GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Guest Blogs

Urban Rustic: Installing a Solar Electric System

With the addition of a photovoltaic system, the house has a shot at net-zero energy performance

With a rating of less than 3 kW, this photovoltaic (PV) system isn't huge, but it may be big enough for net-zero energy performance on a house that's been carefully air-sealed and superinsulated.
Image Credit: All photos: Eric Whetzel
View Gallery 5 images
With a rating of less than 3 kW, this photovoltaic (PV) system isn't huge, but it may be big enough for net-zero energy performance on a house that's been carefully air-sealed and superinsulated.
Image Credit: All photos: Eric Whetzel
Anthony from Rethink runs conduit through the air barrier at the attic floor level. The conduit starts at the electrical panel in the basement and ultimately ends at the panels on the roof. Two Roflex gaskets — one on top, another on the bottom — protect the integrity of the air barrier where a 3/4-inch electrical conduit must pass through. The disconnect box on the right is too close to the house to allow installation of insulation, furring, and siding. It didn't take long to correct the problem by adding an extension and moving the box out a few inches..

Editor’s note: This post is one of a series by Eric Whetzel about the design and construction of his house in Palatine, Illinois, a suburb of Chicago. The first blog in his series was called An Introduction to a New Passive House Project; a list of Eric’s previous posts appears below. For more details, see Eric’s blog, Kimchi & Kraut.

After deciding to pursue a combination of Passive House and The Pretty Good House concepts, which entail careful planning and attention to air-sealing along with a significant amount of insulation, we knew we could have a shot at net-zero or zero net energy (ZNE) — meaning we could potentially produce as much energy as we use with photovoltaic (PV) panels installed on the roof.

To find an installer in our area, we used the website Energy Sage. In addition to publishing useful articles and information about solar, they also work with installers who can provide consumers with competitive bids. It didn’t happen overnight, but in about a week or two, we got three or four bids and ultimately decided to go with Rethink Electric in Geneva, Illinois.

Based on the suggestions from Energy Sage and Rethink, we ended up with the following system:

  • 2.915 kW DC system
  • 4,059 kWh of production annually
  • 11 Canadian Solar panels
  • 265W Module Enphase M250 (Microinverter)

Also included was web-based monitoring of the system’s production.

In theory, this system could produce more energy than we use (it’s just my wife, my daughter, and I who will be living in the house). But there a few caveats: we need to stick to all-LED lighting, use Energy-Star-rated appliances, hope the heat-pump water heater works as advertised, and be careful not to use electricity when it’s unnecessary. That means making sure we turn off the lights as we’re leaving a room, and limiting phantom electrical loads.

Installing the system

Based on other projects I’ve read about, even homes initially built to the ZNE standard sometimes fail, in terms of overall performance, because of occupant behavior. Only time will tell what impact our solar array will have on our utility bills. It looks like the worst-case scenario would be needing to add four to six more panels to get to ZNE or even carbon positive.

Installation by Rethink went really well, and they were happy to work with me on properly air sealing the conduit that runs from the basement at the main panel up into the attic. The conduit eventually terminates on the roof, where it is connected to the panels. (See Images #2 and #3 below.)

It was only after the installation that I realized an important detail had gone all wrong (see Image #4 below). It was my screw-up. I was so worried about getting the air sealing details right on the interior, from the main floor to the attic, that I completely forgot to let Anthony know about extending his disconnect box 6 inches, to what will be our finished surface (once two layers of Roxul and two layers of 1×4 furring strips, along with cedar siding, are installed).

The day after the box went in, I came walking around the corner of the house, saw this, and literally slapped my forehead (while spitting out a few choice expletives), as I realized my screw-up. Thankfully, Anthony was able to come back out and make the necessary adjustment (see Image #5).

The cost

Here’s the cost breakdown on our system (if trends continue, a similar system should be less expensive in the future):

Initial investment:                     $12,519.50

Federal investment tax credit:   $(3,755.85)

Net cost in the first year:            $8,763.65

Solar renewable energy credits: $(3,816.00)

Net cost after all rebates:            $4,947.65

It will be interesting to follow the performance of the solar panels over the course of a calendar year or two, just to find out exactly how well they perform. I’ll come back here and post information on our monthly utility statements, noting output of the panels and our use, to give people a better sense of actual performance. Hopefully this will help others in the planning stages of their own projects to decide if solar (and how much of it) is right for them.


  1. calum_wilde | | #1

    Did you detail the expected
    Did you detail the expected electrical load in any of the previous posts? 4kW seems exceedingly low to me after looking at my power bills, I'd love to see what the plan is to get there.

  2. _Stephen_ | | #2

    For my certified net Zero
    For my certified net Zero home, we basically put as many panels on as would fit. Technically, we're way past net Zero for just the home, but the solar array should now cover the electric car's have as well.

  3. Jon_R | | #3

    Is there any kind of thermal
    Is there any kind of thermal break possible on steel conduit?

  4. user-6944519 | | #4

    net zero on 4k blows my mind
    Hitting Net Zero with only 4MWh of annual production seems really really optimistic. My electric car (Leaf) uses 3.2MWh annually! My GE heat pump water heater with a 3.25 Energy Factor is rated at 1.3MWh/yr (no meter for it). And you've still got heat pumps, fridge, dishwasher, ventilation, and laundry machines to power.

    In 2016 I deployed a 9.8kW array. In calendar year 2017 my array produced 9,288 kWh. The average household in this region has an energy budget of 22,000 kWh/yr and so I was quite proud when we used just 9,180 kWh and achieved net zero. If you can hit net zero with just 4MWh/yr of production, I salute you and yours, good sir. That would be remarkable.

    Comparisons: my house was build in 1955 when energy conservation meant asking someone else to grab you a beer. Seattle has 4,700 heating degree days compared to your 5,500. My house is 2,300 sqft, so 38% larger. We have four occupants (1 more). We have all LED lights, power strips normally off to kill parasitic loads, all EnergyStar or better appliances, R-36 outside-the-sheathing insulation, spray foam and air sealed portions of the floor and roof, new 0.19 low-e windows throughout, air sealed like a madman while adding the aforementioned insulation, HPWH, low-flow (but not too low) faucets and fixtures.

    Many thanks to GBA and Building Science for much of the wisdom, knowledge, and inspiration.

  5. itserich | | #5

    why is net zero energy meaningful?
    I don't understand why people consider net zero energy meaningful.

    If I cut my energy use by half, that is meaningful.

    Whether I install my own solar panels, to the grid, or a utility installs solar panels, which I receive through the grid, does not make a difference in the amount of solar power in the world.

    I can see the appeal of producing power locally with a battery, to be off grid, but when a utility can install solar more cheaply than myself, and I rely on the grid either way, I don't understand the pride in net zero.

    The real pride is energy efficiency, which I understand cannot generally be seen from the street, unlike solar panels.

  6. Expert Member
    Michael Maines | | #6

    Net zero is meaningful
    Net zero is meaningful because it's an easy to define, universally equivalent goal--produce as much energy as you use. Once you improve your envelope to a certain level (roughly "Pretty Good House" levels), it's more cost-effective to put additional money into PV panels.

  7. Expert Member
    Dana Dorsett | | #7

    It's more complicated than that @ Erich Riesenberg
    Solar power from a utility scale solar farm adds substantially to the grid infrastructure requirements to get the power from the PV to the load. Any increase in load takes up capacity on the transmission and distribution grids, pours more power through the transformer feeding your house.

    With behind the meter solar, some of the solar power generated from a rooftop never goes onto the grid at all, consumed on your side of the meter. The rest goes on to the local distribution network on your side of the substation, usually consumed by loads on the same transformer. That lowers the net load on the transmission & distribution grids, the substations, even the transformer feeding your house.

    As PV within the distribution grid exceeds the local load, batteries can alleviate the excess, while reducing the amount of the power generated on your side of the meter from ever reaching the grid, freeing up even more grid capacity. In over-built PV neighborhoods of Hawaii, self-consumption controls are the only way to get a permit to put PV on the house. Whether the excess power is dumped into a battery or an electric hot water heater, it still frees up grid capacity, and never back-feeds the grid. Smart car chargers already exist that will perform that function too.

