Weather-Smart Electric Grids Are Needed for Wind and Solar Power to Surge | Hacker News Reader

Weather-Smart Electric Grids Are Needed for Wind and Solar Power to Surge

scientificamerican.com

46 points0xbxd46 comments

OP is hyperbolic framing of a genuine issue. Pacific Northwest is one of the more extreme cases, due to the confluence of hydropower seasonality with renewables. In general, we will address the variability of wind and solar by building more continent-scale grids, with lots of storage, with overprovisioning of solar PV (which is now so cheap it pencils out even with occasional curtailment in many cases), with load shifting, and for a decade or two by running natural gas speakers in extremis.

Read this article and notice what a key role storage plays:

https://www.vox.com/energy-and-environment/2018/7/13/1755187...

Then recall that at least three major automobile companies are making $1B-scale investments to make solid-state batteries cheaper and higher performing than the Li-ion technology that is cost-effective at grid scale today. Articles like OP that cherry-pick atypical cases and ignore cost curves take a real issue and blow it up into a crisis, which it is not.

Continent-scale grids of the scale we need look very unlikely to happen due to regulatory / NIMBY concerns - see the decade long delays for Clean Line Energy to get anything moving. Maybe we'll get a federal administration bold enough to push it through but I wouldn't count on it. That means we'll have to vastly overbuild solar + storage as you said. Luckily the current cost trajectory makes that look possible, but seems worthwhile to have demand side planning in the mix to balance the overbuilding.

Every article on this topic has the same presumption: that we'll continue to have a power grid. But the grid kinda sucks.

Having lived for 2-years on a school bus with an off-grid solar system, I don't see the advantage of investment in the grid.

I question whether, in 2018 and beyond, the advantages of the grid (sending power where there is none, borrowing power from where there is some) are better than very small batches of battery storage.

How does your gridless plan work for dense urban areas? That need more power than they have rooftops, and the wind blows best elsewhere.

I think it's already the case that not being on the grid is financially best in areas with very low population and reliable sunshine, like parts of the Australian outback.

Yeah, I mean dense urban areas will need to have some unification infrastructure and to be able to get power from areas just outside the city. But I don't think that this means that we need a single universal grid across a landmass as huge as North America.

Also, cities (and all people everywhere) need to dramatically reduce power consumption no matter what the source is.

Residential may very well work best off grid, or better still with highly localised combined heat and power. There's probably still bumps that are better smoothed at grid scale than everyone having peak battery coverage.

Industrial use changes things. Take the grid away and there'll be some unpleasant knock-on effects to prices for everything.

Large industrial users are already smoothing the grid, either with cheaper pricing for off peak use, and deals for intermittently taking surge from the grid. With no grid and no daily varying residential and office loads how do you smelt your aluminium or cook your canned foods other than at normal price? Does every plant need to be local to power generation now?

Electric car charging can smooth the grid, similar to industrial users. If big battery cars (that only need to be recharged every few days) become popular, and most people can plug them in both day and night, and you have a forecast of when the car user might need to actually be fully charged, then you can charge-or-not depending on the weather.

Which part of the country are you in? In the northern parts of the country the supply / demand imbalance between summer solar gen and winter heating requirements are huge, either you're producing 10x more solar than you need in summer or supplementing heat in the winter - or you have an extremely well insulated school bus!

That, or we switch to a safer nuclear. Turns out nuclear is the only CLEAN baseload energy provider. The rest is all intermittent.

Fun fact: Nuclear (in the US) is 50 times safer than hydro-electric (in the US) and 4400 times safer than rooftop solar (global) as measured by deaths per watt-hour.

https://en.wikipedia.org/wiki/Energy_accidents#Fatalities

Say what you will about Chernobyl and Fukushima, but if LFTRs can provide a better and safer nuclear? Then it's about time environmentalists started pulling their heads from their asses and reconsider nuclear before we irreversibly fuck up our entire planet and ecosystem.

"Is nuclear power going to help the United States decarbonize its energy supply and fight climate change?

Probably not.

That is the conclusion of a remarkable new study published in the Proceedings of the National Academy of Sciences in early July — remarkable because it is not written by opponents of nuclear power, as one might expect given the conclusion. The authors are in fact extremely supportive of nuclear and view its loss as a matter of “profound concern”"

https://www.vox.com/energy-and-environment/2018/7/11/1755564...

But it's worth noting that the study focused on the U.S., and much of the problem with advanced nuclear in the U.S. comes from the NRC's blinkered approach to regulating new designs.

A couple years ago I got to sit in a meeting between a former head of the NRC, and representatives from over a dozen advanced reactor startups. The main complaint of the startup people was that the NRC requires them to spend several hundred million dollars on a detailed design before the NRC will even look at their proposal. Then the NRC gives a straight yes or no. If "no" then the project is over; if "yes" then they still have nothing but a paper reactor.

It's a very difficult environment for investors. The startups said just going with a more phased approach would be a huge help.

By contrast, Canada has regulators much more friendly to new technology, without compromising safety. Terrestrial Energy, a molten salt reactor startup in Canada, has spoken highly of their regulators, and thinks they can have reactors on the grid by the mid-2020s.

Most of the grid will be hydro, solar, wind, and utility scale battery storage before Terrestrial Enegry has their first reactor operational.

I don't think we'll make that much progress by 2025, though I'd love to be wrong about that.

Countries are meeting their renewables targets years ahead of schedule, and the cost of wind, solar, and batteries will continue to decline. I don’t see any way we’re still using fossil fuels for electricity by 2030. It’s not optimism, just curve fitting.

Except batteries aren't even close to where they need to be in costs to cover windless nights.

They will be in the next 5 years.

Maybe. And maybe Terrestrial Energy really will deliver a molten salt reactor delivering nuclear at 3 cents/kWh in 2025.

Renewables advocates have a bad habit of comparing the renewables of the future to the nuclear of the past.

Nuclear advocates have a bad habit of never delivering on time and/or without massive cost overruns. Renewables (solar and wind) are already below 2 cents/kwh at utility scale. They, theoretically, should be considerably cheaper in 5 years (and so, nuclear must target 1 cent/kwh to be profitable).

Solar/wind with enough storage to be reliable without fossil backup is nowhere near that cheap.

It doesn’t have to be storage backed at every hour if the day. ComEd in Northern Illinois provides me nuclear at 1 cent per kWh from midnight to 5am because demand is so low in the middle of the night. You can overbuild renewables, and ramp battery production to arrive at a cost below nuclear, coal, and natural gas (and even used load shedding and coordinated storage orchestration using electric vehicles).

Maybe I’m just optimistic based on historical cost decline curves, but I’d bet $1000 on my thread thesis (renewables replace all fossil generation by 2030). It’s the economics.

Stop confusing batteries with wind/solar prices. Until the battery problem is fixed, overbuilding doesn't do jack because you have nowhere to store the bursts.

Stop talking about wind and solar prices and just look at the price of batteries. They aren't on the trajectory you are projecting. It's hard enough to get them cost competitive for a single house, let alone industries with massive 24/7 consumption (e.g. electric mine shovels).

"A study from McKinsey & Co. last summer found the continued energy storage cost declines present a growing threat of disruption for utility business models. Recent bids for utility-scale solar-plus-storage projects may also give a glimpse where the industry is heading.

An Xcel Energy resource solicitation received more than 400 individual proposals last month, including what may be record-low prices for renewable energy paired with energy storage. The median price bid for wind-plus-storage projects in Xcel's all-source solicitation was $21/MWh, and the median bid for solar-plus-storage was $36/MWh. Previously, the lowest known bid for similar solar resources was $45/MWh in Arizona." [1]

(By comparison, the lowest cost conventional technologies were gas combined cycle technologies, averaging $74 per megawatt-hour, and coal plants, averaging $109) [2]

"An acceleration of the 70% cost reduction seen in lithium-ion prices since 2012 will drive global uptake of energy storage over the next few years, finds a new report by IHS Markit.

The analysts forecast prices for lithium-ion battery modules to tumble below $200/kWh by 2019, enabling previously “uneconomical applications” such as the colocation of battery storage and solar PV to surge.

This sharp and sustained cost reduction will help cement lithium-ion as the battery chemistry of choice in all energy storage markets, including grid-scale behind-the-meter storage, residential storage, and micro-grids.

By 2025, the world’s base of cumulative installed storage capacity will reach 52 GW, IHS Markit says, up from around 4 GW today. Last year, 1.3 GW of grid-connected storage was deployed globally, and this rate is poised to accelerate to 4.7 GW a year by 2020, and 8.8 GW annually by 2025." [3]

[1] https://www.utilitydive.com/news/eia-700-mw-of-utility-scale...

[2] http://energyinnovation.org/2015/02/07/levelized-cost-of-ene...

[3] https://www.pv-magazine.com/2017/08/03/lithium-ion-batteries...

What are your thoughts on mofular factory-produced reactors like NuScale [1] ? They use proven tecnology, and walkaway safe.

[1] https://www.nuscalepower.com

I think they’re an excellent solution where renewables aren’t feasible (the arctic circle, for example).

Like most climate hawks, I am a passionate proponent of maintaining existing nuclear and one who hopes that new nuclear will reach the market. But the operative words in your post are "thinks" and "mid-2020s". Every ton of CO2 we put out now warms the earth now and for centuries. Please read this [1] and see that renewables plus storage are doing now what you hope nuclear will do years from now.

[1] https://www.vox.com/energy-and-environment/2018/7/13/1755187...

I completely agree that we should keep rolling out renewables as fast as we can.

But we should also fix the unnecessarily difficult regulatory environment for advanced nuclear, because so far, we don't have a single example of a country running almost entirely on wind/solar/storage, and current storage technology isn't scalable or cheap enough to get large populations through windless nights.

You can say that we'll get there, but that's no less speculative than Terrestrial Energy saying when their reactors will be ready.

Costa Rica, Iceland, and Uruguay run almost entirely on renewables. Other countries are catching up quickly.

I said wind and solar, not "renewables." Countries with large hydro and geothermal resources can easily run on those, because they're always available.

We really need two different words: one for hydro/geothermal which are steady supplies but geographically limited, and another for wind/solar, which can be installed anywhere but are unpredictable and require massive storage or fossil/nuclear backup. The use of the same word for both has been a constant source of misleading rhetoric.

Iceland is a tiny country with lots of geothermal energy. Comparing that directly to the US, makes zero sense.

"[1] we should keep rolling out renewables as fast as we can. But [2] we should also fix the unnecessarily difficult regulatory environment for advanced nuclear"

[1]: sit on hands and watch existing markets follow their current trajectories. Odds of success: excellent. [2]: make a major change in the existing political economy of energy, beating all the entrenched interests who will oppose it. Odds of success: ___________.

You don't answer the parent.

Parent: "[...] but if LFTRs can provide a better and safer nuclear? Then it's about time environmentalists started pulling their heads from their asses and reconsider nuclear before we irreversibly fuck up our entire planet and ecosystem."

Your article: Nuclear isn't economically/politically likely.

Well, if the safest and greenest form of electricity was also the cheapest and most politically friendly alternative then we wouldn't be having this discussion.

If the environment was prioritized enough then neither would pose any obstacle, what remains is the debate of whether nuclear really is the greenest and safest form of energy.

I linked to a write-up of peer-reviewed research which focused on the "if" and demonstrated that the odds are daunting. That ... is an answer. I linked to a second article showing that renewables are reducing emissions from natural gas today. "If the environment was prioritized enough" then either would work. Renewables are making headway given the current political economy of energy, nuclear is not. And to repeat: I wish it were!

Odds are daunting that it is a green and safe source of energy?

That new nuclear will be deployed at scale in the next two decades that wind and solar are being deployed at now.

Too late: the LFTR is still vapourware, although not as much as fusion, and the existing technology of reactors already takes ages to build and is prone to cost and time overruns. We need to keep the renewables buildout going while we wait for someone to produce a working non-pilot commercial LFTR.

Why even the focus on LFTR, what is wrong with proven technology in modular form? Just dont build reactors gigantic, make them walkaway-safe instead - [1], they are quite resilient - [2] and can be produced on a factory, so inspections are easier and costs are lower.

[1] https://www.nuscalepower.com/technology/design-innovations [2] https://www.nuscalepower.com/-/media/Nuscale/Files/Technolog...

The "proven" technology was proven not to be reliable enough, to be too expensive for use, and to create some large problems down the line.

Besides, if you change the size, it's not "proven" anymore.

This new smaller design solves reliability and cost issues of older bigger designs.

France and the US navy both rely on PWRs. France built a fleet to get the benefits of scale. Nuclear works.

Modular would be great, that's a big part of why wind, solar, and batteries are doing so well these days.

Modular nuclear power isn't proven to reduce waste storage costs, for one thing. Look at how long it took renewables to go from a theory to being a mature technology with a declining cost curve and full lifecycle experiences that make predicting cost of ownership very reliable.

It would be nice if there was a zero-carbon source of concentrated power that had a track record of being cost effective and not having nasty end-of-life surprises. This is why nuclear is seductive. But it has never lived up to expectations.

> Turns out nuclear is the only CLEAN baseload energy provider.

Thing is in the long run we don't even need baseload power. Renewables have become so cheap now that power plants need to optimize for flexibility rather than efficiency to stay profitable[0].

[0] https://www.ge.com/power/transform/article.transform.article...

EDIT: emphasis.

Power plants right now have to optimize for working around varying wind/solar supply, but if we tried to run civilization on wind/solar alone, we'd rediscover the need for good baseload supply when the wind stops blowing at night.

Actually no, because this function is already being taken over by gas peaker plants. It's less expensive to have such a plant and fire it once in a while than to keep a large nuclear/coal plant churning.

Both of us were talking about the long run, and what I said was "wind/solar alone." Of course we can run on wind/solar/gas but that doesn't decarbonize the grid.

If solar had the same record of construction delay, cost overruns, and exorbitant hidden costs in waste storage and decommissioning, nobody would start a solar project. Instead, solar projects fall in cost and exceed their lifecycle expectations.

If you can't demonstrate a full life cycle that happens on time and on budget, and without risks that carry costs that wipe out all the value delivered by nuclear, nuclear power is theoretical.

Reply on news.ycombinator.com