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

Solar and Energy Efficiency Need to Work Together

Rather than making an either/or choice, consumers should put these two things together — like peanut butter and jelly

Solar and energy efficiency complement each other as successfully as peanut butter and jelly, says the author, and should not be viewed as an either/or choice.
Image Credit: Matias Garabedian / CC BY-SA 2.0 / Flickr

Editor’s note: This post originally appeared on the ACEEE Blog.

Energy efficiency and solar advocates have on occasion butted heads over which option should be implemented in homes and buildings first and how much should be installed before the other is considered. Here at the American Council for an Energy-Efficient Economy we believe that, like market solutions vs. energy efficiency programs, this is a false choice. Both are valuable and can, and should, work together as an integrated solution to create cleaner and cheaper energy.

While energy efficiency is just as clean as solar when it comes to emissions, efficiency by itself can’t produce energy for customers looking for a clean energy option, and solar without energy efficiency can’t reach the full extent of its potential.

However, in recent years, some solar companies and some consumers have been employing a solar-first strategy in the residential sector — installing solar systems without paying much attention to energy efficiency. This strategy has been spurred in part by substantial solar tax credits, net-metering rules in place in most states, and the availability of solar financing that reduces or even eliminates the initial purchase price, replacing the up-front cost with monthly payments that extend over many years.

Despite these incentives, it still generally makes sense to implement as much efficiency as possible when installing generation. To look more closely at this issue, we conducted two illustrative analyses. The first compares the cost per kilowatt hour produced or saved from solar and energy efficiency when done individually or together. The second compares solar technical potential and residential electricity use, with and without efficiency. We find that when efficiency and solar are implemented in tandem, costs are lower, and solar can meet a larger share of residential loads.

Cost per kWh

For this comparison, we looked at the average cost per kWh produced from a typical solar system today, the average cost per kWh from residential energy efficiency, and the cost per kWh when efficiency and solar are done together. Our results are summarized in the table below. A solar system costs about 17 to 23 cents per kWh produced (the low-end estimate is based on very sunny Las Vegas, the high-end on Washington, DC). Energy efficiency costs less — about 8 cents per kWh. But when solar and efficiency are combined, the cost is 3 to 6 cents less per kWh than solar alone. Energy efficiency has a lower cost, and it also reduces the size and cost of the needed solar system. PB&J (solar and efficiency) is less expensive than PB (solar) alone.

This analysis ignores the federal 30% solar tax credit and also ignores utility incentives that are commonly available for energy efficiency measures. If tax credits and incentives are included, the overall result is still generally the same — a combined approach is less expensive per kWh than solar alone. This is just a simple analysis for typical measures and hence is only useful as a rough approximation.

Solar production relative to residential electricity use

For this analysis, we compared estimates of the technical potential for rooftop solar systems in each state (as estimated in a GIS-based analysis by the National Renewable Energy Laboratory) with residential electricity use (from the most recent EIA Residential Energy Consumption Survey or RECS). We looked for states where the solar technical potential in the residential sector was at least 50% of current residential consumption, or of residential consumption if energy efficiency were to reduce consumption by an average of 30%.

Our analysis only covers 24 states, as those are the states with detailed data in RECS at the single- or two-state level. Results of our analysis are shown in the map below. With efficiency, 23 out of the 24 states could hit the 50% solar threshold, including six reaching 75% solar (California, Colorado, Kansas, Nebraska, New Mexico, and Nevada). Without energy efficiency, only nine of the 24 states could meet at least half of the residential load with rooftop solar. Only in two states (California and Colorado) does solar potential exceed 75% of residential consumption. In other words, solar can meet a much larger proportion of residential loads if efficiency is included.

Source:ACEEE analysis based on EIA residential energy consumption survey and NREL report on technical potential for rooftop solar systems.

This analysis doesn’t include potential growth in electric loads such as from increased use of electric vehicles, or conversion of gas and oil space- and water-heating systems to heat pumps. Details of our analysis, including a case where all gas and oil space-heating systems are converted to heat pumps can be found here. In this alternative case, only two states meet the 50% threshold without efficiency, while 12 states meet the threshold with efficiency.

As with our first analysis, this is a rough analysis that assumes all of the solar potential is achieved and that all homes implement energy efficiency. Also, this simple analysis ignores the fact that some homes can produce more solar power than they use while other homes are not suitable for solar, such as those heavily shaded by trees or that do not face south. This analysis should be considered a yardstick and not a definitive analysis.


Energy efficiency will generally be less expensive per kWh than solar. And by lowering consumption, energy efficiency will stretch the available rooftop solar resource farther, allowing solar to serve a higher percent of residential consumption while also allowing a smaller and less expensive solar system. These are two simple analyses but they make a clear case that jelly (efficiency) is needed to help peanut butter (solar) do its best.

Steven Nadel is the executive director of the American Council for an Energy-Efficient Economy.


  1. GBA Editor
    Martin Holladay | | #1

    Comparing PV and efficiency measures
    I feel compelled to point out that the statement that "Energy efficiency will generally be less expensive per kWh than solar" is just about meaningless.

    Here's how I would phrase it: If you know how much a PV system will cost, and you know the terms of your local net-metering agreement, and you know your cost for grid-connected electricity, and you know what incentives and tax credits are available, it's very easy to determine whether a PV system is a good investment. In many locations, this type of investment is cash-flow-positive from day one.

    When it comes to efficiency measures, some measures are a better investment than PV, and others are a worse investment. It's impossible to generalize about efficiency measures; you have to do the math. Here's a caveat: It's worth noting that many software programs overpredict the savings that will result from efficiency measures. PV savings are much easier to predict.

  2. JC72 | | #2

    I wonder if higher effeciency would help address the "duck

  3. GBA Editor
    Martin Holladay | | #3

    Response to Chris M
    Many energy experts disagree with the position taken by the author of the article you linked to (Gail Tverberg).

    Among those who disagree is Christopher Clack, author of this GBA guest blog: The Cheapest Way to Scale Up Renewable Energy?

  4. Dana1 | | #4

    The $/Mwh of efficiency is measured regularly.(@ Martin)
    While the $/Mwh cost of individual measures will vary widely, in the aggregate it has historically been well below the lifecycle $/Mwh of PV, as measured by everyone from governments, academics and investment bankers.

    Of COURSE you have to do the math on any individual upgrade, and yes, some fall well above the lifecycle $/Mwh cost of PV (even rooftop PV). My favorite example is the last 4" of sub-slab EPS in a PassiveHouse, which has next to zero financial benefit, and subsequently a substantial cost in $/Mwh terms.

    Tax rebates and other incentives will skew those numbers by quite a bit, but the range of (unsubsidized) levelized cost of efficiency measures is tracked by many, and is well understood. One example of levelized cost analysis can be found in brief in (investment bank) Lazard's periodic updates of their original levelized cost numbers, updated to reflect new cost/benefit data as these rapidly evolving markets change. The November 2015 update was the latest revision, and it can be found online here:

  5. Dana1 | | #5

    Just Ducky! (@ Chris M )
    The extent to which energy efficiency can improve the duck curve falls mainly in the form of demand response programs, where the utility or grid operator is allowed to curtail load during peak hours, and dump excess power into non time-critical loads for grid balancing. By manipulating the demand side the belly of the duck rises, and the neck & head drop, which reduces the amount of spinning reserves and low efficiency peaking power are required..

    In markets where aggregators can bid distributed load resources into the demand response, frequency control, and capacity markets, a popular tiny load for aggregation is a controller on an electric hot water heater's heater element controlled by the aggregator. In the PJM region (mid-Atlantic through the eastern midwest) aggregators will pay homeowners $100/year for the ability to use their hot water heater for those services. In one West Virginia community a church paid for their rooftop PV with donated fees from the demand response aggregator (Mosaic Power) to members of the congregation. organized by a group called Solar Holler, who also plucks the cost effective efficiency fruit first on most of their projects:

  6. GBA Editor
    Martin Holladay | | #6

    Response to Dana Dorsett (Comment #4)
    We're in agreement. As you know, I've long been in favor of programs (like the Weatherization Assistance Program) that promote the implementation of cost-effective energy-efficiency measures. (See, for example, my article called Is Weatherization Cost-Effective?)

    I'm simply pointing out that there are two types of energy-efficiency measures: those that are more cost-effective than PV, and those that are less cost-effective than PV. Before anyone makes generalizations about the energy-efficiency measures, it's important to explain this difference.

    If we make a list of energy-efficiency measures ranked in order of cost-effectiveness, it's possible to draw a horizontal line at a point where all of the measures above the horizontal line are more cost-effective than PV, and all of the measures below the horizontal line are less cost-effective than PV. Note that as PV gets cheaper, this horizontal line is moving steadily upward, reducing the number of energy-efficiency measures that are more cost-effective than PV.

  7. vensonata | | #7

    New calculation on PV
    I realize the article is a general proposition about PV vs/with efficiency, and as Martin says, it depends on the specific numbers, not on generalizations. However let us see how specifics do change the scenario quite drastically. The estimates for Las Vegas are PV at 17 cents KWh. But... that is based on the full price of $3.70 watt and without the ITC of 30%. Those prices are out of date, new NREL national average is $2.90 watt. So lets put that in to the math... 22% reduction gives us 13.2 cents KWh without subsidies. Next, the original price is based on a 20 year lifespan. We all know lifespan be at least 30 years. So let us reduce a further 33% to 8.9 cents KWh. This shows a huge difference when using present prices and appropriate lifespans! We haven't even included the ITC which would bring the price to 6.2 cents KWh. It is very difficult to compete with that price through efficiency upgrades...perhaps only swapping out incandescent bulbs for LED would be worth it, and perhaps an increase in attic insulation if there were not complications.

  8. Expert Member
    Dana Dorsett | | #8

    Comparison with PV isn't totally apt from a policy point of view
    PV qualifies for the 30% tax subsidy at any price point, which at the time the subsidy was enacted was nearly 3x as expensive (on average) when it was first enacted (The Energy Policy Act of 2005) than it costs now, and probably on the order of 5x as expensive as it will be by the time all Federal tax credits begin stepping down in 2020.

    Comparing a ranked order of efficiency improvements against the cost of the moving-target cost of PV doesn't reflect the full value of those efficiency improvements. Even when efficiency improvements result in a lifecycle cost of energy saved that is 3-5x the lifecycle cost of PV there is often substantial other value to the efficiency improvements worth pursuing than raw energy cost savings. It's not really an "apples to apples" sort of comparison to make, though it's still worth at least noticing when that threshold is crossed. It's important to have a conscious and rigorous understanding of the rationale behind each upgrade.

    As more PV solar gets built the commercial value of that power goes down, and at some point may become lower than the cost of installation. It also cuts into the value of other power generators too- the wind power industry understands this better than most, since the falling cost of wind against fossil burners allowed them to eat fossil-burners' lunch, and already PV is beginning to eat theirs. Almost all the fossil fueled generator retirement in the ERCOT region (Texas) between 2020 & 2030 is expected to be replaced by solar, not wind, even though wind growth from 2000 to present has gutted the economics of coal burners, and it's lifecycle cost is still falling by double-digit percentage learning curve. These energy-price deflationary sources are at the same time reducing the energy savings value of efficiency, even though well targeted energy efficiency is still cheaper than both.

    Ven: What are the discount rates you're using for the extended life LCOE values?

  9. GBA Editor
    Martin Holladay | | #9

    Response to Dana Dorsett
    In you last sentence, your wrote that "well targeted energy efficiency is still cheaper than both [wind and PV]."

    If that's your definition of "well targeted energy efficiency" -- namely, measures that are cheaper (more cost-effective) than both wind and PV -- of course we are in agreement. Everyone is.

    If an energy-efficiency measure is that cheap, it's a win-win situation for homeowners and utilities.

    Needless to say, a lot of energy efficiency investments aren't that well targeted.

  10. dknowlt | | #10

    woa woa woa
    Assume anyone now deciding to own pv will be there 20 yrs or longer.
    Selling a house with a decent working 6kw system should get something back - TBD
    Many strive for small footprints while keeping a manageable budget
    Making a thermal fortress while still relying solely on commercial energy is..just self serving and a 1 at a time endeavor
    A well attentioned envelope of code+ tightness - pretty alright house, with only modest use efficiency (assuming you're American) with pv and a metering agreement, is a good PB&J sandwich appealing to 1000 times more consumers than building a supertight cave and having to think too hard. Good greenfield development should adopt this sandwich for the masses who will love it, x'ers. millenials, definitely younger boomers like me. Selling like hot dogs or better. Wouldn't that be great? A grand dream. Isn't that we all want? I like PBJ on toasted bread thank you. Great article Sreven

Log in or create an account to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |