Potential for deep geothermal energy in Scotland: study volume 2

8 Hot Dry Rocks (Enhanced Geothermal Systems)

8.1 Introduction

'Hot dry rock' ( HDR) resources typically will be deeper and hotter than HSA resources. Consequently, they have the potential to yield much more energy but they will present a much greater technical challenge. Unlike HSA resources, heat is extracted from 'dry' crystalline rocks by fracturing them, injecting cool water into the hot fractured rock, and extracting the resulting hot water. This requires the development of an Enhanced (or Engineered) Geothermal System ( EGS). A 'loop' consisting of boreholes at either end of a network of connected, open fractures must be developed in the hot rock, through which cold water is introduced and hot water is removed. HDR resources yield hot (100-200°C) water (or steam), and the thermal energy stored therein is converted into electricity at the surface. Where the resource already contains hot water in naturally occurring fractures, the term 'Hot Wet Rock' ( HWR) resource is used.

HDR projects are currently being developed in several parts of the world, including Australia, France, and the USA. Some of these have the potential to yield substantial amounts of energy but as yet none is being operated on a sustainable, commercial basis. Two projects to exploit HDR prospects are currently being developed in the UK, both in the granite intrusions of Cornwall: the United Downs Project ( http://www.geothermalengineering.co.uk/page/projects-and-developments.html) and the Eden Deep Geothermal Energy Project ( http://www.edenproject.com/support-us/future-plans/eden-deep-geothermal-energy).

The considerable technical challenges and associated costs of developing an HDR resource via EGS means they are likely to be considered only if the resource is big enough to generate large quantities of electricity for a number of decades. It has generally been considered that a heat reservoir temperature of at least 150°C will be required to support a commercially viable HDR project. However, the actual temperature that will be required in a future EGS project at a particular locality depends on a range of factors, including the technology available to convert heat to electricity (which is improving all the time), the size and sustainability of the heat reservoir, and the cost of accessing it (which depends, amongst other things, on its depth and the technology used for drilling and hydraulic fracturing). Additionally, the rising cost of other energy sources and the growing environmental need to develop sources of clean, renewable energy, have the effect of reducing the minimum temperature at which an HDR prospect can be considered commercially viable. In recent years the development of binary cycle power plants, in which electricity can be generated using water that is cooler than 100ºC, has greatly improved the potential for recovering geothermal energy from HDR prospects that previously would have been considered marginal or not viable.

The 'regional geothermal gradient' for Scotland described in section 4.2, suggests a temperature of 150°C would be reached at a depth of approximately 4,000 metres, which is within the widely quoted practical lower limit for exploiting HDR resources (5,000 metres). If the more conservative geothermal gradient suggested by borehole temperature data from onshore boreholes is used (30.5°C/km) then 150°C should be encountered at approximately 4,900 metres, still within the 'accessible zone'. However, the HDR concept is most suited to crystalline rocks, and only 21 of the 133 temperature data that define the trend displayed in Figure 9 were measured in crystalline rocks (15 in offshore boreholes and 6 in onshore boreholes). None of the 21 data lies off the trend, so there is at present no direct evidence from borehole temperature data that the regional geothermal gradient in crystalline rocks is different to that in sedimentary rocks. Nevertheless, the lack of temperature data for crystalline rocks in deep (> 2 km) onshore boreholes means that caution should be exercised in applying the regional temperature gradient of Figure 9 to potential HDR settings.

Some parts of the crust contain thermal anomalies, wherein the size of the local heat resource exceeds that which is produced by simple 'background' heat flow. Finding thermal anomalies at accessible depths should increase the chance of developing viable EGS schemes, because the average geothermal gradient in the crust overlying them is higher than the 'background' gradient. This section therefore focuses on whether and where such anomalies might exist.

In Scotland, three situations can be envisaged in which a 'hot' thermal anomaly might be produced in 'dry' crystalline rocks:

(i) where the background heat flow is augmented by additional heat generated in situ;

(ii) where the upward flow of heat is impeded, such that some of it becomes trapped.

(iii) a combination of (i) and (ii).

Intrusions of High Heat Production ( HHP) granite are likely to be the only source of significant additional heat generated in situ within the crust beneath Scotland. For the purposes of this assessment the threshold Heat Production ( HP) value above which rocks are considered to have HHP character is 4.0 μW m ^-3 ( section 2.1.2). Rock units with low thermal conductivity provide perhaps the likeliest situation in which the upward flow of heat might be impeded. The potential for both situations to exist in Scotland is discussed below.

8.2 High Heat Production ( HHP) granite

8.2.1 The granite intrusions of Scotland

Distribution, origin and HP values

Onshore Scotland has more than 200 intrusions of granite (and similar rocks) at surface which have an area at outcrop of more than 1 km ². The largest (the Rannoch Moor intrusion) crops out over an area of c. 380 km ². The intrusions crop out in many parts of the country; however, the greatest concentration (including most of the largest intrusions) is in the block of crystalline rocks bounded by the Highland Boundary Fault and the Great Glen Fault ( Figure 3 and Figure 26).

These intrusions have commonly been referred to individually and collectively as 'granites'. However, many contain, or consist of, rock types other than granite sensu stricto ^[4] (notably granodiorite and diorite, which are related to granite), and some contain no granite (sensu stricto) at all ( Table 11). The lithological distinction is important, because the chemical elements K, Uand Th (which produce heat by radioactive decay) generally only reach concentrations that might have geothermal significance in granite sensu stricto; concentrations of these elements are typically too low in petrologically similar but less geochemically evolved rock types like granodiorite and diorite.

The granite intrusions of Scotland formed in association with several geological events.

A small proportion (less than 10) formed in association with a period of crustal stretching that preceded and accompanied the opening of a major ocean (the Iapetus Ocean), around 600 million years ago. These intrusions are granite sensu stricto, they crop out in the Grampian Highlands and Northern Highlands, and they are typically small (with one exception, the Carn Chuinneag intrusion in Easter Ross). There are no published HP data for any of these intrusions, and no published geochemical analyses that include Uand Th. However, the geochemical data that have been published suggest the intrusions are not as compositionally evolved as HHP granite, and therefore are unlikely to have HHP character.

By far the largest proportion of Scotland's granitic intrusions formed in association with the Caledonian Orogeny, which represents geological events associated with closure of the Iapetus Ocean between 500 and 400 million years ago. Broadly speaking, intrusions of granite and granite-like rocks formed at three different stages of the Caledonian Orogeny, each one associated with a major crustal collision event ( Table 11):

• the Grampian Event (c. 490-460 million years ago) produced several large bodies of granite and granodiorite in the region south of Inverness and near Aberdeen. All these intrusions have low HP values, in the range 0.6-2.2 μW m ^-3 (data from six intrusions).
• The Scandian Event (c. 435-420 million years ago) produced numerous intrusions of granite, granodiorite and diorite that occur in greatest concentration in the Grampian Highlands region. HP values range from 2.2-7.3 μW m ^-3 (data from sixteen intrusions) in those intrusions that consist exclusively, or very largely, of granite sensu stricto. In eight of these (the Helmsdale, Fearn, Abriachan, Bennachie, Mt Battock, Ballater, Cairngorm and Monadhliath intrusions), the mean values of HP data match or exceed the threshold for HHP character used in this assessment (4 μW m ^-3); these intrusions can therefore be designated as having HHP character. Parts of four other intrusions (the Strontian, Ballachulish, Etive and Lochnagar intrusions) also exceed the threshold, though the mean value does not.
• Two large intrusions (Fleet and Criffel) were emplaced in what is now southern Scotland, at the end of the Caledonian Orogeny (c. 410-390 million years ago). The mean HP values are 3 μW m ^-3 and 2.2 μW m ^-3 respectively. However, both intrusions are compositionally zoned, and the central part of the Fleet intrusion may approach the HHP threshold.

Several intrusions c. 350 million years old crop out on Shetland. Each typically contains a diverse range of rock types, from ultramafic (very silica-poor) rock to granite, and their origin is not well understood. There are no published HP data for these intrusions, and no published geochemical analyses that include Uand Th. However, the geochemical data that have been published suggest the intrusions are not as compositionally evolved as HHP granite, and therefore are unlikely to have HHP character.

A number of granitic intrusions formed in Scotland around 60 million years ago as a result of igneous activity that was a precursor to, and accompanied, the opening of the North Atlantic

Figure 26 Bedrock geology map of Scotland showing the location of granite intrusions referred to in Table 11 and elsewhere in this report. Map based on BGS Bedrock Geology UK North 1:625 000 map.

Numbers refer to Caledonian intrusions of granitic-rock: 1 Strath Halladale; 2 Helmsdale; 3 Lairg-Rogart; 4 Grudie; 5 Migdale; 6 Fearn; 7 Abriachan; 8 Cluanie; 9 Strontian; 10 Ross of Mull; 11 Moy; 12 Ardclach; 13 Ben Rinnes; 14 Grantown; 15 Findhorn; 16 Foyers; 17 Strathspey; 18 Strath Ossian; 19 Rannoch Moor; 20 Ben Nevis; 21 Ballachulish; 22 Etive; 23 Strichen; 24 Peterhead; 25 Bennachie; 26 Coull (Cromar); 27 Hill of Fare; 28 Skene (Crathes); 29 Aberdeen; 30 Mt Battock; 31 Ballater; 32 Glen Gairn; 33 Lochnagar; 34 Cairngorm; 35 Monadhliath; 36 Priestlaw; 37 Distinkhorn; 38 Carsphairn; 39 Loch Doon; 40 Fleet; 41 Portencorkrie; 42 Criffell. HBF = Highland Boundary Fault; GGF = Great Glen Fault. High Heat Production intrusions in bold.

Pale pink and dark pink polygons flanking the Inner Moray Firth are Old Red Sandstone sedimentary rocks representing onshore remnants of materials deposited in the former Orcadian Basin. The Orcadian Basin extended across what is now the Inner Moray Firth, the Orkney Islands and Shetland.

Table 11 Heat production values of the largest intrusions that formed in Scotland during the Caledonian Orogeny. Grey shading highlights those that consist largely or entirely of granite sensu stricto at outcrop.

Intrusion (a)	Event (b)	HP (c)	n (d)	Rock type (e)	Area (f)
Strath Halladale	S	(1.4)	(22)	qm, gd, g	250
Helmsdale	S	4.1	97	g	98
Lairg-Rogart	S	1.4	17	qmd, gd, g	70
Grudie	S	3.8*	3	g	6
Migdale	S	2.4	9	g	20
Fearn	S	5.1*	4	g	33
Abriachan	S	4.0*	2	g	4
Cluanie	S	1.1	10	gd	18
Strontian †	S	1.9	37	gd	200
Ross of Mull	S	1.5*	4	g, gd	140
Moy	G	2.1*	6	g, gd	56
Ardclach	G	1.4*	4	g, gd	63
Ben Rinnes	S	3.2*	6	g	55
Grantown	G	1.3	6	g	25
Findhorn	S	1.4	10	gd, qd, d	90
Foyers	S	1.1	72	d, qmd, qd, gd, g	80
Strathspey	G	0.6	5	g	36
Strath Ossian	S	2.2	6	qd, gd, g	112
Rannoch Moor	S	1.7	34	md, qmd, gd, g	380
Ben Nevis	S	1.8	18	qd, t, gd, g	40
Ballachulish †	S	3.2	17	d, md, g	30
Etive †	S	1.9	236	qd, md, g	300
Strichen	G	2.2*	2	g	50
Peterhead	S	2.2	18	g	136
Bennachie	S	5.7 ( 7.0)	32 (8)	g	55
Coull (Cromar)	S	2.6*	2	g	17
Hill of Fare	S	3.9*	2	g	37
Skene (Crathes#)	S	1.6	17	t, gd, g	240
Aberdeen	G	2.2	3	g	90
Mt Battock	S	5.0 ( 4.8)	48 (6)	g	370
Ballater	S	5.7 ( 6.8)	34 (9)	g	48
Glen Gairn	S	2.8*	8	g	62
Lochnagar †	S	2.7*	14	g	150
Cairngorm	S	5.0 ( 7.3)	233 (9)	g	365
Monadhliath	S	5.7*	6	g	113
Priestlaw	S	1.4*	2	d, gd, g	8
Distinkhorn	S	2.0*	4	d, gd	8
Carsphairn	S	2.2*	17	d, gd, g	13
Loch Doon	S	2.5	164	d, gd, g	200
Fleet		3.0	146	g	150
Portencorkrie	S	2.1	4	d, gd, g	5
Criffell		2.2	18	gd, g	200

(a) † Lee et al. (1984) reported that parts of these zoned intrusions have heat production values exceeding 4 μW/m ^-3. # the 'Skene' intrusion is now divided into the Crathes, Tillyfourie and Kemnay intrusions. (b) G = Grampian, S = Scandian. (c) Heat production data from Lee et al. (1984), Tables 5.1 and 8.1. All values are the mean of surface heat production data ( μW/m ^-3), except values in brackets which are the 'preferred' values of Wheildon et al. (1984) and Lee et al. (1984) measured in borehole samples with corrections for topography. Values in bold text exceed the threshold for HHP plutons, as defined here (4 μW/m ^-3). * The distribution of samples and variability of data was considered by Lee et al. (1984) to be inadequate for deriving confidently a satisfactory value. (d) Number of samples used to calculate the mean. (e) Dominant rock type in bold: d = diorite; md = monzodiorite; qd = quartz-diorite; qmd = quartz-monzodiorite; qm = quartz-monzonite; t = tonalite; gd = granodiorite; g = granite. (f) Approximate area at outcrop, in km ².

Ocean. They crop out exclusively off the west coast, and are exposed on the islands of Skye, Rum, Mull, Arran and Ailsa Craig. A significant number of intrusions of this age also crop out on the sea floor off the west coast of Scotland. They all consist of, or contain, granite sensu stricto; however, there are no published HP data and no published geochemical analyses that include Uand Th. The geochemical data that have been published suggest the intrusions are unlikely to have HHP character.

In summary, the existing dataset of HP values (which is far from complete) suggests that HHP character is a feature mainly of the granite intrusions associated with the late Caledonian Scandian Event in Scotland. Granite with HHP character is known to occur in twelve intrusions that formed during this event: in eight of these the mean values of surface-derived HP data exceed the threshold for HHP character used in this assessment (4 μW m ^-3); the remaining four are zoned intrusions in which some part of the intrusion exceeds the threshold.

The distribution of HHP rock in HHP granite intrusions

A typical granite intrusion is an enormous three-dimensional rock mass. The HP capacity of rocks is typically quantified by analysing rock samples collected at outcrop or from shallow boreholes. Such samples only represent a two-dimensional 'slice' through the rock in the intrusion, and it is generally not possible to say how representative the results are of the intrusion as a whole. Single intrusions commonly consist of two or more chemically distinct batches of magma and some of these may not be exposed at outcrop. Furthermore, magma usually changes its chemical composition as it moves through the crust. The net result is that the composition of rocks can change from place to place, horizontally and vertically, within a single intrusion. In intrusions where HHP granite is exposed at outcrop, it is generally not possible (in the absence of a deep borehole) to predict with certainty to what depth and within what volume of rock the HHP character persists. The presence of HHP rock in some part of the outcrop of an intrusion does not mean that the intrusion as a whole has HHP character. Conversely, the absence of HHP rock in the outcrop of a granite intrusion does not mean that there is none at deeper levels (though it is probably unlikely in most cases). Detailed geochemical mapping of the outcrop, or heat flow measurements in shallow boreholes across the outcrop, might help to determine whether HHP rocks are likely to exist at depth, but ultimately a deep borehole will be required to prove it.

The top (shallowest) part of an intrusion is called the roof zone, and the most geochemically evolved rocks in a granite intrusion are commonly found in, or towards, the roof zone. The highest concentrations of the radiothermal elements K, Uand Th are commonly found in the most geochemically evolved rocks, so in some intrusions the rocks with the highest HP values are in the roof zone. The thickness of roof zones is in general difficult to gauge; in intrusions where the roof zone is exposed at outcrop, the top of the roof zone has generally been eroded away ( Figure 27) and the base is commonly gradational rather than abrupt. The thickness of roof zones is likely to vary considerably between intrusions, and probably also across individual intrusions. Based on limited field and borehole information for large granite intrusions in Scotland, the roof zone in some cases may attain a thickness of 1 to 2 km.

Large intrusions can span a depth range of more than 10 kilometres at the time they are emplaced in the crust. All the intrusions that are currently exposed at outcrop in Scotland have been eroded to some degree: in some, the current land surface intersects the roof zone while in others it is below the roof zone ( i.e. the roof zone has been removed by erosion). Intrusions that are currently not exposed ( i.e. are concealed beneath other rocks) may never have been exposed, in which case the roof zone will still be intact at the top of the intrusion ( Figure 27). Alternatively, they may have been exposed and partially eroded in the past, then buried beneath sedimentary rocks; in such cases the roof zone may have been removed by erosion.

It is generally difficult to determine from outcrop information alone what part of an intrusion the currently exposed level represents. However, some of the geochemical and physical features of an exposed granite intrusion may indicate that the exposed level is in, or near to, the roof zone. These include: the presence within the granite of blocks of the rock that was originally above the intrusion and were detached and incorporated by the invading granite magma; features that point to a rapid increase in magma vapour content, such as small cavities in the rock, veins of quartz or coarse granitic rock, and associated chemical alteration; and textural evidence for de-gassing or venting of the magma chamber.

Figure 27 Diagram used on the front cover of all reports in the series 'Investigation of the geothermal potential of the UK' (Downing and Gray, 1986), illustrating two contrasting concepts for HDR projects. On the right, an HHP granite intrusion extends from basement to outcrop (current surface), with no barrier to the upward flow of heat. On the left, an HHP granite intrusion is buried beneath a thick layer of low thermal conductivity sedimentary rocks, which will impede the upward flow of heat, causing it to pond and form a heat reservoir. In this case a thin layer of metamorphic rocks, into which the granite intrusions are emplaced, separates the top of the intrusion from the base of the sedimentary rocks. Superimposed isotherms illustrate the increase in geothermal gradient that would be associated with the two HHP granite intrusions relative to their country rocks. The intrusion on the right has lost some, or all, of its roof zone. The intrusion on the left has never been exposed and eroded; hence, it retains its entire roof zone. The diagram is not to scale, but the long side would be in the order of 20-30 km and the short side (depth) 15-20 km.

HHP granite intrusions can be exposed at Earth's surface or buried beneath other lithologies ( Figure 27). In terms of their geothermal energy potential, the two settings have contrasting advantages and disadvantages: those exposed at the surface are easy to find and characterise, but the heat generated within and passing through them will have been dissipating into the atmosphere for many millions of years, and they may have lost much of their HHP rock through erosion; by contrast, those buried beneath other rocks will be difficult and relatively expensive to find and characterise, but they have the potential to yield far larger stores of geothermal energy because they may retain their original roof zone (which may be rich in radiothermal elements), and the overlying rocks may have impeded the upward flow of heat from them, with the result that a reservoir of trapped heat has developed. An example of one such 'buried granite' setting that is currently being developed for geothermal energy (Cooper Basin, Australia) is described in Appendix 1.

The possibility of applying the HDR concept to exposed HHP granite plutons in Scotland was explored in the late 1970s and early 1980s. A summary of that project is presented below.

8.2.2 Previous investigation of HDR potential in exposed HHP granite intrusions

Introduction

A wide-ranging programme to investigate the geothermal potential of the UK was undertaken with government support in the late 1970s and early 1980s (results are summarised in Downing and Gray, 1986).

A heat flow map based on the Geothermal Map of the UK ( BGS, 1986) shows the highest heat flow values in the UK to be associated with clusters of granite intrusions in the south-west of England, northern England, and the East Grampians region of Scotland ( Table 3). Strongly negative anomalies in regional gravity data (which reflect changes in the density of rock in the crust) are associated with each of these areas ( Figure 28), from which it is inferred that very large volumes ('batholiths') of granite underlie each cluster of exposed intrusions: the Cornubian Batholith in south-west England, the Lake District and Weardale batholiths in northern England, and the East Grampians Batholith (sometimes referred to as the Eastern Highlands Batholith) in Scotland. In other words, the granite intrusions currently exposed at the surface in these areas appear to be just the surface expressions of much larger volumes of concealed granite that underlie each region.

The granite intrusions of Cornwall and Devon (which are of Carboniferous age and believed to be surface expressions of the Cornubian Batholith) were the first in the UK to be recognised as potential HDR targets ( e.g. Dunham, 1974). An HDR research project funded by the then Department of Energy ( km) and the European Commission was initiated in 1977 on the Carnmenellis granite intrusion in Cornwall, which has the highest known heat flow in the UK (~120 mW m ⁻²). The project, based at Rosemanowes Quarry, was essentially a rock mechanics investigation (involving deep drilling, fracture stimulation and flow-testing) of how to create a fractured reservoir (Batchelor, 1987; Richards et al., 1994); the intention was not to produce a power-generating system. The outcomes have been used in all subsequent EGS projects.

The potential of the HDR concept in Scotland was assessed (as part of the UK-wide assessment) through a collaborative investigation by IGS (now BGS), Imperial College and the Open University (Rollin, 1982, 1984; Lee, 1984; Webb and Brown, 1984; Wheildon et al., 1984; Lee et al., 1984, 1987). There are no known major Carboniferous granite intrusions in Scotland, and therefore no direct analogues of the south-west England intrusions. However, a moderately high heat flow value (95 mW m ^-2) for the Early Devonian Weardale granite in northern England (England et al., 1980) focussed attention on other intrusions of similar age, of which there are many in Scotland ( e.g. Brown et al., 1979). The East Grampians Batholith ( EGB) is inferred to underlie an east-west trending zone extending inland from near Aberdeen as far west as Strathspey ( Figure 28 and Figure 29). Many large granite intrusions of Silurian and Devonian age crop out within this zone ( Figure 26), several of which have HP values well above the HHP threshold. Four intrusions were selected for detailed study; that study, which was part of a much wider programme of work carried out under the broad title 'Investigation of the geothermal potential of the UK', is summarised below and is referred to hereafter as 'the km HDR project'.

The East Grampians intrusions

The Cairngorm, Mt Battock, Ballater and Bennachie granite intrusions were selected for detailed investigation in the km HDR project because they were considered to be the most promising HDR prospects in Scotland, based on their medium to large size, accessibility, and high HP values ( Figure 26 and Table 12). Other intrusions with promisingly high HP values, notably the Monadhliath granite intrusion, were not included due to poor accessibility. All the intrusions underlie large upland massifs, and two - Cairngorm and Mt Battock - are among the largest at outcrop of any granite intrusions in Scotland ( Table 11).

The four intrusions share many features. They consist of granite sensu stricto with subordinate proportions of coarser (pegmatitic) and finer (microgranitic and aplitic) granitic rock. Variations in grain-size and in the degree to which large crystals (phenocrysts) of feldspar are developed are characteristic features of all the intrusions, and this textural heterogeneity is the main basis for recognising and mapping internal divisions within the intrusions. Zones of rock that have been altered by hot water (hydrothermally altered rock) are common; these probably formed shortly after the granite magma was emplaced and solidified. The passage of hot water through these zones has produced a range of minerals formed by alteration (notably hematite, epidote and chlorite), veins of quartz, and joints. These act to weaken the rock mechanically, lower its thermal conductivity and raise its permeability.

Approach

The investigations focussed on characterising the thermal properties ( HP, heat flow and thermal conductivity) and lithological character of the intrusions at and near the ground surface, modelling the extent and volume of granite at depth, and combining these datasets to produce modelled thermal profiles through the intrusions. A single vertical borehole was sunk to around 300 metres in each intrusion ( Figure 29). Core was recovered from each borehole in three short (<7 metre) sections at approximately 100, 200 and 300 meters (amounting to ~5% of the total drilled depth). HP values for surface samples were affected by the fact that uranium is mobile in the surface and near-surface environment (leading to variations in measured uranium concentration that don't reflect those below the near-surface zone), so a second set of 'preferred' HP values calculated from unaltered rock recovered in core discs and chippings was used for modelling. Heat flow values were calculated from depths below 100 metres in the boreholes, to avoid near-surface perturbations caused by the influx and movement of rainwater. The reported heat flow values incorporate a small correction for the effect of local topography, but a correction to account for the effect of recent climate change (see section 2.1.4) was not deemed necessary. The distribution of radioelements (K, Uand Th) and the nature and possible origin of spatial variations in rock composition were assessed from whole-rock geochemical analyses of surface samples and core materials (Webb and Brown, 1984; Webb et al., 1985). The extent and shape of granitic rocks in the subsurface was modelled from gravity data (Rollin, 1984). Finally, thermal models drawing on all these data were generated, from which subsurface temperatures were predicted and HDR potential was assessed (Wheildon et al. 1984; Lee et al., 1984).

Figure 28 BGS colour shaded-relief image of the UK gravity field (Bouguer anomaly onshore, free-air anomaly offshore) illuminated from the north. 1- East Grampians Batholith, 2 - Lake District and Weardale batholiths, 3 - Cornubian Batholith.

Figure 29 Inferred subsurface extent of the East Grampians Batholith (labelled 'Eastern Highlands batholith' in this map), based on geophysical survey data. From Lee et al., 1984, Figure 1.4.

Results and conclusions

Thermal data for the East Grampians intrusions are summarised in Table 12, with comparable data for the other UK HHP intrusions that were included in the km HDR project. The East Grampians intrusions have the highest HP values but the lowest heat flow values; the heat flow values are only moderately elevated (~30% higher) with respect to the average value for the UK (54 ±12 mW m ^-2; Wheildon and Rollin, 1986). The range of thermal conductivity values is broadly similar in all three areas, reflecting the fact that all the boreholes are in granite.

The association of promisingly high HP values and surprisingly low heat flow values in the East Grampians intrusions was interpreted to reflect two factors: (i) a decrease in HP capacity with depth that is much more rapid in the East Grampians intrusions than in the intrusions of northern and south-west England, and (ii) relatively low background heat flow in the region. The East Grampians Batholith is modelled from gravity data to be around 13 km thick (Rollin, 1984), ruling out a possible alternative explanation that granite only extends to a depth of 6-7 km. A rapid decrease in HP with depth was ascribed to rapid diminution in the concentrations of U, Th and K with depth; in other words, the HHP character of the intrusions was interpreted to be just a near-surface feature. This interpretation was based to a large extent on a theoretical understanding of granite magma evolution; the geochemical data provided by samples from the relatively narrow vertical range spanned by the borehole and surface exposures did not yield conclusive evidence to support the interpretation (Webb and Brown, 1984; Webb et al. 1985).

Table 12 Thermal data measured in boreholes and cores for selected UK HHP granites.

Region	Pluton	HF (1)	HP (2)	TC (3)
East Grampians	Cairngorm	69.5	7.3	3.5
	Mt Battock	58.7	4.8	3.0
	Ballater	71.4	6.8	3.2
	Bennachie	75.8	7.0	3.5
Northern England	Weardale	95.4	3.7	3.1
	Wensleydale	65.0	3.3	3.6
	Shap	77.8	5.2	2.9
	Skiddaw	100.9	4.2	3.5
South-west England	Carnmenellis (a)	116.2	4.1	3.3
	Bodmin (b)	116.1	4.2	3.3
	Land's End (c)	125.4	5.5	3.4
	St Austell (d)	126.2	4.2	3.3
	Dartmoor (e)	114.0	5.3	3.2

Modified from Table 8.1 in Lee et al. (1984). (1) HF = heat flow, in mW m ^-2; (2) HP = heat production, in μW/m ^-3; (3) TC = thermal conductivity, in Wm ^-1 K ^-1; (a) mean of data from seven boreholes; (b) mean of data from five boreholes; (c) mean of data from three boreholes; (d) mean of data from two boreholes; (e) mean of data from five boreholes.

Models extending to a depth of thirty kilometres were generated from a combination of gravity data and both measured and inferred thermal property data for the granite intrusions and the surrounding rocks (Wheildon et al., 1984). In satisfying the measured HP and heat flow values at and near the surface, the models predicted temperatures of 85-98ºC at 5 km depth and 130-149ºC at 9 km. The geothermal gradient was therefore modelled to be approximately 17 ºC/km, approximately half that in the south-west England intrusions and much too low to yield a viable heat reservoir within an accessible depth range. On this basis, the authors concluded that the HHP granite intrusions of the East Grampians region should be ruled out as an HDR prospect.

8.2.3 Reassessment of HDR potential in exposed HHP granite intrusions

The conclusions of a brief review of the km- HDR project research into the East Grampians intrusions can be summarised as follows.

• Geochemical data from the East Grampians intrusions do not provide good evidence for strong vertical fractionation of radiogenic elements within the sampled depth range. However, field evidence suggests the present outcrop surface is close to the roof zone in each intrusion, so the concentrations of radiogenic elements in the East Grampians intrusions might therefore diminish quite rapidly below the present ground surface; however, the complexity of the magma emplacement history at the exposed levels (as revealed by detailed geological mapping of the intrusions since the km HDR project) means this may only become apparent over a vertical interval significantly greater than that penetrated by the boreholes (300 metres).
• HP values reported for other intrusions that are, in a geological sense, closely related to the four East Grampians intrusions are in the range 2.6-4.9 μW/m ^-3. These intrusions appear to be exposed well below their roof zones, suggesting that HP capacity in the (concealed) main body of the East Grampians intrusions may decrease to within the range 2.6-4.9 μW/m ^-3. Many of the granite intrusions of south-west England and northern England, which are associated with significantly higher heat flow values than the East Grampians intrusions, have HP values within this range. Hence, strongly vertically stratified Uand Th content may account for the relatively high HP values in exposed parts of the East Grampians intrusions, but it probably does not explain adequately why heat flow values are significantly lower than those over the Cornubian and Lake District batholiths.
• The absence of reliable measurements of background heat flow in Scotland means it is currently not possible to confirm whether background heat flow is lower in the East Grampians region (or any part of Scotland) than in south-west England.
• Heat flow values determined from boreholes less than 2 km deep could be underestimated by up to 60%, because of the effect of climate change on the top part of the geothermal gradient, provides perhaps the most likely explanation for the relatively low heat flow values over the East Grampians Batholith compared to more southerly locations in the UK, and particularly south-west England which lies south of the limit of glaciation (see section 2.1.4).

Scotland lay towards the southern limit of the ice sheets, so the geothermal gradient may have been perturbed less here than in more northerly latitudes. An upward correction of 30% (half of the estimated maximum value of 60%) to account for the effects of climate change takes the heat flow value for the Cairngorm intrusion to 90 mW m ^-2, significantly above the mean heat flow across all continents (65 mW m ⁻²), but still significantly below values for the south-west England intrusions. 90 mW m ^-2 equates to a geothermal gradient of approximately 25 ºC/km, which is significantly lower than that likely to yield a commercially viable heat reservoir at an accessible depth.

Thus, on the basis of existing heat flow measurements, the potential for developing commercially viable HDR projects within exposed granite intrusions in Scotland appears to be poor.

However, the borehole temperature data for boreholes sited in the East Grampians intrusions (and in crystalline rocks elsewhere in Scotland) sit on the trend defined by all borehole temperature data ( Figure 9), raising the possibility that the temperature (and heat resource) at depth is significantly greater than is suggested by heat flow measurements. Exposed intrusions that are known, or suspected, to contain a significant volume of HHP granite should not therefore be ruled out as potential HDR prospects until the shape of the geothermal gradient below the climate-affected zone has been established in one or more deep boreholes. Based on surface HP capacity values and intrusion size at outcrop, the most promising intrusions are: Cairngorm, Bennachie, Ballater, Monadhliath, Fearn, Mt Battock and Helmsdale ( Table 11 and Figure 30).

Figure 30 Locations of exposed High Heat Production ( HHP) granite intrusions. See Figure 3 for abbreviations.

8.2.4 Buried intrusions of HHP granite

Introduction

The occurrence at outcrop in Scotland of intrusions with HP values at and above the HHP threshold raises the possibility that substantial heat reservoirs exist where HHP granite intrusions are buried beneath a thick cover of low thermal conductivity rocks.

To date, only two examples of buried large granite intrusions have been proved onshore in the UK: the Early Devonian Weardale and Wensleydale intrusions, both in northern England ( Figure 29). Neither exceeds the 4 μW m ^-3 HP threshold used in this report to denote HHP character, but heat flow associated with the Wensleydale intrusion is similar to that in the East Grampians intrusions, and in the Weardale intrusion it is significantly higher ( Table 12). The Weardale intrusion is inferred from geophysical data to consist of five connected masses of granite (Bott, 1967) and is concealed beneath several hundred metres of Carboniferous sedimentary rocks. The temperature in the top part of the intrusion is estimated to be higher by around 7ºC than the same granite would be at outcrop, because of the relatively shallow burial beneath low conductivity cover rocks (England et al., 1980). A modelled temperature profile, based on data gathered from the Rookhope borehole which penetrates the shallowest levels in the centre of the intrusion, predicted an average geothermal gradient between the top of the intrusion and a depth of 7 km of ~31 ºC/km and temperatures of 150 to 200ºC between 4.4 and 6 km, respectively (Lee et al., 1984). More recently (in 2004), the UK's first deep (995 m) geothermal exploration borehole for more than 20 years was drilled at Eastgate in County Durham (Manning et al., 2007), penetrating 723 metres of the Weardale granite beneath 267 metres of Carboniferous strata (and the Whin Sill, a large sheet-like intrusion of dolerite). The extent to which heat has accumulated beneath the sedimentary rocks was not reported (it is likely to be relatively small given the thinness of the cover), but a reported relatively high mean geothermal gradient of 38 ºC/km from the Eastgate borehole is likely, at least in part, to reflect the 'buried hot granite' setting. The thickness of the sedimentary cover over these intrusions is much thinner than the 3-5 km that would probably be required of a major HDR prospect, but they illustrate well the concept and its potential.

Granite has a relatively low density compared to many other rock types, so large granite intrusions that are concealed in the subsurface may generate negative gravity anomalies that can be identified in regional geophysical surveys. The hot magma within an intrusion can affect the magnetic character of the rocks enclosing it, so concealed intrusions may generate positive magnetic anomalies that can also be detected by regional geophysical surveys. The following assessment of potential 'buried hot granite' settings in Scotland (see Figure 31 for locations) is based largely on an assessment of current BGS bedrock geology maps, the BGS 1:500 000 series gravity and magnetic anomaly maps ( BGS 1997; 1998), and the gravity modelling of Rollin (1984). The largest exposed HHP granites of the East Grampians Batholith (Cairngorm, Mt Battock and Monadhliath) are associated with positive magnetic anomalies of 150-350 nT and negative gravity anomalies.

Figure 31 Onshore parts of Scotland considered most likely to overlie buried HHP granite intrusions. See Figure 3 for abbreviations.

The East Grampians Batholith

Using the BGS gravity anomaly map for the region and published rock density data for key rock types, Rollin (1984, Fig 5.1) generated a three-dimensional geological model of the Eastern Highlands region showing depth contours on the upper surface of the East Grampians Batholith ( EGB) and nearby large intrusions of basic igneous rock. In the model, the top surface of the batholith undulates considerably and extends from above ground (in the exposed intrusions) to depths of more than 12 km below sea level. No substantial areas of non-crystalline (sedimentary) rocks overlie the modelled subcrop of the EGB, so the classic 'buried hot granite' setting (low thermal conductivity sedimentary rocks overlying hot granite) does not exist here. Instead, the model shows granite underlying metamorphosed sedimentary rocks at a range of depths over a very large area. The thermal conductivity of the metasedimentary rocks (~3.5 Wm ^-1 K ^-1) is broadly the same as in the granite (3.0-3.5 Wm ^-1 K ^-1) (Wheildon et al., 1984), so it is unlikely that the metasedimentary rocks will act as an effective 'thermal blanket' over the buried granite. However, the area encompassed by the model includes some settings with HDR potential:

(i) By analogy with the exposed East Grampians intrusions, parts of the concealed roof zone of the EGB may be affected by intense hydrothermal alteration ( section 8.2.2). The hydrothermally altered rock is likely to have lower thermal conductivity than either the fresh granite or the overlying metasedimentary rocks, and if they are thick enough such zones may act to impede the upward transfer of heat across the interface between the granite and overlying metasedimentary rocks. Reservoirs of trapped heat may exist below (and within) such features. Hydrothermal fluids and radiogenic elements are likely to become concentrated in the topmost parts of intrusions, so modelled contours pointing to large circumscribed bulges at accessible depths on the surface of the concealed EGB would be obvious targets for further consideration. Before selecting target areas, a new 3D model of the region should be produced using up-to-date data, knowledge and interpretations.

(ii) Large intrusions of basic igneous rock crop out on the north side of the EGB; the Morvern-Cabrach, Insch and Huntly-Knock intrusions are the largest ( Figure 32). These intrusions are believed to be laccoliths; that is, they are broadly flat-lying and saucer-shaped rather than upright and balloon-shaped (which is more typical of large intrusions). As such, the EGB will extend beneath them, and the interface between the intrusions and underlying rocks is likely to be relatively flat-lying (good for trapping heat). Like the EGB, the intrusions are enclosed in metasedimentary rocks. The basic igneous rock forming these intrusions has significantly lower thermal conductivity (~2.2 Wm ^-1 K ^-1) than both granite and the enclosing metasedimentary rocks, so they have the potential to significantly impede the upward flow of heat supplied from beneath them. The northern margins of at least two exposed HHP granite intrusions - Ballater and Bennachie ( Figure 26) - are in direct contact with the Morvern-Cabrach and Insch masses, respectively. The contacts are modelled (Rollin, 1984) to dip northwards at 45-60ºC, beneath a substantial thickness of basic rocks. The bases of the basic intrusions are modelled (Rollin, 1984) to be at between 1 and 5 km depth. Thus, HHP granite within the EGB may in places be concealed beneath a thick and extensive 'cover' of basic igneous rocks in this region. As noted above, a new 3D model of the region should be produced using up-to-date data, knowledge and interpretations, before selecting areas for further investigation.

(iii) The subsurface extent of the EGB north-west of the Monadhliath intrusion was not well-constrained by the gravity modelling of Rollin (1984), but his modelling did suggest that it may continue sufficiently far north to underlie the Devonian sedimentary rocks that crop out around the margin of the Moray Firth ( Figure 31). The area around Nairn and Forres, and possibly further to the east and west, may therefore be underlain by Devonian sandstones that overlie metasedimentary rocks, which in turn overlie granite in the EGB. A small granite intrusion in this area, the Auldearn intrusion, has compositional and textural characteristics similar to the East Grampians intrusions, and may be the surface manifestation of the buried EGB. Unfortunately, there is no reported HP value for the Auldearn intrusion. Intrusions of 'hot' granite buried beneath a thick cover of metasedimentary rocks and sandstone may exist in the district around Nairn and Forres.

The Orcadian Basin

In the far north-east part of Scotland a thick sequence of sedimentary rocks (sandstone, conglomerate and mudstone) was deposited in a large, lake-filled depression - the Orcadian Basin - during the Devonian Period, in the aftermath of the Caledonian Orogeny. The sedimentary rocks are now exposed onshore across the whole of Orkney, parts of Shetland, and within a variably broad band bordering the Moray Firth. They also occur offshore (mainly beneath younger sedimentary rocks) under much of the Moray Firth and farther afield. The thickness of sedimentary rocks generally increases towards the centre of the former lake, so the thickest onshore sections - possibly exceeding several kilometres (Rippon, 2002) - are probably mainly in the vicinity of the coast. The same crystalline basement rocks that underlie the Northern Highlands and Grampian Highlands underlie the Orcadian Basin strata, both onshore and offshore. Most of the exposed HHP granite intrusions in Scotland crop out on the mainland to the south and west of the Moray Firth ( Figure 26 and Figure 30). The distribution of these exposed intrusions suggests that others are concealed beneath the strata of the Orcadian Basin.

Possible targets for 'buried hot granite' in onshore settings beneath Orcadian Basin strata include the following (in no particular order).

(i) The Orkney Islands. The islands are underlain mainly by Devonian sedimentary strata, though basement rocks crop out locally on Mainland. There is currently no geological evidence or geothermal evidence for buried hot granite. The sedimentary strata on Orkney include substantial thicknesses of mudstone, which has low thermal conductivity ( Table 1) and should be a good thermal insulator.

(ii) Caithness, east of a line from Dounreay to Berriedale. The sedimentary rocks in this area are inferred to be underlain by a concealed granite intrusion, the 'Caithness Granite' (Hillier and Marshall, 1992). The available evidence suggests the intrusion was emplaced into the sedimentary strata (rather than having been buried beneath them), so it may be younger than the East Grampians and other late Caledonian intrusions. The sedimentary strata in Caithness are mainly mudstone, which has low thermal conductivity and should be a good thermal insulator.

Two large granite intrusions crop out in the metasedimentary basement rocks adjacent to this area: the Strath Halladale intrusion (Figure 26) has a very low HP value (1.4 μW m ^-3), but the nearby Helmsdale intrusion ( Figure 26 and Figure 30) has an HP value of 4.1 μW m ^-3 (above the HHP threshold used in this report). Uranium mineralisation occurs at the contact between the Helmsdale granite and the Devonian sedimentary rocks that overlie it locally, and it is not clear to what extent the measured HP value (from samples collected at the ground surface) has been influenced by leaching and local redistribution of uranium. The pink, texturally variable granite at Helmsdale is similar in many respects to that in the East Grampians intrusions, but unlike them the Helmsdale intrusion is apparently concentrically zoned and lacks significant hydrothermal alteration. These features suggest the outcrop level through the Helmsdale intrusion is below the roof zone, hence the measured HP value may be broadly representative of a substantial volume of rock at depth. The Helmsdale intrusion is interpreted to be a steep-sided, stock-like mass, and is associated with a significant positive magnetic anomaly (~150 nT), similar in magnitude to the anomalies associated with the HHP granites of the East Grampians Batholith. North of Helmsdale, a slightly larger (~180 nT) magnetic anomaly centred on the coast just south of the town of Wick, well within the outcrop of the Devonian strata, may point to a concealed, Helmsdale-like intrusion ( e.g. Flinn, 1969). Heat flow has been measured in several boreholes near to this magnetic anomaly ( Figure 4); the values lie close to the UK mean (54±12 mW m ^-2; Wheildon and Rollin, 1986).

(iii) The area bounded roughly by Dornoch in the north, Dingwall and Inverness in the south, and Elgin in the east. Two granite intrusions with HHP character - Fearn ( HP = 5.1 μW m ^-3) and Abriachan ( HP = 4.0 μW m ^-3) ( Figure 30) - crop out in metasedimentary basement rocks close to this area ( Figure 26) suggesting there is a reasonable prospect of at least one buried intrusion of HHP granite in the area. A substantial positive (~350 nT) magnetic anomaly extending from Ben Rinnes to Elgin and northwards beneath the Moray Firth is the most striking in the Eastern Highlands area (Rollin, 1984). This magnetic feature is not associated with a strong positive gravity anomaly and is therefore probably not a body of basic igneous rock. It could reflect a large, buried body of granitic rock (Rollin, 1984).

Midland Valley

The Midland Valley represents the largest onshore area in Scotland where a thick pile of sedimentary rocks overlies crystalline basement. The sedimentary strata in places contain numerous seams of coal, which has very low thermal conductivity ( Table 1) and so is an excellent thermal insulator. The coal seams are, however, very thin compared to the interbedded mudstones and sandstones that will mainly determine the thermal conductivity profile, and the degree to which they would act as a thermal insulator is not clear. The nature of the basement rocks beneath the Midland Valley is very poorly understood, and the extent to which intrusions of granite (let alone HHP granite) might exist in them is not known, and may be impossible to determine. Only one intrusion of granitic rock of significant size actually crops out within the Midland Valley; the Distinkhorn intrusion ( Figure 26) consists of diorite and granodiorite, and has low HP capacity (2.0 μW m ^-3). The potential for locating and exploiting a 'hot buried granite' in the Midland Valley therefore appears to be very small; however, if granite intrusions do exist in the basement rocks, the heat they supply to the overlying sedimentary rocks may contribute to the geothermal energy potential in Hot Sedimentary Aquifer settings and abandoned mine workings.

Southern Uplands

A north-east-south-west trending gravity anomaly near Peebles in the north-east part of the Southern Uplands region has been interpreted to reflect a large, concealed granitic intrusion (the 'Tweedale Granite') whose top lies beneath 2-3 km of weakly metamorphosed sedimentary rocks (Lagios and Hipkin, 1979; Rippon, 2002). By analogy with the large intrusions that crop out further to the west (Loch Doon, Fleet and Criffel, Figure 26), at least part of the inferred Tweedale Granite could have HHP character. The metasedimentary rocks of the region have thermal conductivity values similar to those of granite, and will not significantly impede the upward flow of heat generated beneath them.

In the south-west part of the Southern Uplands, moderately thick piles (probably <1,500 metres) of Permian age sandstones and subordinate basic igneous rocks are preserved in several areas (see section 7.3.1.1). These sequences might act to impede the upward flow of heat from beneath them, but their relatively small size suggests they are unlikely to overlie intrusions of granite.

8.3 Low thermal conductivity rocks in areas of elevated heat flow

For any given heat flow value, units of low thermal conductivity rocks should be hotter than units of high thermal conductivity rocks, yielding a broadly consistent geothermal gradient. Thick units of low thermal conductivity rocks therefore have the potential to contain large reservoirs of heat. This effect will be enhanced if the supply of heat from depth exceeds normal 'background' values. Two areas of Scotland where this setting may exist are described below ( Figure 32).

Basaltic and andesitic lavas form thick units of low thermal conductivity rocks (probably around 2.2 Wm ^-1 K ^-1) on Skye, Mull, in neighbouring Morvern, in Lorne (Argyll), the Midland Valley, and around Cheviot in the Southern Uplands. Elevated temperatures compared to other rock types may exist in these units due to their low thermal conductivity. One of these areas may be associated with slightly elevated background heat flow. The western seaboard of Scotland (together with Northern Ireland) was the last part of the UK to suffer a geological event that involved the transfer of a substantial amount of heat from the mantle into the crust; this occurred during the opening of the North Atlantic Ocean, around 60 million years ago. The resulting voluminous magmatism created many intrusions, volcanic centres and thick piles of lava, which now crop out extensively on islands of the Inner Hebrides and in numerous offshore locations. Residual heat from this event, stored deep in the crust, may provide a small boost to background heat flow in the area. A preliminary assessment by BGS of thermal data from the Atlantic Margin ( BGS unpublished work) has identified no evidence for residual heat flow associated with Tertiary volcanism; however, a more detailed assessment would be required to confirm this preliminary outcome.

The low thermal conductivity (~2.2 Wm ^-1 K ^-1) of basic (silica-poor) igneous rocks forming large intrusions in Aberdeenshire (including the Morvern-Cabrach, Insch and Huntly-Knock intrusions that were described in section 8.2.4), means they may be hotter than rocks of higher thermal conductivity in the same area. As described in section 8.2.4, these intrusions are interpreted to overlie the East Grampians Batholith, with their bases at between 1 and 5 km depth. Parts of the EGB consist of HHP granite, and might provide a boost to background heat flow in the area.

Figure 32 Onshore parts of Scotland where large, thick units of low thermal conductivity rocks crop out at the surface. These rock units may contain HDR prospects. See Figure 3 for abbreviations.

8.4 Geothermal potential in low permeability sedimentary rocks

The 'regional geothermal gradient' derived from data plotted on Figure 9 suggests that a temperature of 100°C should be encountered at approximately 3,000 metres, and 150 °C at approximately 4,000 metres, in parts of Scotland where sedimentary rocks extend to such depths. The same temperatures would be encountered at around 3,300 metres and 4,900 metres respectively if the lower geothermal gradient of 30.5°C/km obtained from onshore boreholes is applied. To date, no onshore boreholes in Scotland have penetrated sufficiently far into the crust to test whether the permeability required for HSA potential is preserved at such depths. However, if the estimated temperatures do exist the geothermal resource may be sufficiently large at accessible depths to have Hot Dry Rock potential if the rocks are insufficiently permeable to have HSA potential. Based on current knowledge, sedimentary rocks extend to depths exceeding 3,000 metres in most of the Midland Valley and possibly around the edges of the Moray Firth.

8.5 Conclusions

1. The widely quoted practical lower depth limit for exploiting HDR resources is 5,000 metres. The 'regional geothermal gradient' for Scotland suggests that temperatures of around 150°C should occur widely around, or slightly above, this depth, hence rocks with HDR/ HWR potential may underlie many parts of the country. However, the HDR/ HWR concept is most suited to crystalline rocks, and the lack of temperature data for crystalline rocks in deep (> 2 km) onshore boreholes means that caution should be exercised in applying the regional temperature gradient described in this report to potential HDR/ HWR resources in crystalline rocks.
2. Some parts of the crust contain thermal anomalies, wherein the size of the local heat resource exceeds that which is produced by simple 'background' heat flow. Finding thermal anomalies at accessible depths should increase the chance of developing viable Hot Dry Rock/Enhanced Geothermal System schemes, because the average geothermal gradient in the crust overlying them is higher than the 'background' gradient. In Scotland, thermal anomalies in crystalline rocks might be produced in three settings: (i) where the background heat flow is augmented by additional heat generated in situ; (ii) where the upward flow of heat is impeded, such that some of it becomes trapped; and (iii) a combination of (i) and (ii).
3. Intrusions of granite containing elevated concentrations of radiothermal elements (U, Th, K) are likely to be the only source of significant additional heat generated in situ within the crust beneath Scotland (setting (i) in conclusion 2). Seven large High Heat Production ( HHP) granite intrusions are exposed in Scotland; a cluster of intrusions lies in the East Grampians region and two others crop out in ground to the north of Inverness. A previous investigation of the East Grampians intrusions concluded that the heat flow values associated with them were too low to indicate HDR potential. However, the recent recognition that heat flow values in Scotland probably significantly underestimate the size of the heat resource at depth (conclusion 3) indicates that the geothermal potential of exposed HHP granite intrusions should be re-assessed (ideally with new heat flow data).
4. The presence of exposed HHP granite intrusions in parts of Scotland raises the possibility that substantial heat reservoirs exist where HHP granite intrusions are buried beneath a thick cover (blanket) of insulating low thermal conductivity rocks. Such settings may exist in Orkney, Caithness, and the coastal zone to the north and east of Inverness, where Devonian sedimentary rocks overlie Caledonian basement (in which reside all of the exposed HHP intrusions). Much of the East Grampians region is underlain at depth by the inferred East Grampians Batholith. Some parts of the buried batholith may have HHP character, and the batholith is in places overlain by large, variably thick intrusions of silica-poor igneous rock with low thermal conductivity. Unfortunately, the buried setting and the 'thermal blanket' effect of the low conductivity rocks means that good 'buried hot granite' prospects are likely to be difficult to identify using only near-surface and remotely-sensed data.
5. Some thick rock units with low thermal conductivity might impede the upward flow of heat sufficiently to create reservoirs of trapped heat (setting (ii) in conclusion 2). This effect will be enhanced if the supply of heat from depth exceeds normal 'background' values. Such a setting may exist in two parts of Scotland. (i) Residual heat from the Tertiary opening of the North Atlantic Ocean may provide a small boost to background heat flow along the western seaboard, parts of which (Skye, Mull, and neighbouring Morvern ( Figure 32) are underlain by thick piles of low thermal conductivity lavas. A preliminary assessment by BGS of thermal data from the Atlantic Margin ( BGS unpublished work) has identified no evidence for residual heat flow associated with Tertiary volcanism; however, a more detailed assessment would be required to confirm this preliminary conclusion. (ii) Some of the large, low thermal conductivity, silica-poor intrusions in Aberdeenshire are interpreted to overlie the East Grampians Batholith, with their bases at between 1 and 5 km depth. Parts of the EGB consist of HHP granite, and might provide a boost to background heat flow in the area.

8.6 Recommendations

The evidence base for assessing the potential for Hot Dry Rock prospects in Scotland is far from adequate. Most significantly, our knowledge of the distribution of granite intrusions with HHP character is limited to those that are currently at outcrop in onshore areas and for which appropriate geochemical data exist. We need therefore to improve our understanding of the distribution in Scotland of exposed and buried intrusions containing HHP granite. The following suggestions for further work focus mainly on identifying and characterising intrusions of HHP granite.

R9 Compile heat production data for all the granite intrusions of Scotland to identify all intrusions that have HHP character at outcrop

A programme of systematic surface sampling and geochemical analysis to augment the existing dataset would provide a complete dataset for granites across Scotland.

R10 Conduct research to identify whether some of the exposed intrusions that do not have HHP character would have had HHP character in now-eroded portions, or may have HHP character in still-buried portions

This would help to constrain the true areal distribution of granite intrusions with HHP character, and to establish whether buried HHP granite intrusions may exist in parts of the country beyond those in which they currently crop out. This could be addressed by: (i) developing a fuller understanding of how and why HHP granite forms; (ii) establishing the typical position and proportion of HHP rocks in intrusions; and (iii) identifying a geochemical (or some other type of) 'fingerprint' that can be used in intrusions lacking HHP character at outcrop to point to the presence of HHP rocks in eroded or concealed parts. A detailed study of intrusions in Scotland and elsewhere could address these issues, drawing on the vast body of published and unpublished granite literature, and gathering new data where necessary.

R11 In offshore areas, use seismic survey data and information from wells to identify buried intrusions and intrusions exposed on the sea floor

This would help to constrain the true areal distribution of granite intrusions.

R12 In onshore areas, re-interpret existing regional geophysical data and 3D geological models using modern methodologies and up-to-date knowledge of the surface and subsurface geology to identify possible buried granite intrusions (and other potential HDR settings)

Specific objectives could include: (i) improving the existing model of the 3D shape of concealed parts of the East Grampians Batholith and low thermal conductivity units that overlie it, and their spatial relationships; (ii) improving the characterisation of known geophysical anomalies that may point to granite intrusions concealed beneath Old Red Sandstone strata; (iii) identifying and characterising other locations where granite intrusions may be concealed beneath several kilometres of sedimentary rock. New, higher-resolution geophysical data collected from a subset of carefully selected prospects would provide a clearer picture of the size, depth, shape and lithology of buried units, and their spatial relationship with overlying rocks.

R13 Characterise the fracture network in exposed HHP intrusions

The 'buried granite' model relies on heat building up over many millions of years beneath a low thermal conductivity layer, but if the granite contains a system of open, connected fractures (which typically form in unroofed granite intrusions) it could suffer significant heat loss via lateral fluid flow. Furthermore, the orientation and disposition of the existing fracture network within an intrusion plays a major role in controlling the shape and nature of an engineered reservoir. A study of fracture patterns in exposed HHP granites could provide an indication of the fracture architecture that will be encountered in geothermal reservoirs developed in buried intrusions.

R14 Conduct a programme of deep drilling

Ultimately, one or more deep boreholes will have to be sunk onshore if the potential for exploiting deep geothermal energy in an onshore setting is to be evaluated fully. There are no zones of unusually high heat flow and no existing deep onshore boreholes in Scotland, so finding an accessible deep HDR geothermal resource will require a dedicated exploration programme. Initially, our ability to identify and quantify geothermal energy prospects will depend on gathering thermal data at the surface and in shallow boreholes, and on building models of the geology in three-dimensions from surface-based and remote-sensing surveys. However, at some point a deep drilling programme will be needed to provide measured and observed, factual data. To provide a clear indication of the deep geothermal regime, and in particular the shape of the geothermal gradient beneath the influence of surface- and near-surface effects, a deep geothermal borehole would probably need to extend to a depth of at least 3 kilometres, and ideally to 5 kilometres.