Currently, as we wander across the early part of the 21st century, most power production is centralized in massive power plants capable of producing hundreds, even thousands, of megawatts of power. Distribution systems then carry that power to where it is needed.
These massive baseload systems are essential to providing the large, and timely, energy requirements of heavy industry. So baseload systems will not be going away. They will, however, change. Fossil fuel systems, driven by the need to remediate increasing climate change, will eventually shut down. The massive coal-fired plants of China, India, and the United States will be the last to shutdown, but they will eventually go as well.
Baseload production will be taken up through a combination of hydro, nuclear, geothermal, and a hefty dose of conservation. Some thermal solar projects will be able to maintain a baseload capability, and widely distributed and linked wind farms can also add to the baseload.
Most readily-available sites for hydropower have already been exploited, and given the ecological damage of building new reservoirs, most new hydropower projects will be run-of-river developments, which aren’t as effective at delivering a baseload, and are far more seasonal. If production from such systems becomes a significant portion of the power generated, it may force some power-intensive industries, like smelting, to develop a more seasonal production model, or else relocate.
Power for residential use will come from more distributed sources, and many new communities will not be part of any power grid, instead generating their own power from a combination of battery-backed solar, wind, geothermal heating, and biogas-fueled generators.
New detached homes will incorporate solar panels as part of their roof construction, with power supplemented by local production. In larger centres, or areas connected to the grid, personal power production feeds back into the grid whenever available.
Higher-density buildings or housing projects will incorporate extensive solar panels along walls and even windows. It would be possible to incorporate a biogas generator and power system using the waste from the building’s inhabitants, but public acceptance of such a system will be slow.
At some more distant point in time, lower costs for orbital access could lead to the development of huge geosynchronous orbital solar power satellites. These would transmit power to earth in the form of diffuse microwaves, to be received by very large ground antennas. It remains to be seen if these microwaves would present any sort of health or environmental issue.
Further off is the prospect of fusion power: unlimited clean energy from the same nuclear reaction that powers the sun. For the past 30 years, however, those working on fusion power have been saying that fusion power is about 20 years away. They are still saying the same thing today, save for those who say that it will never happen.
Power production in the future will be satisfied by a combination of widely distributed residential-level systems, along with centralized baseload systems for industrial purposes. Power systems will be as low-impact as possible, and utilize renewable sources whenever feasible.