Hacker News Reader

Comment by WalterBright | Hacker News Reader

The cheapest grid "battery" is minute-by-minute pricing of electricity. Much of a home's energy use can be time-shifted if there was a point to time-shift - and variable pricing provides that point.

Your home is a "battery" as you can run the temperature of your hot water heater higher when electricity is cheaper, and let it "coast" when electricity is more expensive. You still have hot water on demand.

The same goes for the refrigerator, heating system and the cooling system. This can be made even more effective by increasing the thermal mass of the house, for example, with a pile of rocks. Pretty cheap for a battery, don't you think?

And, of course, there's charging your car when electricity is cheaper.

At last, we actually have a use for the Internet-of-things - an internet device on your hotwater heater to query the current price of electricity.

Pretty darned cheap for a grid battery.

View on HN

> The cheapest grid "battery" is minute-by-minute pricing of electricity.

This works to the extent that it works. But one of the big applications for a sufficiently cheap grid battery would be to store generation from solar to be used at night.

A large fraction of the nighttime load is for heat and light, because night is when it's cold and dark. It can't really be shifted into the daytime.

And another obvious use for a grid battery is to take advantage of that same demand pricing by buying power when it's cheap and selling when it's not.

> It can't really be shifted into the daytime.

The heat can be. Heat can be stored in the thermal mass of the house itself, and thermal mass can be added in the form of rocks.

With the advent of LED lights, the lighting bill isn't that much anymore.

This concept was widely deployed in certain regions with "storage heaters": very heavy (~40 Kg) electric heaters that would accumulate heat during the night when electricity is cheap and release it throughout the day.

They felt out of fashion because apparently they were hard to use as you cannot control how the heat is released. Charge them too much and you will have to open windows the next day wasting energy, charge them too little and you will have to turn them on during the day when the electricity is expensive.

You heat up a big block of concrete during the day. Insulated of course. Then at night you run water through it, and pipe that (now hot) water to radiator around your house. Already being deployed in Germany.

What's the advantage of using the block of concrete instead of directly heating water? Water has a 4x higher specific heat capacity by mass, or 2x higher by volume.

I could be entirely wrong as to the substance(s) involved :-)

I'm quoting from a memory of an article I read about 6 months ago, sorry.

googling for "specific heat high energy storage material" seems to indicate that rock-type materials are used where air heat is required.

They definitely are used in some circumstances. For example, a sauna uses heated rocks. This allows you to control the humidity separately and not have the water evaporate. I guess it would work just as well with sealed metal tanks of water, but rocks are cheaper and more traditionally available. And they can't leak.

So I'm not doubting you, just wondering why an intermediary heat storage medium is preferred in practice. Maybe for some of the same reasons as the sauna.

> And another obvious use for a grid battery is to take advantage of that same demand pricing by buying power when it's cheap and selling when it's not.

Far as I get the actual price of electricity varies a lot, so there potentially a lot of arbitrage to exploit. There's a dozen ways to store energy with varying amount of round trip efficiency, capital costs, and operating costs. Some set of technologies will win out in that space.

I suspect one reason these technologies aren't well developed is because historically the cheapest is baseload power produced by coal and nuke plants. Economically the price is actually subsidized. So utilities did not want knuckleheads buying their cheap power at night and selling it during the day.

This is all nice and well, though I'd like to add the thought that this is much easier to do on industry scale.

Have one large industrial factory like an aluminium smelter or steel plant that is redesigned in a way that it can shift its electricity use a little bit? That's probably having an effect larger than a lot of personal users.

I guess there's a political issue here, as industry often pays very little for electricity to begin with. I know at lesat in Germany most large industries are excluded from a lot of fees that normal customers pay. Providing "cheap and reliable electricity" to industry is often almost a political mantra. I guess it would make sense to say at some point: "Dear industry, you can have cheap electricity, you can have reliable electricity, but you can't have both. If you want to have it cheap you have to support the grid by providing flexibility."

This exists already. https://www.ecotricity.co.uk/business-energy/virtual-power-p...

Providing cheap and reliable electricity to industry is basically a prerequisite to having industry in the first place, which is why it's so politically important in countries which still care about that. Aluminium smelters, being particularly sensitive to electricity cost, are somewhat willing to choose their location and reduce their power usage briefly if it gives them really, really cheap power, but I think even those still need to be running most of the time.

From what I heard there are reasons for both. Industrial electrical users are required to engage in their own impedience matching, demand regularion and other such contractual issues in addition to the ammount of other expenses for say 100MW to N consumer homes being greater than one connection to a widget factory.

It isn't some mustache twirling plot to make the consumer pay more while some abstract industrial fat cat pays more or anything - real economics are at play.

The prevailing obvious alternative to competitive pricing that tries to roll in other expenses would be direct generation by industrial scales running their own fossil fuel turbines or direct thermal alternatives. Which result in worse performance at scale from everyone mid-sized and up rolling their own separate systems and worse pollution control as instead of having four great smokestacks to put scrubbers on and regulate there are now four thousand small ones. Economies of scale have been here for centuries.

Aluminum smelters and steel plants while they may use a lot of power also cannot be just neglected and cut off mid-cycle without wasting a very large batch slag in a way that is very expensive to clean up.

I'm not here for micromanaging my water heater or refrigerator.

Both need to hold their temperature as long as possible and get up(or down) to temperate at some balance point between speed and efficiency. I want my provider to even it out and set a reasonable price between the variable extremes so I don't have to think about it.

You don't have to think about it. Just get an IoT refrigerator.

Sorry, I'm also a picky consumer when it comes to appliances. I want my fridge to be just a simple heat pump controlled by a thermostat.

You'll pay more for electricity, then. Your choice.

Will?

Might.

And probably trivially more, especially when you factor in a software update bricking your fridge and the manufacturer has conveniently disappeared / gone bankrupt / changed hands so many times no one knows who supports it.

You’re assuming an ideal world, which can’t possibly happen.

Sounds like a lot of hellish outcomes to me: Planned obsolescence, selling my data, privacy invasion. Did I miss any?

The only data transmitted to the power company is the query "what is the current price of electricity".

Yeah, but you can bet the manufacturer will avail themselves of every opportunity to make extra money and bundle ads while also selling your data. We've seen it with smart tvs compared to dumb tvs, smart phones compared to dumb phones, smart speakers compared to dumb speakers, etc, etc.

You're not thinking capitalistically enough.

Why blame capitalism?

Call it what it is: exploitation.

Because it is. People can choose to pay more for a device that doesn't do that, but they don't (for various reasons). It's capitalism meets human psychology.

Capital controlled by private owners rather than the state doesn’t necessarily include greed and fraud.

Greet and fraud are encouraged, if not idealised.

That’s not capitalism fault though.

Every other form of economic policy has the same problem, so it can’t be those systems in and of themselves that are the sole issue.

What data? My water heater sends a query every 10 minutes about the electricity price? And where's it going to display the ad?

Smart phones to dumb phones? Seriously? Until someone can post a guide to posting on using a rotary phone that is just an asinine comparison.

The fact that IoT is still a running joke and just a tolerated parasite to lower the sticker price at best is proof that the data is garbage because none of them are making much money off of it. Until they have an actual monetiziation plan for the data it is plain magical thinking.

"People mostly increase demand on their hot water heater in the early morning, evening, and night for showers dishwashers, and laundry on a semi-weekly, weekly, or bi-weekly basis."

"Well I could have told you that without spending six to seven figures on unnecessary expenses on a product line!"

In theory it's possible to do it without any of that. All you need is a standard protocol for the power company to announce the current generation rate (and a prediction for what the rate will be over the next 24 hours), which your device then uses to determine when to use power, with no data ever going back in the other direction.

Of course, the actual implementations can (and in many cases have) been a lot worse than that.

Any you would recommend? All the sales I’ve seen are based around gimmicky touchscreen and phone app controls that are likely to leave service long before the rest of the fridge does.

Working with some power companies in Australasia, I found out there are actually water heaters which receive a signal over the power lines to turn off and on. You would make an agreement with the power company for a lower rate, and a total amount of hours on per day (plus more details). The power company would then schedule your water heater in off peak times.

This was more of a method of managing peak load, however, still seems applicable.

When you shower are you aware whether the water was heated an hour ago or a day ago?

If the temperature is beginning to dip below it's acceptable threshold and it's waiting for a certain price before it kicks back on, yes.

Yea...hot water heaters aren't perfect insulators.

Actually they are pretty good. I did a test where I switched off a completely heated 150L standard home water heater for 36h and after that it only needed to heat for 0,6hWh to become back to the starting temperature.

Also you should always store hot water at a very hot temperature (Nasty bacteria thrive in warm water so you mustn't store it) but you should always mix it (with fresh cold water) to a lower temperature for bathing and washing.

So whether it's a tank of 80°C water or a tank of 65°C water make no difference to the 38°C water coming out of the thermostatic mixer in your shower.

You don't need to keep the water at a high temperature all the time.

I don't have the regulations to hand but it's enough to bring it to a high temperature once every few days for some period of time (to kill the bacteria that cause legionnaires disease).

0.6kWh sorry for the typo.

The majority of electric water heaters in France and Swiss households are controlled by the utility company (Enedis in France, SIG, RECOM, etc. in Switzerland). The signal comes as a simple PLC and thus turns on the relay which turns the heater on.

The advantages are a lower rate and a guaranteed 8 hours straight on-time per day.

I would also add that heat insulation is grossly underrated. With modern heat insulation solutions like wood or hemp concrete, mineral wool one can decrease the heating / cooling need by 60-80%. This can be combined with solar panels. In practice, this means we have ~10 USD heating / cooling cost per month for a medium size house in Central Europe. Initial cost is not trivial though. Same goes for other usage of electricity, for example lighting. Using LED lights reduce the energy need by a large percentage. The best energy optimisation is to reduce the amount of energy needed to be produced. Once you have that it becomes easier even with current technology to have a battery that can buffer your electricity,

Is this a real problem/solution? Wouldn't the current usage pattern of a city block be normalized across all residents such that there's no gain to be made here? Is the current minute to minute power usage fluctuating that much?

Don't things like water heaters and fridge compressors run many times a day. Is shifting all that to night hours even feasible simply because the appliance is smart?

It is. Most people’s air con will switch on at similar times. People mostly shower first time in the morning, or in the evening, cook around the same time. Amusingly and semi-notoriously, the Uk national grid monitors for ‘tv pickups’ like halftime in the football or the cliffhanger in the latest Eastenders - they can spike hundreds to thousands of megawatts. There are special power stations to handle to peak while others spin up, otherwise the frequency drop would cause grid failures.

But that all seems like human behavior/usage that would have to change, not automated processes.

Even still, wouldn't randomly fuzzed automation times be easier to implement than an fully connected system?

The concept is at best to try to use them as a local "lossy capicator" when the baseload power is unused to minimize dispatched power.

I suspect in this area they would get more flies with honey by letting them use power cheaper when they want to get rid of it. A "free/cheap power" signal essentially that tells any device which wants it to disregard efficiency - any utility is better than nothing.

Power banking batteries are the trivially obvious way to exploit and monetize it while fortunately stabilizing the grid as a side effect but well the infastructure costs are expensive.

Fuzzing would be easier to implement assuming the fuzz period is meaningful on a time scale without causing other problems like "power cycle it if it doesn't come on in twenty five seconds" because the end user/technician rightfully thinks it isn't working and when it rolls 5 second delay next time it gets a faster end boot.

Regarding the water heater. Actually the more you heat something the more it also cools down. Unless you have beyond excellent insulation the coasting method will not do good for long. The heater will release more heat the more it possesses and so it would be best to instead increase thermal mass than temparature.

Somebody correct me if im wrong

> The same goes for the refrigerator

What do you mean? Fridges generally run at the highest reasonable temperature. Are you proposing adding more thermal mass to them?

There's a range of suitable temperatures for your refrigerator. When electricity is cheap, it cools to the low bound, when expensive, it only cools it to the upper bound. Of course, increasing the thermal mass of the refrigerator will improve this.

Your freezer can do that, but your fridge can not. The safe range is 40-35 (4.4-1.6), however your food will spoil faster. It's better to keep it 37-35 (2.7-1.6).

That's a very narrow range. You might be able to pre-cool the freezer, and use some of its cold to keep the fridge where it needs to be.

But I suspect you would pay for that in lower efficiency, because the colder you need to go the less efficient the cooling is.

Electricity prices would have to be dramatically lower to make it worth taking your freezer extremely low to store cold.

I suppose you could make a freezer with a special water storage compartment (filled at home to reduce weight), then use that to store cold. But with the extra complexity, and lower efficiency.... just how much lower electrical prices are we talking?

Electricity prices go negative at times.

Only wholesale. Retail prices never go negative.

Never say never.

“Receive SMS alerts whenever prices drop below zero, or use our API to program your smart devices.”

https://octopus.energy/agile/

Thats not a narrow range at all for a very well insulated fridge.

Don't forget that people open the fridge, it's not realistic to make a fridge that you are not allowed to open at certain times.

Not necesary to make that requirement.

The way stuff like this works, is that with a fairly simple set of end-point power controland some pretty simple monitoring, and then some central smart software, you can effect pretty dramatic improvements.

For example, the smart controller allows the fridge temperature to drift (just a little), similarly with central heating, and geyser, and the net effect is that power draw is temporarily reduced during peak times.

Multiply this over your whole year and you have some cost savings.

Multiply it over a whole neighbourhood, and the utility can dramatically reduce their cost of maintenace, and their cost of deployed capital.

(Because of reduced power/thermal swings in their equipment, and reduced peak carrying capacity of the network)

Reply on news.ycombinator.com