Publication - Research and analysis

Hill of Banchory geothermal energy project: feasibility report

Published: 23 Mar 2016
Energy and Climate Change Directorate
Part of:
Environment and climate change

Report of a study which explored the potential for a deep geothermal heat project at the Hill of Banchory, Aberdeenshire.

Hill of Banchory geothermal energy project: feasibility report
15. Carbon Emissions Savings

15. Carbon Emissions Savings


Deep geothermal energy is widely recognised as a very low carbon energy type, particularly when used to produce heat. The calculations set out below suggest that for a system with the same likely parameters as the proposed Banchory scheme each megawatt hour of thermal energy would produce around 4.4 KgCO 2/MWh th. This compares to the commonly accepted figure of 184.45 KgCO 2/ MWh th for natural gas, the most common type of heating in Scotland for both domestic and non-domestic consumers, and 246.57 KgCO 2/MWh th for heating oil, the most common type of heating found off the gas grid.

Where geothermal heating replaces gas or oil heating, therefore, significant amounts of carbon emissions will be abated.

At Banchory, our calculations suggest that for a geothermal doublet system with a capacity of 2.5 MW th meeting 13,140 MWh of heat demand annually, where that demand had previously been met by natural gas, the annual amount of CO 2 abated would be approximately 2,370 tonnes. Over the thirty year lifetime of the geothermal system this equates to an abatement of 71,000 tonnes of CO 2.

In the 'best possible case' under which a positive outcome to the Banchory project results in 250 similar geothermal heat projects across areas of Scotland with suitable granite plutons, we estimate the total carbon saving achievable as 900,000 tonnes of CO 2. Such an outcome would be equivalent to meeting around 5% of Scotland's heat demand.

The carbon footprint of deep geothermal energy

A commonly used unit of carbon intensity for heat is 'kilogrammes of carbon dioxide per megawatt-hour of heat produced' (KgCO 2/MWh th). Various sources provide 'reference' figures for the carbon intensity of a wide range of heat-producing technologies. For example, in the UK Defra publishes the 'Greenhouse Gas Conversion Factor Repository' [10] , an official database that is often used when modelling carbon footprints or evaluating proposals for carbon emission reductions. We have used the Defra figures for this analysis.

Deep geothermal heat generation requires only minimal input of fossil fuels, and as a result it produces a very low level of greenhouse gases per megawatt hour of heat delivered. However, there are relatively few sources in the academic literature offering a detailed analysis of this question, or providing a reference figure for general use. The Defra tables do not quote figures for geothermal heat or power, for example. Internationally, figures are available for the carbon intensity of deep geothermal power production (for example, the standard figure given by the US Government [11] is 26.6 KgCO 2/MWh e). This includes a large component of volcanogenic sources which are naturally rich in CO 2, which is released to the atmosphere in flash steam operations. As geothermal heat production does not require venting, and as many of the mid-enthalpy sources used for heat do not contain natural CO 2 anyway, the equivalent carbon emissions figures for deep geothermal heat-only systems can be confidently be expected to be far lower than the figure for power production quoted above. Although accurate figures for deep geothermal heat production are as yet sparse, a review of various sources presented by Younger (2015) quotes a figure of 4 KgCO 2/MWh th.

Instead of relying on such generalised estimates, we here calculate the likely carbon intensity of deep geothermal heat at Banchory from first principles, using reasonable assumptions and approximations. We do this on the basis of a 'full lifetime' calculation i.e. we consider the potential greenhouse gas emissions that arise in the course of a geothermal heat project, from the start of the borehole construction process through to the 'steady state' production of emissions while the system produces heat from the system over its lifetime. We do not include a calculation for the carbon emissions arising from decommissioning, which we judge to be de minimis in comparison with the construction of the boreholes and the operation of the system over decades.

Project Lifetime

We make a conservative assumption that the lifetime of a deep geothermal system at Banchory will be thirty years. Systems can last much longer than this (one geothermal heat system in Boise, Idaho, has been in operation since the 1890s) but the lifetime will be dependent on local factors. This figure assumes there will be some refurbishment of the system, such as occasional pump replacement (over a thirty year lifetime the contribution of such refurbishments to overall carbon emissions should be negligible).

The potential for venting of gases from the geothermal fluids

For completeness we should note in passing that greenhouse gases could be produced from the geothermal system in ways that are unrelated to energy consumption or heat production. Where systems in other countries are based on volcanic geothermal sources some greenhouse gases such as CO 2 may be vented. However this is highly unlikely in Aberdeenshire and we strongly expect that the geothermal water itself will not give rise to any greenhouse gases in this way. Indeed, the production and re-injection boreholes will be operated as a sealed system.

Carbon emitted prior to system operation

Here we consider the carbon emissions that arise before the system is commissioned. This will include the 'embedded carbon' involved in the manufacture of the different elements of the system, and the carbon dioxide emitted during the drilling of the boreholes. There will be other sources of carbon - for example, that generated by activities associated with the planning phase of the project.

The 'embedded' carbon emissions associated with the system components can be divided into those from the manufacture of the surface elements of the system e.g. the pump, heat exchanger etc. and those arising from the manufacture of the 'casings' - the stainless steel cylindrical lining that is generally used to stabilise the walls of boreholes as they are drilled. We judge the contributions from the surface elements are not large in scale (c.f. the carbon emissions involved in manufacturing a mid-size car are around 15 tonnes of CO 2).

As noted in Section 6, the bulk of the steel casing will be 9 5/ 8" in diameter (dimensions in the drilling industry are still given in imperial units). Casing of this diameter should weigh no more than 100 kg/metre. We note from the World Steel Association's 'sustainability indicators' page [12] that producing a tonne of steel in 2014 can be assumed to release 1.9 tonnes of CO 2. Therefore, 100 metres of casing can be taken to equate to 19 tonnes of CO 2.

If both boreholes were fully cased - which would require 5,000 metres of casing - the associated embedded carbon associated with the casings could be significant, at around 950 tonnes. However, the Banchory project concept is based on drilling through granite, a very stable rock. We therefore expect that for the great majority of the boreholes' depth an 'open hole' approach will be used i.e. no casing will be needed as the walls of the borehole will be highly stable and reinforcement unnecessary. If, for example, each borehole was cased only to 500 metres depth, the associated CO 2 emissions would be around 190 tonnes.

We will take this as the central assumption, rounding up to 200 tonnes of CO 2.

GHG emissions arising from drilling the boreholes

Moving on from the embedded carbon in the system components, another significant source of carbon emissions will be the fuel consumed by the drilling rig (these can vary considerably in size, but almost all run on diesel). For the Banchory project a relatively large drilling rig would be required, and based on consultations with drilling industry experts we expect such a rig to consume around 1,500 litres of diesel oil per day, or 62.5 litres per hour.

There is inevitably some uncertainty in any estimation of the number of hours of operation required for the rig to drill to the required depth. Each drilling operation is different, and time taken will depend on the hardness of the rock strata and any difficulties encountered (lost drill bits, drilling into voids, the need to drill 'daughter' side boreholes etc.). However, we suggest that for a 2,000 to 2,500 metre borehole in granite a figure of 1,500 hours' drilling is a reasonable estimate.

This implies a total consumption of 1500 x 62.5 = 93,750 litres of diesel fuel per borehole. Given that we also need to drill a re-injection borehole we need to double this figure, to reach a final figure of 187,500 litres of diesel fuel consumed.

Burning one litre of diesel fuel produces 2.4435 kg of CO 2 [13] . Therefore the carbon emissions arising from the drilling of the two boreholes equate to 458,156 Kg CO 2. Given the uncertainties involved we can reasonably approximate this to 460,000 Kg, or 460 tonnes of CO 2.

Carbon emissions arising from running the system

During the actual operation of the geothermal system the dominant energy input is the electrical power to the pump (we assume this will operate using grid electricity).

Note that the geothermal fluid (i.e. the hot water) need not be pumped from the full depth of the borehole; artesian pressure will raise the level of water in the borehole to the level of the local water table, a point much nearer to the surface (cf Younger and Manning 2010). This greatly reduces the work that must be done by the pump.

However, without knowing the exact level at which water will settle in the borehole - and therefore the height (or 'head') through which the pump will need to raise it - it is not possible to produce an exact figure for the amount of work done by the pump. By extension it is not possible to give an exact figure for the electricity consumed and, in turn, the CO 2 produced per megawatt-hour of heat that is ultimately generated. However, we can calculate the power required to raise water from an estimated minimum and maximum heads (which we take to be H min = 50 metres and H max = 100 metres) to give indicative figures for the power consumption.

The power used by the pump is given by the equation:

P = ρ g Q h / η

where P = the power in watts; ρ = fluid density (kg/m 3); g = gravitational acceleration (in metres/s 2); Q = pumping rate (in cubic metres per second); h = pumping head (in metres) and η = the efficiency of pump.

We use figures of 0.02 m 3/s - 20 litres per second - as a reasonable proxy for the pumping rate, and 0.6 for the pump efficiency.

We calculate the pump power using the H min and H max figures for the pumping heads.

For H min:

P = ρ g Q h / η

= (1000 × 9.81 ×0.02 × 50) / 0.6

P = 16,350 watts = 16.35 kilowatts

For H max:

P = (1000 × 9.81 ×0.02 × 100) / 0.6

= 32,700 watts = 32.7 kilowatts

Using this set of assumptions suggests that the pump power will range from 16.35 to 32.7 kilowatts. We now estimate the annual CO 2 footprint for pumps with these power ratings.

This part of the calculation is wholly dependent on the carbon intensity of the grid supply. At present this is 462.19 Kg of CO 2 per MWh supplied [14] . If the geothermal system operates 60% of the time (a reasonable assumption given the performance of similar systems in continental Europe) this equates to annual carbon emissions between 39.7 tonnes of CO 2 (min) and 79.4 tonnes of CO 2 (max).

Over the thirty year lifetime of the system, this suggests a range of carbon emissions of approximately 1,290 to 2,380 tonnes in total. However this would be a very conservative approach given that the carbon intensity of the UK grid is widely expected to fall over this period as coal fired power stations are phased out and lower carbon generation from renewables, nuclear and gas is phased in. In fact, the UK Government's working assumption is that their 'Electricity Market Reform' policy changes will achieve a reduction of grid carbon intensity to only 100 grams of CO 2/kWh of electricity generated by 2030 - less than a fifth of the current figure.

Decarbonising the grid this much is ambitious and may, of course, not be achieved for a variety of factors inside and outside the Government control. However it would be surprising if the carbon intensity of the grid did not reduce substantially over the 30 years from 2018 (which we take as the earliest commissioning date for a Banchory geothermal system).

As a compromise we will assume that the 100 grams of CO 2/kWh target is reached, but at the end of the thirty year period of the system lifetime, and that grid intensity falls steadily over that time. In other words, we take the average carbon intensity of power taken from the grid over the plant's operation as (462 - 100)/2 = 281 grams CO 2/kWh.

The minimum and maximum amounts of carbon generated by the pump's operation over the system lifetime would thus become:

(minimum case) 16.35 x 8,760 x 0.6 x 0.281 x 30 = 724.4 tonnes of CO 2.

(maximum case) 32.7 x 8,760 x 0.6 x 0.281 x 30 = 1,448.9 tonnes of CO 2

We can reasonably round these figures to 725 and 1450 tonnes of CO 2 respectively.

Bringing the numbers together

In summary, we have calculated the carbon emissions from the pre-commissioning phase of the project as:

Embedded carbon in the casings 200 tonnes CO 2

Embedded carbon in the system surface elements not significant

Emissions from diesel fuel required during drilling 460 tonnes CO 2.

Total pre-operation: 660 tonnes CO 2

During operation, the carbon emissions from the operation of the pump are projected to be:

Minimum assumed pump rating 725 tonnes CO 2 (over 30 yrs)

Maximum assumed pump rating 1,450 tonnes CO 2 (over 30 yrs)

Bringing the pre- and post-commissioning figures together, we find the following:

Minimum carbon emissions over project lifetime: 1,385 tonnes CO 2

Maximum carbon emissions over the project lifetime: 2,110 tonnes CO 2.

To summarise, the main source of carbon emissions over the 30-year lifetime of the geothermal system is that associated with the electricity driving the pump. Having said this, the amount of carbon produced by drilling the borehole, and that embedded in the system's components, is not insignificant. The relative sizes of these contributions will depend on the exact depth from which the geothermal fluids will need to be pumped, and the depth to which the boreholes are cased.

Heat production over the project lifetime

We now need to calculate the amount of heat produced by the geothermal system over its operational lifetime, being careful to ensure this is consistent with our other assumptions.

We assume that the well produces 2.5 MW of heat, a plausible mid-range figure for a geothermal borehole doublet. We also assume that the annual system loading is 60%. Over thirty years the heat generated by the system will therefore be

2.5 x 8,760 x 0.6 x 30 = 394,200 megawatt-hours (= 394,200,000 kilowatt-hours).

This is the equivalent of meeting 13,140 MWh of heat demand per year.

Heat from a geothermal system at Banchory would be very low carbon in nature

Having calculated approximate bounds for the lifetime carbon emissions from the system, and a figure for the amount of heat produced over the same period, we can divide one by the other to find the average carbon intensity of the heat produced.

Lower bound on carbon intensity:

1,385 tonnes of CO 2 / 394,200 megawatt-hours = 3.5 KgCO 2/MWh

Upper bound on carbon intensity:

2,110 tonnes of CO 2 / 394,200 megawatt-hours = 5.4 KgCO 2/MWh

Taking the average of these upper and lower figures, we estimate that a deep geothermal heat system in Banchory would supply heat with a carbon intensity of 4.4 KgCO 2/MWh th.

Carbon savings versus other heat production technologies

The calculation above has given us a plausible range for the carbon intensity of geothermal heat production of 3.5 to 5.4 KgCO 2/MWh. This compares to 184.5 kg of CO 2 per MWh of heat from natural gas central heating, 246.6 kg of CO 2 per MWh for heating from fuel oil, and 462 kg of CO 2 per MWh of electrical heat.

In other words, deep geothermal heat is exceptionally low carbon compared to other fossil fuel heating types, and in particular natural gas heat - the 'reference' heat type in the UK.

Potential carbon emission savings in Banchory

As we noted above, we expect the geothermal system to produce 13,140 MWh of heat per year, with an average carbon intensity of 4.4 KgCO 2/MWh th.

The amount of carbon abated by building the geothermal system will depend on the origin of the heat that it displaces. As discussed in the introductory chapter, the current heat network at Banchory used biomass fuel. This is widely recognised as having a low carbon intensity, with the exact figure dependent on the fuel variety (wood chip, wood pellets, for example) and on the distance the fuel had been transported. There is therefore no simple reference figure for biomass heat from wood chips. The (now rather dated) report 'Carbon factor for wood fuels for the Supplier Obligation - Final' gives a figure for combustion of ' UK forest residues' as 11-17 KgCO 2/MWh [15] . The carbon intensity of heat from the Banchory heat network can be expected to be at or towards the lower end of this range, because the fuel used is forestry waste and thinnings from local sources, minimising transportation distances.

As will be noted, the figure we have arrived at for the carbon intensity of deep geothermal heat is still about half that of wood chip heat at the lower end of this range. This suggests that even against the current very low carbon arrangement a geothermal well would reduce carbon emissions.

Compared with the carbon intensities of common heating types, both deep geothermal and locally sourced biomass are very low carbon heating technologies. The carbon intensity of natural gas heating - which is used to meet the great majority of domestic and non-domestic heat demand - is given as 184.5 KgCO 2/MWh in Defra's reference tables. In other words, a carbon intensity 39 times higher than our calculated figure for deep geothermal. (The figure given in the Defra tables for heating oil, another common heating type in rural Scotland, is 246.6 KgCO 2/MWh, which is 52 times higher.)

In other words, where deep geothermal replaces natural gas heating there is scope for very large reduction in carbon emissions. Given that the vast majority of heat consumers in the Banchory area are likely to use natural gas heating (or electrical heating, which is even more carbon intensive), this is an entirely plausible scenario.

If we assume that the geothermal system produces 13,140 MWh of heat per year and this displaces the same amount of natural gas heating, the amount of carbon emissions saved annually will be:

(184.5 - 4.4) x 13,140 Kg = 2,366,514 Kg ≈ 2,367 tonnes of CO 2 per year.

Over thirty years operation the geothermal system would thus save around 71,000 tonnes of CO 2.

Potential carbon savings across the rest of Scotland

This section considers how much carbon could be saved if it proved plausible to roll-out deep geothermal heat schemes exploiting radiothermal granites across the areas in Scotland where they are found. (We do not consider other types of deep geothermal schemes, such as those based on sedimentary aquifers, which may present opportunities elsewhere, such as the Central Belt.)

Such an exercise will inevitably carry uncertainties, but it is possible to draw some broad conclusions. Note that here we are considering the technical potential for what could be achieved in the long term i.e. we are not considering the commercial potential - what would be economically viable now.

Is making this estimate we must make one subjective judgement call. This concerns how far we believe heat can be transported before the cost of pipework becomes prohibitively expensive, or thermal losses make the system unviable. Heat can, in fact, be transported over large distances without a problematic drop in temperature; a famous example is the Copenhagen heat main which allows heat to travel tens of kilometres from the city's periphery to reach high heat loads in the centre.

So it is not unreasonable to imagine some larger diameter heat mains carrying geothermal water from the edges of the granite plutons into urban areas - most obviously, Aberdeen.

How much heat could be produced?

We assume that a typical deep geothermal doublet such as that proposed for Banchory will have a capacity of a few megawatts - let's take 3 MW as an average figure (note that if many such projects were rolled out significant expertise would presumably be acquired in maximising well outputs). We also assume that these are able to run at a loading of 90% (not unusual for systems of this kind, which are relatively common in other countries).

Wells must be far enough apart to prevent thermal 'short circuiting', where the re-injected colder fluids from one well find their way through horizontal pathways into the production borehole of a neighbouring system, reducing their effectiveness. Such considerations are important in places such as Reykjavik and Paris, where production and re-injection systems are relatively close together. In the latter location systems have been sited 2 km or less apart with no impact on sustainability.

Assuming that systems can be installed round the edges of all the granite plutons in the Cairngorms, and system separation was 2 km, we could expect to install 250 systems without creating an unsustainable impact on the geothermal potential. This equates to a geothermal capacity of 250 x 3 = 750 MW th the annual heat output would be

750 x 8,760 x 0.9 = 5,913 gigawatt-hours

Given the uncertainties we can take this as approximately 6 terawatt-hours (TWh).

How much heat demand would be within range?

Scotland is a voracious consumer of heat. According to 'Energy in Scotland 2014' [16] total non-electrical heat demand in that year was 86.8 TWh, or 55% of Scotland's total energy consumption.

The Scottish heat map, an online resource, allows us to measure the amount of heat demand in any defined area. Bearing in mind our assumption (above) that meeting heat demands within tens of kilometres - but not more - is reasonable, we will assume that Aberdeen and the surrounding areas can be supplied by granite-based heat systems. This leads us to identify the area shown in pink in Figure 40 below as able to be supplied with deep geothermal heat. It covers the local authority areas of Angus, Perth and Kinross, Highland, Moray, Aberdeenshire and Aberdeen City.

Figure 40: maximum area in which heat demand could be met from geothermal heat from radiothermal granites

Figure 40: maximum area in which heat demand could be met from geothermal heat from radiothermal granites
Image: Dr Alistair McCay

We have interrogated the heat map database and established that the annual heat demand in this area is 6.4 TWh - very similar to the total geothermal heat capacity that we estimated above.

On first inspection this is a happy coincidence, though we must exercise caution. The 6.4 TWh figure will include heat demands at temperatures geothermal heat cannot supply (for example, in industrial processes). Furthermore there will be periods of high demand when the geothermal sources (which are characterised by constant, unvarying heat production) are not able to cover the peaks - just as our financial model at Banchory demonstrates.

Assuming good use is made of heat stores (again, as is the case at Banchory) it should be possible to meet the majority of this heat demand. As a conservative estimate we assume a figure of 70%.

In conclusion, we assume that 70% of the 6.4 TWh = approximately 4.5 TWh, or 5.2% of total Scottish heat demand.

We would again note that we have only considered the potential for deep geothermal heat from radiothermal granites. There may be just as much or more potential to exploit deep geothermal heat to meet demand elsewhere in Scotland, including the central lowlands where the majority of the heat demand is located.

Converting heat production into carbon savings

Most of the heat consumers across the selected area will be from natural gas, though a significant proportion - given that much of it is off the gas grid - will use heating-oil. The relative carbon intensities of these heating technologies are:

natural gas heating: 184.5 KgCO 2/MWh

heating oil: 246.6 KgCO 2/MWh.

These figures come from the Defra tables mentioned above.

Robust data are hard to locate, but we note that while purely in terms of square kilometres most of our selected area is off the gas grid, the majority of the population within it (living in the urban areas around Aberdeen) will use gas central heating. We should thus assume a figure that is closer to the lower, natural gas figure - the obvious round figure being 200 KgCO 2/MWh.

(We note in passing that we have not considered the scope for replacing electrical heating with deep geothermal heat - which for the foreseeable future would yield a larger carbon emission saving per unit of heat delivered than gas and heating oil.)

The amount of carbon generated by meeting 4.5 TWh of heat using a technology with 200 KgCO 2/MWh is:

200 x 4.5 x 1,000,000 = 900,000,000 Kg.

In other words, we estimate the potential carbon abatement from using radiothermal granites is around 900,000 tonnes of CO 2 per year.