[time-nuts] Re: Power line timing -- setting a clock

[Poul-Henning Kamp]

Andy Talbot via time-nuts writes:

> I can't help wondering if solar and wind inverters could be designed to
> push energy into the grid with leading or lagging power factor. 

They can, and for non-consumer-size installations it is already a
requirement pretty much everywhere, although the specific parameters
differ a lot from grid to grid.

Now that battery prices have come down, there are people on the grid-side
arguing for even tougher parameters, such that new installations will be forced to install "at least a couple of minutes worth of batteries"

The big issue is regulation and control.

Once they are electro-magneto-mechanically coupled together via the
grid, big rotating iron cylinders will Do The Right Thing because
the Laws of Physics controls them, and you can concentrate on getting
the power balance right.

Once your power sources are tens of thousands of computers, each
controlling an inverter from some unsynchronized source of power,
it becomes a really, /really/ challenging control & regulation
issue.

Some people have argued for taking the easy way out:  Simply lock the
grid to UTC, so that everybody can have a local "GPS reference" and know
if phase is leading or lagging.

Problem is, as nice as it sounds, it doesn't work:  Both distances
and transformers mean that there are intrinsic phase differences
and what's worse, they deepend on the amount and very much on the
direction of the energy flow through the transformers.

The other easy solution is to make the "backbone" HVDC, which not
only makes the cables (air or ground) improve rather than cause
stability with their capacitance, but also reduces the control
problem to simply one of voltage.

That is probably where we end up eventually, but as of yet, the
cost is still prohibitive, and the idea of having three or more
HVDC lines meet on a big bolt&nut connection, where you have no way
to steer the power flow, makes a lot of people uncomfortable.  With
HVAC you can at least steer the flow somewhat by adjusting the
voltage with the winding-selectors on the transformators.

So right now, everybody are implementing "intelligent algorithms"
in the inverters, and yes, some call it "AI".  As of yet we have
little idea, and even less confidence, that it will Do The Right Thing
when parameters jump through their windows, so everybody is
looking forward to the next grid collapse, and hoping it happens
to a grid somewhat like, but not exactly their own.

One other huge difference between big rotating lumps of iron is
how much abuse they and their copper-windings can endure, compared
to high voltage semiconductors which are uncomfortably sensitive.

That's why some grids are taking no chances and installing "increased
momentum" in their grid, in the shape of big lumps of rotating iron,
configured as flywheels with a single motor/generator.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.