Fault architecture and fracture network evolution (and resulting bulk hydraulic properties) are highly dependent on the mechanical properties of the rocks at the time the structures developed. This paper investigates the role of mechanical layering and pre-existing structures on the evolution of strike–slip faults and fracture networks. Detailed mapping of exceptionally well exposed fluvial–deltaic lithologies at Spireslack Surface Coal Mine, Scotland, reveals two phases of faulting with an initial sinistral and later dextral sense of shear with ongoing pre-faulting, syn-faulting, and post-faulting joint sets. We find fault zone internal structure depends on whether the fault is self-juxtaposing or cuts multiple lithologies, the presence of shale layers that promote bed-rotation and fault-core lens formation, and the orientation of joints and coal cleats at the time of faulting. During ongoing deformation, cementation of fractures is concentrated where the fracture network is most connected. This leads to the counter-intuitive result that the highest-fracture-density part of the network often has the lowest open fracture connectivity. To evaluate the final bulk hydraulic properties of a deformed rock mass, it is crucial to appreciate the relative timing of deformation events, concurrent or subsequent cementation, and the interlinked effects on overall network connectivity.
Differences in the mechanical properties (mechanical stratigraphy) of rock layers have long been recognised as influencing the style and evolution of faults (Anderson, 1951; Donath, 1961; Ranalli and Yin, 1990; Ferrill et al., 2017). However, work has tended to focus particularly on normal faults, with the effect of mechanical layers in sand–shale sequences (e.g. van der Zee and Urai, 2005; Schmatz et al., 2010), interbedded limestones and marls (e.g. Ferrill and Morris, 2003, 2008; Long and Imber, 2011; Ferrill et al., 2012), and ignimbrites (Soden and Shipton, 2013) receiving particular attention. The lithology being cut by the fault influences fault dip, e.g. strands in competent layers have steeper dips than those in incompetent layers (Ferrill and Morris, 2008), with important consequences for vein geometry and mineralisation potential (Dunham, 1948). The ratio of competent to incompetent lithologies thus affects fault style and displacement profiles (Ferrill et al., 2017; Ferrill and Morris, 2008). Fault-related folding of thin competent layers (e.g. limestones) is common in successions otherwise dominated by incompetent lithologies (e.g. shale) (Ferrill and Morris, 2008; Lăpădat et al., 2017). The presence of incompetent lithologies also restricts fault growth with strands terminating at incompetent beds and leads to formation of faults with high length to height ratios orientated parallel to the strike of bedding (e.g. Nicol et al., 1996; Soliva and Benedicto, 2005; Roche et al., 2013).
Pre-existing weaknesses (e.g., joints and faults) also play an important role in the nucleation, orientation, and length of later faults (Crider and Peacock, 2004; Peacock, 2001; Walsh et al., 2002). The mechanical response of a pre-existing joint to faulting will depend on its orientation relative to far field stress (Moir et al., 2010), the ratio of principal stresses (Lunn et al., 2008; Healy et al., 2006; Moir, 2010; Chang and Haimson, 2000; Haimson and Chang, 2000), and local variations in the stress field due to the interaction of joints in the pre-existing network (Crider and Peacock, 2004; Kattenhorn et al., 2000; Moir et al., 2010; Peacock, 2001). Where joints or cleats are orientated perpendicular to the growth direction of faults, they can act as a strength contrast and restrict fault growth (Wilkins and Gross, 2002). Alternatively, where pre-existing joints are orientated favourably, they can act as a plane of weakness and be reactivated to form faulted joints (e.g. Crider and Peacock, 2004; Cruikshank et al., 1991; Wilkins et al., 2001).
Veins are often associated with faulting, providing evidence of the paleo-fluid flow through a fracture network (Bons et al., 2012; Oliver and Bons, 2001; Peacock and Sanderson, 2018) and may act as a baffle to post-cementation basinal fluid flow (e.g. Skurtveit et al., 2015). Additionally, the strength of a rock mass can vary depending on the strength ratio between the host-rock and veins, along with the mineralogy, thickness, and orientation of veins relative to the maximum compressive stress (e.g. Shang et al., 2016; Turichshev and Hadjigeorgiou, 2016, 2017; Virgo et al., 2014). Therefore, the cementation of faults and joints can influence subsequent deformation of the rock mass (Caputo and Hancock, 1998; Holland and Urai, 2010; Ramsay, 1980; Virgo et al., 2013, 2014).
