Dirty panels causing battery issues?

[BenHawaii]

We have a hybrid 48V Outback system with AGM batteries that connects to the utility grid when the battery voltage declines to 48 volts or 60% SOC. The system is managed using HBX mode with the Outback Mate3 controller. We had the system installed to reduce our electric bill with the current rate at .46 cents per Kwh, and to have uninterrupted power when there are power outages, that might occur 3-4 times a year, but can last for multiple days due to downed lines. We are on the East side of Hawaii island which is known for being rainy and cloudy for extended periods of time in the Winter.


7,200 watt array (18 x 400 watt panels) facing 53 degrees NE unobstructed on roof at about a 25 degree angle.
9,000 Kwh (usable) sealed AGM battery bank- 8x l16 AGM 400 AH @ 20 Hrs
System installed Nov 2021


Our daily usage varies from 11-14.5 Kwh. At this time of year we tend to generate 14-16 Kwh from the array. The batteries carry us through from 5 PM to 4 AM, and then system connects to the grid. We buy 1.5 to 3 Kwh daily, depending on the weather. The array receives the most sun in the morning averaging 2.5 Kwh per hour with our hourly usage .7 to 1.2 Kwh also in the morning hours. The system quickly reaches the Absorb state, also early in the morning.

Our usage during the day is usually 500-700 watts per hour, and when we are asleep 400 watts. We rarely exceed 1,500 watts total when using appliances, such as a toaster, microwave, coffee maker, washer. We try to do laundry, run the dishwasher in the late morning or early afternoon. We check the system graph daily and overall during the day the battery graph line is green (above 50V). Between 5 PM - 10 PM we consume 500-700 watts per hour.

Five days ago I began noticing that we were connecting to the grid in the late afternoon, just as the sun sets. Then the past two days we only generated 9.5 Kwh each day, even though it's been clear and sunny and batteries still reaching Float stage. Also during the past two days when the sun was setting the battery voltage dropped from 52V to 48V in under an hour, and the system connected to the grid. The status graph shows we only consumed 700 watts during the same period.

I went out to the yard with binoculars and saw the panels need to be cleaned. I also compared the daily graphs from March 2022 to 2023 and noticed an average decrease in the daily generation from the same period in 2022. However, I don't understand why the daily power generation each of the past two days would decline by 5 KWh, from the usual 14.5 Kwh. Maybe this is typical when panels become dirty. However, it also seems to be affecting the battery bank since the batteries are not holding a charge after the sun sets. We also had an abnormally rainy February. I went through the status log and the batteries only reached Float stage 6 days in February. In January they reached Float 16 days in the month; and so far 10 days in March out of 17.

I've scheduled a panel cleaning and am taking the system offline for the next couple of days, which are forecast to be sunny. Thinking this might help the batteries be fully charged and in float for an extended period without any load.

Just hoping to get some feedback on this issue.

Thanks
Ben

HBX Settings.pngCharge controller settings.pngAbnormal very cloudy rain day Feb 2023.pngRapid discharge 3:16:23.pngTypical day March 2023.pngTypical day March 2022.png

[chrisski]

We have a hybrid 48V Outback system with AGM batteries that connects to the utility grid when the battery voltage declines to 48 volts or 60% SOC.

IME with my smaller RV build of 2550 watts of panels, cleaning makes no measurable difference, perhaps 1%.

My setup is off grid, but what I see makes me think discharging AGM batteries to 50% each night means the batteries may be depleted.

Your batteries manufactured chart should be used, but these are generic charts I found. The first shows that 48 volts = a charge state of 50%.
C15B315E-CFF7-4C7F-8847-4B1B3E5B160C.jpeg
The next shows that cycling AGM batteries to 48 volts 50% every day may deplete them in as little over a year.

05F0F048-2B1D-47CC-B4C7-DE043592D9CB.jpeg
With a 9 kWh battery pack depending on how much is being drawn would actually make the DOD less when the system switches it and further shorten the batteries, this could be much more so if high wattage appliances like air conditioners are running at the time of this .3 hour cutoff.

I’m not sure how your SOC is measured, if its a shunt, that is accurate, if its no shunt and off voltage, SOC can be quite inaccurate and require batteries to sit a while without use.

The only way I know to measure State of Charge with an AGM battery is to let them sit disconnected for a while and use a volt meter. If you can do this you can see what your batteries are. If they settle to 49.1 volts after a full charge, the batteries are at a little under 70% of original capacity. To me the reduced battery capacity is what may be causing the reduced production. Only 2 kWh instead of 5.5 kWh could be going in the battery each night.

Lithium packs are getting much more affordable and there are some that come UL approved like the EG4 server rack batteries that although still more expensive up front end up lasting for thousands more cycles.

[bcroe]

It might be, your batteries are not taking much charge any more,
so panels have generated a lot less KWH energy when the
charging is cut off. That would be a problem with batteries, not
panels. But I would not expect that, if you send excess energy
back to the grid. Bruce Roe

[BenHawaii]

The only way I know to measure State of Charge with an AGM battery is to let them sit disconnected for a while and use a volt meter. If you can do this you can see what your batteries are. If they settle to 49.1 volts after a full charge, the batteries are at a little under 70% of original capacity. To me the reduced battery capacity is what may be causing the reduced production. Only 2 kWh instead of 5.5 kWh could be going in the battery each night.

I appreciate the review of my data and suggestion on testing the battery. There is currently no load on the system and batteries are charging (Outback showing 57.2V) and in Absorb state. Will let system switch to float and sit for remainder of day. I will disconnect batteries, test, then let them sit disconnected and keep testing over the next couple of days and post results.

The system was installed by a local Outback dealer, who lives in our community, and has been in business for a while. I also did extensive research on solar and a indepth energy audit of our home using Kill a Watt meter etc. and I increased original proposal from a 5,340 watt array to 7,200. Batteries were purchased by installer at a local battery dealer, who has also been in business here for 40 years. I will contact the battery dealer and ask for the spec sheet to verify.

Last year we had two volcanoes erupting, Moana Loa and Kilauea, spewing volcanic gasses and ash into the atmosphere. These events likely created extraordinary dirt on our panels.

To clarify another reply, we do not sell back to the grid, we only buy when needed.

[J.P.M.]

I'd be interested in hearing (reading) what the panel cleaning does to restore the output. A 30or so %decline in output from fouling is pretty severe.

[chrisski]

Last year we had two volcanoes erupting, Moana Loa and Kilauea, spewing volcanic gasses and ash into the atmosphere. These events likely created extraordinary dirt on our panels.

Ash from the fires in California disrupted production. I imagine volcanic dust would to.

I deal with desert dust, and even three months without rain, cleaning the panels may make 1% more power, but makes me feel good about them.

[BenHawaii]

I'd be interested in hearing (reading) what the panel cleaning does to restore the output. A 30or so %decline in output from fouling is pretty severe.

I wonder if this could be a charge controller issue? We have two controllers and I have been looking at the log for one of them, and it's been showing minimal solar wattage since 3/12/2023. If you look back at the graph of 3/5/23 in the attachments, it typical of how the system was running up until 3/11/23. Yes, there has been some reduction in generation over the past year, but minimal until beginning 3/6/23. I'll see if I can find some more information on the Outback forum. In any case definitely need to schedule a service call and have all the components checked.

[J.P.M.]

I wonder if this could be a charge controller issue? We have two controllers and I have been looking at the log for one of them, and it's been showing minimal solar wattage since 3/12/2023. If you look back at the graph of 3/5/23 in the attachments, it typical of how the system was running up until 3/11/23. Yes, there has been some reduction in generation over the past year, but minimal until beginning 3/6/23. I'll see if I can find some more information on the Outback forum. In any case definitely need to schedule a service call and have all the components checked.

Understood.

Thank you.

[BenHawaii]

After running the system with no load for two days, allowing batteries to charge, I was advised to test the batteries at night. 7 out of 8 of the AGM batteries read 6.5 volts and one was 5.1 volts. So we seem to have a failing battery. They are supposed to be under warranty. I will follow up once resolved and also report on effect of panels being cleaned at end of the month.