Modeling of biogeochemical consequences of a CO₂ leak in the water column with bottom anoxia

Highlights

• To study the biogeochemical consequences of a release of CO₂ in an anoxic marine environment a FABM family set of models consisting of a transport model, biogeochemical model (including carbonate system processes block) and bubble fate (transport and dissolution) was applied.

• The measurements performed during a controlled 2-h long CO₂ release experiments show elevated levels of pCO₂ and simultaneously decreased values in pH, that was used for the model validation.

• The model analysis of consequences of a CO₂discharge demonstrates that CO₂ bubbles are dissolved shortly after termination of the leak, while changes in pH, pCO₂ and TIC can be detected for several days after the leak, but only at a limited distance from the source (< 10 m in the examples evaluated here).

Abstract

In this paper we investigate the spatial extent and biogeochemical properties of a known CO₂ plume using the pelagic transport-biogeochemical model BROM (Bottom RedOx Model). The model consists of a biogeochemical module, a 2-dimensional vertical transport module and gas bubble fate module, parameterizing bubbles rising and dissolution according to existing approaches. A controlled CO₂ release experiment was carried out in the Horten Inner Harbor, Norway, in September 2018. This isolated bay is characterized by limited water mixing and anoxia in the bottom layer. CO₂ was released at a water depth of 18 m either in a gas phase or dissolved in seawater at leak rates ranging from 0.1 l/min to 15.8 l/min. The chemical response to the release events relative to background variations was measured using chemical sensors mounted on two seabed templates located 4 m and 15 m from the release point, respectively, and compared to the values predicted by the model. The measurements show elevated levels of pCO₂ and simultaneously decreased values in pH corresponding to the controlled release experiments. The model's simulations were in good agreement with the baseline observations and the measured changes forced by the experimental leak. The model predicts that after a continuous leak of this magnitude in stagnant conditions of anoxic bottom water, a 2–3 weeks long restoration period occurs, after which the disturbances disappear. This work confirms that the footprint of a potential CO₂ leak is localized in the vicinity of the source (tens of meters) where it can be detectable with available chemical sensors.

Introduction

Carbon capture and storage (CCS) has emerged as a promising technology for reducing greenhouse gas emissions and reaching international climate goals. During and after injection into an offshore geological reservoir, monitoring is required to verify that the CO₂ stays in the reservoir as intended, and to ensure that there is no negative impact on the surrounding environment (Blackford, 2014). While seismic methods are used to monitor the geological reservoir and the overburden, marine monitoring is intended to verify that there are no indications of leakage at the seabed or in the water column. Anomalies potentially related to leakage include changes in the chemical composition of the sediment pore water and in the water column, the appearance of seep-related features on the seafloor such as bacterial mats or pockmarks, microbubbles in the sediment, and bubble trains in the water column. A comprehensive CCS monitoring program should incorporate multiple technologies to target the different ways in which a leak-related anomaly may occur (Blackford, 2020).

Several recent and ongoing research projects contribute to knowledge building and maturing of marine monitoring strategies. The ECO2 project built a framework for environmental practices related to CCS (http://www.eco2-project.eu/). As part of the QICS project (Blackford, 2015), controlled releases of CO₂ made it possible to understand and document the environmental impact of CO₂ leakage, and to evaluate the detectability of leaks. Lately, the STEMM-CCS project (STEMM-CCS, 2019) covers both the marine environment and reservoir overburden sediment. Results include modeling of natural fluctuations in the CO₂ concentration in seawater, which can be several orders of magnitude, both spatially and temporally and modeling of leakage scenarios and subsequent optimization of sensor locations and movement patterns for mobile sensor platforms (Blackford et al., 2017). The results allowed the influence of potential leaks on the pH distribution in the vicinity of the leak to be modelled (Vielstädte, 2019). A comprehensive controlled CO₂ release experiment was also included in the ACT4storage (Acoustic and chemical technologies for environmental monitoring of geological carbon storage) project, where the focus was on evaluating the capabilities and limitations of different acoustic and chemical sensor technologies for marine monitoring of offshore CO₂ storage sites.

While it is considered unlikely that sequestered CO₂ will migrate from a geological reservoir and reach the seabed, it is important to study the environmental impact that a hypothetical leak might have and to have tools and solutions in place to detect a leak if it should occur. A potential CO₂ leak may be accompanied or preceded by a release of pore water from the sediments. The evacuated pore water will have different characteristics compared to the bottom sea water, such as lower levels of dissolved oxygen, variations in salinity, etc. We can hypothesize that a leak will lead to anoxic conditions at the sediment-water interface, or potentially to formation of a bottom anoxic layer. In this work, we combine experimental data from a controlled CO₂ release experiment with a biogeochemical model to better understand the chemical footprint of a hypothetical leak under anoxic conditions. Anoxic bottom conditions may either be formed during a leakage or be present prior to the leakage event. In this research, we analyze the potential changes in the carbonate system that can be affected by anoxic conditions, from point of view of their detectability with existing sensors. Such data is novel supplementary to the leak experiments conducted in normoxic conditions, i.e. STEMM-CCS. The controlled release experiment was carried out in a sheltered environment in the Oslo Fjord with bottom anoxia (the Horten Inner Harbor, Fig. 1) as part of the ACT4storage project. In the event of unintended CO₂ migration from the reservoir reaching the seabed, the CO₂ may enter the water column in gas phase or dissolved in seawater, or both. To mimic this situation, the controlled release experiment included releasing both gaseous CO₂ and CO₂ dissolved in seawater. During the 6-week period of the controlled release experiment, acoustic and chemical sensors mounted on seabed templates were programmed to continuously record data, capturing natural variability as well as simulated CO₂ leak events. The goals of this experiment were to better understand the effects of a potential CO₂ leak by quantifying background variations in relevant chemical and oceanographical parameters (CO₂, O₂, pH, salinity, temperature, and ocean currents), and to evaluate different technologies' ability to detect the changes in parameters connected with the simulated CO₂ leak. A biogeochemical model was used to analyze biogeochemical consequences of a CO₂ leak in the water column with anoxia. The simulations were in good agreement with the baseline observations and measured changes forced by the experimental leak and could be used as a reliable instrument to simulate natural variability and predict the effects of a potential CO₂ leak.