Solar Eclipse Could Affect the Grid

[S'toon]

Operators of Ontario’s power grid are bracing for a sharp drop in electricity generated by solar panels during a summer solar eclipse coming Aug. 21, CTV News has learned.

Those with commercial or residential solar panels will see a drop-off in generation when the moon passes between Earth and the sun, turning day into night. That will have those users – tens of thousands of them – switching over to the conventional grid, raising worries about overburdening a system that will already be running at a summer peak.

“We have significant quantities of solar in Ontario,” said Leonard Kula, CEO of the Independent Electricity System Operator, which operates the electricity grid in the province. It’s been preparing for the eclipse for more than a year.

“We will see a reduction of about 70 per cent of the solar generators in Ontario that on the day will be about 10 per cent of our supply.”

<snip>

Full article at:
Solar eclipse in August raising worries about Ontario's power grid

 

[Ampster]

The President of the California Public Utilities Commission expressed some concern as well. California is not even close to the main shadow area.

 

[SageBrush]

How long will the eclipse last ?

 

[NikeWings]

The President of the California Public Utilities Commission expressed some concern as well. California is not even close to the main shadow area.

Several months ago I read an interesting bit which revealed a 1/3 of CA power is imported due to growing sources of intermittent power gen within the state. Mostly from the NW and SW.....maybe the shadow will be impacting production zones typically relied upon by CA.

 

[NikeWings]

How long will the eclipse last ?

We'll be watching from Jackson Hole and it's estimated start/finish for about 2:45 mins, although totality is only 3 minutes or so.

 

[nwdiver]

If only there was a way to know precisely when this is going to happen so we can be prepared

 

[GasKilla]

If only there was a way to know precisely when this is going to happen so we can be prepared

Just like "nobody could have predicted the breech of the levees" this is just something we'll have to deal with if it happens

 

[idoco]

Doesnt' this happen every day, aka night, without armageddon happening?

 

[RandyS]

Idoco,
The challenge to this particular situation is the solar power level at the time that will be lost and the short nature of the dip it will cause.

Fossil fuel plants are not known for their instantaneous ability to ramp up, but batteries can. There just aren't enough battery resources yet installed.

In San Diego, the eclipse is only rated at a level of 50%, but we have well over 500 MW of rooftop solar generation online. So within a 3 minute period, if it turns out to be a sunny day - and it probably will, I would guess we could lose at least 300 MW of generation and then have it "come back on" in 3 minutes.

So to keep the power system stable, it's really two event, the loss of that much generation, and the reappearance...It will be interesting, that's for sure....

 

[bxr140]

So basically, it's like nighttime, but only for a couple minutes?

 

[ReddyLeaf]

Been there, done that. Even in a smaller country. I guess John Q Public will just have to wait and see if the American and Canadian grid operators are as smart as the German ones.
German power net survives solar eclipse | News | DW | 20.03.2015

 

[Ampster]

Here is a map of where the totality will be:
http://www.eclipse2017.org/2017/maps/whole-us.jpg

 

[ecarfan]

Here is a graph showing the dip in output of my rooftop solar system during the eclipse as experienced near San Francisco.

 

[Ampster]

Yes the CAISO graph of solar output looks similar. The grid was resilient and I have not heard of any glitches.

 

[swaltner]

@ecarfan, thanks for bringing this thread up to the top. I briefly looked for it, but didn't find it.

It's really interesting in the difference between the visible light change and the solar irradiance during an eclipse. People seem to intuitively judge the impact of the eclipse on solar production based off the rate at which the light drops. The amount of light doesn't seem to change until right around totality, but in reality, there is a gradual decrease in the amount of solar irradiance over the whole ~3 hour event. So the grid is adjusting over a 3 hour period, not a 3 minute period. Not to say the grid operators don't have to plan for it, but it's a much easier problem to deal with than what I was originally imagining.

Here's the graph from my solar system. I was a little close to the center line, so 93% coverage instead of ~73% coverage that resulted in your solar output graph. Partial eclipse time here was 11:36 through 14:04, which lines up exactly with the trough right in the middle of the day. It wasn't a clear day, so my production graph isn't a clean curve, but you can definitely see the impact of the eclipse.

 

[mblakele]

I figure I lost about 4% for the day. I'm sending the bill to the moon.

As expected there was no impact on the grid, which is great. Talking about it can also help raise awareness of the importance of grid-attached storage.