CO2-rock interaction during CO2-enriched brine injection into storage reservoirs using NMR integrated with fiber optic sensors

Author Identifiers

Bruno da Silva Falcão


Date of Award


Degree Type


Degree Name

Master of Engineering Science


School of Engineering

First Advisor

Stefan Iglauer

Second Advisor

Alireza Keshavarz

Third Advisor

Ahmed Yaseri


The combustion of fossil fuels has led to a global temperature increase due to the emission of carbon dioxide, which is the major contributor to global warming. The mitigation of CO2 emissions into the atmosphere is therefore critical, where new technologies need to be urgently developed and employed on a large scale. One promising method is to capture anthropogenic CO2 and store it in geological formations, e.g. depleted oil and gas reservoirs and saline aquifers. This process is referred to as carbon capture and storage (CSS).

The storage of CO2 in depleted oil and gas reservoirs that have trapped hydrocarbons for millions of years, requires the evaluation of their storage capacity as well as CO2 leakage risks to ensure long-term safety. Therefore, reservoir conditions need to be monitored, namely pressure and strain changes to avoid compromising the geo-mechanical strength of the reservoir rock, which could reactivate faults and lead to pathways for CO2 leakages. Rock strain is conventionally evaluated by linear variable differential transformers (LVDT) and electrical resistance strain gauges (ESG) that are attached to a rock sample’s surface. However, the geo-mechanical response obtained under laboratory conditions often mismatches with the response of an underground rock formation due to heterogeneity and anisotropy that are not captured by LVDT and ESG. Therefore, Fiber Bragg Grating (FBG) sensing has become a promising alternative due to its various advantages, such as resistance to electromagnetic interference, high-pressure and -temperature, small size, multiplexing capability, amongst others. In this study, FBG sensing technology was applied to measure strain changes in uniaxial compressive strength (UCS) and hydrostatic compressive tests of a standard plexiglass (PMMA) and limestone rock samples. The FBG fiber was wrapped around the samples in a helix with circumferential loops alongside the sample. To allow strain transfer, the fiber was glued onto the sample surface using Loctite Super Glue. The results showed that the FBG sensors were capable of accurately monitoring strain changes during UCS and hydrostatic compressive tests. Conventional methods were employed for comparison. The measured strain changes by FBG showed significantly higher precision and lower noise compared to the results obtained by LVDT and ESG.

Rock permeability, porosity, fluid saturation and flow characteristics are essential information required to design any CCS project, and are commonly measured by core flooding experiments using nuclear magnetic resonance (NMR). However, NMR is not capable of measuring rock deformation caused by fluid migration. Therefore, the evaluation of the rock sample surface deformation during core flooding has been investigated by associating NMR with FBG sensing technology. An optical fiber with 8 FBG sensors was attached to a Savonnières limestone sample and subsequently mounted to a GeoSpec2 (Oxford Instruments) NMR core analyzer. Brine was injected into the sample at 0.1 cc/min, whilst obtaining saturation profiles by the NMR and simultaneously measuring strain changes by the FBG. The results showed the FBG sensors to be capable of tracking the fluid migration through the sample at a high resolution, and to be in excellent agreement with the saturation profiles measured by NMR. As a result, the association of NMR and FBG has been demonstrated as a very promising tool to monitor fluid injection and strain changes during core flooding experiments.

Tests were conducted to monitor interactions between CO2 and the rock samples. The injected CO2 produces a weak carbonic acid in the presence of formation water that can react with calcite within the rock causing ion dissolution and the formation of secondary minerals. This can in turn lead to significant changes in the geo-mechanical strength of the rock, where it is extremely important for CCS projects to ensure safe long-term CO2 storage. Two sandstone rock samples were used to investigate the injection of live brine – CIPS-2 and CIPS-3. CIPS is an artificial sample consisting of quartz with cement composed by calcite. The CIPS-2 sample was used for the static aging process to evaluate the effect of the CO2 far from the injection point. The CIPS-3 sample was used for the dynamic aging to appraise the CO2 effects near the injection point. The aging process ran for approximately 40 hours, which was long enough to demonstrate that the reaction occurred in the bottom of the samples, with CIPS-3 presenting a stronger mineral dissolution. The changes in the samples internal structure were confirmed by CT scan images and NMR measurements. The FBG sensors were able to track strain changes at this specific location of the sample, allowing for quantification of changes in the geo-mechanical strength of the rock sample caused by the reaction of CO2 with calcite.

Overall, FBG sensing technology has been used in this study to accurately measure geo-mechanical rock strain properties at a significantly lower noise level and higher precision than conventional sensors. Further, the association of NMR with FBG sensing technology has been demonstrated in the study. Strain changes were measured at high accuracy during the core flooding process, allowing for tracking of fluid migration within a rock sample. In addition, the strain changes were shown to be related to the rock structure, confirmed by CT scanning images prior to and after the experiment. Possible changes to the rock structure caused by the reaction of dissolved CO2 and the rock were also tracked via FBG technology, allowing for monitoring of the reservoir’s geo-mechanical strength. This is an essential property to ensure the long-term safety of CO2 storage reservoirs.

Access Note

Access to this thesis is embargoed until Wednesday 10th March 2027.

Access to this thesis is restricted. Please see the Access Note below for access details.