Fault and natural fracture control on upward fluid migration: insights from a shale gas play in the St. Lawrence Platform, Canada

Abstract

Environmental concerns have been raised with respect to shale gas exploration and production, especially in eastern Canada and northeastern United States. One of the major public concerns has been the contamination of freshwater resources. This paper focuses on the investigation of possible fluid upward migration through structural features in the intermediate zone (IZ), located between a deep shale-gas reservoir and shallow aquifers. The approach provides insights into how such an investigation can be done when few data are available at depth. The study area is located in the shale-dominated succession of the St. Lawrence Platform (eastern Canada), where the Utica Shale was explored for natural gas between 2006 and 2010. Detailed analyses were carried out on both shallow and deep geophysical log datasets providing the structural attributes and preliminary estimates of the hydraulic properties of faults and fractures. Results show that the active groundwater flow zone is located within the upper 60 m of bedrock, where fractures are well interconnected. Fractures from one set were found to be frequently open in the IZ and reservoir, providing a poorly connected network. The fault zones are here described as combined conduit-barrier systems with sealed cores and some open fractures in the damage zones. Although no direct hydraulic data were available at depth, the possibility that the fracture network or fault zones act as large-scale flow pathways seems very unlikely. A conceptual model of the fluid flow patterns, summarizing the current understanding of the system hydrodynamics, is also presented.

Résumé

Des préoccupations environnementales ont été soulevées concernant l'exploration et la production de gaz de schiste, particulièrement dans l'est du Canada et dans le nord-est des États-Unis d’Amérique. L’une des principales préoccupations du public est la contamination des ressources en eau douce. Cet article porte sur l’étude de la possible migration ascendante de fluide à travers des éléments structuraux de la zone intermédiaire (ZI), située entre un réservoir profond de gaz de schiste et un aquifère peu profond. Les résultats présentés fournissent des éléments de réflexion sur la façon dont les impacts environnementaux du développement du gaz de schiste sur les aquifères peuvent être étudiés, lorsque peu de données sont disponibles en profondeur. La zone d'étude est située dans la plate-forme du Saint-Laurent (est du Canada), où le Shale d'Utica a fait l’objet de travaux d’exploration (2006 à 2010). Des analyses détaillées ont été faites à partir de données de diagraphies afin d’obtenir une estimation préliminaire des propriétés hydrauliques des failles et fractures. Les résultats montrent que la zone active d'écoulement est située dans les 60 m supérieurs du substratum rocheux, où les fractures sont bien interconnectées. Ce sont principalement les fractures d’une famille qui semblent ouvertes en profondeur. Ces dernières sont cependant mal interconnectées, défavorisant ainsi la circulation sur de grandes distances. Les zones de failles ont été conceptuellement décrites comme une combinaison de conduits/barrières aux écoulements, avec un cœur de faille scellé par de la gouge et des fractures ouvertes dans la zone de dommages. Bien qu’aucune donnée hydraulique ne soit disponible en profondeur, la possibilité que le réseau de fractures ou les zones de failles agissent comme des voies d’écoulement à grande échelle semble très improbable. Un modèle conceptuel des types d’écoulement de fluides, résumant la compréhension actuelle de ce système est également présenté.

Resumen

Se han planteado preocupaciones ambientales con respecto a la exploración y producción de shale gas, especialmente en el este de Canadá y el noreste de los Estados Unidos. Una de las principales preocupaciones públicas ha sido la contaminación de los recursos de agua dulce. Este documento se centra en la investigación de la posible migración ascendente del fluido a través de las características estructurales en la zona intermedia (IZ), ubicada entre un reservorio profundo de shale gas y acuíferos poco profundos. El enfoque proporciona información sobre cómo se puede realizar dicha investigación cuando hay pocos datos disponibles en profundidad. El área de estudio se encuentra en la sucesión dominada por las lutitas de la St. Lawrence Platform (este de Canadá), donde se exploró el Utica Shale para gas natural entre 2006 y 2010. Se realizaron análisis detallados de conjuntos de datos de perfilajes geofísicos someros y profundos aportando los atributos estructurales y las estimaciones preliminares de las propiedades hidráulicas de las fallas y fracturas. Los resultados muestran que la zona de flujo de agua subterránea activa se encuentra dentro de los 60 m superiores de la roca madre, donde las fracturas están bien interconectadas. Las fracturas de un conjunto se encontraron frecuentemente abiertas en el IZ y el reservorio, proporcionando una red mal conectada. Las zonas de falla se describen aquí como sistemas combinados de barrera de conductos con núcleos sellados y algunas fracturas abiertas en las zonas de daños. Aunque no se disponía de datos hidráulicos directos en profundidad, la posibilidad de que la red de fracturas o las zonas de fallas actúen como vías de flujo a gran escala parece muy poco probable. También se presenta un modelo conceptual de los patrones de flujo de fluidos, que resume la comprensión actual de la hidrodinámica del sistema.

摘要

人们对页岩气的勘探和开采,尤其是对加拿大东部和美国东北部页岩气的勘探和开采的环境关切日益提高。公众的主要关切之一就是淡水资源的污染。本文重点论述了液体通过位于深层页岩气储和浅层含水层之间的过渡带中的构造地貌向上运移的可能性研究成果。该方法提供了在深部缺乏数据的情况下怎样开展调查工作方面的认识。研究区位于 (加拿大东部)St. Lawrence台地上的页岩主导的交替层上,2006年到2010年在这里对Utica页岩层天然气进行了勘探。对提供断层和断裂的构造属性和水力特征的初步估算结果的浅层和深层地球物理测井数据集进行了详细的分析 。结果显示, 活跃的地下水流带位于基岩上部的 60米内,这里的断裂连通良好。发现同一层的断裂在过渡带和水储中常常是开放的,致使网络的连接性很差。这里的断层带被描述为具有密封的核心及损伤带中有开放断裂的组合通道屏障系统。尽管没有深部直接的水力数据,但断裂网络或者断层带充当大规模的水流通道的可能性似乎很小。这里还展示了流体流动模式的概念模型,该模型总结了目前对系统水动力学的认识。

Resumo

Preocupações ambientais foram levantadas com respeito a exploração e produção de gás de xisto, especialmente no leste do Canadá e nordeste dos Estados Unidos. A contaminação dos recursos hídricos é uma das maiores preocupações do público. Este artigo foca nas investigações da possível migração ascendente de fluídos através de estruturas na zona intermediária (ZI), localizada entre o reservatório de gás de xisto profundo e os aquíferos rasos. A abordagem fornece percepções sobre quais investigações poder ser feitas quando poucos dados estão disponíveis. A área de estudo é localizada na sucessão dominada por xistos na Plataforma de St. Lawrence (leste do Canadá), onde o Xisto Utica foi explorado para gás natural entre 2006 e 2010. Análise detalhada foi desenvolvida em ambos perfis de geofísica, raso e profundo, fornecendo atributos estruturais e estimativas preliminares das propriedades hidráulica de falhas e fraturas. Os resultados mostram que a zona ativa de fluxo de águas subterrâneas está localizada nos primeiros 60 m de rocha, onde as fraturas estão bem interconectadas. As fraturas de um dos dados parecem estar abertas na ZI e no reservatório, fornecendo uma rede pobremente conectada. As zonas de fraturas são descritas aqui como sistemas combinados de condutos-barreiras com testemunho selados e algumas fraturas abertas na zona de danos. Apesar de nenhum dado hidráulico estar disponível em profundidade, a possibilidade da rede de fraturas ou zonas de falhas atuarem como caminhos de fluxo de grande escala parece ser improvável. Um modelo conceitual do padrão de fluxo de fluídos, resumindo o entendimento atual do sistema hidrodinâmico, também é apresentado.

Appendices

Appendix 1

The term “apparent apertures” is used in the paper to refer to fracture apertures estimated from shallow (ATV) and deep (FMI) well logs. Fracture aperture estimations are in fact affected by the method limitations presented in Table 4.

Table 4 Methods description and limitation for fracture aperture estimations

Appendix 2

Methods used for the calculation of the normal fault core properties of the Rivière Jacques-Cartier fault zone

The Shale Gouge Ratio (SGR) method (Yielding et al. 1997; Freeman et al. 1998; Eq. 5) is based on the length of the throw along the faults (T; i.e. the vertical displacements of the stratigraphic units), the thicknesses of the stratigraphic units (Δz) and the volume of shale (V_sh) within these units to estimate the percentage of shale within a portion of the sedimentary succession that has slipped past a certain point along the fault. T and Δz of this normal fault system were identified using the cross-section presented in Fig. 1. Similar estimations were not possible (or would have been speculative at best) in the case of the thrust faults because these structures often display only shale on shale repetitions with little lithological or stratigraphic contrasts. Gamma ray logs were used to estimate V_sh (Rider 2002). Séjourné (2015) used a similar approach, employing computed gamma ray logs (called HCGR), acquired in the deep shale gas wells and in conventional wells drilled in the area in the Lorraine Group, Utica Shale, Trenton and Beekmantown groups and an empirical equation for pre-Tertiary rocks proposed by Atlas (1982). The V_sh values used in this current exercise are from this dataset: 0.44, 0.21, 0.08 and 0.20 respectively for the Lorraine Group, Utica Shale, Trenton Group and Beekmantown Group. Tightly sealed faults display high SGR values. Threshold values of 18% (Freeman et al. 1998) or 20% (Yielding et al. 1997) were proposed for sealed faults in shale/sandstone successions that display cross-fault pressure differences.

$$ \mathrm{SGR}=\frac{\sum \left({V}_{\mathrm{sh}}\cdotp \Delta z\right)}{T}\cdotp 100 $$

(5)

The Sperrevik et al. (2002) relationship (Eq. 6) is an empirical relationship developed using core samples collected in faulted clastic reservoirs in the United Kingdom and in the Sinai Desert (Knott et al. 1996). These cores include sandstones and some clay-rich units such as shales. Equation (6) is based on the clay content and of compaction and diagenesis effects which strongly impacts rock porosity and permeability. The maximum burial depth (z_max) and the depth at the time of rock deformation/faulting (z_def) were used as proxies. This is relevant for the study area as the geological units were buried under at least 5,000 m of Paleozoic strata before erosion (Héroux and Bertrand 1991; Yang and Hesse 1993). Then, z_max corresponds to the actual depth of the units plus a 5,000 m value. Because a conservative estimate was used for the z_def value, the shallower depths of the stratigraphic units at the fault footwall were used. The parameter k is the fault core permeability expressed in milliDarcies (mD: 1 mD = 10⁻¹⁵ m²), the a parameters are empirical parameters proposed in Sperrevik et al. (2002) (a₁ = 80,000; a₂ = 19.4; a₃ = 0.00403; a₄ = 0.0055; a₅ = 12.5). As proposed in the previous section, the use of k (m² or mD) instead of K (m/s) is used here because of the presence of multiple phases in the pore space (water, gas, brines) of the cover succession. For comparison purposes, the k values are converted into K using the thermo-physical properties of water at 35 °C (the temperature estimated at an arbitrary depth of 1,500 m using the mean geothermal gradient proposed in Bédard et al. (2014) for the SLP)

$$ k={a}_1\cdotp \exp -\left({a}_2\cdotp {V}_{\mathrm{sh}}+{a}_3\cdotp {z}_{\mathrm{max}}+\left({a}_3\cdotp {z}_{\mathrm{def}}-{a}_5\right){\left(1-{V}_{\mathrm{sh}}\right)}^7\right) $$

(6)

Appendix 3

The numerical values of hydraulic properties plotted in Fig. 6 are provided in Table 5.

Table 5 Estimated hydraulic properties: min, geometric mean and max for the values presented in Fig. 6

Appendix 4

Higher open fracture densities values associated with thrust planes were observed in the vicinity of the Appalachian structural front (higher values for well A). The numerical values are presented in Table 6.

Table 6 Open fracture densities for deep wells (fracture densities were calculated using a 15-m window size every 5 m)