| Preview |
Abstract
Geological sequestration of carbon dioxide in deep saline aquifer is proposed
as an option for mitigating CO2 emissions into the atmosphere, but multiwell
injection into the same aquifer will cause pressurization. While many papers
have considered long term CO2 trapping, few have addressed the role of pressure
monitoring during injection as a way to ensure that the injection pressure does
not exceed a regulated value below the fracturing pressure.
This work provides models for the injection falloff response that can be used
for permanent pressure gauges installed at the injection interval for aquifer
characterization and, in particular, for CO2 leak detection. A three region
analytical composite reservoir model with sealing or constant pressure outer
boundary is used to model numerically simulated falloff data for CO2 injection
from a hypothetical 500 MW coal power plant. The analysis shows that regular
falloff tests can monitor the advance of the dry zone and provide a reasonable
estimate for the average aquifer pressure for the well injection area. The
latter can be used for material balance, and simulations show that the behavior
of the average pressure versus time is sensitive to the presence of a leak.
Furthermore, it may be possible to determine whether the leak is located in the
dry zone, two-phase zone, or the unswept brine zone.
Pressure behavior in CO2 storage aquifers has been neglected thus far in the
literature. In reality, analysis of successive pressure falloff tests easily
distinguishes over time whether a well is injecting into a limited volume
exhibiting pseudosteady state behavior, an open aquifer with constant pressure
support, or an effectively infinite aquifer. This paper spells out why pressure
monitoring makes sense during CO2 injection.
Introduction
Deep saline aquifers contain saline water, or brine, in rock pore space at
depths of several thousand feet. The brine is too saline to be used for
industrial, municipal, or agricultural purposes, and studies by NETL [2007] and
the IPCC [2007] suggest that the underground volume in deep saline aquifers is
more than sufficient for envisioned volumes of CO2 to be sequestered.
When bulk CO2 is injected into the aquifer, it accumulates around the injection
well creating a zone of high CO2 saturation taking up nearly 100% of the pore
space. This zone is called a dry zone because nearly all of the water is
displaced away from the well. Because CO2 is less dense than the brine it also
rises near the well forming a plume that spreads at the top of the aquifer.
Immediately above the aquifer is a seal that prevents the brine from flowing to
the shallower horizons, and the seal should also prevent upward migration of
the CO2. However, if the CO2 managed to find a hole in the aquifer seal, it
could rise and flow into other formations. Well above the depths envisioned for
CO2 sequestration may be fresh water aquifers suitable for drinking water,
irrigation or other uses. Clearly, injected CO2 must not leak from the aquifer
intended for storage.
Oddly, there is little evidence in the literature that current demonstration
CO2 sequestration projects have emphasized an obvious approach to monitor
injection wells using pressure measurements. This paper shows how regular
pressure falloff tests can monitor the expanding radius of the dry zone and
would readily show the presence of leaking CO2 or brine from the aquifer. The
CMG-GEM™ simulator [Nghiem, et al 2004] was used to generate simulations for
pressure and saturation behavior during bulk CO2 injection. Periodic injection
falloff tests were simulated to show what behavior can be observed over 30
years of continued injection.
|
