How does load affect frequency on the power grid? This story about the use of battery/freewheel based Frequency Regulators confused me about how the 60hz frequency of the North American power grid was set--saying that it was kept at that frequency by balancing load and supply. I used to think that it was only voltage which was affected by this balance, and that the frequency was determined by the speed of the rotors in the generator. EG see Wikipedia's page on Alternator Synchronization. 
Can someone help me understand the physics behind this?
 A: The physics is actually much easier than it seems at first glance.
Power generators are engines just like the everyday ones we see all around in our cars, lawnmowers, snowblowers, etc. 
Except for new power sources like some wind and solar systems with electronic inverters, the vast majority of power is supplied by large rotating AC generators turning in synch with the frequency of the grid. The frequency of all these generators will be identical and is tied directly to the RPM of the generators themselves, generally 3600 RPM for gas turbines and 1800 RPM for nuclear plants. If there is sufficient power in the generators then the frequency can be maintained at the desired rate (i.e. 50Hz or 60Hz depending on the locale).  
The power from the individual generators will lead the grid in phase slightly by an amount roughly corresponding to the power they deliver to the grid. 
An increase in the power load is accompanied by a concurrent increase in the power supplied to the generators, generally by the governors automatically opening a steam or gas inlet valve to supply more power to the turbine. However, if there is not sufficient power, even for a brief period of time, then generator RPM and the frequency drops. 
This is much like what happens to a car on cruise control if you start going up a hill, if the hill is not too steep you can maintain speed, once you reach the limits of the torque supplied by the engine, the car and engine slow down. If the combined output of all the generators cannot supply enough power then the frequency will drop for the entire grid. All the generators slow down just like your car engine on a hill.
For large grids the presence of many generators and a large distributed load makes frequency management easier because any given load is a much smaller percentage of the combined capacity. For smaller grids, there will be a much larger fluctuation in capacity as delays in matching power supplied are harder to manage when the loads represent a relatively larger percentage of the generated power.
So a battery systems like the one in the article is really designed to keep short-term fluctuations in power requirements from dropping the frequency because of lags in the governors and generators which require a finite time to adjust to the new power requirements. These "frequency regulator" power stations can supply very high power for short bursts to keep the power requirements even so that the other generators don't see too much load faster than they can respond due to mechanical limitations.
A: Applying Newton's equations for rotating masses to a turbine-generator system gives the following expression (Samarakoon, after Kundur).
$J\frac{d^2\theta}{dt^2}=T_m - T_e$
$\theta$: angle (rad) of the rotor with respect to a stationary reference.
$J$: moment of inertia.
$T_m$: mechanical torque from the turbine.
$T_e$: electrical torque on the rotor. 
The Inertial response for a generator is characterised by its Inertia Constant, H, with units of seconds, defined as (Samarakoon, p40):

the ratio of kinetic energy stored at synchronous speed $\omega$ to the
  generator kVA or MVA rating, $S$.

$$ H = \frac{0.5J\omega^2}{S}$$            
An equivalent Inertia Constant for an entire system can be estimated: (Ekanayake, Jenkins, Strbac)
$$
H_{equivalent} = \sum_{gens} H_{gen}/S_{gen}
$$
A value for the GB system (in 2008) was estimated at 9s (by Samarakoon), projected to drop as far as 3s in 2020 with a high wind penetration.
When modelling Inertial response (more commonly referred to as frequency response), a power system can be simplified to a transfer function (Ekanayake, Jenkins, Strbac):
$$
\frac{1}{2H_{equivalent}s +D}
$$
$D$ is known as the Damping Coefficient - the term encapsulates response from frequency responsive demand (Mu,Wu,Ekanayake,Jenkins,Jia).
To maintin system stability, the frequency must be closely controlled. Traditionally this is achieved through droop controllers on steam turbine generators. Increasingly, however, energy storage and demand response are contributing.
See this question also; https://engineering.stackexchange.com/questions/2245/quantifying-inertia-on-the-electricity-grid
A: Higher load in the grid -> lower overall frequency and vice versa
More info here:

Frequency and load
The primary reason for accurate frequency control is to allow the flow
of alternating current power from multiple generators through the
network to be controlled. The trend in system frequency is a measure
of mismatch between demand and generation, and so is a necessary
parameter for load control in interconnected systems. Frequency of the
system will vary as load and generation change. Increasing the
mechanical input power to a synchronous generator will not greatly
affect the system frequency but will produce more electric power from
that unit. During a severe overload caused by tripping or failure of
generators or transmission lines the power system frequency will
decline, due to an imbalance of load versus generation. Loss of an
interconnection, while exporting power (relative to system total
generation) will cause system frequency to rise. Automatic generation
control (AGC) is used to maintain scheduled frequency and interchange
power flows. Control systems in power plants detect changes in the
network-wide frequency and adjust mechanical power input to generators
back to their target frequency. This counteracting usually takes a few
tens of seconds due to the large rotating masses involved. Temporary
frequency changes are an unavoidable consequence of changing demand.
Exceptional or rapidly changing mains frequency is often a sign that
an electricity distribution network is operating near its capacity
limits, dramatic examples of which can sometimes be observed shortly
before major outages. Frequency protective relays on the power system
network sense the decline of frequency and automatically initiate load
shedding or tripping of interconnection lines, to preserve the
operation of at least part of the network. Small frequency deviations
(i.e.- 0.5 Hz on a 50 Hz or 60 Hz network) will result in automatic
load shedding or other control actions to restore system frequency.
Smaller power systems, not extensively interconnected with many
generators and loads, will not maintain frequency with the same degree
of accuracy. Where system frequency is not tightly regulated during
heavy load periods, the system operators may allow system frequency to
rise during periods of light load, to maintain a daily average
frequency of acceptable accuracy.[27][28] Portable generators, not
connected to a utility system, need not tightly regulate their
frequency because typical loads are insensitive to small frequency
deviations.

