Hacker News Reader

Comment by aresant | Hacker News Reader

This is all very encouraging and in particular batteries role in solar

Two interesting data points to that end

1) The "duck curve" for CA is almost neutral - eg the timing imbalance between peak demand and solar power generation - battery utilization is the most straightforward solution here - https://twitter.com/baker_edmund/status/1750644294673748366

2) There has been a massive decline in rooftop solar applications in CA since solar energy reimbursements dropped - https://twitter.com/thomasopeters/status/1750920941868347539 - some of that is potentially pent up demand, but I think illustrates the role state policy has to play in moving towards "renewables"

View on HN

It's super cool you can watch California's grid level batteries "breathe" every day here, https://www.caiso.com/TodaysOutlook/Pages/supply.html#sectio...

Yesterday we peaked out at 3GW discharging rate, and 4GW charging rate. We are plowing ahead in the transition to utilizing all of our excess solar! We peak at 25GW expected today, so we have a little ways to go but it's incredible how far and how fast they're replacing everything. Clean air FTW! Thanks sun!

the CAISO website is super rad in general -- so many real-time charts!! demand trend is really cool to watch during heat waves; things get spiky on both the demand side (for obvious reason) and the supply side (peaker plants & "virtual" power plants coming online)

Kind of unrelated, but I'm wondering why TX uses so much more electricity than CA. Right now (afternoon of 1/26) happens to be a good time to compare:

-- Temperatures are mild throughout most of TX (50s and 60s F); temps in CA are similar, perhaps a bit warmer;

-- It's roughly mid-day in both places (3PM TX, 1PM CA);

-- TX has a population of ~30M, CA has ~39M

..yet somehow right now TX is consuming ~47GW (per ercot) while CA is only consuming ~23.5GW (per caiso). What gives?

ERCOT: https://www.ercot.com/gridmktinfo/dashboards

There are a few reasons.

1. TX has more heavy industry than CA.

2. CA has spent decades and billions on energy efficiency improvements.

3. CA prices are higher, which encourages lower demand, and encourages investments in efficiency. TX has lower prices which encourages more demand.

California takes a much more european approach to energy efficiency.

It's a mix of a bunch of things. CA is more heating day dominated than TX, and as a result of that, more households use natural gas for heating, whereas you're more likely to see supplemental electric heat in TX than, particularly, NorCal.

(2) high energy costs lead to more investment in electrical efficiency in CA;

(3) high energy costs mean that if you're running an aluminum smelter (for example), you don't run it in CA. Or building a new mega data center. so there are fewer electricity-intensive industrial facilities in CA than there would otherwise be.

Natural gas is a by-product of much oil extraction, and commonly used for heating, cooking, etc.... Propane is also commonly used, in more rural locations.

But what Texas has is much higher requirements for air conditioning, which can't be easily gas-powered at the level of individual houses. Thus the higher electrical energy requirements. Now, you can run gas turbines at the utility level to generate electricity, but that's much less common at the level of individual homes.

Does home solar usage count on this graph? The electricity never hits the grid so I think it might not. If not California's massive push for home solar could be part of the explanation.

The amazing thing is there basically were no batteries three four years ago. And they can supply about 10% of the max power demand already. So it feels like the technologies we need are now good to go on an engineering an accounting basis. And adoption can be quite rapid. We're not saying in 50 years, 25 years, 10 years. We're looking at 5 years.

"power". This doesn't mean anything. LiFePO4 scales power to energy storage at a rate of about 1:3-1:4.

1GW of "power" they can supply for 4 hours if you completely discharge the battery.

But that's it. Batteries could supply 100% of the grid briefly, and you still wouldn't be anywhere near having completed a switchover.

You can see by the graph that the batteries don't piddle out that quickly already.

I've looked and not seen numbers on the actual battery capacity. But I think it's about 3-4 hours at name plate power levels. I think from what I know capacity is a function of discharge rate. And wear is also a function of discharge rate and max/min charge levels.

Makes the operating economics non trivial.

Given the dramatically lower costs of utility-scale solar vs. residential rooftop solar, is it not better at a society level for state policy to incentivize utility over residential solar? Utility-scale battery installations likewise should be much cheaper.

Residential solar may still make sense for new construction (maybe?) in the right areas. But there was a lot of scammy behavior around home solar installation at one point. Given that home solar does not effectively give you a backup generator for free (barring a lot of batteries and specific electrical hookups as I recall), it's not clear it's a big win in general. A lot of utility things aren't ideal at the individual house level if there's an option.

A lot of home solar installers now include grid tied batteries as part of a standard installation, which is what NEM 3 was intended to do (secondarily, after its primary goal of giving CA IOUs a massive free lunch).

This argument ignores the cost of transmission and distribution, which are higher costs than the generation itself.

And as electricity prices are driven down even further, T&D will come to dominate over energy generation costs.

Few models account for this, but Christopher Clack made one, and the lowest cost energy path was a small amount of investment in distribution now, paired with massive deployment of solar on homes and industrial/commercial.

This won't happen unless utilities are forced into it or their allowed profit model is changed to deliver ratepayers the lowest cost energy, however.

Unless every home is completely off the grid, you’re going to need to pay for most or the majority of that transmission and distribution anyway. Peak winter and summer I also would imagine many or most residential deployments aren’t going to cover their own need, unless they’re incredibly over built with massive batteries. In any case, I should check out your Christopher Clack reference.

The modeling is a few years old now, but here's a podcast episode diving into details:

https://xenetwork.org/ets/episodes/episode-146-why-local-sol...

(And of course, perhaps one modeler can get it wrong. But over here in California where the majority of our sky high electricity cost is from T&D, it definitet feels very true. And last national stats I saw had T&D as higher than generation costs)

As long as the houses are still connected to the grid you will indeed by paying for distribution costs. But I think that you should expect to see transmission savings since you'll have more of the demand met locally.

Why? If there's any possibility those homes will draw 100% of their power from the grid on any given day, then the cost of infrastructure will be pretty much the same since a grid failure is a disaster.

And there is a 100% chance they will, like my home is today, because 11kW of solar is making about 800W due to cloud cover which has persisted most days of summer.

You don't need as much grid capacity if houses have their own generation and storage.

The grid still needs to be built for the worst case - houses aren't going to have days of storage for a week of cold weather with dark clouds

Worst case is months. In the depths of winter you need energy captured from the sun during summer. That's a drastic rethinking of the scale of storage.

Absolutely. For the sake of the topic of discussion I was trying to think of Californian conditions but for places like Sweden you'll need completely different solutions (Sweden has been powered 100% on Hydro+Nuclear for decades now so that's one way)

My PG&E bill would make you think so.

But that is a PG&E problem (well problem for the consumer and great for their bottom line).

Generation is the majority of the cost across the US: https://www.eia.gov/energyexplained/electricity/prices-and-f...

Wouldn't that mean further subsidizing rich homeowners?

The people best positioned to take advantage of incentives own their own home and have the disposable cash to make the initial outlay for residential generation.

Yep, it definitely would. It doesn't mean we shouldn't still do it -- especially because richer folks tend to use more electricity and pollute more in general.

Generally, yes, but a few items:

- Distribution systems today can only handle some percentage of EV penetration in a given area, not 100%. Charging an EV from the roof skips the grid entirely

- The bulk of the cost is labor and permitting, not the modules themselves. Given that, California requires solar on newly constructed homes

- California updated their net metering program so that ratepayers aren’t subsidizing rooftop solar anymore (in NEM 3.0, homeowners get paid wholesale rates)

"Charging an EV from the roof skips the grid entirely"

As long as you have enough sunshine hours in your region.

Southern Spain is fine year-round, but you won't get any meaningful output out of a solar panel in Finnish winter. The rest of Europe lies between those two extremes.

Northern winter will be made more challenging by addition of EVs to the grid, because a lot of electricity is being used for heating as well.

As far as rooftop solar, it makes a big difference if solar PV is built into the design process at the beginning, rather than tacked on as an afterthought. Similarly, a household battery might be as common in the future as a household hot water heater (with a similar footprint, size-wise).

There's also regional variability - cities absolutely require utility-scale solar, just as they require dedicated agricultural land to feed the population, because there's not enough surface area in a city. Rural/suburban areas on the other hand are ideal for integrated rooftop solar.

> Given the dramatically lower costs of utility-scale solar vs. residential rooftop solar

According to Wikipedia the LCOE on small roof top solar is about twice that of utility solar. [0] So yes, it's dramatically better. But typically the price difference between what the utility is paid and what the customer is charged is 3 times, so your average household is better off installing solar despite generating it at twice the cost of a utility.

The price difference isn't just transmission costs. It's also retail markup, regulatory charges and metering and billing.

[0] https://en.m.wikipedia.org/wiki/Cost_of_electricity_by_sourc... - there are a lot of tables in that page. Twice looks to be generous to the utilities.

State policy is fundamental to the entire green revolution. None of it would have been possible without nations incentivizing it. Eventually once the entire energy generation/consumption cycle is entrenched and we are all dependent on it, they can safely take away the incentives.

It's a bit like changing the tires on a moving bus. Somebody had to pay for the new tires and wheels, and the support truck to run along side the bus, and the extra fuel, and a discount for the new tires, etc. Once the new tires are on the bus can keep driving without support.

That duck curve tweet is disingenuous. That curve in the tweet is for the lowest net load day (net load is actual load or usage minus generation from renewables). In 2023, if you took the day that had the least amount of net load, yes, it was almost entirely covered by solar power. That does _not_ mean the claim made in the tweet that California is run totally by solar power from 10am-4pm every day (today at 11:56 AM PST, it's about 51% run by solar power). California's grid has enough good things going for it that we don't need to lie about it.

You can look this up for yourself: https://www.gridstatus.io/live/caiso

> There has been a massive decline in rooftop solar applications in CA since solar energy reimbursements dropped

People also learned that the cheap Chinese solar cells die in 5-10 years and aren't worth installing unless your electricity costs are really high.

Is this something documented? I haven't heard of this and I've been following the space for a while.

Mostly installers--probably about 7-10 years ago back when the American manufacturers got pushed out of the market by the Chinese ones.

A lot of the installers got out of the business knowing they were on the hook when the warrantees came calling.

Although, American manufacturers have been restarting since Biden adjusted some of the tariffs in 2023.

We'll see if they go back to producing a decent product or if they just try to shovel garbage like China.

Reply on news.ycombinator.com