Bulk Absorption Chg Finish Selltogrid

pcguy2u
pcguy2u Solar Expert Posts: 151 ✭✭✭
All of a sudden my system started going into an infinite loop beginning with the sunrise yesterday morning. Up until that point, when the sun came up, it started selling and the watts being generated slowly showed an increase on the SCP.

So now, for the last two days and as I write this, the system is going through the motions of Bulk - Absorption - Chg Finish - Selltogrid and back to bulk. There are no error messages.

Could a setting in the system have changed inadvertently, or what is really going on here? Yesterday, when I noticed the problem at about 7:00AM, I reset (powered down) the system and initially there was no improvement, but it seemed that after the sun came up and while I wasn't looking, everything started working normally.

Now today, same thing as the start of yesterday - have not reset the system yet and hope this thing will work itself out.

Checked the battery voltage levels and they are all identical at 12.69 volts.

The system is three years old, and aside from the F49 problem that occurs from time to time, has been funtioning normally.

Here are the specifics:
Grid Tied Xantrex XW6048, 32 KC 200's totaling 5000w (after deducting efficiency factor) of PV, 2 MPPT-60 charge controllers, Trimetric meter with System control panel and four Concorde 12V AGM 258AMP batteries in series..............

Kind regards,
Nick
«134

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    I would use a voltmeter (and current clamp meter if you have one) to see if you can find a bad electrical connection somewhere... Just good enough to start the hardware up--but when it starts loading down, the voltage drops and resets everything.

    Could be on the array side, or in the DC wiring. Don't know.

    Just a guess/starting point.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Forgot to mention that yesterday, I shortened the absorption duration to 10 minutes from 2 hours in order to allow the process to complete the cycle more quickly.

    Also, the sell to grid portion of the loop lasts for only 10 or so seconds and the whole loop takes about 25 minutes.
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    BB. wrote: »
    I would use a voltmeter (and current clamp meter if you have one) to see if you can find a bad electrical connection somewhere... Just good enough to start the hardware up--but when it starts loading down, the voltage drops and resets everything.

    Could be on the array side, or in the DC wiring. Don't know.

    Just a guess/starting point.

    -Bill

    Hi Bill, just check the array outputs from the two charge controllers - normally they are somewhat different at this time of day because Array 1 (controller 1) gets more sun earlier than Array 2 (controller 2).

    At the moment it just turned foggy (was sunny a few minutes ago) which means both array should be similar - they are not. Array 1 is battery output is varying wildly at a level as low as 5% of Array 2 when they should be producing approximately equaly output. The harvest shows .4Ah @ .3KW vs. .7Ah .4KW
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    I just noticed that during the 11 seconds that sell begins, the inverter shows the AC1 watts to be 6100W which is impossible, especially with the conditions and time of day.
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    So I removed the distribution panel covers, I'm checking for loose connections, taking voltage readings, comparing the readings to the MPPT controller displays and while I'm doing that the 6048 continues to cycle.

    Then at 9:30 (still foggy out) it starts to sell again and I start watching the watts output. Many seconds go by and it settles at 2.5KW (pretty normal for a bright foggy day) and it is now selling with no problems???

    This is apparently the same thing that happened yesterday, except that yesterday I turned the system off and back on, and today, I didn't do anything except remove the covers and check the tightness of a few screws - none were loose.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    I would focus on that array that is "flaky"... It could be a bad electrical connection on either side.

    Try using a volt meter and see what the voltage looks like. If it is not relatively close to Vmp-array... Low voltage--I would suspect the array has a bad connection (or bad panel). If the Vmp-array is stead to higher than normal--then I would look at the DC output side to the battery bank.

    Watching the power generation on the MPPT controller may be confusing during debugging (unless you do the watch the meter while shaking each electrical connection).

    Just some more guessing on my part.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Joe_B
    Joe_B Solar Expert Posts: 318 ✭✭✭✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    pcguy2u wrote: »
    Hi Bill, just check the array outputs from the two charge controllers - normally they are somewhat different at this time of day because Array 1 (controller 1) gets more sun earlier than Array 2 (controller 2).

    At the moment it just turned foggy (was sunny a few minutes ago) which means both array should be similar - they are not. Array 1 is battery output is varying wildly at a level as low as 5% of Array 2 when they should be producing approximately equaly output. The harvest shows .4Ah @ .3KW vs. .7Ah .4KW

    I run 2 XW SCC's and I sometimes see the kind of behavior that you are seeing. The way my system works is MPPT 0 takes the load and MPPT 1 comes on line only when the system load demands it.

    That being said, there are certain illumination conditions and load levels that can cause the controllers to become unstable and cycle as you describe in your OP. I notice it most on cloudy days where the harvest is about half of the max, SCC 0 will run about 800 watts and SCC 1 will cycle from 0 to about 6-800 watts. It should stabilize in full sun, at least mine does.
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    Joe_B wrote: »
    I run 2 XW SCC's and I sometimes see the kind of behavior that you are seeing. The way my system works is MPPT 0 takes the load and MPPT 1 comes on line only when the system load demands it.

    That being said, there are certain illumination conditions and load levels that can cause the controllers to become unstable and cycle as you describe in your OP. I notice it most on cloudy days where the harvest is about half of the max, SCC 0 will run about 800 watts and SCC 1 will cycle from 0 to about 6-800 watts. It should stabilize in full sun, at least mine does.

    In this case, both controllers are active at all times and it was sunny on both of the last two days. On both days, the system corrected itself at about 9:30AM.

    Go Figure??????????
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    This morning, sunny again, the system is acting normally. It started selling at about 6:30AM and has been operating normally.

    Like Bill suggested, I suspect that the differences between the two arrays/controllers are at the base of the problem. This morning, the outputs are what I expect to see normally with Array 1/Controller 1 producing somewhat more than Array 2.

    In the late afternoon it is just the opposite.

    Can anyone explain the logic of the array output causing the symptoms I'm seeing?

    TIA
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Rain this morning - same problem after the sun-up and automatically fixed itself (while it's still raining) at 10:30AM. Go figure.

    Will continue to report on this until I figure out what's going on. Will report the fix, if and when I find it.

    :roll:
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    More info this morning. Checked both arrays at the combiner boxes and opened the breakers with all inputs on array 1 being very close - like within .2 Volts. I neglected to test array 2 with the breakers open, but it wasn't the array that showed fluctuations early on.

    Found a couple of set screws not totally tight - not likely the problem.

    Also found that the entire cycle from start to finish and back to start is almost exactly 5 minutes, very much like the cycle is set base upon a memory variable???

    Another thing that I noticed is that with the inverter connect to the batteries, they measure 57.8 V and with the battery breaker turned of they measure 53.6 V, even when the system is not selling or in ACPassThru.

    I reset the system about 10 minutes ago and it's going thru the startup process at this moment.

    Just started selling and the same thing occurred.

    Any more ideas?:confused:
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Ok, one more bit of info/test results. Turned each of the charge controllers onto standby mode, one at a time. No change - each time same results?

    Does this point to a comunications problem between the controllers and the inverter????
  • Joe_B
    Joe_B Solar Expert Posts: 318 ✭✭✭✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    I wonder if this has something to do with an AI fault??? If an AI fault occurs, it will disconnect for 300 seconds (5 minutes)!
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Just back from being gone for half an hour and the system is now selling again:confused::confused::confused:
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    Joe_B wrote: »
    I wonder if this has something to do with an AI fault??? If an AI fault occurs, it will disconnect for 300 seconds (5 minutes)!

    That would be convenient if it turned out that way, but there is no fault message that appears in the system? The last fault message was for Over Frequency on 6/21/11.
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    I offer the following which is a complaint to the Calif PUC, regarding the power in this neighborhood that PG&E won't do anything about. The only reason I offer this is because of the timing of all the events. I filed the complaint on 6/21 and the problem started on 6/24 (a Friday).


    I have had an ongoing complaint with PG&E over the fluctuations in the voltage "FREQUENCY" (not voltage level) that occur regularly on our service here. It occurs generally at 8:03AM and later in the afternoon on any day of the week and usually twice a week or more. The problem is not localized to my service and is experienced by others in the neighborhood including one person who has evidence of the problem as I do. I explained more than once to a PG&E engineer that I could demonstrate the problem via my Fluke meter, my APC battery backup and my inverter fault log. His reaction was that I should have my equipment checked - really? All three pieces and my neighbor's as well? The neighbor who also experiences this is on a medical rate with PGE due to the need for a breathing machine (bi-pap). She says she sees the problem almost every day in the afternoon and as often as I do in the morning. The problem causes computers and inverter to go down, some for extended periods.

    The engineer said that the problem I described just can not happen and that the problem must be isolated to my home and being caused by something in my home or that my equipment was defective. I guess he thinks we are all imagining this. He also says they have no way of checking this kind of problem – how would I be able to trap the existence of the problem with my limited budget and yet PG&E would not have similar or more extensive resources available? I basically got kissed off, not once but twice previously when I got to the end of my rope with this situation, just as I am now.

    I would hope that you folks have enough influence to cause PG&E to fix this problem – getting someone, who appears to be retired on the job, to take action seems to be more than I am able to accomplish. Finding the source of the issue will probably be the hard part. The fix should be relatively simple.
  • niel
    niel Solar Expert Posts: 10,300 ✭✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    maybe you could hire an electrician and explain to him what you are seeing and if he can verify what the ongoing problem actually is and with proof? my opinion is that they know you are right, but want you to prove it before they are forced to do anything about it. typical.:grr
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Recieved a reply from the PUC today - it indicates that they have assigned a case number to the complaint and have given PG&E 30 days to respond.

    At the moment the inverter is selling with no problems. I'll be looking at the frequency and voltage levels during the next instance of this problem. It never occurred to me that I should do that previously because there have been no Fault messages.

    Even if this proves to be a power problem, it appears that the inverter is also failing by not reporting the isssue properly, ie, no Fault warning.

    More to follow.............
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Some of you may recall one of the other threads I started in Sept/09 - it's been going on for some time now with no particular resolution. My guess is that all of this is going to come to a head soon with part of the problem being Xantrex firmware and the other being PG&E and/or one of it's commercial customers.

    Here's the other thread: Xantrex XW6048 - fault F49 - DC over voltage

    Just to be clear, there appears to be three issues that may all be associated:
    The issue in this thread that doesn't appear to show any fault message.
    The F49 issue in thread above.
    The ongoing saga with PG&E regarding F25 - this one is often followed up with F49 and usually occurs at 8:03AM

    Yesterday, I took the time to organize all my notes on the inverter/power issues into a spread sheet - included Date, Time, Day of the week and Fault #. Well it turns out that all the issues may be related??? Up until yesterday I never associated the F49 issue with the one in this thread, but after seeing the pattern of dates (in particular, day of the week) and times of the day, and it now appears that there is some association.

    All but one of the problems with F49 (DC over voltage) and F25 (Over Frequency) have occurred on a Monday, Tuesday or Wednesday and most of both have occurred at 8:03AM and at some time in late afternoon when the neighbor, who is also aware of this problem, also experiences the anomaly.

    The most curious thing about the issue in this thread, specifically, is the fact that there is no Fault message - otherwise the symptoms are very similar to F49. I have my fluke meter sitting on my desk and will trap voltage and frequency extremes during any new event. In any case, this morning, a Thursday, I was observing my Fluke and trapping MIN/MAX AC frequency and the high freq came up at 63.97 - the default setting on the inverter at which a fault message should/would be issued is 55min and 65max. In either case, the PG&E engineer told me that such a fluctuation could not happen????

    I do expect to hear from PG&E in the near future and I am now gearing up for that encounter.

    BTW, it's sunny this morning and the inverter is working fine....
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    Joe_B wrote: »
    I wonder if this has something to do with an AI fault??? If an AI fault occurs, it will disconnect for 300 seconds (5 minutes)!

    Hey Joe, you may have hit it on the head with the above. It looks like there may be a phantom fault. No entries of any kind in any of the fault logs, also.
  • Joe_B
    Joe_B Solar Expert Posts: 318 ✭✭✭✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid
    pcguy2u wrote: »
    Hey Joe, you may have hit it on the head with the above. It looks like there may be a phantom fault. No entries of any kind in any of the fault logs, also.

    Welcome to the wonderful world of hack inverter firmware. I have been down this road a few times.
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Here are my notes from this mornings episode - a Thursday with a dry overcast/fog.

    At 6:30AM got the phantom Bulk - Absorption - ABS Finish - Chg Finish cycle for a short time and then SellToGrid. No inverter fault messages.

    Trapped a freq reading of 71.22 and F25 on the inverter at 8:03AM - after the 8:03 fault, got Bulk - Absorption - ABS Finish - Chg Finish - and the cycle repeated for one half hour, and then SellToGrid at 8:33am. Continued to monitor with Fluke meter and found no further incidents of High Freq nor were there any more F25 faults on the inverter.

    I'll continue to post these episodes until a true pattern develops or until a solution is found.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    If you are logging ~70 Hz for Utility Grid Frequency--My first guess is that there is "noise" on the line (arcs, heavy switching of loads, etc.).

    A few decades ago, we had problems with kids throwing metal objects on the local power lines for the "fireworks display".

    Besides the obvious problems... This also would nick/weaken the power lines themselves from the arcing.

    I don't think you would have had a true 70 Hz power variation--that would usually be "next to impossible" (as being discussed in another thread here) without the utility, and the region, having massive issues.

    Would be neat to get a digital storage scope on the AC line with a post trigger from your frequency meter... But not cheap or easy.

    Power company may not even have the equipment/knowledge to do the capture locally...

    I would be tempted to get one of these for $200 (2 analog, 2 digital channels):

    http://www.seeedstudio.com/depot/dso-quad-4-channel-digital-storage-oscilloscope-p-736.html?cPath=174

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Hi Bill, well, the 70+ was on the high side of the problem, but I have recorded frequencies well into the upper 60's over the last couple of years. BTW, these issues primarily happen at 8:03AM on days the work week and never on holidays.

    It seems clear to me that something is on a timer and that timer is manually reset by someone when DST comes and goes. I have trapped the problem at precisely one hour on either side of 8:03AM when on the cusp of the change in times.

    Confused as to what I might find with the reference piece of equipment above that I don't get with my Fluke 115. Granted I can only see Freq or Voltage fluctuations at any given moment depending on what I set it up for. But I am able to trap what the inverter is also seeing.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    Mainly--Get a "Picture" of the wave form that may be causing the problems.

    Grid Frequency is not going to change very "fast" or very much (0.5 Hz or so +/- typically). Seeing a 10 Hz change would seem to me (a power company outsider) to be almost impossible if your is from a region network (as opposed to an industrial genset three blocks down the road in Alaska).

    Typically frequency counters are just looking for "zero transitions" or peak detection... If you have some arcs that took out or add "peaks" -- then a frequency counter would read high frequency because of the noise on the waveform--when what you have is "dirty power" on top of a steady 60 Hz grid power (injected by a local user/unusual event).

    So, if you call the power company and tell them their grid is running at 65-70+ Hz--they will tell you it is "impossible", none of their logs show the grid running at 70Hz, and that will be the end of the call.

    However, if you can show them a picture of the grid power at that point you are having problems--then they cannot deny that something is happening... Especially if you can set the digital scope up at a neighbor's and show the same type of event.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    One thing that I have never accomplished is trapping this problem with the inverter powered off. What are the chances that the inverter is somehow involved in this problem?

    Nothing here in the house is on a timer except for a couple of recirc pumps - one AC and one on a DC wall wart, neither is set to any time near the 8:03 timeframe.

    Doubt very much it's here or in the inverter, but I would be totally embarrassed if it turned out that way.

    The local water company has a plant that is not so far away - I think we've discussed that in another thread. But they have nothing on a timer - it's all dynamic and does what it needs to do based upon usage and water availability.

    That's the only industrial type faciltiy with 4-5 miles.........

    I'll be very interested to hear the feedback from PG&E to the PUC complaint I made last month. They indicated that it would take 30 days, so maybe sometime before the end of the month.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    Perhaps Solar Guppy can offer some insight here...

    A grid tied inverter basically feeds current as it follows the AC wave form... It is possible that a "hashed up" AC Sine Wave may cause the inverter to "amplify" the hash. A lot would depend on the internal control loops and their parameters.

    My guess is that the current output would respond fairly slowly--I.e., to the 60 Hz "filtered" AC sine wave--and much of the "high frequency hash" would be filtered out--So--I would not have expected the GT Inverter to "make things worse".

    If it detects a non-sine waveform on the AC input--the Inverter should shutdown within 5 cycles (again, guessing from what little specs. I have seen).

    Could, during the 5 cycles, a GT Inverter make the AC Grid "noise" problem worse--maybe...

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Joe_B
    Joe_B Solar Expert Posts: 318 ✭✭✭✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Grid power is pretty crappy (at least in my location) I have seet the over frequency issue a bunch of times. I can hear RFI on my ham radio from 11 MHz to about 22 MHz. I have had the power company out here several times and have let them see the noise on my spectrum analyzer (fortunately mine is pretty constant). Its classic power line stuff, when it gets wet, it goes away and when dry...... bad RFI. They went up the poles and tightened things up for me and it fixed it for about a month. But it never really got fixed and it still raises hell with my radio and still gives me AI over frequency faults.

    I would speak with those tech support guys at Schneider about it. If you get the right guy, they usually fess up that they have a problem. Then they will magically come up with some software that will make things better. After talking with them a LOT, I am of the opinion that tere are a lot of problems with the grid software and if you hit a nerve with them they will acknowlege it.

    I dont sell but I do see the same cycling behavior with my system when the RFI is bad so its not a big issue for me. But, the fact that it always happens to yours when its foggy might be a clue.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Re: Bulk Absorption Chg Finish Selltogrid

    I used to live right on the coast--And we had wet salty fogs that just raised heck with phone lines and arcing insulators (as well as rusting out 1970's cars/pickups in a couple years--great place for the car radiator repair/replacement business). As well as eucalyptus trees that would shed bark during storms on to power lines.

    Is that a Bodega Bay Ca. problem? Don't know.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • pcguy2u
    pcguy2u Solar Expert Posts: 151 ✭✭✭
    Re: Bulk Absorption Chg Finish Selltogrid

    Here are some of the notes in my log of events - there doesn't appear to be any association with the weather. Also, it is likely that the F25 fault, or the phantom fault, occurred in conjunction with the F49's. At the time I made notes, I was not conscious of the coincidence. Phantom meaning the inverter went from SellToGrid to the cycling of Bulk, Absorption, ABS Finish and Chg Finish but never issued a fault and eventually went back to SellToGrid without manual intervention. The latter is what happened early this morning followed by F25 at 8:03AM.

    F25, F49 &F70 suunny
    F25 & F49 foggy
    F49 foggy
    F49 foggy
    F49 sunny
    F49 overcast
    F25 & F49 sunny
    F49 sunny
    F49 sunny
    F25 & F49 sunny
    phantom sunny
    phantom sunny
    phantom raining
    phantom raining
    Hi Freq trapped sunny
    phantom dry overcast/fog
    F25 dry overcast/fog