We are testing a new beta website for gov.scot go to new site

Feasibility Report of Fortissat Community Minewater Geothermal Energy District Heating Network

Listen

Chapter 4: District Heating Network

4.1 Introduction

This chapter undertakes an analysis of the potential heat demand in the study area and a preliminary network analysis to identify target areas with potential for medium and longer term connection to a District Heating Network (DHN) powered by minewater geothermal energy. The preliminary network analysis was then refined into three alternative sizes of network for medium term development - ranging from the smallest network which connects part of Allanton only to a network which connects all of Allanton and Hartwood - taking into account the two preferred geothermal energy options. This process involved a detailed analysis of demand to establish peak demand and demand profile for each scenario, to enable the energy centre, thermal store and network to be appropriately sized, and the operating profile of the heat pump to be modelled. This provided the necessary level of detail to form the basis for the cost schedule and inform decisions about the network operating temperatures. In addition to modelling three alternate sizes of network for medium term development, the chapter also identifies a longer term opportunity for extending the network to Shotts, subject to identifying further heat sources to power the expanded network.

4.2 Heat Demand Analysis

4.2.1 Identifying Potential Demand

The Scotland Heat Map was used to identify areas of heat demand in Hartwood and the surrounding area. It was evident from an early stage that the only locations within an accessible distance of a geothermal resource were Hartwood, Allanton and Shotts. Salsburgh was also considered within the scope of the study but the low heat density in the town, lack of underlying geothermal resource and distance from viable sources of geothermal energy do not imply a likely economic opportunity for connection. Of these settlements, Shotts is by far the largest and would therefore be likely to be a main target for heat supply over the longer term. Allanton is also a relatively good target area as it is closer to the JHI Hartwood Farm and has a reasonable property density, with the majority of properties being semi-detached 4-in-a-block.

The overall demand for each area can be seen in Table 4.1 below.

Table 4.1: Summary of study area heat demands

Study Area

Properties

Total Heat Demand

Hartwood

83

1631 MWh

Allanton

567

9454 MWh

North Shotts

2328

38238 MWh

To the north-west of Shotts lies HMP Shotts, a large prison with a capacity for about 550 inmates. No heat demand data for the prison is available in the heat map, and an estimated demand per inmate was assigned using CIBSE energy benchmarks is therefore used. The exact number of inmates could not be confirmed and so the lower of the figures found - 528 inmates - was used giving an estimated yearly demand of 9,959 MWh. However, further research established that the prison has recently had its own CHP system installed, with further plans for wood biomass, and so would unlikely to connect.

4.2.2 Business as Usual Scenario

The alternative business case to district heating is assumed to be the retention of gas supplies, where gas networks currently are present, or electricity, oil or solid fuel. This business as usual (BAU) assessment is important to model a heat sales price from district heating that is competitive and will compel the local community to agree to connect. For the purposes of modelling a competitive heat price the status quo case is modelled based on a gas supply.

The cost of heat from a gas boiler comprises the fuel costs (both fixed and variable costs from the supplier) as well as the cost of maintenance and replacement of the gas boiler. Council and Registered Social Landlord (RSL) tenants do not pay for the boiler replacement, whereas owner-occupiers will include this cost.

Typical unit costs for gas (Based on a monthly billing or pre-payment tariff) were obtained from a price comparison website in January 2016 and are listed below:

Table 4.2: Published fuel costs from price comparison website.

Green Star Energy

British Gas

Scottish Power

SSE

E.On

Unit Cost

p/kWh

2.5

4.0

3.4

4.1

4.2

Standing Charge

p/day

32.8

26.0

20.6

25.8

31.5

Standing Charge corrected to fuel use

p/kWh

1.0

0.8

0.6

0.8

0.9

Fuel supply tariff

p/kWh

3.4

4.8

4.0

4.8

5.1

DECC also publish average domestic gas prices to customers and data for the third quarter in 2015 shows an average gas tariff of 4.9 p/kWh. This figure will be used as the basis for the alternative business case. The gross boiler efficiency of an existing boiler is expected to be 80% and therefore the unit cost of heat is calculated to be 6.125 p/kWh. The cost of boiler replacement can be calculated as a cost per kWh of heat used assuming a typical annual use of heat of 12,500 kWh and a boiler replacement cost[1] of £2000 on a 15 year replacement lifecycle equates to a unit cost of 1.07 p/kWh.

The total unit cost of heat in the alternative business case is therefore:

  • 7.19 p/kWh to owner occupiers; and
  • 6.13 p/kWh to Council/RSL tenants

4.2.3 Qualitative Evaluation of Target Areas

The three target areas are made up primarily of residential properties of mixed ownership. The tenure layer of the Scottish Heat Map shown in Drawing 4.1 illustrates the percentage of properties in an area that are socially rented. Darker shades represent higher percentages and it can be seen that Allanton and some parts of North Shotts have relatively high numbers of socially rented properties.

While exact figures for the proportion of socially rented properties can be extracted from the Heat Map Data for each area, more detailed and up to date information has been obtained from North Lanarkshire Council to ensure that the correct properties are focused on in the analysis.

Hartwood

This is the least populous of the three potential target areas and consists of 83 private residential properties, none of which are connected to the gas grid. Although off-grid properties often have higher heating costs - therefore increasing the saving potential of a DHN - guaranteeing connection uptake from all private properties is usually difficult.

However, during consultations with the local residents, high levels of community engagement and interest were shown in the project. This would improve the likelihood of connecting all properties to the network, and hence the chance of there being enough demand to justify the installation costs.

Allanton

This small village that lies to the south of Hartwood and consists of 567 properties, of which approximately 40% are council owned. There is also a small primary school on the eastern side which would help to diversify the network load, and - due to it being a council owned building - could be considered a very likely connection, therefore helping to increase the overall demand on the proposed network.

North Shotts

This is by far the largest of the three study areas and has a high proportion of socially rented properties; making it a very suitable target for a district heating network. Shotts also contains a large High School and supermarket, which could help to provide stable sources of demand to balance out the heavily residential demand profile on the network.

A network supplying Shotts from an energy centre near Hartwood would at some point have to cross the Shotts Railway line. This can be a very expensive procedure as Network Rail has been known to charge in the region of £20,000 - £30,000 p.a. at an interest rate of RPI + 5% for permission to build across - either under or over - one of their lines. However, these charges may not be applied - project team research with Aberdeen Heat and Power established that no charges are incurred by the heat network in Aberdeen which crosses a railway line.

Salsburgh

Salsburgh is an off-gas-grid town to the north-west of Hartwood Home Farm, with approximately 41% Council owned properties. Salsburgh's potential in this project is severely limited by the high CAPEX costs associated with the large length of pipe - approximately 4500m - that would be required to connect it to the rest of the geothermal network. To evaluate its potential as a stand-alone scheme, the network has been analysed with and without this large section of pipe.

4.2.4 Heating Network and Demand Analysis

The District Heating Opportunity Assessment Tool (DHOAT) developed by Ramboll for the Scottish Government was used to analyse the heat map data and preliminary network designs. This tool provides clear indications of what the peak and annual demands would be, as well as a preliminary set of KPI data including Linear Heat Density[2] (LHD) and indicative network CAPEX and OPEX costs.

The outputs of each scenario were compared against one another to help determine which would be most suitable for a heating network; this decision is based on both the technical figures derived from this modelling as well as knowledge on the local property types, potential connection issues and any other commercial or construction hurdles.

4.3 Preliminary Network Analysis

North Lanarkshire Council (NLC) have a duty under the Energy Efficiency Standard for Social Housing (EESSH) to improve the overall energy efficiency of social housing by 2020. To achieve this, a minimum energy rating was introduced for gas or electric heated social rented homes. Over the next 5 years, NLC will focus on the homes that do not already meet this standard.

North Lanarkshire Council's Housing Department, as project partner with responsibility for council tenants, provided an initial focus on council owned properties for this project. In addition, the Council would be a key stakeholder for connections to a network and building fabric improvements; therefore the networks were designed around areas of high council property density. Two scenarios were analysed for each network; one where only council properties were connected and one where all properties in that area were connected. The effects of this will be discussed fully in Section 4.3.

4.3.1 Preliminary Network Layouts

HW011 - Hartwood

This network would supply heat to all of the properties in Hartwood. Its proximity to JHI owned land and therefore the proposed well locations would reduce the length of connecting pipe required and therefore the overall network costs.

HW021 - Allanton Council Only

This option would supply all of the existing council housing and the local Primary School in Allanton, to the South of Hartwood. As discussed above, these properties are fairly spread out so do not represent a particularly high demand density.

