Clean Power us-solar-costs-590x416

Published on October 29th, 2014 | by Giles Parkinson


Solar Grid Parity In All 50 US States By 2016, Predicts Deutsche Bank

October 29th, 2014 by  


Rooftop solar PV will reach grid parity in 50 US states by 2016 – up from just 10 now – setting the scene for a dramatic increase in the uptake in household and commercial rooftop solar in the world’s biggest economy.

That’s the prediction of Vishal Shah, the leading solar industry analyst at Deutsche Bank, who says that declining system costs, customer acquisition costs, financing costs and rising volumes should drive significant scale benefits .

Shah’s prediction was included in his first report on the newly listed Vivint Solar, which is the number two installer in the US. Deutsche Bank considers its prospects are so good that it will at least double its sale in each of the next two years.

But it is also predicting big things for the US market, based on a continued fall in installation costs of solar, the cheaper cost of finance as new financial models attract more mainstream funding, and assuming that the attempts by utilities to curtail the proliferation of solar are resisted.

This graph below gives an idea (click to enlarge). In 2013, the US installed around 1GW on residential rooftops, little more than was installed in Australia, with a fraction of the population. But in the next three years the installation rate is expected to rise six fold by 2016.

deutsche-us-solarVivint, which had just 130MW of installed capacity at the end of June, is expected to have more than 4GW by 2020.

Vivint Solar is interesting because it is an offshoot of a home security company that has used its customer base and customer service model to attack the residential solar market with considerable success.

Rather than an on-line model favoured by rivals, Vivint goes door to door. It uses a relatively unique power purchase agreement model, that sees it selling solar electricity at an average of 14-15c/kWh – a discount of 15-30 per cent below the utility price in its markets.

Customers enter a 20 year contract with no or little up front fees. There is a 2.9% annual escalator, but this is still expected to track below rising utility prices. The escalator is likely to fall.

Currently, it only operates in 7 states where the grid price is highest – in the north east, California, Arizona, and Hawaii. But Deutsche Bank says its target market will rapidly expand as solar costs continue to fall and grid prices rise in other states. And Deutsche expects the fragmented solar market to consolidate rapidly in coming years.

One of the reasons is the availability of finance. Currently it comes through tax equity structures inspired by the investment tax credit, currently at 30 per cent. Deustche Bank says this ITC could change, but it estimates the cost of finance to fall from between 7 to 9 per cent to around 5.4 per cent by 2015. This will help new financing models such as yieldcos, solar loans, asset backed securities and retail bond offering.


Source: RenewEconomy. Reprinted with permission.

Get CleanTechnica’s 1st (completely free) electric car report → “Electric Cars: What Early Adopters & First Followers Want.”
Keep up to date with all the hottest cleantech news by subscribing to our (free) cleantech newsletter, or keep an eye on sector-specific news by getting our (also free) solar energy newsletter, electric vehicle newsletter, or wind energy newsletter.


Tags: , , , , , ,

About the Author

is the founding editor of, an Australian-based website that provides news and analysis on cleantech, carbon, and climate issues. Giles is based in Sydney and is watching the (slow, but quickening) transformation of Australia's energy grid with great interest.

  • Solar energy is almost illegal in Spain according to president Rajoy in ful agreement with the energy cartel. These caveman economies of conservative parties remaining at service of lobbies and banks established no matter population are depressed, starving and facing even more decline on mid term. Sorry to say so about our beloved Spain and its habitants, but shooting them down would not achieve anything except 100 new caveman capitalists replacing each of them. By this day Spain should have been a macro european provider of clean sustainable solar energy….results and facts but speaking its own language.

  • Patricia Longo

    I already know how to install it and use it-
    It’s sad how it isn’t right where I might SEE IT in “our” stores.

  • Sandy Dechert

    Actual numbers by state from Bloomberg graphed below.

  • Guest

    Bloomberg show the actual numbers for each state below.

  • Steve N

    I love the chart. It just starts taking off and suddenly after 2016 the growth rate just slows to a crawl. Just like digital cameras, cell phones, etc. etc. did. Prices will be dropping while utilities will be raising rates like crazy to capture there cash flow. This chart is right on the money. Up until 2017……

    • Mike Shurtleff

      Yes, exactly. No way it is going to level off like that and grow at a slow linear rate after 2017. It may slow for a few years, but then off it will go again.
      10 states at parity now.
      50 states at parity by 2016.
      That means those 10 states will still be piling on the solar in 2017 with, or without, the ITC. The cost saving margin with be that large by then. …and probably still growing. New developments in efficiency improvement and cost reduction are still coming to the market for Solar PV and Storage. You ain’t seen nuthin yet.

  • Marion Meads

    “Customers enter a 20 year contract with no or little up front fees.
    There is a 2.9% annual escalator, but this is still expected to track
    below rising utility prices. The escalator is likely to fall.”

    And these installers will sales talk you into thinking that that the current utilities are going to continue to raise their prices indefinitely thus bloating your savings, suckering you into the contract that you will soon regret

    Reality check: Do you think that the utilities will do nothing when the residential solar PV installers are eating their revenue stream? There will be price wars and the existing utilities will try to be more efficient and perhaps unbundle the retail rate into various components, and the retail price per kWH will be severely reduced by your utility. So while your utilities prices will be reduced due to competition or changes in regulations, you have just guaranteed the lessor the escalation price.

