Introduction

A key determinant for CO₂ storage in deep saline formations (DSFs) is dynamic efficiency (E factor) – that is the effect that increased pressure caused by fluid injection has on the storage capacity of a formation.  The storage capacity will always be limited by the pressure limit imposed by the geomechanical strength of the caprock, which is defined as the fracture pressure.  If a formation is bounded by faults or other low permeability barriers, then excess pressure could limit the dynamic efficiency, a condition referred to as a closed boundary.  In contrast formations that extend over several 100 square kilometres without significant barriers can enable pressure to be dissipated, a condition known as an open boundary.  In a previous study commissioned by IEAGHG the effects of dynamic efficiency were compared between two contrasting onshore basins (one open and the other closed), but over a long hypothetical time-scale of 2,000 years.  Although the previous study showed the effects of boundary conditions, the dynamic efficiency was based on very large areas extending of several thousands of square kilometres.  The results did not reflect the more likely conditions of much shorter timescales and injection over limited areas that would be experienced in early CO₂ storage sites.  The aim of this second study is to improve the estimated dynamic storage of DSFs based on a modelled 50 year injection period and over comparatively limited areas of ~1,000 km².  Two well researched formations were selected: one from an onshore basin (the Minnelusa Formation in the USA) and the other form an offshore basin (the Bunter Formation in the North Sea).  This study also includes a cost development model to determine how the number of wells affects the cost-effectiveness of each storage site.

Key Messages

• The impact of water extraction on the Minnelusa over a 50 year period raised the storage efficiency from 4.7% to 5.9%.  This is equivalent to an estimated increase in storage capacity from 242 Mt to 302 Mt of CO₂.  Extending injectivity for a further 50 years would increase storage capacity to over 400 Mt of CO₂.

• The impact of water extraction on the Bunter was profound raising storage efficiency from 4.7% to 7.4%.  This is equivalent to raising the estimated storage capacity from 1,770 Mt to 2,806 Mt of CO₂.  The difference between these two formations in terms of storage capacity can be attributed to the highly favourable permeability across the Bunter compared with the Minnelusa.

• As the number of injection wells increases in a designated storage system, more of the wells become influenced by pressure interference from their neighbours and the injectivity rate per well declines. 

• The closer a DSF approaches full development, the more its efficiency approaches that of a closed system, even if it has open boundaries. 

• The differences between open and closed boundaries clearly signifies the importance of defining or conducting a careful preliminary assessment of boundary conditions.

• Well configuration and structural settings can have a significant influence on storage efficiency.

• The annual injection rate profile is a critical parameter in the design of an optimised injection plan for a multiwell project.  The rate of injection will gradually decline with time.

• In both cases 20% of all the wells in the cost model were able to deliver more than 60% of the total CO₂ injected.  In both modelled formations the number of wells was the primary variable in determining the cost factor.  Delivering the amount of injected CO₂ by increasing the number of wells becomes proportionately less cost-effective.

• The E factor only applies to the modelled areas, as in these cases, and cannot be extended to the full aquifer unless the model boundaries are coincident with the periphery of the formation.

• There are variations in modelled predictions based on the model grid cell size for the same level of salinity which is a significant parameter that controls CO₂ solubility.

• Heterogeneity and different model projections can substantially influence the quantity of injected CO₂.  It is important to understand and separate the effects of the choices of simulation parameters from the physical effects in a storage formation.

It is recommended that key parameters used for initial dynamic storage estimates are clearly stated and should include: domain dimensions, formation boundaries, caprock threshold limits and the duration of injection.