Hacker News Reader

Comment by rayiner | Hacker News Reader

Your first paragraph misses the mark, because, unlike a nuclear plant, a windmill or solar plant isn't a standalone power source. To have reliable baseload power with renewables, you need massive battery storage. Right now, you don't factor in the (very high) cost of building battery storage into the cost of wind power because you still have all those coal, gas, and nuclear plants providing baseload power. But the cost of all that baseload capacity that's increasingly sitting around not being used (but has to be there to ensure grid stability) can't be ignored.

View on HN

>because you still have all those coal, gas, and nuclear plants providing baseload power...

That's right. So why would you not decommission all of that incredibly expensive nuclear, and draw down your use of the coal plants to fill in the gaps? That's how most dispatch stacks work. It's just common sense. Consumers want to save money. Utilities want to make enough money.

There's just no way you use nuclear in any scenario you can come up with, unless the government is paying for it. Then you don't care, because you're not paying the costs.

Coal can't dispatch fast enough, and most coal plants are at the end of their lifecycle and need to be replaced. So you have to build new baseload capacity, usually gas. (And really, you still need a bunch of batteries to handle short-term fluctuations in renewables output.) The problem, then, is that you have all new coal or gas plants, which are expensive capital assets, that you're basically using for standby power. (And they incur operating and maintenance costs even when not running.) You can't say that electricity from renewables is cheap without accounting for the cost of that standby capacity renewables require. A nuclear plant, by contrast, doesn't need a bunch of coal or gas plants sitting around on standby. You can shut down that old coal plant, stop paying the workers, reuse the land, etc.

Baseload is not the necessity, the necessity is to match the power to the load.

Up until now, baseload has been a convenient way to do that matching because baseload was usually the cheapest source of power.

Once baseload is no longer the cheapest source, it's time to re-evaluate that model. Also, when we have a global communication grid and lots of flexibility in our load schedule, it's also fine to re-evaluate how we are doing pricing for electricity.

That time is now. Utilities are highly regulated, and both utility and regulator are slow to adapt to the quickly changing technology that they're now confronting, after nearly a century of glacial technology change. But they will adapt. As we must not only change our economic mode but also become carbon neutral.

Those with huge capital investments may resist, but as you say, we can repurpose that old coal plant and reuse its connections to the grid, which are valuable and expensive to recreate. In Moss Landing, California, gas turbines are getting replaced by over a GWh of lithium ion battery. The change will happen as soon as utilities start bringing recent pricing into their planning process.

>So you have to build new baseload capacity...

No, you don't. The coal plant in another Alliant Energy area has been around since roughly 1900? or so. It's been updated several times. Never was there a need to rebuild from scratch any coal plant. (This is Madison Wi btw.) Now we stopped using coal there in 2011, (all gas now). But the point is, all the renovations are much cheaper than building a nuclear plant. Utilities around here, (the midwest), are run by old, stodgy, conservative guys who are generally not prone to rash action. There are very few executives around here who are going to build entirely new plants because the machinery in the old one has reached its end of service life. They are going to replace the machinery, at a fraction of the cost. I don't think executives in other parts of the nation are all that different in this regard. Now all that said, even if they did completely tear down and rebuild their plants, which they wouldn't, but even if they did, it would still be cheaper than building a nuclear plant.

So the coal plants are fine. And as I mentioned elsewhere, whatever you use to fill gaps will be a transitional technology. Wind turbines will become more efficient. (Work at lower windspeeds.) Pumped hydro storage will be built in new and innovative ways. What will fill the gaps in 2069 will bear little resemblance to what is filling the gaps in 2019. Because the new methods will likely be not only more clean, but much cheaper to boot.

And incidentally,

>You can't say that electricity from renewables is cheap without accounting for the cost of that standby capacity renewables require...

the 2 cents per kWh people see on their bill does factor in every source in the dispatch stack.

Just to back up your point a bit:

https://www.lazard.com/perspective/levelized-cost-of-energy-...

Running existing coal and nuclear generation appears to be cheap (2nd chart), but new generation appears to be much more expensive (1st chart). The cost of Solar + Storage is low and dropping, and already beats the cost of new nuclear plants (4th chart).

BTW, I grew up on the east side of Madison, Cheers.

Renewables and storage are already beating natural gas in that gap capacity. As both continue to decrease costs, that niche use will grow. Natural gas and coal tech are cost mature, while both renewables and storage are following natural tech development S-curves down and still have a long way to run.

>And really, you still need a bunch of batteries to handle short-term fluctuations in renewables output.

Not really the case. Renewable output does not fluctuate fast enough to require the fast response of batteries.

Fast responding storage, in the case of the UK pumped hydro, is usually needed due to thermal plants tripping and causing the grid frequency to suddenly nosedive.

Power plant failure also generally happens due to extreme environmental effects. During those events renewable energies often reach their maximum production capacity and therefore actually contribute to grid stability.

> So you have to build new baseload capacity, usually gas.

Since when was natural gas ever good for base load? It's almost always used peak load because it spins up quickly.

The promise of new nuclear plant designs is to cut down construction cost. This is pretty exciting. I think nuclear power will make a come back in popularity in the next few years, as more and more news and promises surface about those designs.

Here's an article shared by Bill Gates last week: https://www.nytimes.com/2019/04/06/opinion/sunday/climate-ch...

Cut down construction costs still won't solve fundamental problems like final waste storage [0] or a lack of proper epidemiological studies into the effects on people, particularly children and pregnant women [1], living near those plants.

Imho the way we've handled nuclear fission so far is far too reminiscent to other environmental disasters that took humanity decades to even recognize and finally act on, even when we had plenty of warnings from the very beginning [2].

[0] http://sitn.hms.harvard.edu/flash/2018/looking-trash-can-nuc...

[1] https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2757021/

[2] https://en.wikipedia.org/wiki/Tetraethyllead#Controversy_and...

> To have reliable baseload power with renewables

You have to be careful baseload power isn't the same thing as battery backed (time shifted) renewable power.

Coal and Nuclear baseload power exists for two reasons. Excess generating capacity at night. And the inability of coal and nuclear plants to shutdown. It's cheap because of oversupply. Time shifted renewable's is more expensive because of under supply.

Some percentage of base load power consumers are only in it for the balance sheet economics. If the pricing structure changes they'll shift their usage to the cheapest source.

Take away: Renewables don't need to generate as much baseload as coal and nuclear plants do now.

You also need base load during the day when wind is blowing and it all of a sudden disappears. Don't say this doesn't happen as I've seen it many times.

Think about it. Wind might be 30% or more of your dispatch and then it just disappears quickly (even with multiple forecast vendors you are lucky to have a 2 hour warning). At this point you have to either have a lot of headroom (excess thermal generation for times like this) or you hope you can startup a resource in time. Therefore you either need storage on a massive unheard of scale, very fast starting resources, or lots of thermal generation for those times of trouble. Another option would be drastically increasing the demand response in the region (basically shut all the A/C's off at a few universities for an hour or so).

You have seen wind disappear from an entire continent during a very cloudy day?

In the USA a very large geographical subsection can lose many GW of instantaneous wind generation across an hour, which can be concerning if the system is currently running pretty economic (not a whole lot of excess reserves).

The way dispatch is done is completely different than the way it was done 10 years ago and also completely different than the way it was designed in so many ways.

We can't really think of the electrical grid as a continental-scale system because it is severely constrained by transmission resources. It as to be modeled as, at best, regional networks with weak interconnection between them.

Even within regions it isn't a well-connected grid.

To give a specific quantitative example within Texas the generation network is modeled as several dozen individual markets (the "nodal market" concept) with generation and load grouped together to form nodes, which are mainly separated by key transmission congestion points.

Lack of transmission capacity sometimes leads to strange artifacts in pricing wind power in Texas, such as negative pricing (due to the Production Tax Credit lowering the floor below zero). These artifacts tend to go away as new transmission capacity comes online and paying loads in more distant markets become reachable.

Additional transmission lowers congestion costs, but you actually increase the "wind artifact" you referred to as you have even more wind online (depending on where the transmission is placed) bidding in below zero as long as the PTC is still in effect. I don't think any new thermal generation has come online in Texas despite their extremely high shortage pricing, so you're really just putting more wind online with new transmission build outs.

It's not really baseload power (which runs 24-7) that's needed. It's backup power that runs much less often. That has different economics.

Reply on news.ycombinator.com