    So if you are a sucker for solar leasing, make sure that you are grandfathered in when the retail prices are unbundled. Otherwise you pay an arm and a leg to the lessor while your utility offers dirt cheap per kWH rates.

    In the future, once the prices of utilities have become cheaper than the escalated price to your lessor, and it is such a regrettable pain to live out the remainder of the lease that you have been suckered into signing, it would be easy to create increasing maintenance headache for the lessor, the system has many weak points for “accidental” or “natural” damages (aka sabotage) to create more frequent onsite service visits and force them into the re-bargaining table.

    • SecularAnimist

      Marion wrote: “the existing utilities will try to be more efficient and perhaps unbundle the retail rate into various components”

      That has already been done in some states. In Maryland where I live, the utility invoice has two separate sections: one for the electricity consumed, and one for the grid transmission charges. Consumers can choose from a variety of electricity producers, including some that provide 100 percent wind-generated electricity. The utility itself is not an electricity producer — they just operate the transmission grid and handle the billing of consumers and payments to producers.

      I think that is likely to become the model for the grid going forward, as it evolves into the Intergrid, connecting a diverse network of microgrids, and both large centralized and smaller distributed electricity consumer-producers.

      When you look at utility resistance to distributed solar, what you’ll see is that the utilities that are really fighting it are those where the utility is not just a grid operator, but also a major owner/operator of fossil fueled power plants.

    • Offgridman

      In some places the rates that the lease companies can charge falls under the jurisdiction of the state PUC’s, just like the utilities rates, If (and that is a big ‘if’ as I have explained before) the PUC sets lower rates for the utilities, they will also do this for the lease companies.
      I know that you strongly disagree with the lease model for rooftop solar, but there are to many families (thanks to our current economy and wealth shifting) that cannot afford to invest in solar, or get the credit to do so. With the continuing reductions in solar installation costs this will decrease the number of people stuck in this situation and leasing numbers will decline. But for right now it does mean increasing numbers of homes with solar on their roofs and reduced energy costs. This is a good thing for those home owners, the environment, and thus all of us.

    • Mint

      I think most (all?) rooftop solar installations go into a long term net metering contract with the utility (like 20 years or something). Even if the utility changes it’s pricing model, your household/business will be locked into existing pricing.

      So from an individual POV, if you can get in on these contracts before the utilities change the rules, they are the suckers, not you. It’ll only last while rooftop solar is a few percent of generation.

    • Mint

      To support your point, I made a numerical example on cleantechnica a while ago, and I’ll repeat that here with some tweaks.

      Consider that PG&E’s financials in 2013:
      $15.6B in revenue in 2013 ($12.5B from electricity sales, ~75GWh)
      $5B in cost of electricity
      $1B in cost of natural gas
      $5.8B O&M
      $3B, depreciation, interest, and tax
      $0.8B net income.

      Now, let’s say net metering makes 7GWh of demand switch to rooftop solar, and naturally it’ll be high-tier demand (maybe averaging $0.25/kWh). PG&E then loses $1.75B in sales (16% of electricity revenue), but they only save $0.45B in costs (9% less cost of electricity). All other expenses are virtually the same, so now it’s a money losing operation.

      • GCO

        You assume that electricity costs PG&E the same 24×7. Your conclusion will not be as clear-cut, or even the same, if the much higher price of generation and/or imports during peak periods (mostly summer afternoons) is taken into account.

        • Mint

          Once you go beyond a few percent solar, peak demand is 5-10PM. That’s what we see in California:

        • jeffhre

          Distributed solar bypasses the high voltage transmission system, and places less stress on the transmission system overall during daytime hours.

  • Mint

    Grid parity assuming net metering.

    So maybe 5% penetration at most before the utilities increase rates on the solarless, who start demanding rooftop solar stops getting a free ride.

    • SecularAnimist

      The claim that rooftop solar “gets a free ride” is nonsense. Households and businesses that install rooftop solar don’t get anything for free from the utility. When they need electricity from the grid, they pay for it exactly like everyone else. And when they sell their excess solar-generated electricity through the grid via net metering they have every right to be paid for it, just like any other electricity producer.

      • Mint

        Net metering credits rooftop solar 2-5x what electricity producers get, depending on the state. So no, it’s not “just like any other”.

        This really isn’t hard to see. If 50% of customers went solar due to “grid parity”, the utility loses half its revenue while still having to pay for 100% of the grid and 100% of generation when the sun goes down. That’s impossible without massive rate increases.

        Of course, things will change before that happens, and I expect it to be at the 5-10% penetration level.

        • Mike Shurtleff

          Yeh, look at SA Australia:

          They have a higher level of penetration than you say is possible and it’s doing terrible things …like supplying a lot of reliable low-cost electricitiy …and reducing their electricity rates …or at least it would if the utilities weren’t making customers pay for their stranded asset costs.

          Kaui is headed toward 50% renewables. A few other islands are already 100% renewables. If small islands can do it, then it’s far easier for a larger, more widely distributed grid to support a large percentage of Solar and Wind.

          Look at low cost battery technologies coming to the market from Ambri, Alevo, Tesla, Aquion, EOS, and others. Really think 10% penetration is the Sound Barrier for Solar? You’re a fool for a fossil fuel & grid utility paradigm. Sometimes paradigms of the past get broken.

          • Mike Shurtleff

            btw Mint,
            I notice you changed from “5%” period to “5-10%” in your next comment. You don’t even believe what you write …or your subconscious doesn’t.

          • Mint

            I said “maybe” 5%. It really depends. In California, that 5% would largely be residences in the top 35c/kWh tier, meaning massive loss of revenue.

          • Mint

            Learn how to read. I’m not talking about total solar. I’m talking about NET METERED solar. And this article is about 50 state solar parity in 2016, not cherry picked solar parity across the globe. In Kauai, solar’s competition is pricey diesel (energy cost 5x that of natural gas).

            But if you want to talk about Australia, it’s undergoing exactly the price hikes I’m talking about. The supply charges have skyrocketed:

            I’m not some FF fool. Bob Wallace feels the same way about net metering: it’s unsustainable beyond its role as a solar kickstarter. Furthermore, the same money spent on big solar will give you far more renewable generation.

            As for storage, it’s too costly for there to be a mass grid disconnect, even with home storage advances from Tesla/Aquion (Ambri and EOS are utility scale). Seasonal variation is too high even for the Bay Area, with winter solar production only 30% of summer, and far worse in the northern half of the US.

          • Mike Shurtleff

            I feel the same way about net metering. I’m in favor of wholesale rates …when the utility can remarket the power to others …nothing when there is already over-production of power. External regulation of the utilities is needed. Separate generators and distributors of power. If utilities are spending any money “fighting Solar and Wind” then replace them with operators who can be cooperative/

            There is no 5% limit or 10% limit, as demonstrated by those other places. No different in Southern USA. An obvious point. Your’s is an already disproven view.

          • Mint

            Once again, the limit I speak of is for net metering. I don’t know how many times I have to repeat that for you to get it. I don’t know any US utility that’s even close to 5% net metered customers yet, so no, it’s not disproven.

            The costs of getting even such high penetration are well studied:

            In California, at 5% NEM, the utilities will face $2.3B/yr in lost revenue and $1.2B/yr in savings. The difference will be charged to non-NEM customers.

            Wholesale has no limit, but parity is far away for rooftop. Big solar is getting close in some states, but certainly not all 50. Wind has a very big capacity factor advantage in most of the US, and will likely dominate renewable energy sales at the wholesale level.

          • Mike Shurtleff

            “Wholesale has no limit, but parity is far away for rooftop.”
            End-of-grid or retail Solar PV parity is here today in some states. If you are already doing this, then you still might want to sell some excess power to the utilities at whatever price you can get. Of course, diurnal storage for your own needs will probably also make sense.

          • jstack6

            so lets stop the subsidies for COAL, NG, Nuclear and see what costs the most. Let’s pay Time Of Day for Solar and FIT, Feed in Tariff so we see the real value of solar. Net-Metering only pays a low price for clean Solar with no environmental value that it deserves.
            Utilities dump excess power at night since they can’t ramp down COAL, Nuclear or even Hydro. There is plenty of excess at night when Solar automatically ramps down and we use less.

          • Mint

            Solar gets rid of the daytime peak after only a couple percent penetration. Net load (i.e. demand minus renewables) is now in the evening when solar produces little to no power. For example:

            Net metering pays a low price? LOL. Net metering credits people upwards of 20c/kWh. Wholesale power, even during the day, is under 6c/kWh on average.

          • Mike Shurtleff

            “As for storage, it’s too costly for there to be a mass grid disconnect in the US”
            Not my point at all. My point is diurnal storage allows for much larger penetration of Solar PV. Those are much lower cost battery storage options than we have had before, some for grid and some for business/residential.

            As cost of Solar PV, Wind, and Storage continue to drop, 50% penetration in winter, with some over-production in the summer sounds AOK to me. Need seasonal storage for Northern half of USA. In NW hydro would work fine. Low cost. More rain, snow, and wind in winter anyway.

            Close minded FF fools think all this is unrealistic, uneconomic, complicated, or hard. It ain’t. It’s coming.

          • Mint

            I’m fully aware of diurnal storage, and that’s gonna happen in areas even without much solar (it’s cheaper than peakers). My penetration discussion is about net metered solar.

            Seasonal storage will never happen in the US on any significant scale. That’s one cycle per year, and would need to be $1-2/kWh to amortize to a reasonable price (that’s a 100x cost reduction).

            Solar PV will eventually reach wholesale parity in many states, but not by 2016.

          • Mike Shurtleff

            “My penetration discussion is about net metered solar.”
            Sorry, that was not clear to begin with, not at all.

            “Seasonal storage will never happen in the US on any significant scale.”
            1. Never is a very long time.
            2. We already have seasonal storage of hydro power. Pumped hydro could do the same. In fact, this could likely solve all the Solar PV and Wind storage problems in the West side of the country …if the utilities had the sense God gave a billy goat.
            3. You do not need much, if any, seasonal storage in the Southern half of the USA.

            “Solar PV will eventually reach wholesale parity in many states, but not by 2016.”
            I agree. This is why residential and commercial Solar PV is more important than utility scale Solar PV. The end-of-grid is where the cost advantage is for Solar PV and for Solar PV storage. Solar PV will be at end-of-grid parity in many states by 2016. It will be below parity by a large margin in some states by then.

          • Mike Shurtleff

            “Solar PV will eventually reach wholesale parity in many states, but not by 2016.”
            Actually, if Vishal Shah at Deutsche Bank thinks we’ll reach grid parity for all 50 states by 2016, then I’m not sure I really disagree with him. On the one hand, if seems too soon to me. On the other hand, we’re about to see a big jump in production scale, accompanied by some large reductions in silicon wafer costs, and increases in PV efficiency. Maybe Vishal is correct.

            Doesn’t matter. Solar PV is already competitive at the end-of-grid …and costs are still falling. That’s all we need for Solar PV to become a large percentage of our electricity production. It’s happening.

          • Mint

            You’re not paying attention. Seasonal storage is, by definition, one cycle per year. Even over 50 years with zero interest, $2/kWh storage would increase energy cost by 4c/kWh. Optimistic cost estimates put pumped hydro at $100/kWh.

            You’d need 400 TWh of storage to shift just 10% of US annual electricity generation from summer to winter. Existing hydro storage in the US isn’t even 1% of that. The Eagle Mountain storage project is only 22 GWh (we’d need 20,000 of them to hit the aforementioned figure).

            Regarding cost, solar has CF of ~20% in the southern US, while wind is 40%. So solar needs to be half the cost of wind per nominal watt installed to match LCOE, and this will never change as CF is due to nature. This means that even if we assume wind won’t drop in price anymore, solar will have to hit $1/W to compete with wind.

            Right now we’re at $3+/W installed for residental/commercial. Rooftops need individual layout and wiring considerations, labor to climb up there, etc. I’ll bet that $1/W installed on the roof won’t happen for at least a decade, if ever.

            Utility scale solar will always be cheaper, and if it achieves end-of-grid parity, it’ll be with robots laying out and wiring massive fields of panels and support structures directly off trucks.

          • Mike Shurtleff

            So you’re telling me Hydro power is not one of the cheapest sources of power? That dams do not store water for irrigation, human consumption, and power generation on a seasonal basis? …and, of course, as a result pumped hydro storage can never be done at low cost for seasonal storage? Who’d a thunk it?
            Forget 50 years, try several hundred. I don’t expect to see this happen in the USA. We’re ignoring our infrastructural needs. I believe it’s feasible though.
            No, residential and commercial Solar PV do NOT compete with source-of-grid wholesale power. They do NOT compete with Wind. Residential and commercial Solar PV compete with retail end-of-grid prices. They will always have a significant competitive advantage. Point of use Solar PV will dominate, not utility scale Solar PV.
            As my father used to say: “everything in this life is relative”. End-of-grid Solar PV has the relative advantage. Ditto for end-of-grid storage. It’s already happening.

          • Mint

            Where did I say hydro POWER is not cheap? Oh, right, you made that up in your head.

            Hydro STORAGE is different. You need giant reservoirs at high altitude change for that. Orders of magnitude more than is used for irrigation and consumption: In developed countries, consumption is <400 liters per capita per day for all uses. At 50 psi, if we used all that energy for electricity instead of pumping water, it amounts to 0.04 kWh/day. Even if we had enough storage to handle an epic 1 year nationwide drought (we don't), its energy content would only be 14 kWh/capita.

            Look at the numbers. No, PuHS cannot be done that cheap unless your reservoirs are free.

            Forget 50 years? Who is going lend you $1B if you promise to pay him only $2M per month for 50 years?


            Sorry, I misunderstood what end-of-grid meant.

            Rooftop solar only competes with end of grid if:
            A) you have net metering or
            B) if you can disconnect from the grid entirely

            Otherwise, it is NOT end of grid, because you need the grid to not only sell that power elsewhere, but also buy power when you need it.

            Distributed solar does not save much transmission cost. On a cloudy winter day with very low solar output, a city still needs the same transmission capacity as a system without any solar. That capacity is always there and needs maintenance from age, weather, trees, accidents, etc, regardless of how often you use it. Therefore the utility sees little value (1-2c/kWh premium at most) in buying your rooftop solar power compared to buying wind power, large scale solar power, or CCGT power.

          • Mike Shurtleff

            I concede you have a good point on PuHS. I’ve been thinking hydro is cheap so PuHS is naturally cheap as well. You point out that the financing picture is radically different. PuHS could still do the job, IF Solar PV and Wind become cheap enough, and IF the PuHS cost is spread out over regular electricity prices for the year.
            You’re pointing out something else to me here. I had not considered that Solar PV might need to pay a premium for grid electricity used for backup power …in order to pay their share of grid maintenance costs.

            A) No, Solar PV competes with end-of-grid prices just fine in many areas even without Net Metering. The problem is the Utilities will have to raise per kWh costs for some group, or everybody, in order to maintain the same cash flow for grid maintenance. Solar PV on the grid creates a diminishing returns problem for the utility. It saves the home owner money until the utility starts to increase charges to try and get their revenue back. …which will drive home owners in some areas right off the grid.
            Interesting. There must be a better way to account for those grid maintenance costs.

            B) Yes, for those with longer transmission line install costs, it makes more and more sense to go off-grid. Obviously, the cutoff line for an off-grid economic advantage is sliding over as the cost of Solar PV and Storage continue to decrease in cost.
            No, you do not have to disconnect from the grid for Solar PV to compete. It saves you money already, unless the utility starts changing the cost structure to take it all back again. It is going to help to have some Storage to go with that Solar PV, 6pm to 10pm storage. That is the time period Solar City is targeting.

            Distributed Solar:
            Look the daily demand curves in your three references below. The first one, for California, clearly shows some drop in peak demand over-all with Solar added in. After fairly low penetration they have now CA will begin to see the same 6pm to 10pm peak that Australia sees, look at those two references. (Notice that the two Australian demand curves already have Solar added in and it has destroyed the more normal daytime load peak. It’s now the lowest demand time of day.) At the Australian level of Solar penetration you can really see the 6pm to 10pm load peak. That is where Distributed Solar PV combined with Distributed Storage lowers demand for grid services. Since a big chunk of that demand is for AC that is higher when the sun is shining (for both S. Cal, and Australia) Solar PV will always be able to reduce that peak. In fact, summer time heat waves are no longer creating the power shortage (brown outs, or outages) that were seen in Australia only a few years ago …this is true in spite of record high temperatures still occurring in the summer there.
            In short, you are wrong. Distributed Solar PV + Distributed Storage does reduce transmission infrastructure required.

          • Mint

            A) Rooftop PV works without net metering only if the utility is paying you a lot more for your surplus power than they pay for other power (4-6c/kWh). They’re only doing it right now because rooftop solar is limited.

            B) Yes, I already told you that after a little solar penetration, peak demand is now at a time of day when there is no solar output.

            Clouds can still cover local areas while the sunshine on nearby land heats up the air. I’ve been through many hot, humid days without much sun shining. Electricity reliability standards are >99.9%. It takes less than one day a year of low solar and high demand to maintain the need for transmission infrastructure.

            In any case, distributed solar + distributed storage is not at US grid parity in 2016 or in the near future aside from corner cases. But in Australia, yeah, it’ll make a lot of people leave the grid soon, because their grid is damn pricey.

          • Bob_Wallace

            Perhaps you need to look at seasonal PuHS storage in terms of extra reservoir size only.

            If the pump/turbine is doing daily cycling and grid stabilization that’s going to cover most of the expenses. The ‘once a year’ stuff is just a large basin to increase capacity.

            The same will hold for vanadium flow batteries. Extra tanks and chemicals will serve for seasonal storage.

            That said, it’s probably going to be cheaper to overbuild and import than to store for months.

          • Mint

            “Just” a large basin? I’ve shown you that $2/kWh is needed to barely make seasonal storage (1 cycle per year by definition) economical.

            No manmade storage reservoir will ever approach $2/kWh. Seasonal hydro storage is only economical when nature gives you 99.9% the reservoir for free and you just have to dam up a small part.

            Here’s an example:
            180 km^2 and 500m elevation (14m swing) to give you 1600 GWh. It’s impossible to build a manmade equivalent for $3B.

            You repeatedly underestimate the cost of building a ‘large basin’. Look at the Taum Sauk reservoir:
            $450M to rebuild just a small section of a 5.7M m^3 reservoir. At 240m head, it works out to 3.7 GWh. Even if the whole reservoir cost only $450M, it’d be over $100/kWh. PuHS is cheap when the reservoir is entirely or mostly built for you.

            Same with vanadium flow batteries. It’ll never reach $20/kWh, and probably not even $200/kWh. $2/kWh is fantasy.

          • Bob_Wallace


            Let’s assume we build a closed loop PuHS site by excavating a reservoir on a high bluff and a second one on the plain below. Bore and line the connecting tunnel. Install the pump/turbine. Run the transmission line.

            Now the cost of adding seasonal storage becomes the cost of making the “holes” larger. Your link is not about enlarging reservoirs, it’s about repairing a failed system.

            $2/kWh is your creation. One would need to calculate how much soil an excavator could move in an hour and how much larger a hole that would make. Based on having a couple of ponds excavated I’d guess that one could easily dig a kWh hole for $2. But that’s a guess, I haven’t done the math.
            Additionally, I’ve never said that vanadium batteries would drop to $20/kWh. I’ve never put a price on the cost of a storage tank full of vanadium. But the comparison to a larger reservoir holds. Adding additional storage to an existing vanadium system would be much less than building one from scratch.

            Finally, I’ve never said that we are likely to use seasonal storage. If you look up only to my last comment I said – “… it’s probably going to be cheaper to overbuild and import than to store for months”.

            Look back to the Budischak study. They found no need to store energy for more than a few days at a time. And by adding load-shifting, hydro, tidal, biofuel, and power swaps their need for storage would have dropped further.

            Just so you are clear – I see it highly unlikely that we will do seasonal storage to any great extent.

            Germany may do a bit with hydrogen since they already have the storage ability, but that’s a unique case. And we may store biomass on a seasonal basis, cutting in the summer/fall for winter generation. Again, not a major input.

            I think it highly more likely we’ll build multiple day storage and build out our generation systems to keep the lights on.

          • Mint

            Everything I wrote here is about seasonal storage. That is the context of the conversation.

        • Poningru

          Rooftop solar produces power at location of consumption, all other generators rely on the transmission system to get the power to the location of consumption.
          Most of the grid costs comes from construction and maintanence of transmission systems (long distance & last mile) that can handle at minimum the highest peak power, if most of their customers go solar this high peak goes down, along with it the associated high cost goes down (assuming the utility made the right infrastructure choices).

          • Mint

            No, rooftop still needs distribution. Residences consume most power in the evening when there’s little solar generation, so there’s no reduction in transmission capacity requirements, and solar production needs to be exported to the rest of the neighborhood or even to commercial/industrial areas. Cloudy winter days also hammer solar output, so the grid must remain capable of providing almost all demand w/o solar.

          • Mike Shurtleff

            “Residences consume most power in the evening when there’s little solar
            generation, so there’s no reduction in transmission capacity
            requirements, and solar production needs to be exported to the rest of
            the neighborhood or even to commercial/industrial areas.”
            Low-cost residential storage is going to reduce this requirement for power from the grid on most evenings.

            “Cloudy winter days also hammer solar output, so the grid must remain capable of providing almost all demand w/o solar.”
            Need grid or back-up generators for this still.

          • Mint

            There is no “most evenings” when it comes to transmission capacity. You either need it or your don’t, whether 1 time a year or 365 times.

            If you want grid backup (and every net-metered customer does) when your array has low output and your storage is empty, and you want your summer noon generation to be exported to the grid, then you need the same transmission capacity as a non-solar customer.

            FYI, net metered customers have no use for storage or generators. It’s a waste of money for them.

          • Jenkins Lee

            Mint. Your whole point will become mute once backup storage prices drop and people can leave the grid completely.

          • Mike Shurtleff

            “There is no “most evenings” when it comes to transmission capacity. You either need it or your don’t, whether 1 time a year or 365 times.”
            Not true. Most of electricity in the Southern USA is for AC. If the sun ain’t out, then you don’t need as much power. This means you have a reduction in transmission required. Again, this is boldly illustrated by experience in Australia.
            Further, low-cost distributed end-of-grid storage combined with satellite weather prediction, smart grid controls, and smart appliances can help avoid infrequent and inconvenient backup power draws.

            “you want your summer noon generation to be exported to the grid”
            No, this does not need much in the way of transmission if your excess power is going to supply your neighbors with no Solar PV.

            “then you need the same transmission capacity as a non-solar customer”
            No, with Solar PV and Storage you will not have the same peak power requirements. With smart grid controls and your peak power user being AC in sunny areas, you will not ever have the same peak demand levels. This will not be the case everywhere, but it will be the case in many areas.

            “FYI, net metered customers have no use for storage or generators. It’s a waste of money for them.”
            Don’t be so self impressed. That’s obvious, but you’ve already said net metering can’t last and I agree. …so Solar PV owners will need a better deal from the utilities or some will opt for storage and in some cases generators for off-grid living. This is going to be the case for many Australians. In some cases it will be with the utilities blessing, and maybe assistance, so they can reduce their transmission costs …and everyone can have lower cost electricity.

          • Mint

            You are wrong about demand profiles. Once you have a few percent solar, the remaining demand peak is in the evening. Here’s an example during SoCal’s sweltering heatwave in Sept:

            Look at the last graph. After you subtract the limited amount of solar we currently have, the “net load” peak use is 7-8pm. At that time, solar and wind are outputting next to nothing.

            You’re also wrong about Australia. Peak use is in the evening:


            End of grid storage will not let you disconnect economically because winter solar production is low (30% of summer or less in most states). You’ll either have to overbuild solar by a factor of 3 or use a generator a lot. Overbuilding will only be economical at 5c/kWh

          • Mike Shurtleff

            See comment above:
            1) Your first reference shows some drop in the max grid demand as a result of added renewables. You are correct that there remains a peak at 6pm to 10pm.
            2) The 6pmto 10pm peak can be seen more clearly in the 2nd and 3rd references showing Australian demand with a larger penetration of Solar PV already added in. A few hours of Distributed Storage can and will reduce that peak, as well.
            The Australians do need to start moving their water heating from night to the middle of the day when the sun is shining. This would get rid of the silly late night peak, level their demand curve out nicely, and allow them to heat their water with additional Solar PV …especially if the utility is going mess with them.
            A huge reduction in grid demand is available to them now. Look for it to happen.

            “End of grid storage will not let you disconnect economically because
            winter solar production is low (30% of summer or less in most states).”
            You must be looking a 48 state average, or something. Go back to your Australian references. It curves don’t change that much with the relatively high penetration of residential rooftop Solar PV in Australia. Same in Hawaii. Same again in S. Cal. Your point does apply for Northern States in US.

            Overbuilding by factor of 2, with Residential Solar PV at 8c/kWh will work fine thanks.

          • Calamity_Jean

            “(assuming the utility made the right infrastructure choices). “

            Not always a safe assumption, I’m sorry to say.

    • Mike Shurtleff

      kind of a cup half empty guy, eh?

      No, grid parity assuming NO net metering, or maybe just wholesale when they need it metering. Please reference where Vishal Shah and his team have made the assumption of always having net metering.

      No, there is no limit at 5% penetration.

      Solar PV, Storage, and Wind are growing globally. No one country can stop it. No number of utilities can stop it. When people see others benefiting when they are being prevented they will demand a change …from utilities (Hawaii) …from politicians (Abbott in Australia).

      • Mint

        I can’t find Vishal’s report, but the article says Vivint is selling for 14-15c/kWh. Wholesale is 4-6c/kWh. So yes, the report is undoubtedly assuming net metering when they say “grid parity”.

        • Mike Shurtleff

          No, they are selling to the retail end of the grid market. Net metering no doubt helps, but at some point it has to go. Solar PV will still be far enough below to keep right on expanding.
          Slow it down briefly and it gets cheaper and grows again. Even if the utilities win a battle (Australia) they lose the war. Utility paradigm is going to change.

          • Mint

            “No, they are selling to the retail end of the grid market.”

            I’m not sure which part of my post this applies to. Shah’s prediction was in the report for Vivint, who is selling to retail with a zero-down Solar-City-like sales model at 14-15c/kWh, so why would ‘parity’ refer to anything but retail price?

            As for Australia, I agree. The utilities there have obscene costs of 20c/kWh not including electricity generation. That’s 3-5x what you see in North America. Grid disconnect with solar+storage is going to explode there. But I brought it up because it proves what I was saying before: high net metering penetration will make utilities jack rates or decouple grid and energy.

          • Mike Shurtleff

            My point is end-of-grid Solar PV can make sense without Net Metering, if Solar PV on your roof is sufficiently below the cost end-of-grid power from the utility. End-of-grid parity is already here in several states. The margin of savings will be significantly larger by 2016. Net Metering will not be needed in those states by then, nor the ITC.

            In places like Hawaii, you probably don’t need it now.

            No Australia is not a good example. The utility is jacking up prices to cover their arrogant and inexcusable over-building of thermal plants (coal & NG). They are covering their losses due to “stranded assets”. Coal and NG thermal plants that are not paying for themselves. They are blaming those high electricity costs on Solar PV and Net Metering and it’s mostly not true. Besides their stranded FF plant costs being contributors, they are usually able to sell that excess residential Solar PV to other customers. In at least some places in Australia, the utility now gives absolutely nothing to some Solar PV owners for the power they put on the grid. Crooked stuff!

            Residential Solar PV is well below half the end-of-grid utility power costs in many places in Australia. It is a prime example of where Solar PV can and has brought down the cost of electricity …but foolish and crooked management by the utilities has kept it higher than it should be.

            You are correct, and they are correct, that Net Metering cannot be maintained at high penentrations of Solar PV. You did not state that correctly above, you wrote something that implied high-levels of Solar PV were a problem period. Some of the utilities in Australia are just lying about their situation, blaming their poor management on Solar PV and Wind in general.

            ….and then there’s Hawaii, HELCO anyway, who cannot understand they need to change their grid, a grid with very short distance connections, to work bi-directionally ….so they don’t get voltage build-ups, so they can redistribute that Solar PV electricity to others on their island who could use that cheap electricity …compared to diesel electricity, Solar PV is very cheap electricity. HELCO needs recalibration if they continue to be obstructive.

          • Mint

            Mike, look at the article. It’s about Vivint solar offering 14-15c/kWh. Does that sound like it’s two years from wholesale parity (4-6c/kWh)?

            Nobody is going to install rooftop solar for you at that price. At best you’ll get 2500kWh/yr for each 1kW of panels, which is worth $100-150/yr at end-of-grid. Who is going to install 10kW of solar for a $1000/yr income stream?

            You can look at AUS cost breakdown here:

            You’re misinformed. It’s not overbuilt NG/coal plants, which would contribute to at most a couple cents per kWh. It’s insanely pricey transmission:

            Environmental policies: 4.49c/kWh
            Transmission & distribution: 14.40c/kWh
            Wholesale: 5.26c/kWh
            Retail sale: 3.8c/kWh

          • Mike Shurtleff

            Vivint is offering 14-15c/kWh for residential/commercial installations, retail electricity at end-of-grid. This is half of some tiered rates in Southern California, so well below end-of-grid parity. The article is about rooftop Solar PV being at end-of-grid parity. It is not talking about source-of-grid parity.

            Never-the-less, we are close to source-of-grid parity when you compare the cost of Utility Scale Solar PV to the source-of-grid cost. Recent bids for Utility Scale Solar PV have come down to 5c/kWh and 6c/kWh (without the ITC that would be more like 7c/kWh and 8c/kWh), …so yes UTILITY SCALE Solar PV looks like it is only a few years away from parity with source-of-grid costs.

            Those are both apples to apples comparisons. Please stop trying to make apples to oranges comparisons, by comparing the cost of Residential Solar PV to the cost of source-of-grid power.


            Residential Solar PV is installed for Vivint costs or even less. This means it is half the cost of the 27.95c/kWh you reference for end-of-grid power in Australia. Thank you for making one of my points.

            If you pay wholesale costs, (end-of-grid) costs, instead of nothing for residential power added to the grid, then you can redistribute this locally and reduce everyone’s electricity bill.

            You are correct that the greatest cost is “insanely pricey transmission” costs. Australia has a relatively small population distributed over a large area. It will be cheaper now for supply electricity to some individuals, and to some towns, using Solar PV, maybe wind, Battery storage and backup generators …and NO grid. If designed correctly the generators will be rarely used. Some Utilities in Australia are already recommending this:
   – October 2014
            “Ergon Energy Wants To Get Some Remote Customers Off The Grid”
            …so two ways there that Solar PV should be helping to reduce electricity costs.

            Net Metering is probably adding some cost, but it’s hard to tell how much from the numbers you reference. Who paid for this study? What does “Environmental Policies” really mean? Does that include anti-pollution scrubbers on coal plants? Where is the accounting for the excessive “meter reading” costs that some utilities are forcing homes and businesses with Solar PV to pay for? …even though they don’t actually do anything different than for regular customers. Lame.

            I do agree that Net Metering paid to Solar PV owners should be somewhere between wholesale and retail costs, and only when the utility has a customer for that extra power, and this should be overseen be a third party with no coal/NG power generation interests.
            The Net Metering paid to Solar PV owners should not be zero, as is now true in some places in Australia.

          • Bob_Wallace

            “Mike, look at the article. It’s about Vivint solar offering 14-15c/kWh. Does that sound like it’s two years from wholesale parity (4-6c/kWh)?”

            Did you not jump from end of grid to wholesale?

            All but three US states have >10c/kWh electricity. The cheapest is Washington at 8.93.


            I’m not sure that I’d predict a 50 state sweep that soon, but a $2/watt and 5.4% discount rate brings the cost to around 11c/kWh which covers about 44 states.

            My guess is that we’ll see 50 state end of grid parity within five years.

            Getting utility solar down to $1/watt will mean ~ 6c/kWh wholesale prices. And then one needs to factor in that the cost will not increase with inflation and changes in fuel prices. A 6c contract means less than 5c over the life of the contract.

          • Mint

            You’re quoting retail prices, and rooftop PV only competes with that if you have net metering.

            Even at end-of-grid, you’re not going to get >10c/kWh for electricity you sell into the grid. Not unless you’re being compensated for reduction of RPS purchases or CO2 emissions, but even then utility solar would get that credit too. End-of-grid is only worth 1-2c/kWh more than wholesale.

            I agree that utility solar could reach $1/W eventually and be wholesale competitive. But rooftop solar has no chance.

          • Bob_Wallace

            Rooftop solar has no chance of reaching $1/watt? It’s already below $2/watt in some places. Panel prices are likely to drop by another 20 – 30 cents. Efficiency will improve which will lower racking/labor expenses. Inverters are likely to drop in price.

            And, yes, I’m talking about rooftop as well as utility. That was the topic before you brought in utility/wholesale.

            The price of both will continue down. At some point rooftop will reach grid parity. At some point utility will drop below 5c/kWh.

          • Mint

            I don’t disagree with any of your individual cost claims, but they will not reduce non-panel costs to $0.70/W, at least not by 2030. That’s wholly unreasonable in a society where we pay people decent wages and have safety regulations for roofing, wiring, etc. Even in Germany, non-panel costs are decreasing really slowly now.

          • Bob_Wallace

            OK, right now rooftop in Germany costs a few pennies over $1.60/watt. (I’m not interested in arguing the second to the right of decimal point number.)

            I assume that Germans are paying around $0.60/watt for panels. Our average is a bit above that. That makes their installation cost about $1 and they pay people decent wages and have safety regs.

            Now, let’s assume an increase in efficiency from 16% to 20% over the next ten years. That’s 25% more power from the same amount of racking, labor, wiring. A drop from $1 to 80 cents per watt.

            Cut inverter prices some, come up with even more efficient racking and you’re tightening down on 70 cents.

            A move from 16% efficient panels to 22% efficient panels would bring BoS to about 75 cents. (Some costs such as wire length wouldn’t be decreased.)

            Panasonic has a 25.6% efficient panel. If panels that efficient become the norm then we’d move under 70 cents.


            I’m not going to bet against 70 cent BoS costs over the next 15 years.

          • Mint

            $1.60/W? Where did you see that? The most recent data I saw was this:

            And aren’t panels $0.50/W now?

            Regarding safety/wages, Germany has no min wage (I’m not knocking them, as their safety net is decent), and they don’t need licensed electricians for installation. I don’t see that happening in the US.

          • Bob_Wallace

            The CT data is for 2013.

            According to this site Germany averaged $1.88/watt in 2013. It’s a blend of larger and smaller installations but heavily weighted toward smaller. It’s a bit lower than the $2/watt on the CT article you linked. 6% difference.

            Has now dropped to $1.63. August 2014 € 1.310.


            Moving from 16% to >24% efficiency panels could cut the hell out of BoS costs. Again, betting against $1/watt by 2030 is pretty risky.

            Weekly prices high/low/average. Average for silicon $0.62.


          • Mint

            That’s for 100kW. You could argue that for large-ish commercial rooftop (which I’ll agree has installation efficiency approaching utility solar), but not residential.

          • Bob_Wallace

            From the photovoltaik site –

            “The current price for a solar power plant to 100 kWp was in August 2014 turnkey 1,310 euros net per kWp .”


            It’s a blended residential/commercial price. Germany is pretty heavy on residential rooftop.

          • Mint

            I can’t find current distribution data for solar installations, but I found this for 2010:

            Seems highly unlikely to me that they’re “heavy on residential”. But if you have data to suggest otherwise, I’d love to see it.

    • Jenkins Lee

      Wrong. It’s called the “Death Spiral” for the utilities. As they keep raising rates, more and more users will switch to solar.

Back to Top ↑