Hacker News new | past | comments | ask | show | jobs | submit login
When will rooftop solar be cheaper than the grid? (theconversation.com)
132 points by leejoramo on Mar 31, 2016 | hide | past | web | favorite | 83 comments

Where I live, it's already gobs cheaper. I live in Shizuoka prefecture in Japan. We get about 2200 hours of sunshine a year at a latitude of about 35N. I pay the equivalent of about 35 cents per kWh for electricity from the local power station (hey, they turned off all the nuclear power stations and the only fuel we have is gas from Russia).

You can buy solar panels in any electronics store. If you go to the mall, you will run into billboards and displays and salespeople wanting to sell you solar panels. Already, if I walk down the street at least 30% of the houses have solar panels. Lately virtually every single apartment building has put solar panels on the roof. All the farmers are putting up vast solar arrays because they can make more money doing that than growing food (which worries me quite a lot... food security is pretty important for a chain of islands off the coast of China).

I'm really curious to see how this affects pricing going forward.

Hawaii is another great example. $0.40-$0.50/kWh for their oil generated electricity. Solar comes in even cheaper than the article mentions. There have been so many solar installations in Hawaii that the power company has effectively put a moratorium on installations that don't include storage because they can't cover the swings in demand when a cloud rolls through town.

They expect things will get better once they install a connection between the islands to connect their grids, but for the moment, the power company can't handle the market penetration.

> the power company can't handle the market penetration.

When it comes to solar power, and anything else with externalities, it's always best to be clear whether "can't" means "has no economic incentive to do so" or not.

In this particular case I'd say we are below the optimum level of solar, and it's just poorly aligned incentives.

edit: just to re-iterate this in clearer language:

If they "can" do something if they spent $X million dollars, and $X million dollars is less than the total amount saved by the economy of Hawaii by doing so, then it's not a technical problem, it's an economic co-ordination problem.

If the utility is going to lose money from doing something, because they can't capture all the resulting profit (like lower pollution meaning healthier people) then their employees will say they "can't" do something, as no-one gets promoted for making children healther on the utility shareholders dime. Even though they could if someone gave them the money.

There's lots of BS surrounding solar, and a main thread of it is to pretend that it's going to blow up the electrical grid. The only thing it destroys is some current utility business models.

When clouds roll in, people are not OK with their TVs turning off. So you'd always need the traditional grid in place, even if it's for a few days a year.

In that case, the utilities would charge a standby fee for keeping their multi-billion dollar infrastructure in place (it's called a "demand charge" in the utility industry). Then, when you use the electrons during the cloudy period, you'd pay an astronomical "consumption charge", beyond $100/KWh because everyone in the city wants/needs those electrons. The utility share-holders would be fine.

But the people of Hawaii would have the cost of the solar installation PLUS the cost of the traditional grid. The calculation in the article misses this point and only shows the $/KWh consumption costs on the utility side.

So, yes, a huge amount of solar can figuratively blow up the electrical grid.

> When clouds roll in, people are not OK with their TVs turning off. So you'd always need the traditional grid in place, even if it's for a few days a year.

Sounds like a great argument for municipal generation and storage. If nothing else, it's both cheaper than "everyone on his own" and could lesser the strain for long-distance infrastructure (which could still be nice for geographic smoothing).

> So, yes, a huge amount of solar can figuratively blow up the electrical grid.

Demand-controlled EVs could take a lot of the load. In fact, it's probably one of the sanest applications of massive solar installations (beyond the potential of future power-to-gas tech).

There's a lot of ways to store energy if you're basically getting it for free, and these are generally more efficient at scale. It would be great if a utility could help buffer this for you.

Yes, that's another reason why a municipal grid would be useful: city-wide storage could work better than each house having a small battery (but the roofs are still useful for panels, of course). But it may not need to be a utility in the traditional sense (unless the US meaning of that word is different from what I understand it to be). Just some local company with a contract. Or a few of them.

Here the power company was split up into a generation unit and a distribution unit. What you're talking about is having a third type, a storage unit, to help manage demand.

If you can pull power off the grid very cheaply, even get paid to "dispose" of it, then later return it at a profit, you could arbitrage solar capacity.

Interesting example about the TV switching off. That would of course be inconvenient but there's already multiple large industries and consumer level smartmeters that will turn things off when electricity demand is high in return for cheaper energy:


The basic reason for this to exist, even before renewable energy was a thing, is that, as you point out in your comment, building enough capacity just to serve a peak that may only last for a few hours a year is stupidly expensive. So you take the money that you would have spent on a power plant that would only run for a few hours and you give it to people for turning down their air con or their blast furnace for a few hours, keeping a portion to yourself for being so clever. Free-market solutions that save customers and utilities money, save energy, and save the planet. Everyone is happy!

Well, consider that the conditions are quite a bit different. With conventional power, you have a small number of sources (plants) and relatively slow, predictable swings in demand. Things like temperature changes throughout the day, or the end of rush hour and resulting spike in entertainment electronics usage.

Contrast this with solar, with a large number of much smaller sources, and on top of the large predictable swings in demand, you add rapid, unpredictable swings in supply to the mix. That is a very different problem. It's solvable, just not using the same techniques and equipment that worked before.

While there is certainly a market incentives problem, accomodating large quantities of rooftop solar does require infrastructure upgrades and an associated massive capital investment. It may not be practical for a relatively small island utility to do this on their own, even if it had a more clear ROI. Ultimately, some government support may be required here, since a fee on solar installations wouldn't be popular.

Not to be too picky but it is actually the swings in supply they can't handle. Demand is inelastic and pretty much constant in the near term.

The problem is that the clouds roll in, supply from solar drops off, conventional gen must compensate because the demand is pretty much constant and price insensitive. When the clouds roll-out the dispatchable gen has to back down so the wires don't melt. There are a bunch of concerns with this: generators often don't like being turned on and off, the generators often can't get paid enough to be there when they are needed, they often can't just be turned on and off (more of a problem for steam), and these peak shaving plants are expensive to fuel, burn dirty fuel, and burn it inefficiently.

Swings in "demand to the utility company".

When clouds roll in, the solar panels still generate power, just less of it. A lot of things in the house could be automatically turned off during such brownouts - the refrigerator, the car charger, the HVAC, etc. - until the cloud passes.

I wonder if capacitors could be used to moderate such short reductions in solar power. Not necessarily at the power plant, but as part of the solar installations.

>swings in demand when a cloud rolls through town

Reminded me of this article about abrupt power swings during commercial breaks in the UK:


They solved it with partly with on-demand power generated from dams.

I live just outside the Sacramento area. Talking to a coworker, I was surprised to learn that SMUD's (Sac area utility) highest "tier" rate was just under PG&E's (most of the rest of CA) lowest "tier" rate. Once you use your "quota" of electricity (easy to do with kids), you move out of tier 1 base pricing ($0.18/kwh for PG&E) into tier 2, which is currently $0.25/kwh, then tier 3, then tier 4 (which I think is in the $0.35 to $0.40 range per kwh). So, putting in at least a partial system is a no brainer.

If you look on the price map in the article, there is a pale colored blob around Sacramento, which reflects that SMUD's prices are a great deal less than PG&E's. I don't know that solar makes sense if you are on SMUD.

During Spring break (nice sunny but not hot) weather, we generated a surplus (although we would not have if it were cloudy). I'm looking forward to being able to use the A/C this summer w/out getting a $300 to $400 electric bill.

I assume you still have to buy electricity from the grid during times when the solar panels can't output enough power, correct? Can you give a sense of how long the payback time is for buying the panels--i.e., how long before you save more in not having to buy grid electricity for at least some fraction of the time, than you spent on the panels?

Australian here. We'll trade you food for solar panels. Plenty of sunlight here, but panel adoption is slowed by the abundance of local coal.

How difficult is it to install and maintain them? Is it something that can be installed by nonprofessionals?

I used SolarCity to finance, install and maintain. Not the cheapest solution, but they are taking care of everything for $0 out of pocket.

Of course, I'm going to be paying interest for decades (and the loan has to be paid off or transfer with my house if I sell), but it is approximately the same per kwh cost, just to SolarCity instead of my utility, plus I can pay it off if I wish.

Geez, on re-read, I didn't mean this to sound like an ad, but they make it so easy.

If you think you're capable of installing something flat on the roof that can withstand strong winds, and then wire it into your house electric wiring without causing a short and killing everyone, why not?

I can't imagine an average person doing this.

If you are going to connect to the grid you need a licensed electrician. If you are not connected to the grid you will need a large battery bank and twice the number of panels to get you through night and cloudy days. This is not to mention building codes and making your insurance company happy.

The article assumes net metering, which is a huge subsidy for solar. Peak solar generation causes low spot prices for power (see the California "duck curve"[1], and the negative prices that Germany has on sunny days[2]. In Texas wind can cause energy prices to go negative[3]).

The costs of storage must to be included for a fair comparison, so that power can either be saved and used by the household, or provided back to the grid. If provided back to the grid, the homeowner should get paid the spot price like every other generator, not full residential rates.

[1] http://www.bloomberg.com/news/articles/2015-10-21/california...

[2] http://www.pv-magazine.com/news/details/beitrag/german-renew...

[3] http://cleantechnica.com/2015/10/01/texas-electricity-prices...

This comes up in every discussion about solar power, but it seems unfair to me. Yes, solar is subsidized by a whole bunch of factors. But so is coal and oil, because the suppliers of those power sources don't have to pay the full cost of the damage their pollution causes.

A fair comparison should either include all subsidies, or none. Not just the subsidies we are used to.

The problem isn't the subsidies. The problem occurs when the proportion of solar users increases such that the utility's margins become low enough to threaten the minimum budget required to maintain the infrastructure.

If you want to subsidize solar users to bootstrap that market and as a general social good, that's fine. But you have to structure it such that the utility still recoups the cost of maintenance. If they're paying market instead of spot, the paid price is not taking into account the cost of transfer and storage to the utility.

When we get to that point the answer is to let the local government take over maintence, much like roads. Subsidize it with a tax.

Energy storage also needs to be encouraged and developed to make solar successful. Like solar cells, the cost of storage will come down as a function of production volumes.

Net pricing does just the opposite - it disincentivizes storage. Is that really the type of subsidy you'd support?

There are still plenty of incentives for grid storage that have nothing to do with residential rates (peak shaving with pumped storage, day/night price arbitrage). Surely not all incentives have to work for every valuable technology.

> If provided back to the grid, the homeowner should get paid the spot price like every other generator, not full residential rates.

Sorta. Because solar is distributed the economics related to transmission are different. If your solar panel is mostly powering your neighbours air conditioner then you are reducing peak load on the transmission system which can be a very valuable thing. If your solar panel is powering someones air conditioner one state over then there is no benefit to the distributed generation.

The catch is that the situation can change from hour to hour and day to day. The proper amount to pay the small distributed generators can be hard to determine.

If you really want to balance out the grid to account for distribution and transmission costs, you still need storage, and net pricing disincentivizes storage.

Spot prices are already location-dependent, and yes a real distributed smart grid would do this on a micro scale, and a correctly designed smart grid would incentivize local storage. Which is what I think you mean by "hard to determine".

> If provided back to the grid, the homeowner should get paid the spot price like every other generator, not full residential rates.

I mean "should" depends on the goal of the pricing policy right? If the goal is a "fair comparison", perhaps the policy makers should include the cost of mitigating externalities in the price of carbon-producing electricity sources.

We should absolutely put a price on carbon, yes. That would penalize carbon emissions.

Net metering is no substitute for a price on carbon. For real green power, we need to storage to balance out solar generation. Net metering /prevents/ storage from being installed.

EDIT: Never going to happen? Actually it's /already/ happened. Examples:

- California already has a low carbon fuel standard[1] that regulates "well to wheel" carbon content of transportation fuels (some sources like tar sands have high carbon emissions in the production process[2]). This is a significant step towards a real carbon price and an opening to provide fuels generated through a green process.

- California' Emission Performance Standard (EPS) no longer allows use of coal-generated electricity[3] which has caused a complete phaseout of coal plants in favor of natural gas (which has half the carbon intensity). Again, this is a restriction of the carbon-intensity of power generation which is also an effective step towards a carbon price.

[1] http://www.wsj.com/articles/california-restores-rule-to-cut-...

[2] http://www.scientificamerican.com/article/tar-sands-and-keys...

[3] http://www.energy.ca.gov/renewables/tracking_progress/docume...

> Net metering /prevents/ storage from being installed.

In a well regulated electricity grid, storage is used at the grid level, not at consumer level. Net metering (which is good regulation) limits the latter, not the former, it actually increases the demand for grid storage by increasing the amount of solar supply.

Ok, let pretend that's never going to happen because it's not going to happen.

Where does solar win in the current political/economic environment and where is it close to winning? Making forward progress in the next decade is going to help more than wishing for a level playing field.

Maybe we can push solar costs lower once the industry grows:


Residential installation in Denver CO at $2.16/W after federal incentive. Just about exactly grid pairity. Their $3.50/watt number is way high compared to the reality of today in the US.

For what value of "never"? Carbon tax on the ballot in Washington this fall - http://yeson732.org/

You got me there!! The state that gets 75% of its electricity from hydro leads the way!


As of today, there are still 50 states to go, plus the rest of the world. Better sell your beachfront property. Time"s up.

>Net metering is no substitute for a price on carbon.

It absolutely is since there is no price on carbon and no realistic prospect of one any time soon.

$30/ton CO2e in BC, in effect today: http://www.fin.gov.bc.ca/tbs/tp/climate/A6.htm

Of course, that's probably not enough. See Greg Mankiw's "invitation to the Pigou club" where he discusses the research on the "correct" value (estimated at between $80 and $300 per ton): http://scholar.harvard.edu/files/mankiw/files/smart_taxes.pd...

>> If provided back to the grid, the homeowner should get paid the spot price like every other generator, not full residential rates.

The home owner then buys a few kW of battery storage and saves the energy for their own use later. This has the same effect on them as net metering once they pay for the storage unit, but only up to the point that they produces exactly as much as they use.

If you got the spot price, I wonder if that would encourage storage development independent of solar installations?

In other words, installations whose sole function is energy arbitrage, charging storage banks when power is cheap and discharging when it's pricey.

Yes if you consider utilities as a business that sells power for more than it pays, net metering is a subsidy.

But if you think a utility has a special responsibility to the public, and I for one do, it's hard to defend charging me more for a kW coming into my home one moment than I get for one going out the next.

Utilities aren't comfortable with the fact that generators will go out of business, and only transmission will remain as a non-profit ISO-type entity (coordinating energy distribution and storage).

>the homeowner should get paid the spot price like every other generator, not full residential rates.

Either the other generators should pay a significant carbon tax or the solar / wind producers should get a significant subsidy for polluting less.

You would of course need to include the externalities associated with the manufacturing of solar panels.

Which is minor compared to the carbon spit out by coal and natural gas plants, as well as uranium released directly into the atmosphere when coal is burned.

Just because your competition has an advantage doesn't mean that you can ignore costs. Even if you choose to ignore the costs, a subsidy is a direct monetary cost and not an advantage that you can just force to compete.

That argument doesn't make any sense.

Whichever way you look at it there's no justification for giving the same tax treatment to both coal and solar. One imposes an enormous externality and the other doesn't.

Whether it's a subsidy for one or a tax for the other is somewhat immaterial. The point is that they shouldn't be taxed the same.

Currently coal and gas actually receive massive subsidies, too, while a 30% tariff is slapped on most solar panels imported from China which is even more perverse. We're literally subsidizing pollution because coal/gas are tight with out politicians, while cheaper, cleaner forms of electricity generation are rendered purposefully less competitive.

> Whichever way you look at it there's no justification for giving the same tax treatment to both coal and solar.

The same? Not even, fossil fuels get significant subsides on top of paying nothing for the emissions and that's just the tip of the iceberg.

In June 2010, the U.S. Energy Information Administration (EIA) said $557 billion was spent to subsidize fossil fuels globally in 2008, compared to $43 billion in support of renewable energy.[1]


Worldwide, yes. In the US, not even close. According to the actual EIA report[1] (see PDF pages 15 and 16, table ES2) US government subsidies for fossil fuel power generation are a drop in the bucket compared to renewables subsidies. Solar alone gets more subsidies than all fossil fuels combined.

If these are the numbers that fed into the $557 billion and $43 billion[2] then it appears the US is responsible for a trivial amount of the fossil fuel subsidies worldwide and over 40% of the renewables subsidies.

[1] http://www.eia.gov/analysis/requests/subsidy/pdf/subsidy.pdf

[2] I know the numbers in the report are for 2010 and 2013, while the big numbers are for 2008. I'm assuming the other numbers have scaled similarly since then.

That report only lists, as it's title suggests: "Direct Federal Financial Interventions and Subsidies in Energy".

Harvard Medical School estimates that coal costs the US half a trillion (yes, trillion) dollars every year in externalities that the public, not the coal industry, or the utilities, pay the price for. The total accounted for in your document, for all energy types is only $40 billion, more than a factor of ten less.

You can have a fascinating semantic discussion about whether those half a trillion dollars, every single year, are a subsidy or not, but either way they're getting paid by the US public when a bit more "sudsidy" for solar could have avoided them.

The post I was replying to was discussing only direct subsidies, not externalities. The report I referenced is the one the poster's reference referenced. I don't see how an HMS report on the cost of externalities is appropriate to the discussion, considering the worldwide costs are certainly in the trillions if not tens of trillions.

I'm not sure I'm parsing this correctly, are you saying that the cost of coal externalities are irrelevant to this discussion, because they total trillions of dollars globally each year?

No, only because they are beyond the scope of direct subsidies.

Taking that as gospel, in a supposed 'free market' economy why does the US subsidize fossil fuels all? Does anyone have a compelling reason? Isn't absorbing the cost of the externalities enough?

Likely because the grid is important enough to the economy that there should be government support of the stability of the energy supply.

I had a 6.7 kW solar system installed here in the Bay Area late last year, and you can beat the costs stated in this article by quite a lot if you shop around. The best bids I got were around $3.30/watt installed -- that's $2.31/watt with the 30% federal tax credit figured in.

For my particular location, assuming a 25 year system lifespan, I calculate LCOE of 5.6 cents/kWh. Way, way cheaper than PG&E. Even without the tax credit, it's still a significant savings at 8 cents/kWh LCOE.

Of course this math only makes sense if you're reasonably sure you're going to stay put for a significant portion of those 25 years.

Even when you sell the house, the solar system's cost can be factored into the selling price. The new owner is getting the reduced electric bill for the next 30 years. Surely that can be reflected in the higher value of the house.

Yeah my system was way lower than $3.50/watt last year too. Their numbers are surprisingly out of date for being 2 years old.

Does this include storage?

No storage. This is a grid-tied system that exports power onto the grid during times when I am a net producer, for which I get credited at retail rates.

Retail rates? Are you sure? (PG&E?)

Yes, quite sure. However, the credits can only offset the "debits" from the times when you are a net importer from the grid (night, cloudy, etc). PG&E sends a once annual true-up statement where you pay for your excess annual kWh (assuming consumed kWh > produced kWh). If you are net exporter on an annual basis (produced kWh > consumed kWh) PG&E only pays wholesale rates for those at ~$0.04/kWh. They used to pay nothing if you were a residential net annual exporter, but after recent change in CA state law, they now have to actually cut you a check.

> Also, residential electric rates, on average about 12 cents per kilowatt-hour in in the U.S., are much higher than wholesale electric rates – the price utilities pay to power generators – which are usually less than 4 cents per kilowatt-hour.

That 8 cent difference is mostly the cost of distribution and storage or variable supply. Distribution from the roof to the home would be inexpensive, but storage is not cheap. Asking utilities to accept solar energy input at this flat rate in exchange for the variable supply and grid backup may work for a small fraction of society, but it will hit a critical mass sooner or later and become infeasible.

Has this been shared on HN yet? You put in your address and Google tells you how much you save by installing solar panels.


Cool site, thanks for sharing. I have never seen it.

According to that site a $20k install would pay itself off in 6 years. That is compelling.

I bought a $15k system in 2013, and if current trends continue it will have paid itself off 7 years after installation. I was interested in solar purely as a means of reducing my carbon footprint; I was surprised to learn that it had already become such a good investment, too. I'm glad to hear that the numbers have continued to improve.

(I could have gone bigger and come close to covering all of my household power needs, whereas the current system gets maybe 60% or so, but I'd have had to cut down the big old cedar tree which shades part of my roof, and that just seemed like it would be missing the point somehow.)

I live in colorado and it says i would lose money by going solar. Interesting since at costco, etc the solar guys are pushing pretty hard. I've held off so far because I think there's still room for dramatic improvement in solar panel and battery technology in the next few years.

Also I seem to get a new roof every 5 or so years due to all the hailstorms we get around here. I suspect solar panels don't like big hailstones very much and I have no idea what would happen if both the roof and panels are damaged. I doubt a roofing company is well equipped to handle the panels.

I'm curious how this is all going to shake out in the housing market. The guys pushing hard have ~20 year contracts.

If you go to buy a house in 2020 that had a system installed in 2012, the panels etc would be fairly dated (plus wear and tear), and there would still be 12 years left on the contract. I'd insist on paying less for the house due to being saddled with the obligation, or the homeowner would have pay (a lot) to get out of the contract early.

Like you I too see dramatic improvements over the next few years, hence committing now not being a good decision.

That's actually pretty surprising to me, I have only visited Colorado a few times but I have always been very (pleasantly) surprised by the amount of sunshine.

A quick look on Wikipedia says Denver has sunshine 70% of the time totaling ~3,100 hours a year. I'm from MI where we see sun about once a week so I'm out of touch, but that seems like it should be more than enough to recoup costs.

Is the amount of sunshine in Denver significantly different from where you're at, or is there something else I'm missing?

might be current cost of energy is low here. and yes we get tons of sunshine, amazing amounts.

Wow, it does 3D modelling of roofs and trees to estimate coverage!

It only works in a few places. That means it doesn't work for me.

"Note that the total installed costs include the federal investment tax credit and any local rebates and tax incentives."

$3.50/W is WAY higher than I paid per watt last year after the federal tax rebate. I had 3.125kW installed on my roof for $2.16/W. Looking at the map, Denver gets grid parity between $2.00 and $2.50 and that has exactly been my experience with the panels.

They conveniently omitted Hawaii from that map. Solar here is already cheaper than the grid, even after the very high installation costs and even if you go totally off-grid.

If solar interests you, we're hiring senior software engineers at Wunder (TS '14).

Wunder was founded with one simple vision: to address the greatest crisis of our generation by accelerating the proliferation of solar energy. Members of our team have raised hundreds of millions of dollars in private equity, developed energy infrastructure solutions for the Department of Energy, and built digital products used and loved by millions of people. Now, we’re saving the world.

Learn more at https://www.wundercapital.com/hustle

If rooftop solar has a net positive efficiency, why aren't companies paying me to put solar panels on my roof?

There are companies who will put the panels on your roof for free, and then sell you the energy (http://www.solarcity.com/residential/how-much-do-solar-panel...)

There are also power companies who pay their users a discount if they're allowed to install solar panels (http://www.komando.com/happening-now/325802/power-companies-...)

It's easier and cheaper for a large company to just find a big plot of cheap land and put the stuff up there. Same basic reason that nobody's offering to pay you to grow vegetables in your yard, even though growing vegetables is a profitable activity in general.

One advantage of Solar on rooftop, is reduction of transmission and transmission losses, also local handling of the peak etc. Utilities love Distrusted solar in limited numbers. They do not love the idea of everyone getting on the bandwagon.

Never IMO

Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact