Re: The Spanish Grid Drop-out - recently released information.

On 5/15/2025 2:37 PM, Glen Walpert wrote:

>
SCADA is used to monitor and control the grid, where control is done
in real time by adjusting the set points for real and reactive power
at all controlled power sources.  A small percentage of sources
being inaccessible degrades control by an insignificant amount.

>
I'll believe it is used to monitor the parameters at key locations in
"realtime", update set points and try to reset tripped breakers.

>
The 'S' in SCADA -- SUPERVISORY.  Like deciding how fas the subway car
will travel but not actually commutating the current in the motor.
You want local control that reports status (Data Acquisition)
and accepts control (Supervisory Control) from a higher level
functionary.

 Right, I had a feeling I was misremembering but was too lazy to look it
up.  Realtime depends on the application - SCADA signaling is fast
compared to the response times of large generators, realtime control of
the grid, not of the local inner loop generator controls (governor and
field exciter).

Yes.  I think Theo's notion was to provide a "reference signal"
to the "network" (via RF) instead of letting the network itself
supply that.  Thinking, perhaps, that <something> could better
hold the network's individual cogenerators in closer "check"
than allowing them to be semi-autonomous.

Indeed. And smaller generation systems can just get crushed like flies
if they try to stop a rampaging elephant as big GW systems drop
offline.
>
If the system did fail due to local over voltage excursions somewhere
and then failed to become stable again after a few suppliers dropped
out then their network stability analysis must be appallingly bad.

>
The problem is all of the distributed "residential" solar rivals "big GW
systems".  E.g., we have ~10GW of total solar, here -- but, of that,
300,000 individual residential systems in the 5KW (avg) power rating.
So, 1.5GW whose location is varied and diverse BUT whose control is
*likely* mandated by regulations/specifications created when solar was
intended to be "the dog's tail" -- tracking an otherwise stable grid.
>
And, if a residential cogenerator goes offline, the residence's LOAD is
still there, no longer being supported by that "local" generation.

 Right, the first generation of solar inverters was like that, the second
generation of 'grid assist' inverters will assist in grid stabilization if
properly used.  At least some of these require a local battery and limit
grid connected inverter output to 80% of stand-alone rating so there is
almost always some reserve for grid support.

How are the "legacy" installations treated?  Are they mandated to replace
(or update) their controllers to comply with "new requirements"?  Or,
is the hope that their effects will be lost in the noise?
(I'm thinking particularly about folks who paid for residential solar
installation and might -- potentially -- be forced/coerced into having
to do an equipment upgrade because of issues with "their version" of
the hardware.)

BTW the reason for using reactive power setpoints to control grid voltage
can be illustrated by considering the connection of a generator to the
grid.  The generator is brought up to slightly above synchronous speed
with the governor in droop mode (power output subtracts from speed
setpoint) and the field excitation will be in voltage regulation mode,
voltage set the same as the local grid connection.  As soon as the output
breaker is closed (in phase!) the generator speed and voltage is locked to
the grid, local control is not possible.  The governor will increase power
until the negative power feedback reduces the setpoint to exactly grid
speed, delivering a small amount of power to prevent the reverse power
relay from tripping; the governor speed knob now controls real power
output, not speed.
 At the instant the main breaker closes on the grid an aux contact switches
the field exciter to reactive power control mode, easily measured in a
balanced 3 phase system as Vac(-Ib), and initially set at 0.  With
reactive power output at 0 the field exciter will provide as much field as
required for real power output as set by the governor.  More excitation
than required for real power produces reactive power, less is a bad idea.
Small cogen may not get paid for or be required to provide reactive power
and won't for maximum efficiency, larger cogen will, and overall reactive
power supply must be balanced with load requirements by the grid
operator.  Increasing reactive power set points above load requirements
will raise grid voltage with the excess field excitation, not enough
reduces voltage.
 Thus the requirement that grid support inverters deliver reactive power on
low grid voltage - it has a much larger effect on grid voltage than real
power, at least when there are enough rotating generators and induction
motors on it.  Plus you can get quite a bit of it with a small reduction
in real power due to the sin-cos relationship.
 Glen