Integrated Data (i-Data), Mining and Utilization Approach for Effective Flare Management Strategies

Abstract

Upset emissions occur during plant startup, shutdown, maintenance, malfunction, and flaring incidents. A wide range of these upsets cannot be managed by standalone control systems; plant personnel intervention is necessary sometimes. The methods needed to assist plant personnel to control and prevent abnormal process operations are gathered under abnormal situation management. Abnormal operations that lead to flare have significant economic, environmental, and safety impacts. Flaring is necessary for managing process upsets, however, it leads to the emission of greenhouse gases (GHG) and volatile organic compounds (VOCs), causing negative social impacts and local transient air pollution. In addition, excessive flaring results in energy and raw material losses. These are valuable commodities that must be sustained. Therefore, flare minimization during normal and abnormal operational situations has great environmental, industrial, and societal benefits. It is not possible to quantify the impacts without understanding the properties and magnitude of these upsets. Such analysis requires extensive amount of historical data. There are large sets of design, operational, and flaring data readily available; however, the challenge when it comes to flare mitigation is in using them effectively and in a timely manner. In this Article, a systematic approach to collect, analyze and utilize historical flaring data based on current industrial practices is presented. An ethylene base case study along with its historical process and flaring incidents data is used to demonstrate the significance of using and integrating data within developed flare management strategies. In the presented case, design and historical process data are used to assess the environmental impacts of abnormal incidents and to identify underlying causes and indicators that lead to process upset, that is, abnormal situations. The data sets are utilized within an optimization algorithm to identify design alternatives to mitigate process incidents and reduce its root causes. The paper highlights the challenges that are faced by environmental agencies in terms of data utilization and documentation.