GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X 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

Planning for Solar Panels and Battery Backup

pico_project | Posted in General Questions on

Still in the middle of a renovation. We’re about to remove the mast / overhead power from the house and bury a new line to a panel in the garage.

I have already planned a 220v outlet for an electric car charger, but have recently started looking into battery backups and solar panels.

I believe a service disconnect switch is required for any electrical backup and as far as I know, we currently don’t have one at the meter (or anywhere).

What kind of things should I have the electrician put in now to make adding something like a Tesla Powerwall or solar panels much easier down the road?


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. Expert Member
    BILL WICHERS | | #1

    I recommend you run your underground service inside PVC conduit. Conduit offers a lot of advantages and isn't usually much more cost to install. You may also want to run a few extra conduits for future telecommunications services to avoid the need for additional trenching in the future. I would recommend 1" PVC or HDPE conduit for this purpose. You'll need at least 2" for the power, maybe more -- the service planner at your power co can tell you what their requirements are.

    If you want a generator, you need a transfer switch. I like automatic transfer switches (ATSes) that can run the entire house, but they aren't cheap and they're also long lead time items right now. This may be something you want to consider.

    Power walls and solar panels are typically hardwired, so they need to wire directly into the panel. The best thing to do is don't recess the panel into the wall, so that the knockouts around the perimeter of the tub are accessible. If you want the panel recessed, I would run a conduit (ideally 1.25" EMT) from the panel to a good size (8" square) junction box in the area where you think you might install the future power wall or solar inverter. This junction box will be easier to access than trying to dig into the wall around a functioning electrical panel.

    Note that I also recommend placing the electric panel and ATS indoors, in the house, if at all possible. Panels exposed to the elements don't hold up as well over time, and a garage is much better than outdoors, but will still be exposed to more extreme temperature swings. Putting things in the house helps to avoid the temperature swings and the condensation that often results.

    Make sure you have a single point ground (which is code now, BTW), which basically means you need to bring all of your services (power, telephone, cable TV, anything with a conductive wire) in at the same place on your house so that all their protectors can ground to the same place. This is very important for lightning protection and also for your safety.


    1. pico_project | | #4

      I believe the panel is already recessed into the wall in the garage. The garage will have a SA WRB (Mavjest 500 SA) and Rockwool in the walls (R15) so it will be somewhat insulated. I'm sure it will still get cold in the winter though. We're in northern Michigan, Traverse City to be specific.

      The walls will be plywood with screws exposed if that helps at all. I think the contractor wanted to glue them though too. Maybe I should avoid that at least in that area so the plywood panels can be easily removed?

      Without an automatic transfer switch... in the event of a power outage, I would have to go manually flip the switch to get it all going, correct?

      What about having a sub-panel of 'critical' circuits, versus keeping it all in one panel and shutting off manually what I don't want to run?


      1. brad_rh | | #8

        Yes to subpanel for what you want on the backup. If you're not adding a panel or doing a lot of rewiring of the panel it may not save much to do it now vs later tho.

      2. Expert Member
        BILL WICHERS | | #10

        Without an ATS, you have to manually start the generator and transfer loads over. You can either use a manual transfer switch (basically a large switch that mounts on the wall), or an interlock kit, which uses a second breaker in the main panel as a "generator INPUT", and the interlock ensures the two sources can't both be on at the same time.

        A subpanel works too, and lets you use an ATS smaller than your main service. This can save some money, but I prefer an ATS that can run the entire house. You can do it either way. I would recommend against the small "generator panels" that have a bunch of seperate switches to manually transfer individual loads. I don't find those to be particularly well made. You'd be better off with a regular subpanel of the same brand as your main service panel and a small ATS if you want to go the subpanel rout. Small 16 to 24 space panels are good for this application, and they can be "main lug" style which is cheaper than the kind that include a main breaker.

        Conduit serves to provide additional protection to the cable, and allows for easier service work if that's required in the future. I try to always run underground cables in conduit for these reasons.

        BTW, another posted recommended "100 amp" cable. You don't want to do that. Run cable good for 200 amp service to ensure you're pretty future proof. 200A service is pretty standard these days, especially if you plan to go with a lot of electric appliances and an EV charger. 200A service is usually run with 4/0 aluminum cable.


  2. Expert Member
    NICK KEENAN | | #2

    In terms of future-proofing, the only thing to run until you know what you're doing is the wire or even just conduit. Plans have a way of changing.

  3. jameshowison | | #3

    Not sure if this is in scope, but consider moving plumbing or other penetrations onto North facing roof segments, clearing the area in which you'd like to put panels.

    If you are replacing roof at all, consider using standing seam metal where you will put panels. You can clip directly to the standing seam without penetrations and rails, which is better looking and less expensive (although the combination of metal and clips might be more expensive, but you'll have a metal roof in the meantime).


    If you'd like to get deep into what's available in the DIY or small installer market (but all UL listed) check out WIll Prowse's channel:

    Also have a think about whether used solar panels make sense for your project. They are already de-rated, but there is a trade off between space (need more for same output), panel cost (much lower), and labor cost (same per panel, so higher overall). e.g.,

    1. pico_project | | #7

      Everything is off the south side except our kitchen vent. I wanted a straight shot on that. However, we just put on a new shingled roof a few weeks ago.

    2. aunsafe2015 | | #15

      I know I'm responding to a months-old post, but James, have you seen houses that have standing seem metal ONLY where the solar panels are located? And shingles elsewhere? I've always thought it was more of an all-or-nothing proposition--- entire roof metal or entire roof shingles?

  4. matt151617 | | #5

    Run 100 amp cable, regardless of what your power requirements are now, and make sure it's run in conduit and not direct-buried. Also, as someone else mentioned, have a second run of empty conduit put in for future cables.

    Now is also a great time to think about having gas (heater), water, or sewer put in while the yard is dug up and the trencher is on-site. Personally I would have gas pipe and a water pipe put in now even if I didn't plan on using them right away.

    You can run gas/water in the same trench as electrical but there's minimum spacing requirements, so check your local codes.

    1. pico_project | | #6

      What's the benefit of conduit other than the obvious added protection? Ability to re-run the line without digging up the yard again?

  5. joenorm | | #9

    It's pretty dependant on your location and what is required by your authorities.

    Get a conduit from the roof to the utility area where the batteries may be and panel is. Depending on what system you go with, that conduit may need to be in EMT. 3/4" will be enough for a pretty large system.

    At the main panel you should create a sub panel with your backed up loads. I like mounting a good sized gutter underneath both panels where the feed runs through. This way, when it comes time to intercept it it's easy to get at. Pipes from the battery and solar can run surface to the gutter. There are a lot of ways to do it and a lot of variables so hard to plan everything without knowing what you want. But that would be a good start.

  6. stuartd | | #11

    Back up battery Without Solar Panels
    Not sure if my reply is specific enough to the original post, but I'm not a subscriber, so cannot reply to:

    I would like to backup refrigerator, well pump, grinder, septic pump, and some circuits, but DO NOT want gasoline, or propane back up generator

    New reasonably tight construction
    wood stove
    Climate zone 6
    all electric
    3.000 SF vacation house
    surrounded by trees
    NO solar panels in the near future

    Battery electric backup (Kohler) quote is about $16k, and apparently not enough juice to run the Fujitsu ducted heat pump system (that's ok have wood stove). Tesla wall is $22K

    One can buy a 3000 w Goal zero battery with wheels, about $3500, plus accessories, which does have an interface with the panel, but it is not automatic switching. I would like to have automatic, since I will not be there most of the time, but will have food that will spoil. There are power outages, range from a few hours to a day.



    1. Expert Member
      BILL WICHERS | | #13

      >"3000 w Goal zero battery"

      That doesn't tell you much. Battery systems have TWO ratings:
      1- the RATED OUTPUT CAPABILITY, which would be '3,000 watts', in this case.
      2- the ENERGY STORAGE CAPACITY, which would be rated in 'watt hours' in most cases, sometimes 'ampere hours', which is more traditionally used for batteries, but not for these self contained systems.

      A 3,000 watt hour backup system can provide 3,000 watts for ONE hour. It can provide 1,000 watts for THREE hours. It could provide 125 watts for ONE DAY. For some perspective, the AVERAGE load for a typical home is around 1,000 to 1,500 watts all the time, so that '3,000 watt hour' battery could run you for about 2 to 3 hours -- not much.

