Why does installed residential PV cost doubble in the US?

[electric_cheeze]

I was looking at some costs of residential solar and cost difference of typical US residential PV 5-10 kW seems to be 2x what is costs most other places;
I am be curious what others see for installed cost per DC watt?

SEIA/Wood Mackenzie put the U.S. residential system price at $3.36/Wdc in Q4 2024.
Here in the US it can even be over that ~$4/Wdc
I was looking at other places and see

• Australia’s 2024 IEA-PVPS national survey put typical 5-10 kW residential rooftop systems at A$1.20/W (about US$0.79/W), excluding GST, and a small-scale subsidy reduced average cost by another A$0.40/W (about US$0.26/W).
• Belgium reported €1.25/W (about US$1.35/W), including VAT, for 5-10 kW systems in 2024.
• Austria reported €1,551/kWp (about US$1.68/W) for residential rooftop systems
• Italy reported €1.2-1.6/W (about US$1.30-1.73/W) for residential projects.

Are those numbers accurate?
thanks

[Stray Electron]

   One word. Greed!  Personally, I haven't been able to find a competent solar dealer or installer anywhere in the state of Florida. Several years ago (pre-Covid) both the state and the US Gov were giving hefty tax incentives for people to add solar to their homes and there were many dealers and they were pushing it heavily based purely on the tax benefits so I spent a lot of time talking to dealers and installers and absolutely none of them knew, or cared, what they were what they were doing. And I've gone down that path twice more since then only to find out that most of the dealers out there today don't really sell you a system. Instead they're sales men for some big company that is leasing solar systems to home owners in order for them to get the tax benefit and the forever income.

  FYI, when I priced buying a system YEARS ago, they claimed that it was a 4kw system and the price was over $14,000 without taxes or any of the other added costs.  The last two sales droids wouldn't even give me a total cost, they just kept talking about the monthly payment even though I told that that I was not interested in financing.  After I called the local power company and talked to them about it they explained that it was a hoax and that they were only leasing the system and that's why they wouldn't give me a price.

   I did get a data sheet for the solar panels from one company and it turns out that they're only called 4 kW panels and that they're actually only rated at between 3400 and 3600 watts per panel. 

   FYI, about 2 weeks ago I saw an exposé on the local news that a local solar company somewhere around Orlando IIRC has been collecting money from people but never preforming any work.

[TimFox]

Several sources I found indicate that in Illinois, before applying government incentives, installed domestic solar costs roughly $3.00 per watt (at roughly 10 kW level).

[electric_cheeze]

Its a big cost difference.
Greed or tariffs, or manufacturers passing along the cost to make UL/NEC/CSA equipment or all 3.
It seems over the last 15 years the solar industry world wide has largely now harmonized around many International Electrotechnical Commission (IEC) standards, especially around the solar safety stuff, like disconnects and such, so OEM's now may have one SKU for stuff sold in many markets across Asia, Africa, Middle East, Europe even most of South America ,
however the US and Canada's very different UL, ANSI and CSA standards (possibly Mexico NoM?) and the respective electrical codes ( NEC Article 690 ) largely went another direction from the IEC, so manufacturers are required to set up a production line of products for just the North American market.
I am curious if this is the case other places as well ?
Do inverters in Australia need to meet a local standard or is the generic IEC ok?
Looking at wages paid by solar contractors it seems pretty similar, Europe may even have higher wages for installers,that again points to material costs, either having to make special runs of equipment like Inverters for just the US and Canada.
If it were just greed I think there would be market pressure unless there is a monopoly, or some collusion?  Someone in the US would offer $1.30 a watt, but its just unherd of below 3.
It cant just be a 60hz issue as there are other places that use 60HZ.
I'd be interested to know if anyone works for a manufacturer what your experience is?
Does complying with North American product standards impact the cost ?
I know the US and UL have recently modernized other standards like HVAC to align well with the IEC and that has reduced the cost to HVAC manufacturers.
The problem with solar is the US electrical code article 690 seems to have some very unique and outdated compliance requirements, that may date from a time when the industry was very young. 

[brucehoult]

I did get a data sheet for the solar panels from one company and it turns out that they're only called 4 kW panels and that they're actually only rated at between 3400 and 3600 watts per panel. 

Whaaat?

I've never heard of anything like that.

First of all, no one makes 4kW individual panels. That's got to be an array. Current common sizes are between around 400W and 650W, or more precisely 2m^2 and 2.7m^2 with wattage slowly increasing with increased cell efficiency. (Or 1m^2 for the ~225W panels)

Second, they absolutely should make the stated power if pointed directly at a sun that is reasonably high (45º+) in a clear sky giving 1000W/m^2 of raw energy and with the solar panel itself at 25º C. (In fact in places with clear air such as New Zealand and Australia the sun can get to 1200W or even 1400 W/m^2)

[jonpaul]

Friend in NL gets reject panels at EU30/kw , has installed 15 panels and 4 inversters so far.

US has lease contracts and liscenced installers, electricians in No Cal rate is $150-250/hr

Jon

[Siwastaja]

All kind of overengineered weird stuff like microinverters (and optimizers) also seem to be a thing in the US. Outside of the US, we don't use microinverters since we prefer reliable, low-cost, battle-tested systems that also produce more energy with better efficiency and less limiting, but maybe this microinverter thing is like a 6-liter V8 engine, it's a symbol of freedom.

[ejeffrey]

All kind of overengineered weird stuff like microinverters (and optimizers) also seem to be a thing in the US. Outside of the US, we don't use microinverters since we prefer reliable, low-cost, battle-tested systems that also produce more energy with better efficiency and less limiting, but maybe this microinverter thing is like a 6-liter V8 engine, it's a symbol of freedom.

In many cases microinverters are a response to the high cost.  Equipment costs in the US are not that much higher than the rest of the world, it's installation costs that are way higher.  I know a lot of people who have done DIY microinverters because it's easier and people are more comfortable with standard 240VAC wiring that the high voltage DC strings. If you want you can hire an electrician to do the breaker panel installation, but that s cheap and any electrician can do it.  The microinverters cost more, but it's a trivial difference compared to the cost of "professional" installation.  On the other hand, most generic electricians won't touch or don't know how to install a string inverter. 

[bsdphk]

   I did get a data sheet for the solar panels from one company and it turns out that they're only called 4 kW panels and that they're actually only rated at between 3400 and 3600 watts per panel.

What you relate is certainly the kind of marketing USA is famous for, but it might not be wrong or even deceptive.

The standardized measurement conditions for solar cell/panel production (Insolation + Temperature) tells you what they can be expected to produce in clear sun in München, Germany.

If you are north of München (latitude > ~48°) you will get less out of them, south, which most of USA is, you will get more out of them.

Colder temperature ?  Higher production.  Warmer temperature ? Lower production.

I would not be surprised if a  3600W set of panels output more than 4000W in clear sun  in most of USA, in particular because most solar panels deliver more than spec, in order to support a "N-year production warranty" for some value of N.

[tszaboo]

I paid 900 EUR for the installation of my panels 5 years ago, 900 EUR in tax, and around 3000 EUR for a 3.5KW Solaredge system. Today I could probably get a 5KW system for the same price.
Why does it cost that much in the USA?
Why do you think the same medicine is 20 times more expensive in the US?

[ejeffrey]

   One word. Greed!

That isn't really an answer.  It's not like people in australia or france or other highly developed country are not greedy.

I don't think for the most part it's differing hardware certifications or standards.  The panels are the same and north american spec inverters are available at reasonable prices.  And wholesale prices of the hardware isn't that much higher, what's higher is the cost of installation.  Laborer wages are probably higher in much of the US than many other countries but I don't think that's enough to explain the difference.  It's really the installer overhead where all the money is going.

[brucehoult]

If you are north of München (latitude > ~48°) you will get less out of them, south, which most of USA is, you will get more out of them.

In summer at noon the sun should be high enough to give rated power to ... I'd say at LEAST 70º latitude.

Temperature is a much bigger problem than latitude. And skies that aren't clear.

You can compensate a LOT cheaply (far more cheaply than optimisers or microinverters) just by buying a few more panels.

My system has 2400W MPPT (actually 2x1200W). I have 2640W of panels (just 10% "over panelling") at a relatively flat 20º angle.

Even now in midwinter when the panels should ideally be mounted at a 60º slope (i.e. I'm 40º too flat, with the sun just 31º high at noon in June, and 34º today) I'm still hitting 2200W-2300W in the middle of the day when the clouds get out of the way.



Those are instantaneous power sampled only once every 20 minutes, so 10 or 15 minute clear gaps between clouds can easily sneak in between readings. That peak today is 2120W and the one two days ago is 2270W.

[brucehoult]

The Sun came out and I caught my PV power at 2251W at 1:10 PM at which time the sun is still at effectively 34º elevation 41 minutes after the 12:29 local noon today. But that 41 minutes means the sun has moved 10º West of North. Combine that with the panels at 20º slope puts the panel normal 36º high and 10º east from the sun, for total off-axis of sqrt(36^2 + 10^2) = 37.4º, so the power level should be cos(37.4) = 0.7944 of 2640W = 2097W.

2251W is actually better than 2097W. I have no complaints at all against these JA Solar JAM54D41-440/GB for which I paid US$65 + tax in September.

Once again, it is mid-winter here.

[bsdphk]

In summer at noon the sun should be high enough to give rated power to ... I'd say at LEAST 70º latitude.

You overlook the length of the light's path through the atmosphere.

Only between ±23.44° latitude, will the Sun reach zenit (Ie: be directly overhead) where it traverses the minimal amount of atmosphere.

The further north or south you are from that band, the more atmosphere the light has to go through, all year, including mid-summer.

But wait!  (drumm-roll!) it gets more complicated:

At lower latitudes the air is warmer and can, and do, contain more moisture, than the colder air towards the poles, and of course once you get to freezing temperatures, there is almost no moisture in the air, so simple trigonometry is not enough

And it's not that simple trigonometry anyway: For instance the sun is not actually located where we see it, unless it is in zenith position.

See also: https://pvlib-python.readthedocs.io/en/stable/reference/solarposition.html

Designing the correct balance between the rated power of solar panels and inverter involves many non-geometrical factors, and the most important factor is almost always how and how much you pay for electricity not generated by your PV-system.

But as a rule of thumb, to get most power out of your hardware, if your latitude is larger than 48° you should have more panel power than inverter power, and vice versa.

And I'll leave the discussion with that: I only wanted to make the point, that the sticker on the PV panels only tells you how they perform in München.

PS: Attached a "photo" I took with my PV panels in 2007-2008:

[Siwastaja]

You overlook the length of the light's path through the atmosphere.

Exactly, and additionally to even the "clear sky" absorbing more on this longer path, the probability of getting clouds in-between - and the mean thickness of the cloud layer - blow up at low tilt angles. So even if you install the panels vertically, you will never get the full nominal power out of them during the winter, and the distribution skews towards very low production days dominating since the weather needs to be absolutely "perfect" over wider area to gain closer to that already derated (by longer clear-sky distance) maximum. Semi-overcast days produce a lot in the summer when the sun is high and the cloud coverage is thin, but not much at all in winter.

[Psi]

I'd be curious to know what you US people pay for a 630W panel.
In NZ a brand new 630W panel is around NZ$170 retail, so that would be around USD$98
But I'm guess you're paying more that due to tariffs ?

[brucehoult]

In summer at noon the sun should be high enough to give rated power to ... I'd say at LEAST 70º latitude.

You overlook the length of the light's path through the atmosphere.

I did not. I EXPLICITLY considered path length, which is why I specified that the sun should be at least 45º above the horizon — which is true at midsummer noon everywhere at less than around 70º latitude (as I specified) — at which point the path length is quite minimal at only 1.4x longer than if the sun is directly overhead.

The effective depth of the atmosphere (i.e. the actual depth if the same mass of air was all at sea level density) is around 28000 ft or 8.5 km.

At a 45º slope the effective path length is only about 3.5 km longer.

When the sun is low the light can have to travel many tens of km through the atmosphere — and has a far higher chance of meeting a cloud. That's why I think it's just about useless to point panels east and west to catch the low sun ... unless you have no more space to put ones facing the equator.

If you like we can forget about 70º latitude. Very few people live more than 55º from the equator (Glasgow, Edinburgh, Copenhagen, Moscow, all the Siberian cities, Edmonton) where the midsummer noon sun gets 58.5º high, for about 1.5km extra path length compared to the zenith. And almost no one at all more than 60º (Oslo, Stockholm, Helsinki, St Petersburg, Yakutsk, Magadan, Anchorage) where the midsummer noon sun gets 53.5º high for about 2km extra effective path length compared to the zenith.

The summer sun in Oslo / Stockholm / Anchorage is more than 20º higher than the winter sun here in northern NZ (31.1º in June 17-26), but I get near maximum power from my MPPT if the sky is clear with just a 10% over-panelling so even at that low elevation the path length attenuation is not a huge thing.

Sure, having the sun at the zenith is optimal, but 30º or 35º or even 45º from the zenith is no big deal in terms of attenuation of the sun.

[5U4GB]

I've never heard of anything like that.

First of all, no one makes 4kW individual panels. That's got to be an array. Current common sizes are between around 400W and 650W, or more precisely 2m^2 and 2.7m^2 with wattage slowly increasing with increased cell efficiency. (Or 1m^2 for the ~225W panels)

Maybe they were counting on them being installed in WA (Western Australia, not that place in the US), which is located about 50km from the surface of the sun, with people crackling audibly as they walk down the street.  You can probably get 4kW out of a single panel there until it bursts into flames.

[bsdphk]

You do know there is a lot of science out there you can consult about this, right ?

But as I said:  The most important factor seldom has anything to do with hardware anyway.

Over and out.

[Psi]

The panels I've played with all seemed to be a bit over-spec'ed
eg, say 545W in DS but you'd get ~560W real world if the panel was brand new, totally clean and under ideal position.  I assumed they adding a bit of extra safety margin to deal with slight degradation from just being installed and not being ultra clean after being on the roof a week.

[brucehoult]

The panels I've played with all seemed to be a bit over-spec'ed
eg, say 545W in DS but you'd get ~560W real world if the panel was brand new, totally clean and under ideal position.  I assumed they adding a bit of extra safety margin to deal with slight degradation from just being installed and not being ultra clean after being on the roof a week.

NZ/AU sun can be up to 1200 W/m^2, or even more, the panels are rated at 1000W/m^2.

[tszaboo]

The panels I've played with all seemed to be a bit over-spec'ed
eg, say 545W in DS but you'd get ~560W real world if the panel was brand new, totally clean and under ideal position.  I assumed they adding a bit of extra safety margin to deal with slight degradation from just being installed and not being ultra clean after being on the roof a week.

NZ/AU sun can be up to 1200 W/m^2, or even more, the panels are rated at 1000W/m^2.

IIR there is 1300W/sqm outside the atmosphere, less than 1% variation of this. And 1120W/sqm as absolute maximum you can have anywhere on earth as maximum. And these values are pretty much constant.

[Psi]

The panels I've played with all seemed to be a bit over-spec'ed
eg, say 545W in DS but you'd get ~560W real world if the panel was brand new, totally clean and under ideal position.  I assumed they adding a bit of extra safety margin to deal with slight degradation from just being installed and not being ultra clean after being on the roof a week.

NZ/AU sun can be up to 1200 W/m^2, or even more, the panels are rated at 1000W/m^2.

ah, that makes sense.

[Someone]

The panels I've played with all seemed to be a bit over-spec'ed
eg, say 545W in DS but you'd get ~560W real world if the panel was brand new, totally clean and under ideal position.  I assumed they adding a bit of extra safety margin to deal with slight degradation from just being installed and not being ultra clean after being on the roof a week.

NZ/AU sun can be up to 1200 W/m^2, or even more, the panels are rated at 1000W/m^2.

IIR there is 1300W/sqm outside the atmosphere, less than 1% variation of this. And 1120W/sqm as absolute maximum you can have anywhere on earth as maximum. And these values are pretty much constant.

Speed reading wikipedia and coming up with a misleading claim?
1120W is typical not maximum:
https://web.archive.org/web/20131029234117/http://www.newport.com/Introduction-to-Solar-Radiation/411919/1033/content.aspx
A variety of peak measurements:
https://journals.ametsoc.org/view/journals/bams/104/6/BAMS-D-22-0215.1.xml
1400+ in NZ, topping out to >2000 in a high altitude desert.

[jonpaul]

All:

Great thread,

Beware of solar spec of panels:

Taken when new or at 100 hrs
On equaor at summer solstice solar  noon (90 deg azmuth)
Taken at 20 deg C temp, no self heat


Expect 30-60% of this rating in typical install and solarization dfrop of 0.5...2% per year of age.

COST: Recent TESLA  install in No Calif USA cost homeowner $36,000 for 25 KW. with 2 -3 batteries/

Jon