beta

You're viewing our new website - find out more

Publication - Research Publication

Unconventional Oil and Gas: Understanding and Monitoring Induced Seismic Activity

Published: 8 Nov 2016
Part of:
Business, industry and innovation
ISBN:
9781786523952

An independent research project into understanding and monitoring induced seismic activity.

93 page PDF

6.3MB

93 page PDF

6.3MB

Contents
Unconventional Oil and Gas: Understanding and Monitoring Induced Seismic Activity
2 Geological and Seismic Characteristics of the Study Area

93 page PDF

6.3MB

2 Geological and Seismic Characteristics of the Study Area

2.1 Overview

The bedrock geology of the Midland Valley of Scotland consists mostly of sedimentary rocks, rich in resources such as coal, limestone, ironstone and shale, which have been extracted throughout the years using underground or surface mining techniques. These rocks were formed between 350 and 300 million years ago, during a time when swamps flourished in a warm humid climate, when Mississippi sized river systems deposited vast quantities of sediments, and sea level fluctuations periodically flooded the land. Over geological time, these sediments were turned to rock and following numerous plate tectonic events, buried at depths sufficient to generate hydrocarbons. Such tectonic forces also deformed the rocks, forming geological folds, faults and fractures within the strata. These structures are mapped across the Central Belt of Scotland at all scales, and fall into three trends: E-W, ENE- WSW and NW- SE.

An analysis of recent instrumental recordings of earthquakes and older historical data confirms that earthquake activity in Scotland is low. On average there are eight earthquakes with a magnitude of 2.0 ML or above, somewhere in Scotland every year. The largest recorded earthquake in Scotland had a magnitude of 5.2 ML and only two other earthquakes with magnitudes of 5.0 ML or greater have been observed in the last 400 years. As a result, the risk of damaging earthquakes is low. Most of this earthquake activity is north of the Highland Boundary Fault, on the west side of mainland Scotland. Earthquake activity in the Midland Valley of Scotland is lower, and most of the recorded earthquakes in this area were induced by coal-mining. Since the decline of the coal-mining industry in the 1990's, very few mining-induced earthquakes have been recorded. Existing catalogues of earthquake activity in Scotland are incomplete at magnitudes below 2 ML from 1970 to present, and for higher magnitudes prior to this. This limits our ability to identify any areas that might present an elevated seismic hazard for any UOG operations. Similarly, limited information about the state of stress in the Earth's Crust mean that it is not possible to identify any particular parts of the study area where faults are more likely to be reactivated and that may present an elevated seismic hazard for any UOG operations.

2.2 Background

The UK is characterised by low levels of earthquake activity and low seismic hazard, in keeping with its intraplate setting. Historical observations of earthquake activity date back to the 14th Century, and show that despite many accounts of earthquakes felt by people, damaging earthquakes are relatively rare (Musson, 2007). This earthquake activity is caused by reactivation of existing faults by tectonic stresses from the ongoing collision of the African and Eurasian tectonic plates and widening of the Atlantic Ocean along the Mid-Atlantic ridge, and stresses from isostatic readjustment due to previous episodes of ice loading. The largest recorded earthquake had a magnitude of 5.9 Mw and there have been sixteen earthquakes with magnitudes of 5.0 Mw or greater since 1650. An analysis of recurrence statistics suggests that an earthquake with a magnitude of 5 Mw or above occurs somewhere in the British Isles on average every 50 years, while an earthquake with a magnitude of 6.0 Mw or above occurs roughly every 500 years. Most earthquakes in the British Isles nucleate in the mid-Crust at depths of 5-15 km, and there is some evidence to suggest that the largest earthquakes nucleate at the greater depths in the Crust. By contrast, there is no evidence of any earthquake in the last five hundred years that has ruptured the surface, although earthquakes with magnitudes close to the expected maximum magnitude (~6.5 Mw) for the British Isles and which nucleate at depths of less than 10 km, may, in theory, be capable of producing ruptures that propagate close to the surface.

What is an Earthquake?

Earthquakes are the result of sudden movement along faults within the Earth that releases stored up strain in the form of seismic waves or vibrations that travel through the Earth and cause the ground surface to shake. Such movement on the faults is generally a response to long term deformation and build-up of stress, caused by processes such as plate tectonics. When this stress exceeds the friction that holds the rocks on either side of the fault together, they slide or slip past each other.

The size of any earthquake depends on both the area of the fault that ruptures and also the average amount of slip or displacement on the rupture plane. Larger rupture areas and larger displacement lead to larger earthquakes. The largest earthquakes occur on ruptures that are many hundreds of kilometres long, with areas of several thousand square kilometres, and that have displacements of many metres.

Earthquake activity is greatest at the boundaries between the Earth's tectonic plates, where the differential movement of the plates results in repeated accumulation and release of strain. These include the edges of the Pacific and other subduction zones, e.g. the Mediterranean Trench between Africa and Eurasia or collision zones, e.g the Himalayan Belt. These are commonly referred to as interplate earthquakes. However, earthquakes can also occur within the plates far from the plate boundaries, and where strain rates are low. These are often referred to as intraplate earthquakes. Large areas of Asia, Australia, Europe and North America all experience intraplate earthquakes, although these events are relatively rare.

2.3 Geographic and Tectonic Context

The UK lies on the northwest part of the Eurasian plate and at the northeast margin of the North Atlantic Ocean (Figure 2.1). The nearest plate boundary lies approximately 1,500 km to the northwest where the formation of new oceanic crust at the Mid-Atlantic ridge has resulted in a divergent plate boundary and significant earthquake activity. Around 2,000 km south, the collision between Africa and Eurasia has resulted in a diffuse plate boundary with intense earthquake activity throughout Greece, Italy and, to a lesser extent, North Africa. This activity extends North through Italy and Greece and into the Alps. The deformation arising from the collision between the African and European plates results in compression, commonly referred to as "Alpine compression", that is generally directed in a north-south direction. The northeast margin of the North Atlantic Ocean is passive and is characterised by low levels of seismic activity in comparison to other passive margins around the world.

As a result of this geographic position, the UK is characterised by low levels of earthquake activity and low seismic hazard. Evidence for this comes from observations of earthquake activity dating back to the 14th Century, which suggests that although there are many accounts of earthquakes felt by people, damaging earthquakes are rare.

The continental crust of the UK formed over a long period of time and has a complex tectonic history, which has produced much lateral and vertical heterogeneity through multiple episodes of deformation (Woodcock and Strachan, 2000), resulting in widespread faulting. Some of the principal fault structures represent major heterogeneities in structure of the Crust and have been the locus of later deformation. Earthquake activity in the UK is generally understood to result from the reactivation of existing fault systems by present day deformation, although such faults need to be favourably orientated with respect to the present day deformation field in order to be reactivated.

Figure 2.1. Black circles show the distribution of earthquakes with a magnitude of greater than 5 across Europe. The red line shows the margins of the Eurasian plate. The great majority of earthquake activity is located at the southern margin in Greece and Italy along the collision zone between Africa and Eurasia. The inset shows the distribution of earthquakes in North America. Plate boundary data from Bird (2003). Earthquake Data from the British Geological Survey World Seismicity Database, © NERC 2016

Figure 2.1. Black circles show the distribution of earthquakes with a magnitude of greater than 5 across Europe. The red line shows the margins of the Eurasian plate. The great majority of earthquake activity is located at the southern margin in Greece and Italy along the collision zone between Africa and Eurasia. The inset shows the distribution of earthquakes in North America. Plate boundary data from Bird (2003). Earthquake Data from the British Geological Survey World Seismicity Database, ©NERC 2016

2.4 Geology and Faulting of the Midland Valley of Scotland

2.4.1 What unconventional oil and gas resources are being considered in Scotland and are relevant to this project?

A range of unconventional oil and gas ( UOG) resources have been assessed in Scotland, including shale oil and shale gas in Carboniferous shales in the Midland Valley of Scotland (Monaghan, 2014), and coal mine methane, abandoned mine methane, coal bed methane and underground coal gasification in coal seams and associated mines in Central Scotland, the Sanquhar and Canonbie coal fields (Jones et. al., 2004). In this report we focus on the shale oil, shale gas and coal bed methane UOG resources. The majority of shale- and coal-bearing strata lie in the geological terrane of the Midland Valley of Scotland (essentially occupying the same geographical area as the Central Belt of Scotland), situated between the Scottish Highlands to the north and the Southern Uplands to the south. These strata are distributed in the west of Scotland in Ayrshire, and Douglas, and through central Scotland (including Clackmannanshire, North Lanarkshire, Stirlingshire) and eastwards to Fife and the Lothians (Figure 2.2).

