Publication - Independent report

Potential for deep geothermal energy in Scotland: study volume 2

Published: 13 Nov 2013

This independent study investigates the potential for deep geothermal energy in Scotland and the steps necessary for commercialisation.

Potential for deep geothermal energy in Scotland: study volume 2
7 Hot sedimentary aquifers

7 Hot sedimentary aquifers

7.1 Introduction

Hot Sedimentary Aquifer ( HSA) settings with the potential to yield commercially viable quantities of warm water require substantial units of permeable sedimentary strata. Most of Scotland is underlain by relatively impermeable crystalline (non-sedimentary) rocks; the Midland Valley is the largest onshore area to be underlain by sedimentary strata. Reviews of the distribution, character and productivity of aquifers in the near-surface zone (to a depth of around 200 metres) in Scotland are presented in MacDonald et al. (2005), Graham et al. (2009), and Ó'Dochartaigh et al. (2011). The properties of aquifers at deeper levels are still largely unknown.

The first, and so far only, successful HSA system developed in the UK is in Southampton (Barker et al., 2000; Opened in 1986, the system exploits warm water (<80°C) at a depth of nearly 2 kilometres in sedimentary strata of the Wessex Basin. A combined heat and power ( CHP) system delivers sustainable supplies of heat (district heating), chilled water and electricity.

7.2 Previous investigation of geothermal potential in Scottish aquifers

The geothermal potential of sandstone aquifers in the Upper Devonian and Carboniferous strata of the Midland Valley of Scotland was investigated as part of the 'Investigation of the geothermal potential of the UK' project in the 1980s (Browne et al., 1985; Browne et al., 1987). The investigations included: identifying areas where thick, permeable and porous sandstone sequences might occur; considering the depositional environment (desert, river-bed and river delta) of different sandstone units, as these typically produce distinct characteristics that will influence aquifer potential; assessing aquifer properties (permeability, porosity, flow rate etc) from outcrop studies, borehole records and lab testing; and reviewing the evidence from geophysical surveys for deep sedimentary basins.

Key conclusions from these investigations are summarised below.

  • The Midland Valley is geologically complex, making it difficult to correlate between boreholes and extrapolate below the ground surface. Interpretation in some areas is complicated further by the influence on groundwater flow of abandoned and active coal mines. The aquifers are typically of variable lithology, intruded by relatively impermeable igneous rocks, fractured, faulted and generally complex.
  • The average temperature gradient (based on borehole temperature versus depth [T-z] data) for boreholes in the Midland Valley was reported to be 22.5°C/km [3] .
  • In general, water moving through aquifers at deeper levels is likely to be of relatively small volume and confined to discrete zones of flow.
  • The Knox Pulpit Sandstone Formation is the main HSA prospect in the Midland Valley. This unit, which is a component of the Upper Devonian Stratheden Group, crops out in northern Fife and may persist beneath other parts of Fife, Clackmannanshire and the Glasgow area at depths of up to 4 kilometres. The formation, which is composed of weakly cemented, wind-deposited (as opposed to water-deposited) sandstone, is around 170 metres thick but the aquifer of which it is part can be considered to include some of the overlying and underlying strata.
  • A borehole drilled in Glenrothes as part of the same project (the Glenrothes Heat Flow Borehole) showed that the favourable porosity and permeability characteristics that occur at outcrop in the Knox Pulpit Sandstone Formation are reduced at depth. However, the average horizontal permeability and transmissivity at a depth of 500 metres remained relatively high.
  • Most of the sandstone aquifers in Carboniferous strata are likely to be of little geothermal interest due to a combination of their physical properties, relatively modest thicknesses and limited regional distribution. However, the most deeply buried sandstone units in some Carboniferous sequences, notably the Passage Formation in Clackmannanshire, Fife and the Lothians, may have geothermal potential.

7.3 HSA potential based on bedrock aquifer productivity

This assessment of HSA potential in Scotland is based on published studies of bedrock aquifer productivity. Figure 24 shows how the 'productivity' of bedrock units within the near-surface zone (less than c. 200 metres below ground) varies across Scotland ('productivity' is a qualitative measure of aquifer quality, and is based on several quantitative hydrogeological parameters). The most productive ('very high productivity') rock units are confined to a number of relatively small occurrences of Permo-Triassic rocks in south-west Scotland and a single strip of Devonian sandstone in Fife. Large areas of somewhat less productive ('high' and 'moderate' productivity) sedimentary rocks of Devonian and Carboniferous age crop out across much of the Midland Valley, in the Scottish Borders area, and on the margins of the Moray Firth to the north and east of Inverness. Virtually all of the Highlands, islands and Southern Uplands are characterised by poorly productive rocks ('low' and 'very low' productivity).

The level of aquifer 'productivity' that would be required to support a commercially viable HSA scheme is likely to vary according to a range of factors, including the depth of the resource and water temperature. For the purposes of this assessment, it is assumed that only units classified as having 'very high' or 'high' productivity on Figure 24 have HSA potential (see Ó'Dochartaigh et al., 2011 for definitions). A brief description of these units is presented below (see Figure 25 for locations of the best HSA prospects described below).

An important point to consider when assessing settings with HSA potential is that the 'productivity' of a rock unit can change with depth. Sedimentary rocks with a significant proportion of intergranular (or matrix) pore space at the time they were deposited (principally sandstones and conglomerates) may be particularly prone to depth-related changes in permeability. With increasing depth, the weight of overlying rock can reduce intergranular porosity and cause near-horizontal fractures to close. The same downward force can cause near-vertical fractures to form or widen. Water chemistry (acidity, oxidising potential, salinity etc) and temperature also change with depth; these changes can cause new minerals to form in pore spaces (thereby reducing permeability) and they can cause existing minerals to dissolve (thereby increasing permeability). The net effect of all these changes on aquifer productivity will vary spatially (and over time) within the rock mass, and cannot be predicted with certainty from surface or near-surface observations alone. Testing in deep boreholes will be required to gauge the actual permeability and overall productivity at depth in any setting with HSA potential.

Figure 24 BGS map of bedrock aquifer productivity in Scotland (from MacDonald et al., 2005; Ó'Dochartaigh et al., 2011). In the key, the terms 'intergranular' and 'fracture' refer to the dominant type of pore space in the bedrock unit. See text sections 3.1 and 7.3 for details.

Figure 24

7.3.1 'Very high' and 'high' productivity aquifers Permian and Triassic rocks

Several relatively small areas of Permo-Triassic rocks in south-west Scotland have been identified as having 'very high productivity'. These rock units are geologically isolated ( i.e. they sit apart from other, similar rock units of the same age) and consist mainly of red sandstone and conglomerate, with occasional thin flows of basaltic lava. Their locations are shown on Figure 24 and summarised here:

  • In the Mauchline Basin of Ayrshire, a thickness of around 450 metres of Permian sandstone overlies up to 300 metres of interbedded igneous and sedimentary rocks.
  • The southern part of Arran is dominated by Permian and Triassic mudstone, sandstone and conglomerate to a depth of possibly several hundred metres.
  • The Stranraer Basin is asymmetric and bounded on its east side by a fault. Geophysical evidence shows that it contains a thickness of up to 1,200 metres of Permian and possibly Carboniferous strata near its eastern margin.
  • The Ballantrae Basin lies mostly offshore, but between Ballantrae and Bennane Lea in Ayrshire a thickness of some 750 metres of lower Permian strata form coastal exposures at the basin margin. The strata contain at least one layer-parallel unit of doleritic rock around 1 metre thick.
  • Geophysical modelling indicates the deepest part of the Thornhill Basin holds a maximum combined thickness of around 400 metres of Permian and underlying Carboniferous strata, in which sandstone and conglomerate are interdigitated locally with thin basalt lava flows.
  • The Dumfries Basin is a broadly symmetrical structure containing a sequence of conglomerate and red sandstone that is estimated from geophysical modelling to attain a maximum thickness of approximately 1,500 metres.
  • From geophysical evidence, the Lochmaben Basin contains a maximum thickness of around 1,300 metres of sandstone, conglomerate and rare basalt.
  • The Moffat Basin is a narrow, elongate trough containing a thickness of around 200 metres of conglomerate and red sandstone.
  • A sequence up to 700 metres thick of Permian conglomerate occupies what appears to be a palaeogeographical depression (as opposed to a fault-bounded structure) in the Snar Valley.
  • Permian to Triassic strata that crop out around Annan and Gretna form the marginal sequence to the large Carlisle Basin, most of which lies to the south of the border with England. The basin extends offshore under the innermost part of the Solway Firth.

The largest and deepest of these basins may have Hot Sedimentary Aquifer potential ( Figure 25). None of the heat flow data or borehole temperature data described in section 4 was derived from boreholes through these rocks, so there are no directly measured geothermal data. Applying the 'regional geothermal gradient' inferred from Figure 9 suggests temperatures of approximately 40-50°C should be encountered in the deepest parts of the Stranraer, Dumfries and Lochmaben basins. Layers of generally basaltic (low thermal conductivity) igneous rock, which occur locally within the mainly sedimentary strata, may act to trap heat and water locally. Devonian rocks

A strip of strata assigned to the Stratheden Group in the northern part of Fife (the Knox Pulpit Sandstone Formation referred to in section 7.2) is the only occurrence at outcrop in Scotland of Devonian rocks with 'very high productivity' ( Figure 24). The Stratheden Group consists mainly of red-brown sandstones with subordinate conglomerate and mudstone. In onshore parts of Fife the unit may attain a thickness of around 500 metres, but it extends offshore into the Firth of Tay and is considerably thicker there. In Fife, the Stratheden Group is underlain mainly by Devonian volcanic rocks.

