Live data from Hacker News

Charging cars at home at night is not the way to go: study

news.stanford.edu

131–140 of 394 posts

Re: Charging cars at home at night is not the way to go: study

#131
post #118

Earlier quoted context omitted.

What are are you using that much power for? I can’t think of much that uses that much per that you can shift to time of day pricing, perhaps pottery kilns? Even 2MWh per month is enough to drive 2 EV’s around 50,000 miles per year each. If you’re very flexible with time of day rooftop solar could probably save you quite a bit.

Just having my 4x3090 constantly running deep learning experiments consumes 1MWh/month. And that's GPUs only.

Except constantly running can’t benefit from time of day pricing. Essentially it needs to be some 15kWh load that only needs to run for 6 hours per day.

Re: Charging cars at home at night is not the way to go: study

#132
post #17

>We're doing it wrong, according to a new Stanford study Wrong according to what metric? Cost? Raw efficiency? A lot of people are more than willing to give up efficiency so that they don't have to actually worry about finding a station during the day. To say that something is just outright "wrong" based on their personal preference of priorities comes off as unhelpful to me. The solution is simple. Make electricity…

That’s how we do it in Denmark. We have hourly prices on electricity, so for me it’s cheaper to start the dishwasher outside peak hours. In some areas they are experimenting with car chargers hooked up to this electricity price info, so it charges when it’s cheap.

Re: Charging cars at home at night is not the way to go: study

#133
post #4

This is simple. Musk just needs to throw 1% of his AI talent at optimizing charging times via realtime grid negotiating and nobody will need to even think about this. Cars will turn into one of our most valuable assets as they help stabalize the grid by being both sinks when energy is plentiful and sources when it's not. There are certain problems we all need to think about, but this is not one of them. It will effec…

This is not simple. The batteries in cars can be used to help load-balance but they cannot offer grid-scale storage.

The following is back-of-napkin figures but there are close enough to make the point.

Take a country the size of the UK. It currently uses 1600 TWh a year. Once fully-electric it should need around 1100 TWh a year, so 3 TWh a day.

The UK, despite being a windy place, still has weeks without wind once or twice a year. We had a week without wind power in the middle of August this year and had to generate 60% of our power using gas. A windless week with a frost snap is something you often see in January and February. A cold week, with everyone using electric heating (when solar power does basically nothing) is why the UK might need TWh of stored energy to be secure without fossil fuels.

7 days at 3 TWh is 20 TWh.

A big home Tesla Powerwall can store 14kWh and if run at full power it will be empty in 4 hrs. You'd need over a billion to store 20TWh.

In the UK an average home currently uses around 10kW of electricity a day, but also has a car, a cooker and central heating, all using fossil fuels. Add those in and were looking at 40-50kWh: 20kWh for the heating, 10kWh for the cooking and another 10kWh for the cars.

There are savings to be made of course, the electric versions of all of those items are way more efficient. There's insulation; there's heat pumps; there's electric cars.

The average UK car sits idle for 23 hours of the day. An average-ish car is around 100kW. Even if that hour is slow and often stationary, it's not unreasonable to expect it to be another 10kWh. And the average UK home has more than one car.

A mid-range electric car has a 50kWh battery, so has a bit more beef than a Tesla Powerwall battery. If they are being used for 1 hour a day they are probably going to need a top-up of around, say, 10kW. So let's say they have 40 kWh to spare.

20 TWh / 40 kWh = 500 million cars.

There are only 33 million cars in the UK.

You can, of course, quibble with these figures. There's subtleties everywhere here, but the orders of magnitude are plain. We need to be able to store renewable energy—desperately, and in enormous amounts. Car batteries will help—no doubt—but more in load-balancing and smoothing rather than long-term storage. We need something like grid-scale hydrogen storage to solve this.

See here for 1100 TWh figure. This document has loads of other really interesting details too:

https://www.nationalgrideso.com/document/264421/download

Re: Charging cars at home at night is not the way to go: study

#134
post #96

Earlier quoted context omitted.

Surprisingly, transmission efficiency (grid to motor) does not appear to be lower with inductive methods than conductive. Numbers above 90% should be expected. A bit lower for the dynamic (in-motion) solutions, but not so much that it’s a deal breaker. There seem to be much larger savings/costs elsewhere in the system that make transmission efficiency of less importance. Cars that are stationary most of the time are…

This source ( https://batteryuniversity.com/article/bu-412-charging-withou... .) says efficiency is only 75-80% when charging inductively. Do you have any better sources for those 90%?

Witricity claims to have inductive charging efficiency that matches traditional plug-in level 2 charging:

https://witricity.com/newsroom/blogs/what-is-efficiency-how-...

Re: Charging cars at home at night is not the way to go: study

#135
post #76
post #69

Earlier quoted context omitted.

Yeah it is. I hate having to have a process in my brain thinking about optimizing everything. So I bought solar panels and batteries. Now I no longer care. The grid can keep getting more expensive and less reliable without causing me grief.

>So I bought solar panels and batteries ...Mission accomplished?

"Profit"

Re: Charging cars at home at night is not the way to go: study

#136

Sounds like the problem is simply that the US doesn't seem to have real time pricing of electricity. In Norway we pay what the market determines is the current price. Most people don't have flat rate charges. I can see the spot price in real time in an app. Why isn't that good enough?

>Residents with variable-rate power plans are being hit the hardest. Such plans charge different prices for electricity depending on how much demand there is. The more demand, the higher the price.

This is a thing for some power companies in Texas, but during peak times or in a crisis like TX experienced in winter of 2021, the prices skyrocketed because demand was out of control trying to keep warm. People ended up with a $5k electric bill trying to keep a modest single family home "warm" (in the article they mentioned they kept lights off and thermostat set at 60F/15C, not very warm).

https://www.npr.org/sections/live-updates-winter-storms-2021...

Re: Charging cars at home at night is not the way to go: study

#137

Sounds like the problem is simply that the US doesn't seem to have real time pricing of electricity. In Norway we pay what the market determines is the current price. Most people don't have flat rate charges. I can see the spot price in real time in an app. Why isn't that good enough?

"The US" doesn't have a single system, not even within each state. Anecdotally, I have Time-of-Use billing with my PG&E electricity, but AFAICT, I was only allowed get that because I have BEVs.

I think the "problem" is that someone needed to publish. This is a complete non-issue and utilities will shift the load as needed by adjusting the billing. Also, the notion that everyone going "in to work" is some antiquated pre-covid thinking.

Re: Charging cars at home at night is not the way to go: study

#138

Earlier quoted context omitted.

it is, but that's the part that technology can actually help us with. all the internet-connected smart home stuff that companies have been trying to sell us is a bit silly when electricity is sold at a flat rate, but if my car charger, my clothes dryer, or my or my dishwasher could sit idle until electricity prices drop into the cheap zone and then turn themselves on, that internet connection becomes useful. an attit…

this just makes me think of the situation that resulted in Global TCP Synchronization flapping problem; when everyone is operating on the same premise without some random delay or other way to 'shard' the load it seems like a Slashdot effect is bound to happen.

Saw the Slashdot effect reference after such a long time. Not sure what % of HNers were there in the heydays of /.

Re: Charging cars at home at night is not the way to go: study

#139

Sounds like the problem is simply that the US doesn't seem to have real time pricing of electricity. In Norway we pay what the market determines is the current price. Most people don't have flat rate charges. I can see the spot price in real time in an app. Why isn't that good enough?

>Residents with variable-rate power plans are being hit the hardest. Such plans charge different prices for electricity depending on how much demand there is. The more demand, the higher the price. This is a thing for some power companies in Texas, but during peak times or in a crisis like TX experienced in winter of 2021, the prices skyrocketed because demand was out of control trying to keep warm. People ended up w…

This is all true, but I think the Texas issue is an outlier. Many things went wrong and all of them were preventable issues caused by mind-boggling terrible governance. Somehow these people are still in power.

Re: Charging cars at home at night is not the way to go: study

#140
post #48

Earlier quoted context omitted.

Future is SAE J2954 (aka Witricity): you have a wireless charging pad under each parking spot. No screwing with power plugs, just park and go. This isn't ideal for fast charging, but for the slower charging you'd get from being somewhere for 6-8 hours, it's perfectly fine. You still need to run power to every parking spot, but now it's embedded in the ground, and you don't have to worry about someone just coming arou…

Losing 30-70% of the power and adding a bunch of cost doesn't seem like the way forward. Cables exist for a reason. They are the best way to move charge.

These can actually be cheaper than the alternative to install, and get greater use (assuming the standard gets widely implemented). Definately going to be a thing. Loading/waiting areas for taxis and delivery vehicles seem like an early adopter market.
Post reply on HN