Balancing Inverter Size and Amount of Solar Panels

[Zeigh]

Hello,

How do I find the best balance between inverter size and amount of solar panels?

First, here are the basics of my system aimed at 100% coverage in a grid-tie setup. My average monthly usage is 2393 kWh, so I am sizing my solar electric system to a 14400 watt inverter severed by forty 350 watt solar panels. All of my solar panels will be installed upon a south-facing 6/12 roof (26.6% angle). The optimum mounting angle for solar panels in a fixed system year around in northern Arizona is 29.7% and provides an average of 71.1% efficiency.

So, that means that my intended system would only be pumping out 9954 watts of the 14000 watts maximum. 56 solar panels would be needed to make up for this loss and serve the maximum capacity of the inverter. Of course all these numbers are averages, so how do I calculate the best balance of solar panels to include? There is allot of solar electric research information out there, but apparently I can't seem to hone in on this particular detail (or give my contact information to be spammed to death).


Peace,
Dr. Z.

[Mike90250]

it's a judgment call. Usually the battery voltage is the starting point for inverter size:

12V = no larger than 1kw inverter
24V = 2kw
48V = 4kw
this is because of the wire gauge needed to safely carry the DC amps to the inverter and the internal wires in the inverter. I trust very few 3kw 12V inverters, they consume 250A to provide 3kw At 48V, that's only 60A DC

You accurately size your max loads and your daily usage. Don't size for motor starting, unless you are over 1/2 hp.
Once your daily usage is determined, you set your battery bank voltage and ah.
Then you size your solar to be large enough to recharge your bank in 1 short day in the winter.
If you have very short solar days in your area, and high usage, you are looking at an expensive array, or a generator to share the loads

[Ampster]

The question you pose can be described as, what is the optimum DC to AC ratio. For example my latest system has 5.7kW of panels and a 3.8kW inverter for a ratio of close to 1.5 to one. The use of high DC to AC ratio has become more common as the cost of panels has dropped. It does result in clipping of the top of a typical sine wave shape of generation during solar Noon.. I questioned my installer and they ran two models of my system, one with my existing inverter and one with a larger inverter. Over a year, using PV Watts insolation data as input, the difference was 100 kWhrs for a system that theoretically would produce almost 8,000kWhrs in a year.

This has been discussed at some length and opinion varies. That opinion may also depend on who you ask. For example, several years ago I had an inverter failure and I chose to replace it with a Solaredge inverter. The panels were 4 year old Canadian Solar panels with an initial STC capacity of 4kW. The vendor recommended a 5.8kW inverter. I asked about using a 3.8 kW inverter and he said it was not good to overwork a smaller inverter. The price difference was $150..He was a salesman working for a supplier.

A year later I bought another home and had an installer design and install the system described in the first paragraph. When I saw the clipping from the Solaredge monitoring software, I began researching the issue. I found that Solaredge allows as much as a 1.50 to one DC to AC ratio. Which design is optimal? Honestly looking back, given that I own both systems I would prefer to have a closer match, but the difference financially is trivial.
In some cases it may depend on your location and the specifics of a net energy metering policy by your power company.

[Ampster]

it's a judgment call. Usually the battery voltage is the starting point for inverter ....

I agree it is a judgment call. In this case, the OP is asking about a grid tie system..

[Mike90250]

Sorry - i missed it's a grid tie setup. I'd have the inverter be 85% of the panel wattage. With expected thermal heating power loss, you should be maxing the inverter only briefly each day. If your site is cooler, you could use 90% of PV wattage.

[Ampster]

.....I'd have the inverter be 85% of the panel wattage. With expected thermal heating power loss, you should be maxing the inverter only briefly each day. If your site is cooler, you could use 90% of PV wattage.

That would be a DC to AC ratio of less than 1 to 1. Some transformerless grid tie inverters like the Solaredge HD Wave series don't have the heat issues that inverters with transformers do. As mentioned elsewhere my installer specified a 1.50 to 1 ratio and I have not seen heat issues in the summer in Sonoma. Others have said it is more efficient to run these type of inverters at closer to 100%. It may be different for hybrid or off grid inverters.

[Ampster]

Sorry - i missed it's a grid tie setup. ...........
.

No worries, I think I missed the real question. I believe the OP wanted to know what size should the system be if his monthly consumption is about 2400 kWhrs a month or about 28,800 kWhrs a year. I saw the roof pitch, the orientation to south and the location as Northern Arizona. These are all inputs that can be used on the NREL PV Watts Calculator at https://pvwatts.nrel.gov to get an estimate of the annual production from a particular solar panel system. You can play around with efficiency but starting with the typical assumption gives you a baseline.

Fo my location PV Watts would have given me a system size of 19.2 kW need to generate 28,800 kWhrs per year. I believe PV Watts figured in average hours of insolation and I suspect Northern Arizona has more sunshine than Sonoma California. I am traveling and only have my phone for input or I would enter his data in PV Watts and paste in a screen shot of that calculation. It is easy enough to do on a computer.
Once you have decided on a system size you can decide if you want to be aggressive on a DC to AC ratio or just go with 1 to 1 or even less as Mike90250 suggested. An inverter is usually less than 10% of the system cost so using a less aggressive ratio costs a little more but allows you to expand the system later by adding panels and reconfiguring strings. As I mentioned, my installer designed my system at near the max that the inverter would allow, so I have no room to expand. The installer did guarantee an annual production number that exceeded what I got by using PV Watts and it has performed to that estimate the first year.

[Mike90250]

> With expected thermal heating power loss, you should be maxing the inverter only briefly

That's the loss of voltage in the PV array when it's hot.

I'd assume the inverter is located to not worry about heat. If in hot climates like Nevada desert, then you should consider some extra headroom for the inverter if it's not in conditioned space.

[Ampster]

> With expected thermal heating power loss, you should be maxing the inverter only briefly

That's the loss of voltage in the PV array when it's hot.
.

Well maybe I'm lucky. My installer extended the warranty for the my grid tie inverter to 25 years. They also designed my system such that it is running at full capacity (and clipping) for as much as 3 hours a day in full sun. Again that may be unique to the Solaredge HD Wave transformerless design. The Outback Skybox that I installed myself is actually running only at 50% of capacity and I don't anticipate pushing it much further but its role is load shifting, self consumption and backup.

As far as the loss of power from the panels when they are hot, I know that is a measurable phenomenon. In that regard, I heard an interesting excuse from the installer when I asked him about the aggressive DC to AC ratio he designed into my system. I was complaining about the obvious flat top to what I expected as a sinusoidal production curve. He said the apparent loss that I was perceiving was not that great because when they ran the scenerio they saw that the program predicted a fall off of production of the panels on hot days. That is just one of the many of the kind or arguments that are used to justify high DC to AC ratios.

[Mike90250]

You don't want too much summertime flat top, in winter on cold days, I caught my 3kw array pushing 3600watts Wooo hooo !

[Ampster]

You don't want too much summertime flat top, in winter on cold days, I caught my 3kw array pushing 3600watts Wooo hooo !

Did some of that have to do with reflectance from snow on the ground or high citrus clouds?

[J.P.M.]

Well maybe I'm lucky. My installer extended the warranty for the my grid tie inverter to 25 years. They also designed my system such that it is running at full capacity (and clipping) for as much as 3 hours a day in full sun. Again that may be unique to the Solaredge HD Wave transformerless design. The Outback Skybox that I installed myself is actually running only at 50% of capacity and I don't anticipate pushing it much further but its role is load shifting, self consumption and backup.

As far as the loss of power from the panels when they are hot, I know that is a measurable phenomenon. In that regard, I heard an interesting excuse from the installer when I asked him about the aggressive DC to AC ratio he designed into my system. I was complaining about the obvious flat top to what I expected as a sinusoidal production curve. He said the apparent loss that I was perceiving was not that great because when they ran the scenerio they saw that the program predicted a fall off of production of the panels on hot days. That is just one of the many of the kind or arguments that are used to justify high DC to AC ratios.

Panel efficiency does indeed change, pretty much inversely and mostly linearly as cell temp. increases or decreases. My temperature measurements of my system' s panels over the years pretty much confirms mfg. temp./voltage coeff. Long, boring story.

[Mike90250]

Did some of that have to do with reflectance from snow on the ground or high citrus clouds?

No snow on ground, no hazy clouds, just really cold and clear. After a minute, the power dropped down as the panels warmed.

[inetdog]

Did some of that have to do with reflectance from snow on the ground or high citrus clouds?

Ought-to-correct for cirrus?

When life gives you lemons, make lemonade.
When life gives you melons, you are probably dyslexic.

[Ampster]

Ought-to-correct for cirrus?

When life gives you lemons, make lemonade.
When life gives you melons, you are probably dyslexic.

Haha, I didn't even notice. I will leave it like that for the humor.