I currently have solar panels on about 35% of the roof area of my house, and I have reduced my electricity cost to near (but not quite) zero. With 80% roof coverage, then, I could provide a surplus equal to what I am using.
According to the chart, Total Electrical Generation is 38.1 quads (with 25.7 going to “rejected”, 67% waste — what’s up with that??).
What’s interesting here is that the electrical infrastructure, unlike other energy infrastructure, is intrinsically bi-directional — if the consumers were turned into producers, they can send power back out over the same wires that they are currently using to receive power.
This means that if residential and commercial customers turned into producers (eg, residential and commercial solar that on net produced as much surplus energy as they had been using), they could completely supply the current industrial use of electricity, even accounting for a 2/3 energy loss (to “rejected”).
So, from a pure capacity and distribution infrastructure point of view, all current electric power generation facilities could be replaced by distributed local generation and storage, using the current grid.
This assessment is correct. There are, however, complications to implementation which we are resolving now. There are fundamental design shortcomings in much of the existing residential grid-tied solar assets. The problem is complex, both technical and regulatory, in adding storage capacity to these existing installations.
Consider the situation of grid failure. This can be as simple as a failed line or transformer, or as widespread as a Carrington Event. The vast majority of existing grid tied inverters, by law (UL1471), must cease production of alternating current in the event of an upstream grid failure. This is implemented in the inverters themselves by the requirement for a 60hz signal from the grid to sync to… no grid, no signal to sync to, no power to the local premises. There are several valid reasons for this, particularly safety in the event of downed lines and equipment protection from local overproduction. To overcome this, we must do all of the following to keep premises with these types of systems installed live in the event of grid outage: Completely isolate the grid from the the local premises. Provide our own battery backed 60hz pure sine wave inverter. Control output of existing grid tied assets or add diversion loads based on local load readings, in order to match production to load.
This kind of retrofit is required to make distributed production and storage nodes a viable architecture in the near future. Not an impossible patchset to apply, to be sure, but it is proving to be an interesting challenge to solve.