Rooftop Solar is Expensive and Inefficient

Noted UC Berkeley energy economist Severein Borenstein writes against the proposal to make solar required on all new residential construction:

Dear Commissioner Weisenmiller:

I just became aware in the last few days of the proposal in the new building energy efficiency standards rule making to mandate rooftop solar on all new residential buildings. I want to urge you not to adopt the standard. I, along with the vast majority of energy economist, believe that residential rooftop solar is a much more expensive way to move towards renewable energy than larger solar and wind installations. The savings calculated for the households are based on residential electricity rates that are far above the actual cost of providing incremental energy, so embody a large cross subsidy from other ratepayers. This would be a very expensive way to expand renewables and would not be a cost-effective practice that other states and countries could adopt to reduce their own greenhouse gas footprints.

Because I, and most other economists studying California’s energy policy, just became aware of this proposal, we have not had time to participate in the policy process or write public documents on the subject. At the least, I would urge you to delay adopting such a rule until independent analysis from energy experts can be made part of the record.

I will add that I have no financial interest in any energy company. I am expr essing my views purely in the interestof moving forward with California’s fight against climate change in a cost-effective way that can be exported to other states and countries.

Sincerely, Severin Borenstein

I agree and would add that allowing more building near transit and other hubs as with California’s rejected SB827 would not only lower housing prices, rather than raise them as with this proposal, it would also be a much better way of reducing carbon emissions and saving energy.


Good energy policy has always taken a backseat to sexy energy policy.

If you think this has anything to do with energy policy, you're out of your mind.

This is 60% about virtue signalling and 40% about kickbacks to companies with large Dem donations.

Actual energy production? Affordable housing? Who cares?


It's also fine 'deliberate hypocrisy' - the goal being to make the power base FEEL like they're "doing the right thing" while actually excluding people from CA, raising or maintaining status differences, and so forth.

One wonders if the general demise of religion creates some kind of public psyche requirement for such nonsense?

You are overthinking this. Its old fashioned banal corruption. A few energy guys, real estate guys looking to make a score from legislation.

Rooftop solar is just a means to an end: piss off enough homeowners who will demand that power companies implement more efficient and cheaper alternatives for renewables. You want incentives, then enrage enough homeowners that power companies cannot continue to make excuses.

Right. Cheap solar and wind energy is just lying around, pretty much everywhere, but power companies invent "excuses" not to use it.


Why would power companies NOT want to install utility-scale solar? In a regulated monopoly model the utility earns a fixed rate of return on prudently invested capital. In a dispatch-to-market environment (like California) generators want to be committed and dispatched at prices that will give them an adequate rate of return on their investment. If solar did that they would build it.

One reason might be the relatively small - actually tiny - output of anything other than huge array facilities. You tend to see stories about 20 MW facilities with the implication it’s a lot of power. Ivavpah is huge, and about 400 MW max. Typical for a nuke plant is about 1000 MW. You need a lot of 20 MW facilities to replace a nuke.

I would speculate that the utility industry is also very conservative, in the sense of wanting to be really sure they can keep the lights on, and may be slow to decide to build the 50 neighborhood 20 MW plants to replace the 1 nuke plant.

I would think there would be merchant power startups who might build half a dozen 20MW plants and sell the power.

Somebody might ask why that’s not an Alphabet business line.

And it is so much worse than that. A PV facility only produces peak power for maybe an hour or two a day. Above 50% of power for perhaps a total of 4-5 hours a day. For 16 hours a day it sits there producing nothing. Another negative factor is that if an electric utility has a major PV system as part of it's power generation it must also have access to conventional electric generation facilities of the same capacity to be able to quickly take up the load if the solar out put drops or, as it does every night, the sun goes down.
Large commercial PV is a scam. It is a big moneymaker to a few large offshore corporations who build them and run them and the rate payers pay about 4-8 times as much for that power then they would for a fossil fuel powered generation plant. It is essentially a tax on all users of electricity that was forced on us by misleading propaganda.

You can improve those output numbers some with one or two-axis gimbals, and the disruption to the system can be lessened by making the solar station curtailable. Not surprisingly, developers don't want their units to be curtailed so they push for 'must take' contracts in lobbying with legislatures and PUCs. Of course, that just pushes more of the cost of intermittency onto the rate payers.

"And it is so much worse than that. A PV facility only produces peak power for maybe an hour or two a day. Above 50% of power for perhaps a total of 4-5 hours a day. For 16 hours a day it sits there producing nothing."

Your numbers work out to a capacity factor of less than 20 percent. The actual capacity factor for utility scale photovoltaics in 2017 was about 27 percent. So it's not as bad as you depict.

There is another way to think about this.

If a utility builds a 100MW gas turbine power plant then what it has is 100MW of generation capacity available basically anytime it is required.

Now consider what you would need to do to get equivalent capacity from renewable sources.

If wind is the preferred option you need say 300MW of wind turbines in widely dispersed locations in order to be reasonably sure of having at least 100MW available at any given time. Your actual output will be anything between 0 and 300MW depending on wind conditions that probably don't correlate very well with electricity demand.

If you'd rather go with solar you need roughly 500MW of solar panels plus a big battery system.

The thing people tend to gloss over is that wind and solar are intermittent energy sources.

A 5MW wind turbine generates 5MW only in ideal conditions. It's average output is much less than 5MW. You also don't get to choose the output you get at any given moment.

100MW of wind or solar capacity is not the same things as a 100MW gas turbine.

Big centralized installations might be more efficient, but would they be politically feasible? Aren't there tons of NIMBY issues involved? Can the government be trusted to handle everything to build those? Mandating residential solar roofs might not be the optimal solution or all that efficient (at least, not in the short term--in the longer term it might trigger a flood of new innovations and cost reductions) but it still seems like a positive step.

NIMBY is not really an issue for building power plants, solar, wind or gas. Outside of urban areas (e.g. San Francisco) I have not heard of developers struggling to find suitable sites. Transmission, on the other hand, can be a real problem to get done, which is why most generation projects, including solar, wind up near existing transmission.

Remember that idiotic fundraising campaign for solar roads? That was the jumping the shark moment for non-utility solar as far as I'm concerned.

hey, rooftop solar at least works.

Solar Roadways did have an entertaining year, between their surface being broken by people stepping on it, catching fire from overheating, and then the creators being caught shoveling snow off it before posting pictures claiming to show that it had melted the snowfall

That rooftop solar is a poor second to large scale (commercial or utility) solar has long been my position, including in these pages.

Here is a PDF on school sized projects. Those make much more sense to me. They can be more rigorously designed, maintained, and monitored for ROI.

Speaking as a Californian, I hope the State wises up. If not, there will probably be messy positive byproducts of the plan, but a great deal of negatives, and general inefficiency.

Get of California out while you can.

Why? It just took over as worlds 5th largest economy from the UK. Theres money to be made. A lot of it.

There's a lot of ruin in a country - or even a state.

California been berry berry good to me.


Yep. I've also pointed this out a number of times over the past few years.

Its interesting to note that even new construction of big box stores and strip centers does not routinely include solar panels, which I view as fairly good empirical evidence that its not economical.

Yet its got to be more economical to install 100 or 200 panels on a roof than 12 on a residence.

"NREL Report Shows Utility-Scale Solar PV System Cost Fell Nearly 30% Last Year" (in 2016 that is).

That should make a big difference in forward projects. Sure, even 2013 projects where in a wholly different cost environment.

Tax carbon emissions from power plants at some rate that offsets the reduced standard of living for future generations due to climate change, and roof top solar becomes very economical.

LOL. Let me know what sort of living standards you get with rooftop solar and giant windmills everywhere.

Are there any solar panel factories with rooftop solar panels? Solar-powered intermodal carrying all the solar panels from China?

'Let me know what sort of living standards you get with rooftop solar and giant windmills everywhere.'

Yep, Germany is a socialist hell hole, no question.

If Germany had to make up that other 80% of its energy from non-renewables you wouldn't even be on the Internet. You'd be too busy chopping wood for your stove to cook your yummy potatoes and cabbage.

67% in 2017, actually, compared to 71% in 2016, when talking about electricity - 'Der Anteil erneuerbarer Energien an der Stromerzeugung hat in diesem Jahr 33 Prozent erreicht gegenüber 29 Prozent im Vorjahr. '

'you wouldn't even be on the Internet'

Well, I would, since I power everything off a solar/battery set up. Sure, there will come a point where the constant growth of renewable electricity in Germany will top out, but till now, everyone who said it would be impossible to have a modern reliable electrical grid using 10%, or 15%, or 20%, or 25%, or 30% renewable electricity has turned out to be wrong.

You've got 20%, not 25 or 30%. Again, all renewables and you will have people freezing to death in apartments. After they've chopped down the Black Forest to build fires.

since I power everything off a solar/battery set up.

You absolutely do not, unless you live in a Kazcinsky-style strugglehut. Or maybe you do.

Tax carbon emissions from power plants at some rate that offsets the reduced standard of living for future generations due to climate change, and roof top solar becomes very economical.

I do not think so, rather I would bet that if you taxed co2 and paid out the money for semi permanently removing co2 from the air enough to reach equilibrium, solar would still not yet be economical.

"Rooftop Solar is Expensive and Inefficient"
Yet, my tan tells another story.

I had to pick myself up off the floor after reading the final line in this post about building near transit hubs lowering housing prices. The experience with the DC Metro system clearly points to a contrary outcome. In my area of Bethesda, every new high rise going up has way above market rents or sales prices in the case of condos.

Those moving into the buildings will leave behind now newly more affordable housing. Supply and demand does work.

Building more housing anywhere is what increases supply and lowers housing prices. But don't expect the brand new housing to be the most affordable. The affordable housing generally will be the older stock, whose price is lowered by the overall increase in supply.

The reason to favor SB827 wasn't because it was going to result in a lot of affordable housing near transit lines, but because the 'lift housing restrictions near transit lines' rationale was the only one that any had hope of cutting through the regulatory thicket of CA building restrictions.

Strangely, solar hot water heating is very efficient, and yet somehow does not get much play in the land of large electric water heaters.

" the land of large electric water heaters"

Actually, the U.S. is more a land of large, natural gas water heaters.

Certainly not where I grew up. Admittedly, I was thinking residential and suburban, not commercial or urban (I believe that many buildings in NYC use oil for water heating, even today). And propane is certainly common in more isolated rural areas.

So let me check - and not much luck, though the Energy Dept. does say that 17% of a home's energy use is heating water - and that is the largest single energy consumer. But then EIA says something considerably different, at least in terms of ranking.

To be honest, this looks like pretty much that even a regional generalization may not be possible. I will say that absolutely none of the places I have lived in or known in NoVa ever used natural gas for water heating. The only exception was a house with its own propane tank way out on Rt 15, but I'm pretty sure that is not what you meant by natural gas.

Just over half of all homes (53 percent) are served by gas water heaters; most
of the rest are served by electric (40 percent). Other energy sources for water
heating include fuel oil (used primarily in the Northeast), propane, wood,
and solar

Given how far natural gas prices have fallen due to fracking since 2009, I'd expect the numbers using natural gas to gone up further.

Makes sense, though not in NoVa - basically, it seems as if the costs of bringing natural gas to the already developed areas of NoVa cannot compete with Dominion being involved in pipelines to bring gas to gigawatt range electrical generating stations.

I think this is extremely regional, even in a place like Virginia, and the fall in the price of natural gas may not play a major role in many places. A pipeline network with a residential component is considerably more expensive than a point to point pipeline feeding a major generating station - and upgrading an existing network is likely much more expensive.

Interesting subject, though of course the savings for the homeowner are unchanged.

The Virginia pipeline link -

Virginia pipeline

Is she related to Florida Pipeline? Maybe you don't know the joke.

I think I've had gas in every place I've lived in NoVa (almost 20 years), including Alexandria and Fairfax.

Yeah it's been pretty bad these past 20 years...

Solar hot water is only inexpensive in areas where it's unlikely to freeze.

It's possible in the frost belt, but now you need to use antifreeze and since you surely don't want to drink this you need a heat exchanger, which reduces both efficiency and reliability. Plus a controller that's smart enough to turn off the external circulation when its heat contribution becomes negative.

Nonetheless, I'd have to agree that solar hot water systems in warm climates are low-hanging fruit. At least if you can DIY it, or find a competent contractor.

Having priced it out recently, I can assert that the fixed costs for retrofitting indirect solar water heat onto existing construction overwhelm any expected fuel savings.

Progger act like religious zealots. They don't care about facts.

Severin Borenstein,

Please work with the large utility companies to stop charging over $.35 for usage during peak hours. For an owner of a single home, Solar saves me $30/month from the prices SCE charge. Remember on a consumer individual basis we are not just paying energy production.

Yes, I agree solar should not be mandated.

Why would he do that? Why would they stop doing that? Peak electricity is expensive.

35-40c/kWh is just a bonkers price to pay. It's going to kick off a death spiral for the utilities, driven by their addiction to monopoly profits. The powerwalls are coming for them. For 16k$ I could cut the cord entirely, I'm just waiting for some superextractive time of use tariff to make it pencil out.

It's surely true that regulated-monopoly electric utilities seldom have much if any incentive to reduce their costs.

Nonetheless, Powerwalls increase the efficiency of the utility's distribution and generation systems. Distribution efficiency is improved because peak current is lower (and losses are proportional to the square of current), generator efficiency is higher because generators with the highest marginal costs are the ones brought online to meet peak demand.

Of course, regulated monopolies seldom lose, as they'll demand (and get) continued ROI on past investments even if these are no longer needed. The only way for them to lose is (a) they become unnecessary due to obsolescence (gas lighting, landline phones), or political winds turn against them allowing regulators to loot them.

They will have a very hard time extracting that ROI if large numbers of middle-upper class residents cut the cord, both mechanically and politically. Cord cutting, at least in California, is within reach of sensible financially.

This is simply nonsense. They're not charging that much to gouge you. they're not even allowed to do that. they're charging that because peak electricity is expensive. If it became less expensive, they would be charging you less. What you are saying is impossible.

In CA, the peak prices are so high because of politics and the regulatory process. They certainly do not reflect actual peak electricity commodity costs, which are dramatically lower. They are made so high by recovering fixed costs (and costs of subsidies for multiple energy-related programs) through usage charges rather than the fixed monthly charges that economics would suggest. The original political justification for this was to protect the "little guy." Lately, the solar industry has fought tooth and nail against pricing reforms because the result would be to reduce the large subsidies to rooftop solar customers.

This is not the doing of the utilities, they understand the problem but have been required to charge this way.

SCE is providing you with a very different product than the product you are selling back to them. They must stand ready to provide you with power whenever you demand it, but you are only providing them with power when you've got it. So if you are net metered at the retail rate, you are being subsidized by the rest of the SCE customer base.

Why only new construction?
Why not mandate that rooftop solar be added any time ownership changes hands?

(Suggesting a way to poison pill the thing, here.)

Of course it is. If it was inexpensive and efficient they wouldn't need to mandate it.

The professor suggests that rooftop solar is only valuable because residential energy prices are higher than marginal generating costs — but if you were to spend the same money on industrial solar generation you'd get much better value.

Then, isn't part of the benefit here that these homes will avoid some of the inefficiencies of the power grid? It sounds like putting power generation directly on top of the consumers has an efficiency advantage over industrial generation. Perhaps that would be true whether it were a solar roof or a teeny tiny coal plant inside your fireplace (of course the coal plant would be environmentally harmful). If it's worth it, why not do it?

Transmission line loss is very small compared to the scale economies that the professor is talking about. There could be benefits from delay of infrastructure investment in distribution areas that are near max capacity, but that benefit could be more than completely off-set by the additional wear-and-tear on substation hardware from increased net load variability that comes with roof-top solar. The people I speak to tell me that there is no distribution benefit discernable from residential solar.

Thanks. Then the question is why rooftop solar will reduce my personal electricity costs rather than purchasing everything from the grid.

Is it a matter of pricing incentives? That is, the utility will charge you per watt-hour, even though there are significant capital costs to them that don't depend on my total usage. Would rooftop solar become relatively uneconomical if the cost of being connected to the grid were separated from the marginal cost of power from the grid?

"It sounds like putting power generation directly on top of the consumers has an efficiency advantage over industrial generation."

Electrical line loss is, typically, very small. Usually, under 1%.


California seems to lose about 10%. Maybe it's 1% on a specific high voltage line, but after you step it up and down a few times it seems like it comes out to more then 1%.

"It sounds like putting power generation directly on top of the consumers has an efficiency advantage over industrial generation."

Yes, the efficiency is that as our energy needs grow, the utility grid doesn't expand automatically and easily. Adding new power plants is expensive, difficult, and not assured.
Whereas adding power generation on every single house is simple and easily done.

It doesn't work that way, however. The utility has to size the distribution network for the amount of power the customers are going to want when they want it, not the amount that they are going to want when the sun is shining. Not many customers will be happy if their AC goes out on a hot, humid, overcast day because the utility didn't build enough capacity to handle the load when rooftop solar wasn't producing. In addition, intermittency puts stress on hardware at the substation level and below, decreasing the effective lifetimes of expensive components of the system. And the distribution network does expand pretty quickly and regularly. Expanding substations can result in some neighborhood squabbles, but they are usually pretty easily mediated. New distribution in new subdivisions is a snap. The impact of rooftop solar vs. utility-scale solar on transmission is generally very negligible, from what I've seen. I've never met a distribution or transmission person who thought that distributed solar provided any measureable advantage to the network. It's all about generation and related ancillary services, and building new power plants is much cheaper, all-in, than solar, either roof-top or utility-scale.

"Hot, humid, and overcast" doesn't describe California.

Never, anywhere in the entire state?

Pretty much. That's one reason people pay a premium to live here. The ocean here is relatively cool. Areas with warm oceans get "hot, humid, overcast".

But the utility (in California we’re talking about CAISO) can’t scale to achieve “pretty much”. Obviously, if solar generation matched load perfectly you wouldn’t need so much infrastructure, but it doesn’t.

More than half the cost of power is the cost of the grid. We can't get by without the grid, yet, but it is headed that way, so there is much improved efficiency in roof top.

I will add that I have no financial interest in any energy company.

wonder what his investments look like.

A cynical person might suggest that consulting or speaking fees, for example, are not a financial interest in any energy company, even when those fees are ever so coincidentally paid by those involved in the energy industry.

"I will add that I have no financial interest in any energy company."

"Wonder what his investments look like."

If I had, for example, an S&P 500 index fund, I would write, "I have no financial interest in any energy company." My meaning would be that I personally do not have any conflict of interest in discussing electric power companies, because my ownership is of an S&P 500 index fund, not some separate particular power company.

It's remarkable that climate policy seems utterly indifferent to developments in climate science. After the IPCC downgraded its estimate of temperature sensitivity to CO2 in its latest reports, dozens of new studies are increasingly suggesting that this sensitivity is much lower than initial estimates.

The recent study by Judith Curry and Nic Lewis suggests the sensitivity of standard models is up to 50% too high.

Has any politician or government reassessed their policies based on the latest science? Of course not.

Incidentally, when I Googled "Lewis Curry climate sensitivity", the first search result for this widely discussed study wasn't the study itself or one of the very popular skeptic sites, but a site called And Then There's Physics known in the skeptic community for its rabid promotion of AGW.

Curry's own site is ranked about 150,000 in the US while ATTP is at 1 million, but Google pointed me to a small and obscure critic rather than the study.

Google has its thumb on the scale.

I have rooftop solar but no battery storage on-site, so I am probably not reducing my utility company's base load very much (though I can run my A/C on full blast in the heat of the day without affecting peak load...)

But what I've wondered is why don't utility companies invest in large scale battery farms that "soak up" the excess solar on the grid and allow them to use it overnight and lower their production at those more predictable hours...? Is the scale of such a battery farm so large that even with a warehouse full of them you could only power a city for 30 minutes? Not sure where that technology is at.

"But what I've wondered is why don't utility companies invest in large scale battery farms that "soak up" the excess solar on the grid and allow them to use it overnight and lower their production at those more predictable hours...?"

I believe the battery cost amortized over the expected number of charge/discharge cycles exceeds the cost of power from a peaking plant by a large margin, thus utilities will only talk about battery backup if the taxpayer is on the hook for the costs.

Depend on how large the warehouse and the small the city, but in general (and real life) I think you will find that any realistic size battery bank is very small when compared to city sized needs. I’d be surprised if as much as 30 minutes was achievable.

Battery technology lags behind, although there are some promising developments in redox flow batteries where energy is stored in liquid form and the capacity can be scaled using tanks and pumps.

I find Borenstein's argument in favor of centralized solar unconvincing. All the studies I've seen this far show they are less efficient for the homeowner when you factor transmission losses and the utility's margin. In Hawai'i at least the case for rooftop solar is overwhelming, and California's weather is not much less sunny.

The biggest cost in rooftop solar is installation, it's quite plausible that requiring it from the get-go would reduce them because the labor costs of laying a roof like Tesla's may not be that much higher than a conventional one.

I've never seen a study that showed that rooftop competes with utility-scale solar. Not that you couldn't construct one. Say I put the utility solar in Minnesota, pointed it north and wheeled the power to San Antonio. The benefits of utility-scale solar relative to roof-top include multi-axis targeting and curtailability. Those are pretty big benefits. As for Hawaii, I doubt that the economic case for rooftop is really that good compared to utility-scale solar given that the line losses should be trivial, much less reasonable thermal alternatives (e.g. LNG).

Battery technology is a long way from being able to provide an economic arbitrage tool. Deep cycling electro-chemical batteries kills their useful life so you can be looking at replacing fleets of batteries (each the size of a railroad car) in a time-scale measured in months. Most of the action in utility-scale battery use is in delaying distribution investment, regulation service, maybe voltage control...things like that, but not shifting generation over hours. Now, if you've got the topography for pumped storage you can arb power, but you generally face some serious environmental conflicts over constructing new pumped storage.

Rooftop solar is kind of silly (who wants to have to decided between power and trees?), but mandated solar panels above arterial roads would huge. Power + driving in the shade!

When i arrived in malibu, ca I was a conservative. But damn these guys know how to live! They DO seem to defy the rules of economic life, but I think it's because its so sunny and beautiful here nobody ever wants to leave. The economy is always booming. Housing prices are (usually) rising and when they fall (as they do spectacularly) they come back up. Yup, the Democrats are utterly corrupt - and yes, I have personal knowledge that the solar industry here is mostly a jobs machine for connected Dems. Yes, people vote for ridiculous socialistic policies that SHOULD kill the economy. But it...just...doesn't. California haters rage and foam but I gotta tell ya -- they really don't know what's going on here. This state is rocking. People are richer than ever. Yes, yes, it'll all fall down. But when, my angry friends? Soon? People have been saying that since I moved out here in 1993! And I'm having a VERY GOOD RUN. guess the jokes on you, motherfuckers!

Agreed. When I first moved to CA, it was the 8th largest economy, now its 5th beating out UK (thanks to foolish Brexit) and France. In that time CA created whole bunch of industries that didn't exist yet like green energy, an electric auto industry, private space travel, social networking, personal communication devices, tech companies galore. Sure Dems are corrupt and clueless, but it turns out that having the right people with the right ideas in your state is more important than who's running the state. People making $30k get priced out and leave but in exchange we get software engineers/dreamers/hustlers from the midwest, the south, India, China, all over really. Let's not forget countries that R's might consider too socialist are also doing well like Singapore, Taiwan, Canada, and Scandinavia. Heck, China is doing well and they're Marxist. There's more than one way to run a society but R's are so adamant about forcing their antiquated views on everybody.

"People making $30k get priced out and leave"

CA has the highest poverty rate in the country

Mississippi might have a word or two to say about that.

I know all the Mercatus folks love to talk about SB 827 but you need to give up on that as a vehicle for lowering housing prices (wow, look how cheap the condos are in SoMA SF!). It would have resulted in developers being able to build more condos in places they've been wanting to build them (e.g., more of SF, Rockridge, Berkeley, Palo Alto, Mt View). It wouldn't have resulted in much more carfree living, at least until CA has more "transit hubs" worthy of the name. They used that word to sound green, and because "upzoned site for condos" sounds unfriendly, but nobody would've been able to live in those places without a car, the way you can in cities with actual mass transit. Build another BART tunnel from the East Bay and expand the network to more than one line on the peninsula and *maybe* then you can talk about transit-friendly development. One sad train station does not a transit hub make.

Here's a question: What saves/generates more net energy in a moderately warm semi-inland California environment such as the San Fernando Valley: rooftop solar panels or a shade tree?

In other words, under what conditions would it make dollars and cents sense to cut down a shade tree to increase the sunshine falling on your rooftop solar panels?

Are you fertilising the tree?

The answer to what conditions is presumably "cool but sunny, with high power prices"

But your point in a very interesting one where I sit here in sunny Bangkok in a small house shaded by trees. Blocking the afternoon sun seems to cut my air con use massively.

Another question: I have about a decade left on the 25-year warranty on my fire resistant roof. I've never had a leak in the decade and a half with the new roof, and I'd like to keep it way. How often does installing solar panels compromise the waterproofness of a roof?

Tyler is right. Rooftop solar panels are expensive and inefficient... for now. Fortunately, technology doesn't stand still. Look up perovskite solar panels - it's 'version 2.0' solar technology that will be rolled out over the next couple decades. 10x lower capex due to cheaper production processes, higher theoretical efficiency, MUCH lighter (so the installation costs are substantially lower) and flexible, so it can be placed on airplane/drone wings, car roofs, and anywhere else that a flexible, light, and cheap solar panel would be useful.

It's only partly about the cost of the panels. The electronic controls are another cost, and those technologies are mature and not getting cheaper. The result has been the tendency of developers to 'over-panel' - that is, to make the panels over-sized to the electronics. This causes a whole different set of issues for control of the system. As I understand it, when the sun rises in the morning a solar array doesn't produce very small amounts of power that slowly increase as the sun gets higher in the sky. They produce nothing until they reach a voltage level and then they come on with a bang. When you over-panel the bang is larger, so you have to have some other resources that can ease the transition. The same thing happens in reverse when the sun goes down. You could use batteries, but they are very expensive and increase the cost of the solar system pretty dramatically. So even if the panels were free, the cost could, in theory, be prohibitive.

There's no bang-bang. Rooftop solar will typically produce a few watts even before the sun rises in the pre-dawn light and a few watts after the sun sets. Solar inverters are very good at this and they are continuing to develop and fall in price. We're seeing price falls every year.

Yes there is bang bang. I’ve spent years examining and modeling generation systems with and without renewables and it happens. This, however, is a bigger problem for utility-scale solar than for roof-top since the roof-top systems are smaller and benefit from spatial diversity.

Have you examined a rooftop solar system? That's not how mine works.

Yes perovskite photovoltaics are the fossil least south of approximately the Mason-Dixon line in the U.S.

Another key development is the positive synergies between photovoltaics and battery-powered computer-driven cars operating as transportation-as-a-service.

Photovoltaics need batteries. Computer-driven cars operating in transportation-as-a-service mode favor battery vehicles, because there is no range anxiety. And large fleet owners of battery vehicles have enough clout with electric utilities to negotiate favorable terms (where individual electric vehicle owners do not).

I'm in South Australia. About 7% of the electricity consumed in the state comes from rooftop solar. We have a lot of rooftop solar because we have high retail electricity prices and we have the lowest cost rooftop solar in the world. We got the lowest prices due to a lack of red tape compared to the US and subsidies allowed the average household to install solar. As the average household was price sensitive they looked for the best deals, keeping the installation business competitive.

High levels of sunshine help but a huge number of Americans live in places that receive as much or more sunshine than the average Australian roof.

Someone has an axe to grind (is he invested academically in molten salt or something?). I just had rooftop solar installed, and it's **much** cheaper then the local utility. I could give up my federal tax credit and it would still be a good deal. The utility scale solar will only be competitive if there is cheap land available, which doesn't describe that much of California. Then there are transmission lines to build and all the NIMBYism that surrounds that...

Installing rooftop solar at the time of new construction is just a no brainer. It's vastly cheaper to run wires on an unfinished house. I'm pretty sure it would be cost effective without the federal subsidy. I can somewhat understand the libertarian argument against requiring it, but /shrug it's California there are a lot dumber laws that could be passed.

Bringing up SB827 is a total non-sequitur. Foolish top-down urbanism lost that round, get over it. There's no reason at all to link the two issues, other then petulant rock throwing.

The simplest way to cut through the question of who's paying what to whom and what's subsidized, etc., is to look at the actual cost of the generated power. For retrofitted rooftop solar, it is on the order of 20 cents per kilowatt-hour. Current wholesale energy prices in California are on the order of 3-5 cents per kwh. That you find the result personally attractive just emphasizes the huge subsidies that you are receiving. The high utility rates are a big part of the problem, because they recover the costs of the poles and wires from the charge for using power. When solar makes your electricity bill go away, those poles and wires are still there and someone else is forced to pay for them instead.

New construction solar is less costly, but still likely uneconomic and also highly subsidized.

Bottom line, every one of these installations that goes in makes us all poorer by raising the cost of electricity to society.

My personal experience is that retrofitted is about 18c subsidized, which would be 27c unsubsidized.

I find it personally attractive because retail rates here run to 50c/kWh. The subsidies are pretty nice, but like I said the project makes sense without them.

If the utilities ran reasonable rates in the 20-25c range, they wouldn't be losing all their customers to solar. They might even have breathing room to install utility solar themselves. As it is, they are shortsightedly cutting their own throats.

"every one of these installations that goes in makes us all poorer by raising the cost of electricity to society." What an odd comment to make in a libertarian blog! None the less, if you have a good policy proposal that increases the reliability of the grid while decreasing carbon output and the cost to the consumer, I will gladly trade this bill and the federal 30% subsidy for it (especially since I already got mine ;) ).

This gets complicated because it depends on where you live. In California generation is not a regulated monopoly. Wholesale prices are set in a market run by CAISO. But the price you pay for retail power covers power when you want it. Your rooftop solar provides power when the sun shines on your house. These are very different products so meaningful price comparisons are very difficult.

"meaningful price comparisons are very difficult"

Not really. Figure for 20 years, two powerwall 2.0's would allow me to go off grid. Couple thousand for installation, maybe have to replace one of them half way through. Works out to about 90$/mo subsidized. I use 600kwh or so per month, and my solar costs 120$/mo financed. So basically I could go off grid for 35c/kwh. That's pretty close to what a nonsolar house pays if you average the tiers out. If the utility can't beat an average of 35c/kwh, they are going to lose the whole market. And they are going to lose it from the top end, as all the heavy users who are paying 50c go first.

Well, that's not what quite I was talking about. I was referring to the product you receive from the utility versus what you could provide back to the utility directly (i.e. two-way metering) or indirectly (net metering). The value of what you receive is different than what you provide and working out the proper relative pricing is difficult. Still, I doubt that you can deep cycle PowerWalls and expect anything like 20 year lives. Also, the average residential retail rate in the U.S. is about $0.13 and the highest is Hawaii at about $0.31, so you are probably not very close to going off the grid, which suggests that the value you receive from the grid is much higher than the value of the power you could sell back from roof-top solar.

After a small allocation at like 22c, SDGE jumps to 40c. If you have AC and an electric car you will pay 50c unless you go solar. I think I know my own power bill, thank you. It pencils out, it pencils out without the subsidy, and it's not far off from penciling out with a powerwall. Is the problem that SDGE is doing a 10x markup on wholesale power? Fine.

Isn't all of this analysis done in the context of fossil fuel energy system that is heavily subsidized by defense budgets and other massive cost externalities?

I don't know about defense budgets, but if coal generation had to pay to remove the CO2 it emitted into the atmosphere then I think it's likely to add a minimum of 5 cents per kilowatt-hour to its cost. Paying for the cost of health effects might add another cent, but that may vary considerably from power station to power station depending on location and pollution controls. This would result in coal being priced out of the market.

This is very economist way of doing anything - quibble about the details and sit on your hands.

Is the policy inherently bad or just sub-optimal? If it's the later they should push forward and actually get something done. Execution eats strategy for lunch.

Regulatory capture is a simple concept that's often difficult to define in practice. But the solar industry has clearly captured California government.

As much as we are quibbling about weather or not this bill is good policy, the fact is they aren't making any new houses in the parts of the state that matter. This bill has big exemptions for condos, which is all that people are building in the parts of the state that matter.

Comments for this post are closed