Hacker News Reader

I got solar last year, so ots been interesting to see how habits change (or dont).

The first stuff was obvious. Our hot water and pool pump were already on timers so those were switched from night to day. Now on sunny days that's a free.

We've moved the dishwasher from night to day. The washing machine was already on day, so no change there.

We gave a small (5kva) battery so a chunk of our evening usage is covered.

There are a few other more subtle habits, harder to spot, and perhaps with more limited gains. Some random oven usage (for not-meal things) is more likely to happen during a sunny day,that sort of thing.

So yes habits have changed, and obviously our grid usage is massively reduced (by about 70%). But equally were not anal about it - things still run when they need to run.

Disclaimer: this is anecdote, not data and ymmv. But saving money is a powerful motivator.

> Some random oven usage (for not-meal things)

Curious: aside from running a cleaning cycle or seasoning a cast iron pan (both of which are at least meal-adjacent), what other non-meal things happen in an oven?

I interpreted "non-meal" as anything except breakfast, dinner, supper. Like a cake

Yeah that was my take on it as well.

Reflowing PCB’s is one thing I can think of.

Drying 3D printer filament too (though I've not tried it myself and not sure I'd want to use my ovens for that).

Drying out 3d printer filament is another.

Also heating up engine casings to get bearings in/out.

"apart from buying batteries" - that is imo precisely enough. We are nearing a time, where a battery that would flatten the avg household demand curve costs under $2000 (without installation costs), and that might be quite enough to change how electrical production works.

Yes, poorly worded from me. I meant more households can't really do a huge amount about load shifting without batteries. Not that batteries are a bad idea or not workable.

If a battery could flatten demand on a household basis, then why not just implement it on a grid basis?

Probably for the same reason many things are the way they are; pushing the cost onto the consumer.

Snark notwithstanding, a thousand batteries distributed close to actual consumers give much more resilience against grid outages, like transmission lines or transformers damaged by a hurricane.

Spreading out the capital costs to those better able to shoulder them.

So rich people? So they can save money while people who can't afford a battery cannot. I guess it's like this for most cost-saving investments, but doing it this way hardly contributes to a more fair society.

Nah just tax the rich and make it illegal for corporations based in a tax haven (like Ireland) to do business in our economies.

>> The real long term idea with smart metering isn't so much demand savings (though I think it was billed as that; and I agree the results are very underwhelming from everything I've seen), it's to start allowing dynamic energy tariffs and getting people incentivized to move load to cheaper times.

The original long/medium/short-term goal of "smart" meters was to eliminate the need to manually check meters and integrate that into the billing systems. It was a direct cost savings to delivery companies. All the fancy variable pricing stuff is an afterthought.

I phrase it as utilities got a gov-mandate to install you new stuff, pass on the cost to you, and save some monthly/bi-monthly opex cost that they can pocket themselves (or at least pat themselves on the back for).

If it costs more than it saves, they don’t mind, actually that’s great (for them).

In rural areas in Ontario, the system used 3G (and for some reason I think only supported 2G in the past) to send readings. So that will be another truck roll (or more likely, cheque to telecoms to keep up some old infra).

I also suspicious of $$$ technical investments to reduce relatively small costs (especially ones where a lot of the dollars leave the area).

It also facilitates half hourly settlement. Previously we used standard profile curves based on the average household to attribute consumption to each 30 minute settlement period, where with smart meters sending half hourly reads you can settle based on the actual consumption.

The transition period includes "elective half hourly settlement" which basically means that as a supplier you can selectively settle meters on a half hourly basis, which created an interesting (but time limited) arbitrage opportunity where you could do some data science to predict which meters would be beneficial to settle half hourly or not

Refs - https://www.ofgem.gov.uk/energy-policy-and-regulation/policy...

- https://www.elexon.co.uk/settlement/profiling/

I’m a bit confused: why do they need visibility into individual household’s load to accomplish this? Can’t you just look at what the whole city is drawing and order based on that?

Or do you mean fully dynamic pricing/billing to end users?

I think this is for situations where customers can choose a 3rd party energy provider. The utility transmits energy to the user and settles with the energy provider.

Makes sense to settle that up near real time, I guess.

This is on the wholesale side - as a supplier your customers consume energy which you're responsible for purchasing, and it's priced in 30 minute chunks. To get visibility into when your specific customers are consuming energy you need smart meter data from the household level

Eating dinner no, but if your washing machine and dishwasher have an internal timer function (most relatively recent models of both appear to), it's pretty easy to load the appliance and tell it to turn on later that night.

> or do washing in the middle of the night

With dynamic tarrifs around here you sometimes make money by doing it during the day, especially on sunny windy days as you get money back for using electricity at that time. My bill is about two thirds of what it would be otherwise (since I cannot install solar panels or batteries it's the only way to have any advantage).

I've automated it so it picks up if the prices goes below a certain set amount, though that's not for everyone.

Btw my parents have been doing off hour washing since the nineties (we've had double meters for a long time). It's nothing new.

The point is that 99% of the contracts have no "dynamic prices", they have few different tariffs established up front an year for the next. In this case you do not need a smart meter to shift load. You simply look at the "current year tariffs" and that's is. If we have dynamic prices than we need a standard data bus for every home appliance connectable to a personal automation or directly to the meter, without that manual load shifting is extremely limited.

In France there is a VERY LIMITED system allow a two wire connection for big appliances were the meter simply tell "cheap or expensive" state and the appliance now that "now the electricity is cheaper, now more expensive". It's way too limited. New homes have since some years a mandatory mitering but without dynamic prices and made to be piloted and throttled loads it's just expensive eye candy stuff.

Personally I have automated via HA as much as I can but such automation is essentially just switches HA can open or close via Shelly integration, there is no real smartness in all appliances even if it's damn cheap adding it from the initial design phase.

My personal experience with "smart" thermostats is that my electric bill went up.

By "smart" I just mean programmable. I had settings for morning, daytime (nobody home), evening, and night. It makes no sense but my total costs went up. This was years ago, so I don't remember exactly what temps I had set for the different times of day. Now I just set the A/C at 76 and leave it there.

Let me paint you a scenario where this is possible: you leave home at 9 AM, and the A/C stops. The house slowly gets hot, maybe 95°F. You plan to arrive at 9 PM, so you program the A/C to get the house to 76°F, which the thing thinks it can do in 1 hour at full blast. But the energy price between 8 PM and 9 PM is the highest of all day: you just programmed the A/C to the worst.

Instead, now you leave it keeping 76°F, and the thing can do it working at 10%. You consume more energy at the end of the day, but the consumption is spreaded through the cheapest hours in the middle of the day, and only a small fraction of it is consumed in the expensive hours from 6 to 10 PM. This is how freezers work, cooling intermitently as the thermostate goes off limits.

I would test if it is worthy money wise, not energy wise, to put the A/C at full blast in the cheapest hours. Say from 4 to 6 PM (the cheapest here in Spain), to get the house as cold as possible, and then try to play with A/C at maybe 5% letting it rise until reaching 76°F at 9 PM.

> I would test if it is worthy money wise, not energy wise, to put the A/C at full blast in the cheapest hours. Say from 4 to 6 PM (the cheapest here in Spain), to get the house as cold as possible, and then try to play with A/C at maybe 5% letting it rise until reaching 76°F at 9 PM.

Depending on how big the house is and how well insulated it is, it very well may make for a big thermal battery and this would work well.

If small and/or poorly insulated or drafty then this will not work well.

Isn't it less expensive and more ecofriendly to turn on the AC when you are actually there? Dumber would be smarter.

Not to mention, you can be late, or early, or not arrive for the whole night.

It might be, you should research how your region produces energy.

For example, where I live the energy produced from 10:00 to 18:00 is almost 100% solar, eolic and nuclear (https://www.esios.ree.es/es/balance?date=04-07-2024&program=... , orange and red bars are solar, light green are wind, dark blue are nuclear). Energy after 19:00 starts to draw gas (yellow bars) to cover the fall in solar production, peaking at 21:00. Incidentally, this also matches prices going from ~5 cents/kWh to ~20 cents/kWh (https://t.me/tarifaluz/3665). So for me it's both ecofriendy and moneyfriendly to shift as much consumption as possible to the 13:00-18:00 hours.

Eco-friendly, possibly, buy cheaper? Maybe not. It depends on when you arrive home. If you arrive home during peak hours when your utility prices are higher, it can be cheaper to run the a/c earlier, while you're not home, to at least get the temperature down into the rough ballpark of where you want it, and then maintain that temperature into the peak hours.

Even the eco-friendly bit isn't always clear, depending on your equipment. It might use less energy overall to maintain your desired temperature 24/7, even when you aren't there. Your particular equipment may run less efficiently if it has to go on full blast for an extended period of time to do a large temperature change when you first get home.

Not always because it depends on many factors such as demand energy pricing. The house also has heat inertia, meaning it could take hours to reach a ceratin temperature, so you'd have to turn on the AC hours in advance before you arrive home.

Cool thing is I don’t care about the temperature of the solid mass inside of where I am, just the air really.

I know when I turn on the A/C in my office, it’s not going to cool the concrete wall/floors/ceiling much in the 8h I’m there before I shut it off for the day. And that’s fine.

The solid mass inside the building, the walls and floors heat the air and the AC is working harder to keep up. Also exposed glass to direct sunshine that creates a greenhouse effect.

It uses the least energy, but here we are talking about demand shifting to make use of variable electricity sources. It can be worth using more cheap solar energy in the middle of the day instead of less natural gas later on.

The key point is that an insulated house acts as a heat/cold storage device, so it's like buying booze and pouring it into a (slightly leaky) bucket during happy hour, to drink later. Even if you loose a bit to the leaks, you're still saving money overall by buying the booze when it's cheap.

An A/C may also operate rather more efficiently when it’s modulated down than when it’s running at full blast.

With heat pumps (like AC) and condensing boilers, it is often more efficient to run for longer and slower as long as that makes your appliance run more efficiently (lower flow temperatures for heating, higher for cooling). Requires the system to support modulation though.

Is there any defined way to do the math on this?

I haven't been able to find even specific case examples showing which consumes less energy: turning down heating/ac when no one's home or just keeping it on constantly.

I'm sure it's pretty complicated because buildings vary, but someone who understands heat transfer should be able to do it at least for a few ideal spherical cows...

> daytime (nobody home)

But that's so last century :) We WFH now.

> It makes no sense but my total costs went up

It does make sense if you were scewing around with the temperature values in excess of 2°C, because the HVAC would work harder to reestablish a new steady state. Most houses have thermal mass and heat inertia and sometimes it takes hours to reach a new temperature set point, the HVAC working at full load to do so. It depends on many factors and it's not easy to optimize without logging the temperatures, the operating cycles of the HVAC and understanding the whole process.

> It makes no sense but my total costs went up.

Due to inflation, the expected state is your costs will go up, even in the case of a small improvement in energy use. A number of western countries also seem to be in a state of energy crisis that could explain the price going up beyond inflation.

So that is less informative than it could be. Did you see your real energy use rise?

> it's to start allowing dynamic energy tariffs and getting people incentivized to move load to cheaper times.

That’s what they rolled out with them, but insufficient change to habits.

> I suppose AC/heating would use enough that it would be worth turning it down at certain times of day.

This is generally the big win, and why they’ve had to still create a program to subsidize smart thermostats and payments to enable the back off features.

Mass installing smart meters to incentivize the public at large to buy+install smarter thermostats is a less effective and more expensive strategy than giving the thermostats away and writing users a guaranteed cheque in exchange for remotely controlling it.

(There are other dynamics at play: electricity here is generally cheap and I pay more in fixed charges than consumption and the heating side is almost universally fossil fuel/wood but that’s achangin)

> That’s what they rolled out with them, but insufficient change to habits.

If that's the case, then that means they didn't set the incentives strongly enough, or didn't educate people well enough about how they can save money. While people do grumble about their electricity bill, they might not realize that there are things they can do to reduce it.

Making up numbers here, but let's say the average/"real" cost of electricity is X. To incentivize different behavior, the utility sets the price for peak hours to X+20%, and drops the price during off-peak hours to X-10%.

They observe behavior for a few months. No meaningful changes in behavior. That just means people value their ability to not think about when they use electricity more than they value the cost savings.

So, set the incentives harder. Set peak hours to X+60%, and off-peak to X-30%. Observe behavior for a few months. Maybe people will start changing their behavior now.

Unless electricity is actually that much more expensive during peak you're falling into the trap of simply artificially making the world worse for the purposes of controlling people's behavior.

People aren't stupid and will immediately recognize when the prices aren't connected to reality and resent you more than change their behavior for you. Most workers don't have access to cheap electricity because they're not home during the daytime hours, and setting up an elaborate series of timers is a non-solution.

You'll never get people to vote for the "everyone's electric bill goes up because fuck you for having a job" bill.

I think electric cars could change this equation. The savings are high enough and the inconvenience low enough.

This is definitely the case - if you have a time-of-use tariff and an EV you charge at home, there's a strong incentive to schedule your charging to the cheapest time of day.

Especially if your EV could be smart enough to do that by itself.

> However, I don't know how much consumers can do about this apart from buying batteries

If one has a water boiler or a heat pump with a buffer, they can run it at full load during cheaper periods. If they have an EV they can also charge that.

The load curve is also changing with renewables.

https://www.enerdynamics.com/Energy-Currents_Blog/The-Electr...