Hacker News Reader

Comment by IshKebab | Hacker News Reader

Storing energy from the summer for the winter is a really inefficient way to do it. It's much better to massively over-provision the solar so you have enough energy - on average - for the winter. Then you only need a couple of week's worth of storage to account for extended cloudy periods.

Much cheaper, and you get a ton of extra free power in the summer. The only downside is a typical house roof doesn't have enough space. But a typical house doesn't have enough space for a 1 MWh battery either so...

View on HN

Yup if you really need to be off grid in a climate that has cold, cloudy, snowy winters, you’re probably going to need a generator that runs on fossil fuels. For everyone else, use the grid.

Agreed. You can increasingly over-provision the solar generation to reduce the proportion of time when you will need a fossil fuel generator or grid input, and install lots of battery to allow the system to smooth over multiple dull days. But chasing that 100% is going to be very expensive, and at some point it'll be much cheaper to have a fossil fuel generator that you need to run 1% of the time.

Yeah, w/o grid fallback, I’d much rather aim for 98-99% w/ solar and have an alternate source to close gap, rather than aim for “five 9s” on solar+batt. It’d take a lot to talk me out of a multi-source approach.

At that point, a small backup generator (or even just staying grid-tied as insurance) makes far more sense

Or you find somewhere with terrain amenable to hydropower. It’s how we bridge the gap in the winter.

you have personal hydropower? that sounds pretty cool

A lot of great YouTube videos on personal hydro setups on small sized creeks. Even just a few hundred watts running 24/7/365 is an incredible resource.

And it's especially great if you have a neighbor with a son who are willing to do the labor for you. [1] [2]

[1] https://ludens.cl/paradise/turbine/turbine.html

[2] https://news.ycombinator.com/item?id=20075110

An insane amount of work to build though ...

Tell me about it.

We don’t have a permanent stream, but we do have enough intermittent flow in the winter to keep a 55,000L tank full. So our install entailed building a huge tank, a filtration system for water ingress (as it’s also our potable water supply, and a firefighting reserve in summer), digging 400m of trench over nightmare terrain with 70m of vertical drop, crossing a road twice, burying 90mm HDPE water line, fibre and 4x25mm2 power (latter two not necessary for hydropower but useful to have, and if I’ve got a trench open I’m putting everything in it at once) - and then building a hydro shed, installing the turbine, connecting it to our grid via a grid tie inverter, configuring our grid to accept power from it, setting up automations to turn it off and on depending on power demand and the level in the tank, and of course all sorts of side quests to achieve the above.

It has been neither cheap (about €12,000) nor easy (perhaps six weeks of full days for me, if added up over the year it took), but it has given us enough extra power in the winter that the petrol generator is now under a pile of crap in the shed, getting dusty.

That's some dedication

> digging 400m of trench over nightmare terrain with 70m of vertical drop, crossing a road twice, burying 90mm HDPE water line, fibre and 4x25mm2 power

This seems like a $50000 bit of work. Will it ever pay off or was it more of a hobby project?

Reminds me of this great house/greenhouse in Norway with a small "power station" in the stream outside. https://youtu.be/irp_HPzfxbQ?si=ZR3PAXvUyjsSSZx5&t=1658

Or run it on biofuels.

Woodgas would work as well

It's uncommon, but a wind generator can help a lot: in some climates, cloudy days tend to be windy days. Not really practical in a city though.

The AMC White Mountain huts have been doing this for years. The croo don't tend to maintain the systems, so I've never gotten a sense for what their storage capacity, generating capacity, and loads look like, but from a visitor perspective, the system works well.

Reportedly, even the fairy stout wind turbines they use up there have short, brutal lives. I heard the story of a croo that had to lasso/tangle/jam the blades of theirs in a storm because it lost the ability to control its speed and the alternative was letting it overspeed and possibly tear itself apart. They aren't large in diameter, but at the speeds they turn even in normal conditions up there, catastrophic failure could be really bad.

> It's uncommon, but a wind generator can help a lot

It's uncommon for a reason. Wind generator capacity rises with the square of the rotor diameter. That means small-ish generators (let's say "small enough to be roof mountable without additional mechanical supports") are significantly below 1 kW of power. Seriously, the systems people by for their sailing yachts make around 50W from a nice breeze - enough for lights and to trickle charge the battery while docked, not nearly enough for a fridge.

Combine that with quite a number of moving parts, changing loads and exposure to weather, you get very short maintenance intervals and final lifetimes.

If you have any other option for power, its almost always economical to just use that.

There's the passive haus highly insulated stuff. Guess that might work?

If you wanna spend 2-3x more, yes. Otherwise solar or grid battery is cheaper.

Try 30-50% more.

It's so obviously better to reduce your need for heating and cooling than it is to increase your panel. battery, and HVAC size.

30-50% over 500k build is 10x more than 10k solar or 5k worth of batteries.

I've just setup electrical heating for my bedroom (HA PID sensor). Uses about 450KWh - $90 NZD worth of grid power per winter. Heat pump would take 20+ years to pay itself. Double glazing probably 30-40 years.

To make same amount of solar power per year I need a single $130 NZD panel.

That's an interesting way of looking at it. I remember in the 70's baseboard heaters were very common. They use a lot of electricity, but electricity was super cheap back then. It would be interesting to compare baseboard+extra solar to heat pump+less solar. The baseboard is more reliable, so potentially would last longer.

Full on certified passive house looks very silly with solar this cheap, and frankly the math used to figure out a lot of the requirements is basically faulty.

2025 Code minimum is pretty decent if it's actually complied with, and 'net zero' middle ground with triple glazing is a worthwhile upgrade.

Wood burner is the best companion imho.

I would use wood power if I was offgrid and needed a storable fuel backup. It is less than ideal, but if you live in a cold place and live off grid with trees around you likely already burn wood for heat.

Please note if your goal is minimizing carbon footprint, burning wood instead of other biomass is probably not a great idea [1].

[1] https://www.pfpi.net/carbon-emissions/

It's literally burning surface carbon that's part of the regular surface land, sea, air, carbon flow that's existed for all of human history (and human existence).

It's not adding to that cycle by reaching down into the depths of the earths crust to bring up carbon captured and sealed away for longer than human existence .. you know, that additional carbon that is referred to when increased carbon footprints are seriously talked about.

This distinction only makes sense if those threes were going to be burned anyways. A can of diesel in a good generator and letting the trees decompose or be used for lumber should be far better in terms of emissions than burning the required amount of wood.

> far better in terms of emissions

Particle emissions isn't what I responded to .. in terms of carbon and greenhouse gases what matters more is trees not being replaced.

In the course of, say, plantation growing timber for lumber generates sufficient burnable wood for landowners and a wider community - the final lumber trees are the ones that weren't weeded out earlier (and burnt) and have been routinely lopped of branches (more burnable wood) to minimize knots, etc.

Forrest management is a thing, timber for lumber, coppicing for regrowth, et al has been going on for several thousand years and has been part of traditional surface carbon cycle.

As has large scale grassland (and forest undercover) burning off for fire management.

Wood that decomposes is wood that is turned into CO2 anyways. If you are selectively harvesting, and preferring dead or dieing trees, the only "downside" is that the forest itself doesn't occasionally catch on fire.

The firewood can be harvested as part of a coppice rotation too, e.g a woodland broken up into 8 coupes with one harvested selectively each year, then starting again at the beginning after regrowth is sufficient. Friends of mine do this and it works well. They replant as necessary as they go.

Does anyone actually use generators for primary power rather than backup? Even the really nice Generac propane backups are crazy noisy. I was in a neighborhood on Cape Cod during a power outage and because about 1/3 of the houses had backup generators going it was unpleasant to be outside.

I live off grid in inland northern California, and a have a solar panel array large enough to run my air conditioner continuously while the sun is up. It's just large enough to run the blower for a gas furnace in the winter so sizing turned out to be pretty even. Using a heat pump in the winter would require several times larger panel array, and running it at night would require the battery be much, much larger.

My 5 kW solar array and 24 kWh battery is fine for my 1300 sq ft house, all summer long, even when it's cloudy, and it works great for clear winter days, but as you mentioned for extended days of bad weather, the battery runs empty. Fluffy white summer clouds aren't a problem, but thick winter rain clouds let in so littl light that the whole array can't even run a 200 W refrigerator, so a few days of rainy weather depletes the battery and I have to top it off with a generator.

A 1 MWh battery isn't actually that big. There's electric trucks on the market right now with 600 kWh batteries sitting on the frame between the front and rear axle. That would easily fit into a basement room.

A good size comparison might by approximately the size of a stand up freezer, a common basement sight.

That small? I was imagining burying a shipping container under the driveway/front yard or something.

Most grid-scale batteries that large will have bigger inverters (usually it'll be inverters that can dump that energy in 2-6 hours so 500-150 KW for 1 MWh of battery) and require cooling systems and such, but if you're putting that in your home then cooling will be negligible and the inverter will remain small. The batteries themselves are fairly compact, it's the support systems that get large.

What's the degradation cycle on batteries that size?

If you're using daily, do you get... three? five years?

Looks like - https://cartroubleshooters.com/how-long-does-a-tesla-home-ba... - ten to fifteen years with a guarantee of 10 years at 70% from Tesla

We're to the point calendar aging instead of cycles is the existential threat.

We can roughly estimate lithium ion batteries as 500 watt-hours per liter. Which makes a million watt-hours 2000 liters, which is two cubic meters. Add in extra overhead and it's still not all that much.

Shipping containers should never be buried. They are not designed to tolerate forces pushing in on the sides and can collapse.

I wouldn't want a battery in my basement. if there is a fire in the battery your house will turn into a smoking hole, in the literal sense. Maybe if it was an iron-air battery or something safe, but not the current generation chemistry batteries.

Yeah, I'm not saying it's a good idea. But you could do it if you wanted to.

" but not the current generation chemistry batteries."

Check for "Saltwaterbatteries", they are starting to reach consumer markets and literally cannot burn as the energy is stored as ... salt water.

Seems like the peak was around 2017 but they never performed particularly well?[1]

The problem is if the promise from the name was true, they'd be everywhere - they're not, so invariably much like vanadiun-redox or iron-flow batteries it turns out all the other details make them more expensive and less performant.

[1] https://en.m.wikipedia.org/wiki/Aquion_Energy

" if the promise from the name was true, they'd be everywhere"

Not necessarily. Lithium is still quite cheap, safety is not the number one demand - and it is mainly about optimizing production to achieve competive pricing.

So yes, mabye there are some blocking details I am not aware of, but otherwise I expect their time will come.

LFP should be about as safe as other stuff people tend to store in their basements.

LFP's biggest issue is people aren't used to "Phenomenal Cosmic kW/h... itty bitty living space!" Yet.

Even if it was a safer chemistry, I would still put it outside of my house, likely in a dedicated structure. Its a shit ton of energy potential regardless, and it can make maintenance and modification down the line way easier.

A sand battery would work[0] if you have the space ;)

https://polarnightenergy.com/news/worlds-largest-sand-batter...

It's possible to store energy for full seasons, just not electricity.

It's been done with heat. Using cheap electricity in the summer to generate heat and store it in basalt. There's a small block of houses in The Netherlands that gets their heat that way: https://www.ecodorpboekel.nl/basaltaccu-is-opgebouwd-uit-duu...

There's more systems like this around the world, although they use different storage methods.

https://en.wikipedia.org/wiki/Seasonal_thermal_energy_storag...

Berlin just uses a really big thermos can to store heat: https://group.vattenfall.com/de/newsroom/pressemitteilungen/...

Off grid is silly unless you actually require it. Massive PV overprovision to ensure there's always something on the table is better than insane battery capacity. A couple of weeks worth of storage is a wild amount for a normal household.

I have a 22*980Ah 3.2Vn LiFePo4 array, and it holds a theoretical 13kWh at the 60% "safe" cycling rate (not below ~20%, not above ~80%, 3.0V min to 3.4V max). Taking DC->AC conversion losses into account, that ends up around 11kWh of 230VAC, which is enough for a single "normal" 24h period without generation: that doesn't include hoovering, welding, or running the dehydrator or dehumidifier. The batteries alone were USD$3500; BMS, balancer, cabling, etc. hundreds more. If I take $4000 as the unit price, then 14 days worth of power for us would represent $56k into a depreciating investment. I don't think most people are going to go for that. $56k would pay a lot of electric bills.

I'm in Ireland, which is fairly temperate, and we heat with wood (including the hot water). If you heat with electricity and you want to float that load on battery through a dim February...brutal.

EDIT: holy shitballs, that's $141,189.74 if you buy it as Powerwalls from Tesla rather than parts from Alibaba.

Huh, your numbers are weird. 22x980Ah of 3.2V cells would give you over 60kWh.

And the price of LiFePo4 continues dropping, it is not a good deal if your cells are aboe $80 per kWh (which at 22kWh should be below $2000).

Fuck. This is what I get for posting on 3 hours of sleep. My hallucinated bullshit came out so close to actual that it looked correct enough.

3.2v nominal per cell, 305Ah capacity: .976kWh per each. Call it .98. Not 980Ah, but 0.980 kWh.

.98 * 22 = 21.56kWh total pack cap.

*0.6 = 12.936 kWh available before conversion losses

We burn about 11kWh daily, so there's about a day in a full battery. Spring/mid summer worked well, but lately we aren't managing to store enough to get through the night, so I will probably add another 10 panels (5s2p) once I can find a grand I don't need.

I knew the numbers had to match something! Anyways, thanks for sharing.

And please take some proper rest, you protect the batteries using only 60% capacity, but allow your body go below 0%. I'm pretty sure thats out of spec.

> $56k would pay a lot of electric bills.

In california, if you have AC and electric car that's 56 months.

EVs cost ~$2 to charge a day for an average amount of driving.

The cost for AC dwarfs the cost to charge an EV.

My bill was €316 bimonthly at €0.31/kWh before I put in panels. Last bill covering the two sunniest months I've ever seen in Western Europe was €100 exactly, of which in excess of €50 was the standing charge that is due regardless. The remaining ~€25/mo overage is the delta of our demand and the inverter's peak output. Next purchase will be more panels; another inverter would offer only marginal gains.

Is electricity insanely expensive, or do californians use an insane amount of electricity? I just received my yearly bill in central europe, and $56k would pay for approximately 51 years, I'm having trouble reconciling the numbers.

Don't think reasonably sized house, think mansion with 6 bedrooms, a hot tub, a sauna, and a multiple EVs. With peak rates being $0.71/KWh, $1000/month is only ("only") 1400 KWh. High, but for a household of 5 adults, each with their own TV on top of everything else, it's not inconceivable. There's also a discount if you're poor, which translates to a surcharge if you're rich. A smaller household's bill is gonna be a fifth of that, closer to $200 than $1000.

To give a concrete datapoint, last month's bill was $820 for just the electricity generation and transmission (so no gas included) - that was for 1900 kWh for 1 month. That handles a 2000 sqft house with central AC, with two EV cars that are driven about 20mins/day, and we only charge those at night when the rates are low (low=$0.3/kWh)

That is what I did. 90kWp PV to power a house with 6000kWh yearly consumption. It works. And this was really cheap compared to buying expensive batteries that need replacing after 10 years. The 90kWp were about €90,000.00. I installed the modules and DC cabling/inverters myself and let the electrician handle the AC work.

Most people don't have enough land for that unfortunately.

Also, ironically now you have enough power to go fully off grid, it doesn't make any sense to do that because you can't sell your mountain of extra summer power!

Yes, going off the grid would be a waste of resources and somewhat anti-social. I calculated that I would need a minimum of a 300 kWh battery to fully bridge the dark days in winter. This is not sustainable. I would much rather see all my neighbours' electricity prices go down because of my energy input to the local grid. This would help to connect people and combat jealousy regarding the unfair distribution of available land and space.

That's a lot of solar panels, right? Where did you put them? I'm currently shopping for a house and am thinking of having enough of a backyard for pv.

A solar carport is on my wish list: https://www.youtube.com/watch?v=B0lCJ0cK6hg

Several garages inherited and mostly not in use, a large backyard, half of the greenhouse (my mother complains that the tomatoes now don't grow as good), the summer kitchen/garden house, the main house. 96x Solarwatt 60M-310, 160x Heckert Solar NeMo 120. Pretty much all buildings and everything I could fit in. See some pictures here [1].

[1]: https://www.photovoltaikforum.com/thread/226950-bericht-60kw...

> It's much better to massively over-provision the solar so you have enough energy - on average - for the winter.

Keep in mind that the UK is really far north, much farther north than most people expect, really, and the nights are really long in winter (there's up to 16 hours of night in London in December). And even your daylight hours aren't going to be sunny in average

So you'd have to over-provision a lot (like 10 or twenty times).

One of the best ways to ensure you have energy is to reduce use and dependence. A huge amount of energy goes to heating and hot water so insulation and shorter showers with on demand can drastically reduce battery and solar panel needs.

Seasonal storage sounds elegant, but in practice it's a nightmare of diminishing returns

If you have good TOU rates (some even offer free) it's far cheaper to skip solar and just get a battery - they are options for less than $100 / KWh. That's $1.5k for battery + inverter + 1-2 hours of labour. Equivalent solar system can cost 10x that.

Do you have some suggestions for manufacturers? Thank you.

Tons on this channel https://www.youtube.com/@WillProwse

Reply on news.ycombinator.com