What If the UK Had a Second Sun?
We built a model of the proposed Morocco -> UK 'Xlinks' project to test the power of abundance and connection.
I read once that, in theory, it would only take about 1% of the Sahara desert to power the whole world with solar. Obviously that’s ignoring major constraints on transmission and storage, but it’s true that we would only need to harness a tiny fraction of the sun’s beams to power the global economy. The sun in the Saharan desert offers consistent, high irradiance sunlight throughout the entire year - the tricky part is capturing, storing, and transporting that power. Low transmission efficiency and high costs have kept Europe from tapping into Saharan solar energy, but as transmission lines become more efficient and affordable, for the first time, projects like Xlinks are nearing reality, offering potential shifts in climate action.
Xlinks is a £22 billion proposed project in the United Kingdom (UK) that is essentially an extension of the UK grid into southern Morocco, where they hope to build 11.5 gigawatts of solar and onshore wind, along with battery storage, and send it 4,000 kilometers through a transmission line into Southern England:
Besides the massive size of the project, the interesting idea behind it is it could solve a big piece of the renewable intermittency issue for the UK. The wind doesn’t always blow in the North Sea, and we all know the UK gets little sun - but tapping into Saharan solar and wind would offer the UK renewable energy during more hours of the day.
This map shows solar energy potential around the world, and you can see the stark difference between Morocco and the UK:
The project would be a first of its kind, and a signal that grids will start to become more interconnected into the future.
Xlinks is still trying to secure financing for the project - but once built, they say it will provide enough energy to cover 8% of the UK’s energy needs. So, a few colleagues and I built a simplified model to test this claim. We were also interested in how the Xlinks project will change the UK’s emissions trajectories, in absolute terms and relative to the other levers that policymakers have at their disposal.
Specifically, we want to know:
Does a market exist in the UK for the electricity that Xlinks would generate?
How would UK emissions change after the project?
Could the project reduce the carbon tax needed for the UK to reach their emissions targets?
Today’s post is a summarized version of the paper we wrote about our model. I was lucky to work on this with the very talented Natalie Ziklova, Euan Gilchrist, and Kai Kempf, and you can download the full, more academic version of our paper here:
If you're mainly here for the big-picture takeaways, feel free to skim through the modeling detail and jump to the results below. But if you’re curious about how we built the model and the assumptions we made, read on and download the full paper above. Also, if you’d like an entry-level explanation on how electricity markets work before diving into this post, my post from last week would be a good place to start.
The Model
Background and the UK grid
We built a simplified model of the UK grid, electricity market, and the Xlinks project that allows us to tweak variables and test hypotheses. We did this using the coding language Julia, which I won’t get too much into in this post.
To build our model of the UK grid, we used data on wholesale hourly prices, demand, and generation for our ‘fixed production’ technologies (nuclear, hydro and biomass) provided by Elexon. For UK solar, onshore wind, and offshore wind capacity factors, we used hourly averages produced by the European Commission’s Joint Research Centre. To finalize our model inputs, we incorporated UK government data on marginal costs for each technology.
The Xlinks assets
The price used for the Xlinks-generated electricity in the model is based on the most recent negotiated prices. Xlinks approached the UK government to negotiate a ‘contract for difference’ with a strike price of 70 pounds per megawatt hour, which is represented in our model. Contracts for difference are the UK government’s primary tool for decarbonizing their grid and helps companies like Xlinks de-risk their projects.
We simplified the scope of the project in response to data limitations. We only considered Xlinks’ planned solar assets (7GW) and did not model the planned wind assets (4.5GW) or battery storage (22.5GWh capacity). We included realistic constraints on the solar and transmission technologies including a 13% transmission loss factor to account for efficiency losses in the 4,000km interconnection to the UK, which is standard.
Model & Assumptions
To solve the model, we utilize market clearing functions based on cost minimization, where the lowest marginal cost technologies are utilized first. Solving the model allows for a comparison of price, welfare and emissions levels, and also provides a description of the optimal composition of the UK electricity production mix.
First, we solved the model without the Xlinks project included to see how our ‘do nothing’ scenario looks compared to the real world. Figure 2 shows the similarities between the real-world technology generation mix from 2023/2024 and our model-generated generation mix when minimizing costs. Our modeled share of wind generation is significantly lower than in the actual generation mix, which may be attributed to technological advancements in the efficiency of renewables over the past decade not captured in our model.