Shale oil and shale gas are produced from an organic-rich rock known generically as 'shale', but more precisely termed mudstone, carbonaceous mudstone, oil-shale and fine siltstone. Shale is a sedimentary rock, composed of fine-grained (clay and silt sized) particles and which may contain organic matter ( e.g. derived from land and aquatic plants, bacteria and animal remains). Shale is typically considered as the source rock in conventional oil and gas accumulations: however, with unconventional hydrocarbon accumulations, shale acts as both source and reservoir rock, with hydrocarbons extracted using hydraulic fracturing techniques.

Coal bed methane ( CBM) is produced from a sedimentary rock known as coal. Coal is formed from organic matter ( e.g. prehistoric vegetation accumulated in swamps and peat bogs), and as a result contains a high proportion of carbon and consists largely of organic carbonaceous molecules. Coal seams in Scotland typically contain one or more sets of sub-parallel near-vertical fractures known as cleats.

2.4.2 What is the geology and characteristics of that geology in the Midland Valley of Scotland?

Coincident with the geographically low-lying limits of the Central Belt is the geological terrane of the Midland Valley of Scotland, a fault-bounded Late Palaeozoic sedimentary basin. It is composed of an internally structurally complex arrangement of Devonian, Carboniferous and Permian sedimentary and volcanic rocks, with up to over 5,500 m of Carboniferous strata locally (Cameron and Stephenson, 1985; Read et al., 1996).

Potentially prospective Carboniferous shales and coal beds are buried beneath an area from Glasgow to Edinburgh (Figure 2.3), to the Lothians, Falkirk, Clackmannan and Fife, and extending into Ayrshire and Lanarkshire (Monaghan, 2014 and Jones et al., 2004). These prospective resources belong to a total of eleven Carboniferous stratigraphic units, in descending stratigraphical order: the Coal Measures Group, Passage Formation, Upper Limestone Formation, Limestone Coal Formation, Lower Limestone Formation, Pathhead Formation, Sandy Craig Formation, Pittenweem Formation, West Lothian Oil Shale Formations, Anstruther Formation and the Gullane Formation. The units containing shale and coal prospective resources are summarised in Table 2.1. The units are distinguished based on their age and lithological composition ( e.g. variations in the amount of mudstone, sandstone, limestone etc).

The Coal Measures Group are the uppermost stratigraphic unit of interest for potential UOG resources in the Central Belt of Scotland. Where the Coal Measures Group are not present at surface, they have been eroded away and therefore do not overlie any other Formation (Figure 2.4). The same rule applies to all stratigraphic units beneath the Coal Measures Group. Therefore, where any Formation is present at surface, it will be underlain by units stratigraphically beneath them, unless that unit is not present (due to non-deposition in that area during the Carboniferous) or unless the stratigraphy is interrupted by faulting.

Figure 2.2: Geology of the Midland Valley of Scotland from 1:625 000 scale DigMap BGS© NERC. Potentially prospective units are shown.

Figure 2.2: Geology of the Midland Valley of Scotland from 1:625 000 scale DigMap BGS©NERC. Potentially prospective units are shown.

Figure 2.3. (b) Extents of potential coal bed methane resource from Jones et al. (2004). (c) Extents of potential shale gas and shale oil resource from Monaghan (2014).

Figure 2.3. (b) Extents of potential coal bed methane resource from Jones et al. (2004). (c) Extents of potential shale gas and shale oil resource from Monaghan (2014).

Table 2.1. Summary of the Carboniferous stratigraphy of the Midland Valley of Scotland (modified after Browne et al. 1999 and Monaghan 2014; dates from Waters, 2011). The eleven potentially prospective intervals are colour shaded blue for coal-rich interval, green for shale-rich interval and yellow for combined shale- and coal-rich interval. Note that the coal-rich intervals highlighted blue also contain shale but were excluded from the shale resource assessment of Monaghan (2014) due to their current day burial depth largely being < 1 km.

Table 2.1. Summary of the Carboniferous stratigraphy of the Midland Valley of Scotland (modified after Browne et al. 1999 and Monaghan 2014; dates from Waters, 2011). The eleven potentially prospective intervals are colour shaded blue for coal-rich interval, green for shale-rich interval and yellow for combined shale- and coal-rich interval. Note that the coal-rich intervals highlighted blue also contain shale but were excluded from the shale resource assessment of Monaghan (2014) due to their current day burial depth largely being < 1 km.

Figure 2.4. Schematic cartoon illustrating relationship between erosion, surface exposure and strata at depth. Cartoon not to scale. The equivalence of the stratigraphic groups shown to component coal and shale-bearing formations is listed in Table 1

Figure 2.4. Schematic cartoon illustrating relationship between erosion, surface exposure and strata at depth. Cartoon not to scale. The equivalence of the stratigraphic groups shown to component coal and shale-bearing formations is listed in Table 1

In Ayrshire, and the Sanquhar and Douglas Coalfields, in areas where the Coal Measures Group are present at surface, all potentially prospective stratigraphic units are present either at surface or at depth down to the Lower Limestone Formation (Figure 2.2). The West Lothian Oil Shale, Gullane and Anstruther formations are not present at depth in this location.

Throughout the remainder of the Central Belt, to the east of the mapped Clyde Plateau Volcanic Formation, potentially prospective stratigraphic units are present either at surface or at depth down to the Gullane, Anstruther or West Lothian Oil Shale formations, where the Coal Measures Group are present at surface (Figure 2.2). The West Lothian Oil Shale Formation and lateral equivalents crops out in West Lothian and is interpreted in boreholes to the north-west and west at depths of up to 5100 m (Monaghan, 2014). The Anstruther and Gullane formations are present throughout the eastern part of the region at depths of up to 5800 m. The West Lothian, Gullane and Anstruther formations are not overlain by any other Formation where they crop out at surface between Edinburgh and Linlithgow, and in the north-east of Fife and in East Lothian.

All formations described above are intruded (in varying volumes and intensities) by Palaeozoic and Palaeogene aged igneous sills and dykes.

2.4.3 What particular character of these formations are relevant to unconventional development, in particular induced seismicity?

Each stratigraphic unit defined in Table 2.1 has its own unique lithological (rock type) character ( e.g. amount of sandstone, coal, mudstone). Different rock types have different strengths and mechanical properties which are likely to vary with pressure, depth and temperature. This affects the way in which they fail as a result of stress or deformation. For example, limestones and sandstones are mechanically strong rocks composed of cemented grains and interlocking crystals, which at high levels in the crust, would tend to deform in a brittle fashion ( i.e. tensional failure by formation of natural fractures) when deformed (Figure 2.5). Shales and mudstones are mechanically weaker rocks, composed predominantly of organic material and clay, and therefore tend to deform in a more 'ductile' fashion ( i.e. slip occurs along bedding planes as well as fracturing the rock mass) when deformed (Figure 2.6). When these rock types form a layered succession ( e.g. such as in the Limestone Coal Formation), they behave and respond accordingly to their lithology. When studied at surface, the dense fractured nature of shale and bedding parallel slip is not observed or repeated in an overlying strong limestone for example, just as the widely spaced natural fractured nature of a limestone is not repeated in the underlying shale.

Figure 2.5: Faulted limestone pavement, note brittle nature (discrete fault planes) of deformation within the limestone. Lower Limestone Formation, East Ayrshire © BGS, NERC.

Figure 2.5: Faulted limestone pavement, note brittle nature (discrete fault planes )of deformation within the limestone. Lower Limestone Formation, East Ayrshire © BGS, NERC.

Figure 2.6: Interbedded mudstones and ironstones have responded differently to faulting. The mechanically strong ironstone remains relatively undeformed, whereas the mechanically weaker mudstone above has been tightly folded (highlighted by dashed white line). © BGS, NERC.

Figure 2.6: Interbedded mudstones and ironstones have responded differently to faulting. The mechanically strong ironstone remains relatively undeformed, whereas the mechanically weaker mudstone above has been tightly folded (highlighted by dashed white line). © BGS, NERC.

As a general rule, the formations are dominated by the following lithologies (Cameron and Stephenson 1985, Midland Valley Memoir):

Formation Name

Dominant Lithologies

Gullane and Anstruther

Sandstone and shale

West Lothian Oil Shale

Mudstone, oil-shale, siltstone

Lower Limestone Formation

Limestone, mudstone, sandstone

Limestone Coal Formation

Sandstone, siltstone, mudstone, coal

Upper Limestone Formation

Limestone, sandstone, mudstone

Passage Formation

Sandstone

Coal Measures Group

Sandstone, siltstone, mudstone, coal

Limited published datasets are available on rock properties, such as mechanical strength and behaviour, in the Carboniferous rocks considered as UOG resources in Scotland. Engineering geology borehole datasets from site investigation reports provide one data source for relatively shallow cored strata (commonly less than 100 m drilled depth). A brief overview of data held in the BGS geotechnical database suggest the uniaxial compressive strength varies between 3.4 (weak) to 195 MPa (very strong) in the Clackmannan Group (Upper Limestone, Limestone Coal and Lower Limestone formations) and fracture index [1] in the Coal Measures and Limestone Coal Formation strata varies from 0-200. Figure 2.7 and Figure 2.8 demonstrate the variability within some of the measured and described rock properties. This dataset requires further specialist interpretation as to whether it is likely to be applicable in the deep subsurface and how it varies with lithology.

Figure 2.7: Percentage of described strength from borehole core for six stratigraphic units from engineering geology descriptions in site investigations from the Glasgow area. Image courtesy David Entwisle, BGS.

Figure 2.7: Percentage of described strength from borehole core for six stratigraphic units from engineering geology descriptions in site investigations from the Glasgow area. Image courtesy David Entwisle, BGS.

Figure 2.8: Box and whisker plot of measured uniaxial compressive strength of different lithologies from Carboniferous strata described in engineering geology site investigations from the Glasgow area (note engineering geology classes 1.25-5=weak, 5-12.5=moderately weak, 12.5-50=moderately strong, 50-100 strong, 100-200=very strong). Image courtesy David Entwisle, BGS.

Figure 2.8: Box and whisker plot of measured uniaxial compressive strength of different lithologies from Carboniferous strata described in engineering geology site investigations from the Glasgow area (note engineering geology classes 1.25-5=weak, 5-12.5=moderately weak, 12.5-50=moderately strong, 50-100 strong, 100-200=very strong). Image courtesy David Entwisle, BGS.

Some description of eastern Scotland rock mechanical properties are given in Olden et al (2014) though this study considered the behaviour of rocks buried more deeply than the units of interest for UOG and modelled subject to injection of large volumes of CO 2.

Other indicators of in-situ stress orientation, of drilling induced tensile fractures and of variation with lithology in the heterolithic succession could be investigated by examining borehole/well breakouts, as in Kingdon et al. (2016a, b), Williams et al. (2015). However no such study is currently known from Scotland. The age and quality of well datasets, along with their spatial distribution could limit work of this type.

There are also limited datasets available on baseline levels of natural fracturing within Carboniferous strata. For example, cleat, a network of orthogonal extensional joints, is a characteristic structure in coals. Cleat azimuths form perpendicular to the minimum principal stress and parallel to the maximum horizontal stress during fracture formation (Rippon et al., 2006). Cleat data are necessary for modelling bulk rock strength behaviour for coal gasification technology and for optimum in-seam drilling directions in coalbed methane exploitation (Gayer & Harris, 1996 & Kent 1996). In general, the main cleat azimuth will vary near faults and the cleat will increase in frequency as will associated mineralisation. Fox (1965) noted cleat frequency was greater in bright coals than in durain or cannel coals. The main cleat azimuths in the Lothians and Fife are broadly NW- SE, with a range from 287° to 328° in the Lothians, and in Fife from 285° to 335° (Fox, 1965: Rippon et al., 2006). The main cleat in coals of the Clackmannan area has an E-W trend (parallel to the prominent normal fault set).

Baseline levels of natural fracturing will change across the area depending on local tectonic background, proximity to faulting, fold structure and bed thickness. The presence of igneous dykes or sills (mechanically strong and crystalline rocks) will likely have a local effect on rock strength and deformation characteristics.

In addition to natural fracturing and bedding planes between sedimentary units, the formations are also affected by other discontinuities including faulting and mine workings. Much of the Central Belt is undermined where coal-bearing strata are present (see Jones, 2004; Gillespie et al., 2013), leaving voids, collapsed workings or packed waste in the sub-surface space previously occupied by the coal. In mines that underwent 'stoop and room' (where pillars of rock are left unmined to support the mine roof), mining voids can be 50% (Gillespie et al., 2013). In 'longwall' mines (where the coal seam is worked between two parallel access roadways and the roof is allowed to collapse as the workings advance), only 20% of voids may remain (Younger and Adams, 1999 in Gillespie et al., 2013). An increase in fractures associated with rock collapse and changes in rock stress is likely in the strata above the mines in a zone of extraction-related subsidence (Younger and Adams, 1999). The presence and extent of mine workings should therefore be factored into detailed studies on UOG potential. Detailed mine plans of the workings can also be used to gain a more detailed knowledge of faulting at the local scale.

2.4.4 What is the geological structure of the area?

The Midland Valley of Scotland, a series of ancient rift basins, is bounded in the north by the Highland Boundary Fault and in the south by the Southern Upland Fault, both ENE- WSW trending structures. The graben structure was developed in the early Devonian (Bluck, 1978) in a zone of crustal weakness inherited from the Lower Palaeozoic (during the Caledonian Orogeny). The area underwent extensive tectonic deformation over the Palaeozoic Era, forming a series of inter-related depocentres and intra-basinal highs. There is little direct tectonic evidence for any subsequent major tectonic event affecting the area (Underhill et al., 2009). The faults within the Central Belt are generally normal ( e.g. East Ochil Fault) or oblique-slip structures ( i.e. some component of strike-slip movement, e.g. Paisley Ruck), with occasional instances of reverse slip ( e.g. Pentland Fault). They are typically steeply dipping structures e.g. between 50 and 85 (near-vertical) degrees (Monaghan 2013), although low angle thrust faults of approximately 30 degrees in dip have been identified in surface coal mines during field visits by the BGS. Fault dips can be measured from direct observation in the field, from coal mine plans, seismic interpretation and borehole data. Several of the faults across the Midland Valley have quartz-dolerite dykes intruded along them ( e.g. Slamannan Fault in Falkirk).

Faults are typically drawn on BGS maps as linear planar structures. However, in nature faults are not simple linear structures but comprise a zone of intensely variably fractured, broken rock ( e.g. Caine et al., 1996). Fault zone complexity is dependent on a number of properties including: host rock lithology, displacement and pre-existing structure in the rock mass. A single fault may change in complexity, structure, strike and dip along its length ( e.g. Childs et al. 1997; Schulz and Evans 1998; Wibberley et al., 2008). Subsidiary faults are often found within the fault zone of large fault structures, although their resolution may be such that they are not mappable, or not identifiable in seismic. Subsurface coal mine abandonment plans may be of high enough resolution to identify these smaller offset faults.

Faults mapped by the BGS have only been occasionally directly observed in the field ( i.e. along coastlines, quarries and opencast sites). They are often interpretations based on landscape features ( e.g. zones of weakness utilised by drainage) or are used to solve geological problems where there is a lack of continuity across geological strata. The faults on the BGS maps are interpreted from field mapping, borehole information, mine abandonment plan data and seismic data and as such may not capture the full structural complexity of an area. The seismic data is of variable quality in the Midland Valley of Scotland, with the result that in some areas of poor data quality it is difficult to tie subsurface seismically interpreted structure to surface geology.

2.4.5 Regional Fault Trends

There are three dominant trends of major faults within the Midland Valley interpreted on BGS maps: E-W, ENE - WSW, and NW- SE. These structures are interpreted as having been active at times during the Devonian-Carboniferous to Permian periods (419 to 299 million years ago). The majority of the structures trend E-W and ENE- WSW, with a lesser population to the NW- SE. In general the Ayrshire coal fields are dominated by ENE- WSW faults, the Central Coalfield by E-W faults and Fife by NW- SE and ENE- WSW faults. BGS maps at 1:625,000 and 1:50, 000 scale show the positions of major faults mapped at surface (Figure 2.9 , Figure 2.10). However, numerous smaller faults have been identified in mine plans and by field mapping. In this section the three dominant trends of major faults in the Midland Valley at the 1:625 000 and 1:250 000 scale geology maps are described, before briefly considering the distribution of faults at the 1:50 000 scale. Note, description of the main faults within each 1:50 000 geological sheet are provided within the relevant geological memoir of the area ( e.g. for Glasgow: Hall et al., 1998; for Falkirk: Cameron et al., 1998; for Edinburgh: Browne et al., in press).

ENE - WSW trending faults

This fault trend predominates in the western side of the Midland Valley of Scotland. Major faults which trend more or less parallel to the Midland Valley terrane bounding faults include the Dusk Water Fault, Paisley Ruck, Kerse Loch Fault, Pentland Fault, Ardross and Dura Den faults. These faults are interpreted as related to inherited crustal weakness from the Lower Palaeozoic tectonic events, as are the major Highland Boundary and Southern Upland bounding faults ( Figure 2.2). The Paisley Ruck has a wide fault zone of up to 180 m, and has been described as a positive flower structure (Gibbs, 1984). Near Linwood, the fault downthrows strata to the north and has a displacement of up to 550 m, though usually less (Hinxman et al. 1920). The Dusk Water Fault dips at 85 degrees toward the south-east (Monaghan 2013).

E -W trending faults

There are few E-W trending mapped faults in the Ayrshire Coalfield at the larger scale e.g. between Stewarton and Kilmarnock three E-W unnamed trending faults are marked. In the Clackmannan and North East Stirlingshire Coalfields, the general trend of faulting is east-west and is considered neither severe nor closely-spaced (Jones, 2004). The major faults include the Abbey Craig, Alloa, Clackmannan and Kincardine Ferry faults, which all throw down to the south with maximum values in the range of 120 - 240 m (Jones, 2004). The East Ochil Fault (Rippon et al., 1996) forms the northern margin of the coalfield and has a vertical throw of at the most 1300 m (Browne and Woodhall, 1999). The East Ochil Fault was exposed during excavations in the Westfield Surface Coal Mine: the fault zone is steeply dipping at 65 degrees and of normal throw with a shatter zone up to 15 m wide (Browne and Woodhall, 1999; Rippon et al., 1996; Underhill et al., 2008). Other E-W faults in the coalfield south of the East Ochil Fault have moderate fault plane dips of around 45° to the south (Rippon et al., 1996).

Figure 2.9: Major onshore faults mapped at 1:625 000 scale BGS© NERC DigMap within the Midland Valley of Scotland. Faults are colour coded by orientation.

Figure 2.9: Major onshore faults mapped at 1:625 000 scale BGS©NERC DigMap within the Midland Valley of Scotland. Faults are colour coded by orientation.

Figure 2.10: Major onshore faults mapped at 1:250 000 scale BGS© NERC DigMap within the Midland Valley of Scotland. This includes the major faults from the 1:625 000 scale. Note additional complexity and density of faults which is not captured at the 1:625 000 scale.

Figure 2.10: Major onshore faults mapped at 1:250 000 scale BGS©NERC DigMap within the Midland Valley of Scotland. This includes the major faults from the 1:625 000 scale. Note additional complexity and density of faults which is not captured at the 1:625 000 scale.

Faults of this trend are a marked feature of the Fife coalfield region, including the Durie Fault, and the Balcormo Fault (Forsyth and Chisholm, 1977). The Campsie Fault, which bounds the north of the Central Coalfield, has a maximum throw of 1600 m but decreases markedly towards the east. The fault splits into northern and southern branches that extend for over 3 km and are up to 600 m apart (Forsyth et al., 1996).

NW- SE trending faults

The Sanquhar Basin lies in a half-graben bounded to the north-west by a splay of the Southern Upland Fault zone, and to the north-east by the Sanquhar Fault, a normal fault with overall downthrow of 580 m (Smith, 1999). The Sheardale and Arndean faults in the Clackmannan and North East Stirlingshire Coalfields have large downthrows to the south-west: the Sheardale Fault has a maximum downthrow of 200 m, and the Arndean over 1000 m (Francis et al., 1970). The Dechmont Fault is the largest NW trending feature in the Midland Valley of Scotland, with a downthrow of 650 m to the north-west (Hall et al., 1998).

2.4.6 Palaeozoic Fault Timing

There is evidence that the Midland Valley of Scotland underwent active tectonism including periods of faulting from the late Devonian through to the Late Carboniferous-early Permian ( Table 2.2; Cameron and Stephenson, 1985; Read et al., 2002; Rippon et al., 1996; Underhill et al., 2008).

The ENE- WSW faults are interpreted as reactivated pre-existing structures formed during the Caledonian Orogeny. Many of the faults of east-west and NW- SE trend cut across the youngest Carboniferous rocks in the area and therefore date from latest Carboniferous or early Permian times (Paterson et al., 1998). By virtue of their association with latest Carboniferous-early Permian dykes, faults and extensional fractures, E-W striking faults and associated intra-basinal structures are dated at Latest Carboniferous (Browne & Woodhall, 1999; Rippon et al., 1996; Stephenson et al., 2003; Monaghan & Parrish 2006).

2.4.7 Faulting - a closer view

The BGS 1:625 000 and 1:250 000 geology maps give a regional overview of the major faults within the Midland Valley of Scotland. However, they cannot provide an accurate representation of the complexity of structures within the subsurface, such as providing intensity, density, spacing and extent, due to their large scales. The 1:50 000 geology map provides a more representative indication of this complexity (Figure 2.11), however it still does not capture the detail for small displacement ( e.g. less than 10 m offset) faults that may be present in the subsurface but not extend to surface. This is in part due to scale, and also due to lack of natural exposures within the Midland Valley of Scotland. As such, small displacement faults have mostly been recorded in surface coal mines, or underground coal workings. In some circumstances, mining information can therefore provide a greater level of understanding of faulting in a particular coal seam for example. An examination of mine plans following coal extraction by Jones (2004; pages 112, 115, 117, 118, 119) provides an indication of the fault intensity in the subsurface of some of the coalfields:

  • Ayrshire Coalfield: Intensively faulted
  • Douglas Coalfield: Faulting is significant and may be closely spaced.
  • Clackmannan and North East Stirlingshire coalfields: Faulting is neither severe nor closely-spaced
  • Fife Coalfield: Faulting is neither severe nor closely-spaced
  • Lothian Coalfield: Faulting is not severe but may be closely-spaced

Table 2.2. Summary chart of Midland Valley of Scotland Carboniferous tectonic history (modified from Monaghan, 2014).

Timescale Lithostratigraphy (after Browne et al. 1999) Tectonic events
Permian  

NW- SE trending faulting

 

E-W trending faulting

Carboniferous Coal Measures Group

Scottish Upper Coal Measures Formation

Scottish Middle Coal Measures Formation

Scottish Lower Coal Measures Formation

 
Clackmannan Group Passage Formation  
Upper Limestone Formation
Limestone Coal Formation
Lower Limestone Formation Dextral oblique-slip on reactivated Caledonide bounding structures. growth over E to ESE trending faults. NE-faults active in west.

Strathclyde Group

Lawmuir Formation

Kirkwood Formation

West Lothian Oil-Shale Formation  
Clyde Plateau Volcanic Formation Gullane Formation
Inverclyde Group Clyde Sandstone Formation
Ballagan Formation
Kinnesswood Formation Possible extensional rifting, NE- SW trends
Devonian Stratheden Group

Possible sinistral oblique reactivation

Acadian deformation, uplfit and erosion

Sinistral oblique reactivation of Caledonide faults generates Devono-Carboniferous basins

Field studies of the formations of interest provide a better understanding of fault scaling relationships within the rocks. An example is given from the Spireslack surface coal mine where an exposed Lower Limestone Formation limestone pavement which sits stratigraphically above a shale layer is faulted across the length of its exposure (Figure 2.12). Upon that fault plane, cross-cutting fault displacements range from 10 m to cm-scale. The detail of local scales of faulting cannot be captured on the regional maps presented here. The evidence from opencast coal sites visited by BGS geologists over many years and mine plans suggests great local variability in faulting density and style; some opencast sites examined have been largely free of natural faults, others have been complexly dissected by faulting at a variety to centimetre to kilometre scale.

Figure 2.11: 1:50 000 scale faulting BGS© NERC DigMap . The three dominant trends are still prominent, but the reality of fault spacing and density is better visualised at this scale. Some gaps in the fault dataset reflect the vintage of BGS mapping of the area.

Figure 2.11: 1:50 000 scale faulting BGS©NERC DigMap . The three dominant trends are still prominent, but the reality of fault spacing and density is better visualised at this scale. Some gaps in the fault dataset reflect the vintage of BGS mapping of the area.

Figure 2.12: Faulted limestone pavement, East Ayrshire. In addition to the faults which displace the limestone, there is a set of orthogonal fractures (with no offset) across the limestone. © BGS, NERC.

Figure 2.12: Faulted limestone pavement, East Ayrshire. In addition to the faults which displace the limestone, there is a set of orthogonal fractures (with no offset) across the limestone. © BGS, NERC.

2.4.8 Cenozoic faulting

The Cenozoic Era encompasses the Palaeogene, Neogene and Quaternary periods and includes the present day. During the Palaeogene, a major phase of rifting occurred to the west of Scotland related to opening of the Atlantic Ocean. It is possible that some of the minor NW- SE structures and tensional fractures may be related to the intrusion of NW- SE orientated Palaeogene dyke swarms, formed due to sea-floor spreading in the north-east Atlantic (Cameron and Stephenson, 1985). Dykes of this age commonly occupy pre-existing Palaeozoic structures.

However, there is little direct evidence for Cenozoic faulting within the Midland Valley. Several master-joints in a NW to northerly trend in the Ochil Hills have been found to display recent movement, along lineaments less than 1km (Davenport et al., 1989). The Southern Uplands Fault has also been found to display evidence for movement during the late Palaeogene to Quaternary, with a suggestion of recent movement during post glacial times where the River Nith crosses the fault (Sissons, 1976; Davenport et al., 1989). Davenport et al., (1989) note that continuing minor activity at sites of post-glacial faulting is evident within the Midland Valley. These major structures along which recent faulting is observed are of an ENE- SWS to E-W trend (Davenport et al., 1989).

2.5 Spatial and Temporal Characteristics of Earthquake Activity

2.5.1 Earthquake Data

Earthquake information can derived from two sources: historical archives containing references to felt earthquakes; and, earthquake source parameters derived from instrumental recordings of ground motions from earthquakes. The primary source of data for historical earthquakes in the British Isles from 1382 to present is the catalogue of Musson (1994), along with subsequent updates. This contains locations and magnitudes determined from the spatial variation of intensity, a qualitative measure of the strength of shaking of an earthquake determined from the observed effects on people, objects and buildings ( e.g. Musson, 1996b).

The primary sources of data from 1970 to present are the annual bulletins of earthquake activity published by BGS each year ( e.g. Galloway et al., 2013). These contain locations and magnitudes determined from recordings of ground motion on a network of sensors around the UK ( e.g. Baptie, 2012). The bulletins also contain error estimates. Bulletin data are updated with revised parameter data published in BGS reports or peer-reviewed journal publications on specific earthquakes ( e.g. Ottemoller et al., 2009).

Estimates of earthquake source parameters determined from historical data are generally subject to larger uncertainties than those determined from instrumental data. In addition, the completeness of these catalogues varies strongly with time. For example, Musson (1996a) states that prior to 1700 only the largest earthquakes are known about, whereas, from the 19th century on many smaller earthquakes are known about. The completeness of instrumental catalogues may also vary as a function of time as a result of changes to the monitoring network, e.g. an increase in the number of sensors in a given area.

Source parameters for both historical and instrumentally recorded earthquakes are combined in the BGS earthquake catalogue for the UK and immediate offshore area. Figure 2.13 (a) shows historical (yellow circles) and instrumentally recorded (red circles) earthquakes from this catalogue in Scotland. It is clear that there is significant spatial variation, with most earthquakes lying north of the Highland Boundary Fault and on west side of mainland Scotland. Northern and eastern Scotland shows relatively little earthquake activity. Earthquake activity is also lower in the Midland Valley, particularly in the instrumental period, though there is some evidence for historical earthquake activity.

The magnitude 5.2 ML Argyll earthquake in 1880 is the largest known Scottish earthquake and was felt along the west coast of Scotland, east as far as Perthshire and throughout the Hebrides. Earthquakes with magnitudes of 5.1 and 5.0 ML occurred in Inverness in 1816 and 1901. The 1901 earthquake was felt over much of Scotland and caused substantial amounts of minor damage in Inverness, including falling chimneys and masonry. The earthquake was followed by an aftershock sequence that lasted some months.

Comrie, on the Highland Boundary Fault experienced earthquake swarms (sequences of earthquakes clustered in time and space without a clear distinction of main shock and aftershocks) between 1788 and 1801, and again between 1839 and1846. The largest event in this sequence was a magnitude 4.8 ML event in 1839. The magnitude 4.4 ML Kintail earthquake in 1974 is the largest instrumentally recorded earthquake in the catalogue. This was the largest in a swarm of over 20 events that occurred over several months.

Our initial analysis is based on the instrumental catalogue (1970-present). This contains 4526 earthquakes, 329 of which have magnitudes of 2.0 ML or greater. The minimum and maximum recorded magnitudes are -0.7 ML and 4.4 ML, respectively. There are five earthquakes with a magnitude of 4.0 ML or greater including the magnitude 4.4 Kintail earthquake, as well as two other earthquakes in the Kintail sequence. The other two are the magnitude 4.1 ML Oban earthquake in 1986 and the magnitude 4.0 ML Arran earthquake in 1999.

Earthquake Magnitude

Earthquake magnitude is a measure of the amount of energy released during an earthquake and is determined from the amplitude of the ground motions caused by the earthquake. We also need to know how far away the earthquake was because the amplitude of the seismic waves decreases with distance, so we must correct for this. The first magnitude scale was developed by Charles Richter in 1935, based on observations of earthquake ground motions in California. Although this magnitude scale is only strictly applicable in California, it has been used all around the world and is commonly referred to as Local Magnitude, ML.

Magnitude scales are logarithmic so that each whole number increase in magnitude represents a tenfold increase in measured amplitude and about 32 times the energy released. Seismic moment is usually estimated directly from recordings of earthquake ground motions.

