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

Community and Q&A

Sizing solar system for new construction

Kevin Chan | Posted in General Questions on

Hi guys, I need some help on estimating the electricity usage for my new construction in order to size for PV system. The house will be heated by 3 Fujisu ducted heat pump. Outdoor unit AOU12RLFC, indoor unit ARU12RGLX. Below are manual J load cal
Zone 1, 1fl, 1150 sf, heating btuh: 11204, cooling btuh: 16475
Zone 2, 2fl, 1150 sf, heating btuh: 9580, cooling btuh: 9224
Zone 3, 2flmaster over garage, 704 sf, heating btuh: 7886, cooling btuh: 5145
The house is located in Westwood, MA, zone 5

is there a formula to estimate the electricity consumption based on the heating degree days per year? All other electricity consumption I could establish from my current house.
Thanks in advance.

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. Roger Berry | | #1


    I'll kick off the discussion and hope others with real life experience in minisplit efficiency chime in. I live in a 2700 sf home with basement (that I allow to be cooler) in CZ6B. Massachusetts has changed a lot since I left 50 years ago, but my perception is that now winters are much milder and more mixed snow/rain. Summers are pretty humid if I remember correctly.

    The energy consumption for heating and cooling will be tricky to predict given the pretty crazy weather patterns I have heard about there. (relatives) Heating degree days will only get you part of the way to an answer, since the COP for heat will be as variable as the weather. I think most of the commenters on GBA would propose a COP of 3 being a reasonable average over the heating season. You would still need to expect heavier electric demand on the truly cold parts of winter, much as you will on the hot humid days of summer. However, on an averaged demand basis, you should be able to guesstimate your BTU losses for a value a bit above the design temp and use that as a start point. I hope Dana shows up to do the math, it will be much more trustworthy than mine. And he lives just a ways west of you and know the AC load well.

    As for non heating usage, I can say with certainty that, try as I might, we seem to use 1kwh per hour for non-heating purposes. Or I should say, not heating the house. We do heat our water, food and clothing with all electric.

    The water heater is an energy pig, but one that suits our home's design parameters. We have electric resistance heat (long story), so stealing heat with a heat pump water heater would be a net loss for us no matter what the COP of the unit. Even placing a hybrid water heater in the garage to take advantage of the summer warmth would not likely create a positive balance. I have not figured out how to meter just the water heater usage, since my spouse won't let me flip the breaker long enough to get a good reading on other demand. I would guess at least a third is going into the water heater.

    We are 95% LED lighting and only two people in recent years. With one fridge, freezer, oven, dryer, washer, induction cooktop and dishwasher eating another third or so, the remainder seems to be the TV, two computers and an ungodly number of wall warts. We do not have any AC so the usage you might see there during summer months is likely to exceed the 1kwh per hour I have.

    So, 1kwh per hour is 8760 per year which is very close to the best extrapolation I can make from my summer bills. I would split our annual usage at 9000kwh for all lighting and living needs and 10,000 for heating. That sounds ungodly high and it is compared to using minisplits, but my whole heating system cost about $5,000 added to the electrical contract. I also don't have to argue with a certain occupant about fan noise. As noted, long story.

    If you are trying to size your solar panels appropriately, you will need to take into account the metering options you do or don't have as well as future expansion needs; ie. electric cars. The obvious other factor is your local sun values and average annual output. Even out here where we are told we have more sunshine than most, the annualized output for an array is about 1.5 face wattage for a fixed ground mount. A offgrid neighbor with a tracking array does better on a per panel output basis, but the post will only take so many panels.

    Going solar for me would necessitate an array of about 14Kw worth of panels as I stand. To add electric vehicles in the future would mean a bigger system still. Fortunately, for me, the local power company is very pro solar and will allow excess kwh to accumulate during the summer and offset my winter demand. My age and the payback frame doesn't make a lot of sense for me to go solar.

    Be very sure of what the power company is offering now and what guarantees there are for the future. Nevada tried to short circuit solar a few years back with very disadvantageous rulings. I think they got rescinded, but in any case check local rulings closely and don't rely on sales people.

    If you are blending in gas appliances or water heaters your calculations will become even trickier. At least the design numbers you posted sound very much in the PGH range. Hope this info helps.

  2. Expert Member
    DCContrarian | | #2

    I would start with your electricity company. Around here about 90% of the benefit from solar comes from various subsidies and incentives, and the local utility limits the size of the system they will accept to a certain percentage of actual usage. If there is no usage history they will calculate it based upon the features of the house.

    Here at least, it is very much in your interest to get the largest system the utility will accept, the larger the system the larger the subsidies.

  3. Roger Berry | | #3


    Realized that my description of output potential might be confusing. A 6.5 Kw panel array can be expected to produce about 9,750 Kwh of power over the course of a year. A common figure of 4 hours of full sun per day per year is a rule of thumb that seems to work here. (4x365=1460 hours) So it is more helpful to say "multiple the face wattage by 1.5 then by 1000 to get yearly kwh production". In the famous words of fine print, your mileage may vary.

    Peak production obviously occurs in the summer when the sun shines longest. However, local cloud cover percentages, siting constraints, angle of presentation, etc all influence yearly output. Winter output is always lower and local cloud/snow conditions compound the lower output. Here, it is very sunny much of the winter, so while snow on low angle roof panels can be a problem, ground rack mounted ones tend to shed fairly quickly. Thanks to our intense sun, the aluminum frames and dark glass warm up enough to slough off the snow by gravity. Most of the time. The occasional 20" drop is best moved off.

    How the panels are connected and the power processing managed is best discussed by professionals. Having an inverter on each panel helps with shading problems, but more components mean more fail points. There is a rich history of pro and con discussion on the internet if you really want to investigate.

    One good news point, I believe that the 26% rebate is conserved for two more years. That is if the good people in Washington have actually signed off on it.

    1. Kevin Chan | | #5

      Hi Roger
      I have a solar company do a design for me based on my plan. My roof could fit a 20kw system. Annual production is about 19000 kwh due to shading and orientation of the roof, which should takes care 100% of my household electricity consumption minus my EV. They confirm the 26% rebate is extended for 2 more years, also MA also have a $1000 rebate plus the SMART credit from the utility.

  4. Kevin Chan | | #4

    Hi guys, thanks for the replies. After some google search, I use the following fomular to calculate the rought kw usage:
    Delta=heat load at 70F - Design temp at 8F=62F
    Heatload 34000 btuh / 62F=548 btu/deg F x24=13152 btu per day per degree
    heating HDD for my zip is 6109
    13152x6109=80,345,568 btu which equal 23547 kw
    Assume the heatpump COP of 3, my estimate annual electric consumption for heating roughly = 7849 kw.
    Does my math sound right?
    For cooling:
    Cooling BTUH 30000, Total CDD annual is 539, Heatpump SEER 21
    Design temp 88 outdoor, 75 indoor
    30000/13 x24x539=29,852,307 btu
    29,852,307/21 seer/1000=1422 kw.

    1. Charlie Sullivan | | #6

      Your math is right, but I think you are missing latent cooling—dehumidification. That might actually double the cooling requirement.

      But also, manual J is often done deliberately conservatively to avoid a customer being unhappily cold. Without knowing how that was done, we don't know how conservative this is.

      You will need to decide whether to aim being sure your cover 100%, being sure you use all you generate, or somewhere in the middle. A modular design with microiverters that can be expanded in the future is one option.

      1. Kevin Chan | | #7

        The manual J was done by my HERS rater, on the report, the total sensible cooling is 26623 and latent cooling is 3199. The manual J is calculated based on 1.5 ACH.

        1. Charlie Sullivan | | #8

          Oh, good. that includes the latent then. And I would trust a HERS rater manual J over an HVAC contractor manual J. Still, it's not an exact prediction and you will likely have some error.

  5. Jon R | | #9

    For being "green", I'd look into the availability of programs that encourage the utility to add more renewable power and then skip residential solar entirely. The former has the potential to be about 5x more efficient, ie, do much more for the environment for the same cost. These residential solar subsidies are a waste of taxpayer dollars.

    1. Joe Norm | | #10

      I have always wondered about the effectiveness of rooftop solar in the effort to reduce demand on fossil fuels. Utility scale has always seemed better, but I have no metrics.

  6. Roger Berry | | #11


    I think your math must be off. I am showing ~38,000,000 btu consumption for a heating season of 7200HDD. Admittedly, my insulation levels and the sunshine here may be working in my favor, but I don't think by 40,000,000 btu. My design temp is -10F and I am keeping 2700 sf of living and 1400 of basement comfortable.

    I am first to admit my math skills are the weakest I have, so please have others check my next thoughts. I believe you took your largest delta t and load to find what one degree would look like. I suspect that the losses are a sliding factor that would require something like calculus to properly express. Try finding your heat loss for 1 degree with your projected insulation/window package values and see if it agrees with the number you derived. If it is sharply different, then maybe using the average of the two will be more accurate. Again, I may well be off here, so my thought is based solely on my known usage.

    For the record, 23,000 plus kilowatts is what I used the first year, which included 2 months of running multiple heaters with the windows cracked to let all the painting moisture out. And then including 10 months of occupancy with all the hot water, cooking and clothes drying going on.

    For ultimate energy demand, your additional needs for air conditioning may be a fly in the ointment, but for heating I think you will do better than the 80,000,000 projected.

  7. Expert Member
    Zephyr7 | | #12

    It's worth pointing out that kw and kwh are NOT the same things.
    1000 BTUH (btu PER HOUR) = 3.412 kw
    1000 BTU = 3.412 kwh (kilowatt HOURS)

    The first is a rate, the second is a unit of energy. The equivalent terms in water would be gallons PER MINUTE for the first, and GALLONS (as an absolute, like how many full gallon-size buckets you have sitting next to you) for the second. Essentially the first doesn't have a time component, so it goes on indefinitely, forever. The second does have a time component, so it is a quantified unit of energy. It can be confusing since the "hour" component is different between the BTU and the KWH, but that's the way the units are defined.

    I see this mixed up all the time, especially in green energy publications that sometimes want to make things look better (or worse). It can also result in BIG math errors if you leave out important parts of the units. If you find you're off by 24 (hours in a day), or 1,000 (watts in a kilowatt), it's possible you're mixing up these units somewhere.

    It reminds me of my college chemistry lab where the professor wrote a number up on the board when we got into class and then glared at us. We all thought we did really bad on the last lab or something. Then he whacked the board with his hand and said loudly (in his thick German accent) "WHAT IS? IS COWS? IS HOUSES? UNITS! UNITS! UNITS!" He did have a good point :-)


  8. Roger Berry | | #13

    Kevin and Zephyr,

    I guess the professor would have a "cow" with my sloppiness. Yes, I do realize and should properly note that kwh are a measure of electric current applied over time. That said, I have taken it as meaning 1ooo watts applied to a resistance heater for one hour will cause 3,412 BTUs of energy to be created-given off-discharged or however the proper terminology is stated over that same hour. If that is not right then I am indeed in big doo doo.

    Am I wrong in using the kwh used to derive the btu output? I use approx. 9500 kwh of energy to cook, wash, heat water etc., and based on four heating seasons, 9,000 to 11,100 kwh to heat. I took the kwh to btu output value of 3,412 btu per kwh used and multiplied the high value to get the approximately 38,000,000 btus.

    Are the underlying math assumptions in Kevins projections off as well?

    1. Expert Member
      Zephyr7 | | #14

      Your heat pump will have a COP number that will effect your calculations, and that number will vary based on operating conditions. 1kwh = 3,412 BTU is an absolute, and is approximately what you'd get with resistance heating. Heat pumps will give you some amount more BTU per unit input energy, so you're calculating a very worst-case scenario. That's not entirely a bad thing, since it's nice to have excess power available -- but you'll be oversizing your system. Oversizing an off-grid system gives you some extra margin for extended cloudy periods, but with a net metering system, it's probably wasted money.

      You need to know your kwh needs by season too, since the winter is going to produce less solar output but you might actually use more electricity during that time. You'll end up with an educated guess as to required system capacity, but I wouldn't count on it to be super accurate.


Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |