By Dave Elliott
A report from Carbon Tracker and the Grantham Institute says that by 2050 solar PV could supply 29% of all global electricity. Its 2050 scenario still has a lot of nuclear in it, supplying about the same as PV, but not much wind (about 7%). That’s a bit odd, especially since the report indicates that onshore wind is currently cheaper in capital cost terms than either nuclear or PV. That stays true up to 2050 on its baseline scenario. But in its lower cost prediction, it sees PV undercutting all by 2050 – falling to M$390-643/GW, about a tenth of its rather low estimate for the current (and future) cost of nuclear (M$3706-4200/GW) and also much cheaper than the baseline capital costs for onshore wind (M$1640/GW) and offshore wind (M$2970/GW) by 2050. That does ignore the likelihood that wind costs may fall faster too and, given the low capacity factor for PV, in order to get to 29% of total output, they say 65% of global generation capacity would have to be PV by 2050 – around 10TW!
The report says the cost of solar PV has fallen 85% in seven years, and it could supply 23% of global power by 2040 and 29% by 2050, around 14,000TWh p.a, entirely phasing coal out and leaving natural gas with just a 1% share. It also predicts that electric vehicles will account for more than two-thirds of the road transport by 2050. Carbon Tracker senior researcher Luke Sussams said ‘Electric vehicles and solar power are game-changers that the fossil fuel industry consistently underestimates.’ Moreover, ‘further innovation could make our scenarios look conservative in five years’ time, in which case the demand misread by companies will have been amplified even more’.
Well yes, that may be so, but they seem to be overly enamoured of solar PV at the expense of wind. Even on their conservative data, the capital cost of offshore wind will be under half that for nuclear, and also biomass/BECCS, by 2050. Though it does seem likely that PV will continue to get even cheaper per kW, maybe dramatically so as they predict, which is good news. However, it has a low load factor (they say 15-20% compared to a rather conservative 20-40% for wind – 50% offshore is more likely by 2050), so costs per kWh will still be relatively high. What’s more, due to the low load factor, the cost of balancing variable PV on the grid will be higher than for other renewables – it will need more backup or storage. In their scenario, 0.25GW of storage is added for each GW of solar PV and wind when they get above a 20% market share. Given the large PV element expected by 2050, that may not be enough. More storage or other balancing measures may also be needed.
In addition, PV offers no rotational inertia. That will be an increasingly important issue for grid stability as renewables replace large coal and nuclear plants with their heavy rotating turbo-generators, which provide lock-step synchronous frequency stability. So some way will have to be found to maintain frequency stability, pushing up the system cost more: it is hard to beat angular momentum! It may be possible for battery storage systems to be linked with inverters to offer a form of ‘synthetic inertia’. More directly, so-called synchronous condensers might be used, basically the same as the back end turbo-generator part of a power plant, but without a power source other than from the grid. Old power station units can be used in this way. It is also possible to run live power plants with no power being produced, just freewheeling against the load, to provide some frequency stability/inertial load.
It could be that the quite large biomass input (about 8,000 TWh p.a.) in the scenario is meant to supply gas turbines or CHP plants to act as balancing plants and also provide a bit of synchronous rotational stability, along with the smaller (around 5,000 TWh p.a) hydro input and even smaller wind input (around 3,400 TWh p.a.). And of course there is still a large nuclear input – around 14,000 TWh p.a. This may be fine for rotational stability, but it’s inflexible and little use for backing up variable solar.
A more balanced mix, with more wind, as well as tidal and wave energy, would make grid management easier and cheaper, with less need for synchronous backup and massive amounts of storage – PV only works in the daytime, whereas wind (and wave and tidal) can be available all through the day and at night, and can add some synchronous inertia. Directly AC-coupled wind plants can help with this, as can tidal current turbines. Lagoons and barrages, like hydro plants, with their large turbine generators, are even better. And they provide more stable supply backup options
Certainly, quite apart from the issue of inertia, trying to balance so much PV could be very expensive. Unless you believe storage will get very cheap. That is possible, for short term battery storage especially (including EV batteries), but it’s not certain, and a key need anyway would be long-term bulk storage, for when there is no sun and no wind for a long period. Batteries can’t do that, EV batteries especially – EV users will want to use their cars daily! Assuming it is not possible to import much power via supergrids from other locations where there is sun and or wind, the best option for long term storage might be hydrogen stored in underground caverns. There’s a way to go on that. Pumped hydro is another option, but sites are even more limited than sites for underground gas storage.
Energy storage is the topic of a new short booklet I have been working on. It looks at all the options, their potential and limits. It concludes that there are no clear winners, with some being suited to short term balancing and others to longer term storage, and that the mix we may end up with will be shaped by the pattern of development of the other balancing measures, including smart grid demand management and super-grid imports and exports. Certainly storage isn’t the only balancing option. But the balancing mix will also be shaped by what supply options we choose. If PV is to dominate, along with nuclear, as in this new scenario, at the expense of wind, then storage may have the edge. Whereas if wind and other renewables are expanded, as in many other scenarios, nuclear could be excluded and less biomass could be used – not everyone is keen on that either!
The potential is certainly there. The downgrading of wind in this study to around 7% of the total by 2050 (despite it being said it had a 12% market share by 2050), is in fact just a feature of the way the modelling was run. As already noted, the report accepts that onshore wind is already cost competitive and offshore will be soon and that wind could expand more rapidly, but its growth ‘has been artificially capped’ by the model used: ‘Given our focus on solar PV and its potential growth from further cost reductions, such a flooding of the power mix by wind would mask our area of interest’. It would be interesting to see the results if that constraint was relaxed.