Hacker News Reader

Comment by coryrc | Hacker News Reader

Grid stability/storage, plus sunk costs.

Say it's a cold cloudy day with no wind and it's been so for a week. You'll need the capacity of that coal/nuclear plant those days. And let's say every other day of the year solar and wind make enough power. So who pays the extremely high fixed cost of the nuclear and coal plant so it's available that one week you need it? No one wants to, so they're shutting them down. The power company could still be paying the bond for the plant, but ultimately that's the consumer's money. Then, surprise, you end up with blackouts or you pay for new natural gas plants to be built. Prices and fragility go up.

It's likely cheaper to make demand more flexible, but that has a cost too.

View on HN

Grid scale batteries have become cost effective in the last two years.

It'll take time to build out capacity but it'll take more time AND cost to build new nuclear plants.

Those batteries only have a few hours of storage at max power, they're for solving the duck curve not a week-long supply shortage (let alone seasonal fluctuations). If we got the Midwest on heat pumps, how you planning to keep places heated when there's a week of blizzard? Right now we have no plan for addressing the occasional bad week besides overbuilding natural gas. That's worse than keeping our nuclear running and expanding, but the solar and wind is subsidized but not the natural gas. The way the market for power works, bidding on a daily or hourly basis, doesn't account for seasonal variation. The solar and wind builders are off the hook on blizzard week but get subsidized enough to run the reliable plants out of business the rest of the year.

>how you planning to keep places heated when there's a week of blizzard?

When the wind is blowing really hard? Wind turbines?

The texas winter was a bit of an anomaly. It got so cold that every power source ending up failing because none were set up to deal with it even though all of them in theory could have (& would have if they were in, say, alaska).

I think their point still stands - even if every town/city has both solar panels and wind turbines, there are times when you won't get enough power from them to meet demand (remember demand can vary too), just due to natural fluctuations in weather. You either need an absurd amount of battery storage, or some sort of backup power source.

I think if you combine demand shaping, overbuilding, grid scale LIon batteries, better grid interconnects (another thing texas failed miserably at) & pumped hydro you'll find that demand can be met for most people.

It'll take time to shift to using green energy and there will be problems but the same would be true for any other kind of shift to any new sources of power.

Probably not 16 years like the new UK nuclear plant will take though :/

Yes, we could do that, but that's not what is incentivized by the market so it's not being done on a large scale or with any urgency. Changing the way the market works is what is needed to make those things the action chosen.

This is something people get confused about. Grid scale batteries are definitely not cost effective or in existence. The media does their best to obfuscate this. What has been built around the places are battery packs to absorb dangerous power fluctuations caused by windmills. They destroyed the grid in South Australia a few years ago, and Elon Musk famously went in and installed batteries (maybe that is what you are thinking of).

If you don't believe me go back and have a look at how big the batteries were. The 500 million dollar project installed 130 MWh worth of electricity storage. That is enough to produce the same energy as a power station for 10 minutes before being fully drained.

> Grid scale batteries are definitely not cost effective or in existence.

California has more than 1.5 GW of utility-scale chemical batteries interconnected to it's grid with another 1 GW anticipated to enter service by the end of the year. Most of these batteries can sustain full output for four hours. This is in addition to the more than 3.0 GW of pumped hydro "batteries."

1 GW is roughly the output of one nuclear power plant.

Source: I develop power plants in the western US.

1GW is roughly the output of one nuclear reactor (most plants have 2 or more) for one hour.

We'll need at least half a day's worth of storage, likely much more if we want to use wind for a significant portion of power generation.

I remember people saying something similar about the amount of deployed solar and wind shortly after it became cost competitive. There used to be a tiny amount deployed but it was the cheapest form of power.

Then over the course of a few years there was a lot.

All that really matters is A) is it cost effective B) will there be scaling problems C) will there be political impediments.

Solar is not cost-effective today for applying power after the sun sets (duck curve) or at night. Wind is not cost-effective at supplying power on calm days. If they were required to build the rest of the system to be reliable (longer interconnects, batteries, etc) it suddenly wouldn't be cost-effective, this explaining why energy cost continues to go up while this supposedly cheap power ramps up.

Combined solar+battery is not just cost effective at dealing with the duck curve it is downright cheap. Hawaii canceled an inter-island grid interconnect because of this.

If you develop power plants then how come you don't know the difference between energy and power?

I'm thinking of battery backed solar farm in Hawaii that cost around $70/MwH and supplied about 1/5th of the power for the island. There's a few new projects in California/Nevada/Colorado that are allegedly costing around $30-40MwH (although I'm skeptical of how much storage they have for that price).

Either way, conservatively speaking, $50-60/MwH for battery backed wind/solar on the mainland seems fairly doable. Prices will likely keep dropping too.

By contrast, in the UK Hinkley Point C is being paid $124/MwH guaranteed for 20 years. It doesn't exist because it's a cost effective way of supplying green energy that's for sure, but it does bolster the nuclear economy. The military do like to hire nuclear technicians and buy parts for PWRs in nuclear subs, etc. and somebody has to pay for these things. Rate payers, I guess.

These numbers are way off though, maybe those are just the cells? If we look at the 2021 estimates by the National Renewable Energy Labs for utility scale 4h batteries [0] We get cost estimates of 250 $/kWh for the high case, 200 $/kWh for the middle case and 150 $/kWh for the low case by 2030. [0] https://www.nrel.gov/docs/fy21osti/79236.pdf.

That seems to be the cost of the physical batteries, not the price the grid would pay the farm for.

That makes sense I guess.

Assume enough energy is supplied to charge the batteries. So lets say we want 1kWh during 4 hours, that puts our capital cost at 4 kWh*200$/kWh = 800$. Now we assume this system runs every day for 4 hours during ten years. Ten years sounds reasonable, but it could be less. 7 years maybe? So about 1kWh*365,25*10=14610kWh in total. That gives us 800$/14610kWh = 0,0548$/kWh or 54,8 $/MWh in 2030 if I didn't make a major mistake somewhere and assuming no efficiency losses. That's pretty good.

Your link says a lifetime of 15 years. Capacity would degrade over time though.

>Say it's a cold cloudy day with no wind and it's been so for a week.

What if it's been so for a month?

What if the country is Sweden?

What if the month is January?

What if the average temperature is -20C?

>So who pays the extremely high fixed cost of the nuclear and coal plant so it's available that one week you need it? No one wants to, so they're shutting them down.

Those who don't want to freeze, they want to pay.

Change the way the market works, pick a baseline power and bid it yearly. Then the nuclear stations can compete. But our current one results in shutdowns while we pay for extraneous wind capacity and new natural gas.

Reply on news.ycombinator.com