      Battery systems aren't very practical for whole-home backup because they just don't last very long, and they are very expensive. Generators are usually much cheaper, and they can run as long as the fuel supply holds out, which can be replenished periodically to deal with longer outages.

      I recommend generator systems over battery systems for cost and usefulness reasons. I do agree with post #12 though, to a point. If you're in an area where no one can remember the last time the power went out, maybe a generator system isn't money well spent, and if you did get one, maybe a cheaper portable unit would be a better fit. If you're in an area with a lot of power outages (I'm in this category, being in a semi rural, heavily wooded, area), then a permanently installed generator system makes a lot of sense.

      You also need to consider the value to you of keeping power running. If you work from home, and no power means you can't work, that makes a generator more valuable. If you expect a typical outage to only be a few hours, and you don't depend on your power for your job, then maybe that backup generator isn't as valuable. Typical freezers can easily ride through a few hours without power without loosing anything, so it's not necassary to keep them powered for short outages, for example.


    2. stuartd | | #27

      Thanks for replies as well.

      I went into my project thinking I wanted solar and battery backup, but physical location, as well as not full time use house, and expense, so probably not going with solar, and battery back up alone super expensive for the few times it would be needed.

      Since my panel is now full, the ducted air handler has to use a 'jumper', which is approved and included with the Fujitsu equipment... something like it takes two circuits and combines into one breaker. Otherwise, ran out of panel space and would need sub panel. I also have a Siemens whole house surge protector, but no room left on main panel.

      Cost to install 100A sub panel is $1500. Doesn't seem worth it, since there may be as yet unknown future electric loads, and panel may have to be moved, rewired whatever. I'm willing to give up whole house surge for now. (Alternatively, I can get a quote from another electrician after house has a COO.
      What do you think?

      1. Expert Member
        BILL WICHERS | | #28

        You can put in a whole house surge suppressor in one of two ways, neither of which uses any extra breaker spaces:
        1- Get a surge protector that is built into a double pole circuit breaker. Siemens makes these, probably other manufacturers too. This way you can still use the two spaces to supply loads.
        2- Use a small surge protector like the Ditek device I usually recommend, then tap it off of two circuit breakers, making sure each breaker is on the opposite "leg" of your service. This is the way these Ditek devices are typically installed, and it doesn't require any extra dedicated spaces in the panel.

        $1,500 to install a 100A subpanel seems pretty high to me, assuming the subpanel would be going next to the main panel. The 100A subpanel alone should only be around $100-200 or so, a 100A breaker to feed it should be well under $100, and even with today's crazy copper prices, I'd expect under $100 or so for wire. That means total on the materials would be around $400 tops, and it shouldn't take more than a few hours to install if you're not also transferring a bunch of circuits around.


        1. Expert Member
          DCcontrarian | | #29

          Bill --
          Off-topic question: my heat pump manufacturer recommends a surge suppressor but doesn't provide info on how to do it. Is a surge suppressor breaker the way to go? I assume that's what they're meant for?


          1. Expert Member
            BILL WICHERS | | #30

            That's the way to go for any hard-wired device. If you have a small plug-in heatpump, then Tripp Lite's small 2 outlet Isobar (ISOBLOK-20) is the easiest way to go. If you have hardwired units, then a permanently installed surge protector (TVSS) on the panel is the way to go. As long as that panel-mounted protector is wired into both busbars, as all should be, it will protect all circuits in the panel -- you don't need seperate devices for each branch circuit.

            My recommendation is always a central TVSS on the panel, plus additional plug-in devices at the ends of any circuits right before any sensitive equipment (TV, computer, etc.). I sometimes add additional hardwired TVSS devices on other things if they have very long circuits feeding them.

            BTW, BE SURE you have all your electrical service grounds tied together at the service entrance. This means power, phone, cableTV, satellite -- everything -- should all be connected together and grounded at ONE point where they enter your home. This is code now, but it's also VERY VERY important for protection against electrical transients like lightning.


  7. nynick | | #12

    I think whole house back-up power is overrated. Unless you live in a high storm, high power outage area, think how many times per year and for how long you may be out of power?
    As someone who has been out of power a half a dozen times for days at a time over the span of a few decades, my 5K gas generator has gotten us through. You don't need to power up the entire house to full capacity to live. You only really need water, heat and refrigeration, and half the time water arrives via public utility.

    These whole house ATS generators are overkill and crazy money. It's not the end of the world to be short on power for a few days. A regular gas generator and a few gas cans will get you through.

    1. pico_project | | #14

      With battery backup you can run off your batteries during the day and then recharge them in the middle of the night when the rates are lower. So it's not just for when the power goes out. It's also about purchasing electricity when demand is low and saving it to use later when demand is high.

      Many systems will do this for you automatically and have an ATS built in. It's only getting better too.

      1. Expert Member
        BILL WICHERS | | #16

        That is known as "load leveling", and is a Good Thing. The only problem is that you'll never save enough money doing this on a residential scale to pay for the battery unit. Even commercially, rates have changed such that it's very difficult to justify for larger projects too. This is unfortunate, because "ice chillers" used to be more common, which made ice (literally) during the night using cheaper off-peak power, then thawed that ice during the day to handle building cooling load so that the chillers didn't have to run during more expensive on-peak times. I haven't been able to financially justify those ice chillers for commercial projects for over a decade now.


  8. Expert Member
    ARMANDO COBO | | #17

    I'm having "issues" with a Tesla distributor and installer here in Dallas for a large house I designed, who after a year-and- a half of planning a Tesla solar tiles, now they are coming up that their batteries cannot be charged with a gas generator, which means that in the event of loosing power for more than two days and no sun, you are screwed!!!
    FYI, Dallas has a new fire code that "requires" no more than 5 powerwalls (or 80kW) and in conditioned space. That means my clients need to buy a $90k-100k generator to supply the house with juice. Its a 15k sf house. Not that my clients can't afford, but the system was bid at $375k, now is $500k + the bigger generator. Oh, they had originally bid 8-9 powerwalls.
    FWIW, there are other solar PV that works with batteries other than Tesla, and Enphase System works with generators. Buying "old" fashion batteries are more economical and if maintained properly, outperforms Tesla... plus you save money in "their" smart panels ($4.5k @), plus gateways ($1.5k @), plus batteries ($8k @) to be replaced every 7-10 years.
    Anyone dealing with Tesla will learn sooner or later that they share NO INFORNMATION AT ALL with the clients, Architect or Builder... and by the time you find out about "stuff", its too late, and huge down payments were made....!
    It maybe great products, but good marketing over-sales and over-promises. Sound familiar??? BE AWARE OF DEALING WITH TESLA!!! 😈😈😈😈

    1. Expert Member
      BILL WICHERS | | #18

      The powerwalls probably don't like the less stable (in terms of line frequency) power coming out of smaller generators. That's a common problem with "smart" power gizmos, including computer UPS systems. All of those systems track grid power to keep in sync, and they can't synchronize to a "wobbly" waveform, which is what you get from a small generator. Regular electrical stuff is fine on smaller generator power though.

      Armando, I suggest putting the powerwalls AHEAD of an ATS for a generator. In this way, the generator would power the house, but NOT the powerwalls. The powerwalls would connect to the utility, and function normally, even providing "backup" power if the utility goes down (if they have that capability). The ATS wouldn't know about any of that, it would just see "utility" power. If the utility failed, and the batteries ran down and stopped providing power, THEN the ATS would see an outage and trigger the generator. The generator wouldn't need to deal with powering the powerwalls this way, and you would have a simpler system. That might be an option for you.

      BTW, $90-100k for a residential genset? I have a commercial project right now with a 150kw diesel generator going in and I sold that genset to the customer for ~$37k or so. Installation obviously adds to that, but even with all the service work on a 1,200A 480V commerical service, the installed cost to the customer is a little under $150k for a much larger system than you'd probably be doing on a residential project. You might want to double check your generator contractor isn't overbidding the job.

      I'd really push people away from conventional batteries too. I deal with thousands of those at work, they have lots of exciting ways to let you down. Much better to just put in a peak shave system with no batteries, and use a generator for backup power. Less maintenance, no hazardous materials, and much simpler installation that way.


  9. Expert Member
    ARMANDO COBO | | #19

    Thanks, Bill. I'll pass your information to "our" people. This is my first, and maybe only, Tesla job I'll participate, at least until they fix their ways.

  10. jberks | | #20


    Tesla does discuss integrating a backup generator. But you're right, the gen can't charge the batteries. However, Just as Bill mentioned, you can integrate the backup gen with an ATS between the Powerwall and the main circuit panel.

    I think there will be an interesting market for this issue of PV battery and gas gen backup in the future.

    As PV's and battery tech gets better, I imagine people will rely on those more and tie less houses to the grid. I'm thinking rural and remote houses.

    So with that, there's still a need for redundancy planning for scenarios like poor sun, or just general higher electrical demand. I hope the polished systems like Tesla will eventually incorporate this, or include some sort of rectifying so a gen can charge the batteries so we don't have to rely on the gen running constantly until the sun comes back.

    I assume the occasional run time of a propane generator would still be greener than fossil fuels burned for grid generation and all the transformers and wires necessary of tieing to a grid if a dwelling is very remote.


    1. Expert Member
      BILL WICHERS | | #21

      >"I assume the occasional run time of a propane generator would still be greener than fossil fuels burned for grid generation and all the transformers and wires necessary of tieing to a grid if a dwelling is very remote."

      That's actually pretty unlikely. Small generators ("small" here means less than multi-megawatt range) typically using piston-type engines, similar to what is in a gasoline-fueled car. Those engines are usually around 16% efficient or so. That means you're only getting about 1/8 of the energy in the fuel out of the generator as electricity. Those small generators are also most efficient when running at full load, which usually isn't the case. If you look up a typical fuel consumption chart, you'll see that most small generators use about half as much fuel when running at 1/4 load, which means you use twice the fuel per watt at 1/4 load as you would at full load. That's pretty crummy efficiency overall.

      Power plants usually use steam turbines. Some of these can get up over 60% efficient. This means you're getting around 2/3 of the energy in the fuel out of the generator as electricity -- which might be up around 8+ times more efficient than that small engine generator running on propane (or natural gas, they are about the same for efficiency)! That's a BIG difference! These turbines are also most efficient when run up near full load, but power grids exist to allow power plants to run up near full load as much as possible for maximum efficiency. Power grids really exist for two reasons: to maximize efficiency, and to increase reliability by allowing for multiple power sources in case of failures or maintenance shutdowns.

      Large transformers are very nearly 100% efficient. 99.9+% is common for very large transmission transformers (the 345+ kilovolt units), but even the smaller distribution transformers are typicall well over 90%.

      Where you do have losses is in the lines. Typical transmission lines are designed for around 94-95% efficiency. Typical system losses are usually estimated to be around 8-15% overall, according to the EIA, which would be a worst case of about 85% overall system efficiency for transmission and distribution lines.

      If we allow 60% for the turbine, and 85% for the power lines, We have a total electrical power system efficiency -- including both transmission/distribution (power lines) and generation (turbine) -- of about 51%. Compare that to the 16% or so efficiency of the small engine generator on it's own, when running best-case full load, and the total efficiency in terms of fuel consumed per unit of electric power delivered is over 3 times better for the "power grid" than it is for a small generator.

      BTW, you have maintenance needs every 100 hours or so on a small generator too, which aren't an issue you have when you buy your power from the grid.

      Solar and batteries would be better than a small generator if your primary concern is energy efficiency, but you're better off with a grid connection than a small generator as described above.


      1. DC_Contrarian_ | | #23

        So total efficiency of about 50%. Thinking of heating, burning gas directly at the source for heat you might get 90+%. But with a heat pump you have a COP of 2 or more, whereas with direct combustion your COP is stuck at one. On the other hand, that COP varies with temperature, and might go below one when it's cold out.

        Sound fair?

        1. Expert Member
          BILL WICHERS | | #25

          Yes, exactly right, and this is why I recommend against electric resistance water heaters if your goal is reducing emissions. In most cases, a natural gas fired water heater actually uses LESS energy, and produces LESS emissions, to heat the same amount of water. Exacty amounts vary a bit depending on the dominant sources of electricity for any particular area, but it's usually close.

