Low SG readings when voltage is still above 50%...

[Mr Vee]

Hello there all.

Before I start I shall give the details of my system.

4 x 240w Bosch panels
Outback FLEXmax60 MPPT Charge Controller
Outback VFX 3024 inverter
4 x Rolls s530 6v Batteries run is series at 24v

Panels receive full sun from 9am until 6pm in summer and 10am until 3.30 ish in winter

System was installed in mid July 2012

I have recently purchased a hydrometer to monitor specific gravity levels on my battery bank as after a long winter and a very wet spring the system had been somewhat under powered. The readings were around 1.22-1.23. I ran an eq cycle on the charge controller at 31.4v over a couple of sunny days and was able to get this up to 1.25-1.255 before I ran out of sun. The system seemed to be performing better until we had a small run of bad weather and the voltage ran low (23.0v). After this it seemed that the battery bank did not want to charge. A full day of sun with 850-940 watts consistent during peak hours and yet it would not rise above 26.2v.

I took a reading with the hydrometer and it was below the scale (the scale being 1.15-1.30) and not just by a bit but by a lot. The top of the glass tube was bobbing in the water. At this I panicked and ran out and bought gas for the generator ( a rather inefficient 6kva petroleum that pushes out 2000w charge and up to another 2000w for ac loads) and ran an eq cycle at 32.0 volts over the day until the readings were 1.26-1.265 across all cells. This is what rolls recommends.

The low battery cut off on the inverter is set at 24.0 volts and the full charge cycle has been completed several times since this (2 hour absorb at 29.4) and hydrometer reading were showing 1.23 in the morning with a voltage of 24.7 easily rising to 1.25 and above with the absorb cycle.

After another couple of cloudy days the system is down to about 24.0 volts last thing at night but has not gone so low as for the cut off to occur.
Yesterday I took a hydrometer reading while the the voltage was above 24 volts and the sun was not shining and my reading was 1.15-1.16 across all cells.

This has confused me as I thought the system was working better now and yet my SG reading indicate that my batteries are way below 50% SOC even though the voltage is above 24 volt. This SG has remained consistent throughout the day even though levels have gone up to 26volts despite the lack of sun. If it was not for the hydrometer I would not know that I have a problem.

What can I do? Should I run another eq and continue until all cells read 1.26-1.27 (I was only short by a hare's breath). Could I have a faulty cell or are my batteries shot after one winter?

I wish to expand the system in the near future doubling the bank and adding more panels to match. I am worried that if these batteries are now compromised I will not be able to see the full potential of any new batteries that I add.

Sorry to be so long winded but I wanted to answer the obvious questions before they are asked.

Thanks in advance.

V

[Sunking]

First thing is forget using State of Charge voltages to determine the state of charge as it does not work on a working system. SOC voltages are only somewhat useful on a fully rested battery that has been disconnected. The only way to measure the true SOC is with a temperature corrected hydrometer.

So if you are seeing low SPG readings means you are over discharging your batteries severely well beyond 100% DOD as 0% SOC SPG is 1.110 to 1.130. To get the most cycle life out of the batteries you should never go below 20% DOD or 1.215 to 1.235.

What you should do is get the SPG back up to where it is suppose to be, and keep it there. The root cause of your problem is you are using more power in a day than your panels will generate period. If you have a bad cell it would show itself with both SPG and voltage readings.

[Mr Vee]

Thanks a lot for your reply.

I forgot to specify my loads on the system which are as follows.

800w water bore hole pump - runs for 15 minutes a day
150 watt fridge on lowest setting
2 laptops
5 led lights

On sunny days the water is used more and we have a washing machine that does a cold wash

On cloudy days the fridge goes off

Our energy usage is low. Most of my electrical equipment is in storage until the system can cope with it.

The charge controller claims to input 5 - 6.5 KW Hours on a sunny day and 1.5 - 2.7 KW Hours on a cloudy day on average.

How can I stop the system draining so much of the battery? The inverter only has the 24.0 volt cutoff which goes no higher.

So are these batteries shot? It seems that we spend thousands of watts of solar energy all day to charge a bank that is flat in a day of no sun.

[inetdog]

So are these batteries shot? It seems that we spend thousands of watts of solar energy all day to charge a bank that is flat in a day of no sun.

If batteries spend to long a time at a low SOC they will sulphate (a formation of insoluble lead sulphate on one plate). This will reduce the effective AH capacity of the battery and will also reduce the SG, even when fully charged.

if your batteries are sulfated, they will not be able to supply as much energy during the night without going to a low SOC. They will also not be able to accept as much Bulk charge from the panels during the day before switching to Absorb or Float. The result is that you just have less capacity to work with and they will charge at a slower average amp rate.

The only way to tell for sure what condition your batteries are in is to do a controlled load test down to a fairly low SOC (as determined from the SG, not the voltage) and calculate the effective AH capacity. Your chronic undercharging may have prematurely killed your batteries.

[Mr Vee]

I thought that running an eq cycle until the SPG reaches 1.265 would deal with the sulphation or at least improve it. These batteries are only 9 months old.

Rolls suggest corrective equalisation for heavily sulphated batteries. The manual says to to set the eq voltage as high as the charger will allow and running at, whilst keeping an eye on battery temperature until the specified SPG of 1.266 is reached. The high charge will make the thick plates of the rolls battery dispel the sulphur build up al send it back in to the electrolyte My charger voltage can go above 48 volts. Surely it is not wise to raise the voltage so high? How high can I go with the voltage as I do not wish to destroy the batteries?

Also I have already reached 1.25-1.265 across all cells. This would indicate that I already have them in a reasonable state but as I am new to all of this perhaps I do not understand.

Before I ran the first eq the circumstances were as you describe. Absorb voltage was reached in a very short time and the capacity was reduced.

Since eqing it is taking longer for the absorb voltage to be reached and the capacity seems to be improved but after a couple of days of cloud they are discharging to a SPG of 1.15-1.16 which then takes much much longer to charge up to absorb voltage on the next sunny day. So it is discharging nearly 100% within before reaching 24volts. The inverter can not be set higher than 24 volts for battery cutoff. How else can I prevent the SOC going so low?

If I am nearly 100% discharged by 24 volts what if I lowered the cut off to 23, or 22 or even 20? Am I in to the minuses then? When the system has been allowed to go lower than 24 volts the charge is held for a long time. It would take much much longer to go from 24 to 23 than from 25 to 24.

Do I need to live with a hydrometer in my hand so I can know the state of the batteries?

[inetdog]

Do I need to live with a hydrometer in my hand so I can know the state of the batteries?

Yes, if you want accurate SOC indications. The saving grace is that you only have to check a couple of pilot cells if you can assume that the cells are well balanced. Since you have a single series string you only have to check balance infrequently.

But as a practical matter, once you get your batteries stable you can measure both voltage and current (charging or discharging) and make up a table which relates the voltage/current pair to the SOC as determined from the SG. This will be unique to your batteries and your installation, and will change with time as the batteries age and degrade.

The SOC also cannot be determined from just looking at total current in and out as some battery monitors do, since the capacity of the battery will go down with time.

[Mr Vee]

So is adding another battery bank a worthwhile idea?

When the sun shines consistently we complete absorb by 2.30 pm, sometimes earlier.A larger battery could be charged with the additional hours of sun as our energy usage would remain the same, providing less of a strain on the batteries as the size of the bank would be higher AH. Was also considering 2 more panels to accompany this to make up the extra power needed to fully charge the bank.

Is it worthwhile with the current bank in the current state?

[Sunking]

So is adding another battery bank a worthwhile idea?

No. Not at this time For now my best guess is you are using more power then the panels can supply. The clue is you said it was fine last summer, but failed in winter. Your story does not match up. If the batteries go to float by noon, there should be no problem with SPG. Are you checking SPG when it reaches FLOAT. That would tell us exactly what is happening. For example if at say noon the charger switches to float and you have low SPG tells me immediately you have sulfated batteries. If you reach Float by noon and th eSPG is good at that moment but low in the morning tells me you are using too much power at night. We need to nail things down.

You also need to quit applying the EQ charge as EQ charge does do irreversible damage to the battery in the form of positive plate corrosion. The only time you apply an EQ charge is when the cells become unbalanced or need to be recovered from DEFICIT CHARGING.

Again the problem appears to be you use more power than your panels can replace. In other words undersized. Hard to say at this time without more data.

You batteries are very good batteries. With a single string @ 24 volts @ 400 AH are capable of delivering roughly 2 Kwh per day. The panel wattage you have is appropriate for a 24 volt 400 AH battery. At full 960 watts should deliver a charge current of roughly 37 to 40 amps or a solid C/10 charge rate. However that does not answer how much power you use in a day. I have no idea.

What is your approx location so I can determine how much power your panels are capable of producing in a single day?

What I am beginning to suspect is you have a problem with either the panels or charge controller. Can you measure the charge current on a bright sunny day. Is it up around 35 to 40 amps? It should be. If it is not a panel or CC problem, then you are just simply using more power than your system can provide, and will need to be upgraded.

So do some homework and try to answer my questions so we can get to the core of the problem and get you fixed up OK? Still too many unanswered questions to make a determination.

[Mr Vee]

Power usage as follows

800w water bore hole pump - runs for 15 minutes a day - Daily usage .2kwh
150 watt fridge on lowest setting, watt meter said it uses 80 watts for 10-15 minutes every hour and 7 watts the rest of the day - Daily usage .38kwh
2 laptops 40w x 12 (hours) x 2 = daily usage .48 kwh
5 led lights daily usage .21kwh
Inverter .035kw x 24 daily usage .84 kwh

total daily usage estimate 2.01kwh

The laptop numbers are largely overestimated as is I feel the inverter load on the system but maybe not? How much should I deduct for the inverter.

When the sun doesn't shine I switch off the fridge, only to run once absorb voltage is reached.

Also I am reaching float by 2.30pm ish on the second day of sun (first day brings batteries back up to absorb but does not reach float)

At 800 watts the charge controller was saying 29 amps. On a sunny day I can get over 6 kwh. My location is central Portugal, level with Madrid. Full sun on the panels from 9am to 5.30pm at the moment when it shines. The panels face due south, are positioned for the sun's position on the equinox and are not on a tracker.

Thanks a lot for all of you help. It is very much appreciated by a somewhat worried offgridder.

[Sunking]

Also I am reaching float by 2.30pm ish on the second day of sun (first day brings batteries back up to absorb but does not reach float)

That tells you never reached 100% and are in deficit. With your location should be no problem providing 2 usable Kwh/day at this time of year. If you are certain the panels and controllers are operating correctly, means you are using a lot more power than you think.

I can certainly understand why your system failed in October (3.2) November (2.16), December (1.67) , January (2.02), February (2.82), and early March (4.2) with only 960 watts of panel power. This is where your system was mis designed. To generate 2 Kwh/day minimum year round you need a 1800 watt system and a 80 amp charge controller for the winter months. Nearly two times more panel wattage than you have now.

I can only conclude you have two issues:

1. You are using more than 2 Kwh/day
2. System is undersized both in panel wattage and battery capacity.

You have a very unusual solar insolation pattern in Portugal. Sun Hours I rarely ever run across. In winter you dip all the way down to 1.67 Sun Hours which is what th esystem has to be designed for as it is worse case. 1.67 is very little sun. No wwhat is unique is is Summer the months of June and July you peak at 7 Sun Hours or roughly 4.3:1 ratio which is unheard of. What that means is in winter with a 960 watt panel you can only generate roughly 1 Kwh/day, and in Summer generate as much as 4.3 Kwh/day. I have no idea how you are calculating the panel is generating 6 Kwh/day as it is not possible.

FWIW my data is using Coimbra Portagal Solar data.

It is not all terrible news because right now you got Summer to get things figured out. You need to really nail down your daily Kwh usage as I suspect you are using more than you think. During this month in your area with your panel wattage you can generate up to 3.8 Kwh/day of usable power. Double check the Flexmax 60 Controller and set Bulk/Absorb to 29.4 volts and Float to 26.8 volts. Monitor the SPG daily and make sure it reaches 1.265-1.275 when the system switches to FLOAT, otherwise start the generator and set to 29 volts and charge til it gets to the correct SPG. Quit using EQ voltages for now.

Once you get a handle on daily usage, then we can determine the correct panel wattage, bu tI can tell you right now more panels and a new Charge Controller are going to be required. What I am affraid of is if you go beyond needing more than 2000 panel watts, requires moving up to a 48 volt battery system, and that means a new inverter and additional batteries. At a minimum you are looking at a new 80 amp controller and up to 800 more watts of panels. Your controller is only good up to 1600 watts @ 24 volt battery, or 3200 watts @ 48 volt battery.

[Mr Vee]

Hello there. Thankyou very much for you detailed reply and for taking the time to research sun insolation in my area to gain an understanding of what power I am to expect from the sun.

I had a quick look at the site you used to be honest it did not make total sense and the data I collect myself here on the ground does not seem to match and I am akin to trusting what I can see with my own eyes.

in Summer generate as much as 4.3 Kwh/day. I have no idea how you are calculating the panel is generating 6 Kwh/day as it is not possible.

At 7pm sunrise the cc wakes up and gives a reading of 0.020 Kw.I stand by my panels in the morning and watch the shadow of the neighbouring trees pass across my panels at around eight am. The cc is reading 0.050 kw rising to 0.150 as it approaches 9 am. After 9am the last shadows leave the panels and it gets full sun exposure. The cc reads 0.350 Kw quickly reaching 0.500 - 0.650 Kw by 10am. By 11 we are at 0.800-0.850 Kw reaching 0.900-0.950 over the hours of 1pm-2pm. Falling a little to 0.800 around 3pm. 0.500 by 4pm. 0.350 at 5pm. 0.200 by 6 until neighbouring trees give shade and it drops to 0.150-0.050 and by 8 it is sleeping.

The sum of the average Kw hours equates to more than 6.0 Kw.

Here is the log from the cc from 15 days ago where it states exactly 6.0 Kw

15 days ago.jpg

Here are the logs from the days on each side of that day.

16 days ago.jpg14 days ago.jpg

In fact anything less than 4.0 Kw h I consider to be a below average day and would most likely have been partially cloudy. I have watched this system closely as I have tried to learn it's patterns so now I know roughy how many Kw the panels are pushing by looking up at the sky.

The winter numbers are even more different.

What that means is in winter with a 960 watt panel you can only generate roughly 1 Kwh/day

Here is a picture of the log from 128 days ago from today (10-May-2013) which I guess to be the first week of January. This is pretty much mid winter. I wish I could go back further but 128 days is the limit.

128 days ago.jpg

4.2 Kw h? And 127 days ago we have 2.2 Kw h

127 days ago.jpg

4.0 Kw h has not been that rare in winter but on some days we have gone as low as 0.03 Kw h.

Here's my favourite. From 56 days ago which is around mid march

56 days ago.jpg

An all time high for me of 6.2.

All of this makes sense to me so I feel I should assess my solar potential from real world data such as this, collected from on site at the panel location.

So taking the above numbers in to consideration perhaps a slightly different assessment is required.

I can add 2 more panels to the FM60 raising me to 1440 peak and essentially adding around 50% to the above figures. Surely this is enough to fill the batteries?

Obviously a run of bad weather giving me anything below 2.0 Kw hours will require the generator to bring things back up.

What I am affraid of is if you go beyond needing more than 2000 panel watts, requires moving up to a 48 volt battery system, and that means a new inverter and additional batteries.

I was of the understanding that I could run another FM60 with potentially another 6 panels giving a further 1440 giving me 2880 watts peak. To expand the batteries I would add another string or 2 in parallel bringing me up to 1200 amp hours. I know 3 strings in parallel is not the ideal situation but I am limited by finances. I was never informed I would need a new inverter or to bring things up to 48 volts nor have I seen in the outback documentation anything that specifies this. From the manual for the VFX3024 i got this

VFX3024.jpg

So 150 amps maximum would allow 2 FM60s to be run together (120 amps max) but not 2 FM80s (160 amps max) so buying an FM80 leaves my FM60 redundant (rather an expensive doorstopper) and means I cannot expand any further without another inverter, again too much money for me.

This is all rather hypothetical at the moment as adding this many panels is a fantasy for me at the moment anyway.

Today I used the genny to bring the batteries up to absorb voltage 29.4 but as soon as I switched off it dropped back to 27.5. The SPG was 1.17. At the start of the day it was 1.15. As I write it is sat at 28.4 and the genny was turned off almost 2 hours ago so we are still awaiting absorb and I guess SPG is at around 1.18-1.19, nearly 50% no?

hopefully by tomorrow we have completed an absorb cycle and the SPG will be up in the 1.20s approaching 1.25 but for that I will have to wait.

Anyway just wanted to clear up where I got the figures from as I am just trying to read the data that is in front of me on the cc and the hydrometer.

Thanks again for taking the time to try and help me out with this.

V

[Sunking]

At 7pm sunrise the cc wakes up and gives a reading of 0.020 Kw.I stand by my panels in the morning and watch the shadow of the neighbouring trees pass across my panels at around eight am. The cc is reading 0.050 kw rising to 0.150 as it approaches 9 am. After 9am the last shadows leave the panels and it gets full sun exposure. The cc reads 0.350 Kw quickly reaching 0.500 - 0.650 Kw by 10am. By 11 we are at 0.800-0.850 Kw reaching 0.900-0.950 over the hours of 1pm-2pm. Falling a little to 0.800 around 3pm. 0.500 by 4pm. 0.350 at 5pm. 0.200 by 6 until neighbouring trees give shade and it drops to 0.150-0.050 and by 8 it is sleeping.

The sum of the average Kw hours equates to more than 6.0 Kw.

First understand Watt Hours = watts x hours.

So at 7 am your panel wake up and generates 20 watts and by 9:00 are generating some 150 watts. So lets say that average some 50 to 60 watts for 2 hours = 100 watt hours.

By 10:00 you are up to 650 watts and by solar noon you are up to some 900 for an average of lets say 700 watts for a total of 1400 watt hours.

By three you are down to 800 watts so an average 850 watts 2400 watt hours

From 3:00 to 6:00 down to 350 watts for an average of around 400 watts or 1200 watt hours.

So the panels are generating around 5 to 6 Kwh.

OK now factor in the losses of the charge controller and batteries at roughly 30% and we are looking at about 3.8 to 4 Kwh of usable power supplied to the batteries. With me so far?

OK using Coimbra solar data right now you are receiving around 6.3 Sun Hours so let's see if that jives. For MPPT system efficiency is 66%. So on paper you should be able to harvest roughly 960 watts x 6.3 hours x .66 efficiency = 3991 watt hours or let's just say 4.0 Kwh/day. Guess what? The numbers jive and are accurate.

What this means is you are using way more than 2 Kwh a day, and your system is way undersized. You are operating at a deficit. Not much more I can do to help you other than to say you have to make some serious changes.

Assuming you are using 4 Kwh/day will require:

Panel wattage = 3500 watts
MPPT Charge Controller = 80 amps
48 volt battery capacity = 400 AH.

Your current system can only be relied upon to provide you with only 1100 watt hours per day. Sorry I know that is not what you want to hear but the numbers do not lie. Right now the only advice I can give you is to run the generator everyday to make up for the shortages to keep those batteries fully charged up, or else you will destroy them. You are using way more power then you think you are.

[Mr Vee]

Okay. Thanks a lot for clearing that up for me. Now I understand the figures that you quoted.

I did not take in to account the 66% efficiency of the cc. This is rather lame and makes a farce of the figures displayed in the logs, but hey ho, that's life I guess.

Out of interest what would a typical 4kwh of usage consist of? Or, how much stuff is normally powered by 4kwh? It sound quite big to me, as I say we run 2 laptops, the water pump and the fridge. Could it be worth putting the fridge on a timer so it switches off overnight and is not a drain on the batteries.

I will experiment with different loads and test SPG throughout.

Any opinions on adding another FM60 cc to the mix in future?

Or running more batteries in parallel?

Thanks again.