A Trip Round Swansea Bay

During his recent visit to Swansea David Cameron was quoted as saying: “From the moment I heard about the (Swansea Bay Tidal Lagoon) project I have always been personally very keen on really examining it because it seems to me it has real transformational potential for Swansea — there’s obviously the energy side of it, the clean, green energy, but also the recreational transformation and economic transformation. I am excited by projects that can really transform.”

Tide power is a technology that Energy Matters hasn’t looked into in any detail, so here I will briefly review its potential as an energy source (I ignore its recreational benefits) using Swansea Bay and the “pipeline” of larger tidal lagoon projects that are scheduled to follow it as examples of the approach that tide power in the UK seems destined to follow. Is this approach really transformational? Or is it just another green pipe dream going nowhere?


Tidal Lagoon Swansea Bay (credit Renewable Energy World)

“Tidal Lagoon Swansea Bay plc is developing a 320MW tidal lagoon power project in Swansea Bay. The company aims to begin construction in the first half of 2015 with first power generation in the second half of 2018. The Swansea Bay project is the first of a pipeline of tidal lagoon power projects identified by the parent developer Tidal Lagoon Power plc (TLP), with five subsequent full-scale lagoons at various stages of development which could be operational by 2023. TLP anticipates that the total potential electricity generation from this pipeline could match or exceed 25TWh/year … equivalent to around 8% of UK electricity demand.”

(The five subsequent lagoons at various stages of development are Cardiff, Newport, West Cumbria, Colwyn Bay and Bridgwater Bay.)

What do we know about the Swansea Bay Tidal Lagoon project? One thing is that as a stand-alone project it’s neither efficient nor cost-competitive. With an installed capacity of 320MW and annual generation of 495GWh it has a load factor of only 18%, about half that of offshore wind. A report by Poyry estimates capital costs at £913 million (£2,853/kW), a levelized electricity cost of £150/MWh and a strike price of £168/MWh, much higher than the £92.50/MWh strike price at Hinkley Point. The strike price is, however, projected to decrease to parity with Hinkley for the much larger Lagoon 3, which is scheduled to be operational in 2023.

There are also questions as to whether these estimates are realistic. A recent back-of-the-envelope assessment by Thomas A. A. Atkins concludes among other things that mean power generation is overestimated by 25%. If so the load factor drops to 13.5% and the levelized cost and strike price increase substantially.

A more fundamental issue, however, is dispatchability. With solar and wind the world has an abundance of non-dispatchable renewable energy, but integrating large quantities of it with the grid poses serious problems. What the world needs is a source of dispatchable renewable energy that can be used as baseload or load-following power. Can tide power provide it?

A report entitled Tidal Lagoon Swansea Bay, Project Introduction says that it can. This report contains a map showing high tide times at the prime tide power sites Tidal Lagoon Power plc has identified around the UK coastline, part of which is reproduced below as Figure 1. The caption alongside the map makes the following claim: “Difference in high time tides around the UK creates potential to produce 24-hour base-load electricity from a network of lagoons.”

Figure 1:  Prime UK tidal lagoon sites identified by Tidal Lagoon Power plc

But does it?

To evaluate this claim we must look first at how the Swansea Bay Tidal Lagoon project will generate power. Tides in UK are semidiurnal, meaning that there are two high tides and two low tides a day. Figure 2 shows tides at Swansea for a 24-hour period around March 20th, 2015 (data from Swansea tide times). The 10.4m tide range during this period is about as high as it gets at Swansea:

Figure 2:  Swansea tides, 24-hour period around March 20th, 2015

Figure 3, also reproduced from the Project Introduction report, shows how energy will be generated. On every ebb tide water is stored in the lagoon and released when the head relative to sea level outside the lagoon reaches an optimum level, and on every flood tide the same thing happens in reverse (note how the claim regarding base-load power is repeated in the caption):

Figure 3:  Swansea Bay 48-hour reservoir holding and power production sequence, reproduced from Tidal Lagoon Swansea Bay, Project Introduction.

The result can be considered as square-wave power output with an average of 3½ hours of generation followed by 2½ hours of no generation. This gives four bursts of tidal power a day with nothing in between, as shown diagrammatically in Figure 4:

Figure 4:  Daily generation relative to tidal cycle, Swansea Bay, using March 20th 2015 tide data.

This on-off generation cycle can indeed be smoothed out by combining it with the output from a tidal plant of equal size where the tidal cycle is shifted by three hours relative to Swansea. Some fine tuning would be needed to avoid spikes but this shouldn’t create too many difficulties.

But here’s the problem. Figure 5 is a histogram showing high tide time differences for each of the 66 possible paired combinations of the 12 sites for which tide times are given in Figure 1. There are no two sites where the difference is three hours or even within an hour of three hours. The differences cluster around zero and six hours, meaning that combining output from any two sites or group of sites will tend to accentuate rather than smooth out the intermittent power delivery:

Figure 5: Histogram of differences in high tide times for all 66 possible paired combinations of the 12 potential tidal lagoon sites identified by Tidal Lagoon Power. 

Of most interest is what the combined output from Swansea Bay and the five other proposed tidal lagoon projects will look like with all of them operating at full capacity (30TWh/year). Figure 6 shows the daily generation curve from the six projects. It’s about as far from baseload generation as it’s possible to get. (I estimated megawatts by assuming that the generation from each lagoon is proportional to the area of the lagoon and by factoring the results so that total generation equals the average daily generation (30TWh/365 = 82GWh)):

Figure 6:  Combined daily generation from Swansea Bay and the five other proposed tidal lagoons operating at full capacity, using the March 20th 2015 tide data for Swansea

With a more judicious selection of lagoon sites it would of course be possible to combine output from the sites into something that does resemble baseload generation. But it can’t be done with the sites Tidal Lagoon Power plc has selected. The fact that Tidal Lagoon Power plc haven’t acknowledged this can only charitably be called an oversight.

And now it gets worse.

Figure 7 shows Swansea tides for the entire month of March 2015. The tidal range varies from 10.4m during spring tides to 3.5m during neap tides, but the concomitant variations in generation can’t be smoothed out by combining output from different installations because spring and neap tides are controlled by the orbits of the sun and the moon relative to the Earth and occur once every two weeks everywhere:

Figure 7:  Swansea tides, March 2015

The variations in generation are also proportionately much larger than the variations in tidal range because the energy generated by the turbines is a function of some higher power of the flow rate. How high this higher power is depends on a number of factors specific to the operation, but it will probably be somewhere between the square and the cube of the tide range, so I have used these as upper and lower limiting cases. Applying them to the March 2015 tide ranges for Swansea gives the results shown in Figure 8, which displays the cube and the square of the tide range for each of the 119 ebb-and-flood cycles during March 2015: (Note that the scales are adjusted so that the averages plot at the same place on the Y scale and that the graph plots total generation during each ~6 hour ebb or flood tide cycle. Output during each of these cycles will actually consist of 3 ½ hours of generation followed by 2 ½ hours of no generation, as shown in Figure 4.)

Figure 8:  Generation per tidal cycle analogued by tide range cubed (red) and tide range squared (blue), Swansea Bay using March 2015 tide data

Figure 9 converts the Figure 5 data into MWh so that the total generation during March matches the average monthly generation of 42,000MWh from Swansea Bay (495,000Mwh annual generation times 31 divided by 365). Note again that the generation totals are for 6-hour tidal cycles and do not show daily fluctuations:

Figure 9:  Figure 9 Swansea Bay data converted to megawatts. Tide range cubed = red, tide range squared = blue

As discussed above we can’t smooth out these spring-neap fluctuations by combining output from different sites. We can do it only by storing the power for re-use. So how much storage do we need? To smooth out Swansea Bay generation to the point where it provides constant baseload power we would need about 7.5GWh for the tide range squared case and about 11.7GWh for the tide range cubed case (note that we need only consider the larger peak in the second half of the month). In short, we would need another Dinorwig -sized (9.1GWh) pumped hydro facility. And constructing another Dinorwig for a project that generates only 495 GWh/year is clearly not viable.

But that’s just Swansea Bay. What would combined generation from the full-scale 30GWh/year suite of tidal lagoons look like when the semidiurnal and spring-neap variations are included? Figure 10 gives my best assessment. To make it more readable the Figure covers only the period from March 19th through March 28th, 2015, i.e. the peak and trailing edge of the higher-amplitude spring/neap tide cycle shown in Figure 4, a cycle that repeats itself once every lunar month (29.5 days). Assumptions and procedures were:

  • Output from Swansea Bay and the five lagoons currently under consideration for development (Cardiff, Newport, West Cumbria, Colwyn Bay and Bridgwater Bay) is combined.
  • Spring-neap variations are taken from Figure 9.
  • The semidiurnal variations shown in Figure 6 are factored back in.
  • Megawatt output is adjusted to match total tide power generation over the ten-day period (240 hours/8760 hours times 30TWh annual generation = 0.82TWh).

The demand and wind power generation curves are from Gridwatch:

Figure 10:  Combined generation from Swansea Bay and the five other proposed tidal lagoons operating at full capacity, using Swansea Bay tide data for March 19th through 28th, 2015. Wind generation is added for comparison purposes.

About all that can be said in favor of the tide power generation curve is that it’s predictable. As a source of baseload energy it makes wind look good.

Finally, how much storage would be needed to convert the tide power generated over this period into baseload generation so that it can compete head-to-head with nuclear, as some of its backers claim it can? It comes out to approximately 500GWh, over fifteen times current UK pumped hydro capacity, or if you like five million 100kWh utility-sized Tesla storage batteries. And even with this much storage tide power supplies only about 8% of total UK electricity demand.

We are now in a position to answer the question posed at the beginning of this post; is tide power really transformational or is it just another green pipe dream going nowhere? Clearly the latter, although this of course doesn’t mean that the politicians aren’t going to pursue it.

This entry was posted in Energy and tagged , , . Bookmark the permalink.

55 Responses to A Trip Round Swansea Bay

  1. Willem Post says:


    Most politicians will not be able, or too busy, to understand it. Whether economical or not, they will vote for it anyway, because it is renewable.


  2. Willem Post says:


    The CF of the one in France is 0.257

    240 MW of wind turbines, costing about $600 million, would produce about the same energy, but those turbines would have a life of about 20 – 25 years, whereas the tidal plant could last at least 100 years.

    • Willem Post says:


      Much of future renewable energy, such as wind and solar, has practically no rotational inertia, which provides stability to the grid.

      Just providing balancing energy is a necessary, but not a sufficient condition for grid stability.

    • oldfossil says:

      CF? The acronym is used nowhere in Roger Andrew’s article. I might guess at Capacity Factor with Cost Factor second, but these are only guesses and leave me with no better understanding than before I read your comment.

    • How long are the Swansea Bay turbines going to last?

      • Euan Mearns says:

        The appropriate question is how long will the concrete wall last?

      • Willem Post says:


        The ones in France are exposed to brackish water. The impellers likely are made of Navy-grade bronze. O&M of such plants likely is very high.

        Right now, low-cost fossil energy would be used to build such plants.

        In the future, such plants would need to be built with materials produced with renewable energy at 3 – 4 times/kWh.

        A brave new world!

        • It doesn't add up... says:

          There is a presentation on the experience of La Rance available here:


          It reveals that they dealt with corrosion extremely well through anode protection. The maintenance burden over 40+ years is described, and seems relatively modest. Not quantified, but mentioned in passing is that silting has reduced the effective tidal range.

          The operational regime and typical power curves over the tidal cycle are also discussed

      • louis says:

        Also maintenance and repair schedules are not predictable. An interruption to supply that would take engineers a few hours to rectify on land can suddenly take several days if the support vessels are unable to put to sea.

  3. Leo Smth says:

    One of the most tricky concepts that needs to be absorbed, is the concept that a unit of electricity that you dont need right now, is a very expensive unit if you have to pay for it.

    Or to look at it another way, just because generator X can generate so many TWh per year, doesn’t mean that the impact of its generation pattern is cost free elsewhere, or indeed always positive.

    It is customary to ignore externalised costs when dealing with renewable energy and play them up when dealing with non renewable energy, so that the costs of transmissions lines and co-operation plant are ignored when talking about levelised costs of renewables, whereas the lack of taxation for CO2 emitters is portrayed as a giant subsidy…

    In essence the externalised costs of intermittent renewables fall into several broad categories, each of which is inherent to the generating technology itself, and is therefore unavoidable. Solar wind and tidal all suffer from these effects.

    (a) Low energy density leads to large structures. That means high environmental impact of one sort or another and a lot of scarce land (or sea) used up generating not very much. The reason the land area is needed is simply the low energy density of the source itself – wind wave tide or solar.

    (b) Low capacity factors lead to inefficient use of resources. A low capacity factor means that most of the time your generator isn’t generating nearly as much as it can do. However, it has to be sized for the worst (or best?) case when it is generating that much. So wind turbines need to have gearboxes and structures capable of standing the peak output, even though this is seldom realised in practice. Likewise transmission lines connecting such devices to where the demand is (and in the case of Scotland, a lot of the demand is in England) must be sized for peak flows, not average.

    What we see instantly is that capital cost not just of the generator structure, but its connectivity, relates to peak output capacity, whereas income relates to average output… What that means is that the lower the capacity factor the more expensive the power so generated.

    (c) Low capacity factors necessitate co-operation with storage or a dispatchable technology, whose fixed costs should be borne by the renewable operators when the renewable plant is in operation and the dispatchable plant is idle. Of course those costs are not. They are thrown back in the teeth of the (usually gas) operators who then have to raise prices, leading to the call that ‘renewables are now competitive’.

    Is tidal in some way different? No. Tidal output is still low energy density, still intermittent, still low capacity factor, still expensive and still – as this excellent analysis shows – not sufficiently distributed in its output potential to in anyway reduce the need for dispatch elsewhere on the grid.

    Which is why it probably will get built at enormous cost to be yet another white elephant and environmental disaster…

    • clivebest says:

      The only positive point for tidal power is that the output is 100% predictable both in time and in magnitude. In that sense it is the only ‘reliable’ new renewable source, excluding hydro.

      Why did Cardiff Bay not include provision for tidal power ?

    • Another externalized cost applicable to tide power is the added wear and tear on load-following gas plants, which are going to have to scramble to keep up with the four daily ~10MW spikes during high springs (see Figure 10).

      • oldfossil says:

        I have read elsewhere, perhaps at Paul Homewood’s page, that gas receives more subsidies than does wind. The gas subsidies are paid (wait for it) when the turbines are idling as spinning reserve for wind. But the subsidies are charged nonetheless not to wind but to gas. This isn’t logic, it’s dumfukkery.

        • Leo Smth says:

          Its political lobbying and leverage.

          Renewable energy passes its costs on to everyone but itself – to the rest of the generating community, to the grid, to the local communities that bears the brunt of the inconvenience, to the local environment and ultimately to society as a whole.

          The justification is that it isn’t emitting CO2, except of course in order to keep up with its vagaries, everyone else emits more.

          It doesn’t matter how predictable it is – the daily demand curve of the UK is entirely predictable – fluctuations in demand or supply always result in an excess of fuel being burnt to balance the two.

          When wind and solar variability were small compared with demand fluctuation, there was some justification in ignoring this effect: Now they are capable on a sunny windy day of meeting maybe half UK demand, they are absolutely not to be ignored

          • PhilH says:

            > The justification is that it isn’t emitting CO2, except of course in order to keep up with its vagaries, everyone else emits more.

            Euan, in his excellent 2013 post
            that looked at renewables penetration in GB up to 2012, concluded that “There is absolutely no evidence from these numbers that the efficiency of large coal and CCGT plant is being impaired through cycling to balance the increasing load from wind and solar.”

            My cruder analysis of the data for 2013 & 2014 show there is still no measurable effect from this.

  4. A C Osborn says:

    I live in Swansea and the lagoon itself has some attraction if it provides for fishing etc.
    It will provide an extension of the current Cycle Paths.
    But as far as Power Generation is concerned it is yet another variable that the grid will have to balance and at massively inflated costs. The Tax Payers will pay twice for this one.
    More Green Madness.

  5. Peter Shaw says:

    Roger – another quotable post.
    Leo – insightful comment. I think you’re asking “Is there such a thing as too much electricity?”; which is an excellent problem-solving question.

    However, you both use a widespread framing of renewables (unreliability, intermittency, etc) which may not be best.
    I suggest that renewables are *wild* energy, by contrast with “domesticated” FF energy. Exploitation of a domesticated resource is usually high (80-90%), wheras wild resource use doesn’t exceed 20% and is typically ~10%. The current performance of renewables should be seen in that light, which suggests there is little development space in them.
    Wild resources are where (and when) you find them (migrating game, target-of-opportunity game, etc). We do (of course) know how to exploit these: We revert to the hunter-gatherer modes of our ancestors. Can a modern state function thus? It’ll be an interesting experiment.

    • oldfossil says:

      Wild vs tame: a very illuminating way of looking at it. You should mention your concept to Dr Judith Curry who, I think, is struggling a bit to communicate her definition of climate as a “wicked” problem. It would help her.

    • Leo Smth says:

      I prefer the following analogy.

      You are a busy company that does things. Let’s say you process and input data on computers. Most of your costs are staff costs plus the cost of the buildings and the computer equipment and the electricity.

      One day a politician phones you up and says ‘look, there are loads of people out there who will actually work for nothing, do you want to employ them on that basis?’

      ‘Sounds too good to be true, what’s the catch?’

      ‘Well sometimes they won’t turn up, and they wont actually do much when they do’

      ‘Oh… but that’s OK – if I dont need them they can stay at home ‘

      ‘Er no, the condition of their being government subsidised is that when they do turn up you have to use them’

      ‘Well ok, but what happens to my permanent staff?’

      ‘Oh put them on zero hours contracts- so you can tell them not to turn up instead’.

      ‘well that’s all very well, but they won’t like it, and they will charge more anyway – they still need to make a living you know: I am not sure these free workers are going to save me any money. Hang on you said that I have to take them on when they do turn up for work, and they aren’t very productive?’

      ‘That’s right’

      ‘But that means I will need more office space and more computers all lying idle just in case they DO turn up.’

      ‘That’s right’

      ‘And in fact there could e a time when I need 10 times as much office space and computers, and I have no permanent staff, just very expensive zero hours contractors and there freeloading interns?’

      ‘That right’

      ‘Matey, this isn’t profit, this is cost all the way. I thought you said that their labour was free?’

      ‘Well it is free – you just have to pay for all the other stuff you need to take advantage of the freeness’.

      ‘Do you mind if I don’t?’

      ‘Er, yes, in fact we mind very much. In fact you are going to be forced to take them on, because that’s part of our ‘full employment initiative”

      ‘I think I will move to another country’

      • Good one, Leo 🙂 🙂

      • Ed says:

        Simple solution Leo. Lay off all your permanent staff. Give these other people who want to work for free your computers and let them work from home when they can. You can get rid of all office space. Your only costs are then just the computers. No point moving to another country because they are in the same position as you. Tough.

        The main difference between your world view and mine is: you think there is a CHOICE between renewable energy and finite fossil energy; while I think there is NO choice because fossil energy will eventually run out because …. you’ve guessed it …. it is finite.

        • Euan Mearns says:

          Leo is strongly pro-nuclear.

          • Ed says:

            Thanks Euan. I misrepresented Leo’s world view. Leo’s world view is that there is choice between renewable, fossil and nuclear energy. Being pro-nuclear is a valid position to take. Sorry Leo.

            Quite rightfully a lot of very difficult questions are being asked about renewable energy but nuclear has it’s own set of difficult questions of it’s own. An interesting debate.

  6. Willem Post says:


    “Is there such a thing as too much electricity?”; which is an excellent problem-solving question.

    Rich Germany has found the answer by exporting its excess energy, during windy/sunny periods, at near-zero wholesale prices, after having paid of it about 20 eurocent/kWh, per ENERGIEWENDE data in this article.

    Germany is telling poorer EU countries: “Be like us” and tells Brussels: “No backsliding”.


  7. Is there such a thing as too much electricity?

    As things stand the only feasible way of integrating large amounts of renewable energy with the grid is to grossly overdesign the systems so that they can fill demand during low wind, low sun, low tidal flow etc periods (and to keep enough conventional backup capacity to fill demand if they can’t). But the massive amounts of surplus electricity generated when the wind does blow, the sun does shine and the tide does flow will all have to be wasted.

    Which poses an interesting question. Wind, solar and tide power all have zero fuel cost. So which gets curtailed first?

    • Willem Post says:


      This reminds me of a baby learning to walk and a parent constantly guiding it so it will not fall.

      There is a small difference, the RE baby will ALWAYS need that parent.

    • Leo Smth says:

      So which gets curtailed first?

      Gas, the lowest emitting and most efficient of them all.

      Which is why its gas stations that are closing.

      The net effect of on grid renewables has been to push the best generation kit off the grid.

      That’s why the legal measures to make coal so unprofitable have been needed.

      Germany runs on coal and nuclear really. They didn’t go so draconian on coal as Miliband did when he was climate and energy wallah.

      Now we are going to have capacity payments… another subsidy.

      The way it works is this:

      Coal is made unprofitable either by restricting running hours or insisting on carbon capture.

      Nuclear is made impossibly expensive by regulatory ratcheting.

      Renewables drive a cart and horses through all regulatory and planning processes because they are deemed ‘good’.

      Renewables are mandated to be used instead of anything else when they are available.

      Gas is simply not profitable anymore because its more expensive than coal, and its running hours are slashed by making way for renewables.

      The net effect is that all that is being built is renewables, and they cant do the job alone, so in a few years there will be complete disaster.

  8. Euan Mearns says:

    Buy shares in the concrete companies evidently loved by Greens. The most interesting and relevant stat here is the 120 year design life.

    Predictability of the tides has great value when combined with storage – so long as cost is no obstruction.

    0.4 TWh / year compares with UK 323 TWh / year. So we need 808 of these and job done. @ £913 million a pop no problem that £738 billion can’t solve + the cost of storage. At risk of sounding green tinged, £1 trillion capex for 120 years of electricity ….

    UK GDP is about $2.5 billion = £1.3 trillion

    • Leo Smth says:

      Compare the cost of 36GW of nuclear, even at £5bn/GW


      • Ed says:

        Are you including ALL your costs there, Leo? After all, tidal can be just left in place after it’s useful lifespan. Also how do you price “risk” ?

  9. roger in florida says:

    “Is there such a thing as too much electricity”?
    Well, what do they do with it?
    Easy, they burn it.

    • Leo Smth says:

      one clip is enough to cast doubt on that – 67.5% lost on transmission? No way!

      • roger in florida says:

        Clearly the statement “67% is lost in transmission” is incorrect (EPRI says an average of 6% is lost in transmission, that sounds a little low to me but they should know). In the comments below the video Andrew Dodson explains where his 67% losses come from, it was not well explained in the actual presentation.
        There is a lot of useful information in the presentation, what it tells me is that in the US, as is probably the case in the UK, energy decisions, crucial to the economic well being of our society, are being made by people who are, to put it charitably, not qualified.
        There are other fascinating presentations from the Thorium Alliance group available on youtube.
        Are LFTRs “the answer”, I don’t know but they clearly are better candidates for research than fusion.

