Problems when using PVWatts

How are you there!
My name is Jiakai, an entry level PV engineer. I met problems using PVWatts V5 when select DC/AC ratio.
I did three tests each has only one/two Parameters change either DC size or DC/AC ratio.
Please refer to the attachment for results.
test2 has a 2 times DC size(0.54kw) of test1(0.27) and also 2 times in DC/AC ratio(which is 2).
test3 has a 2 times DC size(0.54kw) of test1(0.27) but everthing else are same.

The result shows a nearly same annual AC power for test2(608kw) and test3(654kw). But big difference between test1(327kw) and test2(608kw). The result value is weird from my side. Because in test1 and test2, system actually share the same rated AC power, so that give me a feeling that PVWatts is DC size dominated. However, after refering the instruction of PVWatts V5, it says that "When the predicted AC output exceeds the nameplate rating, the output is clipped to the nameplate value." Well I didn't see a obvious clip here. I know the a DC/AC ratio of 2 is two big in reality, but just for verification.

Please help with that. Thank you!
New attachment added.

Hello,

Your question is about the PVWatts implementation on the web, but I will answer it here because PVWatts is also implemented in SAM.

Here's what I see:

Test 1: 0.27 kW with DC/AC ratio of 1 generates 327 kWh annually.
Test 2: 0.54 kW with DC/AC ratio of 2 generates 608 kWh.
Test 3: 0.54 kW with DC/AC ratio of 1 generates 654 kWh.

Test 2 has a smaller inverter than Test 3 (the array's DC capacity is the same in both cases, but the Test 2 inverter capacity is half of the Test 3 inverter capacity), so its output is less because the system reduces the array's DC output during periods of high output to avoid overloading the inverter.

From the solar radiation data, you can see that the solar resource during the winter months is much less than the summer months (December is 2 kWh/m2/day, while June is 5.9 kWh/m2/day), so for this location, a system with a higher DC/AC ratio (smaller inverter) may be more economical than one with a lower ratio (larger inverter) because you do not need the extra inverter capacity during the winter months.

Best regards,
Paul.

Thank you so much for your reply Paul! You information are really appreciated. By your answer, I have understood the relation between the tests and the practical use of DC/AC ratio. But I still have one more concern: In either PVWatts or SAM, it brings too much AC output increment when we increase DC size and keep the same inverter size (In other words, the AC decrement is too small when we decrease inverter’s size and keep the same DC size).

If we still refer to the 3 tests we did: Test3 AC output is twice to Test1, this is reasonable because both the DC rated size and AC rated size are doubled in Test3. So the AC increment is 100% here. But if we take look the AC increment of test2 to test1, it is (608-327)/327*100%=85.9%. This value is ridiculous because actually the inverter in these two tests is actually the same one. Can the inverter make use of so much overbalanced DC power and convert to AC?

These days I was trying other platforms to perform the same tests (1 and 2) for the results. And I found a more reasonable result in SMA’s sunny design:
Panel: Yingli 270w
Inverter: SMA SB240 (rated AC 240, Max DC input 250w)
SMA Test1:Panel*1(270w), AC annual production:363.9kWh
SMA Test2:Panle*2(270w), AC annual production:364.8kWh
Although the AC output here is little bit lower than expected, but still holds the line.

Paul, please refer to the new attachment in my original post floor for the SMA results.
And one more article supporting my opinion was found: “DC / AC Ratio Optimization” by Gordon Woodcock on SAM Virtual Conference 2013. It’s an article on NREL’s website:
sam.nrel.gov/sites/default/files/content/virtual_conf_july_2013/07-sam-virtual-conference-2013-woodcock.pdf

And actually it also has the presentation on Youtube:
www.youtube.com/watch?v=nvN6s1EpdS8

In this slides, page 4 shows the daily performance of different DC/AC ratio on an inverter.
Although the DC size is increased from 1.16 to 1.33, the AC gain on the diagram is still limited.
I also include this screenshot in the attachments please check.

Is it my illusion that the PVWatts result for Test2 is too high? Or it is supposed to be like this? I’m confused on this for long time, please help me out Paul. Again, Thank you so much!

Best,
Jiakai