          Heat pumps are the big exception, and the COP is the reason. Heat pump water heaters need to be looked at carefully, since they scavenge heat from the room they're in, which may or may not help your overall home efficiency. Heat pumps for space conditioning though are typically a win, even down to pretty low outdoor temperatures now with the newer hyperheats.


      2. jberks | | #24



        Can you also add a factor of usage time to your calcs?

        The scenario being off grid pv with the occasional gen use (at 16% efficiency) vs constant grid tied use (at 50% efficiency)

        Also consider 'greeness' of cost, materials, build, machinery, tree clearing, labour, etc, as most remote locations for grid power, you need to place transformers and run lines and whatever else you electricals people do. I believe factors like distance and location (like on an island) play the major role on this. I have seen prices for running lines 1for minor distances, I can't imagine what it would be for major distances and across water.

        Vs a set PV+battery+gen backup system is relatively the same purchase cost regardless of location.

        This word comparison is obviously situational and varies greatly in real life. But basically what I'm getting at is there's more to it than just mechanical efficiency.

        Btw, I still don't think pv's are quite there yet in terms of efficiency vs cost. But I think they'll become cost viable soon.

        I'm not an expert on this shit so I could be wrong.


        1. Expert Member
          BILL WICHERS | | #26

          Usage time doesn't matter, and this is why:
          A typical small generator will run at light load most of the time, because you have to have a generator sized for your peak load, which usually doesn't happen very often. Running a small generator with a light load is the worst case for efficiency, so your average efficiency in terms of converting fuel to electricity is lower than the full load scenario. This is average efficiency can then be considered to be steady state, which means it continues over the entire time you run your small generator. Even at full load for the entire run cycle, you still don't exceed about 16% or so in terms of efficiency of energy conversion.

          The power grid is there to even out loads, so when you're not running your peak load, maybe your neighbor is, or maybe someone in the next city. Average over millions of people, all those "peak" loads average out, and you end up with a fairly smooth curve (see attached graph for the midwest region's grid). You can also run the generators up closer to full capacity, since an extra 10kw of load is a big deal for a 15kw small generator, an extra 10kw is so small as to be immeasureable to a 100MW generator (0.01%). This means you can run the big power grid generators at much higher average load levels, where they are most efficient, and still have enough excess capacity to handle occasional peak loads that isn't an option with smaller generators.

          All of this means the power grid can maintain it's much higher average efficiency while your small generator cannot. Useage factors don't apply, because we're talking averages on both systems. I think if you include that, along with the maintenance needs, you'd still find the power grid to be better overall than lots of small generators would be. While the grid needs to be built and maintained, small generators do too, so neither really has an advantage there.

          PVs systems can be had for around $1/watt these days, as long as you don't try to go off grid and use battery storage. Payoffs are usually 5+ years, so not unreasonable for smaller systems sized for peak shave (which is what I usually recommend). If you try for net zero, the numbers get more complicated, and depend on some outside factors like the net metering offered in your area, if it's offered at all.

          BTW, costs to install cabling over longer distances is usually less than short distances when worked out as cost per mile. Some of this is because short distances tend to be in urban areas, where the cost of construction tends to be much higher (lots of roads to deal with, etc.). Larger projects tend to go through more rural areas, which are usually cheaper to build in because it's easier (for example) to trench through grassy fields than paved city blocks. For fiber optic construction, which is what I usually work with, common budgetary numbers are around $35,000 per mile in the city (for aerial on existing poles), about $18,000 per mile for projects further out of the city. Underground in the city can be well over $100,000 per mile, but in the country maybe half that or less. These assume multimile projects. Power line construction costs are higher, but still less per mile for large projects than for small ones.


    2. Expert Member
      ARMANDO COBO | | #22

      Thank you, guys. This thread has been very useful and educational.

  11. Lettenyano | | #31

    I went through something similar recently. When I was renovating, I didn't think about solar panels and battery backup either. My advice? Talk to your electrician about adding a service disconnect switch – it's handy for future upgrades. Also, plan a good spot for the Tesla Powerwall if you're considering it, or use these options instead. I hope that helps!

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |