Tweet
CSA C22.3 No.9 defines OV1 as > 110% for 2 seconds, or 264V for a nominal 240V split phase service. I imagine your local utility follows a similar standard. I find most utilities distribution zones are designed to be a few percent hot, to help compensate for voltage drops from high loads. On a cloudy day when my PV system is putting out little power, L-L voltage at my panel is usually between 245 and 249 Volts. When the sun is out, that goes up by about 2 V, and at the inverter it goes up by a few volts more. My inverter monitoring has never recorded more than 255 Volts.
-
-
My power company is Indiana/Michigan Power. I live in Michigan, 15 miles north of the Indiana border.
Last December the power company put a data logger on my meter for a week. Average was a little over 124v, standard deviation about 1.5v, absolute peak 127v. The inverter routinely (as in multiple times per sunny day) records voltages in excess of 260v. Solar Edge says that they expect a 3% voltage rise over a run like mine. Two standard deviations on the logged data should be about 126.5 and adding 3% gets one to 260.6. So under these conditions I should expect to see some shutdowns, but nothing like I experience with the inverter shutdown so frequently, and often for hours, that production is off 50%. I can't see it because the inverter shuts down, but it looks like voltage routinely heads north of 260v on any reasonably sunny day.
Are there perhaps three problems here? One, that given the slightly high grid voltage and the inverter threshold set at 260v (if, indeed, it is), there is not enough headroom for the expected voltage creep. Second, voltage at the inverter routinely exceeds the expected 3% increase. And third, the inverter seems to vary considerably the voltage at which it shuts down, and for how long it stays down.Comment
-
My guess, your 6 ga is running 40A, dropping some 8V in the
loop, that gets added to the line voltage. Power loss of over 3%,
much like mine was. If going to this trouble, I would not go to
4 ga, I would go to 2 ga, recovering most of that loss and making
the most impact on the V trip problem. 1/0 ga aluminum will do
as well and cost less, direct burial would be the simplest. Look
up 1/0-1/0-2 Triplex Aluminum URD Direct Burial Cable. Here
the original cable was not disturbed, no service lost time.
Certainly address if the line voltage can be lowered. Or find
someone who can bump up the inverter V monitor settings. I
was able to do that on Fronius.
A clumsy solution is use a dual lead auto transformer to knock
off like 5 volts on each line to the inverter. But it would need to
handle like 500W, and would draw idle current 24/7.
good luck, Bruce RoeComment
-
Check your inverter specs.....most inverters will not accept large gauge alum. A conversion to copper will have to take place somewhere.Comment
-
The installer did give me an alternative proposal for a #1 aluminum wire for $4400. The proposal for the #4 copper is $4100. Cost aside, is the aluminum preferable?Comment
-
Here the inverter AC is wired to a disconnect box, that to a circuit breaker,
the breaker plugged into a dedicated distribution box. So the box is what
must interface to the long run. Here a 200A 20 circuit box was a couple
hundred $, I used one which easily connects even to my 4/0 aluminum.
understand gauges
6
5
4
3
2
1
0 or 1/0
00 or 2/0
000 or 3/0
0000 or 4/0
6 to 1 is 5 steps larger, which gave the improvement I wanted. BUT
that is for similar metal, going from copper to aluminum you must go
2 more steps to get the same performance.
15KW, I went from 4 copper to 0000 or 4/0 aluminum. The 300 ft roll
of aluminum triplex was (in 2019) $650 delivered. I estimated the
equivalent performance 00 or 2/0 copper would be $2600. No the
aluminum is not better, but it is WAY cheaper. These items were about
my only expense for a run of about 260 ft.
I do not know how the installer arithmetic works, but it is way different
from mine. Given your shorter run and and the overall smaller wire
(and your voltage issues to date), I think the installer is shorting you
on the wire, wasting energy, and might not cure your trip problem.
250 ft of the wire I mentioned is $700 delivered today, cannot believe
copper of similar performance could come close. Bruce RoeComment
-
I just analyzed the data on inverter shutdowns over the last week and I'd appreciate your comments on it.
There were 41 shutdowns for excess voltage (plus 3 for hitting the inverter 10kw limit). That's about 6 per day. The range was 1 to 13.
Average voltage at shutdown was 260.1v, but the range was 251.0v to 262.5. Fully 22 of 41 shutdowns were at less than 260.0v.
Shutdown time averaged 20 minutes but, again, the range was large: 5 minutes to 3:35 hours. Three shutdowns were for more than 1:50 hours.
Does this sound like appropriate inverter functioning? Why would it be shutting down at 251.0v? Or for that matter anything less than 260? And why would it shut down for such extended periods?Comment
-
I'm not an expert in power line monitoring, but regarding the power line monitor the power company left on site, was it a calibrated device (regular certification)?
How was the device set up? Trigger on event or continuously monitor the entire test period at a xx microsecond sampling rate?
My SMA inverter follow UL 1741 rules for frequency and voltage ride through requirements. Some of these parameters are distilled in the attachment SMA thresholds.
Previously, peakbagger brough up the point that sometimes inverters have internal parameters tighter than standard voltage range. As stated previously, my SMA inverter has separate settings for grid reconnection limits vs grid connection limits. Once a fault happens, the inverter goes through an unknown set of boot up processes that checks and tests the grid. One of those variables exposed to me in my parameters menu is reconnection upper voltage limit, 127vac per leg or 254vac between legs.
It is entirely possible that SolarEdge's grid checking routine compares the grid to a set of internal inverter variables that you have zero visibility to seeing.Comment
-
The line is not a perfect lab clean sine wave. There are bumps as
things cycle, it could easily bobble a few volts and shut down the
quick reacting inverter electronics. Cross that line and my inverters
were off 5 minutes. Look for a significant operational improvment,
a small one may not cure it, just reduce the shutdowns. Bruce RoeComment
-
Okay first off I am not an electric guy but everyone seems to be talking voltage fluctuations in this thread. If you look at the parameters that Oregon_Phil posted there is a very narrow margin for the frequency threshold. Could the problem be unstable frequency causing the issues? I ask because I just saw a YouTube on this but I can't remember where and the topic.Comment
-
I just checked my frequency over the last two days. 60 Hz, +/- 0.02 Hz.Comment
-
I have a hunch when the solution it found it will be multiple causes adding up to create your troubles. "Normal voltage" 250v but on the high side.......acceptable voltage drop but also at the upper limit, perhaps a connection that's loosen just a tad, just enough to be the provable straw that broke the camel's back.Comment
-
This may not help your issue but years ago I had an old Xantrex inverter go offline because of higher grid voltage. When I called SCE about the issue they fixed it quickly and the field guy explained that they routinely adjust things at the substation. In this case they had upped the voltage because of summer loads and did not ramp them down in the Fall which created higher voltage than they intended.9 kW solar, 42kWh LFP storage. EV owner since 2012Comment
Comment