A number of different magnitude scales have been developed; however, the most standard and reliable measure of earthquake size is moment magnitude, MW, which is based on seismic moment, which is related to both the area of the rupture and the displacement on the rupture.

How the Richter’s magnitude Scale works. The amplitude is measured from the recording of ground motions at a specific site, as is the distance from the earthquake. A line connecting the two values on the graph gives the magnitude of the earthquakes.

How the Richter's magnitude Scale works. The amplitude is measured from the recording of ground motions at a specific site, as is the distance from the earthquake. A line connecting the two values on the graph gives the magnitude of the earthquakes.

Magnitude

Earthquake Effects

2.5 or less

Usually not felt, but can be recorded by seismograph.

2.5 to 5.4

Often felt, but only causes minor damage.

5.5 to 6.0

Slight damage to buildings and other structures.

6.1 to 6.9

May cause a lot of damage in very populated areas.

7.0 to 7.9

Major earthquake. Serious damage.

8.0 or greater

Great earthquake. Can totally destroy communities near the epicenter.

Figure 2.13. (a) Historical (yellow circles) and instrumentally recorded (red circles) earthquakes from the BGS catalogue for Scotland. Circles are scaled by magnitude. (b) seismograph stations operated by BGS between 1970 to present. Note that not all station were operational at the same time. (c) Cumulative number of earthquakes as a function of time from 1970 to present. Blue line shows all recorded earthquakes. The red line shows earthquakes with magnitudes of 2.0 ML and above. (d) Annual number of earthquakes from 1970 to present. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.13. (a) Historical (yellow circles) and instrumentally recorded (red circles) earthquakes from the BGS catalogue for Scotland. Circles are scaled by magnitude. (b) seismograph stations operated by BGS between 1970 to present. Note that not all station were operational at the same time. (c) Cumulative number of earthquakes as a function of time from 1970 to present. Blue line shows all recorded earthquakes. The red line shows earthquakes with magnitudes of 2.0 ML and above. (d) Annual number of earthquakes from 1970 to present. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.13 (b) shows seismograph stations operated by BGS between 1970 and present. Note that not all stations were operational at the same time. Figure 2.13 (c) shows the cumulative number of events from 1970 to 2015. The blue line shows all recorded earthquakes, while the red line shows earthquakes with magnitudes of 2.0 ML and above. Since the natural earthquake activity rate is expected to remain constant over long periods of time, the cumulative number of events should show a linear gradient. This is approximately the case for earthquakes of magnitude of 2.0 and above, so the observed changes in gradient may be related to changes in the detection capability of the network or the inclusion of dependent events such as aftershocks in the catalogue. For example, there are jumps of seismicity rare following the occurrence times of the Kintail and Oban earthquakes, which may be related to aftershocks. Figure 2.13 (d) shows the annual number of events from 1970 to present.

2.5.2 Mining Induced Seismicity

The coalfields of Britain have frequently been the source areas of small to moderate earthquakes and tremors in these areas have been reported for at least the last hundred years, for example the Stafford earthquake of 1916 (Davison, 1919). With the growth of instrumental seismic monitoring in the UK in the 1970's, many more tremors were recorded in mining areas across the UK (Redmayne et al., 1988) and a number of temporary networks of sensors were deployed to study these events in more detail. This led to the conclusion that these events were related to ongoing mining activity and that these were quite distinct from the natural background seismic activity of the UK. Tremors around Stoke-on-Trent, Staffordshire in the period 1975-1977 were shown to originate from a region above active mine-workings (Westbrook et al., 1980). A network deployed around Rosslyn Chapel in the Midlothian coalfield recorded over 250 earthquakes between 1987-1990 that were shown to have a close spatial and temporal association with mining activity (Redmayne et al., 1998). Between July 1989 and August 1990 over 130 tremors were felt and reported by people in the Edinstowe district of Nottinghamshire. A temporary network of sensors detected a further 785 microseismic events in the following 11 months helping to establish a causal relationship between the local microseismicity and coal production (Bishop et al., 1994).

In the 1980's and 1990's mining events accounted for approximately 25% of all the earthquakes recorded in the UK (Browitt et al., 1985). Since the rapid decline of mining activity in the UK there has been a general decrease in the number of these events. The identified area of potential UOG development coincides with areas of high-density coal-mining in the Midland Valley, and since these mining induced events represent a temporary perturbation they need to be removed from the earthquake catalogue so that an accurate measure of natural earthquake activity rates can be established. We do this by defining a simple spatial filter based on the Mining Reporting Areas, as issued by the Coal Mining Authority. All events from within these areas are removed from the catalogue.

Figure 2.14(a) shows the spatial distribution of both natural and mining induced seismicity (red circles) in the Midland Valley region overlaid on the Mining Reporting Areas (grey shaded areas). Also shown are events that were identified as mining-induced during analysis (black circles). There are a large number of earthquakes within the Mining Reporting Areas, particularly in the Midlothian and Clackmannanshire coalfields. It is also clear that there are many events in these areas that have not been classified as suspected mining induced events but are likely to be so, given the strong spatial and temporal correlation with other mining induced earthquakes.

Figure 2.14(b) shows the cumulative number of both natural (green line) and coal-mining (red line) events as a function of time. The rate of both natural and mining induced events is approximately constant through the 1970's, 1980's and 1990's, with the majority of the events being of suspected coal-mining origin. By the late 1990's the number of mining events begins to fall resulting in little subsequent increase in cumulative number of events, whereas the cumulative number of natural events continues to increase at approximately the same rate as previous. With this in mind, we consider that any events in the Mining Reporting Areas post-2000 are natural events rather than mining-induced, although stoop and room collapses may occur after mining has stopped.

Figure 2.14. (a) Red circles show instrumentally recorded earthquakes (1970-2015). Symbols are scaled by magnitude. Grey shaded areas show the Mining Reporting Areas (Coal Authority data). Black circles show earthquakes identified as mining-induced during analysis. (b) Cumulative number of earthquakes as a function of time from 1970 to end of 2015. The blue line shows all recorded earthquakes. The red line shows earthquakes removed by the spatial filter and the green line shows the earthquake data after all events in the Mining Reporting Areas have been removed. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.14. (a) Red circles show instrumentally recorded earthquakes (1970-2015). Symbols are scaled by magnitude. Grey shaded areas show the Mining Reporting Areas (Coal Authority data). Black circles show earthquakes identified as mining-induced during analysis. (b) Cumulative number of earthquakes as a function of time from 1970 to end of 2015. The blue line shows all recorded earthquakes. The red line shows earthquakes removed by the spatial filter and the green line shows the earthquake data after all events in the Mining Reporting Areas have been removed. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

In theory, earthquake focal depth could also be used to discriminate between natural and mining-induced earthquakes since we expect mining-induced earthquakes to occur at shallower depths (0-2 km), whereas natural earthquakes are generally deeper. However, examination of the focal depth distribution of earthquakes in the Mining Reporting Areas (Figure 2.15(a)) shows that depths for earthquakes prior to 1982 are primarily fixed depth hypocentral solutions rather than seismicity concentrated in the first 2 km. The subsequent improvement of the network resulted in independent hypocentral determination of seismic events with focal depths spanning the active shallow crust (0-12 km). The majority of events (>80%) are shallower than 5 km depth indicating many fewer earthquakes at larger depths over this time period. However, the large uncertainties mean that the use of depth as a discriminant becomes unreliable.

Figure 2.15 (b) shows the magnitude distribution over time of earthquakes in the Mining Reporting Areas. The maximum observed magnitude is 2.6 ML, with an average reported magnitude approximately ~ ML=1.0. Furthermore, we observe a lower activity rate post-2000, suggesting that activity is returning to the background activity rate for natural earthquakes.

The maximum observed magnitudes from coal mining induced seismicity in the UK (Bishop et al., 1994 and Redmayne et al., 1998) is around 3 ML. The three largest events had magnitudes of 3.1 ML and occurred in Mansfield and Stoke. These were felt with an intensity of 4 EMS. However, other mining induced earthquakes with smaller magnitudes have been felt with higher intensities, for example, a magnitude 2.3 ML event at Rosewell in Midlothian was felt with an intensity of 5 EMS.

Figure 2.15 (a) Focal depths as a function of time for earthquakes in the Mining Reporting Areas. (b) magnitudes of mining induced earthquakes as a function of time. The deeper earthquakes may be natural seismicity. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.15 (a) Focal depths as a function of time for earthquakes in the Mining Reporting Areas. (b) magnitudes of mining induced earthquakes as a function of time. The deeper earthquakes may be natural seismicity. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

2.5.3 Earthquake Activity Rates

After removing the mining-induced seismicity, we use the revised catalogue from 1970 to 2015 to determine activity rates for background natural earthquake activity both for all of Scotland and for the Midland Valley study area. To ensure that the results are credible, we compare them with activity rates determined for historical earthquakes over the last few hundred years. These estimates of past activity rates can be used to make forecasts of future earthquake activity.

The relationship between the magnitude and number of earthquakes in a given region and time period generally takes an exponential form that is referred to as the Gutenberg-Richter law (Gutenberg and Richter, 1954), and is commonly expressed as

where N is the number of earthquakes above a given magnitude M. The constant a, is a function of the total number of earthquakes in the sample and is known as the earthquake rate. This is commonly normalised over period of time, such as a year. The constant b gives the proportion of large events to small ones, and is commonly referred to as the b-value. In general, b-values are close to unity. This means that for each unit increase in magnitude, the number of earthquakes reduces tenfold. Plotting earthquake magnitudes against the logarithm of frequency (Figure 2.16) gives a straight line, where the slope of the line is the b-value and the rate, a, is the value where the line intersects with a given reference magnitude (often zero). An observed roll-off in the number of earthquakes at low magnitudes shown by the blue crosses in Figure 2.16 is typically seen due to inability of regional seismic networks to detect small earthquakes.

This roll-off in the magnitude-frequency relationship at low magnitudes leads to the concept of a completeness magnitude, Mc, which can be defined as the lowest magnitude at which 100% of the earthquakes in a space-time volume are detected (Rydelek and Sacks, 1989). The use of too low a value of Mc will generally result in underestimates of both a and b values. With this in mind it is important to select appropriate values of Mc when calculating these parameters.

Figure 2.16. Schematic showing the number of earthquakes above a given magnitude plotted against magnitude. This shows an exponential distribution leading to the Gutenberg-Richter law. The slope of the lines is the b-value and determines the relative number of earthquakes of different magnitudes. An observed roll-off in the number of earthquakes at low magnitudes shown by the blue crosses is typically seen due to inability of regional seismic networks to detect small earthquakes. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.16. Schematic showing the number of earthquakes above a given magnitude plotted against magnitude. This shows an exponential distribution leading to the Gutenberg-Richter law. The slope of the lines is the b-value and determines the relative number of earthquakes of different magnitudes. An observed roll-off in the number of earthquakes at low magnitudes shown by the blue crosses is typically seen due to inability of regional seismic networks to detect small earthquakes. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Firstly, we use only the instrumentally recorded earthquakes in the Scottish catalogue from 1970 to 2015, with the mining earthquakes removed, to estimate earthquake activity rate by applying a penalised maximum likelihood procedure ( e.g. Johnston et al., 1994). We assume that the catalogue is complete for magnitudes of 2.0 and above. The results are shown in Figure 2.17 (a), and give an activity rate of 2.79 and a b-value of 0.95. This is equivalent to eight earthquakes with a magnitude of 2.0 or above occurring somewhere in Scotland every year.

Similarly, analysing seismicity in the Midland Valley, we find 6 events with M≥2.0 for the time period 2000-2015, corresponding to an activity rate 0.3750 events/year. Figure 2.17(b) shows the frequency magnitude distribution in Midland Valley for the time period of 1970-2015. This gives an activity rate of 2.75 and a b-value of 1.33, equivalent to 1.2 events with a magnitude of 2 or above somewhere in the Midland Valley every year. The large number of low-magnitude mining earthquakes in the Midland Valley results in the much higher b-value.

Next we determine an activity rate using only historical earthquakes in the Scottish catalogue from 1597 to 1969 using the same method, specifying different magnitude of completeness thresholds for different time intervals of the catalogue ( Table 2.3) following Musson and Sargeant (2007). In this case, we consider a minimum magnitude of 3.0. The results are shown in Figure 2.17 (b), giving an activity rate, of 2.987 and a recurrence parameter, b, of 1.015. These values are broadly consistent with the results for the instrumental catalogue only, and are equivalent to nine earthquakes with a magnitude of 2.0 or above occurring somewhere in Scotland every year.

Table 2.3. Magnitude of completeness for different periods of time used for Britain by Musson and Sargeant (2008).

Magnitude

Date

3.0

1970

3.5

1850

4.0

1750

4.5

1700

5.0

1650

6.5

1000

Figure 2.17. Magnitude-Frequency distributions for: (a) instrumentally recorded seismicity in Scotland from 1970-2015; (b) instrumentally recorded seismicity in the Midland Valley from 1970 to 2015; (c) historical seismicity in Scotland from 1597 to 1969; and (d) historical and instrumentally recorded seismicity in the British Isles. The blue squares show the observed data. The blue straight lines show the best-fit to the data for a Gutenberg-Richter distribution. The a and b values are given in the top right of each plot. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Figure 2.17. Magnitude-Frequency distributions for: (a) instrumentally recorded seismicity in Scotland from 1970-2015; (b) instrumentally recorded seismicity in the Midland Valley from 1970 to 2015; (c) historical seismicity in Scotland from 1597 to 1969; and (d) historical and instrumentally recorded seismicity in the British Isles. The blue squares show the observed data. The blue straight lines show the best-fit to the data for a Gutenberg-Richter distribution. The a and b values are given in the top right of each plot. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Finally we calculate an activity rate using the historical and instrumental earthquake catalogue for all of the British Isles using magnitude of completeness thresholds for different time intervals in Table 2.3. Again, we consider a minimum magnitude of 3.0. The results are shown in Figure 2.17 (d), giving an activity rate, of 3.67 and a recurrence parameter, b, of 1.0. Both values are in keeping with the results obtained by Musson and Sargeant (2007) using only instrumental data.

2.6 The Regional Stress Field and Fault Reactivation Potential

The pre-existing state of stress on a fault partly determines how close it is to failure. Although the magnitude of the ambient stress field in the Earth cannot be directly measured, borehole experiments from stable continental regions suggest that a critical state of stress is expected at any depth throughout the Earth's Crust (Townend and Zoback, 2000). This suggests that even small stress perturbations may cause an active fault to fail.

Earthquake focal mechanisms provide both fault geometries and principal stress directions that can be used to constrain our understanding of the driving forces of current deformation. In areas of low seismicity and sparse station distribution, determining reliable focal mechanisms can be problematic. This means that available mechanisms for earthquakes in Scotland are limited to larger events of ML≥3.5/4.0, as it is generally not possible to calculate mechanisms for smaller events. Figure 2.18 (a) shows focal mechanisms available for earthquakes in Scotland (Baptie, 2010), along with the mechanism for the 1979 Carlisle earthquake. The blue shaded areas show directions where the initial motion of the seismic waves is up (compressional), while the white areas show directions where the initial motion of the seismic waves is down (dilatational). The directions of the maximum and minimum compressive stresses are shown by the blue and white squares, respectively.

The resulting focal mechanisms are mainly strike-slip with N-S compression and E-W tension. This results in either left-lateral strike-slip faulting along near vertical NE- SW fault planes, or right-lateral strike-slip faulting along near vertical NW- SE fault planes Some of the events also have an oblique component to the slip ( e.g. Dunoon, 1985; Aberfoyle, 2002).

The World Stress Map (Heidbach et al., 2010) contains s H (maximum horizontal compressive stress) orientations for the British Isles determined from a variety of stress indicators including borehole breakouts ( e.g. Williams et al., 2015), drilling induced fracturing and hydro-fracturing as well as previously published focal mechanisms (Baptie, 2010). Borehole breakouts are enlargements or elongations in the cross-section of a wellbore in a direction parallel to the minimum (least) horizontal stress. Figure 2.18 (b) shows smoothed stress orientations on a 0.5° grid from the World Stress Map. The orientation of s H is roughly N-S in the north of Scotland, rotating to a more NW- SE direction in the north of England. This agrees reasonably with the focal mechanisms in the north of Scotland, however, the smoothed s H values are based on sparse data, most borehole breakout data are from offshore areas ( e.g. Williams et al., 2015), and there are no good stress data for the Midland Valley of Scotland.

Horizontal maximum and minimum compressive stress directions will result in strike-slip faulting on near vertical fault planes, as is widely observed in the British Isles (Baptie, 2010). If the maximum compressive stress direction is N-S then faults that strike NE- SW will be optimally oriented for left-lateral strike-slip motion, while faults that strike NW- SE will be optimally oriented for right-lateral strike-slip motion. This means that faults such as the Highland Boundary Fault, or the Great Glen fault, which strike NE- SW are most likely to show left-lateral strike-slip.

Given the observed N-S compression and E-W tension, the reactivation potential for near-vertical E-W or ENE- WSW striking faults in the Midland Valley region will be low. However, if the maximum horizontal compressive stress direction does rotate to a more NW- SE direction then the reactivation potential for these fault populations will increase.

Figure 2.18. (a) Focal mechanisms available for earthquakes in Scotland (Baptie, 2010). The blue and white areas show the compressional and dilatational quadrants and the lines between the quadrants show the strike and dip of the two possible fault planes. The axes of maximum and minimum compression are indicated by the blue and white squares respectively. The blue squares on the map show the location of the earthquakes. (b) a comparison of the stress field with mapped fault orientations. The black lines show the orientation of the maximum horizontal compressive stress, sH, taken from smoothed stress orientations published in the World Stress Map (Heidbach et al., 2010).

Figure 2.18. (a) Focal mechanisms available for earthquakes in Scotland (Baptie, 2010). The blue and white areas show the compressional and dilatational quadrants and the lines between the quadrants show the strike and dip of the two possible fault planes. The axes of maximum and minimum compression are indicated by the blue and white squares respectively. The blue squares on the map show the location of the earthquakes. (b) a comparison of the stress field with mapped fault orientations. The black lines show the orientation of the maximum horizontal compressive stress, sH, taken from smoothed stress orientations published in the World Stress Map (Heidbach et al., 2010). Green lines show mapped faults. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

Focal Mechanisms

Seismologists refer to the direction of slip in an earthquake and the orientation of the fault on which it occurs as the focal mechanism. This can be determined from recordings of seismic waves and is typically displayed as a "beach ball" symbol. The dividing lines between the quadrants on the beach-ball define the orientation of the fault planes and the directions of slip.

The schematic shows examples of focal mechanisms for four different types of faulting. If the rocks on either side of the fault slide horizontally past each other this is strike-slip faulting (a). This can be either left or right-lateral depending on which direction the rocks on the far side of the fault move. In this case the focal mechanism projection divides into four quadrants. Vertical dip-slip faulting (b) occurs when the fault plane is vertical and the fault motion is in a vertical direction. In this case the focal mechanism projection divides in two. Normal faulting (c) occurs when the upper fault block slides down relative to the lower block. Reverse faulting (d) occurs when the lower fault block slides down relative to the upper block.

Commonly, shaded or coloured areas show directions where the initial motion of the seismic waves is up (compressional), while the white areas show directions where the initial motion of the seismic waves is down (dilatational).

Green lines show mapped faults. Earthquake data from the British Geological Survey UK Earthquake Catalogue © NERC 2016.

The schematic shows examples of focal mechanisms for four different types of faulting

2.7 Discussion

The earthquake catalogue for the study area and for Scotland consists of earthquake information determined from both historical records and instrumental recordings of ground motions. The completeness of this catalogue varies in both time and space. The earliest historical record of an earthquake dates from 1597 and there are regular reports of earthquakes throughout the 17 th and 18 th centuries. The magnitude 5.2 ML Argyll earthquake in 1880 is the largest known Scottish earthquake, while the magnitude 5.1 ML Inverness earthquake in 1901 is the most damaging. The catalogue is assumed to be complete for earthquakes with magnitudes of 4.0 and above from 1750 to present.

The installation of sensors to measure ground motions from earthquakes from 1970 on resulted in a dramatic increase in the number of earthquakes recorded. Over 4000 earthquakes were recorded in Scotland between 1970 and the end of 2015. Although the completeness of the catalogue also varies in space and time, we estimate that it is reasonably complete for magnitudes of 2.0 and above. There are 363 earthquakes in the catalogue above this magnitude from 1970 to the end of 2015.

The identified area of potential UOG development coincides with areas of high-density coal-mining in the Midland Valley and there are a large number of instrumentally recorded earthquakes within the Mining Reporting Areas, as issued by the Coal Mining Authority. The dramatic decline in the number of mining induced earthquakes in the last two decades shows that the mining induced earthquakes have been a temporary perturbation, which needs to be removed from the earthquake catalogue so that an accurate measure of natural earthquake activity rates can be established. We attempted to remove these by defining a simple spatial filter based on the Mining Reporting Areas. Over 2000 mining-induced earthquakes were removed from the catalogue this way.

The low number of earthquakes observed in the Midland Valley from 2000 to present suggests that activity is returning to the background rate for natural earthquakes. While the number and frequency of mining induced earthquakes is expected to further decline in future, such earthquakes may still occur. As a result, it may be difficult to discriminate such events with events induced by other future industrial activities such as hydraulic fracturing using seismic data alone.

Earthquake activity rates for Scotland determined from the instrumental catalogue only suggest that, on average, there are eight earthquakes with a magnitude of 2.0 or above somewhere in Scotland every year. This is reasonably good agreement with activity rates determined using historical data. Activity rates in the Midland Valley area are less that the average for Scotland, although the small number of earthquakes in catalogue for this area means the values are poorly constrained. However, it is clear that there is less than one earthquake with a magnitude of 2 or above in this region annually. Activity rates in Scotland are lower than the average for the British Isles.

It is difficult to associate earthquakes with specific faults because of: (a) uncertainties in the earthquake location, especially depth, which are typically several kilometres; (b) uncertainties in fault distributions and orientation at depth; and, (c) the limited size of the earthquakes means the rupture dimensions are small. A few of studies ( e.g. Ottemöller and Thomas, 2007) have used the alignment of earthquakes from a specific sequence, along with earthquake focal mechanisms, to identify causative faults. However, most of the earthquakes in the catalogue cannot be associated with a specific fault.

The limited number of focal mechanisms for earthquakes in Scotland, along with other reliable stress indicators such as borehole breakouts means that there is a lack of detailed information on the spatial variation of stress directions across Scotland. The observed orientation of the maximum and minimum compressive stresses means that the reactivation potential of faults that strike either NE- SW or NW- SE are interpreted, based on existing data, to be highest, with the former optimally oriented for left-lateral strike-slip motion. This means that fault systems that follow the widely observed Caledonian trend may have a higher reactivation potential than other orientations. However, the reactivation potential for near-vertical E-W or ENE- WSW striking faults in the Midland Valley region is interpreted to be low based on current data. This seems in keeping with the low levels of observed seismicity in this part of Scotland. However, if the maximum horizontal compressive stress direction does rotate to a more NW- SE direction as suggested by smoothed values of s H from the World Stress Map over the Midland Valley of Scotland, then the reactivation potential for these fault populations is higher. Given the limitations of the existing data (focal mechanisms as well as other stress indicators) it is not possible to identify any particular parts of the study area where the fault reactivation potential is higher or that may present an elevated seismic hazard for any UOG operations.


Contact