Not paying electricity for an entire year with Tesla Solar Roof [video] | Hacker News Reader

Not paying electricity for an entire year with Tesla Solar Roof [video]

youtube.com

43 pointsteleforce58 comments

Meh. Tesla has been randomly changing the price of a solar roof. I can't remember the last quote they gave me, but it was slightly less than the second quote which was twice the price they initially quoted.

While solar roofs are cool, I sorta need to know how much it will cost and I need that cost not to change between the time we sign a contract and the time 8 months later when Tesla decides they don't have enough local installers in my area and cancel the contract (and mysteriously deliver a an unsolicited quote for twice the original price.)

So sure, if you're a YouTube influencer or Tesla employee, you can get one installed. But the rest of us schlubs will have to make due with unfashionable solar panels from reliable installers.

from the looks of it, his CTs are mis-reporting/double counting. A very common problem with tesla installs. My solar roof had the same issue initially, and electricians really thought it was generating more then the rating of the inverters, until I fixed it (software controlled). With 3 inverters he can't have more than ~23kw AC (assuming each is 7.8kw, which is the largest inverters tesla had at the time). My solar roof is a bit smaller (~21kw), in a similar region, but the daily peaks are slightly less than half of what he was showing.

>My solar roof had the same issue initially, and electricians really thought it was generating more then the rating of the inverters, until I fixed it (software controlled).

What was the software fix? Was there some config that wasn't set correctly?

yeah exactly, had to go into "installer mode" and configure the CT that's on a secondary powerwall, so it wouldn't count that value as part of the generation (forget the exact cryptically named option). Otherwise it considered energy pushed towards that powerwall as "additional energy in the system", and since the grid ct didn't show it was coming from there, it was assuming that the generation must be coming from solar. Or so it seemed.

200+ kWh per day consumption. In ~6 days this guy has used more electricity than me for an entire year. I get he also charges his electric car, which takes up a big chunk, but it still blows my mind how much electricity the average American home consumes.

There's nothing 'average' about using 200 kWh/day in the US. Where I live in the US, the average temperature (not the high!) is in the upper 80s right now. My home is not large by US standards, but it is old and not nearly as well insulated as a newer home. The A/C runs continuously for hours on end each day, and right now we are using 35-40 kWh/day.

So yeah, I'm pretty sure most of that 200 kWh/day is going to charge the car.

200kWh is about 700 miles of charge (and he says he doesn’t charge every day). There’s something weird going on, but it’s definitely not typical usage.

I suspect an instrumentation/configuration problem before I’d suspect an average usage of almost 5x mine.

rich guy with a big air conditioned house?

Or rich guy with a heated pool.

Edit: Sorry I made a stupid mistake. Should be 30kWh/day! That tracks much better with your own figure. Ignore the blow.

---------------------------------------------------

You're right, most use less because they don't charge a car. The average seems to be about double that though at ~74 kWh/day: https://www.eia.gov/tools/faqs/faq.php?id=97&t=3

Taking that average it would take around 2 weeks to reach my yearly consumption. Should be noted that's for something like 2.7 people, mine is a two person household.

charging a car is ~20kwh a day, not 150

even if he fully charged the car from 0 to 100% every day, that isn't half of 200kwh

And the fact that he gets it for free, but the electricity company still needs to build out the network (and profit) can tell you: solar power is going to be taxed heavily.

Which, ironically, tells you the major advantage of the tesla solar roof: not easily visible that it's solar panels, and thus easy to avoid tax with.

> not easily visible that it's solar panels, and thus easy to avoid tax with.

I doubt that. If you're consistently consuming a lot less electricity than your neighbors, that will be pretty easy for the utility company to notice, at which point they can send a technician out to your house to have a look. They probably already have to do this from time to time to investigate people who engage in meter tampering.

Around 20 minutes in, he gives his year total as around 54MWh (which is still insane to me, but less than a third-higher figure would be).

Looking at my bill, we average between 36kWh/day and 45kWh/day on all months except August (where I do put in the window shaker ACs and the usage soars to 65-70kWh/day). That’s for a family of four, lots of computer usage, working from home, and a 250W basement computer rack (6kWh/day on its own).

You are using about 1200 kWh per year? That seems quite low.

My desktop computer (2017 27" iMac), router, network switch, a Thunderbolt drive enclosure with 3 SSD drives, and a Hue hub averages 80 watts. That's 700 kWh per year right there.

If I'm quick in the shower I'm done in 3 minutes, using around 6 gallons of hot water. It takes 740 watts to heat 6 gallons of water from the temperature it comes out of my well to my shower temperature and I shower at least daily, so if my electric water heater was 100% efficient and its insulation was so good that there was no loss of heat from the tank to surrounding air that would be another 270 kWh per year.

My electric clothes dryer is 5600 W. That takes about an to dry a week's worth of clothes. That's another 290 kWh per year.

So that's brought me to 1280 kWh per year for just some modest computing, a short daily hot shower, and drying my clothes after washing.

Add in a 50 watt fan to circulate air in the house, running 12 hours a day. That's another 220 kWh.

How are you getting by on 1200 kWh per year?

Setup: Tesla Solar Roof (I thought it was discontinued) + 3 powerwalls + net metering 29 kilowatt array, 40 kwhr powerwall.

Cost: 120k with 30k credit, projected payoff: 9.6 years

Location: New Jersey

Usage: Tesla car, electric AC, but gas heating (and probably gas water heating)

He could probably go electric water heating too, because he is running a surplus. Full electric heating in the winter maybe, depending on insulation.

He pulls from the grid in the winter, but is net-negative in all other months.

The cost seems high, he says solar panels could be 19 kw total. The interesting thing is that by his viewpoint, as much solar as possible is good because as much of your appliances can be put on electric.

The near-future view of this seems to be (this is not in the video, this is my personal opinion):

- perovskite + silicon solar panels will increase efficiency and drop costs

- 150-200 wh/kg sodium ion batteries will drop the cost of the storage arrays by 50-60%, but the utilities will continue their political war on net metering, so you'll need a bigger array

- heat exchangers will drop the energy need of heat/AC so it can go electric

I think the federal incentives kind of still suck. They need to make this a no-brainer option for almost all homes, be it financed zero-down or similar schemes. Right now this is a luxury feature, but that's not going to get us to low-carbon energy independence.

There is also a housing crisis practically everywhere. The Feds will eventually stimulate construction to increase supply, why not make all the new supply solar enabled? Especially since the new supply is generally intended for the newer generations?

I'll retire tot the corner to keep taking the hallucinogens that gave me that temporary optimism.

> the utilities will continue their political war on net metering

And they'll probably win, because it makes no economic sense for a utility to provide a valuable service (reliable energy storage) for free.

It's worth noting the delta between this price, versus buying more normal panels on the open consumer market - for similarly optimistic reasons!

Based on a couple of panels I recently picked up on Aliexpress, they're running around 550-600 EUR (delivered) per kW, based on 200w panels. (You could probably go lower with more careful shopping and/or a bulk discount.) At these prices, a 29kW array (like MKB's) would cost 16-17.5k EUR, panels alone.

Of course, this price is without fitting or any other required parts (noting the Powerwalls especially!), would be considerably uglier on your roof than Tesla's offering, and would likely not even fit on the same roof area (so you'd have to settle for fewer kWs) - but it's exciting how cheap solar panels have become - and looking to the future, they'll become even cheaper and more efficient over time.

Funny that he complains about the 5.75$ per month the electricity company charges. You still used the grid, so how are they going to maintain it? How to pay the guy who fixes the power lines in your street?

This is actually the way it should be, and obviously not the way it's done here in Germany. We don't have net metering "because you'd be using the public power grid as your private battery then". I already pay 15€ every month as base fee, and it's not like the power grid is wearing down like a tire if you use it more. (Yes, technically it does, but practically negligible). Also if there would be more power produced in my section of the grid than gets consumed, I simply couldn't push any power back into the grid, so it's not like I'd get paid for producing power that nobody is consuming.

But no, not in Germany... current prices for consuming is around 30-40ct per kWh, while you only get payed around 6-10ct per kWh depending on when you got your system installed. And that only applies to bigger setups on your roof, the smaller ones you can mount on your balcony for example, you get nothing. Energiewende made in Germany.

Presumably, the smaller ones you mount on your balcony end up paying you the full retail rate for electricity you generated and consumed locally and that the grid never even knew about, right?

Theoretically yes, but that requires you have an old meter. Anything installed since ~2010 is either digital or at least has a mechanical blockage that stops it from running backwards. I think there is even a plan to update every single household to a digital one in the next few years

A small balcony unit is unlikely to surpass the electricity demand of the household. If it doesn't, your meter would never be "asked" to run backwards.

Mine often does as I'm not at home during peak hours. Around 1pm I get about 200W currently and the only things running is my router and a tiny home server, and occasionally the fridge kicks in at about 100W. I could have installed about twice as many panels, but why should I?

I already schedule my washing machine to run around that time, but that's like once a week, and the dish washer every two to three days.

Now wait for the Electrical Companies to change the reset month to Nov. Then you have to pay Nov, Dec, Jan bill and get credits for summer months. Only for them to get reset again in Nov. An MBA will be made partner for this suggestion.

Depends on the company. PNM (Northern New Mexico) doesn't ever reset unless you opt to get paid cash. If you decline that, you get a (so far) endlessly available energy credit, with which you get 1 free kWh for every banked kWh, so it is a much better deal than the cash paybacks, at least if your use exceeds your generating capacity for part of the year

People talk about how the grid wouldn't be able to handle everyone switching to electric cars. I wonder how residential solar (which works better in some places than others obviously) fits into that.

Meanwhile I'm over here in my 100+ year old house wondering which part of my tube-and-knob cracked-rubber insulated electrical system is going to fail if I decide to use the microwave... there's plenty of interconnected pieces in an electric system that just aren't built for the transition, and there's a lot of work in getting them ready that might not be immediately obvious.

That’s true for any transition.

The EPA is considering declaring no safe level of lead in buildings, and that is going to fuck with a lot of older homes. https://www.nytimes.com/2023/07/12/climate/epa-lead-dust-hom...

Over time, will we cover all the roofs of the world with solar panels?

I would guess that a solar roof generates enough energy to cool the rooms below it, no matter how high the temperature rises on the outside. So with climate change continuing and solar roofs becoming more efficient and more affordable, it seems inevitable.

That depends. Climate change might very well (seasonally) increase cloud cover, vastly decreasing the power produced.

Hypothetical

It is physical fact that each degree of temperature increase means the air can transport 7% more moisture and recent studies confirmed that the change in the last decades perfectly match the theoretical value.

price per watt decays similar to moores law. thus cloud cover, latitude etc soon become negligible though

Night time temperatures are also thing to consider. Not sure how viable thermal sinks cooled during sun-light hours would be...

Sounds more difficult to implement than battery storage to keep the heating pumps running at night (given that we already have all that working well).

Possible, but you're only moving heat. You're locally cooling, but the net effect of AC is more heat. If a lot of that heat can get out of the atmosphere through radiation, then we'll start to cool the planet. Greenhouse gases make that difficult.

    the net effect of AC is more heat

Hmm... I don't think so.

Yes, operating a fridge in your kitchen will heat up the kitchen. And yes, if that fridge is powered by burning coal somewhere, then it heats up the planet too.

But solar energy is already coming in the form of energy. It is not as if the sun is raining coal down on earth and we decide if we want to burn it or not.

I would think the amount of radiation given of by a solar powered AC will equal the amount of radiation it takes in.

You may want to retake a middle school education on "work" and "thermodynamics"

Before you get snide, consider what you are proposing: that 1W of insolation can cause >1W of power to be generated as waste from a solar-powered AC system.

It’s pretty clear that this is not the case.

> "thermodynamics"

Isn't part of that 'the law of conservation of energy'?

The basic principle of the greenhouse effect is that light from the sun hits the earth. Some of it is reflected out into space (because the atmosphere mostly lets it, some of it is absorbed and converted into heat. The heat can be radiated as IR but IR tends to get trapped by the atmosphere.

Solar panels are all black. So from a very nitpicky perspective, the energy used to power the A/C creates heat that is not radiated in the same way as, say, having a giant mirror on your roof would. The heat is also released locally rather than being aimed at the sky.

But from an overall perspective you are correct, because the alternative to solar is just having a regular roof which will convert some of the light into heat without doing any useful work, and because you'd probably get your energy from some less clean source.

The Earth's climate is broadly stable (a random walk within certain ranges for hundreds of millions -if not billions- of years!) because of the water cycle. The water cycle need only speed up a wee bit to cool the planet if forcings get a bit hotter, and the water cycle need only slow down a wee bit to warm the planet if forcings get a bit colder. Condensation leading to precipitation radiates a great deal of energy into space. Clouds also increase the albedo, though at night they also block the escape of radiation to space from the surface, so this effect not as large as the condensation effect.

At reasonable latitudes in secure enough countries? Possibly. A lot depends on balance-of-system: the cost of inverters, electricians, and permitting. Even if the solar panels were no more expensive than regular tiles. But it's a very simple capital investment with a real payback. That then starts making time-of-use metering important, too.

How many years before you pay off your roof and when will it need to be replaced?

If you watch the video, 9.6 years to payoff and 25 years to replacement in this particular case.

And I imagine 25 years from now, the new solar roof be cheaper and more efficient.

Prying the old tiles off of the ruins of an post-apocalyptic wasteland can be very cost efficient.

It also makes sense to use freely available slave labour.

25 years is just what they guarantee, it probably lasts much longer.

My solar roof (not Tesla, of course) is meanwhile 21 years old, and works almost like on day 1. Since I dont expect a sudden failure in 4 years, my guess it that it will last another 10-20 years.

What leads you to believe a solar roof will last "much longer" than 25 years?

>Since I dont expect a sudden failure in 4 years, my guess it that it will last another 10-20 years.

Oh, you're just guessing.

For the record, I had 4x some small panels (165W) from a reputable company that shall remain nameless. They had a 10 year warranty.

Turns out that (a) the 10 year warranty actually meant "generating at least 90% of rated capacity in full sun after 10 years, but pro-rated across 10 years if not" and (b) they fell under this level after less than 5 years.

This is what I'm curious about. There seem to be really large differences between panels. Got two 150W panels for 78€ each this year. Peak output so far was around 240W, if performance stays like that for about five years I might break even in about 5 years for the full setup.

Bear in mind that most residential roofs (of any sort) only last 10-20 years anyhow.

This question can't really be answered without knowing the location. The differences can be significant.

It really depends on where you live. Aside from the obvious electrical cost savings, you typically have state/federal incentives plus the opportunity to sell back excess power to the grid. In addition, and most importantly, there may be SREC markets that pay you to generate EVEN if you consume everything you produce. In Washington DC for example, my $30k traditional panel system should be paid off in about 5 years due to (ordered by magnitude): SREC sales, federal tax credits, excess power generation sold to the grid, electrical savings. Currently in year 3 and about 2/3 paid off.

I'm afraid that in my location by the time the roof pays off, the roof will be so badly polluted with pollen, moss, and residue from air pollution applied by the rain, that the fall in efficiency will render the entire system inoperable.

How hard is it to rise/wash it off?

Depends from the flora in the region. Once lived in a city district with trees producing sticky syrup (maple?) and yellow powder (birch?). This mixture was penetrating every aperture in e.g. a car. Went to car wash, the next day car was sticky again. Spruce and pine pollen result in a moss-like layer on the roof quite quickly.

Reply on news.ycombinator.com