I've just received two quotes for installing a rooftop solar system at my residential address (near Sydney, Australia). This caught my eye:

Company 1:

  • Panel: 22 x ET panels 300w
  • Inverter: Growatt 5 kW (5000 MTL-S)
  • Size: 3 - 6.6 kW

Company 2:

  • Panel: 20 x Canadian 300w
  • Inverter: Sungrow 5 kW (SG5KTL-D)
  • Size: 3 - 6 kW

Both offers are for 6.6 kW systems, but both come with an inverter that is only 5 kW. Math tells me there's a 1.6 kW overage here that is unaccounted for.

Is this ok? On the off-chance I manage to produce over 5kW of power, do I run a risk of burning it out? Is there something I'm missing here?

  • 1
    OP, is this a final setup or do you intend on growing the installation more in the future with additional panels?
    – Criggie
    Commented Jan 6, 2020 at 20:18
  • 2
    @Criggie - This is a final setup. I thought it would work out easier to do the phased install but honestly it seems to be just as easy to do it all at once
    – Robotnik
    Commented Jan 6, 2020 at 22:17
  • 3
    The fact that you're asking this question tends to indicate you're assuming you're going to routinely get 6.6kW out of your system with panels rated at 6.6kW. That's unlikely to happen, and it certainly won't happen most of the time. We don't know your use case. Just be sure the power which is expected to be generated from your system (i.e. far below 6.6kW on average), will meet the needs you actually have under the conditions which prevail in your location and with the installation you're having done.
    – Makyen
    Commented Jan 7, 2020 at 0:58
  • 1
    @Makyen - I've added a sentence addressing why I thought this might be a problem. :)
    – Robotnik
    Commented Jan 7, 2020 at 2:45
  • @Robotnik: With which orientation? Towards the North? Do you know the slope of the roof? Commented Jan 9, 2020 at 7:30

8 Answers 8


It could be that the system will spend such little time providing above 5 kW, once shading and losses are taken into account, that it's not worth the cost of getting the next-largest inverter.

It could be a restriction of your electricity network operator: often, they have bandings based on the maximum power that a PV system will deliver to the network, with different types of permitting applying to different bands.

Typically, there will be a cheap, easy permit for systems below a set small power rating, typically around 5 kW. And then a more expensive permit with a more time-consuming application process, for systems 5 kW - 100 kW (for example) . And one for 100 kW - 5 MW, and so on ...

The combination of these two things means that it's probably not worth your while having a system at 6 kW. A 5 kW inverter will be the most cost effective. And the extra 1 kW of panels means that you'll get best use out of your inverter - it will spend more time running at its peak output.

And no, you won't burn it out: the inverter won't draw more power from the panels than it can handle. It's completely standard to have a set of panels with higher nominal power than the rated inverter output.

  • 9
    It's possible that given the latitude, roof pitch, and the direction the roof faces the panels can never achieve their rated output, or will only do so briefly around noon around the summer solstice, even if you assume perfect weather and no shading. I would expect manufacturers' planning software to build in a derating factor based on the site survey, or it could be done with a set of tables.
    – Chris H
    Commented Jan 6, 2020 at 11:44
  • 2
    Or twice a year, 23 days before and after the Summer Solstice, at 2:31pm. Commented Jan 7, 2020 at 2:09
  • 3
    @Harper-ReinstateMonica: Max power output is usually achieved in may (in the northern hemisphere). Ambient temperatures aren't too high yet but the sun is already close to its highest declination. Commented Jan 7, 2020 at 8:34

Standard Test Conditions vs Real life

Photovoltaic panels are rated at "Standard Test Conditions" (STC):

  • Irradiance = 1000 W/m²
  • Cell temperature = 25°C
  • Air Mass = 1.5

Those conditions are achieved in testing laboratories, and basically never happen in real life. Modules need to be cooled down in order for them to stay at 25°C.

1000W/m² is a lot of irradiance, and if it ever happens, it will be at a high ambient temperature. Photovoltaic modules get hotter under high irradiance, so a module under 1000W/m² at 25°C ambient temperature might have a cell temperature of around 60°C. Cell performance decrease with increasing temperature, so your 300Wp modules will almost certainly never achieve 300W once installed on your roof.

Growatt 5000MTL-S

Here is the datasheet for this Growatt inverter: enter image description here

Growatt 5000MTL-S has a maximum recommended DC Power of 6150W and a max AC output of 4600W.

Depending on your location, this inverter might indeed be too weak for the solar panels. At the very least, the company in charge of the design should check if there's a mismatch.

Is it dangerous?

An undersized inverter shouldn't be dangerous for your inverter or your house, though. According to this SolarPowerWorld article ("Solar inverters and clipping: What DC/AC inverter load ratio is ideal?"):

During times when the DC input power is too high, the inverter will raise the operating voltage of the modules to pull the array off of its max power point and reduce the DC power.

With a Euro-efficiency (a.k.a. average efficiency) of 97.4%, the Growatt 5000MTL-S will radiate at most ~150W as heat.

Orientation, strings and MPP-tracker

You wrote in a comment:

The house is slightly L-shaped. I assume the majority of the panels will face north or south, with some on the east portions of the building. Slope I'd say is ~30° but that's a guesstimate

It means that not every panel will produce the same power at the same time and it might lead to further losses:

  • panels in series (a string) have to deliver the same current. They will only produce as much current as the worst panel. The proposed panels have bypass diodes in order to mitigate this effect, but it would be better to only connect together modules which are unshaded and which have the same orientation.
  • The proposed inverter has 2 MPP-Trackers, which can optimize the power output of two strings separately. It shouldn't be used with 3 different orientations. South orientation wouldn't make much sense in Australia anyway.

Ask for a simulation

There are are handful of reliable photovoltaic system simulation softwares (e.g. PV*SOL, PVSyst, NREL SAM, INSEL, ...).

Any serious PV company should have at least one license and be able to quickly simulate your planned installation, in order to check if there's any mismatch: not just for peak power, but also for maximum current, maximum voltage or MPP voltage range.

As a result, you'd get a:

  • potential energy yield [kWh/a]
  • a performance ratio [%]
  • a specific energy yield [kWh/(kWp.a)]

The energy yields depend on your location so it's not possible to give any estimate without more information. There are some easy to use online tools if you want to check yourself (e.g. PVGIS, PV*SOL online, NREL PVWatts Calculator).

If the performance ratio is lower than 85%, it probably means there's a problem with the design.

  • 6
    For a simple estimation, the EU has a free and easy to use tool called PVGIS here: ec.europa.eu/jrc/en/pvgis, which also appears to work for locations in a good part of the Americas. Commented Jan 6, 2020 at 16:05
  • 3
    @AlexanderKlauer: Indeed, it's a really good tool. It assumes that the PV system is correctly designed though, right? As far as I know, you cannot specify inverters and panels in order to detect mismatch, for example. Commented Jan 6, 2020 at 16:09
  • 1
    Correct, you simply give it some geometry info (geocoordinates, mounting position), installed peak power and an overall "system loss" percentage. But it's good to get a general idea of what the actual peak production might be (use the "Hourly data" tool and load the resulting CSV file into your spreadsheet app). Commented Jan 6, 2020 at 16:20
  • 2
    @AlexanderKlauer: Thanks, I updated the answer. Commented Jan 6, 2020 at 16:28

Solar systems are rated at peak power. It's normal for inverters to be rated less since you will very rarely have ideal conditions to reach peak power. For example, my inverter is rated at 80% of peak power and I live in a sunny area. Only during the best days in the summer does my inverter get maxed out for a few hours at a time.

Regarding your edit, there should not be a risk of burning out the inverter. It will just sustain at its maximum power during the very rare times your panels could provide more. Search the Internet for "inverter clipping" to get more details about this, the loss is typically minimal or none.

  • 1
    +1. This is the only one that answers whether there are risks in using lower-rated inverter.
    – justhalf
    Commented Jan 7, 2020 at 7:19

There already are some answers addressing the fact that you probably will not reach or will rarely reach the rated power of your panels. That explains why you don't need a higher rated inverter. Now, you could want a higher rated inverter just in case, assuming it doesn't hurt either. But that assumption is often wrong.

Inverters are not perfect; they have a certain efficiency which varies with the load they get. Generally, they are less efficient at low loads, where 'low' is relative to the maximum load. This means that using an oversized inverter which runs at say 10% of its capacity most of the time will waste a lot of energy because of the reduced efficiency. So you want to size the inverter in such a way that it runs at high efficiency most of the time. Even when this means you lose on the moments where your panels peak above the maximum you will likely still gain overall.

Obviously there is an optimum somewhere, which very much depends on the specific inverter and local circumstances. But generally, solar panels spend far more time producing a relatively low output, making the most of that gains you way more than getting the most out of the occasional peak.

To get some insight into the specific offers, you should look up the efficiency curve for those specific inverters. (Similarly, you might want to look up how the specific panels perform under less than optimal conditions.)


In some jurisdictions (like Flanders, Belgium), you have to pay a fee based on the maximum power your inverter can deliver to the grid.

As EnergyNumbers notes, slight undersizing won't necessarily reduce the output of your system that much, and smaller inverters are cheaper. When you add this fee, an undersized inverter may become more profitable than a 'properly' sized one, if that size is even available. Oversized inverters can take large chunks out of your profits in these cases.


While this has already been correctly answered I'll expand upon the fact that the panels don't produce their rated power:

I live in a very solar-friendly climate, there are solar installations all over the place. I have yet to see rooftop solar in any orientation other than flat against the roof. Since snow is not a factor roofs here generally have a low pitch--by eyeball ours is less than our latitude, thus guaranteeing a panel flat on the roof can never produce full power. A panel's output goes at the cosine of the angle between the panel and the sun.


It might be ok. Check a map to see whether sunlight in your region ever gets stronger than 760 Watts/m^2. If it does not, then 6.6 kW(peak) ideal rating of your solar panels won't deliver more than 5 kW ever. If you still have concerns, ask a rival solar panel installer. He'll be more thorough than some internet forum. As a bonus, you might get better inverter efficiency from running a slightly undersized inverter. Mine is slightly oversized - 3.0 kW rated inverter with 2.45 kW(peak) rated solar panels. Beware, as some installers tend to try too hard to sell the inverter which they had a training course on.

In the uk, I found it well worth paying someone to fit panels properly on the roof. They get much more hours of light than the DIY ones near ground level on my shed.


The two reasons why it often makes sense to use less inverters than panels:

  • Inverters are expensive, nearly as expensive as panels
  • To put that expense into useful work, you want big enough capacity factor

Let's consider that you have 10 kW of panels, and you are deciding how many 1 kW inverters you want to put there. Let's also assume the inverters share the panels, in such a manner that even the first 1 kW inverter can use power from all of the 10 kW panels.

Investing in the first 1 kW inverter is easy. 10 kW panels will create more than 1 kW of power from sunlight nearly always except during the night. They will create more than 1 kW of power probably even in cloudy weather and the sun doesn't have to be perfectly aligned with the panels. So the first inverter has a very high capacity factor: it is creating lots of energy because the utilization hours occur often.

The second 1 kW inverter (for a total of 2 kW capacity) is somewhat tougher decision. There are times when it isn't fully used, times when the first 1 kW inverter would be enough. For example in very cloudy weather, or near sunrise or sunset you will find the second 1 kW inverter is fully used.

Then let's skip the third to ninth 1 kW inverters and consider the tenth 1 kW inverter. What use is it? Well, you need quite many simultaneous conditions to hold for it to be any use when comparing it to 9 kW of inverters:

  • The panel array needs to be in perfect condition to get that last kilowatt. For example 20-year old array could have reduced power production so it can't produce 10 kW at all, it could max out at 9 kW.
  • The panel array needs to be clean. Dust over it can reduce production.
  • The sun has to be perfectly aligned with the panels.
  • There has to be absolutely no clouds.
  • The air mass effect of low sun can't be there. For example in a very northern location it's possible that a 10 kW panel can only produce 9 kW even in optimal situations because in northern locations the sun is quite low even in the summer, meaning sunlight has to be travel over a longer path in atmosphere

I'd say these conditions occur very rarely: the requirement of perfect orientation limits it to at most few months in a year and only few hours per day on those months, the requirement of needing perfectly sunny weather also removes most of the hours in those few months, and you have to keep your panels clean. Also even if those requirements are satisfied, it's possible the panels have aged or you live in a location where the panels can't produce full power.

So, in all cases, it actually makes sense to install less inverters than panels. How much less depends on the relative pricing of panels and inverters, on the price of electricity, on the location (typical weather and distance from equator), and also on whether you can orient the panels well or if their in-optimal orientation alone limits power production.

5 kW inverter for 6.6 kW panels sounded about right in most cases in 2020. I won't say that was perfect because prices of inverters, panels, installation work and electricity prices varied.

Inverter capacity should be full only in cases where inverters are trivially cheap, yet panels are horrendously expensive. Few years ago, the situation was such that solar cells and panels were very cheap (only their installation had real costs) and inverters cost a lot of money. Today panels are getting more expensive and due to popularity of solar increasing installation also becoming more expensive in many areas, so it no longer may be the ideal to have 5 kW inverter for 6.6 kW of panels. Also, in many areas electricity is now so expensive so the additional missing 1.6 kW of inverters could be easily paid back by high electricity prices, even if the last 1.6 kW of inverter capacity is producing electricity only rarely.

This question was asked in 2020 when panels were very cheap.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.