Applying the 'regional geothermal gradient' to this setting suggests a temperature of approximately 22°C would be encountered towards the base (c. 500 metres deep) of the Stratheden Group beneath its outcrop in Fife. Stratheden Group strata are overlain locally by extensive layers of doleritic (silica-poor) igneous rocks up to several hundred metres thick forming the Lomond Hills, and these may to some degree trap heat and water in the Stratheden Group rocks beneath them.

Units of Devonian sandstone with 'high productivity' crop out within substantial areas of ground bordering the southern edge of the Moray Firth, in the northern half of the Midland Valley, and around Jedburgh in the Scottish Borders ( Figure 24).

There are no deep boreholes through the Devonian rocks adjacent to the southern edge of the Moray Firth, so there are no directly measured temperature data and the thickness of sedimentary rocks sitting on crystalline basement rocks has not been proved. The sequence in general thickens towards the coast, and the total thickness of Devonian strata is likely to reach at least 3,000 metres in places close to the coast. Applying the 'regional geothermal gradient' to this setting suggests a temperature of 107 ±23°C would be encountered towards the inferred base of the Devonian strata (at a depth of ~3 km) close to the coast.

The northern part of the Midland Valley is underlain mainly by Devonian strata, much of which has 'high productivity'. The strata originally filled the Strathmore Basin which has subsequently been deformed into a broad fold, the Strathmore Syncline. The strata consist dominantly of sandstone, with subordinate but locally substantial units of conglomerate and mudstone, and in places thin to thick, silica-poor to silica-rich lavas. The thickness of the sequence probably varies significantly across the outcrop, and in places may exceed five kilometres. If the 'regional geothermal gradient' applies to this setting, there is clearly considerable potential for HSA resources (temperatures may reach or exceed 100°C at a depth of 3 km and 150°C at 5 km), but more detailed investigation than has been possible here is required to identify specific targets with HSA potential beneath this large area.

A large area of Devonian strata assigned to the Stratheden Group crops out around Gordon and Jedburgh in the Scottish Borders region. The unit is estimated to be up to 200 metres thick and to consist of sandstone and siltstone above a basal layer of conglomerate. The temperature at the base of the unit is probably no higher than 15°C, and it is therefore unlikely to have significant HSA potential.

Figure 25 Rock units which on geological grounds appear to have good HSA potential. The Devonian and Carboniferous lavas of the Midland Valley do not have HSA potential, but locally they may overlie sedimentary strata that do. See text for details. See Figure 3 for abbreviations.

Figure 25

7.4 Onshore margins of offshore basins

The average temperature gradient in two of the boreholes represented on Figure 6 and in Table 4 is notably high compared to others in the same region: a temperature of 40.6ºC at a depth of 736 metres in the Lothbeg borehole (on the coast just south of Helmsdale in Caithness; see Figure 4) equates to an average temperature gradient of 42.9 ºC/km; and a temperature of 61.2ºC at a depth of 1,365 metres in the Archerbeck borehole (around 20 km east-north-east of the Solway Firth; see Figure 4) equates to an average temperature gradient of 37.9ºC. Both values lie above (on the 'hot' side of) the trend defined by all of the data in Figure 6 and Figure 9, suggesting the boreholes have penetrated local thermal anomalies. The Lothbeg borehole was drilled by an oil company in sedimentary rocks of Jurassic age in the coastal zone south of the Helmsdale granite intrusion, while the Archerbeck borehole intersects a thick and lithologically variable sequence of Carboniferous rocks a few miles inland of the Solway Firth. The Helmsdale granite has HHP character, and the high average temperature gradient at Lothbeg might therefore reflect elevated heat flow above buried HHP granite; however, the evidence from geophysical survey data does not indicate the presence of buried granite at depth. An alternative explanation may lie in the proximity of both boreholes to deep offshore sedimentary basins (the Inner Moray Firth Basin and the Solway Basin, respectively). Hot water from deeper parts of the offshore basins may have migrated to shallower levels in the onshore margins of the basins. The hydrogeological 'driver' for migration onshore of offshore water is not clear, but the greater density of saline (offshore) water compared to fresh (onshore) water may play a role. A further possible explanation for high temperature at the bottom of the Lothbeg borehole may lie in the proximity of the Great Glen Fault ( Figure 4). The outcrop of this major geological discontinuity lies a short distance off the east coast of Caithness, between the main part of the Moray Firth Basin and the Lothbeg borehole, and the fault may play an important role in controlling the movement of water around the basin margin. The high temperature at the bottom of the Lothbeg borehole may indicate that the Great Glen Fault in this area acts as a conduit for hot water moving upwards from depth.