        • Leo Smith says:

          Right now any sort of fission is the answer.

          Its reliable, its safe and it buys us about 1000 years.

          Which is time enough to look at what comes after.

          Heck, in 1,000 years we may have finally cracked fusion.

          in many senses the more energy per capita we have, the ‘richer’ we are: there probably isn’t enough ‘renewable’ energy falling in the planet to give everyone a Western life style, and there probably isn’t enough surplus to get us off planet to collect it elsewhere.

          Fossil fuels is Gaia’s gift to humanity to enable it to build the first nuclear power stations..:-)

  10. Pingback: A Trip Round Swansea Bay | The Conscience of A Libertarian

  11. Tony says:

    I’ve been watching this project from afar with some interest. In the current environment the major justification for renewable projects seems to be reducing CO2 output. So, if a project doesn’t make economic sense (as this one clearly doesn’t’) and also ignoring any recreational or other non commercial benefits if a project doesn’t result in a net reduction of CO2 emissions is there any remaining reason to proceed? Given the huge upfront effort (read energy) needed for this project, most of which will be fossil fuel sourced, does this project result in a net CO2 reduction? Anyone care to take that on? As a corollary, the huge CO2 release will be mostly upfront whereas the reduction will trickle in over time. What are we to make of that?

    • Ed says:

      NOTHING will reduce our CO2 output because all our fossil fuels will be burnt eventually. It is pie in the sky to believe any of it (that economically extractible) will be left in the ground unburnt. The real question is: does the EROEI of tidal make it worth while?

      Even if the answer is as low as 2, it is still worth while. Every unit of fossil energy we put in will give us two back. Admittedly in a less useful (less dense) form. It buys us time to prepare and adjust.

    • It doesn't add up... says:

      There are some numbers and a lot of waffle from the Swansea project here:


      I’ve not had time to review them for credibility. Perhaps someone should?

      • It doesn't add up... says:

        P.S. I did note that they calculate the wall’s contribution to emissions by reference to a 100m segment of wall. As the wall varies in height between 5 and 20m it perhaps conceals fudge.

        • Tony says:

          Thanks for the reference. They frame the question much better than they deal with it. Lots of reliance on assumptions etc and assuming a 120 year life to do their calculations over. Does anyone else see a problem with a 120 year life of a sea wall made largely of plastic bags stuffed with silt. They say they haven’t finalized their design but the alternative design uses much more offsite material and would be something like 5 times the CO2 released according to their comparison. Still unless they are completely dreaming or misleading it seems like a reasonable reduction of CO2 over project life. What isn’t addressed nor likely to be is the impact on releasing all the CO2 upfront during construction and reducing over the 120 years of operation.

          • It doesn't add up... says:

            Indeed: there is likely to be a very substantial deficit for many years – and should we happen to succeed in “decarbonising” power supply (for example via cheap nuclear technology in say 20-50 years’ time), the notional savings in the future are not merely undiscounted, but will never be realised.

  12. It doesn't add up... says:

    Prof Mackay’s technical discussion of marine energy is here:


    On pages 318ff he discusses ways of smoothing out power variations in tidal pools including using pumping with neap tides (in use at La Rance), and twin pools, in effect also potentially turning it into pumped storage.

    For more on the practical experience at La Rance, see my post above.

    • Roger Andrews says:

      IDAU: Thanks for the links to La Rance and David Mackay’s book. I think Ed linked to Mackay as well.

      If might be possible to smooth out the daily tide power fluctuations at Swansea Bay into something resembling baseload generation by pumping water back and forth, but not to flatten out spring-neap variations. As discussed in the text we would need ~10GWh of storage capacity to do that, and the Swansea Bay lagoon has a capacity of only about 1GWh.

      • It doesn't add up... says:

        I suggest you consider Mackay’s more detailed paper here:


        The claim is not to completely flatten the differences via pumping, but to reduce them significantly (his example suggests reducing variation from a factor of 4 to a factor of 2 without building extra wall height beyond that required for a Spring tide). The key idea is that the moon does most of the heavy lifting, with the pumps simply shifting the water almost sideways into place for gravity to do its work.

        Without having attempted any more detailed calculation, I suppose we may assume that his more elaborate schemes are not economic, in that they are missing from the Welsh proposals: the proponents are hardly likely to be unaware of them, given his position. However, it might be entertaining to show this.

        Incidentally, I think his assumptions about the costs of walls are somewhat off. I watched the construction of the breakwaters for the Brighton Marina in the 1970s, with cylindrical caissons being built and filled (you can see the same technique was used for the coffer dams during the construction of La Rance), and then protected from wave erosion through building sloping buttresses under water (again, the construction at La Rance). These have a triangular cross section, and therefore scale with the square of wall height.

        • What David Mackay is saying is that the highest and best use of a tide power plant is as a pumped hydro facility to smooth out intermittent wind generation, not as a power generation facility in its own right. Building pumped hydro lagoons in the sea would certainly be a more palatable approach than flooding scenic glens in the Highlands, particularly when flows between the reservoirs can be increased by making use of natural tidal ranges. I might look into this in more detail when I get time.

          • It doesn't add up... says:

            I’ve recently found an old thesis here:


            It goes into some depth on the real engineering issues of tidal power (it was done at an earlier time when the perennial Severn Barrage was under consideration) and contains this conclusion about different modes of operation:

            At the time of these Severn Barrage studies (1978/79) there was a belief that a tidal power barrage should be developed to extract energy at lowest unit cost and the ebb-generation mode of operation was the way to do this. Comparative studies for the Severn Barrage Committee of the alternative forms of development, to which the writer contributed, confirmed this to be so, although it is true to say that more in-depth studies were carried out for the single-effect ebb-generation mode of operation than for double-effect operation or any of the double-basin alternatives proposed which could provide a degree of firm power.

  13. Jamie says:

    “The result can be considered as square-wave power output with an average of 3½ hours of generation followed by 2½ hours of no generation.”

    Can it? Surely if the lagoon is generating for 14 hours per day (58% of the time) but the overall load factor is 18% then there must be a fair degree of ramping up and down of generation. The only alternative would be that they run at part load all the time which would be odd.

    • It doesn't add up... says:

      Indeed: if you look at the power curves at La Rance you will see they are quite peaky. I suspect these are somewhat stylised, since tide heights are approximately sinusoidal, and instantaneous power proportional to the square of the head and the efficiency factor of the turbine, which will vary as in some inverse manner with the head.

    • Figure 3 shows an on-again-off-again generation pattern that looks very much like a square wave. There’s also a difference between “can be considered as” and “is”.

  14. Jamie says:

    Also wouldn’t there be scope for sacrificing some peak power in order to advance or retard the time of peak generation?

    • It doesn't add up... says:

      The first optimisation is simply to extract maximum energy given the basic physics and engineering realities. A variant is to extract maximum value by weighting output against marginal grid price. This may get complex as the relative phase of tides and rush hours keeps shifting, or if peak power prices become particularly spiky. Of course, guaranteeing a fixed price for output via a CFD means no economic optimisation is needed. However, there is instead a variation in the effective subsidy.

      If you want to play, it’s probably not too difficult to set up a spreadsheet to be optimised via e.g. simulated annealing: discontinuity implied by on/off switching makes gradient techniques unviable.

  15. A C Osborn says:

    Why bother, every time you add some pumping or any other type of conversion you affect both efficiency and cost.
    Just don’t build the things in the first place, build instead some far cheaper 24/7/52 Base Load in the first place.
    How about thinking of and putting the CUSTOMER first?

Comments are closed.