What are the costs of solar power intermittency?

A key problem with solar energy is intermittency: solar generators produce only when the sun is shining, adding to social costs and requiring electricity system operators to reoptimize key decisions. We develop a method to quantify the economic value of large-scale renewable energy. We estimate the model for southeastern Arizona. Not accounting for offset carbon dioxide, we find social costs of $138.40 per megawatt hour for 20 percent solar generation, of which unforecastable intermittency accounts for $6.10 and intermittency overall for $46.00. With solar installation costs of $1.52 per watt and carbon dioxide social costs of $39.00 per ton, 20 percent solar would be welfare neutral.

20 percent solar for Arizona as the social welfare break-even point does not strike me as especially impressive, but of course the infrastructure integration technologies may yet advance.  In gross terms, intermittency costs exceed carbon costs.  Note also that forecastable intermittency accounts for most of the costs, and so with perfect storage solar would be a much more efficient technology.

Here are ungated versions of the paper.

Comments

By the way, though carbon dioxide social costs are really hard to measure, but they may already be -- and, especially, could become -- much higher than $39/ton, at least according to French experts: "En 2008-2010, le Centre d'Analyse Stratégique (CAS) a estimé que la valeur de la tonne de CO2 pourrait être de 100 € en 2030 et de 200 € en 2050 (scénario médian dans une fourchette de 150 à 350 €/t)" (https://fr.wikipedia.org/wiki/Prix_du_carbone).

*...hard to measure, they...

What does an Angel weigh?

How much do a ton of angels weigh?

Metric or Imperial?

Imperial Ton of Angels sounds too Warhammer

Mike Trout weighs 230 lbs

actually the costs are negative. Carbon dioxide is extremely beneficial to agricultural yields. That's why greenhouses pump it in at 1000 ppm. It;s probably worth trillions of dollars in benefits since 1960. Global GDP is ~$75T . Agriculture is ~ 3% or $2.25T . a doubling of carbon dioxide increase agricultural yield by sqrt(2) or $0.65T.

As far as warming ( assuming it's all due to CO2, there is natural variability and the earth was already warming before the Industrial revolution) , the most pessimistic data set Hadcrut and Giss show 0.17 C per decade since 1979 and 0.164 C since 2001 Nothing to worry about. The earth has seen it all before

The satellite data sets ( UAH/RSS) show 0.13C and 0.12 C per decade since 1979 and only 0.068C and 0.045 C per decade since 2001. The facts is, there isn't much warming in this century.

The guardian , a pro warming newspaper had to admit Antarctica is actually cooling

https://www.theguardian.com/environment/2016/jul/20/antarctic-peninsula-temperatures-have-fallen-study-shows?CMP=twt_gu

Didn't read it, but has any such study yet accounted for the costs of building sufficient alternative power plants to account for intermittancy/night? The grid MUST be able to serve a peak load, and if solar/wind cannot be depended upon 100%, alternative generation must be built. Is this "back-up" capacity then to stay idle until needed. Who will invest in a part-time, unprofitable power plant?

Until these costs are counted, solar studies are just fantasy.

Nope. Carbon is virtually never the limiting factor for crop growth. The stuff you claim comes from lab studies. Under real world conditions you don't see such growth advantages, and when you do get better plant growth you often get worse yields of what you actually want.

Steve

With perfect storage, any energy source could become more efficient! That would allow you to optimize production for efficiency rather than for demand peaks. Besides that, there are no practical energy storage systems that come near the density and accessibility of fossil fuels.

"With perfect storage, any energy source could become more efficient! "

Everything's awesome when you ignore the Laws of Thermodynamics.

http://image.wikifoundry.com/image/1/KWcvTFC-YYapbq81KYDKdA33611/GW280H320

Aww, there's that darned Second Law again!

Density no, but economically yes https://en.wikipedia.org/wiki/Aquifer_thermal_energy_storage

You store heat, which without conversion costs is only locally useful. Which may be fine for the local loads, but not widely useful with commodity pricing.

Here is an article from the NY times about renewable energy driving out other energy sources, pushing nuclear power into bankruptcy

http://www.nytimes.com/2016/07/20/business/energy-environment/how-renewable-energy-is-blowing-climate-change-efforts-off-course.html?_r=0

It's really cheap natural gas from fracking that's doing it.

+1

If you're planning your energy future, you've got to be thinking about natural gas.

No it's not. It's a consequence of solar intermittency. Too much power in the middle of the day, not enough at 6 pm when everyone is back from work, but you can't turn nuclear on and off, so how can nuclear power be profitable if there are times of the day when they can't sell their power.

You clearly don't know all that much about the economics of nuclear power plants. Short-run marginal costs are very, very low, so they would run full around-the-clock regardless. What nuclear plants care about is 24-hour average pricing, and they don't care about intraday pricing differences so long as they are high enough on average. That average price is all about natural gas. The problem nuclear plants face is that they have comparatively high fixed operating costs that they need to cover, and at current power prices the short-run profits (and capacity payments in markets that have them) are just not enough for them to do that.

They seemed to do fine before solar came along in significant numbers and 24 hour average pricing is going down because they can't sell for half the day.

Nationwide, solar isn't yet to "significant quantity." It's only going to crack 1% of overall output this year. There are zero places in the United States where nuclear plants "can't sell for half the day" because of solar. Furthermore, there hasn't actually been a significant overall decline in nuclear output yet. Only 3 units in have closed in the past 5 years I believe, somewhat offset by 1 new unit that started up earlier this year. The concern is that absent a substantial change in power prices, substantial numbers are likely to retire in the upcoming 3-4 years. Only in CA has solar reached a penetration level high enough to seriously impact dispatch patterns. Solar gets huge amounts of attention not because of anything it's done to date but what people expect/hope it will do in the future. Look at the numbers for yourself: http://www.eia.gov/electricity/data/browser/

Here are the numbers on natural gas: https://www.eia.gov/dnav/ng/hist/rngwhhdm.htm .Note that those prices are not inflation-adjusted, so the drop in real terms in the past 8 years is even larger in real terms than indicated by prices there.

@Plucky . OK I agree with you now , it seems California dominates the rest of the US for solar capacity generation and they only have one nuclear plant operating ( Diablo Canyon) which is slated to close by 2024.

http://www.eia.gov/todayinenergy/detail.cfm?id=24852

The key "system" show-stopper in getting more power from renewables-- is that their intermittancy REQUIRES a huge 24/7 backup generation capability from Coal/Gas/Nuclear. And these backups can’t switch themselves on and off on a dime-- when clouds drift in or the wind diminishes unpredictably.

This conventional backup source must have an overall generation capability equal to several times that of the renewable sources on the grid, but must be turned off (or sell power to the grid at a loss) when the renewables are workin well. The economics doom renewables as a primary power source, under current technology -- the "backup" costs are prohibitive at large scale. Plus, U.S. government regulatory policies are now destroying existing conventional power generation capability on a large scale.

CA used to have another (San Onofre), which closed in 2012/3. That closure preceded the big solar buildout and was primarily due to scandalously shoddy (as in, an actual scandal in which still is an issue in CA politics) maintenance that was exposed during a refueling. That one had nothing to do with either solar or gas price

Had a friend who worked at San Onofre and said they had to use special white paint inside the reactors, but when they were built some painter ran out and just bought regular white paint...started peeling in a few years.

They traced the work order back to the guy though.

Natural gas is driving nuclear out of business because once you must install natural gas to pay the social cost of nuclear operators demanding the grid take nuclear power unconditionally at a high fixed cost, you might as well replace all the nuclear power at a current lower cost.

Nuclear power costs in France where they varying plant output during the day and have a utilization rate of 60% is so high it's a French government secret.

Thanks to economists like Milton Friedman who sold electric power deregulation and competition, it is illegal to require a natural gas power plant to enter into a contract for a fixed price of electricity for the next 40 years.

Conversely, it is illegal to require ratepayers to commit to paying the cost of nuclear power in a 60 year contract.

So, when the longest term for a power contract is a couple of years, how do you convince investors to commit billions to build a power plant when the price it must be at is unpredictable at the midpoint of the debt service. If you charge high interest rates on 40 year bonds to address that risk, the debt service prices nuclear power totally out of the market.

Milton Friedman attacked the PUC because it was too conservative in being committed to never having service outages and never being required to hike rates rapidly due to economic cycles. Thus they set rates high enough that 20 to 50 year commitments could be made for capital assets, and if an alternative cheaper source of power came along, it was shut out of supplying power because prior contracts must be met by all rate payers, and rate payers must not be allowed to leave.

France is overpaying for power now. But when Putin decides to use Europe's gas supply as a political weapon France will be extremely well posiotiined. Oil supply used as a political weapon in the 70's is what lead them to this place. Think of it as an insurance premium.

The article notes in a single sentence that what's actually pushing nuke plants out of business is natural gas, not renewables. Gas is almost always what sets the overall power price, and at present (and likely future) prices, nuclear plants can't cover their fixed costs. The article gives a somewhat misleading impression because most of it is about the dispatch and grid management problems created by renewables, but those are (at least in the US) mostly irrelevant to nuclear power because nuke plants here are designed to run full-out around the clock and not cycle up and down. It's a similar story with coal. "Renewables putting coal out of business" is a story that's just too good for the NYT to examine very closely, and "EPA putting coal out of business" is a story that's just too good for Fox to examine to closely. It's all about natural gas

20% sounds like a pretty good best case for solar. Why shouldn't we be happy with that? It mostly overlaps with when power demands are highest (although evening is an outlier).

If we can copy France with their 77% nuclear, that's a pretty reliable and low-carbon grid.

We can't. France can only do that because they have neighbors to whom they can sell the excess power at night. They can only throttle down the nuclear plants to about 40% of full power at night while maintaining the ability to meet daytime peak demand. And they can only do that because their reactors were designed for variable load. In the U.S., nuclear plants are only designed to meet base load -- the variable load is met by other sources.

You mean "we don't", not "we can't". Any large-scale approach to addressing AGW is going to involve new nuclear designs and a lot of building.

I'd suggest a mixed fleet of nuclear and natural gas power, because gas is cheap, easy to make responsive, and relatively AGW benign. For political reasons we're moving in the direction of a more gas-based policy than I'd believe optimal. Nuclear should be the backbone.

If you design nuclear reactors to vary output, the cost will be much higher.

How much higher is unknown because the real costs are French state secrets.

Estimates of the NSA budget are likely more accurate that the amount of taxpayer subsidy to French nuclear power.

Global warming is a real problem or it isn't. If it is, you're stuck with nuclear as the only solution. Even if it costs more. If it isn't, then we're fine with fossil fuels for quite some time.

But, there's reason to believe that nuclear costs are only where they are because of legal and political issues which could presumably be swept aside in the event of a real climate problem. From an engineering and safety perspective, it should not be an expensive technology. It should be too cheap to meter.

"I’d suggest a mixed fleet of nuclear and natural gas power, because gas is cheap, easy to make responsive, and relatively AGW benign. For political reasons we’re moving in the direction of a more gas-based policy than I’d believe optimal. Nuclear should be the backbone."

The current net summer capacity of nuclear power plants in the U.S. is about 100,000 MW. My prediction is that the net summer capacity of nuclear plants in the U.S. will be less than 30,000 MW by the middle of this century, i.e., more than a 70 percent reduction. The average age of nuclear reactors in the U.S. is over 30 years, and there aren't likely to be significant additions to nuclear capacity (i.e., greater than 10,000 MW) in the next three decades.

It's purely a question of how real and serious the global warming threat is. If it turns out to be a big deal, we'll be a nuclear world. If it doesn't, it seems like natural gas is the fuel of the future for electrical power generation.

Right now things certainly seem to be going the way of gas - I don't dispute that at all. It would take a real political push to overcome the obstacles nuclear faces, and we'll only do that if the powers that be perceive global warming to be a real and economically important problem. Right now they don't, hence they're not doing anything about it. They might be right. I'd probably bet they're right, but I'm not an expert so I'm not very confident in my projection.

The trouble with gas is that, even though we have a very large supply of it, that supply is domestic, the technology with which we can extract it is developing rapidly, and it's relatively benign from a global warming perspective, is that it's still a source of CO2, and for that matter, leaked methane. It's a lot better than most alternatives, but it's not a perfect global warming solution. If, and I realize this is a big if, global warming turns out worse than expected, we're going to need to come for the gas generation capacity as well. A power system designed for that world will be largely nuclear with gas to handle peaking.

Why does anyone need to throttle? Can't they disconnect the turbines?

Anyway, peak load doesn't get more than twice above nighttime base load, so the grid could be 50% nuclear even with no throttling plans.

