How Much Would Nuclear Power in Ireland Really Cost?

There are many good reasons not to build nuclear but cost isn't one of them

In a previous blog, I wrote about the tremendous changes that would need to be enacted immediately across all sectors of society if countries were serious about greatly reducing fossil fuel dependance in order to avoid economic and societal breakdown. This dependence works in two directions - the downstream problem of too much carbon in the atmosphere leading to climate change, and the upstream problem of the ever-decreasing availability of the fossil fuels needed to build a replacement system (the EU is projected to lose 12% of its oil supply by 2030 and 50% by 2050 in the best-case scenario).

In this blog I will compare two non-carbon emitting replacement grids that do not rely on imports. Designing self-sufficient grids is essential, as most European countries will rely heavily on wind and solar by 2050. During continent-wide periods of windless, sunless days - the now infamous dunkelflaute - there would be no external energy to import.

The first of the grids is one dominated by renewables with long-term storage, and the second is a grid made up of existing renewables plus nuclear power.

Starting in 20022, here’s a breakdown of Ireland’s electricity generation by type:

The non-renewable sources - natural gas, coal, oil, non-renewable waste, peat - need to be replaced, along with imports, for a total of 20.08TWh.

The first grid will be based heavily on offshore wind. Since wind is variable but our lifestyles are constant, we’ll need to store excess energy from the wind when production exceeds demand, and release it when demand exceeds production. Currently, the only realistic option for long term storage is pumped hydro.

Unfortunately, Ireland isn’t blessed with many potential pumped hydro spots due to our low topology - the greater the elevation difference between the lower and upper lakes in a pumped hydro systems, the great the power output - but for the purposes of this exercise, we’ll assume that there are enough suitable locations to store all our energy needs.

First let's take stock:

Capacity Needed

Generating 20.8TWh of electricity in a year would require adding 2.37GW of capacity (20.08 * 1000 to convert to GWh and divide by the number of hours in the year) if the generator worked all the time at its maximum.

But the load factor for offshore wind is about 50% i.e on average 50% of the time offshore wind is generating electricity at the maximum power a turbine can produce.

Therefore, the amount of installed capacity to generate 20.08 TWh of energy would be 4.75GW (dividing the previous result by 50%). With nuclear, the average load factor last year was 81.5%, meaning that we would need to install 2.91GW of capacity.

Since this is a first order calculation, we’ll be generous and assume that during periods of low demand and excess production, all that excess energy can be stored without any losses. We’ll also assume that it is the new offshore wind farms that power the pumped hydro stations during periods of excess production.

During periods of low demand in Ireland, demand drops to about 2.5GW. On a reasonably windy day, offshore wind can be expected to produce at about 70% of its capacity. This means 1.575GW of excess power will be generated (4.75GW * 70% - 2.5GW). Therefore, 1.575GW of pumped hydro storage will be needed.

There is an ongoing debate about how long energy in a renewables dominated grid needs to be stored for, with some arguing just one day, and others arguing up to 6 weeks. That would determine the volume of water that needs to be stored, but for this exercise, I will park it and just deal with the pumped hydro capacity.

Unfortunately, the second law of thermodynamics rears its head, meaning that, as with any form of energy storage, losses will occur (the 2nd law states that when going from one form of energy to another, losses will occur - which is why a perpetual motion machine is impossible!). When going from the wind turbine to the pumping station, about 4% of electricity is lost. Pumping the water up to the top lake incurs losses of about 20%. Then there are losses of about 15% when converting the water in the top lake back to electricity for overall losses of 39%. We’ll be generous and assume optimisations can be made and the overall losses come to 35%. So an extra 35% of electricity will need to be produced to make up for these losses, meaning installing 35% extra offshore capacity. This means we’ll need to install a total of 6.18GW of offshore wind capacity (4.75GW plus 35%).

The good news is pumped hydro stations last a long time, up to 100 years.

The bad news is wind turbines do not. Onshore turbines last between 20 to 30 years, while offshore turbines last 20 to 25 years. We’ll go with the more optimistic 25 years for offshore. On the other hand, nuclear reactors last between 60 and 80 years, and potentially longer. That means the wind turbines will need to be replaced at least 3 times in the lifetime of a nuclear reactor.

Offshore Cost

For offshore wind turbines, installation costs are about €2600/kW. Replacing centralized power stations, such as natural gas, and coal stations, with decentralised stations requires a lot more grid infrastructure (it costs much more to install 200 light cables than it does to install 10 heavy ones). That means more pylons, more holes dug in the ground, and a lot more cement. We’ll use an estimate and say that for every 1 euro invested in building a wind farm, 50 cents is needed to upgrade the grid infrastructure (€1300/kW). We’ll assume this only has to be done once over a century (unlikely of course, there will need to be constant maintenance of the grid for both systems we are evaluating).

In OECD countries, the cost to build a pumped hydro project is between €2,000 and €3,500. We’ll take the midpoint of €2,750. For the purposes of this exercise, I won't include any of the extra costs of building pumped hydro - flooding of valleys, legal issues etc.

Nuclear Cost

Capital costs for nuclear power vary wildly from €2,600/KW in South Korea to €7,100 in Finland to €8,600 in France, all the way to €12,000 in Hinkley Point in the UK where the second most expensive nuclear project of all time is still under construction. We’ll be ultra conservative and use the higher cost of €12,000.

Turning to fuel, 1 kg of natural uranium produces 45,000kWh at a cost of €103 per kg. So to produce 20.8TWh of electricity per year, the total cost per year will be €47,608,888.89. Since we are using the same cost per KW to rebuild wind turbines every 25 years over the century, we’ll also assume the price for uranium won’t change over the century (uranium from seawater and thorium reactors likely means there will be no shortage of nuclear fuel). The gives a total cost of fuel of €4,760,888,888 over 100 years.

The Final Result

Let’s have a look at the results when comparing both systems over a century:

So over a century, even if Ireland built a nuclear reactor as exorbitantly expensive as Hinkley Point in the UK, offshore wind would still end up 1.21 times more expensive than nuclear. So much for the “nuclear is too expensive” argument!

Inertia

One other important factor I have pointed to in other blogs is the problem of inertia. The power coming out of our sockets in Europe has a frequency of 50hz. This means the electrons change direction 50 times per second. Why? Because in a generator, the rotor's speed, combined with the number of magnetic poles, results in a magnetic field oscillating 50 times per second. All our electronics depend on that frequency being 50, or staying really close to 50. If there is suddenly extra load on the grid, the frequency will reduce slightly. With traditional grids, the large spinning mass of the turbines from coal or gas power plants absorb this extra load, buying a few seconds to increase fuel to stabilise the grid back to 50. This is called inertia.

With grids dominated by renewables, the problem of a lack of inertia arises. Wind turbines and solar panels so not have heavy masses directly connected to the grid. Power from a turbine going through an inverter before joining the grid. This leads to a frequency that destabilises the more renewables that are added to it. A crucial element for 100% renewables grid is hydro - as this has a large spinning mass directly connected to the grid. In Ireland however, we are not blessed with with locations naturally suited to hydro. There are plans to add “artificial” inertia in the form of synchronous condensers. These are spinning masses that use existing electricity to get up to the correct rotation speed, and stabilise voltage and frequency when extra demand occurs. It is unclear though if synchronous condensers can work at large scale across the European grid. This is another benefit of nuclear as it provides the crucial inertia required.

Conclusion

Ireland is one of the few places in the world that does not have a decarbonisation plan that involves hydro or nuclear. This appears to be due to an aversion to nuclear due to previous safety issues, a misunderstanding at the decision-making level of the necessity of dispatchable power for base load and grid stabilisation, and the perceived cost of nuclear. However, this analysis shows that, over the long term, nuclear proves to be cheaper than renewables. So while there are some reasonable arguments against building nuclear in Ireland - cost isn't one of them.