Hacker News Reader

Comment by jacquesm | Hacker News Reader

I've done a lot of spreadsheets on this kind of solution, but on a slightly smaller scale: a single family home. While it will work for a while it is not enough to meaningfully offset the seasonal cycle which is the thing that needs solving. Storing energy for a few days up to two weeks is (relatively) easy, storing it cheaply for up to 6 months is very hard unless you are willing to invest massively offsetting much of your savings. The pile of rock required to heat an average home for a couple of weeks handily outweighs the house itself. And that's without a double conversion, it is used and stays as heat, the idea was to moderate the leakage upwards as a source of heat by blowing air in a controlled manner rather than to convert it again.

So I really hope these guys will succeed where I can't even get it to work on paper, sometimes scale really is a requirement to make something work and this could very well be one of those.

View on HN

Isn't it cheaper to just buy more panels, enough to meet your winter needs? Then who cares if they sit mostly idle for most of summer; energy not used is harmless.

Then your storage model becomes "a cloudy week" rather than "a whole season", and the storage scale changes significantly.

This is certainly the solution in somewhere like Australia where the winter production is about half of the summer production. It gets more problematic in places further north/south where its 5x or worse and then you need quite a bit of land to produce enough power and battery storage than can go multiple days at least.

The difference between net zero across the year and 95% unlikely to need the grid on solar and batteries is a staggering difference, about 10x the panels and battery storage. But the equation changes drastically depending on local weather patterns and the solar irradiance difference from summer to winter.

How about keeping an ICE-powered generator for those several days in a year when the sky is stubbornly cloudy and dark? Could make sense for remote houses.

BTW keeping your house connected to the mains costs $10-20 / mo, even if you consume nothing. Connection cost is one-time, but likely several thousand.

> How about keeping an ICE-powered generator

Or an EV. I only need enough instant battery capacity to last a couple hours. That would handle the vast majority of outages I ever experience, all by itself. If it looks like it is going to be longer, I grab the cable from the transfer switch and plug it into my Lightning. That would get me through a hypothetical week long outage.

No, (1) you need too much space and I'm pretty much maxed out already (50 panels), during the wintermonths (2) the angle is really bad, (3) the days are shorter and (4) you're overcast a lot of the time. So that simply will not work. But in the summer I have so much excess I get penalized for it, which is ridiculous.

Once homeowners have done everything they can (as you have, in your example), the remaining optimization is to drive down storage costs (LFP is ~$52/kWh in China at utility scale) until seasonal storage is solved for (which is likely a combination of electrical transmission [to be able to shift low carbon energy from areas that are constantly plentiful] and longer term storage like ammonia or similar fuels created and transported [pumped around] with low carbon energy). Of course, if you have the space, keep putting solar PV up until it's no longer an issue when paired with heat pumps.

> electrical transmission

I think the current consensus is that transmission to solve for regional solar differences is not cost effective. At least in the US, most places have plentiful land available nearby and it costs way less to overbuild the solar for winter than it does to try and bring it from some place with more plentiful sunlight.

HVDC is definitely a game changer but requires international cooperation. North-South for seasonal and East-West for daily.

> the angle is really bad

Optimize the angle for the Winter. The Summer will do well enough anyways.

Assuming rooftop, there are too many cons with angled installations so it may not be realistic to change the angle. Flat mounted is the way to go (IMO).

Indeed, at this point I'd plan on going for maximum coverage, the efficiency loss from a non-optimal alignment is completely overwhelmed by being able to put half again or twice as many panels in the same area.

Flat mounted has some issues with drainage though leading to a messy surface on the low edge of the panel. You can clean that but it doesn't take much before it is really dirty again.

The panels themselves are cheap and getting cheaper.

However the mounting systems, solar charger controllers, and inverters are mostly not.

The bulk of the cost of panels now isn't the panels themselves, but all the supporting infrastructure.

The seasonal variation can be as high as seven to one, so the total install has to be very very cheap to support a 700-1000% over build.

The last time this was discussed, someone mentioned you only needed as much inverter capacity that you wanted to process. So in winter all panels were hooked up but in summer some were disconnected or bypassed somehow?

Though if you are on a grid that allows it, you might want to send summer power back into the grid for credits

There's basically two kinds of systems.

Micro inverters or mppt with inverters.

With micro inverters you cannot over provision the panels in a useful way, each panel or pair of panels has a micro inverter.

The mppt plus inverter system you absolutely can do that... but it scales poorly. Shading of a single panel effects the entire string, so the more panels you add to over provision the worse that effect gets.

So yes but actually no.

It seems like there should be some intermediary per panel smart bypass circuit for pairing up with mptt architecture for this to keep system cost down in these seasonal areas.

Edit: turns out there are products for mptt strings called power optimizers that seem to help with this very thing I hypothesized.. at least one company offering them is SolarEdge

They're almost as expensive as per panel micro inverters.

There's no free lunch here.

Panels are surface area limited. The maximum my house and land can support is the 10kW I already have.

Honestly I'm pretty skeptical of residential rooftop solar - it's not even close to the solution, it's just a cost optimization we shouldn't actually need.

In almost all cases you lack the area for doing that. And that's not even considering exposure, as in winter most north facing panels on a roof won't get any direct sun because it's lower on the horizon

Depending on your case. Where I live (NYC, a barebones Great Depression-era building) my electric bills skyrocket (3-5x) in the summer due to the house being heated by the sunlight. The roof is white, but it helps moderately. Heating in the winter takes much less energy, especially if we factor in all the electric devices inside that also release some heat.

Storing a half-day's worth of electricity, and powering stuff from a solar panel at daytime, would likely let my apartment stop consuming external electric power in the summer.

Assuming my apartment consumes no more than 15 kWh during the hottest days (according to the meter), a moderate 8 kWh battery, and $1000 worth of solar panels, would suffice, given the room and a permission to mount the solar panels somewhere.

You could potentially skip the storage by cooling aggressively on the solar and then not cooling when it is dark.

> The pile of rock required to heat an average home for a couple of weeks handily outweighs the house itself.

Maybe I'm failing to follow the intended argument here, but I do not see what is expensive about this. Houses are hollow; they do not weigh all that much. Dirt is cheap, especially when sourced locally.

The only place where you could conceivably store this without having another conversion or lossy transport loop is under the house. It's not exactly trivial to excavate that much space under an existing structure without having the whole thing come down. I think I've figured that part out but it would still be legally quite shaky and it might destroy the resale value of the house unless it can be made compact enough that it would fit within a standard basement. Insulation is the key problem to solve here.

The article never frames this as off-grid storage for individual homeowners. They say

> A typical site is a factory, power plant, or town with a large earthen mound at the edge. The mound might be the size of a house for a smaller factory, and up to many football fields for a large power plant. Surrounding the earthen mound will be high-density, low-profile solar arrays.

I agree that trying to give every suburban house its own rock pile would not be very practical.

> The article never frames this as off-grid storage for individual homeowners

I know. See my original comment at the top.

That depends on how hot you are prepared to get your rock.

I’m also off grid, also suffer in the winter (although a spot of hydro power is going to alleviate that somewhat), and have also thought about thermal storage - I actually calculated that a 20ft container filled with sand could give us thousands of kWh, as long as the sand could get up to ~600C.

At that temperature you of course have to think about making really damned sure no water gets in there, but it makes the volume of material more manageable. Would still work with resistive heating, energy out would be trickier but not impossible to engineer something that wouldn’t explode.

Insulation volume was the issue for me. Sand works well enough, the ideal setup would be - as far as I've figured this out - a sinkable construct that has insulation in the bottom and the walls that you can put together in a modular way by slipping one part after another into a basement. Then you sink it in the same way that you sink a well. Indeed, keeping water out of it essential but you can test where the water table is and keep a very healthy margin. The 'instant steam explosion' angle in case of a water leak in the house is one of the reasons I've never actually done it.

> I've done a lot of spreadsheets on this kind of solution

As an investor - what RoIC do you want to see when doing initial analysis

(for example, $10M capex per system, with 10,000 systems TAM )

I only looked at this from the homeowner perspective as a DIY project, but if I were to look at it from a utility scale commercial perspective I'd say the first thing I'd do would be to compare it to wind power because where I live that is probably one of the most efficient ways to generate renewable energy even if there is no storage option. It nicely dovetails with solar. When I did this at the residential scale in an off-grid house I built I managed to get it - fairly easily - to the point where we did not need a generator at all, even though we had one.

Reply on news.ycombinator.com