HW022 - Allanton All Residential

This option is the parallel analysis of HW021 that connects to all available properties, not just council owned ones. This more than doubles the number of connections for the same length of pipe, which would have a similarly large effect on the LHD.

HW031 - North Shotts Council Only

Although Shotts is the largest area of demand, its council properties are also very spread out, once again resulting in a low LHD. The area also includes a fairly large supermarket, a Primary School, and a large High School that will help to raise demand and provide a more diverse demand profile.

HW032 - North Shotts all properties

This variation to the North Shotts network would connect all the properties that are in reach of the network that would be designed solely for the council properties. As a lot of the buildings are multi-residence, connecting these extra properties would not incur a huge added costs over the work already suggested for option HW031. This option would require a similar length of pipe to supply a much greater demand, raising the LHD significantly.

HW041 - Shotts Council Only & Shotts Prison

These final variations on the Shotts network analysed the effect of connecting Shotts Prison. However it was discovered that the prison has its own CHP system and so no further analysis was carried out beyond this initial opportunity model.

HW042 - Shotts All Residential & Shotts Prison

Again, this option is a variation on HW041 that includes all of the available properties. However as mentioned, the potential for connection to the prison is very unlikely so will not be analysed any further than this stage.

HW051 - Salsburgh

The proposed Salsburgh network can be seen in the Drawing 4.2. This includes connections to all the properties that are in reach of the network that would be designed solely for the council properties. As discussed, the network will be analysed with and without the long connection that stems from the Northern end of the Hartwood section. The analysis without it relates to network HW052 in Table 4.3.

4.3.2 Modelling Results

The linear heat density (LHD) offers a useful rule of thumb to assess the potential economic viability of a network, and while networks with low LHDs can be made to work commercially, their viability is sensitive to factors such as the heat sales price and finance costs and they are likely to have to operate on a low IRR thus potentially reducing the finance options available. Where the prevailing heat source is not gas then there is a greater carbon and cost benefit to consumers to switch to district heating and the heat supply cost to consumers can increase while remaining competitive with oil or other alternatives.

The scenario that has the highest LHD is the extended North Shotts network that theoretically connects to the Prison, which as previously mentioned is highly unlikely due to their recently installed CHP and wood biomass aspirations. Despite this option not being likely, its analysis does show the effect that one large source of demand can have on a network's performance, and should be kept in mind as any large future developments in the area could offer up large demand and similar performance benefits for the network.

Some of the council housing in both Shotts and Allanton has been sold off, with the remaining properties now interspersed with owner-occupiers. This results in a more widely distributed demand, requiring longer lengths of pipe per property, hence a lower LHD in a scenario where only Council properties are connected.

A side by side comparison of the various scenarios can be seen in Table 4.3 and clearly shows the negative performance impacts that result from only connecting council owned properties. When comparing the model results for Allanton - projects HW021 and HW022 - it can be seen that connecting only council properties results in a LHD that is almost three times lower than that of the mixed tenure network. For all other options the decrease is approximately half. This would suggest that if a network in this area is to have any chance of being economically feasible, it would have to supply private properties as well as Council-owned ones.

Table 4.3: Summary of technical parameters and potential revenue for each of the nine heat supply options.

Short Name Reference

HW011

HW021

HW022

HW031

HW032

HW041

HW042

HW051

HW052

DHN Design Option

Hartwood

Allanton: Council Only

Allanton: Mixed Tenure

Shotts: Council Only

Shotts: Mixed Tenure

Shotts & Prison: Council Only

Shotts & Prison: Mixed Tenure

Salsburgh

Salsburgh: Stand-Alone

Proposed Supply Asset

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Mine-Water Geothermal

Stand Alone System

Technical Parameters

Network Length [m]

1,308

4,535

4,535

12,937

12,937

13,300

13,300

9,030

4,565

Total Heat Demand [MWh]

1,600

2,600

6,700

9,600

24,100

20,200

34,800

8,400

8,400

Peak Demand (MW)

0.8

1.3

3.3

4.8

11.7

8.7

15.6

4.2

4.2

Primary Supply Asset Capacity (MW)

0.192

0.312

0.804

1.152

2.892

2.424

4.176

1.008

1.008

No. of Connections

83

152

371

687

1527

727

1567

445

445

Linear Heat Density MWh/m

1.25

0.56

1.47

0.74

1.87

1.52

2.61

0.93

1.83

Required Source Flow Rate (l/s)

6.9

9.5

25.0

49.1

108.5

82.7

142.1

36.0

36.0

Potential Revenue

Weighted Average Heat Selling Price to Customers £/MWh

£63

£60

£61

£62

£60

£31

£43

£62

£62

Revenue

£101k

£155k

£409k

£596k

£1,457k

£636k

£1,497k

£520k

£520k

4.3.3 Land Ownership

An important consideration will be the location and route of the energy centre and district heating network respectively. The acceptability of locating the energy centre and production and/or re-injection wells as well as the cost of wayleaves and civils cost will be influenced by the land use and ownership. The James Hutton Institute own a large portion of land around the Hartwood area - as shown in Drawing 1.2 - which is a consideration for any network to be proposed there as it could guarantee an area for the energy centre and potentially the production and/or re-injection wells.

4.4 Preliminary Scenario Appraisal

While a high level analysis of several network scenarios was useful in the initial stages of the project to outline the overall scope of demand potential, a detailed look into the more feasible networks was required to meet the final project requirements and ensure the results from any financial analysis and other modelling were as accurate as possible. The network options were therefore narrowed down based on their KPI (as mentioned in Section 4.3), how their capacity matched the geothermal resource and other non-technical factors.

4.4.1 Scenario Flow Requirements

To provide an initial estimate of the well requirements for each scenario, outputs from the early stage modelling were converted into required heat pump capacities and hence required flow rates. A diversification[3] factor of 0.6 was used for all scenarios and the Heat Pump was assumed to provide 45% of this. The table below was constructed using the following equation based on a coefficient of performance (COP) of 3 and a temperature drop of 5°C.

Mathematical Equation

Table 4.4: Estimates of required mine-water flow rates for heat pumps at 45% of the diversified peak.

DHN Design Option

Total Annual Demand (MWh)

Undiversified Peak (MW)

Diversified Peak (MW)

Required Heat Pump Output (MW)

Flow Rate (L/s),
COP = 3

Hartwood

1600

0.8

0.48

0.22

6.9

Allanton: Council Only

2500

1.1

0.66

0.30

9.5

Allanton: Mixed Tenure

6000

2.9

1.74

0.78

25.0

Shotts: Council Only

11300

5.7

3.42

1.54

49.1

Shotts: Mixed Tenure

25700

12.6

7.56

3.40

108.5

Shotts & Prison: Council Only

21900

9.6

5.76

2.59

82.7

Shotts & Prison: Mixed Tenure

36300

16.5

9.9

4.46

142.1

Salsburgh

8400

4.2

2.5

1.13

36.0

Salsburgh: Stand-Alone

8400

4.2

2.5

1.13

36.0

4.4.2 Determining Geothermal Heating Potential

Whereas a traditional heat network would have its extent defined either by choice or by available demand, the extent and capacity of a geothermal based network is mostly defined by the heating potential of the target resource. This is determined by both the temperature - which influences the achievable COP of the heat pump - and extraction flow rates from the resource.

Initial research from the geology members of the project (Chapter 3) revealed that the target mine workings are under artesian conditions, resulting in an overflow from an old well shaft close to Allanton of approximately 18l/s. Further to this, when Kingshill Colliery No.1 was previously being pumped to prevent overflow, it was at a flow rate of 41.7 l/s for several years. These figures were used to calculate a range of heat production rates, which can be seen in Table 4.4. A temperature drop of 5°C was assumed and values calculated using the following equation:

Mathematical Equation

Table 4.5: Indicative heat pump outputs in kW for various flow rates and COP values.

Heat Output (kW) for Varying Flow Rate & COP

COP

Abstraction Flow Rate From Well (l/s)

10

15

20

25

30

35

40

45

50

2.0

418

627

836

1045

1254

1463

1672

1881

2090

2.5

348

523

697

871

1045

1219

1393

1568

1742

3.0

314

470

627

784

941

1097

1254

1411

1568

3.5

293

439

585

732

878

1024

1170

1317

1463

Upon comparison of these initial estimates with the required flow rates and heat demand values in Table 4.5 it can be seen that there is likely to be enough flow to supply the smaller networks proposed at Hartwood and Allanton. In theory there may also be enough for a network in North Shotts supplying only Council properties. However, as was discussed in Section 4.3.2; for the Shotts proposals to be economically viable, they would need to connect to as many properties as possible, therefore increasing the network demand and required source flow rate to over 100 litres per second which is not considered to be a sustainable abstraction rate from a single borehole and therefore raises the capital costs and risk profile.

Figure 4.1 uses a load duration curve to illustrate that one geothermal production well producing up to 45 L/s can only provide a small proportion of the annual Shotts heat demand (blue area). It is estimated that three geothermal production wells would be required to provide enough heat for a Shotts district heating network. This should be explored in future analysis with a view of expanding the district heating network to Shotts, but the required flow rates from the geothermal resource may not be achievable (see Section 3.4.6.). Once the demonstrator minewater geothermal project has been proved, this may provide evidence that the mines can sustain additional wells with higher abstractions to allow expansion of the network to Shotts.

Figure 4.1: Shotts district heating load duration curve showing heat provided from one geothermal production well in blue, and the heat that would need to be provided from additional heat sources in red.

Figure 4.1: Shotts district heating load duration curve showing heat provided from one geothermal production well in blue, and the heat that would need to be provided from additional heat sources in red.

4.4.3 Re-Evaluation of Project Extent

North Shotts

Due to the uncertainty around achieving flows above the previously extracted rates the network supplying North Shotts would probably be too large a network for the geothermal resource being targeted. As the largest study area in the project, its associated capital costs were estimated to be as much as £34m, compounding the financial risks of the well not meeting the required production rates. Because of these factors, Shotts was not considered as a final design option and was not included in any further analysis. However this study has shown that North Shotts is a potentially economically viable location for a district heating network with alternative or complementary renewable heat sources to the minewater geothermal resource.

Salsburgh

The network up to Salsburgh was analysed including all properties and with the full length of pipe required to connect it indicated a LHD of 0.97. This would suggest that it is not a viable network option for this scenario as the heat sales could not be enough to cover the high CAPEX costs of installation. Added to this, the heat loss in such a long section of pipe would be fairly significant, further increasing the running costs of the network. For these reasons, Salsburgh was not considered as a final option for more detailed analysis although it should be noted that as a network which was able to secure its own nearby supply (from a non-geothermal resource), it has significantly more potential.

Minimum Network Extent

As there are high CAPEX costs associated with both the geothermal well and the DHN, a very small network is unlikely to provide enough revenue to pay these back over a feasible timescale. The smallest of the scenarios previously analysed was the network at Hartwood. This would only provide heat to 83 properties, none of which are council owned; therefore excluding NLC as a potential project partner. Despite the fact that Hartwood is not on the national gas grid - and therefore likely has more expensive heating costs - this added value was deemed too small to overcome the other large barriers, therefore Hartwood would not be considered for a network on its own, only as part of a larger network including Allanton.

Target Property Types

As discussed in Section 4.2.2, networks that only connect to council housing would tend to perform poorly due to lack of demand relative to the size (and cost) of the network. The scenarios taken forward into further analysis were designed on the assumption that all properties, private and council owned would be connected.

4.5 Final Network Design

4.5.1 Energy Centre Location

Typically in district heating design, it is good practice that the energy centre should be located as close to the demand as possible to minimise the heat losses and costs associated with long lengths of high capacity insulated transmission pipe. However, due to the high iron content of the water in the target mines of this project, it is desirable to extract heat as close to the production well as possible to minimise the risk of exposure to oxygen and hence fouling of the equipment.

In a situation where the production well is not located close to the heat demand, this creates a set of opposing requirements. To meet both of them, the systems have been designed with an intermediate section of pipe that hydraulically separates the energy centre from the well by extracting heat at the well head and transporting it via this loop of clean water to the energy centre.

Option 1

Of the two most suitable production wells, the closest to Allanton is Option 1, located at the site of the old Kingshill Colliery No.1 mine-shaft. The energy centre for the two potential Allanton networks (Networks A and B) would be located on the outskirts of the town as shown in Drawing 4.4 with an interconnection loop of 650 m to the production well. It should be noted that neither the production well or Energy Centre are located on JHI's land in this options.

Option 2

For Network C supplying both Allanton and Hartwood, the energy centre would be best located equal distances from either source of heat demand to reduce heat losses. There is an area of JHI owned land that is ideally suited for this, just off Hartwood Road between the two villages as shown in Drawing 4.4. This makes Option 2 the most suitable for this network as it is located just 450m away, also on JHI's land. It should be noted at this point that Option 1 is also compatible with Network C.

4.5.2 Final Network Design

To allow for some variation in the final analysis, networks were designed for three different sizes of system. The results of these would then be compared to determine the most economical scale for the district heating scheme. The options can be seen in Drawing 4.5 and Drawing 4.6 are described below.

Network A

This is the smallest network proposed, located in the West side of Allanton and connecting to all houses located within the extent of the council owned properties.

Network B

This proposes a larger network in Allanton that supplies the same properties as Network A, but also includes an extension to connect the Primary School and properties over to the East.

Networks A and B are displayed in Drawing 4.5.

Network C

This option is the largest of the proposed schemes and would supply all properties in Hartwood as well as the majority of properties in Allanton through a common network.

Network C is displayed in Drawing 4.6.

4.5.3 Matching Demand to Supply

As discussed in Section 4.3.2. the heat demand must be proportionate to the potential heat supplied from the geothermal system. Given the demonstrator status of the proposed project only one production well is suggested in each system, providing a maximum of 45 L/s of minewater. As Figure 4.1 illustrated how Shotts can be ruled out at this stage due to heat demand being too high, Figure 4.2 below shows how Network C can be "ruled-in" as an optimal heat market for a single minewater geothermal production well.

Figure 4.2: Network C (Allanton and Hartwood) district heating load duration curve showing heat provided from one geothermal production well in blue, and the heat provided from the thermal store and back-up boiler.

Figure 4.2: Network C (Allanton and Hartwood) district heating load duration curve showing heat provided from one geothermal production well in blue, and the heat provided from the thermal store and back-up boiler.

4.5.4 Detailed Analysis of Demand

Upon finalisation of the network layouts, the heat map data for the properties to be supplied was analysed to determine the peak demand for each scenario as well as the demand profiles. This enabled the energy centre to be correctly designed to meet the network demands without being oversized. The daily undiversified peak demands for Network C can be seen in Figure 4.2. These profiles are simply the sum of all individual properties and so represent a very high peak.

Figure 4.3: Peak daily demand profiles for Network C.

Figure 4.3: Peak daily demand profiles for Network C.

Diversification of Demand Profiles

Due to the fact that not all individual property peaks will occur at exactly the same time, the demand profiles must be diversified to prevent oversizing the system. As the number of properties on a network increases, the diversity also increases. This leads to a significant reduction in the peak demand seen at the energy centre.

Table 4.6: Technical parameters of network options.

Network

Total Annual Demand (MWh)

Peak Demand (MW)

Diversified Peak Demand (MW)

Average Property Demand (kWh)

Total Network Length (m)

Linear Heat Density (MWh/m)

A

3860

2.79

1.45

11767

1730

2.23

B

5713

4.48

2.33

12439

3328

1.72

C

9670

8.38

4.36

15250

6175

1.57

The energy centres and heat pumps were sized based on this diversified peak to ensure that the top-up boilers and overall plant capacity could always meet the entire network demand.

It should be noted that the peaks visible in Figure 4.4 are lower than the overall diversified peak demand as it is based on a monthly average, which will be lower as the peak will not be reached every day of a month.

Figure 4.4: Yearly diversified demand profile for average day in month, Network C.

Figure 4.4: Yearly diversified demand profile for average day in month, Network C.

4.5.5 Sizing Network

The pipe sizes and hence costs for the proposed networks were determined using Ramboll's thermal and hydraulic modelling software, System Rørnet. The heat demands from individual buildings were grouped, then diversified and applied to nodes on the main branches of the network. This simplification has very little effect on the end results, but speeds up analysis by not sizing each individual property connection.

The software calculates the appropriate diameter for each section of pipe based on the demand on its nodes. This generates a pipe schedule that shows total lengths required of each pipe size and enables accurate heat losses, pressure drops and costs to be determined for each network option and temperature scenario.

4.6 Energy Centre Design

4.6.1 Heating System Overview

The water to be extracted from the mines below Allanton is expected to have a temperature of about 18 °C which necessitates a heat pump based system.

Heat will be extracted from the minewater using a heat pump and upgraded to the required network flow temperature, covered further in Section 4.6.2.

The system will require back up gas boilers that meet the full output capacity of the system, both to meet peak demands and to cover for any down-time that the heat pump may experience.

Figure 4.5: Indicative energy centre block diagram for all scenarios. Exact operating conditions such as flow rates, temperatures and outputs will vary.

Figure 4.5:  Indicative energy centre block diagram for all scenarios. Exact operating conditions such as flow rates, temperatures and outputs will vary.