This study utilises an exceptional succession of faulted fluvial–deltaic exposures of the Limestone Coal Formation exhumed at the Spireslack Surface Coal Mine, Scotland. Coal-bearing, fluvial–deltaic sequences are commonly mechanically stratified. Fluvial–deltaic sequences are characterised by cyclical sequences of limestone, sandstone, siltstone, seat earth (paleosols that are often found beneath coal seams), shale, and coal (Thomas, 2013, and references therein). The competent lithologies in the sequence (limestone and sandstone) commonly contain joints. Coal has a distinctive blocky texture due to the presence of two roughly perpendicular fracture sets referred to as cleats (Laubach et al., 1998). Cleats form in coal beds during diagenesis and act as pre-existing weaknesses that can influence the location, orientation, and length of faults. We investigate how the internal structure of strike–slip faults at Spireslack Surface Coal Mine depends on the lithology, presence of pre-existing weaknesses (e.g., joints, cleats), and synchronous cementation. Our observations contrast small offset, self-juxtaposing faults and faults with larger offsets that cut multiple lithologies.
Spireslack Surface Coal Mine is located in the Midland Valley of Scotland, a 90 km wide, 150 km long, ENE-trending basin that opened during the Late Devonian to Early Carboniferous in response to back-arc extension within the Laurussian Plate (Leeder, 1982, 1988). This was followed by a period of thermal subsidence that continued throughout Namurian and Westphalian times, leading to the deposition and preservation of thick coal measures across much of the UK (Fig. 1a) (Leeder, 1982).
The Midland Valley is bound by two major faults: the Southern Upland Fault to the south and Highland Boundary Fault to the north (Fig. 1a) (Bluck, 1984). Carboniferous basins that have axes oblique to the main trend of the Midland Valley (e.g. Central Scottish Coalfield; Francis, 1991) can reach over 6 km in thickness (Dean et al., 2011). Faults with associated, localised folding have a complex history of reactivation caused by sinistral strike and oblique-slip movement during the Tournaisian and dextral strike and oblique-slip movement during Viséan to Westphalian times (Browne and Monro, 1987; Rippon et al., 1996; Ritchie et al., 2003; Underhill et al., 2008).
Overview photographs of Spireslack SCM:
Spireslack Surface Coal Mine (SCM), next to the now abandoned coal mining
village of Glenbuck in South Ayrshire, Scotland (Fig. 1a), provides an
exceptional exposure of Carboniferous rocks in a 1 km long residual void
(Figs. 2 and 1c). Shallow, southerly dipping (20–40
The stratigraphy is comprised of a continuous succession of Viséan to Namurian strata, including a complete section through the Limestone Coal Formation (LCF) (Fig. 1b, c) (Ellen et al., 2016, 2019). Bituminous coal is found in cyclical fluvial–deltaic sequences that outcrop across much of the dip-slope and high wall, bounded by the Upper and Lower Limestone Formations. The Lower Limestone Formation represents more marine-influenced facies, including extensive, fossil-rich limestone units (e.g. the McDonald Limestone) (Davis, 1972). The Spireslack Sandstone is exposed above the Limestone Coal Formation and is comprised of one channelised and two tabular sandstone beds (Ellen et al., 2019).
Several faults with shallow slip vectors and variably complex internal structures offset the stratigraphy. Additionally, at least five Paleogene basaltic dykes are observed trending NW–SE to WNW–ESE, which Leslie et al. (2016) suggest intrude along pre-existing faults. The rocks exposed at Spireslack SCM are part of the southern limb of the upright, WSW–ENE-trending Muirkirk Syncline that formed in response to mid- to late- Carboniferous sinistral transpression (Davis, 1972; Leslie et al., 2016). Leslie et al. (2016) attribute the faulting and folding observed at Spireslack SCM to this deformation, and have observed no evidence at the site of the later widespread dextral deformation found elsewhere in the Midland Valley (e.g. Underhill et al., 2008).
Geological mapping of the dip-slopes captured all sandstone and shale units
below the McDonald Limestones and the sandstone bed above the Muirkirk Six Foot
Coal. Mapping was undertaken at a
One way to describe the topology of a fault or fracture network is as a
series of branches and nodes (e.g. Manzocchi,
2002; Sanderson and Nixon, 2015, 2018). A branch is a fracture trace with a
node at each end. Nodes can occur where a fracture terminates into rock
(I-node), abuts against another fracture (Y-node), or crosses another
fracture (X-node). The proportion of different node types (I, Y, and X) can
then be plotted on a triangular diagram to characterise the connectivity of
the network (Manzocchi, 2002;
Sanderson and Nixon, 2015). In this work we recorded faults and fractures as
orientation sets and report fracture and branch trace length (tl), 2D fracture
intensity (I), and the percentage of connected branches (
Fault and fracture mapping were undertaken using two datasets: (i) a drone-derived photomontage of the McDonald Limestone bedding plane, provided by
Dave Healy of Aberdeen University, and (ii) an auto-rectified photomontage
of the high wall collected by the British Geological Survey. In order to
understand the geometry, topological properties and cross-cutting
relationships of fault strands and joint sets, interpretation areas were
selected from both the dip-slope and high wall for analysis. Due to its
instability, the high wall is generally unsafe to access, so any
interpretations of the high wall are made principally on the photomontage.
We outline our workflow in detail below.
The digitisation and analysis of the fault network separately from the
“joint” dataset meant that where faults terminated against pre-existing
joints (i.e. a Y-node in the combined network), this was classified as an
isolated node. This was done to provide the network properties (i.e. connectivity, trace length and fracture intensity) of the “active” fault
network where evidence of shear and mineralisation is present. Because the
mineralised fault network will be sealing to flow and therefore not
hydrologically connected to the joint network, it is not appropriate to
classify joint–fault abutting relationships as connected nodes. Therefore,
where a joint terminates against a pre-existing fault in the “joint” dataset
this was also classified as an I-node. The combined network represents the
fault–fracture network that is typically digitised and analysed for
topological analysis.
Typical fracture properties for McDonald Limestone and McDonald
Coal:
Fractures at Spireslack SCM can be classified as either joints (barren open mode fractures), faulted joints (joints that show evidence of reactivation, e.g., mineralisation or cataclasis), or shear fractures, with the latter two often found in proximity to faults. In this study “shear fractures” refer to a fracture with displacement below map scale and can be either mineralised or barren. Cross-cutting relationships are often complex and display several age sets. For example, in the McDonald Limestone bedding plane (Fig. 3a) there are two generations of joints: an early set of NE–SW trending joints (dashed black in Fig. 3a) and a later set of N–S trending joints (black) that abut the earlier set. Both generations represent the pre-existing fracture set at the time of faulting. These pre-existing joints are then cut by a set of NNE–SSW-trending mineralised shear fractures (dashed blue) that are restricted by favourably orientated joints and are locally associated with new barren shear fractures (lilac). Finally, several of the N–S-trending joints become reactivated (maroon) and are interpreted as faulted joints (cf. Zhao and Johnson, 1992).
Calcite mineralisation at Spireslack SCM (Fig. 3b, c), which is often found associated with faults, occurs as two styles: (1) amorphous, where no growth structures are present and occasional fragments of limestone are observed within the vein, or (2) with syntaxial growth textures suggesting both sinistral and dextral motion during the mineralisation of a single vein (Fig. 3c). Along fault planes and within a few metres of faults, composite veins commonly occur, with multiple growth stages and evidence of reactivation (Fig. 3c).
Fractures in the coal layers are commonly filled with a buff to orange-coloured mineral, identified in the field as ankerite (iron-rich carbonate)
(Fig. 3d–f). Fractures in coal occur as the following features.
Other lithologies observed in Spireslack SCM display a strongly developed
fracture stratigraphy (cf. Laubach et al., 2009).
For example, the McDonald Seat Earth exposed in the dip-slope towards the
west of the site (Fig. 4a) lacks a well-developed joint pattern. Instead,
shear fractures are observed in relation to small stratigraphic offset,
strike–slip faults (Fig. 5a, b). Fractures are only found in close
proximity to fault strands and are either sub-parallel to fault strands in
the hanging wall block or oblique to the fault strands in relay zones and
fault tips. Fractures commonly display small sinistral and dextral
stratigraphic offsets (mm to cm) and are typically barren, although
occasionally pyrite is found along the fracture plane. Sandstones display
bed-bound joint sets in a similar manner to the McDonald Limestone. However,
there was limited bed-parallel exposure to explore the age and orientation
of fracture sets in sandstone lithologies. In contrast to the dip-slope,
seat earth in the high wall displays a well-developed bed-bound fracture
network. This suggests that mine-related stresses may have caused
deformation of these lithologies and that the natural network has been
altered by both subsurface and surface mining activities.
Geological map of Spireslack SCM:
Like the McDonald Limestone bedding plane (Fig. 3a–c), a complex chronology of fractures can be observed in the Muirkirk Six Foot Coal (Fig. 3d–f). In Fig. 3d, dextral en-échelon vein arrays (red) cross-cut earlier sinistral sets (blue), with the former abutting against mineralised shear fractures. Barren shear fractures then abut against both sets displaying a curvature indicative of a dextral fracture array. Abutting relationships suggest the barren shear fractures likely formed at the same time as the dextral en échelon vein array; however, given the lack of mineralisation it is likely they were isolated from the source of mineral-rich fluids.
In Fig. 3e, multiple phases of mineralisation and reactivation of veins
can be observed. Veinlets of ankerite both abut against and cut through
the calcite vein associated with a nearby small ( Ankerite veinlets formed along the N–S striking face-cleats. Faulting led to the development of coal breccia and calcite veining, which
either cut across or abut against pre-existing structures. Brecciation of the calcite vein and coal led to the development of a
chaotic fault breccia (following the classification of
Woodcock and Mort, 2008). The breccia contains
angular clasts of coal and calcite within an amorphous calcite matrix. Finally, mineralisation returned to ankerite with dextral en-échelon
arrays developed alongside barren tip-damage zones.
These observations suggest that initial deformation and associated
mineralisation occurred over a wide zone of en-échelon arrays (Fig. 3d), which was strongly influenced by the pre-existing cleat network (Fig. 3e). En-échelon arrays then began to interact leading to the development
of localised mineralised shear fractures (Fig. 3f). As the trace length of
the shear fracture increased so did the thickness of the zone, leading to
the formation of a dense array of small stratigraphic offset (
In order to understand the role of lithology on faulting style, we describe and compare fault characteristics between faults that cut the same lithology (self-juxtaposing faults) and faults that juxtapose multiple lithologies of the stratified sequence. Additionally, in order to elucidate the role of pre-existing joints on faulting style, we focus on the interaction between faults and fractures within the McDonald Limestone formation because of exceptional, laterally extensive bed-parallel exposure on the dip-slope.
Faulting at Spireslack SCM formed either under early sinistral (Phase 1) or
late dextral (Phase 2) shear (Fig. 4c). Phase 1 dextral faults are
interpreted as having formed concurrently with normal faults in the Six Foot Seat
Earth and thrusts in the shale. The south-dipping bedding is consistent
with the regional fold axis inferred from British Geological Survey (BGS) maps (40
Self-juxtaposed fault characteristics by lithology.
Self-juxtaposing faults with small stratigraphic offset (
Characteristic observations of self-juxtaposed faults:
The fault dip depends on the lithology cut by the fault. Dips in the
McDonald Limestone range from 45 to 88
Characteristics of faults that cut multiple lithologies.
Key features observed along faults that juxtapose multiple lithologies (i.e. that are not self-juxtaposing) are summarised in Table 2. Based on cross-cutting relationships, we observe two phases of faulting at Spireslack SCM.
Digitised fault strands of sinistral faults cutting the Limestone
Coal Formation exposed along the high wall.
Summary of the key features observed along faults that juxtapose multiple lithologies. Please see S3 for full field descriptions. LLF stands for Lower Limestone Formation, and LCF stands for Limestone Coal Formation.
Larger stratigraphic offset (
Fault core thickness is typically low (
Fracture maps with increasing intensity of faulting. For each digitised map the exported fault (red lines) and “joint” (dark grey lines) maps, along with the interpretation areas used for the analysis (light grey), are provided. Please note that while it is possible some joint's and/or faults acted as trailing segments (cf. Nixon et al., 2014), no direct field evidence was observed.
The style of the fault and fracture network in the McDonald Limestone changes across the site (Fig. 8) with the chronology and network properties of each sample area described below. In this section mineralised shear fractures, which are often faulted joints, are classified as faults for the network analysis.
Network characteristics of the joint and fault datasets presented for the three sample areas outlined in Fig. 8. Please note that because the fault network is superimposed onto the joint network, I-nodes (i.e. where a fault terminates) can represent a Y-node in the combined network. Similarly, where a joint terminated against a fault, due to the sealing properties of the fault, it is no longer appropriate to classify this as a connected branch and as such is classified as an I-node in the “joint network”. The percentage of each node classification is provided in brackets following the number of node counts. Trace length data are presented as orientation sets that were derived following visual assessment of length-weighted rose diagrams and do not relate to the age sets outlined in Fig. 8. For the combined network fracture statistics and trace length distributions for all datasets please refer to Fig. S2.
The interpretation area in Fig. 8a is dominated by large trace-length NE-trending joints and smaller trace-length NNW-trending joints. Abutting relationships suggest these formed as four distinct phases, with two phases occurring at each orientation. The fault network displays two orientation sets (N and NNW) of sinistral faults with low connectivity, trace length, and intensity (Table 3). Both fault sets abut against favourably orientated Phase 1 or Phase 3 joints, indicating they formed later. Abutting relationships of Phase 3 joints against NNW-trending faults suggesting Phase 2 joints were reactivated as faulted joints (following Zhao and Johnson, 1992) during the first phase of faulting. Phase 5 and 6 joints, which display variable orientations in Fig. 8a, abut against the faults, suggesting that they formed later.
The interpretation area in Fig. 8b, which is located slightly closer to
the NW-trending dextral fault zone that cuts the middle of the site (Fig. 4), displays a similar intensity of faulting (I
The interpretation area in Fig. 8c is located close to the major
NW-trending dextral fault (Fig. 4) and includes two self-juxtaposing
faults towards the bottom and top of the studied section (Fig. 8c). The
panel displays a complex fracture evolution; however, many of the features
observed in the previous panels are visible. Phase 1 to 4 joints are still
easily identified; however, their trace length has further decreased due to
increased fault intensity (I
As fault intensity increases, the complexity of age relationships in the
fault-fracture network also increases (Fig. 8). Phase 1 to 4 joints are
identified across all three panels and are interpreted as the “pre-existing”
joint network. As fault intensity increases, these “pre-existing” features
become segmented through faulting and their recorded trace length decreases.
While fault intensity is similar in Fig. 8a and b, faults with a N–S
strike are only present in Fig. 8a. This is probably due to the subtle
anticlockwise rotation of the pre-existing joints relative to the stress
field that enabled the reactivation of Phase 2 and 4 as faulted joints
(Fig. 8b, c) and promoted the formation of Phase 5 and 6 joints (orange
and purple lines in Fig. 8). The number of faulted joints drastically
increases with increased fault intensity, with joints becoming linked
through the formation of new fault strands. In agreement with the void-scale
mapping (Fig. 4), two phases of faulting have been identified in Fig. 8b
and c, with an earlier sinistral and later dextral phase. The sinistral
phase appears to preferentially reactivate Phase 2 joints, whereas the
dextral phase preferentially reactivated both Phase 2 and 4 joints. The
increase in reactivated joints and two clear phases of faulting in Fig. 8c explain the large increase in joint intensity in this panel (I
Summary of the structural features observed at Spireslack SCM. References in the table are as follows: (1) Leeder (1982); (2) Underhill et al. (2008); (3) Haszeldine (1984); (4) Ellen et al. (2019); (5) Coward, (1993); (6) Anderson (1951); (7) Soper et al. (1992); (8) Browne et al. (1999); (9) Read et al. (2002); (10) George, (1978); (11) Thomas (2013); (12) O'Keefe et al. (2013); (13) Rippon et al. (2006); (14) Ritchie et al. (2003); (15) Leslie et al. (2016); (16) Caldwell and Young (2013); (17) Emeleus and Gyopari (1992); and (18) Mykura (1965).
The exceptional 3D exposures of the Limestone Coal Formation and surrounding lithologies have informed a five-stage conceptual model for the development of the structures (Table 4). While this model is based on observations from Spireslack SCM, the model could be improved by utilising data from nearby open-cast sites (Leslie et al., 2016), legacy subsurface data as introduced in Ellen et al. (2016), and additional correlation with the larger-scale structures observed in the Midland Valley of Scotland.
The mechanically stratified succession at Spireslack SCM has led to the
development of a fracture stratigraphy (Laubach et
al., 2009). While joints across the site locally display two “orientation
sets” (Fig. 8, insets), abutting relationships discussed in Sect. 4.4.3
identified 8 “age sets” punctuated by two phases of faulted joints (Fig. 8). Different orientation sets have previously been attributed to separate
tectonic events (e.g. Vitale et al., 2012) or situations where the
intermediate (
There are several examples of joints or cleats influencing fault growth at Spireslack SCM (Figs. 3, 5, 8). Jones and Tanner (1995) found that transpressional strain can often become partitioned across pre-existing structures. At Spireslack, joints appear to be accommodating the shear-strain component, with pure shear accommodated through the tightening of the Muirkirk Syncline. Throughout both deformation phases, faults abutted against NE-trending joint sets. However, during the sinistral phase of faulting, larger trace length N- to NNW-trending cleats and joints (e.g. Phase 2 joints, Fig. 5a) were reactivated. As the principal stress orientation changed to enable the formation of phase 2 dextral faults (Fig. 4c), faulted joints associated with the first phase of faulting became reactivated (Fig. 8c), with Phase 4 joints preferentially reactivated (Fig. 8b, c). Phase 2 joints only became reactivated in the vicinity of self-juxtaposing dextral faults (NW–SE-trending feature cutting Fig. 8c). The preferential reactivation of specific joint sets could be due to the following reasons.
Changes in the mechanical properties of lithologies at Spireslack SCM due to mineralisation associated with Phase 1 faults. For example, coal cleats, which previously acted as a weakness in the rock (Li et al., 2016), now act as strength inclusions following ankerite mineralisation, enabling barren shear fractures to develop (Fig. 3d).
Subtle differences in joint orientation between sets (Fig. 8, inset) change the relative orientation of features to the stress field (e.g. Moir et al., 2010; Zhao and Johnson, 1992) and alter the stress ratio across the fracture (Chang and Haimson, 2000; Haimson and Chang, 2000).
Differences in mechanical properties of the fracture surface, e.g. due to their longer trace length or fracture roughness (Nasseri et al., 2009; Tsang and Witherspoon, 1983), could increase in Phase 1 joints in comparison to smaller trace length Phase 3 or 5 joints (Reed et al., 2008).
The fact that some joints show evidence of preferential reactivation and
subsequent cementation, while others remain barren, suggests that certain
joint sets indicate the past connectivity of mineral rich fluids through the
network, which at Spireslack SCM was dominated by faults (Fig. 8). Barren
joints typically post-date mineralisation
(Peacock, 2001; Peacock and Sanderson, 2018); however, Phase 1 and 3 joints at
Spireslack are often offset by faults or reactivated as faulted joints
(Fig. 8). This suggests joints were present at the time of faulting;
however, not all joints sets were hydraulically connected to the
mineralising fluids. This could be either due to fluid flow being dominated
by vertical flow associated with fault–valve behaviour during slip events
(Sibson,
1990, 1992), micro-cataclasite and/or mineralisation along joints that were
not visible during field observations or had been weathered out during
subsequent groundwater flow, or mineralisation that occurred under a
stress-induced flow pattern that had a relatively high stress ratio (
Network topology data. Node and branch triangle (following
Sanderson and Nixon, 2015) for the
joint, fault, and combined fracture networks for the three sample areas
shown in Fig. 8:
Groundwater flow within Carboniferous aquifers is dominated by bed-parallel
fracture flow (Dochartaigh et al., 2015). While the combined
fault–fracture network across the McDonald Limestone displays very high
network connectivity (
The drop on connected joints is shown in the trends (pink arrows) on Fig. 9 and is caused by the gradual increase in abutting relationships between faults and joints. As more joints become reactivated as faults, the fault network becomes more connected as splays (i.e. Y-nodes; Fig. 8) develop, whilst reducing the number of connected joints (i.e. X- and Y- nodes in the “joint” dataset) (Fig. 8). Similarly, as the intensity and connectivity of the fault network increases, the number of abutting relationships between joints and faults increases. The increases the number of I-nodes in the joint network and gradually decreases the number of connected branches as the intensity of faulting increases. This leads to the counter-intuitive observation that although joint intensity increases in areas associated with faulting (Table 3), the cementation of faults and faulted joints causes the connectivity of the modern-day network in these areas to be lower (Fig. 9).
The internal structure of faults at Spireslack SCM is greatly affected by the level of lithological juxtaposition, with different properties observed for self-juxtaposing faults (Sect. 4.2.1) and those that cut multiple lithologies (Sect. 4.2.2). Self-juxtaposing fault strands cutting lithologies without pre-existing joints are typically relatively planar, develop relay zones, and only display local iron staining along fault planes (e.g. the Six Foot seat earth; Fig. 5a, b). Conversely, in lithologies where pre-existing weaknesses influence the growth of faults, multiple sets of lineations on fault planes and the presence of compound veins provide evidence for multiple slip events (McDonald Limestone; Figs. 3a, c, 6b and 8). This suggests that faults in these lithologies initiated as a segmented fault–fracture mesh (Sibson, 1996), with field evidence suggesting mineralising fluid flow in the McDonald Limestone and coal occurred as multiple crack–seal events (Fig. 3c, e). This implies that self-juxtaposing faults cutting the McDonald Limestone and Six Foot Coal at Spireslack behaved in a similar manner to other faults in carbonates with fluid pathways only remaining open for a small amount of time and probably closing following fault slip (cf. Billi et al., 2003; Sibson, 1990, 1992). Mineralised Phase 1 faults that cut multiple lithologies also display multiple slip events (Fig. 6c) and matrix-supported (calcite) fault breccias located within relay zones (Fig. 6d), intersections between Phase 1 and 2 faults (Fig. 6c), and asperities along the principle slip zone (Fig. 6a). This suggests fault valve behaviour was also present along non-self-juxtaposing faults (Peacock et al., 2019; Sibson, 1990).
Where faults cut multiple lithologies, shale accommodates the rotation of
bedding, leading to rotated blocks and multiple generations of curved
slickensides (Fig. 6). As shale is buried and compressive stresses
increase, the ratio of pre-consolidation stress and compaction-related
stresses control the behaviour or shales and mud rocks
(Yuan et al., 2017; Nygård et al., 2006). As a general rule,
shales are ductile during burial and brittle during exhumation, where they
experience stresses below the maximum stress they have encountered. Ductile
behaviour of the shales at the time of faulting suggests that both phases of
faulting occurred prior to maximum burial, which is estimated at
Fault cores at Spireslack SCM also differ between self-juxtaposing faults (Fig. 5) and those that cut multiple lithologies (Table 2). Wilkins et al. (2001), studying growth of normal faults through jointed lithologies, found similar observations to those at Spireslack SCM with little fault rock development (Fig. 3a, 5f) and considerably smaller displacement-to-length ratios than expected for faults that do not cut jointed lithologies (Figs. 5f, 6a, d). While the fault core at Spireslack SCM is typically thin (Table 2), similar to previous studies (e.g. McKay et al., 2019; De Rosa et al., 2018) thickness was found to be highly heterogeneous both along strike and down dip. Much of this variability is caused by the lithological juxtapositions observed across the fault (Fig. 7), asperities on the principal slip zone (Fig. 6), the degree of folding (Fig. 6), and the presence of fault core lenses (Fig. 7). In agreement with the fault growth model of Childs et al. (2009), the highly segmented network of self-juxtaposing faults (Fig. 8) and differences in fault dip between lithologies (Fig. 4d) contribute to the heterogeneity observed in the fault cores of faults that cut multiple lithologies.