"Why does anyone need to throttle? Can’t they disconnect the turbines?"

Well you can just vent Steam for that matter. But really the fundamental issue is that you are literally throwing away expensive power. Grid power storage seems the most likely long term solution to these type of Base Load and Intermittency issues.

And waste water. A lot of water. Generating electricity condenses the steam back to water. In physics, nature forces zero sum on engineers.

Nature forces zero sum on economies, too, but economists keep trying for perpetual motion....

mulp, most nuclear power plants are built along rivers with large water flows. The steam produced would be insignificant.

That is where hydro electric fits nicely. Its a great storage system.

"Not accounting for offset carbon dioxide, we find social costs of $138.40 per megawatt hour for 20 percent solar generation"

The costs of solar/wind intermittency are high, but they are capped at the cost of energy storage. IE, if you can store electricity at $100 per MWh then you'll never pay more than that.

Note: Currently Lithium Ion battery storage is roughly $400 per MWh, but falling. Pumped hydro is roughly $100 per MWh, but is limited by regional water sources and terrain.

Cost of storage is 99% labor cost which means storage adds to jobs and gdp.

That's totally different than than oil at $100 a barrel or the price of an iPhone.

So, conservatives see paying workers to be an unacceptable social cost. Unemployment or slavery are preferred.

How is storage 99% labor cost? You propose to use the surplus electricity to run people up an escalator, and then drop them down a chute to run a turbine during peak hours? Seems inhumane to me, but if the unions are ok with it, I won't object.

Beats creating jobs by hiring pairs of people to dig and fill a hole. I guess.

"Cost of storage is 99% labor cost which means storage adds to jobs and gdp."

This is ridiculously untrue.

"So, conservatives see paying workers to be an unacceptable social cost. Unemployment or slavery are preferred."

And this is a sign of someone who's mind has a tenuous connection to reality. Do you literally think that conservative have slave plantations hidden in the hinterlands of Red states?

Keep on fighting the good fight, JWatts. I admire your persistence.

JW, you seem to think mulp can be reasoned with.

He doesn't just spam this message board, he spreads this drivel across *dozens* of different econoblogs. I have no idea why he chooses to spend his free time blaming Reagan for EVERYTHING but here we are.

Is there some compelling reason academics cannot make their papers pretty? It's as though academics, particularly economists, still lived in the Soviet Union. OK, we've graduated from courier to Times Roman, but, gees, life goes on.

What if renewable energy is for now just conspicuous consumption? Is that a big problem?

Perhaps only when mandated by governments

That's the problem with renewables, it feels imposed. When voters support local government decision, no problem: https://www.texastribune.org/2015/03/18/georgetown-goes-all-renewable-energy/

"What if renewable energy is for now just conspicuous consumption? Is that a big problem?"

For example, Leo DiCaprio's $100,000+ electric cars:

http://www.dailymail.co.uk/tvshowbiz/article-2027585/Leonardo-DiCaprio-new-toy-A-100k-hybrid-2012-Fisker-Karma-supercar.html

In his defense, the Karma is a spectacular looking car.

In engineering terms, it makes no sense at all. But it sure looks nice.

Interesting. A real discussion.

In gross terms, intermittency costs exceed carbon costs. Note also that forecastable intermittency accounts for most of the costs, and so with perfect storage solar would be a much more efficient technology

It's striking to me that car companies are launching efforts to recycle electric car batteries into stationary renewable energy storage batteries. Here is an article on BMW doing the same. Over time that may reduce the cost of storage and improve the intermittency problem.

The nice thing about solar power in a place like Arizona is that most peak electricity demand is driven by air conditioning, which coincides closely with sunny daytime periods. As long as demand and supply are closely correlated, the need for expensive storage or alternative backup sources (like natural gas fired power plants) can be pretty modest.

There is a positive correlation between load and solar production peaks, but the paper notes (p9) referring to figure 2.

" Finally, note that during these three days, all the solar installations except for the San Diego facility show large fluctuations over a fine time scale. These fluctuations are not easily forecastable and hence will lead to increased reserve operation costs"

That will result the need to use Natural Gas plants, but it won't reduce the requirement to have them on standby. It can always be cloudy and hot on any given day. So, solar will, on average, cut into peaking costs but it will not totally prevent the peaks, just reduce the frequency.

That (AC demand match to sunshine) is generally true, particularly where the nighttime temps and humidity drop off rapidly after sunset. But I think that's a limited area. Tthere are large areas (Houston - 4th largest city in the US for example) where the AC runs most of the night to make it reasonably comfortable.

I live near Houston, and have been running numbers recently on how much storage I would need to run the AC off grid, and how much solar I would need to run the AC during the day plus charge the batteries. I have 21 SEER AC, and its a lot solar and a lot of storage.

I've been looking as systems that can use excess power during the day to freeze water to provide cooling at other times. Its currently designed and marketed to use cheap nighttime power to freeze the water, and then use that to reduce power requirements for AC during the day. Haven't seen a cost comparison to batteries.

Texas has a lot of wind power. Wind plus solar offers some advantages; the wind tends to blow at night. But it is also highly variable.

Unless you are willing to live with an increased level of system wide power interruptions, you are still going to need a lot of nuclear or fossil fuel base load capacity.

Of course, if you live in the North, its really a severe problem if your electric heat doesn't work at night.

"I’ve been looking as systems that can use excess power during the day to freeze water to provide cooling at other times. Its currently designed and marketed to use cheap nighttime power to freeze the water, and then use that to reduce power requirements for AC during the day. Haven’t seen a cost comparison to batteries."

I've worked on such a system on an industrial scale. It's noteworthy that the project did not reduce the cost of electricity. The capital cost plus the cost of O&M exceeded the costs savings in electricity.

However, the system was still economical, because it allowed the plant to run using 4 industrial Chillers rather than 5 Chillers. The 5th Chiller was only necessary for very hot days of the year for a few hours. Adding the Ice System (ORE skid) was cheaper than added a 5th Chiller.

ORE = orbital rod evaporator
http://www.icesynergy.com/L3-8-PDFlibrary/Ice%20Crystal%20Slurry%20Tes%20System%20Using%20The%20Orbital%20Rod%20Evaporator.pdf

Thanks for the reference. Interesting paper.

The problem with freezing is your eer drops off compared to AC. Especially if you are freezing during the day when ambient temperatures are high. You would be better off freezing during the night and buy off peak power.

Another thing to explore is solar absorption air conditioning. Instead of producing power you use the heat to separate either salt/water based or ammonia water based refrigeration. The higher the sun load the more effective the process. They are typically installed where there is excess heat from industrial processes. There was one I read about in Ontario where they used solar to augment their winter hearing and absorption cooling in the summer.

"The problem with freezing is your eer drops off compared to AC."

Note: That an ORE skid attempts to avoid this issue by creating an ice slurry.

I don't disagree. The Southwest has a big daytime v. nighttime temperature spread compared to the Southeast, and the Southwest also has many more sunny days than the Southeast which has much more cloud cover and more days of precipitation. Solar power is a quite nice solution for a pretty narrow geographic niche, and isn't nearly as good for other purposes - although it is also possible with pretty modest adjustments in cultural practices and in the way systems are organized to shift demand that is currently scheduled for off peak periods in order to get low prices at times of low demand to periods of high solar capacity (e.g. recharging batteries for electric vehicles designed to switch out freshly charged batteries for batteries that are out of juice, and non-time sensitive manufacturing processes).

In the Great Plains and off the coast of the Atlantic and Pacific in places where solar isn't very attractive, wind power can be a good solution, and wind is very price competitive right now, and while intermittent, isn't intermittent on the same schedule as solar is.

Electric heat is not a material concern. Something like 98% of home heating is via either natural gas/propane outside the Northeast or Alaska, or in the Northeast and Alaska, heating oil, because electric heat is much, much more expensive for fundamental technological/engineering reasons. Electricity makes sense for space heating only in very niche applications.

"Electric heat is not a material concern. Something like 98% of home heating is via either natural gas/propane outside the Northeast or Alaska, "

I don't think this is correct. Heat pumps are widely used in the Southern states and they are electric heat.

"Heating fuel choice shows electricity and natural gas roughly equal in newer homes

Homes built since about 1970 use electricity and natural gas as their main space heating fuel in roughly equal proportions, a stark contrast to homes built before 1970, when natural gas dominated heating fuel choice. "

http://www.eia.gov/todayinenergy/detail.cfm?id=7690

OK, after looking up the relevant numbers from the EIA for 2015 found at:
http://www.eia.gov/forecasts/aeo/data/browser/#/?id=4-AEO2016&cases=ref2016~ref_no_cpp&sourcekey=0

The amount of space heating from electricity in the U.S. is actually 8%, small but not quite as small as I had recalled (and the last time I looked it up was a while ago when the percentage was smaller). Heating oil is 11% (principally in the Northeast). Propane is 7%. The balance (about 73% after de minimus amounts from wood and direct burning of coal) is natural gas.

While lots of Southern home do apparently have electric heat pumps, the amount of heating demand in the South in the winter is far less than in the North.

Relevant maps here http://www.nrel.gov/gis/solar.html show that there are much more solar power resources on a line that runs roughly along the Western boundary of the Texas panhandle than to the east of the line. The Northern boundary of the region with lots of solar is in Colorado and points west of it.

Here's the company I was referring to - Ice Energy. They market something called an Ice Battery

https://www.ice-energy.com/grid/

Also worth mentioning that photovoltaic solar power systems are not the only ones that are economically viable. Systems that directly use solar power to heat or pre-heat water that doesn't ever have to reach a boiling point (e.g. household and commercial tap water heating and swimming pool heating) have very high efficiency at converting solar power to useful energy per system cost.

Yes, if I was building custom, I'd definitely look at solar heating for water and space. Particularly if I had any ambitions to ever go off grid. Less clear on a refit to existing construction. Although my gas usage is now so low I'm not sure its compelling.

We moved from HVAC with gas + 2 tanked water heaters to HVAC with gas + 1 tankless a little over a year ago. I only have data for one year, but the current system used only about 45% of the gas the previous system used, both summer and winter.

My summertime use (with no space heating) is only 400-500 CCF. Only 16% of the summer bills is actually gas, 84% is the flat connect fee. Winter use (Jan-Feb) is 8200 CCF; 75% of the bill is gas, 25% connect fee.

I heard of someone developing a solar/electric heat pump. The outside evaporator is a solar collector. I haven't seen one but an intrigued.

It depends on what you mean by "closely" but there's less overlap than people think. We're peaking between 4:00-6:00pm and solar is peaking close to noon. Fixed axis solar (what people put on their roofs) is in the 30% ballpark when the system is peaking. Single axis solar (what utilities build) does much better and can still be 2-3x that during system peak.

It's good that some academics are studying this, but this is a very fuzzy cost/benefit analysis. There are industry analyses that are much more detailed and realistic but mostly not published. Also all such analyses I'm aware of are for specific projects or grids. That's the only way to realistically measure back-up costs, as it depends very much on the sun and weather of the location.

For example, Germany's got weak and highly intermittent sun, which greatly increases the cost of intermittency, but lots of solar installations spread out across the country, which somewhat balances and reduces overall intermittency, and a well-interconnected grid with lots of peaking power available. Another country may have much better sun, but a lousy grid and no extra peaking power available, so big centralized solar installations are actually less economically feasible. When you're putting solar plus gas back-up up against gas alone, gas alone always wins hands down.

"A key input into our analysis is the installed cost per unit of solar PV capacity. Based
on results for 2012 from Barbose et al. [2013] we find that the installed cost is $3.93 per watt
of DC power. We add the net present value of the cost of replacing inverter equipment, to
arrive at a capacity cost of $4.41 per watt of DC power."

Tyler, these 2012 numbers are way off. It's more like $1.25 a watt now and fast falling. http://www.greentechmedia.com/articles/read/solar-pv-prices-to-fall-below-1.00-per-watt-by-2020. http://www.economist.com/news/business/21696941-solar-power-reshaping-energy-production-developing-world-follow-sun.

From p32 "We find that solar capacity costs would have to drop to $1.52 per W for 20% solar to be welfare neutral at the central 2015 U.S. government figure". Well, guyess we're past that point then.;

"Tyler, these 2012 numbers are way off."

Yes, exactly. I was just going to point that out, using slightly different numbers.

http://www.nrel.gov/docs/fy15osti/64746.pdf

From Table ES-1 of that NREL report, the cost for Utility Scale Benchmark in Q1 2015 is $1.77/Wdc. That includes hardware costs (module, inverter, rack, BOS) and soft costs (installation, other).

If they published the paper in 2016, why wouldn't they at least mention that much lower numbers are available for 2015? In fact, I have a hard time even believing their numbers were accurate for 2012. I know PV prices have come down quickly, but geez...

Aha! Check out Figure 24 from that NREL report, on the utility-scale benchmark (in 2015 dollars):
2009 = $4.39/Wdc
2010 = $3.76/Wdc
2011 = $2.56/Wdc
2012 = $1.96/Wdc
2013 = $1.81/Wdc
2015 = $1.77/Wdc

So from that NREL report, one needs to go all the way back to 2009 to find a number like $4.41/Wdc.

Imagine a small liberal arts college investing in a solar panel array in Upstate New York. The only time there's reliable sun is when school is not in session (i.e., latter May, June, July, August). Yeah, we're building it! It goes with our "100% of HWS electricity is from wind power" position. I wonder how many of those water heaters in the theme houses are electric....?

Well, Michael, I googled "map upstate new york" and saw there was a place near the top called Malone. Using PVwatts I see it is about one of the worst places on earth for solar as a south facing system will only have a capacity factor of around 5.7%. But using a discount rate of 5%, a 30 year lifespan, and 2% of the capital cost for yearly operations and maintenance, then at around the current average Australian installation costs the levelized cost of electricity will come to about 23 cents a kilowatt-hour.

I don't know what the retail cost of electricity is in Malone, but it does seem like it could potentially pay for itself provided the installation cost is around the average for the leading developed country. If the college receives a high price for electricity sent into the grid it doesn't matter if solar electricity production at times exceeds consumption. If they don't receive a high price for electricity sent into the grid then they could install a small system. I assume the college will always draw some electrical power during the day.

Solar nearly anywhere would look pretty expensive compared to Oz :)

Australians quite intentionally built their cities in the cloudiest parts of the continent. As a result the United States has many cities that are sunnier than any Australian state capital. For example, Los Angeles. For the most part our solar is on roofs, not off in a solar farm,

Solar power is now being installed for an average of around $1.40 US per watt in Australia, before tax or subsidy. This is much less than the $4.41 US used in the paper:

https://onestepoffthegrid.com.au/solar-choice-pv-price-index-july-2016/

This price is for rooftop solar which does not require long distance transmission and reduces the need for local distribution. Distributed solar is the cheapest source of electricity available to most Australians. As we're not doing anything magical here and large parts of the US receive plenty of sunshine the cost of US solar will fall to Australian levels.

Thanks, RB.

& great comments upthread-MR @ its best.

One way to store solar power is to use excess solar power at noon to pump water uphill to a reservoir that releases water downhill past turbines for, say, air conditioning at 6 pm.

One general problem with using water to store power is that solar power works best in hot sunny climates where water is in short supply and evaporation is a big problem.

We already do this in Australia with our existing pumped hydro capacity. For example, at the moment, at around noon on a Friday, the wholesale price of electricity is around 1.8 US cents a kilowatt-hour in Queensland but it is expected to go up to around 7.3 US cents at 7:00 in the evening. As a result, water is probably being pumped from its lower storage into Wivenhoe Dam at the moment to be used to generate power in the evening.

Australia is a place where water is in short supply and there is a lot of evaporation, but what really uses a lot of water is inland coal generation. Cutting that, which our solar power does, saves water.

Also, a pumped storage facility would have to be in an extremely dry location to not capture more water than it loses due to evaporation. There is one proposed pumped storage system in the outback semi-desert using the pits from an old gold mine for reservoirs. It would have to be filled with bore water, but even in its dry location it should be able to maintain its own water level from rainfall.

But solar power has reduced the likelihood of new pumped storage being built in Australia by lowering the wholesale price of electricity here, so we're not likely to get any new pumped storage. This could possibly change if we get a carbon price, but by that time home and business battery storage, which looks like it might be about to take off here, may still make it unprofitable.

Biomass is another way to store solar power, just grow energy crops.

The relevance of this issue declines with the size of the grid.

How often is it cloudy in every sunny place in the USA in the same day? So, maybe the 20% can become 30%.

Awwwww, no silver bullets. I can't believe anyone even bothered to go down this dead end alley in the first place. it is not a silver bullet ergo it is a complete failure.

Comments for this post are closed