4.6.2 Network Operating Conditions

Two network scenarios have been analysed: a low temperature and a high temperature system. The low temperature network could operate on 75/45 flow and return temperatures, resulting in a temperature raise of 57°C, and the high temperature network will operate on 85/60 flow and return temperatures, requiring a temperature raise of 67°C (these scenarios are highlighted in bold in the table below). These different temperature raises affect the COP that a heat pump can achieve and in turn the cost of heat production.

Table 4.7: Indicative cost of heat production for various values of temperature raise. COP data from supplier.

Temperature Raise (°C)

COP at Flow Temp

Indicative Cost of Heat for Varying Electricity Price (p/kWh)

6

8

10

12

14

52

3.46

1.73

2.31

2.89

3.47

4.05

57

(low temperature 75/45)

3.06

1.96

2.61

3.27

3.92

4.57

62

2.70

2.23

2.97

3.71

4.45

5.19

67

(high temperature 85/60)

2.36

2.54

3.39

4.24

5.09

5.93

72

2.05

2.93

3.91

4.89

5.86

6.84

The network operating temperatures can also affect other aspects of the energy centre design, network capacity and heat losses and associated costs. Lower network temperatures will result in lower heat losses from the pipe network. A wide delta-T reduces the flow rate in the network and therefore offers lower pipe diameters.

To be eligible for the renewable heat incentive (RHI) heat pumps must perform with a COP of 2.9 or greater. In the High Temperature scenarios this is not achieved and the sensitivity of this has been tested in the financial model. It is likely that qualification as a Deep Geothermal heat source and the resultant RHI tariff will influence the economic viability of the project (see Chapter 7).

The most significant impact of operating a low temperature network is the way that it affects customers. This is further discussed in Section 4.6.4 below.

Ruling-out of High Temperature Network

The results of the financial analysis prepared to inform Chapter 6 indicate that the financial performance of all three network designs is similar when deploying a high temperature and low temperature network. The higher capital costs of the low temperature system are off-set by the higher operating costs of the high temperature system for a network of this (relatively small) size.

In addition, the COP of the heat pump's performance in the high temperature network is consistently below 2.9 (see Table 4.8), which is the minimum required COP to be eligible for the RHI. Without the RHI the project is not economically viable.

The high temperature network option is therefore ruled out at this stage.

It is possible that bespoke heat pumps could achieve high temperatures whilst performing above a 2.9 COP. If so, this could be explored in the next stage of design.

4.6.3 Reducing Fouling Risk

The risk of fouling caused by iron precipitation will be minimised by installing an additional brazed plate heat exchanger right at the well head, reducing the length of pipe minewater travels through and hence the likelihood of a leak. The heat will be transferred into a secondary source loop of clean water that will then act as the cold side of the heat exchanger, eliminating any contact between the minewater and more expensive plant equipment.

4.6.4 Building Requirements

Traditional gas central heating systems are designed to operate on flow/return temperatures of 82/71, which does not provide a high enough temperature drop for either a low temperature or high temperature DHN to be effective. For a property to maintain the same level of thermal comfort while reducing the mean temperature of the radiators, extra insulation is required or larger surface area emitters (radiators) are required.

Whatever network temperature is selected, the thermal comfort in buildings must be maintained. If network temperatures are reduced to optimise the network and heat generation efficiency then the investment to allow compatibility of connecting buildings to these temperatures must be included in the financial modelling for the project. Analysis of these costs can be found in Appendix A4.

4.6.5 Thermal Store Sizing

The thermal stores for each scenario design have been sized to hold 3 hours' worth of the peak heat pump output. This is on the higher side of some guidelines and estimates; however this larger capacity will allow the heat pump to cover a higher proportion of the peak network demand, hence reducing the need for additional gas boiler top-up.

The store's physical size requirements are based not only on the capacity desired, but the temperature difference at which the water can be stored. This should ideally be the same as the difference between the network flow and return temperatures, provided the store is thoroughly insulated. This required size was calculated on a per MW basis, to be applied to all scenarios.

Mathematical Equation

Table 4.8: Required volume of thermal store for low and high temperature networks.

Network

Heat Pump Output (kW)

Store Volume(m3): LT, ∆T=30C

Store Volume(m3): HT, ∆T=25C

A

700

60

72

B

1000

86

103

C

2000

172

206

4.6.6 Heat Pump Operating Profile

An excel model was created to determine the operating profile of the heat pump for each network option. This was done to provide detailed operating costs as well as a profile that RHI payments could be applied to in the financial analysis.

The main operating cost for a heat pump consists of the electricity used. Due to the tariff-based structure that many electricity providers operate on, this can vary largely depending on what time of day the heat pump is operating.

The benefit of using a heat pump along with a large thermal store is that the fuel costs can be minimised by charging the thermal store during off-peak times while electricity is cheaper, then turning off the heat pump during peak times and using the stored energy to cover the gap.

Electricity Tariffs

The main variation in electricity price for non-domestic customers comes from the Distribution Use of Service (DUoS) charges. These are shown for the local Distribution Network Operator (DNO) in Table 4.9:

Table 4.9: DUoS time bands and charges for high voltage connection with Scottish Power Energy Networks, assumed to be half-hourly metered.

Charging Tariffs

High Tariff

Medium Tariff

Low Tariff

Charge (p/kWh)

7.73

0.42

0.02

Time Band

16:00 to 19:00

09:00 - 16:00
19:00 - 20:30

00.00 - 09.00
20.30 - 24.00

It should be noted that these charges are additive to the standard electricity prices. This base price was taken as an average over the two most recent quarters from the DECC quarterly non-domestic fuel prices for the appropriate electricity consumption bands as 10.47 p/kWh.

Table 4.10: DECC quarterly non-domestic electricity prices. Averaged values shown in bold.

Size of Consumer (MWh)

Cost of Electricity (p/kWh)

Q1 - 2015

Q2 - 2015

0-20

13.90

13.41

20-499

12.35

12.10

500-1999

10.98

10.82

2000-19999

10.08

9.99

20000-69999

9.73

9.84

70000-150000

9.56

9.55

150000+

9.16

9.13

Average

10.62

10.46

This base electricity price plus the DUoS charge results in an energy cost of 18.2 p/kWh and - with reference to the process used in Table 4.7 - an equivalent cost of heat produced during peak times of approximately 6 p/kWh, compared to approximately 3.4 p/kWh during off-peak times.

Heat Pump Model

The excel model created was based around the premise of avoiding heat pump operation during the peak time for electricity rates. The technical parameters of each scenario such as the thermal store size and heat pump capacity were used to determine a daily operating profile, an example of which can be seen in Figure 4.6.

As previously mentioned, the initial comparison of peak time heat pump operation and heat from gas boilers did not include the RHI. This was due to several factors. Firstly, the future of the RHI was somewhat uncertain at the time of the model's construction - government budget announcements were pending - and secondly, it was deemed beneficial for the modelling that the thermal store was completely emptied at least once a day, thereby utilising the cheaper heat generated during off peak times.

Figure 4.6: Heat pump modelling output for February of network option 1A - breakdown of heat utilised.

Figure 4.6: Heat pump modelling output for February of network option 1A - breakdown of heat utilised.

As shown in Figure 4.6, the heat pump is running at full capacity overnight to charge the store, which is then partially depleted during the morning peak. The heat pump will then continue running at peak output during the afternoon to ensure that the thermal store is as full as possible and able to provide a large portion of the evening peak demand while the heat pumps are off.

4.6.7 Pressure Losses and Distribution Pumps

The pressure losses for each system were calculated as part of the System Rørnet analysis. The required hydraulic power was also generated in this analysis then the pump power calculated through use of the total pump efficiency. This was estimated to be 65% based on the efficiency of the pumps recommended in the Grundfos online sizing tool.

Table 4.11: Network pressure losses, pumping power and annual power consumption.

Network Temperature Scenario

Network Option

Network Flow Rate (l/s)

Total Pressure Loss (kPa)

Hydraulic Power Required (kW)

Pump Power (65% efficient)

Total Power Consumption

Low Temperature

A

11.6

265

5.1

7.8

3138

B

18.6

375

8.5

13.1

5249

C

34.8

612

21.1

32.5

12997

High Temperature

A

13.9

191

6.0

9.2

3692

B

22.3

368

10.0

15.4

6154

C

41.7

549

28.2

43.4

17372

The pumps were assumed to run for approximately 400 Equivalent Full Load Hours (EFLH), giving the annual power consumption. This is notably lower than the heat pumps' EFLH as the relationship between flow and pump power is cubic; i.e. a pump running at 50% flow will be using just 12.5% of its max rated power.