Skip to content

Swansea Bay Costings

May 12, 2018

By Paul Homewood



Following the Select Committee’s meeting to discuss the Swansea Bay tidal lagoon, I have looked more deeply into the costings.

One of the complaints made at the Committee concerned the lack of detail provided to MPs and the public about the pricing structure. Given that the public will end up paying the bill, this seemed a fair criticism.

Under questioning, Mark Shorrock of Tidal Lagoon Power offered up a few tidbits, but still left many questions unanswered.

He told us that he would need a 60-year CfD contract at £92.70/MWh, making it competitive with Hinkley Point. However, things are not as straightforward as they appear.

The strike price for Swansea Bay would not be fully index linked for inflation, only partially so. As the Hendry Review explained, this will mean that the “real price” will reduce over a period of time:



The logic of partial indexation is sound. As the bulk of the costs are upfront, it would be unfair to consumers to inflate that portion of the strike price each year. Instead strike prices would only need to be inflated to cover running costs.

When Shorrock talks about a strike price of £92.70/MWh, he means the average over the contract period of 60 years.

In other words, the starting price will be much higher, and the price at the end lower, than £92.70. But it is prices in the early years that are of most importance to us, not what they may theoretically be after we are dead.

And on that matter, Shorrock has remained totally silent on just what that opening price will be. He has also not said how much of the strike price would be indexed.

But we can make some informed guesses.

Shorrock did tell us that running costs would be £16.5m a year, based on output of 550 GWh. That equates to £30/MWh. From this calculation alone, we can see that it is questionable whether the strike price will ever fall below the market electricity price of around £45/MWh.

I have run a spreadsheet, based on:

1) Running costs of £30/MWh to be index linked

2) Annual inflation of 2.5%

3) Total opening strike price of £150/MWh, (leaving £120/MWh as fixed.)


Over the 60 years of the contract, the average price (at current prices) would be £93.40/MWh, in other words almost the same as Shorrock’s figure of £92.70/MWh.

Of course, he might be assuming different inflation rates and/or a different element of fixed costs. But I doubt whether he is far away from my assumptions.

If I am right, that opening price of £150/MWh will surely be unacceptably high to the government. Furthermore, the price only falls below Hinkley in Year 30. In current price terms, it will still be £58/MWh in Year 60, still higher than the market price of electricity.

If Shorrock really is looking at a starting price anywhere near to my numbers, it is hardly surprising he is reluctant to tell anybody. It is a pity that the Select Committee did not push him further on this. I suspect they are not even aware of the implications of how the pricing mechanism works.

As to the “benefits” for consumers in 60 years time, there is no guarantee the lagoon will even be working by then.


Swansea Bay’s Viability?

Shorrock reckons he has about £25m from share capital at the moment, but will rely largely on commercial bonds for the £1.3bn needed to build the lagoon.

Based on the above costings, he would need to borrow at around 3 to 4% for the project to be viable. This is far lower than anything I am aware of on the normal commercial market. Both Hendry and Shorrock hope that they can present this scheme in the same way as utilities, such as water companies, who have long life assets and reliable income streams, and thus can borrow relatively cheaply.

Whether investors see Swansea Bay quite this way is another matter.

Presumably most of the bonds will be long maturity, but the question remains – how will Tidal Lagoon Power be able to redeem the bonds at maturity?  Companies usually rely on issuing new bonds, to pay off the old ones, on a roll over basis. If Swansea Bay has a limited life span of say 60 years, that will eventually be impossible.

Even if they manage to roll over shorter term bonds, what rate of interest will they have to pay on the next issue? Market rates may well be much higher in ten or twenty years time, which would make the whole financial model unworkable.

There is also the question of major refurbishment in years to come. Even Hendry recognised that major work might be needed to keep the turbines and seawall in working order as they get older. We could well see a situation where the company simply does not have, or cannot borrow, the funds needed to keep the lagoon operational. At best, we could well see output dropping substantially, which again would leave the whole business unviable.

And we have not even touched on the question of silting.

All in all, there must be a very great risk that lenders’ money will be in danger. As for the shareholders, a few fat dividend cheques will see them recoup their money before deserting the sinking ship.


Cardiff Bay and the rest

The only real logic to the deal from the government’s point of view is that Swansea will act as a pathfinder, which should enable much bigger lagoons to be built at Cardiff and elsewhere at much lower cost/MW.

However, there is another large snag here, even if Shorrock’s figures are right.

As the bulk of the cost is upfront, the level of interest rates is crucial, as Hendry illustrated:


Allowing for the fact that Swansea Bay will cost £1.3bn, rather than the £1bn used as an example, an increase in interest rates of 5% would add £130/MWh to costs.

Interest rates are, of course, ridiculously low by historical standards. Whether they stay at this level or rise during the next decade or so is anybody’s guess.

But even a rise of 2% would totally destroy the business case for the likes of Cardiff, which are unlikely to be built before 2030.

There has to be a very real risk that these bigger lagoons will never be built. In this case, shelling out what would amount to £1.6bn over its life in subsidies to Swansea Bay makes no sense at all.


Final Thought

I showed this calculation in my post yesterday, but it goes to the heart of the economic argument about the value of Swansea Bay.

By Shorrock’s own admission, running costs will be £30/MWh, all to generate electricity worth £45/MWh.

In other words, there will be added value of £15/MWh, which based on annual output of 550 GWh  equates to £8.25m a year.

On what planet would such a paltry return on an investment of £1.3bn even be considered?

  1. swan101 permalink
    May 12, 2018 6:41 pm

    Reblogged this on UPPER SONACHAN WIND FARM.

  2. May 12, 2018 6:43 pm

    “On what planet would such a paltry return on an investment of £1.3bn even be considered?” The planet on which the B Ark crashed.

  3. It doesn't add up... permalink
    May 13, 2018 1:33 am

    If you look at their latest projections, you find that they plan to use double action turbines from Andritz that are capable of pumping as well as generating. Now, close to high or low tide, pumping can make sense (because the moon has done most of the work through gravity, and otherwise you are waiting for the tide to turn and change enough to have a useful head for generating). However, that requires buying grid power to do the pumping. Of course, if you can get grid power at market price, and convert it (less an efficiency round trip penalty) into highly subsidised CFD price power, it’s a winner – and an incentive to over-invest in turbines, and pump beyond what would be economic e.g. at Dinorwig or at La Rance, where they do do some pumping. Once the differential falls, it may no longer pay to do the pumping. Pumping is of course a market linked cost. It also exacerbates the grid profile of generation, drawing from the grid before switching to a burst of power output.

    Incidentally, I found that Ecotricity’s Solway Firth partners, Tidal Electric, made some interesting comments:

    They are of course exaggerating on a number of issues, but all is fair in love and war, eh?

    • Ben Vorlich permalink
      May 13, 2018 11:48 am

      Won’t pumping only work at high tide to raise the water level inside the barrier? Pumping at low tide will be of restricted by the position of turbines relative to bottom of lagoon and the fact that pumping the lagoon out won’t raise sea level. Or have I miss understood how pumping works in this case?

      • It doesn't add up... permalink
        May 13, 2018 1:27 pm

        You may find the following diagram of double-effect-with-pumping operation at La Rance useful:

        In practice, they don’t operate like that much of the time, because they are restricted on how much power they may draw for pumping. Also note that they appear to optimise their output for grid market prices – seeking to increase output for peak demand, while not boosting significantly for periods of lower demand. An expensive CFD provides no such incentives.

      • It doesn't add up... permalink
        May 13, 2018 1:43 pm

        There’s a good discussion in this paper:

        Click to access CIWEM-fifth-reworked-draft-3.4.10.pdf

      • It doesn't add up... permalink
        May 13, 2018 10:04 pm

        Here’s a diagram from TLP themselves:

      • Ben Vorlich permalink
        May 14, 2018 6:05 am

        Thanks for the links. The output without pumping is pretty much what I expected, reminiscent of rectified AC with chopped off areas where the water levels are similar and stationary at high and low water.

        I’ll have to reread as I’m still puzzled by the “gains” at low water. The only way is that by over filling at high water you can continue to run the turbines until after the sea has started to rise, delaying the point at which there is no potential energy available.

        It still depends on there being surplus energy at the point where the tidal plant requires input. To make a profit the operator then requires a shortage of power when the additional head of water is available, not a given. If there isn’t then there is a potential loss (assuming no subsidies) when the additional potential energy has to be used otherwise the rising tide cannot be used because the lagoon is already partly filled. A large river estuary will be different due to the input from the river.

      • It doesn't add up... permalink
        May 14, 2018 5:57 pm

        The gain at low water arises from lowering the level inside the lagoon by pumping it out. The height through which the water has to be pumped is low, so the energy required is also low, so long as it is done close to low water – there is little head. It increases the volume of lagoon to be filled as the tide comes in. By lowering the level inside the lagoon, the time at which the head reaches an adequate level to start generating is brought forward. That allows the extra volume to flow through the turbines, now at a much higher head than was involved in the pumping phase, generating extra energy. How much this is worth can in practice depend on the slope of the shore and the extent of the tide, since at very low tide the lagoon will have a lower area, and therefore a smaller volume for a given amount of lowering of the water level.

  4. dennisambler permalink
    May 14, 2018 4:53 pm

    “There has to be a very real risk that these bigger lagoons will never be built.”

    Make that “hope”…

Comments are closed.

%d bloggers like this: