HOT AIR RECIRCULATION (HAR) DIGITAL 1.0 – IMPROVING EFFICIENCY OF LNG PLANTS USING AI AND IOT

Abstract

PETRONAS (PETRONAS Research, Group Technical Solution and LNG Plant), in collaboration with JGC Corporation (Japan), developed a predictive algorithm-based application named HAR Digital 1.0 that utilises Artificial Intelligence (AI) and Internet of Things (IoT) technologies. The system provides real-time weather data prediction to forecast power margin setting for gas turbine and allow the plant to maintain stable operation with the opportunity to increase LNG production. Hot air recirculation (HAR) is a phenomenon that affects air-cooled LNG plants due to its geographical location, in which the hot air from air-cooled heat exchangers (ACHE) flows back into the ACHE intake or the intake of other equipment such as the gas turbine (GT). This reduces the turbine helper motor power margin availability, hence forces higher power requirements to the gas turbines, making the operation energy inefficient, exposes them to instability, and subsequently may cause plant trip. To date, there is no accurate method to predict the occurrence of HAR that could allow the plant’s advanced process control (APC) to adjust the power margin setting to optimise GT operation. This innovation uses Real-Time Qualified Plant Data and Predictive Data Analytics to provide Forecasted Temperature and Power Margin Set Limit to allow for early ACHE/GT temperature prediction. This will assist APC in power margin setting, which allows the plant to use any available power margin to boost LNG production safely and reduces process fluctuation. A solid and compelling economic case is confirmed from the economic evaluation conducted based on the new power margin value. The production increase is estimated to translate to more than MYR 12 million per annum for two LNG trains.
 Keywords: Hot air recirculation, LNG Production