High Specific Gravity Unexplained

mike95490 · June 2018

I'd try makeing a Reference solution for testing your hydrometer, distilled water and sugar, find a chart to create a solution close to your battery target SG. Some might say Salt is a good standard, but if any salt water came into contact with your batteries, it will be an unhappy day. At least the sugar solution won't kill you . Rinse well, then rinse again !

MrM1 · June 2018

I use 3 different types of Hydros to get a base line rather than using a test/calibration solution. I have a Hydrovolt, a glass type and a refractometer. All (well except for the HydroVolt) are reasonably cheap. I test with each one. The Hydrovolt reads consistently 5 points lower than the refractometer. I use the hydorvolt on a daily basis because it is the easiest to read -5 points, but I compare with each about once a month.

petertearai · June 2018

my rolls battery's have some cells at 1300.
when new 1280 was the max . . mine are 8 years old . so about to get replaced . can't eq without temp climbing significantly.the cell closest to the positve terminal is always the highest now.

jonr · June 2018

> the difference between being 1cm over the plates to being 3cm over the plates

I calculated SG would be about 1.30 if you are missing 2cm from a S-605. Maybe it was overfilled with acid mix.

https://www.thesolarbiz.com/media/catalog/product/r/o/Rolls_Surrette_S-605_Data_Sheet.pdf

mcgivor · June 2018

jonr said:

> the difference between being 1cm over the plates to being 3cm over the plates

I calculated SG would be about 1.30 if you are missing 2cm from a S-605. Maybe it was overfilled with acid mix.

https://www.thesolarbiz.com/media/catalog/product/r/o/Rolls_Surrette_S-605_Data_Sheet.pdf

The temperature the suggested level is specified would be at the standard 25°C, I would assume, being that the temperature of the batteries in question are 16 degrees higher than this, there would be expansion, so filling to this level may in fact be too low, resulting in a higher SG. To figure this out one would need to know the expansion coefficient of the electrolyte.

Ako · June 2018

I intended to take a reading just before charge started this morning however MSview crashed late last night so by the time i rebooted and checked everything the Voltage was already over 25 volts and 9 Amps had gone in this morning although the temperature was normal that time at 30c , my readings were all between 1.24 and 1.25 which was encouraging as yesterday the TS60 was working on the new settings ( Duty Cycle from under 30% for 1 hour to under 50% for 30 minutes - Voltage reduced from 29.76 to 29.36 ) It had reduced Absorption time to three and a quarter hours ( from the normal 5 or 6 ) and batteries had spent just under Four hours in Float ( doesn't normally go into float ) . I had used 1600 Watt/Hours ( 64 amp hours - 10% battery capacity or 15% of C20 rate ) while in Night Mode during the Non Charging period . Wouldn't know if this lower reading could be partly due to being weaker mix above the cells as to early to be bubbling even gently .Think before i do anything else i will compare everything over the next couple of days . I am reading and taking into consideration all of the suggestions and implementing where possible so thank you everyone for the continued support . The only thing i have eliminated as a possible cause so far are my Hydrometers and the accuracy of the RTS and the strength of the mix as the batteries are 18 months old and the 1.30 readings have happened before but only on a few occasions and certainly not withing the frst 6-8 months , normal readings were as they should be . I do have a new TS60 PWM but its not easy to change so will leave that as a last resort as i doubt its the cause .

mcgivor · June 2018

@Ako did you add water?

Ako · June 2018

Yes i did add water 24 hours earlier yesterday morning ,just as charging started , increased the level to about 6mm below vent protrusions as per manual , took more than i expected as i think i filled to a much lower level previously and the Rolls do have a large reserve above the plates , i know it was wrong time but i wanted batteries to have plenty of time to bubble and mix but haven't added any since

Sorry to have another question but on a different subject . I know temperature effects Panels and have been reading to find out by what extent . I notice the effect of hot panels is an Increase on Current and a decrease in Voltage . My system is down 20 to 25% now my panels are hot to the touch but thats incoming current which should be slightly up not down , im now getting 32 -33 amps coming in when i had 44+ during the colder months , according to what i read it shouldn't reduce Amps , just Volts which i am noticing on Adsorption is only around 1 volt down possibly .

Have i misunderstood something .

mcgivor · June 2018

Think in terms of power, Power = voltage × amps, lower voltage, lower power, read this article https://www.civicsolar.com/support/installer/articles/how-does-heat-affect-solar-panel-efficiencies

BB. · June 2018

You are correct about Vmp and Imp... The temperature related changes to Imp are much smaller than Vmp changes, and we ignore Imp changes for our estimates.

You have a PWM TS-60 charge controller (I believe), and Vmp changes do not really affect the output of a PWM controller as long as Vmp is higher than Vbatt+wiring+controller drop. On very hot days, Vmp-array can fall enough that the available (falling) charging voltage reduces current available to the battery bank.

In theory, when a charge controller is in "absorb charging" mode, the limitation to current flow is the battery bank itself. The charge controller is limiting current to the battery bank to hold the 14.7 volts (aprox) to the battery bank.

Only when the charge controller is in "bulk" charging is the charge controller passing all available current from the array to the battery bank.

Remember that PWM charge controller are really nothing more than an On/Off switch (transistors in modern controllers). When the switch is closed, all available array current flows (and current is limited by battery bank voltage/current acceptance). When the switch is off, no current flows to the battery bank. The controller's computer adjusts the switch's On:Off ratio. The more "on time", the more average current flow.

-Bill

Ako · June 2018

.McGiver thats one of the articles i read , they are talking about voltage i thought being reduced by .258 per 1c for a particular panel as an example , mine are 10 - 15 degrees hotter so if i had those panels or one with same specifications i would loose between 2.5% and 4% roughly . " As the temperature of the solar panel increases, its output current increases exponentially, while the voltage output is reduced linearly " , i only have a small drop in voltage but my current ( amps ) are within what the example gives for voltage .

Odd thing is that even with a lot less time in asorbe and spending plenty of time in float when it hardly ever did im not getting any reduction in my daily Amp/Hours , just seem to get more in Bulk and less after , strange with 20% less amps showing as coming in during Bulk im getting about a total of 20% more amps before going into asorbe and amps reducing as before in absorption . The voltage has also been reduced by 0.4v , i would have expected it to finish Bulk when amps are not throttled earlier , everything seems to be working back to front .

BB , I have been using the amps coming in during Bulk when there is no throttling but Voltage during Asorb as voltage is throttled to around .3 volts above battery bank during Bulk to alow maximum current but when in asorbe its the reverse so i use that mode to check the Voltage .

Estragon · June 2018

With pv connected to bank via PWM controller, voltage will be at/near battery voltage, so you wouldn't really see pv temp effects on Vmp. With mppt, you can see the effects on the controller input voltage.

Current increases exponentially in the passage quoted above because it's referring to the panel warming as the sun hits. Around dawn, pv is near temp adjusted Voc, current near zero. As the sun hits, voltage drops as the pv warms, and current rises quickly (initially), then slowly as the angle of the sun to the array gets closer to perpendicular near noon. I don't think it refers to changes in current between winter and summer temps.

If you're seeing lower current in bulk in hotter weather, this makes me wonder if panel tilt was set for max winter production (a reasonable thing to do, as you need as much current as possible in shorter days).

Ako · June 2018

Panels at 45 degrees pointing in correct direction and has given us sufficient for our needs so i dont need to squeeze every last drop out but i do get concerned when something is obviously wrong as the cost of things is so expensive here . Not sure actually when the input Amps in Bulk dropped as i have no readings sinse December but looking back to last year the amps coming in for this time of year and when the temperature was the same as now were only 1 or 2 amps higher than now while the colder months show an increase in amps coming in although the daily totals are similar in them all and remain so now .. The only time my total amps for the day drop is during overcast or cloudy days . Possibly i dont cycle very deep so im getting more than enough coming and the variation in consumption some nights is only small so not enough to make a noticeable different to the way the controller is working in as the voltage just before charging is always above 24.5 closer to 24.8 most days .

mcgivor · June 2018

You do realize you've derailed you own thread, the focus was on SG, although interrelated im still curious if you added water to dilute the higher reading previously experienced, all part of the learning curve.

Vic · June 2018

Ako said, " ... Looking at the fact sheet it stated theres 57mm reserve fluid above the plates , that is a lot however it shows 12kg difference between dry and wet weight , thats 4 kg of fluid a cell so 4 ltrs per cell ...".

A bit of nit to pick, but, the above is true if the fill fluid is pure water. For the speced electrolyte -- 1.280 --, the electrolyte weight would be about 3.125 kg.

The speced Electrolyte Reserve, is most often the total distance from the top of the plates to the top of the shoulder of the fill tube -- the point of overflow of the jar. This is customarily a electrolyte height is not something that anyone could use, in practice.

Trying to calculate the effect of adding make-up water on a cell's SG is fairly difficult, due to having too little real data, and the difficulty of testing the effect of adding water upon measured SG, IMO.

It would be nice to know the number of Ah removed from these batteries on a typical night. This would help in determining the actual DOD of the batteries. Deep-Cycle batteries DO need to be cycled. Frequent charging of lightly-cycled batteries can cause some SG anomalies, although, perhaps not from nominal 1.265, to 1.300 SGs.

There are probably several things going on with this bank. FWIW, Vic

Ako · June 2018

McGiver , Yes i did add water 24 hours earlier yesterday morning ,just as charging started , increased the level to about 6mm below vent protrusions as per manual , took more than i expected as i must have filled to a much lower level previously and the Rolls do have a large reserve above the plates , i know it was wrong time but i wanted batteries to have plenty of time to bubble and mix but haven't added any since . Its possible i guess that the addition of water reduced the SG however i added about the same amount last week and it didn't do anything after a days charging on the old settings .

Vic , Last night between the TS60 going into NIGHT mode and from NIGHT mode to Bulk my inverter shown I had used 1600 Watt/Hours ( 64 amp hours - 10% battery capacity or 15% of C20 rate ) . That is normal throughout the year as is the daytime consumption of around 1600 watt/hours to 2200 watt hours . I never look at the daytime consumption due to the fact my previous Trojan Batteries of the same capacity were 10 years old and regularly caused the Inverter to LVD late in the evening or before morning however it never even once let us down in the daytime .

Today after 6 hours of charging they went into Absorption showing they had 142 Amp/Hours of charge gone in at that point . They stayed in Absorption for 3 hours before going into Float with 174 Amp/Hours . They stayed in float for 3 and 3/4 hours until going into Bulk with 191 Amp/Hours showing . TS60 has now just gone into Night mode with a total of 192 Amp/Hours total input . Amp/Hour wise would be midway between the parameters of a normal days charge throughout the year although before the problem with the high SG and adjusted settings would be made up with Less while in Bulk and more In absorption , float never happened enough to bother checking .

Another obvious effect of adjusting the settings is the voltage at the point TS60 goes into Night mode they have now dropped to 25.5 volts , previously they had been showing as anything between 25.6 and 25.8 volts and took several hours to drop to 25.20 , Rolls said it was surface charge but i was using 64 Amp/Hours which seemed more that surface charge .

The SG reading after batteries had been in Float for an hour was 1.260 to 1.275 with 3 cells at 1.280 @ 34c

mcgivor · June 2018

It may take time to ballance out, but as you pointed out its not the easiest thing to get exact levels in all cells, so a slight difference would skewer results, but it sure seems as though it was a concentration issue. To establish a bench mark, you could fill to the vent protrusion when warm and fully charged, even if it's a little more than required, it will eventually return to a point where SG is normal, then you'll know where summer levels are.

Vic · June 2018

Ako, Thanks for the info on your readings on Ah consumed and returned ... did not know/forgot that you had the ability to report this.

One other thing; for whatever reason, it appears that your CC is doing a "Re-Bulk", after the end of Absorb -- this is when the CC begins another charge cycle, after the first on a given day.

Generally, most of us try to avoid this, unless the system has a relatively small battery, and/or large, infrequent loads during the day (like pumping water).

This could easily add battery charging, that you really might not want/need.

Usually, there is a setting, like Re-Bulk voltage that could fine-tune this to the needs of your system.

Just guessing, and perhaps you have carefully set this function, for the needs of your batteries, but it seems that this is adding some uncertainty/unpredictability to the operation of the system, IMO.

Later, thanks, Vic

mcgivor · June 2018

The return to bulk with a TS60 is not the same as a rebulk, when the battery voltage drops at the end of the day, the controller will display bulk charging, capturing the dregs of sunlight until the array is unable to overcome the battery voltage. With other controllers the rebulk is a setting where, if the battery voltage drops below a certain value in float, usually around 25v then a new bulk absorption cycle is triggered.

Vic · June 2018

mcgivor, Thanks for that info, the function is similar, although the effect of "overcharging" is PROBABLY reduced, but it seems to create uncertainty.

If this undesirable function is not triggered by battery voltage, then perhaps it is based upon PV in voltage, but, to me, it would be of NO use, and undesirable, particularly in a system that appears to suffer from "overcharging" the batteries. If a system has adequate battery Capacity, then Floating out the end of available sun would be preferable than increasing the Float voltage (or however one chooses to look at it.

IMO, for this system, as it appears to be operating, REDUCING the Float voltage seems in order.

The Surrette batteries in use here, use a Vflt of 52.4 (26.2 per 24 V), as has been recommended for a number of years. This recommendation has been increased in the past few years. But, we do NOT need that boosted Vflt, let alone, a second attempt at a Bulk near the end of the day.

As always just IMO. FWIW, Vic

jonr · June 2018

> A bit of nit to pick, but, the above is true if the fill fluid is pure water. For the speced electrolyte -- 1.280 --, the electrolyte weight would be about 3.125 kg.

12kg is 12kg. So you must mean volume = 3.125 liters.

It would be interesting to see confirmation that SG on a fully charged balanced battery doesn't rise further if you overcharge. I expect that the reason no further SG rise isn't used for determining full charge is that it's impractical.

Vic · June 2018

jonr said:

> A bit of nit to pick, but, the above is true if the fill fluid is pure water. For the speced electrolyte -- 1.280 --, the electrolyte weight would be about 3.125 kg.

12kg is 12kg. So you must mean volume = 3.125 liters.

It would be interesting to see confirmation that SG on a fully charged balanced battery doesn't rise further if you overcharge. I expect that the reason no further SG rise isn't used for determining full charge is that it's impractical.

YES, jonr, had wondered if I had made that mistake, later, but got wound up in projects and forgot to check.

Surrette specs many, many things, but the reserve volume to some stated height might be useful, but there are variables in the way these batteries are made that might make that less than useful.

Thanks, Vic

Ako · June 2018

The return to float was something i had not adjusted , i left it at the defult setting which was after 1 hour at 100% duty cycle . I noticed today when it was in float i was watching the instant read out , it was showing as 4.5 Amps going in , i turned on a 50 watt fan and it immediately went up to 7.5 , turned it off and back to 4.5 , did it many times at different inputs 7.5Amps went to 11.5 with the same fan when something else n the house was also drawing power in and so on , was always the same and instantly rose and dropped .

Looking at the settings it appears there are two methods of ending Float , either through the Duty Cycle algorithm or when the Voltage drops to a certain Voltage . Looks like you only have the option to cancel Float when voltage drops to a preset voltage if you select the Duty Cycle method you can then either enable the voltage method also or choose not enable to it . Attached a screenshot as its explainse better than i can .

Looking at the log i see that it went back to Bulk after 1 hour at 100% duty cycle , I had not realized that returning to Bulk was something to avoid but i will do in the future .however fortunately it didn't have any effect on the charging at all ,current going in was still around 2.5 Amps both in Float then in Bulk due to the time of day so i wont change that for a while but eventually change the settings to 3 hours at 100% which will keep it in FLOAT until it goes into Night Mode .

Last night my consumption went up slightly to 1700 Watt/Hours ( 68 Amp/Hours ) ( went to bed an hour later then normal ) and the readings were in line with what i would have expected , voltage at the lower end of the normal parameters , Amps coming in when entering Absorption at just 2 Amp/Hours more than Yesterday , into Float just 8 amps more than yesterday , temperature the same and will remain that way for a couple of months at least which is good as i have that as a constant . Total Amps input for the day was close also at 205 Amp/Hours ( 13 more than Yesterday )

Unfortunately the SG has risen 0.05 on 5 cells but a few days monitoring everything should see a pattern and help to give me a better understanding of these batteries

Hard to know what could be relevant so finding the line with information between including everything relative and boring irrelevant repetition ..

Vic · June 2018

Thanks Ako, for the info.

Guess, that if when in Float later in a day, and a larger load comes on, that the CC returns to Bulk -> Absorb, until the end Absorb parameters are again met ... dunno, we like mysteries.

Back to work now, will re-read your post above. Thanks, Vic

BB. · June 2018

Actually, lead acid electrolyte (sulfuric acid plus distilled water) weighs around 1.280 x pure water density (sg 1.000 for water).

Bill

Vic · June 2018

Ako, that 12 = 12 comment from jonr, am sure, was meant for me and my error in a previous post.

YES BB Bill, that 1.280 was my assumption for the nominal electrolyte density, as the relevant specs are based on 1.280 SG fill electrolyte . That was what was used to estimate the nominal volume of 4 kg of electrolyte ... There still be errors in my previous post ... just buried in plumbing projects ... no time ...

FWIW, Later, Thanks, Vic

jonr · June 2018

Yes, meant for Vic. I read his stuff very carefully (because he is so knowledgeable).

Ako · June 2018

Took readings this morning just as TS60 changed from NIGHT to BULK with input of .05 Amp/Hours and voltage was below Battery sense wire . The battery temperature was 29C but have not added the .005 adjustment so cells must be an 100% or more so still above the SG there supposed to be as the batteries cant be more than 85-90% . The thing that struck me firstly was the same Cells that were high in Float ( 1.275 - 1.290 ) yesterday and the previous day were normal this morning . The voltage was 24.6v and the consumption during the Non Charging period was 60 Amp/Hours . ( 10% @ C100 15% @ C20 ) .

4 Batteries each with 3 cells .The top lines are during Float yesterday , line underneath this morning ..

1.270 1.270 1.290 1.275
1.240 1.245 1.255 1.255

1.275 1.260 1.270 1.270
1.250 1.240 1.245 1.245

1.285 1.255 1.280 1.280
1.255 1.230 1.255 1.250

Vic · June 2018

Ako, Thanks for providing the Float and early Bulk SGs.

There is more variation in the SGs than I would have guessed. AND, for those measurements, it seems that your high SG situation, is not as extreme, as you noted before -- perhaps some of your changes to the charge parameters are helping (?),

If this were my battery bank, would try to discharge it deeper than normal -- to about 65-ish percent SOC. Watch the voltage on the second battery in the table (the one with the lowest SG cell), or the SG of that battery/cell during the discharge. Deep-cycle batteries DO like to be cycled relatively deeply, on occasion.

After this discharge, fully charge the bank, and EQ it in the lower range of the recommended EQ voltage, if possible with the TS-60. Would recommend trying to recharge the battery immediately after this discharge, as letting a fairly deeply-discharged battery sit, off-charge is not the best thing to do.

If you could try charging only that battery with a 6 V charger, that might be even better. If doing this, the output of that charger must not have its output connected to the safety ground, or either AC input connection ... am not certain where you are located, and the standards used for safety grounding, and weather the battery has a connection to any safety ground in your system. If you have a 6 V charger, you could quite possibly charge this lagging battery while the system is running in Float. This might be an adventure that you might not want to try.

FWIW, just some opinions, Vic

Ako · June 2018

Thanks Vic , im in Spain and i do have a DC earth ( Ground ) , a 3 mtr x 20mm steel bar laid horizontally in the ground 1 mtr deep and connected to a swiveling Bus Bar with 2 Inverters , batteries , TS60 and 2 Wind Turbines ( turned off almost permanently , not needed ) and which i find evens everything out .

I will leave the system as it is for a few more days because i want to see if the trend continues to reduce the SGs . After that i will do as you suggest cycling them deep one night , the next morning will take SGs to get an idea how low they went during the night and turn on Equalize straight away . The TS60 will do its thing and start off in Bulk and will continue until it reaches 31.2v then spend 2.5 hours equalizing which will be the lower end of the advised voltage for equalizing .

Unfortunately i dont have a 6v charger so will have to hope its enough to bring all the batteries closer as long as it doesn't raise the SG in the already high cells .

I think your right and the lower SGs im getting are due to the new settings although not reducing the input Amps daily totals im getting , in fact im getting more than weeks prior to changing the settings as the attached screenshots show , the amps coming in while in bulk for both screenshots were consistent around 32-33 Amps as they have been for several weeks although the early morning voltages are now consistently less ( higher inputs leaves lower morning voltages ) while consumption has remained the same . Im also noticing lower voltage when Night time Mode comes in , i also see the voltage dropping once in night mode as i would expect it to and not like before having to wait hours for it to get down to 25.2 . I would have expected the new settings and the long time spent in Float to have had reduced my daily total though in order to have given me the lower SGs im seeing now . I titled the attachments OLD SETTINGS and NEW SETTINGS because if i had not done so it would be natural to assume they were the other way around .

The variation between batteries is about what its always been .04v think thats 1/25 volt .

My heads hurting .

High Specific Gravity Unexplained

Comments

Categories