Need help with power disruption

[krementz]

To me, it sounds  like a very quick voltage drop. Cause could  be turning old old fluorescent lights, water pump starting, microwave,  furnace or hvac startup. These  types  of load do NOT show up on regular multimeter or kilowatts meter - they are too brief. If the wiring from the battery is undersized or has loose connections, startup loads would definitely drop the voltage momentarily, even though  the static load looks fine. Some motors and electronics have startup loads 5 to 10 times the static

Clocks losing time usually means low frequency, which is usually cause by underpowered inverter or loads too high. ( same thig).

[FNG]

Clocks losing time is sort of normal on Magnums PAE. The PAE runs at 59.9Hz so the clock will run slightly slow. And if using the 6.0 or newer firmware and using custom battery mode this enables frequency shift that is designed to control the MicroGT. This starts shifting frequency down if the battery voltage exceeds the absorb setpoint by ~1 volt. So if the Classic was set for say 59.8 and the Magnum was set for 58.8 the PAE would drop the frequency more as it sees the battery as full.

[Matrix]

Absorb and EQ at 57.2 (2.38v /cell) for 3 years - Isn't that below gassing voltage?  ... Sounds like Sulfation to me.  At least as a starting point.   Has the OP never EQ'd the  batteries at the Rolls spec of 63.6v?  Seems if it has always been done at 57.2v this could be at least problem.

http://support.rollsbattery.com/support/solutions/articles/428-state-of-charge-charging-flooded-lead-acid-batteries

I would suggest a corrective EQ among all the other ideas noted  here.

http://support.rollsbattery.com/support/solutions/articles/430-corrective-equalization-instructions

[TrustinHim]

The system is on grid.
I have not seen any red LEDs light up on the remote during these quick surges.
The temperature on all the connections (battery, sub-panel, inverter, everywhere are between 65 - 68 F. (The room is 65 F.) Except for 3 GFCIs which are between 77 - 80 F. These GFCIs are new 20A with a pig tail for common, and only run three rooms drawing less than 4 amps total. Strange but I guess this is normal?

The batteries dive from 54 v to 49 after a load is put on them. They seem to stay at the 49 - 48 V range under load until the sun reaches high noon. There is enough juice from the panels to bring that number back up with just a moderate amount of sun.

I also have a fridge on this sub panel, and will disconnect during later experiments.

The SG was measured using a glass bulb type, during Absorb charge at 59.4 v 8 A.

Battery 1 (Neg side of the string) (all A's are on the neg side)
a. 1.205
b. 1.210
c. 1.220

Battery 2
a. 1.210
b. 1.220
c. 1.230

Battery 3
a. 1.240
b. 1.210
c. 1.215

Battery 4
a. 1.230
b. 1.215
c. 1.215

Battery 5
a. 1.215
b. 1.220
c. 1.225

Battery 6
a. 1.220
b. 1.225
c. 1.220

Battery 7
a. 1.210
b. 1.220
c. 1.235

Battery 8
a. 1.220
b. 1.230
c. 1.235

The last time I added water (good filtered) (not distilled I know) the level was just above the plates, not below. The level is now right at 1/4" below the vent tube.
The batteries hit low voltage disconnect twice in their lifetime, at 40 V, then spent most of their life sitting at about 54 V.
All battery cables are 4/0 awg crimped and soldered, except where it was pre wired to the inverter they are 2/0 awg.
The connections are tight, with only the tiniest amount of corrosion on one spot of 2 terminals. There is at least 1 sq. in. of clean contact area on these terminals, so there shouldn't be an issue there. (Although remove, clean and re-torque terminals is on the list.)

I am currently adjusting the settings as listed by Vic.

Thanks for the help!

Seriously, there is some expertise in here. Thanks for sharing.

Trust-

[Vic]

Hi Sparky,

Thank you very much for the SG readings.   Those SGs show,   at that point during Absorb,   that the SOC is approximately 75%

After your next full-charge,   suggest that you do an EQ for at least 90 minutes at about 62.5 volts,   as a starting point.  This is primarily due to the variations in SG readings between the cells.

If these batteries have never been equalized at a voltage above 62 V,   ever,   then it might take some time to reach the target EQ voltage.

Since you are Grid-connected,   you could try using the Magnum inverter to do some part of the EQ,   as,   you may not have enough PV power/time to get up to the EQ voltage (Veq)  after a full charge from PVs.

Is your system considered a Grid Backup system,   or do you run part of your home from PV/inverter  most/all of the time?

Because batteries are expensive,   and Distilled Water,   not so much,   if at all possible,   would suggest that you use a quality Distilled Water when adding to your nice battery bank.

