[Greenbuilding] Using the grid as a battery

[Nick Pine]

John Daglish <johndaglish at gmail.com> writes:

>Look at the video to see were the Danes are going :)

>Smart Energy Systems: 100% Renewable Energy at a National Level (Full 
>Version)
>https://www.youtube.com/watch?v=eiBiB4DaYOM

They use the grid as a battery with the help of agile load-following CHP 
generators.

Dan Johnson <danjoh99 at gmail.com> writes:

>... there is no heat when the sun's not shining.  For this, Nick suggested 
>a heat pump and PV panels, "using the grid for storage". I pointed out that 
>the grid has no storage...

Yes, the grid has no storage, but it can essentially store energy if utility 
generators can react quickly. I suggest solar air heaters with enough 
overnight heat storage for houses on average January days.

Ryan Carlyle at 
https://www.quora.com/Is-nuclear-power-a-viable-alternative-to-natural-gas-peaking-power-plants 
writes:

"There is a popular (universal?) belief that pressurized water reactor 
nuclear power plants must take hours or days to ramp up or down and are 
therefore only suitable for base load power. However when you look at the 
facts it becomes clear that this is NONSENSE!"

His technical discussion concludes:

"We have non-load-following nuclear power today because we chose to build 
base-load nuclear plants. If you want load-following nuclear reactors, you 
just have to build them."

http://www.zdnet.com/article/why-baseload-power-is-doomed/ continues with:

"The baseload fallacy

"The notion that renewables cannot provide baseload power is really an 
artifact of the way the grid and its regulators have evolved. If all 
generators were able to ramp up and down on demand, and if grid operators 
were able to predict reliably when and where the sun would be shining and 
the wind would be blowing, accommodating any amount of power from renewables 
would be no problem."

With sufficient utility generator agility, prediction would not be required.

"A 2010 study called "The Base Load Fallacy" by Australian researcher Dr. 
Mark Diesendorf, an expert on integrating wind into power grids, fingers the 
"operational inflexibility of base-load power stations" as the main obstacle 
to further integration of renewables. "The renewable electricity system 
could be just as reliable as the dirty, fossil-fuelled system that it 
replaces," he observes, if demand were more efficient and intelligent, and 
supply were made up of a wide variety of renewable sources plus a small 
amount of gas-fired capacity to cover the peaks. The perpetrators of the 
baseload fallacy, he argues, are mainly the industries who benefit from the 
status quo: coal, oil and gas companies, the nuclear industry, power 
generators, and industries who depend on them like aluminum and cement 
manufacturers.

"Claims that renewables could never generate more than a few percent of grid 
power without taking down the grid have been given the lie by the real-world 
experience of areas that deliberately adapted their grids."

A few simple simulations could estimate grid storage efficiency as a 
function of utility generator capacity (eg 1 GW), utility generator response 
time (eg 10 minutes to 10 hours), renewable energy source power (eg 1 kW to 
1 MW), number of renewable sources (eg 10 to 1000), renewable source time 
correlation (eg 0 to 100%), average renewable source duration (eg 1 to 10 
hours per day), and load management ability (eg 10%-50% of the utility 
generator capacity.)

Intuitively, a 1 GW utility generator with a 10 minute response time and 
1000 10 kW uncorrelated renewable sources with 10 hour average durations 
could have close to 100% storage efficiency, especially if the utility can 
instantly turn on some off-peak loads like electric car chargers or 
municipal drinking-water well pumps. The storage efficiency would be lower 
with a 10 hour response time and 1000 1 MW correlated sources with 1 hour 
durations.

To simulate source correlation, a simulation could have 1 minute time steps 
and 100 virtual pulse generators activated by individual accumulators that 
add a fixed amount from a uniform random number generator with each time 
step, proportional to the daily source frequency, plus another amount 
proportional to source correlation from each pulse generator with an active 
output. When an inactive pulse generator accumulator reaches a fixed 
threshold, we could dump its accumulator and fire its output for a random 
interval with the average source duration.

Nick