Guardbridge geothermal technology demonstrator project: feasibility report

Geothermal Energy Challenge Executive Summary

This feasibility study investigates whether a geothermal district heating system, which accesses Hot Sedimentary Aquifer ( HSA) resources underlying a brownfield site at Guardbridge in northeast Fife, can be developed in a cost-effective manner. This project's scope is to assess the available geological information and estimate the hot saline aquifer heat supply, calculate the current heat demand at the Guardbridge site, Guardbridge village, and the nearby towns of Leuchars and Balmullo, and to incorporate future Guardbridge development plans (provided by the University of St Andrews) and anticipated growth in housing stock (from Fife Council) to estimate future heat demand. The capital, maintenance and repair costs for the geothermal well and designed district heating network are used to develop economic models for a number of district heat network scenarios. A key aspect of this study is an evaluation of the opportunities to cost effectively de-risk deep geothermal exploration in Central Scotland, and to outline the potential for developing geological heat storage systems. The study identifies the key legislative and environmental issues, risks and uncertainties associated with any exploration and production, involves stakeholder engagement, and makes recommendations for a Phase 2 stage for geothermal heat development at Guardbridge.

Two of the key outputs from this feasibility study will be an economic model and business case based on different heat demand options, and an optimised model of well design based on different exploration strategies. Both are transferable to similar operations at other geothermal sites. The key objectives are therefore to:

(a) design a geothermal well that will be drilled in Phase 2 of the project, and secure valuable information on Fife regional sub-surface geology and geothermal properties of the primary aquifer,

(b) explore how advanced drilling techniques, such as directional drilling, can be deployed to improve geothermal recovery,

(c) demonstrate how a geothermal system can integrate with an existing biomass heating installation to optimise both schemes and provide a district heat network for on-site industries and the local community,

(d) evaluate the potential for storage of seasonal heat energy in the subsurface (a first in Scotland), and

(e) assess the relative merits of water treatment and on-site recycling, re-injection or disposal to sea.

A regional geological model was constructed using available data from the British Geological Survey, published data and academic theses. The sub-surface geology was interpreted from surface geology and extrapolating the local behaviour of geological structures into the Guardbridge area. Modelling the geology involved defining the orientation and width of a natural fault zone, which could be a significant influence on the behaviour of the Hot Sedimentary Aquifers. The rock units of interest in this study are the Upper Devonian Scone Sandstone, Glenvale Sandstone, Knox Pulpit and Kinnesswood formations, and the latter two units are previously identified as having the highest potential to be highly productive aquifers. The presence of a major fault near the Guardbridge site means that the target aquifers are at very different depths on either side of the fault. The report therefore investigates and evaluates three well options to target the different aquifers at the varying depths on either side of the fault.

Hydrogeological modelling was conducted using FEFLOW ^® to evaluate the behaviour of the fault on fluid flow rates, and to predict the necessary conductivities to produce reasonable, economic and sustainable rates of fluid extraction. Although not an accurate model of the Guardbridge site, and limited by a significant lack of data constraining the important parameters, the flow simulations suggest that fracture permeability in the aquifers and underlying rocks is needed to sustain the flows recommended by this study, and re-injection would be required if a producing well was to be sustainable over many decades.

Regionally developed rock quality predictors have been used to estimate the permeability and temperature of the target aquifer intervals in the three selected well options at, or near, Guardbridge. Oil field well simulation tools have been used to estimate water flow rates, temperature profiles, and circulating rates from different geological models of the wells. Two of the wells, GB-1 and ES-1, are not expected to penetrate enough high permeability sandstone to support the minimum water flow rates of 5 l/s and so are ruled out as viable aquifer producers.

GB-2 is a deviated well that penetrates the Kinnesswood and Knox Pulpit formations, the best quality regional aquifers, in a zone where the fault may enhance the permeability even more, and has potential to supply 5 to 20 l/s of water at a surface temperature of 25 ^oC (± 2 ^oC). Such a well will be produced using an electric submersible pump which will require 20 - 40 kw of power to deliver 15 l/s of flow (although the volumetric rate will vary with the rock quality). GB-2 is taken forward and drilling designs are provided with three outcomes: 1) a dry hole scenario; 2) a 5 l/s scenario; and, 3) a 15 l/s scenario.

The vertical wells have been modelled as heat pump circulating wells, and therefore would not produce any aquifer water at the surface. Only deeper wells, up to 2500 m, have the potential to give surface temperature increase of 5 ^oC at reasonable circulation rates (e.g. 8 l/s). A deep GB-1 well as a heat pump could be taken forward in Phase 2 as an alternative heat source.

The proposed GB-2 deviated well can be drilled across the fault from the Guardbridge site to a depth of 1200 m. A casing string set will isolate the shallow geology and a slotted liner used to prevent hole collapse of the target intervals. Such a well will require a 100 tonne conventional drilling rig and well control, logging and coring tools will assess the aquifer quality. In the most likely case, the drilling phase will take 24 days, including rig mobilisation and demobilisation. If coring and logging demonstrate that the well will not flow adequately, then the well will be suspended. Low cost options have been investigated that would allow exploratory wells to be drilled and this could result in the recovery of regionally significant data on the performance of the aquifers at depth, although none of the boreholes could be completed to production stage due the drilling technology employed.

The drilling scenarios investigated do not include a re-injection well, in order to create an economically viable district heating network project, even though very preliminary hydrogeological modelling demonstrates that re-injection is required if the geothermal well is to be sustainable over 30+ years. Alternative management of produced water investigated in this report are: water disposal-to-sea and partial-full water recycling and re-use on site. The first option could have environmental consequences on the adjacent Eden Estuary, which is part of the Tay River and Eden Estuary Special Protection Area, and these potential impacts would need formal assessment by a competent authority (Fife Council and SNH) as part of a Habitat Regulations Appraisal, and an Environmental Impact Assessment is most likely required. The second option reduces the environmental impacts on the estuary, but has additional CAPEX and OPEX costs which are estimated. The opportunity to be innovative about partial water recycling and re-sale should be investigated in Phase 2.

The heat demand is based on preliminary district heating network layouts at different scales, based on the demand analysis. Demand has been assessed at Guardbridge and the nearby towns of Leuchars and Balmullo, using the Scotland Heat Map and future development data provided by the University of St Andrews and the Fife Development Plan. These various options provide an indication of the potential annual and peak heating demands that can then be compared against the geothermal heating potential, and an economic modelling tool was developed to analyse the performance of the overall system, including key performance indicators to evaluate the financial viability. This analysis leads to a preliminary network design and an economic model of the potential scheme.

The District Heating Opportunity Assessment Tool ( DHOAT) designed for the Danish Energy Agency analyses the Heat Map data and preliminary network designs and provides peak and annual demands and key performance indicators, namely total heat demand and indicative CAPEX, OPEX, REPEX and heat sales. All input parameters are modelled with an uncertainty of ±10%. Based on this analysis, the proposed development of one well and estimated heat supply is not sufficient capacity to provide heat outside of the Guardbridge site itself. All district heating network designs and economic models were therefore based on the aggregated customer base of the Guardbridge site. The economic model assumes that geothermal heat can supply 50% of the Guardbridge site needs (2,867 MWh/a), with a capacity of 0.42 MW, and the other 50% would be provided by the biomass plant. Revenues from heat sales are based on a heat sale price scaling (MWh and p/kWh) and costs of heat from the biomass plant.

An Excel model calculates the profitability of the scheme based on a CAPEX of £530,000 for the heating network and £1,517,000 for the well completion, flow tests and water treatment. OPEX and REPEX costs are principally power consumption for the heat and distribution pumps (£280,000), and a ESP and heat pump replacement after 10 years (£250,000). NPV and IRR are used to demonstrate viability for potential investors over a 21-year period; the best case scenario shows that the scheme might achieve a 10% IRR and a positive NPV. However, the heat sale price is too low to create sufficient margin to make the economic performance attractive. This is principally due to the cost of the geothermal heat. The capital cost of the geothermal well is a significant portion of the project CAPEX and does not vary with the well heat potential, which is a relatively modest value given the temperature and flow rate estimates presented. Flow rate is highly uncertain, while temperature is better constrained and low due to the shallow depth of the proposed well. The district heating network requires higher temperatures and the addition of a heat pump increases the capital costs and adds a relatively high operating cost for the electricity to run the pump.

The carbon emissions reductions are compared to an individual gas boiler alternative (business as usual [ BAU]) and the geothermal-biomass heat network shows an 84% reduction in carbon emissions, assuming that the biomass boilers and geothermal heat pumps each supply 50% of the network demand. About 58% of the emissions reduction (13,878 tonnes CO ₂/kWh relative to BAU) is attributed to heat generation from the biomass plant and the remaining 42% (9,812 tonnes CO ₂/kWh relative to BAU) is attributed to the geothermal well and the heat pump. These figures are based on a model lifetime of 20 years. The value of this carbon saving has not been included in the economic model, however it could be considered to represent an additional savings compared to the business-as-usual alternative.

The heating network can be enhanced at a subsequent stage to provide combined heating and cooling for the site. This would increase the utilisation of the heat pump by operating in combined heating and cooling mode during inter-seasonal periods. Although not explored in any extensive technical or economic sense, the system could also potentially be used to fill separate hot and cold seasonal heat stores.

Requirements for Phase 2 would begin with a non-invasive geophysical survey to provide imaging of the fault and the target aquifers in the subsurface. This could be completed in three months. Phase 2 would most likely require the preparation of an Environmental Statement before any drilling could commence on site, particularly addressing the viability of disposal of water to the sea. However, current developments at Guardbridge have required Environmental Statements (i.e. since 2014) and much baseline data already exists. The time required to complete an EIA range from 12 weeks to prepare the report, or up to one year of time if SNH and Fife Council require additional new data. A benefit of the Guardbridge site is therefore its status as an industrial site with a pre-existing history in terms of Environmental Statements. Ideally, Phase 2 would culminate in revised well designs, procurement of the drilling rig, and test drilling to intercept the fault and target aquifers. The time and costs are estimated and depend on the choice of drilling option. A positive outcome from a test borehole would lead to the design of a full production well and progression of the project as a Technology Demonstrator. Regardless of whether the test borehole proves that the Guardbridge District Heating Network project is viable, the data recovered as part of the test drilling (core samples, flow tests and water chemistry) will be highly significant for de-risking hot sedimentary aquifer exploration across central Scotland.

The economic feasibility of the Guardbridge geothermal heat project is dependent on the best case scenario for flow rates, along with a large number of other poorly constrained variables. It could be economic, but there is a very large uncertainty in the geothermal heat estimates. However, the additional value in the potential research that can be achieved at Guardbridge in de-risking hot sedimentary aquifer exploration in the Central Belt of Scotland, as well as integrating low carbon heat source exploration with other technologies, including dual heating and cooling and water recycling, should be considered when deciding to progress this project.