EKF and UKF for estimating the liquid level inside gas-liquid cylindrical cyclone

Abstract

The oil and gas industry is focussing more on subsea separation and processing and the need for compact equipment such as gas-liquid cylindrical cyclones (GLCC) is increasing. As compactness of the equipment increases, they become more sensitive to variations in the input conditions and the dynamics become faster. Improving the control aspect is one of the solutions to ensure efficient operation of these devices during all field conditions. However, the measurements available are limited due to higher instrument cost or the absence of proper sensors. Virtual sensors or estimators is an alternative technology which could solve the problem with lower investment cost. In the case of gas-liquid cylindrical cyclones, there is no proper sensor to measure the level inside it. Therefore, this paper focuses on estimating the level inside a GLCC using two well-know nonlinear estimation techniques, the EKF and the UKF. The estimates are given as a state feedback to a level controller.