The possibility that thermal anomalies exist in some onshore margins of large offshore basins needs more investigation. However, if it proves to be the case then other parts of the Scottish coastline may harbour potential HSA resources. For example, the Stranraer, Dumfries and Carlisle basins mentioned in section are all connected geologically (and presumably hydraulically) to much larger, deeper basins offshore ( i.e. they represent the onshore margins of deeper offshore basins), as do the outcrops of Permian rock at Ballantrae (Ayrshire) and in south Arran ( Figure 25).

It may prove possible in some places to access HSA settings at depth in offshore (near-shore) sedimentary basins by drilling inclined boreholes from onshore coastal locations, though this is likely to be expensive and may not be economically justifiable.

Extracting geothermal energy from salty water (from offshore settings) would involve challenges ( e.g. metal corrosion and potential contamination of fresh water aquifers) that are not presented by fresh water, but these can be overcome.

7.5 HSA prospects beneath low thermal conductivity rocks

Thick sequences of basaltic and andesitic igneous rocks (lavas and pyroclastic deposits) overlie sedimentary rocks with HSA potential in several parts of the Midland Valley. These igneous rocks typically have low productivity, and as such are very unlikely to be HSA prospects themselves. However, they also have relatively low thermal conductivity, and they may therefore act to trap both heat and water in sedimentary aquifers beneath them. Thick sequences of Devonian basaltic and andesitic rocks underlie the Ochil Hills and Sidlaw Hills in the north-east part of the Midland Valley, and the Carrick Hills and Pentland Hills (amongst other areas) in the southern part ( Figure 25). Thick sequences of Carboniferous basaltic rocks underlie the Renfrewshire, Kilpatrick, Campsie and Gargunnock hills in the west part of the Midland Valley, and the Garleton Hills in the east. In other parts of the Midland Valley, the same sequences of igneous rocks will be buried beneath younger sedimentary rocks; in these areas, sedimentary strata with HSA potential may be encountered at relatively deep levels beneath the igneous strata.

It is emphasised that there is at present no direct evidence to support the existence of thermal anomalies beneath igneous rocks in any of these areas. New heat flow measurements may reveal the presence of local thermal anomalies, and new boreholes could test whether temperature and/or aquifer productivity increases beneath a thick pile of igneous rocks. The igneous strata in the Midland Valley typically form substantial upland massifs, and it may be possible to access the sedimentary rocks beneath some massifs by drilling inclined boreholes from low ground beside the massifs rather than drilling vertically through them.

7.6 Conclusions

  1. Based on geological factors only, the Devonian strata to the east of Inverness (forming the southern onshore margin of the Moray Firth Basin) and in the northern part of the Midland Valley (notably the Knox Pulpit Sandstone Formation), and Permo-Triassic strata filling the Dumfries, Lochmaben, Stranraer and Carlisle basins in south-west Scotland, have the greatest HSA potential in Scotland. However, the permeability (and hence productivity) of a rock unit can change with depth, and the nature and effect of such change cannot be predicted with certainty from surface or near-surface observations alone.
  2. Relatively high temperatures reported from two boreholes drilled into onshore outcrops of mainly offshore basins (at Lothbeg near Helmsdale and Archerbeck near Dumfries) suggest that hot water sourced from deeper levels offshore may have moved to shallower levels in onshore parts of the basin margins; if so, then these and other parts of the Scottish coastline representing the onshore margins of offshore basins may harbour potential HSA resources. It may prove possible in some places to access HSA settings at depth in offshore (near-shore) sedimentary basins by drilling inclined boreholes from onshore coastal locations, though this is likely to be expensive and may not be economically justifiable.
  3. Virtually all of the Highlands, islands and Southern Uplands regions are characterised by rocks with 'low' or 'very low' aquifer productivity, and these rock units are likely to have little or no HSA potential.

7.7 Recommendations

For Hot Sedimentary Aquifer ( HSA) settings we recommend:

R6 More detailed investigation of parts of the Midland Valley (possibly including the Strathmore Syncline and buried parts of the Knox Pulpit Sandstone Formation) to identify specific targets with HSA potential.

R7 Further investigation of the possible thermal anomalies in some onshore margins of large offshore basins such as the Moray and Solway firths and the Firth of Clyde.

R8 Testing in deep boreholes to gauge the actual permeability and overall productivity at depth in settings with HSA potential.