Ah, the apparent paradox of energy. Don't forget, electricity goes equally back and forth. It's an AC system after all.
How then could we ever feed any electricity to any homes? Answer: we don't have to.
Instead, things become much clearer once we start start employing physics terminology. (Sometimes the answers even seem simple and obvious.) First, never say "electricity." If you mean energy, say energy, or perhaps "joules." If you mean charge, say charge, coulombs, or "electrons."
Instead of selling "electricity," the utility companies are actually selling electromagnetic energy. The electrons inside the wires may wiggle equally back and forth, but the energy behaves very differently.
As with any transmission line, the energy in the AC power grid is traveling out in the space surrounding the wires. You may already know that both signals as well as "electric power" travel at nearly the speed of light? That's because they're the same thing AS light; same as radio waves propagating along a waveguide, same as magnetic fields around inductors, or the e-fields between the plates of capacitors.
Unfortunately, all these correct ideas are only taught in engineering school, while back in K-12 grade school, instead we're all fed "lies to children," a collection of over-simplified explanations which have little to do with reality. Electrons zooming through solid copper at the speed of light? Batteries and capacitors "storing charge," etc.? These distorted ideas must first be abandoned, if we want to have any hope of deeply understanding the AC grid.
To understand basic EM physics, even before treating equations, we first need to realize:
The word "conductor" means "materials which are full of mobile
electricity." For example, metals contain an enormous amount of
free electric charge, roughly 10K coulombs per cubic cm of copper. In copper,
each atom of the metal donates one mobile electron to the "common
sea." Physicists call it the "electron sea" or the "ocean of
charge." A piece of wire is like a pipe that comes pre-filled
with electricity, and no bubbles allowed.
In AC systems, electricity moves back and forth only. It doesn't
flow across the grid, and it doesn't flow across circuits. The electricity
(electrons) were already inside the wire, even before the power supply was connected.
Generators don't create any electricity, and loads don't consume any
electricity. Instead, generators are charge-pumps. Instead, electrons inside wires just wiggle back and
forth. (Or in DC circuits, all the electrons move slowly in closed circles,
with no electricity being gained or lost. They move like a slow flywheel; a slow drive-belt.)
Electrical energy flows almost instantly from a source to a load. In an AC
circuit, this energy doesn't wiggle back and forth. It flows continuously
forward, as waves.
Electrical energy, or electromagnetic energy, is composed of e-fields
and b-fields, corresponding to volts and amperes respectively. When electric energy is sent through coaxial cable,
all the energy is located inside the plastic dielectric. When it's
sent across circuits, or along transmission lines, it's located in empty space
surrounding the wires (see diagram of fields below. That's a diagram of
the energy flowing along your AC power line.) While the amperes may truly be inside
the wires, the joules are all found in the space outside.
Electricity wiggles slightly back and forth, while electrical energy flows forward. In other words, two separate things are flowing along any circuit. There is no single stuff called "electricity." This isn't a complicated concept; it's all about Waves-versus-Medium. In circuits, the slow electrons of the copper metal act as the medium for the traveling energy-waves, while the lightspeed waves themselves are located in the EM fields just outside the wires.
During engineering school, in our fields/waves class about transmission-line theory, note well that transmission lines have no low-freq cutoff. They function the same at microwave freqs, HF, audio, and all the way down to DC. Transmission-line theory isn't just about radio, it's about all circuits everywhere. The EM energy-flow is composed of b-fields created by circuit currents, and e-fields created by circuit potentials. After all, a high-freq AC supply can drive an incandescent lightbulb, just as easily as it can power a dipole antenna. (But in the case of the dipole, rather than being absorbed by the lightbulb, the traveling EM fields instead leap off the wires and propagate away into empty space.)
Above: b-field of a circuit, e-field of a circuit, and the watts of EM energy-flow
What happens to excess energy fed to the power grid? Well, imagine that all the dynamos are just conventional DC supplies, or even batteries. Imagine that the loads are just resistors, or perhaps incandescent light bulbs. Now, what happens when excess energy is fed to the grid? How can we even do such a thing?! Easy: just slightly raise the voltage of one of your DC supplies. All the light bulbs will burn slightly brighter. Ideally, the added energy flows to all the bulbs on the entire grid (but if the grid conductors have slight resistance, the closer bulbs will receive a bit more of the overvoltage/ energy.) The same thing happens on the AC grid. A simple way to feed "excess energy" is to raise the voltage of one of the AC dynamos. The grid voltage will rise slightly, and the injected energy will end up flowing to all the connected loads.
Note that if we raise the voltage of one battery, then it fights with the other batteries, and it charges other batteries hooked to the grid. Or at least for wires having some ohms, it reduces their output current, if not actually reversing it. For an AC grid with dynamos, the same thing happens, and if one dynamo has a too-high voltage, it will spin the other dynamos as if they were AC motors, or perhaps just reduce their output-currents, if the grid wires aren't perfectly conductive.
Is there a minimum amount of (joules) which can be fed into the grid? No, not unless you're talking about single EM photons! For example, if your DC power supply is set to exactly the same voltage as the entire battery-grid (say 120VDC,) then its supply-current will be zero. The wattage at that DC supply will be zero, and no energy is flowing. Now try turning up the voltage of the one DC supply. This produces a current (as if the entire grid was a giant resistor,) and now there are some watts of energy-flow, directed out of your DC power supply and going out into the grid. There is no lower limit to this (you could adjust it to produce an amp, or a mA, or a picoamp.) Just multiply this current by 120V, to find the rate of energy-flow being injected into the grid.
Photovoltaics? Connecting DC supplies to AC systems is conceptually fairly complicated.
Therefore, in order to clarify, either pretend that the entire power-grid is batteries, or DC generators (and let the questions answer themselves.) That, or for an AC grid, first hook your PV panels to a DC motor, and connect the motor's shaft so it can spin a small AC dynamo that's connected to the AC grid. Note that such a dynamo will constantly spin at 3600RPM, "idling" or "freewheeling," with the DC motor acting as a generator, the DC motor-current ideally being zero, and no energy being sent into the AC grid.
With the above setup, whenever the PV panel is producing zero output current, the DC motor is spinning without torque, while your AC dynamo freewheels (no torque is on the connecting shaft, for ideal frictionless bearings.) In that case the AC dynamo is "idling" at 3600RPM/60Hz, and doesn't send any energy into the grid. The back-EMF of the spinning DC motor will match the PV panel's output-volts, with zero drive current. Next, add brighter sunlight, drive the freewheeling DC motor a bit. This pushes the AC dynamo slightly ahead, slightly increasing the AC output voltage. Math: if we add up the dynamo's instantaneous wattage over one AC cycle, we'll find that electrical energy is now flowing out of the dynamo and into the grid. (Heh, or instead put your thumb against the shaft between AC dynamo and DC motor when it's freewheeling, and instead, the energy-flow reverses, and the dynamo will start taking energy from the grid, functioning temporarily as an AC motor.)
If we connect the shaft of a DC motor to an AC dynamo, that's called a Rotary Converter, and it performs the same task as modern solid-state inverters. To understand energy-flow, it pays to first understand DC motors which drive dynamos. Then later, figure out the details of the transistor-switching circuitry which simulates a motor-driven AC dynamo.
In physics, what is Electricity? What's the narrow technical definition of the word "Electricity?" We could just ask JC Maxwell, Einstein, Faraday, that whole crowd. They all agree: Electricity isn't the flow of the charges, and Electricity is not a form of energy. Traditionally, Electricity is the coulombs! That way, a flow of coulombs is the same as a "flow of electricity," or in other words, an electric current. Current isn't electricity. Current is a motion of some pre-existing electricity.