Gas source attribution techniques for assessing leakage at geologic CO2 storage sites: Evaluating a CO2 and CH4 soil gas anomaly at the Cranfield CO2-EOR site

Abstract

23At the Cranfield CO 2 enhanced oil recovery (CO 2 -EOR) site, a localized area of 24 high concentrations of CO 2 (up to 44%) and CH 4 (up to 47%) in soil gas was detected 25 near a plugged and abandoned well. The complexity of attributing this anomaly, 26 especially in a CO 2 -EOR setting, underscores the need for careful attribution techniques 27 and provides rare and valuable experiential knowledge on attributing blind anomalies. 28An extensive geochemical monitoring program utilizing process-based soil gas ratios, 29 stable and radioactive isotopes of CO 2 and CH 4 , light hydrocarbon concentrations, 30 noble gases, and perfluorocarbon and sulfur hexafluoride tracers was undertaken from 31 2009 through 2014. The goals were to attribute source, assess the usefulness of 32 various attribution techniques, and begin to develop a framework for attribution in 33 complex CO 2 -EOR settings. Initial process-based assessment indicated an -exogenous‖ 34 source meaning that it was not the result of natural in-situ processes (Romanak et al., 35 2012). We report on the additional analyses used to determine the degree to which the 36 anomaly was related to CO 2 injection. This work included characterization of potential 37 non-reservoir gas sources within the overburden using mud-gas samples collected 38 during a new drill and downhole fluids collected from wells within the field. Two 39 hydrocarbon gas sources, one within the reservoir (Tuscaloosa) and one in the above-40 zone (Wilcox) were geochemically distinct. Stable carbon isotopes (δ 13 C) of CH 4 in the 41 anomaly were similar to those of the reservoir, but stable hydrogen isotopes (δD) 42 indicated that anomalous gases originate from an undetermined microbial source rather 43 than either of the subsurface gas reservoirs. Hydrocarbon geochemical parameters 44were therefore not only useful for attribution, but were also found to have a high 45 potential for leading to inaccurate conclusions because of alteration via CH 4 oxidation. 46Noble gases and introduced tracers proved least effective for attribution in this case. 47 followed for site selection, risk management, and environmental protection. However, 53 any indication of potential leakage at GS sites will need careful assessment. Indications 54 of potential leakage may range from soil gas anomalies located by reconnaissance 55 monitoring to public complaints about visual land changes, to environmental damage. 56Once an anomaly has been identified, source attribution will be critical to determining 57 the need for next steps such as establishing liability, quantifying emissions for credit 58 accounting, or instigating remediation activities (e.g. Dixon and Romanak, 2015). 59Accurate gas source attribution techniques will clearly be required independent of the 60 potential for GS sites to leak. 61Whereas source attribution is critical to identifying whether an anomaly 62 represents leakage, a lack of anomalies at GS sites means that few empirical 63 opportunities for developing attribution techniques exist. ...

Light hydrocarbon and noble gas migration as an analogue for potential CO2 leakage: numerical simulations and field data from three hydrocarbon systems

“…The concentrations of dissolved hydrocarbons were below 0.04 ppm in all groundwater samples. In April 2012, two wells completed in the overburden Wilcox Formation were sampled for gas at the wellhead and analyzed at Isotech Laboratories (Champaign, Illinois) . In December 2009, nine wells from the Tuscaloosa reservoir were sampled for gas and analyzed at University of Texas at Austin laboratories .…”

“…Thus, our goals in this investigation were threefold: (i) to use evidence from natural systems to define the dominant earth processes that cause geochemical changes to CO 2 , light hydrocarbons and noble gases during migration, (ii) to develop numerical solutions to define those changes, and (iii) to use these solutions to assess more accurately whether a geochemical signature observed at ground surface represents leakage (and if so, then the volume leaked) from a storage formation. We consider preparation for such signal attribution to be a skill that will be needed increasingly to meet regulatory and public acceptance expectations, in that any incidents or allegations of surface leakage will need to be addressed …”

“…We consider preparation for such signal attribution to be a skill that will be needed increasingly to meet regulatory and public acceptance expectations, in that any incidents or allegations of surface leakage will need to be addressed. [5][6][7] Our approach for the first goal was to measure vertical profiles of light hydrocarbons and noble gases at three CO 2 EOR sites. We then attempted to constrain the degree to which these intermediate gases may represent migration from the reservoir to shallower intervals and interaction with the natural system.…”

Impacts of Methane on Carbon Dioxide Storage in Brine Formations

“…Each of these candidate reservoirs may contain elevated levels of potent greenhouse gases such as methane and higher order aliphatic hydrocarbons (e.g., ethane, propane), which necessitates the need to evaluate and quantify the dynamics of hydrocarbon gas, CO 2 , and brine interactions in the subsurface during CO 2 injection. Brine‐saturated formations offer the largest storage volumes in the context of GCS, however brines are commonly associated with near saturation levels of methane (Darrah et al , a, 2015b; Anderson et al ; Gilfillan et al ; Harkness et al ). Thus, it is not surprising that several important GCS pilot projects involve CO 2 injection into brine‐saturated formations that already contain different dissolved gases, particularly CH 4 and CO 2 (Van der Meer et al ; Hosseini et al , ; Hovorka et al ).…”

“…As an example, we consider the large‐scale pilot test conducted as part of the Southeast Regional Sequestration Partnership (SECARB) near Cranfield, Mississippi, in 2009 to 2010, located in a brine‐saturated formation down‐dip of an oil field, where CO 2 was injected for a carbon storage test project (Figure ). The formation brine was saturated mainly with CH 4 and CO 2 (see Lu et al a, b; Verma et al ; Anderson et al ). There are well‐known effects of residual CH 4 on the fate and transport of injected CO 2 .…”