Our data demonstrate that the evolution of faults and fault zone structure, and therefore the bulk hydraulic properties of the rock mass at Spireslack SCM, varied both through time and as faults cut multiple lithologies. The abundance of faults within competent lithologies that cannot be traced into shale interbeds suggests faults at Spireslack SCM initiated as segmented fault strands within competent lithologies (e.g. limestones) and the coals (Figs. 5, 8). Shale interbeds acted to restrict fault growth and instead accommodated ductile deformation. Despite faulting being dominantly strike slip, the oblique orientation of faults to bedding across the site meant that many fault growth models derived from observations and modelling of normal faults in mechanically layered sequences appear to be valid (e.g. Childs et al., 1996; Ferrill et al., 2017; Schöpfer et al., 2006, 2007, 2016). However, it is also clear that the initial segmented fault network within the competent layers was strongly controlled by the presence and evolution of the joints and mineralised fault zones (Fig. 8). It is therefore helpful to consider the concept of lithological juxtaposition, the presence and behaviours of shale interbeds, and the relative timing of deformation, when considering the growth and internal structure of fault growth in mechanically layered sequences.
The exceptional exposures of the Limestone Coal Formation at Spireslack SCM provides an excellent opportunity to examine the role of lithology and pre-existing structures on fault evolution, internal structure, and connectivity. Careful mapping to unpick cross-cutting relationships has revealed a five-stage, complex geological evolution for the Spireslack SCM succession consisting of two phases of faulting and eight phases of joint development.
While the overall fracture density increases around the larger faults,
counter-intuitively the modern-day network connectivity decreases in these
areas due to the cementation of faults and joints.
We find that the fault zone internal structure at Spireslack SCM depends on (a) whether the fault is self-juxtaposing or cuts multiple lithologies; (b) the presence and ductility of shale layers, which in turns leads to bed rotation and fault-core lens formation; and (c) the orientation of open and mineralised joints or coal cleats at the time of faulting. Self-juxtaposing faults are strongly affected by the orientation and mineralisation of pre-existing joint sets and coal cleats, causing them to grow as multiple segmented fault strands within competent lithologies. Self-juxtaposing faults only become well connected where fault intensity is high. Faults that cut multiple lithologies are strongly affected by the presence of shale interbeds and display a complex and heterogenous fault structure with fault length limited by the presence of pre-existing faults. Therefore, it is crucial to appreciate the relative timing of deformation events, concurrent or subsequent cementation, and the degree of lithological juxtaposition when considering the mechanical and hydraulic properties of a mechanically stratified succession.
The data and metadata associated with this publication are available from the University of Strathclyde KnowledgeBase at https://doi.org/10.15129/4556163e-e417-4bd4-94d2-fc96ba9eb725, (last access: 11 November 2020, Andrews, 2020).
The supplement related to this article is available online at:
Initial discussions and planning of the paper were undertaken by all authors. BJA, ZKS, and RL were involved with conceptualization. BJA and LMcK undertook the collection of field data, with BJA undertaking lineament mapping and data analysis. The paper was prepared by BJA, with contributions from all authors. Funding and supervision of the project was provided by ZKS and RL.
The authors declare that they have no conflict of interest.
This article is part of the special issue “Faults, fractures, and fluid flow in the shallow crust”. It is not associated with a conference.
This work was funded through BJA's PhD studentship, supported by the Environmental and Physical Sciences Research Council (EPSRC, award no. EP/L016680/1). LMcK is supported by a University of Strathclyde Environmental and Physical Science Research Council (EPSRC) Doctoral Training Partnership (DTP) award (reference no. 1904102). We would like to thank Dave Healy for the use of the high-resolution photomontage of the McDonald Limestone dip slope and the British Geological Society for the use of the photomontage of the high wall. The authors would like to thank Bailey Lathrop and David Sanderson for their detailed reviews that helped greatly improve the quality and clarity of the manuscript.
This research has been supported by the Environmental and Physical Sciences Research Council (EPSRC) (grant no. EP/L016680/1) and the University of Strathclyde Environmental and Physical Science Research Council (EPSRC) Doctoral Training Partnership (DTP) (grant no. 1904102).
This paper was edited by Fabrizio Balsamo and reviewed by David Sanderson and Bailey Lathrop.