPA) Compilation of Air Pollutant Emission Factors (AP-42) Fifth Edition Chapter 7 Liquid Storage Tanks that covers fixed roof tanks (vertical and horizontal) and floating roof tanks (internal, external, and domed) with a respective combination of the roof, shell, rim seal, fittings and breather vent settings. It is also linked to the monthly Malaysian Climate Condition covering domestic municipals’ temperature (minimum and maximum), Solar Irradiance Factor, and wind speed that improve the tank emission calculation accuracy. iCON Tank Emission lets users calculate thousands of tank emissions simultaneously by using iCON Machine Learning feature and displays the monthly and yearly summary reports for report standardisation among multiple business units. The simulation convergence time is instantaneous, making it time efficient for HSE engineers to perform strict monthly calculations and reporting online. Case studies and whatif scenarios could also be run to reduce tank emission via the introduction of vapour recovery and optimising tank mechanical parameters, i.e., type of rim seal, colour, and fittings selection. iCON Online Tank Emission links the iCON model with real-time plant information data for real-time tank emission calculation, transparent to management and regulators. The findings from this compositional first principle thermodynamic base tank emission simulation study are considered more efficient, faster, and cheaper than the conventional method.