What is the duck curve, and what does it mean for utility-scale battery storage?

DanKegel

Hi folks!  I'm a relative newbie here, but I posted for a long time in that other forum.  
I tend to take a wonky look at solar energy policy (though I do have solar on my own roof, so I talk nuts and bolts sometimes, too).

I don't think the "duck curve" has been mentioned on this forum yet.  That's the name for the funky shape the total load on fossil fuel generators looks as more and more solar power is installed.  See http://www.raponline.org/knowledge-center/teaching-the-duck-to-fly-second-edition/ for the article that popularized the term.

This week, the EIA released its latest graph of how load and wholesale energy prices in California vary during a typical day, and how that's changed over the last few years: https://www.eia.gov/todayinenergy/detail.php?id=32172
That price spike now reaches $70/MWh in the evening.
Is that high enough to attract battery peaker plants yet?

A recent study, https://www.lazard.com/media/438042/lazard-levelized-cost-of-storage-v20.pdf, says not yet.
But that same study says a battery might pay for itself when used for Frequency Regulation + Demand Response, which doesn't require as much storage, and values batteries' nimbleness more.
And hybrid peaker plants are indeed starting to be built; https://www.bloomberg.com/news/articles/2017-05-05/slap-a-battery-on-a-gas-turbine-and-make-an-extra-1-4-million suggests the payback time for the battery is 3 years (because of quick-start payments).
So, at least in a very small sense, battery peaker plants are here already.  

Further reading:
http://www.utilitydive.com/news/is-battery-energy-storage-at-a-turning-point-for-us-utilities/440055/

Let me know if that was too wonky for this forum.

BB.

At this point, most short term grid storage (minutes to 15 minutes of capacity) are expensive and fairly high costs. In Europe (another thread here), the UK is paying ~8x normal power costs for a bunch of relatively small (to avoid the whole CO2 regulations, of undetermined NOX output) to support the loss of Coal Plants, the lack of new natural gas turbine generators, etc. to carry through the variability of green power).

In Germany and other countries with high green energy content, their CO2 generation (aka fossil fuels like coal and such) are running in hot standby to keep the grid stable) is pretty flat (not declining).

In California, we have (relatively) high green content. And we are paying Arizona to sink our excess "green power" into their grid, to keep California grid stable. And from all of this, we have among the highest retail power costs in the US.

Very interesting that your link shows that we have more variable (peak and valleys) of power in 2016/17 than we had in 2012. Presume that solar/green power during the day is taking away the base loads--Which makes power manage with large plants (nuclear, coal) more difficult and expensive. While needing more hydro and alternative storage (battery, flywheel, etc.).

You can see the California Independent System Grid information (about 85% of total grid) here. Also has solar/green power mix information:

http://www.caiso.com/Pages/TodaysOutlook.aspx

-Bill

Estragon

Interesting stuff. My son works for a company that started out doing load projections for utilities but is now mainly speculating in the day-ahead power markets using the prediction software. He's in the computer end of it, but I'll have to ask him if this has changed their trading any.

Is there an appetite for any of the solutions (eg real-time utility control of water heating)?

BB.

30+ years ago, you could get interuptable service where the utility would give you a radio controlled relay for an electric water heater or an A/C system.

Don't know if residential program still exist, but here is the commercial program (100 kwatt average load).

https://www.pge.com/en_US/business/save-energy-money/energy-management-programs/demand-response-programs/base-interruptible/base-interruptible.page

- Bill

bill von novak

Estragon said:

Interesting stuff. My son works for a company that started out doing load projections for utilities but is now mainly speculating in the day-ahead power markets using the prediction software. He's in the computer end of it, but I'll have to ask him if this has changed their trading any.

Is there an appetite for any of the solutions (eg real-time utility control of water heating)?

Sure; we were enrolled in such a program for a few years.  Utilities could increase thermostat settings for up to four hours in response to high demand.  You got $75 a year for signing up (as well as all new thermostats.)  It was called the SDG+E Reduce Your Use Thermostat Program.

DanKegel

BB. said:

30+ years ago, you could get interuptable service where the utility would give you a radio controlled relay for an electric water heater or an A/C system.

Don't know if residential program still exist...

Here's a residential water heater demand response trial I'm kind of excited about:
https://www.bpa.gov/Doing Business/TechnologyInnovation/TIPProjectBriefs/2017-DR-TIP-336.pdf

It's neat because it requires no communications installation; the load shedding signal is delivered over FM broadcast radio.
Here's a bit about how that can be done securely:
http://www.sgip.org/wp-content/uploads/SGIP_WP_Smart_Grid_System_Security_with_Broadcast_Communications_FINAL_March_3_2017.pdf
and here's the module they're using in the trial:
http://www.e-radioinc.com/solutions/

In areas with good FM reception (like my basement :-) ), the idea of bypassing WiFi and Zigbee and all that crap for a secure, simple, FM RDS approach is pretty attractive.  They could probably add the needed electronics to water heaters for $50 or so.

Of course, water heaters are all natural gas in Los Angeles, but if we are going to move to 100% carbon free energy, that probably has to change... and when it does, we need all those water heaters on a cheap demand response system so they can be used to help shave the peaks off the load curve.

petertearai

Here in New Zealand we have pilot control of hot water . Either a pilot wire, or a thing called ripple control . a frequency is injected into the supply line . this turns a relay on or off to drop the power to the hot water cylinder when load is high . Typically early evening when everyone get home and turns on ovens and heaters . This will give reduction in the cost per unit of electricity.

Solray

> @bill von novak said:
> Estragon said:
>
>
> Interesting stuff. My son works for a company that started out doing load projections for utilities but is now mainly speculating in the day-ahead power markets using the prediction software. He's in the computer end of it, but I'll have to ask him if this has changed their trading any.
>
>
>
> Is there an appetite for any of the solutions (eg real-time utility control of water heating)?
>
>
>
>
> Sure; we were enrolled in such a program for a few years.  Utilities could increase thermostat settings for up to four hours in response to high demand.  You got $75 a year for signing up (as well as all new thermostats.)  It was called the SDG+E Reduce Your Use Thermostat Program.

Now everyone is using programmable wifi thermostats to control use very effectively. You can set up several programs easily and change the settings when you are out of the house and going to be late getting home to save even more.