Abstract

Fluids are injected in subsurface permeable formations for various purposes including waste disposal, gas storage, CO2 sequestration, and enhanced oil/gas recovery. Containment of the injected fluids is needed to meet the regulatory requirements and/or to ensure efficiency of the intended processes. The injected fluids can leak to overlying formations in presence of leakage pathways. Improperly plugged and abandoned (P&A) wells are considered as the main potential leakage pathways. In a previous work, we introduced a vertical pressure transient interference test and presented an analysis methodology to detect and characterize leaking wells. The analysis methodology was based on an inverse modeling algorithm that can be highly instable and computationally expensive. Here, we propose an easy-to-use fully graphical methodology to characterize leaking wells. The pressure measurements are graphed in three different forms. The slopes and intercepts of the line-fitted graphs are used to determine the leak location and transmissibility as well as the transmissivity ratio of the connected zones. The graphical method is applied to an example problem to illustrate its application procedure and effectiveness.

Introduction

The Leakage through abandoned wells and improperly plugged boreholes can create vertical communication between otherwise hydrologically isolated permeable zones. The driving mechanism behind the leakage can be the hydraulic gradients created by injection into one of the zones. Zeidouni and Pooladi-Darvish (2012a, 2012b) introduced a vertical interference test to detect and characterize a leaking well connecting the operating zone to an overlying non-operating zone which is otherwise separated by a sealing caprock. The test involves injection (production) into (from) the operating zone (OZ) and observing the pressure at a distance both in the OZ and the monitoring zone (MZ). We use injection throughout this paper for consistency. Several researchers attempted to analyze the pressure observations through inverse modelling approach and data assimilation (Wang and Small 2014, Jung, Zhou, and Birkholzer 2013, Sun et al. 2013, Zeidouni and Pooladi-Darvish 2012a, b, Chabora and Benson 2009, Jung, Zhou, and Birkholzer 2015, Keating et al. 2014). While inverse modeling can be very useful, it requires robust and computationally expensive inversion techniques that may not be easy to implement in practice. Also, inverse models can be very instable if the unknown parameters are not fully independent. It would be useful to develop graphical approaches such as those used in conventional pressure transient analysis that can be conveniently used in analyzing the pressured data for leaking well characterization.

This content is only available via PDF.
You can access this article if you purchase or spend a download.