    Distributed power has marginally higher efficiency than centralize power utilizing the grid infrastructure, and by freeing up grid capacity it lowers the wear & tear on the grid, for an overall more reliable, more efficient power grid system.

    But you're right in the sense that Net Zero is just a stake in the ground on a ENERGY ONLY basis. A Net Zero house with no self-consumption storage (thermal or battery) and an electric tankless water heater or an 8 ton air conditioner is making HUGE demands on the grid infrastructure. (A tankless electric water heater is one of the least-green way to heat water- worse than an oil fired water heater or tankless coil in an oil boiler by many measures.) Net Zero LOAD would be the ideal, but smart grids and load management (on either side of the meter), can still maximize the locational value of the distributed PV generation.

    Unfortunately utility regulations still need to catch up to be able to fully utilize & remunerate distributed PV or distributed batteries, but that's coming, and sooner than most people are aware of. Already in Denmark electric vehicle (EV) owners that allow the utility to use their smart car charger and car battery as a controllable load to provide ancillary frequency and voltage control on the distribution grid get paid for that service. Those that also allow 2-way power flows both to & from the car battery & grid get paid even more.

    As more EVs go onto the grid, the need for smarter charging and load management goes up. The existing grid has PLENTY of capacity for the average load, even in an all-EV universe, but the potential for EVs to take down parts of the distribution grid is very real if left unmanaged. Distributed PV can help alleviate some of that load, but grid-aware chargers can make a mostly-PV grid possible, even with 1-way power flows (grid to car) by having distributed controlled loads to keep from backfeeding. A centralized PV generator & battery does nothing to facilitate throughput of the distribution grid in the EV-laden grid, but distributed PV does.

    In most locations the grid benefits of distributed PV still exceeds the higher cost of net-metering, but as more distributed PV gets built the grid benefits dwindle, but don't really disappear until it's creating chronic back feeding. As batteries get cheaper (the learning curve is comparable to that of PV), distributed battery + PV will become commonplace, and quite valuable to the grid if the battery can be utilized by the grid operator to flatten the demand peaks, as well as the valleys created by excess PV. The utility regulations have to change to make that all happen in a way that pays the battery owner fairly, but it will happen. A Net Zero house with a smart battery can be a very effective grid stabilizer, more so than centralized grid assets.

  8. itserich | | #8

    what is the point Michael?
    Michael, I can't comprehend the point you think you are trying to make. Sure it is an easy to understand goal. Doesn't make it worth anything.

  9. itserich | | #9

    Dana, I think I understand
    Dana, I think I understand your point, at least part of it. But there are benefits to large scale energy deployments, including economies of scale and siting. Surely a professional large scale deployment is going to be produce more power more efficiently than residential rooftop applications.

    I definitely agree with your last sentence, residential solar plus a battery makes a lot of sense and is a worthwhile goal on its own.

    I live in a state that is a leader in wind power. I can't see how installing solar on my east west roof would make any sort of sense, and think that would be true for most homes if the utility generated its own wind or solar.

  10. _Stephen_ | | #10

    In Response To Erich
    You'd think there would be economies of scale, but from a purely economic perspective, unsubsidized solar on my roof is objectively cheaper than grid power in my area.

    The problem with centralized production is that it is produced somewhere it isn't needed. So, you have to pay to move that power to places it IS needed. Distribution costs are non-trivial.

    Furthermore, I graciously allow myself to use my roof free of charge. However, commercial solar plants don't have this luxury. They have to pay to rent land or rooftops for this purpose. This rent increases every year.

    In addition, I justifiably don't charge myself for monitoring and maintenance of the array. Commercial solar plants don't get that either.

    All in, solar on my roof comes in cheaper than centralized grid solar in my area, by a large margin, while reducing peak grid loads.

  11. Jon_R | | #11

    See #20 below for support for
    See #20 below for support for utility scale solar typically being much more cost effective, even when you account for distribution costs. But sure, in some special cases, it may go the other way.

    I think a "atmospheric carbon per resident" metric would be a much better for determining "green" than residential NZE.

  12. itserich | | #12

    the grid
    Stephen, obviously most homes with solar panels still use the grid. I don't know the marginal cost difference between using the grid to bring all the energy to the home versus producing energy at the point of use and shipping some or all of it back to the grid.

    But, it would help keep the discussion honest if the fact that the grid is still required is not ignored.

    You may not calculate your cost to monitor and maintain the array, but the cost still exists.

    You conclude: "All in, solar on my roof comes in cheaper than centralized grid solar in my area, by a large margin, while reducing peak grid loads."

    Have you actually calculated this cost difference? Please, do share.

    Does this include the costs of storing the energy for use when the sun goes down? I think some utilities store some of that wind and sun for use at night.

  13. _Stephen_ | | #13

    Erich: I provide utility to
    Erich: I provide utility to the grid by trimming the summertime peak, which is why they're happy to provide me with a net metering agreement. I give them expensive peak power during the summer, in exchange for cheap night time baseload power in the winter.

    My costs all in for power are $0.077 / kWh, versus grid prices of around $0.11 / kWh for grid power. Inflation and rising costs modeled out over the 30 lifespan of the panels suggest a total savings of $110,000 once you take into account the $30k investment.

    Economically, its not even close. As panel prices fall, and grid prices rise, that's only going to get worse.

  14. GBA Editor
    Martin Holladay | | #14

    Response to Matt Simerson (Comment #4)
    Every family is different. Your family uses 9,180 kWh/year. Right now, I'm attending a building conference in Burlington, Vermont, where I heard a presentation by Marc Rosenbaum, an energy consultant who lives in Massachusetts. At their all-electric house, Marc and his wife used 4,099 kWh of electricity last year. There are two people in the house, and they don't have a television. Their house is heated by a single Fujitsu ductless minisplit unit.

    There are all kinds of reasons why other families use more than 4,099 kWh per year -- they have a bigger family, they live in a bigger house, or they live in a colder climate. But a family that uses only 4,099 kWh a year doesn't need a 9-kW photovoltaic array. Some families can hit net zero with a much smaller PV array.

  15. itserich | | #15

    basic facts
    Stephen, I asked you to substantiate your claim: "All in, solar on my roof comes in cheaper than centralized grid solar in my area, by a large margin, while reducing peak grid loads." You also claimed your cost preference included no subsidies.

    To substantiate this claim, you wrote: "My costs all in for power are $0.077 / kWh, versus grid prices of around $0.11 / kWh for grid power."

    Your all in cost is, in fact, heavily subsidized, right? So it bears no relation to your original claim. And of course, the grid price is not the cost of a utility solar project. Right?

    I won't get into the other details you ignore, because, I guess you have never done the calculation you originally claimed. Which I understand, because it would be very hard for a consumer to calculate the utility's cost.

    I understand we live in an ego driven society, where, somehow, claiming to be "net zero" signals some sort of virtue. I just don't think the value of the ego is too great.

  16. itserich | | #16

    There is something wrong with
    There is something wrong with the Captcha. I hit the Post button once, and it came back with a message about a "reuse" attack.

  17. GBA Editor
    Martin Holladay | | #17

    Response to Erich Riesenberg
    Thanks for the feedback. GBA is aware of problems with our spam detection system and our CAPTCHA feature. We've been promised that these problems will go away when our tech team does a web site redesign and relaunch -- scheduled for this year.

  18. GBA Editor
    Martin Holladay | | #18

    Response to Erich Riesenberg
    Clearly the economics are complicated -- and changing monthly. Your skepticism is already or will soon be obsolete, however. While the installation costs for a homeowner-owned PV system are higher than the installation costs for a large utility-owned system, the benefits to the utility of distributed PV (as pointed out by Dana Dorsett in Comment #7) are significant, and may well justify the higher installation costs.

    Meanwhile, as we debate the finer points of this economic analysis, the unsubsidized cost of installed PV continues to fall, and in many locations, new PV and wind installations are less expensive than coal-burning power plants.

  19. itserich | | #19

    not complicated
    Thanks Martin. The economics seem simple. Most existing homes are not built with any thought to positioning solar, and a utility project is going to have siting, product margin, and knowledge benefits. I would guess fewer than 50% of homes have any solar potential due to roof orientation.

    My local utility is claiming to be toward 100% renewable energy, ignoring the huge natural gas they use, while also making clear he would not be building any renewable without tax subsidies.

    As to Dana's comment about the grid, installing a battery at home, without the solar, could be a solution to a lot of problems. Just an uninformed guess.

  20. GBA Editor
    Martin Holladay | | #20

    Response to Erich Riesenberg
    Battery costs are high right now, and will have to fall quite a bit before they'll be attractive to homeowners. Eventually some utilities may offer battery owners a lot of money for peak power on hot summer afternoons -- and when that happens, the economics may change. My guess is that the first homeowners who will take advantage of those opportunities will be owners of electric vehicles plugged into the grid.

  21. GBA Editor
    Martin Holladay | | #21

    Response to Erich Riesenberg
    Battery costs are high right now, and will have to fall quite a bit before they'll be attractive to homeowners. Eventually some utilities may offer battery owners a lot of money for peak power on hot summer afternoons -- and when that happens, the economics may change. My guess is that the first homeowners who will take advantage of those opportunities will be owners of electric vehicles plugged into the grid.

  22. itserich | | #22

    Yes, battery costs are very
    Yes, battery costs are very high. When electric cars are common, there will be more incentive for excess energy to be stored at home during the day, to recharge cars at night.

    Of course, as the cost example in this article shows, home solar is heavily subsidized, so real world costs are not the only test of feasibility. Perhaps batteries can be painted in glow in the dark green and mounted to the roof to show the world the homeowner is respectable.

  23. Jon_R | | #23

    A more realistic view of
    A more realistic view of net metered residential solar is that you are selling the utility power at a high price when they can generate it themselves with utility scale solar at a much lower cost (say $.11 vs $.045). A losing deal (paid for by other customers), even in the unlikely case that there is a $.05 savings in distribution cost.

    The good news is that every time you drive by a neighbor's net metered solar array, you can say "I helped pay for those".

  24. Expert Member
    Dana Dorsett | | #24

    Response to various.
    PV is extremely scalable, and not more efficient at the large scale than residential scale.

    Battery costs are only high if the only purpose (and remuneration) is energy storage. When used for ancillary grid services, and competitively PAID for those services, batteries are rapidly becoming the go-to solution at the grid scale, and will be financially rational at the residential scale before 2030.

    Batteries used to manage peak draws are financially rational RIGHT NOW for commercial ratepayers being assessed demand charges. A recent rate case in Massachusetts approved (for the first time in that state) demand charges for residential customers with PV, which instantly created a market for residential demand-charge management batteries. That rate case is being appealed, so the market isn't really begun yet, but if it survives appeal I'm expecting scaled down versions of the commercial demand-charge management systems to start showing up. If/when regulations allow distributed asset aggregators to participate fully in the wholesale electricity markets, paying the asset owners a return for those uses, the number of distributed battery (and PV) situations that will be financially viable will be MUCH higher than it is right now. But over a year ago there was already a good financial case for demand charge management batteries for commercial ratepayers in at least a handful of states:

    Home solar is currently heavily subsidized, but utility scale solar is too. So?

    The all-in UNsubsized levelized cost of residential solar in the US is between 19-32 cents/kwh (with many exceptions to prove the rule), compared to about 4.5-5.5 cents for utility scale:

    But residential solar doesn't have to be that expensive. It's about 7-8 cents/kwh (pre-subsidy) in Germany & Australia, using comparable panels, inverters & racking systems. The US residential & commercial PV market suffers from truly bloated "soft costs", everything from marketing to permitting/inspection etc., some of it driven by utilities protecting their vested interests as well as local/state politics.

    The learning curve a batteries is well into double-digits percentage cost reduction per doubling of capacity, and production capacity is booming worldwide. Hawaii is seeing exponential growth in their behind the meter battery market, due to a combination of high retail electricity pricing, the ending of net metering for residential PV, and in some over-solarized neighborhoods a requirement for self-consumption for any new PV. It's only a matter of time before the learning curves of batteries & PV catch up to YOUR residential rate structures. See:

    The disruptive aspects of these double-digit learning curves and scalabilitiy for both PV and batteries is profound, and utility business models and how the grid is managed is in for a HUGE overhaul over the next decade.

  25. itserich | | #25

    so, utility scale is cheaper
    Dana, I am confused by these two statements which seem to contradict themselves:

    "PV is extremely scalable, and not more efficient at the large scale than residential scale."


    "The all-in UNsubsized levelized cost of residential solar in the US is between 19-32 cents/kwh (with many exceptions to prove the rule), compared to about 4.5-5.5 cents for utility scale"

    I am not going to be able to comprehend how those two statements are reconciled. It just seems obvious to me that a properly sited and professionally maintained solar array is going to achieve better results even with the same solar panels, than a rooftop installation which was not designed to capture the most sun. Maybe I am wrong and the equipment is all that matters and the cost is similar retail and wholesale.

    As to Dana's claim that the primary difference in cost between home solar and utility is solar is soft i.e. "unnecessary" costs, that would mean 75%-85% of the cost to install home solar is truly soft, as implied if the actual hard costs to install home solar is 5 cents, increased to 19-32 cents by soft costs. Reminds me of the Mid American guy who said the company is cutting energy efficiency programs by 50% because of administrative reductions. Not credible.

    Net Zero reminds me a lot of Energy Stars for buildings - the reality does not match the hype.

    Net Zero reminds me of common claims that Passive Houses use "up to 90% less energy" or "no active heating" when these statements are almost always wrong.

    And, at the point where it really matters, I think clearly, a more efficient use of energy is more important than where the solar is captured. Which is my primary point. There may be a benefit to both, but I think efficiency is more important. Maybe I am wrong.

  26. itserich | | #26

    how to face solar panels
    Here is an article about the efforts to try and convince homeowners to face their panels west, instead of south. I would guess this is the tip of the large iceberg about the problems with home solar.

  27. Expert Member
    Dana Dorsett | | #27

    Efficiency isn't the same as cost-efficiency @ Erich Riesnberg
    The raw electrical efficiency from photons hitting the panel to power consumed by the load of utility scale PV is lower than residential PV, since there are grid losses, and sometimes curtailment losses due to maxed out grid-capacity.

    There is some curtailment losses to residential PV too, but that has to do with sizing the inverter optimally for the array, not grid congestion. Most inverters are maxed out on bright late-winter days, and not delivering the full panel output capacity to the load (or grid), but those are pretty small on average compared to utility scale PV.

    Many residential systems in my area come with free monitoring of the performance by the installer to detect any issues or early failures during the warranty period (it's cheaper to fix something before it fries and takes down other parts of the system.)

    Many utility scale arrays boost efficiency with tracking mounts and siting them in locations with higher insolation, but the latter comes with additional grid infrastructure costs that are accounted for in the levelized cost of energy. The levelized cost of the ENERGY is lower for utility scale PV, as reflected by the Lazard analysis, but that's measured at the inverter output, not panel-to-load, and doesn't include any distribution or curtailment losses. The levelized cost numbers don't reflect the grid cost of delivery of that energy.

    The delivery cost for distributed PV is very low, since it doesn't use much grid resource (none, for power consumed behind the meter), and is indeed often cost negative, since frees up grid capacity for other generators & loads. The cost or enhanced value of the distributed PV is very grid-location specific. With utility scale PV there's always an additional grid cost (often substantial), but again, it varies by location within the overall grid structure. Consolidated Edison in NY is saving the ratepayers money in reduced grid infrastructure cost a the Brooklyn-Queens substation using ratepayer-owned distributed assets and other "non-wires alternatives" (the industry buzzword) to reduce grid congestion rather than building more substation capacity.

    That particular project isn't without well informed critics, even though the concept in general has merit:

  28. itserich | | #28

    peak loads
    Is the article incorrect that nearly all home solar is positioned to maximize solar at non peak times?

    It seems that producing energy when it is not needed is more wasteful than the friction of grid losses, but that is just my uninformed speculation.

    If houses with solar do in fact add to the demand on the grid during peak loads, it is confusing how one can conclude those homes benefit the grid. But, I am no expert.

    "Houses with solar systems consume less than half as much utility-delivered electricity as ordinary houses, the study found. But from about 4 p.m. through the night, they consume more, and they add to the system’s peak demand, which comes around 5 p.m.

    Pointing panels to the west means that in the hour beginning at 5 p.m., they produce 55 percent of their peak output. So a 10-kilowatt system would make 5.5 kilowatts. But point them to the south to maximize total output, and when the electric grid needs it most, they are producing only 15 percent of peak, or 1.5 kilowatts."

  29. _Stephen_ | | #29

    In reponse to Erich
    I'm not in the US, so my costs are not subsidized at all. I payed fair market value for the array.

    The only "subsidy" you could argue that I received is the net-metering agreement, except the local utility is actively encouraging as many solar arrays as possible, because most of the generation they displace is extremely high cost summer peaking generation at $0.40 / kWh. They in turn get to "repay me" with $0.03 / kWh electricity as baseload winter.

    Trust me, they're getting the better deal here.

  30. itserich | | #30

    Thanks Stephen
    Sounds like your situation is the inverse of America's, where home based solar is much more costly and produces a lot of energy that cannot be used. Wonder why there is such a difference.

  31. itserich | | #31

    utility scale more cost and energy efficient
    Dana, I don't understand this comment, beyond of course it is a simple fact. I don't understand its application. "Efficiency isn't the same as cost-efficiency"

    Assuming utility scaled solar is both cheaper in dollar terms, and does a better job creating energy during peak periods, it seems better in terms of both energy and cost efficiency.

    I wish I had the common sense to understand what I am missing.

  32. _Stephen_ | | #32

    I'm sure on an installed per
    I'm sure on an installed per kW price, my install is higher, because of labor and permitting costs. I paid about $2.70 / watt installed. I believe commercial installations in my area are around $2.10. / watt

    I think the reason why it is the case is that we are up north. The result is that the cooling season is much shorter, people don't spend a lot of money on good air conditioners and people expect to be cool especially at night. Setting the thermostat to 70 is common. This pushes us into a very heavy summer peaking situation. Solar installation power matches up with those peaks nearly perfectly, especially mine that is angled south west which pushes peak production further into the evening.

    Also, it's only been recently that solar arrays have been cheap enough to make sense at all up here, so solar penetration is pretty low overall.

  33. Expert Member
    Dana Dorsett | | #33

    An artifact of how it's remunerated, re; peak loads.
    To date most residential PV has been simply net metered, with no financial incentive for producing more during peak hours, just as with flat-rates there's no incentives for residential ratepayers to shed load during peak hours. Until PV gets well into a double-digit percentages of the total grid power delivered, it's always producing when there is electrical demand, and NEVER "... producing energy when it is not needed...".

    There is at least a reasonable overlap with south oriented PV and peak air conditioning loads, but not with the evening peak grid loads when people go home, crank up the AC, take a shower, cook dinner, then settle in for a night of binge-watching K-dramas on Hulu. A more westerly orientation would align better with the average day's peak, but at the cost of lower total kwh produced. If you're paying only based on energy, not peak draw (demand charges), or time of use rates, there is no incentive to do anything with the PV than orient it for maximum energy production. But even south facing solar is still producing some power at 5 pm half the year, offsetting some of the absolute peak and shifting the net grid peak load into later hours (at a lower net load than it would have been without the PV.)

    In most US regional grids on an average day with minimal (or no) air conditioning the grid load is lowest between midnight and 4AM, then rises pretty rapidly until 8AM, stays pretty flat until 4-5PM, then peaks in the early evening. A south facing array frees up grid capacity during the middle part of the day during the flat part of the grid load. As more PV goes onto the grid there's a noticeable mid-day dip in the apparent grid load, due to the PV covering part of the real load, not visible to the grid operator. This makes the rising ramp-rate of load that needs to be covered by other generation in the late afternoon leading up to the evening peak quite a bit steeper, something that has to be managed by the grid operator. This has been dubbed the "duck curve" in California, "Nessie curve" in Hawaii, but it's known artifact of a high PV grid (distributed or utility scale), and why a 100% PV grid isn't likely to be the most cost effective solution. There are many ways to tame the duck, and electric vehicles are likely to play a major part of that.

    Changing the rate structures with dynamic rates or time of use rates or demand charges correlated to use during PEAK hours can also incentivize people to do something other than come home at 6PM, plug the car in to the car charger, turn on the 5 ton central air conditioner and jump into shower heated with a tankless electric while dinner is simmering on the electric stove. Pre-cooling the house in mid-day when the PV is putting out uses more energy, but it lowers the peak grid load. Heating water to a higher setpoint in an electric tank with an element under the grid operator's control increases standby losses at the home, but lowers the amount of spinning reserves necessary for managing grid voltage & frequency, and lowers the peak grid load. In the PJM region distributed aggregators (eg Mosaic Energy: )are paying people with electric water heaters to operate them as a "virtual powerplant", bidding into the ancillary services, demand-response & capacity markets. Since the Supremes blessed FERC Order 745 that sort of thing will become more common, as demand response is now required to be remunerated at the same rates as generators in the wholesale spot markets and capacity markets (in places where those markets already exist.) In my region the ISO-New England grid operator's demand response market will be going live this coming June. It'll take a couple of years, but I expect aggregated demand response to become a significant factor in reducing grid peaks.

  34. itserich | | #34

    glad we agree
    Thanks Dana, glad we agree!

  35. Expert Member
    Dana Dorsett | | #35

    Efficiency vs. cost efficiency explained
    A typical utility panel has about 15% conversion efficiency from sunlight to DC power at the bus bar. They're big, and cheap in terms of $/DC-watt. With some additional money spent on tracking mounts they can squeak more power out per DC-watt than with a fixed tilt. But the raw efficiency is still 15%.

    Residential panels (in my area at least) are smaller, but run about 20% efficiency, photon-to-busbar-DC, but usually mounted fixed tilt. But they are also more expensive than utility type panels in $/DC-watt.

    But cost-efficiency ($/watt-DC or $/kwh-AC) isn't the same thing as electrical efficiency or conversion efficiency- they're not even related. One is physics, the other is economics.

    Taking the total delivered-cost of photon to load is an important factor that doesn't appear in an LCOE ( levelized cost of ENERGY ) analysis, since it's not the delivered cost measured at the load, only at the generator's connection to the grid. PV can be more expensive in $/DC-watt terms for a rooftop residential system, but still end up lower in cost in delivered kwh, as is the case currently in Hawaii and Australia, where both the insolation levels and grid costs are high. As PV becomes ever cheaper, this will become true pretty much everywhere by 2030 or so.

    PV has gotten cheap very quickly, though it might not seem that way. But the unsubsidized $/watt currently paid for residential rooftop ($2.89/watt, the most recent average: ) is less than what utility scale PV was costing 10 years ago. Currently utility scale PV is averaging less than a buck a watt, all-in, installed price, and by 2030 that will probably be true for residential scale solar too. At that point the delivery costs will dominate the total cost, and even if utility scale PV were $0/kwh their total cost of delivery would exceed the cost of self-produced power on the rooftop.

  36. itserich | | #36

    not that dumb
    Thanks so much Dana. I am not that smart, but not that dumb either.

    "But cost-efficiency ($/watt-DC or $/kwh-AC) isn't the same thing as electrical efficiency or conversion efficiency- they're not even related. One is physics, the other is economics."

    When I commented on economic versus energy efficiency, I was acknowledging the difference. I also said I don't understand why you make that comment, because utility built solar seems to be more efficient in both respects.

    You do seem to keep moving the goal post.

  37. Jon_R | | #37

    > then peaks in the early
    > then peaks in the early evening

    And the utility has to build the grid to handle this load without any help from solar. Until we see significant use of batteries and different demand curves, don't expect much from the claimed distribution/transmission savings (highly distributed residential solar vs smartly distributed utility scale solar).

  38. _Stephen_ | | #38

    In Reponse to Jon R
    Interesting... My locality has time-of-use rates. As a result, our summertime peak is in fact during the day, which my solar array trims quite nicely.

    Everyone schedules their EVs to charge after 7, and they also do their laundry on the off-peak as well.

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |