A year without hydrocarbons

So what would the Texas ERCOT grid look like if we didn't have natural gas and coal?

While some people are calling for a complete abandonment of fossil fuels, that is not something that can happen rapidly (see post on extreme energy policy), but we also have to face facts that natural gas won’t last forever. There are people saying we have massive natural gas reserves and using that for the argument that we don’t need renewables, but the problem is they are using reserve numbers which are actually very unlikely to happen (see post on natural gas reserves). My view is we will begin to see natural gas supply issues in the next ten years. So, is it time to start talking about what the future grid will look like?

So this is the present ERCOT grid for November 2022 through October 2023. We had two freezes, one in December and one at the end of January/beginning of February. We had our summer peaks in July and August, that even extended into September. Wind (green line) shows a lot of variability, as does Solar (yellow line). Consumer demand (black line) also changes a lot from day to day and it varies within a single day. Day time usage can be 30,000 megawatts more than night time usage.

When you look at a one year snapshot of ERCOT, you see a lot fluctuation in consumer demand and renewables (wind & solar). All that volatility is corrected by natural gas power plants reacting. But in a future where natural gas is not available, what does that look like? If we pull out natural gas (maroon line) and coal (grey line), what will fill the ever changing gap between renewables and consumer demand? What are we going to replace fossil fuels with? And here is my best guess at a solution.

For the new non-hydrocarbon grid, we didn’t increase wind by much, only adding 34% to wind to reflect the amount in construction and planning at this time. While wind is the cheapest source of electricity, it’s high variability and not being in sync with consumer demand means that adding more wind makes the grid harder to balance. So adding a lot more isn’t a good idea.

For, solar, we added a lot, increasing it by 350%. The reason is that solar is almost as cheap as wind and makes the most electricity in the summer months and during the time when consumer demand is higher. In December solar makes less than a third of what it makes in July and August, so that helps with the big change in consumer demand from winter months to summer months.

But, you need something that can supply a base-load generation and is adjustable, and for that we increased nuclear. ERCOT presently has about 5,000 megawatts of nuclear, for this simulation we increased it to 30,000.

We also added daily energy storage to move excess solar power from the day into the nighttime hours. And we added a seasonal energy storage to take excess electricity in plentiful months and store it for winter freezes and for the energy intensive summers that Texas experiences. The result was a system that worked.

Now, this used consumer demand from the past twelve months and took the performance of wind and solar and increased it by the percentages above. It doesn’t reflect load growth from more electric vehicles or from population growth. There is a whole lot of optimization that could be done with this design, but the whole point of this was to see if we could redesign the grid to meet consumer demand without wasting a lot of electricity and over building.

Let’s consider two other scenarios - all renewables and all nuclear.

For the all renewables scenario you need to massively over build on wind and solar and on seasonal storage. During the winter freezes, seasonal storage is supporting the grids as wind and solar are producing nothing during the cloudy calm days that are the heart of each winter freeze. In the nuclear and renewable scenario, electricity is supplied by seasonal storage and nuclear ramping up to full power. Remove nuclear and all that electricity supply falls to seasonal storage.

In the all nuclear scenario, we assume you keep the existing wind and solar, but add nuclear and storage. For this you need to increase from 5,000 megawatts to 49,000 megawatts of nuclear and that is if you still build a large seasonal storage system to get you through the winter freezes and the worst of summer. Don’t build the seasonal storage system and you are looking at needing more than 65,000 MW of nuclear at a cost higher than than of building the seasonal storage.

The challenge is that to build one megawatt of nuclear capacity is about $7.5 million based on numbers from the Department of Energy. Wind is about $1.7 million per MW and solar is $1.3 million per MW. And while you need to build a lot more wind and solar to get the equivalent megawatt-hours that nuclear delivers due to the volatility of wind and solar, they are still cheaper. As a result the all nuclear scenario is the most expensive.

But coming in second is the all renewable scenario. Needing a lot more wind, solar, and energy storage causes this to be more expensive than the nuclear/renewable scenario, but slightly cheaper than the all nuclear scenario.

And all three scenarios require seasonal storage. This seasonal storage could be hydrogen, compressed air energy storage (CAES), or iron-air batteries like those made by Form Energy. They can store over 100 hours of electricity with medium efficiencies (35-50%). The reality is that the efficiency isn’t critical for seasonal storage, it is the depth of storage that the unit can provide per megawatt. In the nuclear/renewables scenario above it is supplying 13.75 million megawatt-hours(MWh) through the year to fill in peak demand days. But with a medium efficiency that comes at a price of needing 30.5 million MWh to charge during days of excess electricity. That means that 16.8 million MWh is used up in the charging, but that is only 3.8% of the total yearly electricity generated. Don’t build the seasonal storage, and you need a lot more nuclear to get through the winter freezes and heat of summer and that is going to cost more than the seasonal storage. For the renewable scenario, it isn’t even possible to operate the grid without a seasonal storage.

So, if you have the seasonal storage, why do you need the daily energy storage? Daily storage is based on technologies similar to pumped storage hydroelectric and has a much higher efficiency (75% to 80%), but only stores ten to twenty hours of electricity. That higher efficiency comes at a price as these tend to be more expensive per megawatt than seasonal storage. But in the nuclear/renewables scenario above it is moving 31 million MWh. Taking excess solar during the day and shifting that electricity to the night when it is needed. And it only requires 38.8 million MWh to charge up, meaning it is only wasting 7.8 million MWh. Daily storage is moving two-and-half times the power that seasonal storage is moving at less than have the cost in electricity. If you did that same day-to-night shifting with the seasonal storage it would costs you an additional 30 million MWh due to seasonal storage’s lower efficiency. That would require you to add power plants to make that additional 30 million MWh and would waste more electricity over twenty years than it would cost to build the daily storage in the first place. That is why you need daily storage. It is your workhorse that is moving electricity in a 24 hour cycle making the day to night transition work with solar. Whereas seasonal storage is moving electricity between weeks and months to make the overall year work.

So, you can change the ERCOT grid so that it works without hydrocarbons. But, it’s going to take building nuclear, seasonal storage and daily storage, which at the moment none are in construction within ERCOT.

And that is the challenge. If domestic natural gas production will begin to decline in the next ten years, than the Texas Legislature needs to stop encouraging more natural gas power plants and start encouraging daily storage, seasonal storage, and nuclear. The reason is none of these new solutions can be built overnight and if we will be needing them in 10-20 years, we really need to be thinking about them now.