I was surprised to learn that China built more wind-powered electrical generating capacity last year than coal, and also more solar than coal. I posted about this on the orange website.
Specifically, in 02020, the People’s Republic of China installed 71.7 GW of new wind capacity, 48.2 GW of new solar capacity (which was already larger than the rest of the world combined), and 38.4 GW(e) of coal capacity. Assuming typical capacity factors of 40% for wind, 25% for solar, and 60% for coal, that would add up to 23 GW average new coal, 29 GW average new wind, and 12 GW average new solar. (But China’s capacity factors are lower; see below.) New solar installations worldwide double on average every three years, which has slowed down from every two years in the 02010s.
Solar capacity factors vary widely by region. In California they’re 28.1%, but in Germany and the Netherlands only 10%.
For scale, total German energy use was about 3800 TWh/year over 02007–02013, including things like transport fuels. This works out to about 430 GW. Of this, 576 TWh/year (65.7 GW) was produced as electrical energy, which had reduced to (+ 60.94 81.94 35.56 59.08 131.69 50.7 45.45 18.27) = 484 TWh/year (55 GW) by the year 02020.
But China is a larger country than Germany. Chinese marketed energy consumption was 28 PWh/year (3.2 TW) in 02010, of which 3.9 PWh/year (440 GW) was electric. In 02019 they produced 7330 TWh electric calculated as (+ 4554 233 148 349 1270 32 405 224 113) rounded to three places. That’s 836 GW. (The 32 TWh of pumped-storage hydro may be double-counted.) In 02019 224 TWh/year (26 GW) was produced from solar and 405 TWh/year (46 GW) from wind, using 204 GW of solar capacity (capacity factor 13%) and 209 GW of wind capacity (capacity factor 22%). Also the 4554 TWh/year from coal (519.5 GW) is on a 1.041 TW basis, so their capacity factor is only 50.0%. Hopefully they’ll start installing their energy plants in more propitious places, like the Gobi, and the capacity factor will go up.
So probably last year’s new installations of 38.4 GW (coal), 71.7 GW (wind), and 48.2 GW (solar) will produce on average 19.2 GW (coal), 16 GW (wind), and 6.3 GW (solar). The resulting 22 GW (average) of renewable energy added last year amounts to 2.6% of the total current electric energy use of China. If we assume that China’s total energy use has increased by 90% since 02010, just as their electrical energy use did by 02019, it would now be 6.1 TW, and 22 GW is 0.36% of it.
Even though wind turbines have a lower cost per kilowatt and higher capacity factors, I think solar is the more interesting thing here, because it lasts for many decades and taps a much larger resource, so I’m going to focus on solar.
The relation between new installations and existing installations gives us an estimate of the growth rate of solar capacity in China: it’s increasing by 48/204 = 23.5% per year, giving a 3.3-year doubling time, similar to the way new solar capacity in the world has doubled every three years over the last couple of doublings; we can expect this to remain roughly exponential for a while. We can estimate the current installed capacity as 204 + 48 = 252 GW, or 0.252 TW. We can also perhaps estimate that China’s total and electrical energy usage each continue to grow at the same exponential rate they have been; 7.4% per year gives us the 90% increase we seem to be observing from 02010 to 02019. We can write this model down as follows:
installed = 0.252
cf = 0.13 # capacity factor
electric_usage = 0.836
total_usage = 6.1
fmt = '| %5s | %7s | %7s | %8s | %8s |'
print(fmt % ('', 'solar', 'solar', 'electric', 'total'))
print(fmt % ('year', 'TWp', 'TW', 'TW', 'TW'))
for i in range(40):
print(fmt % ('%05d' % (i + 2021),
'%.3f' % (installed * 1.235 ** i),
'%.3f' % (installed * 1.235 ** i * cf),
'%.3f' % (electric_usage * 1.074 ** i),
'%.3f' % (total_usage * 1.074 ** i),
))
With this model, China’s solar energy production exceeds its 02021 current electrical energy consumption of 836 GW in 02037, but doesn’t exceed its contemporary electrical energy consumption until 02045 (at which point we extrapolate that it will use 4.6 TWe) and finally catches up to its total energy consumption in 02059 at 92 TW.
| | solar | solar | electric | total |
| year | TWp | TW | TW | TW |
| 02021 | 0.252 | 0.033 | 0.836 | 6.100 |
| 02022 | 0.311 | 0.040 | 0.898 | 6.551 |
| 02023 | 0.384 | 0.050 | 0.964 | 7.036 |
| 02024 | 0.475 | 0.062 | 1.036 | 7.557 |
| 02025 | 0.586 | 0.076 | 1.112 | 8.116 |
| 02026 | 0.724 | 0.094 | 1.195 | 8.717 |
| 02027 | 0.894 | 0.116 | 1.283 | 9.362 |
| 02028 | 1.104 | 0.144 | 1.378 | 10.054 |
| 02029 | 1.364 | 0.177 | 1.480 | 10.799 |
| 02030 | 1.684 | 0.219 | 1.589 | 11.598 |
| 02031 | 2.080 | 0.270 | 1.707 | 12.456 |
| 02032 | 2.569 | 0.334 | 1.833 | 13.378 |
| 02033 | 3.173 | 0.412 | 1.969 | 14.368 |
| 02034 | 3.918 | 0.509 | 2.115 | 15.431 |
| 02035 | 4.839 | 0.629 | 2.271 | 16.573 |
| 02036 | 5.976 | 0.777 | 2.439 | 17.799 |
| 02037 | 7.380 | 0.959 | 2.620 | 19.116 |
| 02038 | 9.115 | 1.185 | 2.814 | 20.531 |
| 02039 | 11.257 | 1.463 | 3.022 | 22.050 |
| 02040 | 13.902 | 1.807 | 3.246 | 23.682 |
| 02041 | 17.169 | 2.232 | 3.486 | 25.434 |
| 02042 | 21.203 | 2.756 | 3.744 | 27.316 |
| 02043 | 26.186 | 3.404 | 4.021 | 29.338 |
| 02044 | 32.340 | 4.204 | 4.318 | 31.509 |
| 02045 | 39.940 | 5.192 | 4.638 | 33.840 |
| 02046 | 49.326 | 6.412 | 4.981 | 36.344 |
| 02047 | 60.917 | 7.919 | 5.350 | 39.034 |
| 02048 | 75.233 | 9.780 | 5.745 | 41.922 |
| 02049 | 92.913 | 12.079 | 6.171 | 45.025 |
| 02050 | 114.747 | 14.917 | 6.627 | 48.356 |
| 02051 | 141.713 | 18.423 | 7.118 | 51.935 |
| 02052 | 175.015 | 22.752 | 7.644 | 55.778 |
| 02053 | 216.144 | 28.099 | 8.210 | 59.906 |
| 02054 | 266.938 | 34.702 | 8.818 | 64.339 |
| 02055 | 329.668 | 42.857 | 9.470 | 69.100 |
| 02056 | 407.140 | 52.928 | 10.171 | 74.213 |
| 02057 | 502.818 | 65.366 | 10.923 | 79.705 |
| 02058 | 620.981 | 80.727 | 11.732 | 85.603 |
| 02059 | 766.911 | 99.698 | 12.600 | 91.937 |
| 02060 | 947.135 | 123.128 | 13.532 | 98.741 |
Extrapolating an exponential trend over 40 years is very likely to be wrong, particularly when the trend has only been in effect for a few years. If we look back to 01993, 28 years ago, we see a significantly faster exponential trend in worldwide photovoltaic installations: 508 GW (peak, DC, nameplate capacity) in 02018, grown from maybe 130 MW in 01993, which works out to 39% growth per year, arithmetic mean, which is a 2.1-year doubling rate. [The estimate for 02019] was 627 TW, though, which is only 23.4% higher than the 02018 estimate, in line with China’s growth rate.
If China’s photovoltaic installations were to suddenly start growing at this faster rate, the model looks like this instead. They’re generating all their electrical energy from solar by 02034 instead of 02045 and all their energy from solar by 02042. Their total production in 02050 is 480 TW, 80 times more than their current energy consumption, 25 times more than the current world marketed energy consumption of some 18 TW, and an order of magnitude larger than their projected energy consumption at that time.
installed = 0.252
cf = 0.13 # capacity factor
electric_usage = 0.836
total_usage = 6.1
ygps = 39.2 # yearly growth percent, solar
ygpt = 7.4 # yearly growth percent, total
fmt = '| %5s | %10s | %10s | %8s | %8s |'
print(fmt % ('', 'solar', 'solar', 'electric', 'total'))
print(fmt % ('year', 'TWp', 'TW', 'TW', 'TW'))
for i in range(40):
print(fmt % ('%05d' % (i + 2021),
'%.3f' % (installed * (1 + ygps/100) ** i),
'%.3f' % (installed * (1 + ygps/100) ** i * cf),
'%.3f' % (electric_usage * (1 + ygpt/100) ** i),
'%.3f' % (total_usage * (1 + ygpt/100) ** i),
))
| | solar | solar | electric | total |
| year | TWp | TW | TW | TW |
| 02021 | 0.252 | 0.033 | 0.836 | 6.100 |
| 02022 | 0.351 | 0.046 | 0.898 | 6.551 |
| 02023 | 0.488 | 0.063 | 0.964 | 7.036 |
| 02024 | 0.680 | 0.088 | 1.036 | 7.557 |
| 02025 | 0.946 | 0.123 | 1.112 | 8.116 |
| 02026 | 1.317 | 0.171 | 1.195 | 8.717 |
| 02027 | 1.833 | 0.238 | 1.283 | 9.362 |
| 02028 | 2.552 | 0.332 | 1.378 | 10.054 |
| 02029 | 3.552 | 0.462 | 1.480 | 10.799 |
| 02030 | 4.945 | 0.643 | 1.589 | 11.598 |
| 02031 | 6.883 | 0.895 | 1.707 | 12.456 |
| 02032 | 9.581 | 1.246 | 1.833 | 13.378 |
| 02033 | 13.337 | 1.734 | 1.969 | 14.368 |
| 02034 | 18.566 | 2.414 | 2.115 | 15.431 |
| 02035 | 25.843 | 3.360 | 2.271 | 16.573 |
| 02036 | 35.974 | 4.677 | 2.439 | 17.799 |
| 02037 | 50.076 | 6.510 | 2.620 | 19.116 |
| 02038 | 69.706 | 9.062 | 2.814 | 20.531 |
| 02039 | 97.030 | 12.614 | 3.022 | 22.050 |
| 02040 | 135.066 | 17.559 | 3.246 | 23.682 |
| 02041 | 188.012 | 24.442 | 3.486 | 25.434 |
| 02042 | 261.713 | 34.023 | 3.744 | 27.316 |
| 02043 | 364.304 | 47.359 | 4.021 | 29.338 |
| 02044 | 507.111 | 65.924 | 4.318 | 31.509 |
| 02045 | 705.898 | 91.767 | 4.638 | 33.840 |
| 02046 | 982.611 | 127.739 | 4.981 | 36.344 |
| 02047 | 1367.794 | 177.813 | 5.350 | 39.034 |
| 02048 | 1903.969 | 247.516 | 5.745 | 41.922 |
| 02049 | 2650.325 | 344.542 | 6.171 | 45.025 |
| 02050 | 3689.252 | 479.603 | 6.627 | 48.356 |
| 02051 | 5135.439 | 667.607 | 7.118 | 51.935 |
| 02052 | 7148.531 | 929.309 | 7.644 | 55.778 |
| 02053 | 9950.756 | 1293.598 | 8.210 | 59.906 |
| 02054 | 13851.452 | 1800.689 | 8.818 | 64.339 |
| 02055 | 19281.221 | 2506.559 | 9.470 | 69.100 |
| 02056 | 26839.460 | 3489.130 | 10.171 | 74.213 |
| 02057 | 37360.528 | 4856.869 | 10.923 | 79.705 |
| 02058 | 52005.856 | 6760.761 | 11.732 | 85.603 |
| 02059 | 72392.151 | 9410.980 | 12.600 | 91.937 |
| 02060 | 100769.874 | 13100.084 | 13.532 | 98.741 |
Of course no exponential curve representing a real-world phenomenon can go on forever. China is only 9.6 million square kilometers; conservatively assuming a 35% “capacity factor” for its sunlight, thus 350 W/m², China only receives 3.4 petawatts of sunlight, which this projection would have it crossing in 02056. Covering all of China’s territory with 21%-efficient solar panels would produce only 700 TW, which this projection would have it crossing in 02052. Continuing to increase energy production past this level would require putting the solar panels somewhere that isn’t currently China, such as on the ocean, on the Moon, or in orbit around the Earth or the Sun.