De-carbonising electricity generation

1. We'll need both wind and solar, plus hydro, tidal, small hydro, and wave power, and any legacy nuclear.   The reason we'll need both wind and solar is that the wind blows at night, when the sun isn't shining, and also that they are negatively correlated.  In summer in high latitudes, the wind drops off, but solar is surprisingly strong.  In winter, the sun in high latitudes is dim, but the winds tend to be strong.  Together, wind and solar can produce much less variable output than either on its own, except in the tropics.
2. We'll need 4 hours of storage.  4 hours will be enough to handle the morning and evening peaks in demand, and to do some short-term smoothing of the random fluctuations in wind, though geographically diversified wind farms will help with this.  But 4 hours will not be enough to take wind and solar to more than 90% of total generation.  We'll need long-term storage for that.
3. We'll have built-in generation overcapacity.  Remember, that wind and solar variability can mean not just too little generation but also too much.  Up to now, when it's potentially too much, the grid operators curtail output of wind and commercial solar farms.  They're also getting the ability to curtail output of rooftop solar too, in many locations.  In future, this output will not be lost but used to generate green hydrogen.
4.  Storing hydrogen is hard.  Its atoms are very small and easily escape through the interstices between larger atoms, meaning the using hydrogen for seasonal storage will be much less effective than using methane.  It also makes pipelines brittle.  Green methane can be made from hydrogen using the Sabatier method, with a small additional loss of energy.  Methane is already routinely stored for months with minimal losses; there are already methane (natural gas) grids in most mid- to high-latitude developed countries, where seasonal storage will be needed; gas peaking plants will not need to be retro-fitted to handle hydrogen; methane is much easier to store and transport than hydrogen; and it can be used for heating, though electric heat pumps may be cheaper and more efficient.
5. It's possible, though I am not convinced, that we will need some small percentage (10- 20%) of nuclear in the grid.  But we are a long way from 80% wind and solar penetration in most global grids--the most recent data have the global average at 10%.  Setting aside legacy nuclear, new nuclear, if it happens, will tend to be smaller plants, because the giants are just so expensive.  I've talked about SMRs, micro nuclear and the great potential for hydrogen-boron fusion reactors, but these are all at least 5 to 10 years away from them being successfully deployed.
6. Micro-hydro, tidal power, wave power and electricity generated from incinerating/gasifying rubbish will all be handy additions to the grid, because they  are uncorrelated with wind and solar.  Even waves, which are driven by wind, can be created by winds hundreds of kilometres away.

OK, how does the actual transition happen?

1. We could leave it to the market.  Wind and solar are much cheaper than new coal and gas, and comparable to existing coal and gas in most places on the globe.  As existing fossil fuel power stations age, they will be replaced by renewables plus storage.  This will take decades, however, so, since we want to substantially reduce emissions from power generations as quickly as possible, we'll need to give this process a nudge.
2. We could set targets for renewables.  Each utility/electricity producer will be required to achieve an annual/quarterly percentage from renewables, rising steadily over time, aiming for 100% within a decade.   Those utilities which exceed this target will earn carbon credits, those utilities which don't reach this target will have to buy credits.  This carbon penalty will encourage utilities to build out more wind and solar farms, as well as other renewable supply, including nuclear.
3. We could run a reverse auction where individual power stations would offer to close down for a fee.  Let's say a decline of 10% a year in emissions from electricity generation is required.  Each year, the government would offer a lump sum payment to the power stations which would close down, choosing those with the lowest offers until the total reaches 10% for that year.  Conditions could be applied to any deals: redundancy payments to workers; requiring the output of the power station to be replaced with new "firmed" output from renewables; making sure the new wind and solar farms are located close to the communities where coal power stations are ceasing to operate.

Source: Our World in Data