Comment by mtgx | Hacker News Reader

Comment by mtgx | Hacker News Reader

Parent Full thread

> said the cost of its technology may halve within four years, potentially boosting its uptake over lithium-ion units.

Chances are that Li-Ion prices will be reduced by at least as much, if not much more within 4 years, as the whole carmaking industry starts mass-producing EVs, and other huge Tesla, European, and Asian battery factories go online to serve that market of millions of EVs a year.

I think some Chinese battery maker recently said that they expect Li-Ion battery cell prices to fall to $50/kWh by 2025 (I think they are around $130-$150 now). It may actually happen a year or two earlier, at least for the industry leaders, with everyone else following soon after.

Personally, with the rise of all-electric vehicles and solar roofs/home battery systems (for which Li-Ion is much more convenient), I'm quite skeptical that other technologies will be able to beat Li-Ion in being used by power companies. The ecosystem and economies of scale will greatly favor Li-Ion, for now, and whatever new tech the battery makers decide to adopt later on (like solid state batteries).

The nice thing about flow batteries is they have the potential to store energy for long periods while being relatively cost effective. All you need is larger tanks for the electrolyte. With Li-Ion batteries, the number of batteries you need scales with the length of time you need to store energy for. So, at some point you reach a crossover point - if you need to store a month's energy, you're probably not going to do that with Li-Ion, but it's not inconceivable with flow batteries (at least assuming the electrolyte itself is relatively cheap and stable).

Exactly, if flow batteries can convert a solar power plant to a base load plant, that would be the ultimate savior of mankind's energy needs.

Nothing can do that. The amount of energy storage needed is in the Terawatt-Hours. Lithium Ion / Flow Batteries are in the hundreds-of-Megawatt-hours area.

Solar + Wind will get us pretty far however. Wind picks up at night, while Solar is best during the day. Natural Gas Peaker Plants seem to be required in any situation I foresee in the next 20 years, but minimizing the use of peaker plants should be the goal.

Alternatively, we build higher base-load plants like nuclear. That would help a lot.

Not quite true.

If you over-build solar by a factor of 3 (above what you'd expect from an energy-only basis), then you can convert it into a 100% constant output baseload plant using about 50 hours of storage, based on a year of hourly sun data for a single site I found. You don't try to store seasonal energy, you over-size the solar panels so there's significant energy production even on cloudy winter days.

Costs less than you might think, considering the spot price for solar cells is like 10 cents per Watt right now (20 cents for panels).

Works in the continental US. Doesn't work in northern Europe because the winter days are too short (too high latitude).

For a SPV plant to be a base load plant, it would have to store just a days worth of energy.

Say we have a 100 MW plant.

Use 50 MW to charge batteries and 50 MW to supply power during the day. Having a battery capacity of 50 MW * 12 h = 600 MWh, will convert this plant to a 50 MW base load plant.

Tesla's recent 25/50 MWh battery cost 25 Million USD, so a 600 one would cost about $300 Million.

A Terawatt hour is about a years production from 30 Tesla gigafactories. We're going to have that amount of production within a few decades. Probably before small modular nuclear reactors have gotten their first production reactor off the line. Building new nuclear is riskier than other existing technologies, as evidenced in South Carolina, Georgia, Hinckley, and others. Therefore it currently seems that building nuclear is less plausible than building many terawatt hours of storage. SMR may change that, but I have a feeling that current companies are not planning for the grid of tomorrow, and are stuck in the last century when it comes to the grid's needs.

From a grid storage standpoint longer energy storage just means you have fewer cycles to make up your investment. More generation tends to be vastly cheaper then storage in such situations. Some reserve is very useful, but lithium ion batteries should not be fully discharged regularly which allows for deeper discharges at higher cost and thus some reserve capacity naturally follows.

Flow batteries might work to replace backup generators, but the cost would need to fall much further.

In some places you could make money trucking electricity in tankers or shipping it.

> From a grid storage standpoint longer energy storage just means you have fewer cycles to make up your investment.

Only true if you are selling on the spot market.

You can also get paid for grid security - your ability to potentially deliver power during extreme load or deliver power to constrained nodes.

For example, you may be paid to install and maintain a whole power station that is only expected to be used rarely (or ideally not at all if designed for emergency loads!)

Load-shifting is the biggest innovation actually. My utility company is giving out deals to basically control the air-conditioning of some houses. During emergencies, they prevent the air-conditioner from running... either 50% or all the way down to 0% cycling (for lowest cost to utility bills)

It caused a bit of controversy when they used their authority to shut off people's air conditioner, but allegedly it saved the grid of a potential brownout.

Batteries / Energy Storage is the obvious way to solve the issue. But creative solutions (ie: controlling people's air conditioners) is more feasible in the short run, and probably should be deployed more widely.

> From a grid storage standpoint longer energy storage just means you have fewer cycles to make up your investment.

It also means you can wait to charge at the cheapest possible times, and sell on the absolutely most expensive ones.

And since it's cheaper to add further storage, it means that you can do both the faster cycle with low margins and the slower one with higher margins.

More generation tends to be vastly cheaper than storage in the absence of a carbon tax. If carbon emissions and other externalities were fully priced in then that would favor storage.

The daily min vs max power output variation for wind is huge, but the median daily power output over large areas vs the min daily power output is far less so. Even more so when combined with solar which again has a significant minimum output.

If you look at 20% over capacity as adding 20% to the cost of wind and solar it’s still cheaper than coal. But the minimum is now 20% larger reducing the gap very significantly especially over several days.

Hydro naturally provides long term storage, so you can generally over any remaining gaps before looking to multi day battery storage. Plus your dealing with that 20% over capacity so the next day is very likely to be a net gain.

Li-Ion is hard to beat now in places where strong discharge occurs every day, but even a more pricey cell with lower self-discharge rate and longer life might be at the end of the day more economically feasible than Li-Ion in some areas since the guy climbing up 100 sensor poles to swap exhausted cells will likely cost more than the different technology.

When doable I recover working cells from old laptop battery packs, to use them in non critical applications with protected chargers (laptop cells rely on the pack circuitry for protection). Last year I recovered all cells from two new packs that failed to charge because of a failure of the old laptop I bought them for. So I recovered all cells, put them in my charger at a slow 500mA charge (don't use fast charge if not needed, it shortens the cell life) and did a few cycles to test their capacity which was fine. Then I put them in their safety bag and pretty much forgot about them because I was reorganizing my lab. A few weeks back I saw the safety bag and recalled about the cells, so I put them in the charger and tested them: about half of them were dead and all attempts to recover them failed. Of the remaining half some were dangerously close to excessive self discharge but still recoverable and the remaining ones were still fine. In the end all working ones were recovered, but I lost about half of the cells I had; one year of storage albeit in proper environment, no excessive temperature and no load, was enough to kill essentially new cells. Had rechargeable Li-Ion a shelf life comparable to primary Li-Ion cells, it would make them suitable for almost every use, but as of today they sadly aren't even close to that.

>... and solar roofs/home battery systems (for which Li-Ion is much more convenient)...

You may need to substantiate that claim.

If you mean it's easier to find a vendor for lithium batteries for home use, you're probably correct.

In terms of life span and constraints around usage, discharge, and recharge, I'm not convinced that Li-ion is more convenient.

Home-scale battery systems is clearly a bad idea IMO.

Utility-scale battery systems is how you democratize the idea for real. Have the entire city own multiple, giant 200 MW-hr Battery that all the citizens can use. Then use net-metering credits to have Solar-homes charge the communal battery.

Not everyone can afford $20,000+ systems to store a few hours of electricity. The people who need reliable home-electricity get $500 generators with some gasoline or propane fuel (which is sufficient for days of electricity during a disaster).

-------

The city can choose whatever battery best fits its requirements. For the near term, Lithium Ion is available, as well as Pumped Hydro. Flow batteries still seem to be "5 year away" technology, but flow batteries just make a lot of sense at the 200MWhr utility scale.

I don't believe you are, but this line :

> Home-scale battery systems is clearly a bad idea IMO.

.. makes me think you're trolling.

There are plenty of places where people need storage on their side of the meter, and the idea that you can better democratise power supply by having it owned by a large multinational or nation state owned entity sounds ludicrous.

The $20k (I'm assuming that's usd?) for 'a couple of hours power' is just plain disingenuous.

For anyone plonking down $800k for a bespoke abode, spending $30k for gen + storage, with or without grid connection, seems a minor addition to the capex.

How many years can Li reserves last?

Loooong enough. https://minerals.usgs.gov/minerals/pubs/commodity/lithium/

Reply on news.ycombinator.com