Phenomena exposure from the large scale gas injection test (Lasgit) dataset using a bespoke data analysis toolkit

Abstract

AbstractThe Large Scale Gas Injection Test (Lasgit) is a field-scale experiment designed to study the impact of gas buildup and subsequent migration through an engineered barrier system. Lasgit has a substantial experimental dataset containing in excess of 21 million datum points. The dataset is anticipated to contain a wealth of information, ranging from long-term trends and system behaviours to small-scale or ‘second-order’ features. In order to interrogate the Lasgit dataset, a bespoke computational toolkit, designed to expose difficult to observe phenomena, has been developed and applied to the dataset. The preliminary application of the toolkit, presented here, has resulted in a large number of phenomena being indicated/quantified, including highlighting of second-order events (small gas flows, perturbations in stress/pore-water sensors, etc.) and quantification of temperature record frequency content. Localized system behaviour has been shown to occur along with systematic aberrant behaviours that remain unexplained.