Item — HN Reader

> The theory goes that as solar panels and battery prices drop, it will eventually make sense for homeowners to produce their own power, and "defect" from the electrical grid completely.

I think it already makes sense in California (given the amount of sun we get).

A Tesla solar install of 8 kW solar production + 2 PowerWalls (27 kWh capacity) has a loan payment of $188/month, which is probably lower than what your cost would be from the utility company for an equivalent "level of service".

The only tricky part is to figure out the answers to "what size panels + storage do I need so that, given elements (weather mostly), the probability of my storage being depleted any point during the day is less than X%".

Solar companies probably already know the answer to this given the data they already have, but the other part of the equation is changes to the owner's consumption (buying an EV? new child on the way? changing your range/dryer/heaters from gas to electric? etc)

That's the real pain point (it can be mitigated with a generator, but that gets complicated), which leads to a certain amount of oversizing, and as a consequence, waste (since you can't feed it back to the grid).

Also the occasional failure in your own infrastructure could lead to blackouts that could be more frequent than if you were tied to the grid (Mean Time To Repair is higher with your own infrastructure for obvious reasons), although I have to admit that the grid hasn't been very good in that respect. A solar system is generally maintenance free, but failures do happen.

Still, I think it's worth it (that's why I took the plunge a few years ago).

3 comments

You can do a lot cheaper than a PowerWall. But you can also depend heavily on propane if you are off grid. You can get propane based central heating, propane stove tops, propane refrigerators, propane freezers.

Going to propane also helps with the overbuilding, because that's more or less pay-as-you-go instead of having to build a finite supply all at once.

It will almost certainly be necessary to significantly overprovision both panels and batteries in order to avoid running out of power or needing to add more later. Hopefully the cost decreases will reduce the financial headache of this... time will tell.

I think there is an unrealized business model out there where a company builds, installs, and then maintains (incl. emergency fixes) solar+battery installs for a monthly fee, just like what you pay to the electric utility today. Such a business would hopefully achieve some cost savings by buying parts in bulk and be able to quickly fix issues by using similar system designs. It would be a bit like the electric utility, but they wouldn't own any grid or power plants.

What's the term of the load in your calculation?

A PowerWall has a 10 year lifespan, where they guarantee that batteries will have 70% capacity or better during that time. I take this to mean you'll need to size for a 30% capacity drop, and plan for battery replacement in 10 years.

Does your loan factor in installation costs, which can be significant?

Since the goal is to reduce the chance of fire, what would the probability of starting a fire be with thousands of powerwalls installed all over the place be in comparison?

> What's the term of the load in your calculation?

What do you mean by that? The amount you can draw from the power walls?

This page has a chart with specs that cover this: https://www.solarquotes.com.au/blog/tesla-powerwall-2/

I think each powerwall can draw 5 kW continuous and 7 kW peak.

> A PowerWall has a 10 year lifespan, where they guarantee that batteries will have 70% capacity or better during that time. I take this to mean you'll need to size for a 30% capacity drop, and plan for battery replacement in 10 years.

Well, it has a 10 year warranty, not life span. Yes, for sure there will be some degradation with time (although for my car, a Model S, I'm surprised that after close to 5 years and 25K miles I barely "lost" 1% of range). It depends on how you draw and charge.

> Does your loan factor in installation costs, which can be significant?

I believe it does, and also includes the rebates/credits. But you can answer those questions yourself by going to their website and doing a "pretend order" - that'll give you a bunch of knobs to play with and see the various options between buy cash vs loan.

> Since the goal is to reduce the chance of fire, what would the probability of starting a fire be with thousands of powerwalls installed all over the place be in comparison?

Well, the goal isn't just to reduce the chance of fire: it's also to offer availability at reasonable cost.

Some will also like the "independence" aspect of it, like "stick it to the man" - but that isn't particularly important to me.

But back to your probability question, I frankly have no idea. I imagine we'll learn it from deployments, like we are learning about "spontaneous fires" that have been happening with electrical cars.

It's certainly a tradeoff, but one of the interesting things about this technology is that a hot environment is generally the time when you can produce good energy with the panels, which you can use to cool the infrastructure that needs cooling (batteries for example).

That being said, there is no 100% safe system. We're dealing with Kilowatts of power moving across a bunch of lines, so we have to be humble and realize we can't control everything to a zero risk environment.

What we can do, however, is measure, learn, and get it safer iteration after iteration.

> What's the term of the load in your calculation?

Sorry - misspelled 'loan'.

Thanks for the link!