#6: The Intermittency Problem with Renewables
We’ve done a lot of work to reduce our carbon emissions, let’s take a step back to see how far we’ve come.
This is the 2019 US energy system:
And this is the one we’ve developed over our last 4 posts:
Renewable energy grew from 11.5% to 37.8% of our energy supply. Let’s calculate how much we’ve reduced our carbon emissions.
So how much did we reduce carbon emissions?
Natural gas, coal and petroleum release different amounts of CO2 per kWh.
Coal: 0.74 pounds of CO2 / kWh
Petroleum: 0.55 pounds of CO2 / kWh
Natural Gas: 0.4 pounds of CO2 / kWh
Based on the 2019 energy system we emit approximately TODO pounds of CO2. With this new wind + solar powered energy system, we’ve reduced our CO2 by TODO%.
But, not so fast
This system we’ve developed works great when the sun is shining and the wind is blowing. When it’s night-time and there is no wind, we will lose over 30% of our energy production.
This creates the need to find the lost energy somewhere else, like a backup natural gas plant to generate electricity. Which, of course, emits CO2. What’s more, the variability in supply creates the need to scale up and down these backup electricity generators very quickly, which causes instability.
Short term intermittency
Let’s look at some historical graphs of wind variability at a single location.
As you can see, even within a month there are a few days with almost no wind.
Solar power has a different problem which is that it predictably disappears at night. Here is a chart showing the solar intensity curves in Pittstown, New Jersey. The random dips are due to clouds.
What effect does this have?
The amount of electricity supplied into the grid always needs to be equal to the amount of electricity consumed. It’s much harder to store electricity than it is to store, say, barrels of oil, and so typically, electricity is produced on demand, and consumed immediately.
electricity supply = electricity demand
Think about the interaction between:
Solar PV producers
Other electricity generators (ex. coal fire plants)
When solar power is at its peak during midday, there is an abundance of solar energy, and there is less demand for non-renewable sources of electricity. This causes fossil fuel electricity generators to scale down during midday.
However, the sun will go down right around the time people are getting home from work (and using more electricity). This effect puts stress on other electricity generators to quickly scale up electricity production. This often isn't easy to do. For example, coal fire power plants take hours to fire up and power down.
It also messes with the economics of electricity (kind of)
I want to devote an entire post on the details of energy markets, so I won’t go into too much detail here. But, basically, because fossil fuel plants need to ramp down for most of the day, it only really makes sense to keep the most economical ones around. If you remember from our second post, coal fire plants are able to produce electricity much cheaper than other types of fossil fuel plants, despite the fact that they are the worst CO2 polluters. Therefore, increasing renewable capacity might actually lead to an economic motive to build more coal fire plants as backup sources.
We’ll dive deeper into the operations of modern energy markets, and exactly how electricity is priced in a future post. Energy markets today likely need to be redesigned to accommodate a sustainable future.
California as a case study
This is already happening in places that have significant solar installations. Here’s a graph of demand vs net demand (think of net demand as just electricity that needs to be generated by non-renewable sources) in California.
This phenomena is commonly referred to as the “Duck Curve” (if you plot out the net demand curve over multiple years, it resembles a duck).
This ramp up in demand is being fulfilled by
Importing electricity from out of state (which is likely generated by fossil fuels)
Thermal plants (which are natural gas or coal plants).
This also adds instability in the price of electricity. In most states, electricity is cheaper at night than during the day, but in California, it’s the opposite.
Long Term Intermittency
Wind is caused by differences in temperature in different areas of land. Hot air moves towards colder air, producing wind. Wind speeds also change seasonally, but the direction of change is different depending on the local terrain. Most areas in the US experience less wind during the summer, as compared to the winter, by as much as 50%.
Solar capacity decreases during the winter (obviously, since that’s literally the cause of winter) by as much as 60%.
This is a far more challenging problem than the short term variations, because it requires either:
Developing 3 times as much solar and wind capacity so that even when production is reduced 60%, we still have enough energy production to meet demand. This means for much of the year, we are under utilizing our solar / wind plants. This is not economical since most of these plants would be non operational (and therefore, non profitable) for much of the year.
Develop long term battery solutions that can store massive amounts of energy that is produced during the summer, so it can be distributed during the winter.
Run back-up fossil fuel plants during the winter. This is also not economical since most of these plants would be non operational (and therefore, non profitable) for much of the year.
We did a good job so far building a more sustainable energy system but there are some blind spots.
The biggest of which is the intermittency of solar and wind.
Both solar and wind face short term intermittency challenges (daily / weekly) and long term intermittency challenges (monthly / seasonally).
Before we can really move on, we need to figure out this intermittency stuff. Most solutions being proposed today involve some form of battery. The basic idea is to over-generate electricity when the sun is shining, and the wind is blowing, and store the excess energy somewhere that can be distributed back into the grid when needed. In the next post, I wanted to explore some existing and proposed short term batteries, and evaluate how many we need to build, to meet demand.
If this was interesting to you, and you want to stay in the loop, sign up for the newsletter: