The Skaggerak subsea power cable connects Norway with Denmark. The NorNed cable connects Norway with the Netherlands. By 2019 the Nordlink cable will connect Norway with Germany and by 2021 the NSN cable will connect Norway with the UK. And now Scotland wants to connect with Norway via the NorthConnect link:
Figure 1: Existing, in progress and planned interconnectors with Norway
Why are these countries so anxious to connect to Norway? Because Norway’s hydro reservoirs are regarded as a large-scale storage battery that can be used to smooth out large quantities of intermittent renewables generation. The 2013 Joint Norwegian-German Declaration says as much:
Thanks to its natural endowments and previous investments, Norway possesses 50% of Europe’s entire power storage capacities. Therefore, Norway is in a position to provide large-scale, cost-effective, and emission-free indirect storage to balance wind and solar generation in other countries ….. In times of high wind or solar production, Norway can import cheap electricity from abroad, thereby saving water in its reservoirs. In times of low wind production, Norway can use the stored water to export power at higher prices. In this way, excess wind or solar production can be stored and used later.
On the face of it this looks like a win-win proposition. Germany and its renewables-heavy, storage-challenged neighbors get to store the intermittent wind and solar power they couldn’t otherwise use in Norwegian reservoirs and Norway makes money selling it back to them. But how is it going to work out in practice? Here we look into this question.
First a brief review of Norway’s electricity sector. Figure 2 shows generation by source between 1997 and 2013 (data from Statistics Norway):
Figure 2: Annual generation by source, Norway, 1997-2013
And the table below gives installed capacity and generation totals for 2013, the latest year for which data are available:
Norway’s electricity production remains dominated by hydro (note that this is almost all “conventional” hydro; pumped hydro contributed only 786GWh of generation in 2013.) However, Norway is not a major net power exporter. Figure 3 compares Norway’s export-import balance with its hydro generation over the 17-year period between 1997 and 2013. For ten of these years Norway was a net power exporter and for six years it was a net importer (exports and imports were balanced in 2006).
Figure 3: Annual hydro generation, demand and net imports/exports, Norway 1997-2013
As one would expect the export-import balance is strongly correlated with hydro generation (R squared = 0.92) with net power exports when annual hydro generation exceeds ~120,000GWh and net power imports when it doesn’t.
Figure 4 plots exports and imports separately. Because interconnector flows on the Nordic Grid are governed largely by short-term price differentials Norway still exports “cheap” hydro even in years when it’s a net power importer:
Figure 4: Annual imports and exports, Norway 1997-2013
It would nevertheless be a mistake to conclude that the Norwegians are making no use of the flexibility that hydro offers. Norway in fact already makes copious use of its hydro to balance diurnal and short-term demand fluctuations. Figure 5, which reproduces Statnett’s hourly production, consumption and export-import graphs for the past week, shows how production/consumption imbalances of up to 3GW in both directions were offset by interconnector flows that fluctuated between 2GW of imports and 3GW of exports. Relative to Norway’s 11GW average demand during the week these are large numbers. (During the same week the UK, with an average demand of 28GW, imported between 2.4 and 3.5GW of power and used none of it for load balancing.)
Figure 5: Hourly electricity production, consumption, exports and imports, Norway, 27th July to 2nd August, 2015.
The question is whether these imports and exports were an artifact of the Nordic Grid’s pricing mechanism or whether Statnett’s generation curve was the closest Norway’s hydro reservoirs could come to matching demand during the week. I suspect it was mostly the former, but there are the following reasons to suppose that it could have been the latter:
• Norway has no significant pumped hydro capacity, so when water goes through a Norwegian dam it doesn’t come back up again. Norway must therefore “store” power by replacing hydro generation with imported power whenever the imported power comes in, thereby preserving water above the dam. This makes the system less flexible than a pumped hydro system, which can be turned on and off at any time.
• Generation from Norway’s hydro plants is further constrained by maximum and minimum acceptable reservoir water levels, minimum acceptable watercourse flow rates and variable permitted release volumes at different times of the year. This adds another layer of inflexibility.
• Norway’s grid presently can’t provide reliable electricity to large parts of Northern and Central Norway. Statnett proposes to spend around 7 billion Euros over the next ten years on grid upgrades, but until they are completed the ability of the Norwegian grid to “wheel” power from place to place will be limited. Another complication is that there are reportedly no fewer than 178 separate grids in the country (Statnett controls only Norway’s “stem net” grid, which transports electricity over long distances. The other 177 are regional and local grids operated by local authorities and county councils.)
For the time being, however, we will ignore these problems and assume that Norway’s reservoirs really are the enormous storage battery many consider them to be. If all goes as planned the Skaggerak, NordNed, NordLink, NSN and NorthConnect links will be charging and discharging this battery with up o 5.9GW of electricity by 2021. Is the battery large enough to provide balancing services for this much input power, which we can reasonably assume will consist dominantly of wind and/or solar spikes? To answer this and related questions I constructed a spreadsheet algorithm that simulates how interconnector flows might work in practice. It makes the following assumptions:
- Power is exported to Norway from Denmark, Germany, Netherlands and the UK (DGNU for short) during periods of high wind/solar output and imported back to DGNU from Norway during periods of low-wind/solar output.
- The minimum duration of power flow in either direction is one hour.
- Power is always available for export from Norway when and in the amounts needed.
- Annual interconnector flows into and out of Norway are the same.
- Interconnector flows are adjusted to produce a result as close to constant baseload generation as possible. No attempt is made to balance output against DGNU demand.
- Power flows along Interconnector links with other countries are ignored.
- Transmission and efficiency losses are ignored.
Another disclaimer before proceeding. Optimized annual plans of the type you are about to see can be formulated only if complete wind generation data for the coming year are available, which in real life of course they won’t be.
To make the plan as realistic as possible I applied the algorithm to combined DGNU hourly wind output for 2013 (Figure 6, data from PF Bach). The amplitude of the wind spikes greatly exceeds the 5.9GW capacity of the interconnectors, which doesn’t look promising, but we’ll see what Norwegian hydro can do about it anyway:
Figure 6: Combined hourly wind generation from Denmark, Germany, Netherlands and UK, 2013
Figure 7 plots the surplus wind power sent to Norway during high-wind periods. This is the largest fraction of the spikes that can be sent within the limits imposed by 5.9GW interconnector capacity and the need to balance exports and imports. It represents 34% of total DGNU wind generation during the year:
Figure 7: Surplus wind power exported to Norway during high-wind periods
Figure 8 shows the 66% of total wind generation that remains:
Figure 8: Wind generation remaining in Denmark, Germany, Netherlands and UK after export of surpluses to Norway
And Figure 9 shows what Norway sends back to fill in the holes during low-wind periods:
Figure 9: Generation returned to Denmark, Germany, Netherlands and UK from Norway
Summing Figures 8 and 9 then gives Figure 10, which shows what combined 2013 wind generation from DGNU looks like after provision of Norwegian balancing services:
Figure 10: Final “balanced” wind generation
DGNU should be happy with this result. In its journey through Norway’s reservoirs the wildly erratic wind output shown in Figure 6 has been transformed into baseload generation delivered at a constant 7.9GW for 95% of the time (the lowest output over any one-hour period is 6.5GW). The residual spikes above the 7.9GW line, which presumably would have to be curtailed, make up only 15% of total generation, and interconnector utilization is a respectable 62%.
But we haven’t considered how Norway is positioned to handle the imports and exports. Figure 11 shows hourly exports and imports over the two-week period from June 16th to June 30th, 2013 and compares them with Norway’s electricity demand over the same period in 2015, which would presumably be similar to 2013 demand (data from Statnett). The green-shaded imports from DGNU are fed into the Norwegian grid(s) to offset hydro generation and thereby retain water behind the dams and the red-shaded exports back to DGNU are generated by releasing the retained water:
Figure 11: Norway’s hourly imports, exports and domestic demand, June 16th – 23rd
To match exports and imports to demand over this period Norway’s hydro generation would have to fluctuate in accordance with the curve shown in Figure 12, and it’s difficult to say whether it could do this without knowing what the impacts of reservoir level and other constraints might be. The range of fluctuation is, however, roughly twice that achieved during normal operations (see Figure 5) so the system could be sailing close to the wind.
Figure 12: Fluctuations in hydro generation needed to match exports, imports and demand, June 16th – 23rd
I next turned to Germany. As stated in the Joint Norwegian-German Declaration quoted at the beginning of the post Germany plans to use Norwegian hydro to balance not only its wind but also its solar, so I ran 2013 hourly wind-plus-solar generation for Germany through the algorithm to see how much difference the 1.4GW Nordlink interconnector might make. Germany’s combined 2013 wind plus solar generation is shown in Figure 13 (solar data also from PF Bach):
Figure 13: Hourly wind + solar generation, Germany, 2013
And found that a 1.4GW interconnector is far too small to make any significant difference. The output (Figure 14) is in fact almost indistinguishable from the input.
Figure 14: Final “balanced” wind + solar generation with 1.4GW interconnector capacity
But with five times as much interconnector capacity (7GW instead of 1.4GW) Germany gets near-continuous baseload output at 7.7GW, 12% curtailment and 62% utilization (Figure 15):
Figure 15: Final “balanced” wind + solar generation with 7GW interconnector capacity
Increasing interconnector capacity would appear to be the solution in this case. Eventually, however, a point will be reached where Norway’s reservoirs will no longer be able to handle Germany’s exports. At the 7GW level they would already be jumping through hoops to handle the solar spikes, which as illustrated in Figure 16 arrive once a day in summer with peak-trough amplitudes usually exceeding 10GW and very steep sides:
Figure 16: Norway’s hourly imports, exports and domestic demand, June 16th – 23rd, German wind + solar
Are there any other potential problems that I haven’t mentioned?
Well yes. First, like many other sources of renewable energy hydro is hostage to the weather. A series of dry years could leave Norway’s reservoirs scrambling to meet domestic demand with nothing left over for anyone else.
Second, we are considering the situation as of 2013. Over the next few decades forests of offshore wind turbines are scheduled to be planted in the North Sea and linked to mainland centers of consumption through the proposed North Sea supergrid. When the wind blows these forests of turbines will deliver far larger amounts of power than the centers of consumption can possibly use, and the bulk of it will have to be wasted if nowhere can be found to store it. But where will this immense amount of storage be found? One thing is for sure; it won’t be Norway.