We then use our model to determine how UK electricity prices, emissions, and welfare would change if the Xlinks project were built and implemented into the UK grid. First, we consider the current UK energy generation assets (the ‘do-nothing’ scenario) and then incorporate the additional assets provided by the Xlinks project (the ‘intervention’ scenario).
For more on the assumptions and data used in the model, feel free to check out the full paper.
Results
Prices, welfare, and emissions
We find that the overall effect of adding the Xlinks solar to the UK grid to be significant, especially on emission reduction. Relative to the “do-nothing” scenario, the Xlinks project:
Decreases the price of electricity by 1.2%,
Increases welfare by 0.05%, and
Decreases emissions by 4.5%.
In contrast to Xlinks’ ambition to meet 8% of the UK’s energy needs, we estimate that Xlinks would meet 2.5% of energy needs with the solar alone. This 2.5% understates the entire proposed project given we did not model the planned wind and battery assets. With that in mind, Xlinks may actually get close to the stated 8% with all of the assets built.
Figure 3 illustrates the price impact of adding Xlinks solar relative to the ’do nothing’ scenario. At a marginal price of £70/MWh, the only technology with a higher marginal cost is gas. This means supply from Xlinks is unable to compete with any domestic renewable energies, but can replace gas when it’s present in the production mix. Notably, gas only has a higher marginal cost than the Xlinks solar due to the existence of a carbon tax in the UK.
Figure 4 illustrates how Xlinks solar is available for 51% of the representative hours. For these hours where Xlinks supply exists, energy from Xlinks is fully utilized 98% of the time. This illustrates the effectiveness of the Xlinks supply at undercutting gas, which is the UK’s price-setting technology 97.5% of the time.
To assess how sensitive our outcomes are to our assumptions, we conducted sensitivity tests on additional interconnector capacity to the Moroccan solar and varying Xlinks solar prices and carbon taxes. We find, with no surprise, that adding interconnector capacity between the Moroccan solar and the UK grid increases solar delivered to the grid, but the cost of an additional line may not outweigh adding batteries to the project. We also find that lowering the contract for difference strike price from £70 to £40 doesn’t change prices, welfare, or emissions much. And, quite interestingly, we find that increasing the carbon price from £32 to £97 has negligible impacts due to the inelastic nature of electricity demand.
Policy comparison: Xlinks vs. carbon taxes
Now to answer our final research question: What equivalent increase in carbon taxes would be required to achieve the same emissions reductions as Xlinks?
As noted before, we found that the Xlinks project (with just solar) would decrease emissions by 4.5%. To find the required increase in carbon taxes that would do the same, we rerun the model without Xlinks and increase the carbon tax until we see a 4.5% reduction in emissions.
We found that it would take a 64% percent increase in carbon taxes (to £69/MTCO2) to achieve this same reduction in emissions. This increase in carbon prices would result in a 10.5% increase in electricity prices, compared to the 1.2% decrease in prices generated by the introduction of the Xlinks assets.
As established earlier, we find emissions are only weakly sensitive to changes in carbon taxes, meaning that using carbon taxes to achieve a small reduction in emissions is extremely difficult. This is due to gas being both the only emission-generating technology in the model and having the highest marginal cost - it is only used as a technology of last resort.
Therefore, reducing gas supply requires encouraging consumers to consume less energy at hours when gas is in the supply mix. Given that energy lacks short-run substitutes, it would take extremely large increases in prices to generate this change in consumer behavior. Whilst carbon taxes are often viewed as the “first best” policy to tackle emissions, this finding shows that renewable energy abundance is likely to be a far more politically feasible solution to decrease emissions.
Final thoughts and takeaways
As discussed, there are limitations to our model: we model the UK grid as static, taking historic data as representative of the UK grid into the future; and we do not consider wind or battery assets. Despite these limitations, our findings show that the Xlinks project, if implemented, would:
Reduce emissions (-4.5%) and energy prices (-1.2%).
Allow for relatively cheap reductions in emissions compared to an increase in carbon taxes.
Economists, including myself, often call for carbon taxes as the first-best policy for decarbonization, but that’s unlikely to happen in today’s political climate. Perhaps the abundance agenda is more attainable.
Unfortunately, it still doesn’t mean the Sahara desert will ever power the entire world, and the UK won’t ever have a literal second sun, but it means Northern Africa could play a role in decarbonizing Europe and lowering electricity prices. Most continents have areas like the Sahara nearby which offer high-irradiance sunlight year round - the American Southwest, Northern Chile, Western China, Australia, etc. - we just need to build longer transmission lines.