The Morecambe Bay barrage is promoted by Northern Tidal Power Gateways (NTPG) (2020b). Their initial proposal employed 125 × 8 m dia. turbines with 30 MW generators.

The published barrage length is 17 km stretching from east of Heysham, on the southern shore, to west of Ramspide in Cumbria in the north. The seabed level along the line of the barrage is approximately −5 m OD (ordnance datum Newlyn) for 12 km of the length and −10 m OD for the remaining 5 km. In the cost model, the rated head of the turbine is taken as 75% of the spring tidal range. The mean spring tidal range is 8.5 m, giving an approximate rated head of 6.4 m. The mean high water springs (MHWS) level is 4.77 m OD (NTSLF, 2023). The published estimated capital costs are shown in Table 2, reproduced from NTPG (2020b). The published figures have been updated to 2022 prices using the CPI for new infrastructure (ONS, 2023); the increase from January 2019 to January 2022 is 1.14.

Table 2. Summary of costs for Morecambe Bay in £m as published (2019) and at current prices (2022)

This scheme is promoted by North Wales Tidal Energy (NWTE) (2023). NWTE proposes up to 125 × 8 m dia. 20 MW turbines. The seabed level along the line of the barrage is approximately −5 m OD for 12 km, −10 m OD for 8 km and −15 m OD for 12 km. The mean spring tidal range is 7.2 m, giving an approximate rated head of 5.4 m. The MHWS level is 3.51 m OD (NTSLF, 2023). The published estimated cost was £7.0 billion (George, 2020).

Using Equation 16 from Vandercruyssen et al. (2022b) with the updated rates from Table 1 gives the estimated costs of T-Gs in Table 3 at 2022 prices. The rated heads of the turbines, H_o, are approximately 75% of the spring tidal range as used by Fay and Smachlo (1983) and Vandercruyssen et al. (2022a) – that is, H_o = 5.4 m for North Wales, H_o = 6.4 m for Morecambe Bay.

Table 3. Estimated costs per T-G set (£m, 2022)

The efficiency of bulb turbines increases with the runner diameter. Those at Lake Sihwa were 7.6 m in diameter and were manufactured over 10 years ago. All the proposed runners are 8.0 m in diameter, which is about the largest considered to be available to date. The mean spring tide ranges for North Wales and Morecambe Bay are 7.2 and 8.5 m, respectively. The costs for a powerhouse (Table 4) are estimated using Equation 3 from Vandercruyssen et al. (2022b) with the updated rates from Table 1.

Table 4. Cost in £m of powerhouses and sluices for North Wales and Morecambe Bay

The definition used here for the SR is the total area of sluice aperture divided by the total area of the turbine runners. The sluices are assumed to be 15 m wide × 15 m high, giving an area of 225 m². The turbine runners are 8.0 m in diameter, giving an area for each of 50.3 m². Thus, for an SR of 1 the total area of sluices matches the total area of turbine runners with approximately nine turbines for two sluices. Using Equation 4 from Vandercruyssen et al. (2022b), the cost of a 15 m square sluice is also calculated in Table 4. Thus, for an SR of 1, there will be 0.22 sluices for every T-G unit.

The cost per metre of the cofferdams is taken from Equation 5 of Vandercruyssen et al. (2022b). The height H_b is the same as the crest level minus the level of the seabed at the turbines. The cost of the cofferdam per metre length is shown in Table 5.

Table 5. Cost of cofferdam per metre for North Wales and Morecambe Bay

The length of the cofferdam is considered to be proportional to the length of the powerhouse units and sluices (L_c). The width of the powerhouse unit (W_p) is taken as 16 m, for 8 m dia. turbines, and the width of each sluice (W_s) as 15 m. Thus, the total length of cofferdams for various numbers of T-G units (N_t+g) is calculated by Equation 1, giving the total cost in Table 6.

$$L_c=N_{t+g}\left(16+\left(15\times 0.22\right)\right)=N_{t+g}\times 19.3$$

1

Table 6. Total cost of cofferdams (£m) for various numbers of units with a SR of 1

The crest level of the bund is assumed to be the MHWS level plus 2 m for storm surges and 1 m freeboard for waves. This figure will need to be a few metres higher if a public road or railway is required as part of the scheme. Also, provision will be required to allow increasing the crest level in line with rising sea levels.

Using the published equations (Vandercruyssen et al., 2022b) and the data in Sections 3.1 and 3.2, the resulting cost/m of the alternative bunds are given in Table 7. Obviously, the bund with the 1:3 slope costs more than the one with the steeper 1:2 slope as it requires more fill material. However, assuming the same materials, the steeper slope is likely to require better compaction so the rate may vary slightly. Both options include an allowance for rock armour protection.

Table 7. Estimated cost of bunds for the Morecambe Bay barrage and the North Wales lagoon

The estimated costs of the components for the two case studies are given in Table 8. In a previous paper (Vandercruyssen et al., 2022b), the authors initially increased the capital costs by 30% of the civil engineering costs to allow for preliminaries (prelims), surveys, design and contingencies as used in the government-funded study of the River Severn Interim Options Analysis Report (Parsons Brinckerhoff Ltd, 2008). However, given the scarce data on turbine costs, efficiencies in reverse flow and triple regulation, the authors now believe the 30% figure should be applied to all costs. Inaccuracies will arise from errors in the rates and the assumed depths; published costs are usually overestimated due to pre-feasibility conservatism. However, the method shown should be suitable for pre-feasibility estimates and ranking schemes in order of financial return.

Table 8. Summary of estimated scheme capital costs for Morecambe Bay and North Wales

Both estimates are close to the developers’ published figures. The details of cost estimates of these and any other proposed scheme cannot be tested against existing values as the components are not published due to commercial concerns. The following text shows how the estimated costs can be reduced by optimising the components.

For best performance, the diameter of the turbine runners must be as large as possible to maximise the flow and turbine efficiency. The maximum diameter currently considered practical to manufacture is 7.6–8.0 m. Figure 1 shows the relationship between the generation output and the number of T-Gs of different ratings. For 125 × 30 MW machines, the predicted annual generation from Morecambe Bay is 6.58 TW/ha. The generation from the same number of 20 MW machines is 6.39 TW/ha, representing only a 3% reduction in output. It has been shown that the cost of the T-G is a function of the generator rating for a given rated head (Table 3). Thus, the cost of a 20 MW T-G with the same 6.4 m rated head is 70% that of the 30 MW machine. From Table 8 the 125 turbines represent 69% of the total capital cost. Reducing the generators to 20 MW saves 52% of the overall Capex for only a 3% reduction in annual generation.

The AEP is asymptotic, gradually flattening as the number of units increases. Figure 1 shows this consistently for all scenarios. The costing approach employed here enables the number of units for a particular scheme to be optimised against cost. The Morecambe Bay scheme has proposed both 125 units (NTPG, 2020a) and 160 units (Baker, 2021). Table 9 shows the calculation of costs and AEP for both schemes with various numbers of units. For an SR of 1, a single 15 × 15 m sluice will be required for every 4.5 turbines of 8 m diameter. It is assumed that the costs of bunds and contingencies will be the same for all options.

Table 9. Calculation for estimated costs and annual generation for different numbers or turbines and generator ratings

Table 9 shows that costs per TWh are significantly lower with smaller generators for both schemes. In terms of costs, the optimum for Morecambe Bay involves 120 turbines with 20 MW generators. However, 120 turbines for Morecambe Bay are not capable of maintaining the existing low tide levels against the higher predictions of sea level rise; the relationship will be examined in a subsequent paper. For North Wales, the most cost-effective option is 100 turbines with 15 MW generators. The cost per TWh for the estuarine barrage is 74% of that for the coastal lagoon.

The last column in Table 9 shows the estimated annual carbon dioxide (CO₂) offset, valued as the equivalent power generation from combined cycle gas turbines (CCGTs) operating at maximum commercial rate of 350 kg/MWh (Bass et al., 2011). Bass et al. (2011) measured the carbon dioxide emissions from a grid-connected CCGT under various operating conditions over a period of 3 months. During cold and hot the carbon dioxide emissions increased to 470 and 590 kg/MWh, respectively. If the goal is to generate as much renewable electricity as possible and maximise the carbon dioxide offset, then the optimum arrangements are different, as highlighted in italics in Table 9, at a slightly increased cost per TWh. Should a carbon tax credit system be available, then the economics will change in favour of more installed capacity to displace gas generation. The optimised generation from Table 9 would save 2.18 Mt (million tonnes) of carbon dioxide per annum from Morecambe Bay and 1.55 Mt from North Wales.

Carbon dioxide payback periods are another parameter to be considered in all new constructions. Hammons (2011) studied this for two of the proposed Severn estuary schemes and predicted carbon dioxide payback times of 5–8 months. This is the most rapid payback for power generation and compares favourably against other low carbon dioxide technologies such as nuclear power (SDC, 2006).

Sluices influence the efficiency of operation of a tidal barrage and the ability to maintain the tidal range over the seasonal cycle. Figure 1 includes the AEP for North Wales with SRs of 1, 2 and 4 for 20 MW machines. The costs of sluices and cofferdams and the total scheme costs are taken from Tables 4, 5 and 9 respectively. Assuming 125 turbines with 20 MW generators, the costs per TWh are given in Table 10 for various SRs. For this configuration the minimum cost per TWh comes from an SR of 2.

Table 10. Effect of various SRs on annual generation for North Wales

Currently, the government has two potential support mechanisms that could provide public finance to assist renewable energy. The principal one is contract for difference (CfD), which has been used extensively for wind farms and gives the developer a guaranteed price per MWh for electricity generated. The agreement is for a defined period (usually for 20–40 years) that is negotiated with the government regulator before detailed designs are drawn up. The developer works on the build, own and operate (BOO) principle. The developer and their financial backers carry all the risks of design, construction and operation and no income is received until the scheme is operational. For tidal range schemes, this could be 4 years for design and 6 or 7 years for construction. For mega projects the risks are high and finance will be expensive, discouraging private investors. Investors are reluctant to consider projects with an IRR of less than 10–15%.

The alternative support mechanism is called regulated asset base (RAB) and has been used for major infrastructure projects, such as London Crossrail and Heathrow Airport Terminal 5. It is being considered for new offshore wind and has been approved for the Sizewell C nuclear power station (Makovšek and Veryard, 2016). The mechanism employs a risk-sharing approach with backing from the government regulator. The risk sharing and profit margins are agreed between the developer and regulator before detailed design. Consequently, the investors carry less risk and the available interest rates will be about half that of the CfD mechanism. RAB is better suited to a 120-year tidal range project, benefitting both parties and saving money for the electricity customers in the long run. Under RAB, income is available from financial closure of the agreement (i.e. before construction starts) so that the effective debt built up in the project is reduced. Price support is unlikely to be required after a period of 40 years when a plant upgrade would be required. Another benefit of RAB is that the regulator can stipulate broader conditions such as tidal range management for specific objectives. Constraints could include stopping generation early to provide flood protection or pumping to match existing low water levels for ecological reasons.

To highlight the impact of funding costs and support mechanisms on the economic viability of schemes, the IRR has been calculated for the case studies. The support price (electricity support price in £/MWh, see Table 14) is adjusted to give an IRR = 10% for CfD and 5% for RAB.

Table 14. IRR analysis

In addition to the assumptions detailed in Section 6, revenue and costs are assumed to increase at roughly the same rate so inflation is ignored in the analysis. Under these assumptions, the IRR is calculated in a spreadsheet as shown in Table 14. The analysis is carried out over 40 years of operation.

While not comprehensive – for example, tax is not included – the approach is adequate to indicate the impact of methods of funding. The RAB model gives a price of electricity less than half that from the CfD model. The RAB price, for both schemes, is below the average wholesale electricity price since 2020, from Table 13. This analysis demonstrates that tidal range is economically viable when the RAB method of funding is used.

Even with CfD funding the support price is lower than the average for 2022. Currently there is much concern that some wind generation operators are making excessive profits. The payment system allows some of them to benefit from high wholesale electricity prices due to the increased cost of gas generation.

For the foreseeable future, electricity will always be required. The RAB model of funding is better suited to long-term infrastructure projects. It is vital to plan for large infrastructure projects as it is likely to be at least 11 years before such schemes are productive.

After the first 40 years the electro-mechanical equipment will be refurbished or replaced on a rolling programme. The cost will be about half the original capital cost, while the revenue will continue at about 80–90% throughout the refit period of 5–10 years. Thus, the IRR for the remaining 80 years of the project will be about double that of the first 40 years. No further subsidies will be required.