All in all,   I would have guessed that your SG readings would have been considerably lower than your measurements.

If your battery bank sits on Float much of each day,   then your Vabs was not horribly low,   verses a system that is cycled almost every day ...

More later,   Thanks for the info,   Vic

[Vic]

AND,   Sparky,

Here is a good article from Surrette on a very good procedure for Measuring Specific Gravities:

http://support.rollsbattery.com/support/solutions/articles/4347-measuring-specific-gravity

I good glass tube/glass float Hydrometer  is a fine way to measure SGs  --   that is about all that I ever use.

Be certain to;   RINSE,   RINSE,    AND RINSE AGAIN WITH DISTILLED WATER,  AFTER EACH SG MEASURING SESSION.  This will help keep your Hydrometer accurate for a long time.   Without thorough rinsing,   a sticky substance will build up on the inside of the Hydro.  Air bubbles will stick to this residue,   causing inaccurate SG readings.

Please do let us know how your EQ goes.

YES,  please do remove the refer as a load,   to see if it might be one of the causes of your power drop-outs.

I know nothing of the Magnum Inverters,   BUT  they do have a reputation of being a bit slow in regulating AC output voltage ...

AND,   I have never seen any issue with clouding of the LCD display on the Classic MNGP control panel.   So cannot advise.

Later,   FWIW,   Vic

[FNG]

With that much voltage sag with that little load I would agree sulfation is an issue, Also the filtered water could be an issue as it will likely still have some minerals in it but nothing you can do about that at this point. Definitely, use distilled water in the future it is pretty cheap compared to new batteries.

So this is grid connected, Are you using the grid all the time and setting up the Drop AC function in the magnum? Basically what I am asking is, When this issue is happening is the grid physically present to the Magnum or do you manually disconnect it?

[TrustinHim]

This issue occurs only when the system has been disconnected from the grid and is running on the inverter. I don't notice any flicker from table lamps for example, but other appliances are more sensitive to the fluctuation. The modem and desktop computer however, resets and or goes onto an APC backup.

The battery bank is currently at 80 minutes of EQ. The battery voltage went to 62.8 V right away, and has since been slowly dropping to about 60 V at this time. The SG of each cell is rising and is averaging 1.240. The temperature of the cells is low, just about room temperature. Their is no risk of boiling over at this point, so I am hoping to reach 1.265 as the Rolls manual suggests. Then charge an additional 2 hours as recommended. (Batteries permitting.)

Thanks for the help!
It is much appreciated.

Sparky-

[Vic]

This issue occurs only when the system has been disconnected from the grid and is running on the inverter. I don't notice any flicker from table lamps for example, but other appliances are more sensitive to the fluctuation. The modem and desktop computer however, resets and or goes onto an APC backup.

The battery bank is currently at 80 minutes of EQ. The battery voltage went to 62.8 V right away, and has since been slowly dropping to about 60 V at this time. The SG of each cell is rising and is averaging 1.240. The temperature of the cells is low, just about room temperature. Their is no risk of boiling over at this point, so I am hoping to reach 1.265 as the Rolls manual suggests. Then charge an additional 2 hours as recommended. (Batteries permitting.)

Thanks for the help!
It is much appreciated.
Sparky-

Thanks Sparky for the update.

OK,  so you charged the batteries,   and then started the EQ.   Is the EQ being done by the Classic?

For now,  seems that you need to try to get the battery bank charged as much as possible,   during an extended Absorb.   Vabs about 60 V,   and set the Absorb timer to about 6 hours.

You might want to do this more than once  --  until the average SGs no longer rise,  during the Absorb.   This should define the maximum amount of charge that the battery bank will accept at that voltage.

Then do an EQ.  It is generally a good idea to limit the battery charging current during an EA to not more than 5% of 20-hour Capacity.   This would be about 21 Amps,  if your single string of S-550s have a C of about 420 Ah   ...   guessing,  for now.

If the EQ that you noted above was from the Classic,  then it is probable that the descent in Veq  would be due to too little PV power being available,   and the charge stage would show as,   EQ-MPPT.

If the batteries were quite Sulfated,   this could cause the immediate rise in the Veq to the target voltage.   It is possible that the descent in Veq could be attributed to  some sulfate being returned to the electrolyte,   and there being insufficient current available to maintain Veq due to the charge source not having enough current available.

Generally FLA batteries that were in need of a good charge,   and EQ  will emit a strong aroma of sulfur,   and this is a good thing  --  indicating that sulfur is being removed from the battery plates.

Charge stages that cause Gassing  raise the battery temperature.   Heavy gassing from EQs will cause a greater temperature rise,  for a given time period.   This heating  can cause increase the electrolyte temperature enough to cause some expansion.   This expansion plus aggressive gassing can cause some electrolyte to flow out of the cell caps.   People do call gassing,  "boiling",  but of course this not the actual case.

L-16 batteries,  like your S-550s  do not have a large space for electrolyte,  above the plates (sometimes called electrolyte reserve),   so,  if  some cells happened to have been over-filled,  there could be chance of this overflow.

Also,  it is very important that you use a Battery Temperature Sensor (BTS) on ALL charge sources.    Place then on a single battery at the center of the bank,   about half-way down the side of the battery (not on the top of a battery).

Enough for now,  please do let us know how you are doing.  Thanks,   Vic

[TrustinHim]

I believe the EQ has finished, but the EQ screen still says EQ started.
The timer view shows
2:00 absorb
:28 EQ
5:10 float

This was charged while the inverter was off, still connected to the grid. I assume this gave the Classic all the power it needed.
There was a lot of gassing for quite some time. Based on how the charge cycle went, it gassed heavily for over an hour.
Are those numbers above an accurate account of what happen?

I am looking at starting an absorb time of 6:00 hours, and will post a new batch of SG numbers.

Thanks for everything.  I will keep you posted.

Sparky-
 

[TrustinHim]

I did a manual EQ charge, but I don't see how to initiate an absorb charge. Do I need to set this through auto EQ?

Thanks,

Sparky-

[ClassicCrazy]

For the Classic you can use the Local Status app to force Absorb in the Basic Config settings

Larry

[Vic]

Hi Sparky,

OK,   In the previous post,  you said that the inverter was off,   and you were connected to the grid.

So,  the way I read what you said,  is that the only charging would come from the Classic,  which would only be able to use PV as a source of power.

I would go to the EQ screen,   and Stop the EQ.

If you have ARST set to ON,  the Classic will begin a new day at Midnight,   and a new charge will begin when there is enough solar to begin the Bulk.

You can manually start a new charge  by going to Tweaks>More>(force)Bulk.   This will start a new Bulk - Absorb - Float charge cycle.

(NO,  Auto EQ is different)

And,   on those Timer numbers   ...   is that the second display page?   The upper right Soft Key (rectangular button)  takes you to the second page,  where  the remaining time is shown.   Looks like the EQ still had 28 minutes remaining on the time set.   Am at an alternate location,  without a Classic to manipulate.   Am having a bit of trouble with the 2:00 Absorb and the 5:10 Float time.

    Later,   Vic

[FNG]

This issue occurs only when the system has been disconnected from the grid and is running on the inverter. I don't notice any flicker from table lamps for example, but other appliances are more sensitive to the fluctuation. The modem and desktop computer however, resets and or goes onto an APC backup.



Thanks for the help!
It is much appreciated.

Sparky-

Ok something doesn't make sense If the modem and stuff resets that would take a significant drop in AC output for a period of time, BUT if that was the case you would definitely see that in a lamp. I would make sure you have a lamp that's on the same circuit as the modem and see if it flickers. I wonder if there is a bad connection on one side of the 120/240 or maybe something wrong either in the inverter. Logic would tell us if it only does it when you are off grid it would have to be in the "inverting" portion of the PAE. If this issue doesn't resolve for you soon I would call Darren or Glenn in tech support and tell them this info.

[Matrix]

For now,  seems that you need to try to get the battery bank charged as much as possible,   during an extended Absorb.   Vabs about 60 V (or 2.5v /cell),   and set the Absorb timer to about 6 hours.

You might want to do this more than once  --  until the average SGs no longer rise,  during the Absorb.   This should define the maximum amount of charge that the battery bank will accept at that voltage.

Curious Vic,
Are you saying that SG would be higher at a higher voltage?   And that even though you could get a higher SG at a higher EQ voltage,  that a fully charged battery SG would be the SG number that is arrived at when charging for 6 hours at this lower absorb voltage (2.5v /cell)?

I had believed one should charge fully,  then EQ at 2.7v /cell until SG stops rising and that this SG number would be the SG that would indicate fully charged and be the fully charged target SG.  Perhaps I am mistaken.   You seem to indicate that an EQ SG would be higher than fully absorb charged SG and that this lower absorb SG would be the target SG for fully charged. 

At what point do you consider a battery fully charged?  Is the SG that is achieved using higher EQ voltages actually an over charge?  Interested to hear what your thoughts are on this.

Mike