Power mismatch between bank and PV panels; is it dangerous?

Saipro

I live in the more tropical regions. For practical purposes, I'm off-grid with two generators (3.5kW and 1.1kW), a 1.4kW inverter and (2) 12V 200AH batteries in parallel. Takes care of most things. That's the old.

The new: I recently acquired (6) 240W 24V (about 30V Voc) panels with a 60A MPPT charge controller and a 3kW pure sinewave inverter. The panels are in two strings of three panels sending a peak of about 90V to the MPPT (can handle up to 150V DC). I'm getting adequate illumination (up to 9 hours daily). The problem is I have a paltry (4) Trojan T105-RE 6V 225AH batteries (wired as 24V in series). I'm looking at up to 51A from the MPPT in peak sunlight. The setup is new and has undergone test loading but not yet put to use. The batteries have not been hooked up to the charge controller yet and are still being charged from the bigger generator.

The PV array outsizes the bank because I have a freezer rated 580W which I want to power during peak sunlight (9am - 3pm). The freezer goes off at 3pm (that's the plan). I intend to tie the CC, inverter and battery bank together (not the best of practices, I'm aware).

Here' the concern: Is there a risk of damaging the batteries? By my calculations, the excess power should be drained away by the freezer during operation. Otherwise, I'm facing a scenario in which up to 40A of current would be reaching my batteries (calculated safe max should be 29A for the bank). Increasing the size of the bank (a replicate bank tied parallel) isn't currently in the budget.


Kindly tell me how to go about it if I'm doing wrong - don't wanna see my investment go up in smoke.

froggersix

yes too much current will wreck the batteries. if it is a good controller you can program it to limit current.

Saipro

My CC is an EP Solar iTracer ET6415N. I can easily control both max current and voltage leaving the CC. I'm looking at the power draw by my appliance and figuring it should more than offset for the excess incoming from the PV array. Am I working on a wrong notion?

EDIT: My CC is an EP Solar iTracer IT6415ND

BB.

As in most things "engineering"... The details matter and get confusing.

If you are asking about flooded cell batteries, if they are below ~80% state of charge (like your Trojans), they can actually take quite a high rate of charge (20%-25% or even more) just fine... However, when they get near 100% state of charge 2.5 to 5% rate of charge can overheat the battery and cause damage (basically, when the battery is "gassing", about 1/2 the energy goes into making hydrogen and oxygen gas and the other 1/2 goes into heat).

Next problem is as lead acid batteries get hot--Their charging voltage falls... So it is possible to overheat the batteries and confuse the charge controller into thinking the battery bank is actually not fully charged or being discharged... That is why we recommend that you have a charge controller with a remote battery temperature sensor... If the battery starts getting warm/hot, the charge controller backs off on the charging voltage (typically -0.005 volts per degree C per Cell). For example:

• -0.005 volts per C per cell * (35C - 25C standard temperature) * 12 cells [for 24 volt battery bank] = -0.6 volt reduction in charge set point

You can use base loads to reduce battery charging current... But you need those loads to be "reliable"... If they cycle (like a refrigerator) or are unreliable (large heat lamp that burns out)--Then you are left with a potentially over charging condition for your battery bank.

You can get a more complex/expensive charge controller (like the Midnite Classic with a Wizbang Jr. battery shunt) that can actually measure the current going to and from the battery bank--And adjust the "true/measured" charging current as needed.

The typical solution--Use a remote battery temperature sensor on your battery bank with your charge controller. That will keep the batteries from overheating/over charging current (in general). And, you can always drop the battery charging set-point voltage a bit if your battery bank is otherwise keeping charged at the end of a typical day.

-Bill

Saipro

Thanks. From the last reply, the solution is obvious - get another set of batteries!
Sadly, I have (as you put it) "unreliable" loading as refrigerators and freezers indeed cycle.
Doubling the bank (in parallel, I mean) would effectively double the required optimal charging current. That way, I'd never have to worry about "overfeeding" my bank with current. The new drawback would be wondering if my bank is getting enough charge especially on days we have an overcast sky.

PS: My CC has both temperature and voltage sensors

BB.

For cloudy days--Either reduce the loads and/or fire up the genset--Pretty much your only options. Lead Acid batteries do not like to sit with less than ~75% state of charge for days/weeks/months... However, if they are actively cycling (being charged/discharged, just not to >90% every day)--They can do alright.

You only need to hit >90% state of charge once per week (one battery vendor suggested once per month). If you can keep the batteries cycling >~50% state of charge--The batteries are doing their job and will have a descent life. It is an alternative method of using lead acid batteries--But they do need to be good quality deep cycle batteries. Automotive/Marine batteries will probably not survive this type of cycling very long.

-Bill

Johann

I am new to solar so, what I am saying may be a little off the wall, but it it may be doable. On the other hand it may not be doable.
I like to see what BB has to say about it. I do not have to much experience in this and I may be wrong.

Here it goes.

Take part of the string and and feed into the charge controller and then into the battery.
Hook a dump controller sensor circuit that checks the battery voltage.... to the battery. Set your on and off voltages for that dump controller. Your dump controller relay would act as a switch and not as dump to do this.
If the battery voltage goes down because of load, your dump controller would turn on and dump/switch the second solar panel string into your charge controller. And if the battery is topped off the dump controller would turn the second solar panel string off. So instead of dumping, you just switch the second solar array on or off to go to the charge controller.

BB would that be doable and possible? Saipro, that is just a thought that could /may possible work without raising the battery voltage over the limit, wait and see what BB has to say about it.
Just a thought.

BB.

The problem with charge controllers is each brand/model can behave differently. Using the older designs, it appears that some of them do a "sweep" of the solar array every 10-15 minutes... Basically they adjust from zero to maximum available current from the array, and log the array voltage at the same time--Maximizing the Pmax=Vmp*Imp -- And the problem appears to be is that some controllers are "unregulated" during these sweeps... So the charge controller can feed "excessive" current to the battery and bascially "over voltage" the battery bank (especially if the battery bank is mostly charged). This can knock AC inverters off line, trip over voltage warnings, and possibly damage the batteries too.

The problem with a "Rube Goldberg" type device--I don't know the behaviour of the charge controller or your loads.... So the chances that one can design a successful solar string series controller that would respond fast enough/reliably enough with off the shelf controllers is an open question.

Certainly, you could design a voltage detector that controlled a relay (easier than controlling a FET, but not as fast a FET)--It may work OK--Or it may not. Also, note that if the charging voltage problem is a result of a "sweep" for Pmax calculations, turning off 1/2 the array would leave the controller with a badly calculated Vmp-array figure until the next sweep.

Can you make it work--Probably. Is there a chance of damage/safety issues--A bit. Will you get some "confusion" when you have Voltage offsets due to changes in battery temperature (hot batteries should have lower voltage set points), having to disable during battery equalization or a different set point, etc.--probably..

I do not have any information on the behavior of charge controllers or the test equipment/time to measure this.... Or what how your batteries respond to changes in charging current. etc...

I am probably less worried about your 51 amps into a 225 AH @ 24 volt battery bank (it is high, but if you monitor the voltage/behaviour/temperature of the batteries and see nothing wrong)--I would not have a heart attack over that part of the configuration.

I am a bit more concerned about a 3 kWatt inverter on a 24 volt @ 225 AH battery bank... The nominal maximum current for flooded cell batteries is around C₈ or ~12.5% discharge rate:

• 225 AH * 24 volts nominal * 0.125 rate of discharge * 0.85 AC inverter efficiency = 574 Watt (long term) nominal maximum AC load

So, at this point, your AC inverter is something like 2-3x what I would suggest as your max. continuous power (reliable, not over heat/stress the batteries). As long as you are drawing ~ 1.2 kWatts or less continuous, and roughly for a C/2.5 or ~40% current draw:

• 225 AH * 24 volts nominal * 0.40 rate of discharge * 0.85 AC inverter efficiency = 1,836 Watt starting surge (few seconds) of AC load

As long as you are not exceeding the above ratings--Your battery bank should be relatively happy (assuming you are properly recharging, checking electrolyte levels, etc.).

-Bill

Saipro

Frankly, you guys have surpassed my expectations in terms of both ideas and solutions. As a bonus, BB even tossed in current draw and load ratings for keeping my batteries happy. For this, I'm truly grateful. I'll go with a little of everything in a single wrap - bigger battery bank. Solves all problem (excessive current, load draw, rogue spikes, etc)

BB.

Let us know how it all works out for you Saipro... Feedback is always appreciated and new readers can learn along with you.

-Bill

Saipro

It might take up to a week to be sure of the short-term effects of running a system like mine; perhaps months to see the cummulative long-term effects. Will surely revert to the forum. Thanks

Saipro

A few days have gone by. Nothing unusual asides a less than anticipated duration of insolation. Cloudy days are here once again. Having said that, my system is doing pretty well (had to use the backup gen to do some heavy work) otherwise, inflow of energy from the panel is actually slightly less than the total consumption so I have to cut back on the load during cloudy weather. I guess my fears were unfounded. Data collation however in progress and I'll give regular feedback. Still way too soon to tell what my equilibrum point might be.

BB.

Cloudy days--The bane of any solar power system...

-Bill

Saipro

As a plus, I had full insolation as early as 9am. By midday, the bank was fully charged because drain wasn't much and I had minimal load connected. On disconnecting the load, I noticed my MPPT was dishing out 0.35A to the bank but on connecting my heavy stuff, that went up to about 40A in a few seconds. This means, once the battery state is assessed (in addition to drain), the MPPT adjusts power supply accordingly. I had no reason in the first place to worry!

Saipro

More data (I've promised to give feedback till the system attains equilibrum - and beyond). New discovery: I was a bit off on my calculations.

Considering 100% system efficiency, my lowest operating voltage is 26.4V with a corresponding max current of 54.5A. Incidentally, that's the point at which my system is mopping up the least power so I can step up my loading to match. The CC also drops the current to attain equilibrum.

At absorb voltages (28.8V - 29.6V), the max current output from the CC is 48.6A (assuming 100% efficiency once more). Weirdly, I've seen 48A register on my CC; that's by the way. My batteries are capable of mopping anything the CC can deliver in bulk mode.

The crux: I'm fearing (with my load), my system might be charging sub-optimally. I'm thinking of cutting my loading so I can give my batteries enough charge to make them happy. However, my system shows it's giving at least 95% between 12pm and 3pm. Independent measurements using a meter confirm this. It would appear the tropics are much sunnier than I anticipated. This is in spite of the corresponding rise in ambient temperature (you would expect a lower system efficiency.

Saipro

So, the four new batteries (T105-RE) have arrived. A new state of data collation will commence next week.

Saipro

A new concern has arisen. Despite doubling my banks, I seem to have about 50 - 75% extension of runtime before my DoD reaches 50%. I was literally expecting at least a doubling of battery runtime. My estimation of DoD is based on the battery bank's voltage. Or am I missing something here?

BB.

The new batteries will take something like 20-100 cycles before they develop maximum storage capacity (maybe 20% additional capacity).

If the battery bank temperature has changed (warmer weather), the voltage will drop a bit (and the capacity will be higher).

-Bill

Saipro

Thanks, I eventually got that info from the Trojan company too. Was about panicking. Where I live, it's not unusual for sly/crafty dealers to sell used/recycled/refurbished batteries (indeed all products) as new. As promised, I'll be in touch when I reach that number of cycles.

Brlux

Do you have any pictures or videos of your system. I skimmed the manual for the ET6415N and it seems to be the most feature packed Chinese charge controller I have seen. It seemed to indicate the user could set custom battery voltage settings but did not go into how to do that. In fact it did not go in to how to change any settings. Are you using the default flooded settings or have you been able to set it specifically to the manufacture recommendations of the T-105-RE?

It looks like it attempts to report a % state of charge without doing coulomb counting (Using a Shunt to the battery)?

For as long as this charge controller has been on on the market I am surprised I was not able to find any pictures or videos of user installed operation. All the videos and pictures I could find were marketing shots of it still in it's box.

Out of curiosity where are you located?

Saipro

I live in Nigeria (smack in the tropics).

I use the iTracer (slightly different from the eTracer by having separate terminals for loads) is highly customizable. Mine is the IT6415ND (not ET6415N as errorneously stated earlier). The available preset battery settings for the different battery types should work fine as they're the same as recommended by many manufacturers. For the more exacting requirements of manufacturers like Trojan, there's a "custom" option under which literally everything can be modified. More important is the fact that you can set the lower cutoff limits (by voltage) to limit your DoD to anything you desire. Mine is set to 24.2V which rougly translates to about 50% DoD with moderate loading. That way, even with light loading, I'm somewhat protected against unanticiapted drains.

DO NOT USE THE SOC feature as it seems grossly inadequate when charging and loading are going on simultaneously, even with the voltage and temperature sensors in place. I've tried tweaking it but it's calculations of SOC seem dependent mostly on the absorb phase voltage of the charging algorithm and as such unreliable. If it exits the absorb phase into the float successfully while still being loaded, the calculations normalize but that doesn't often happen as my settings force it to go back into absorb mode in daylight once SOC drops below 90%. Theoretically, with minimal loading, the SOC feature should work just fine. The summary is: the SOC cutoffs vary with the stage/phase of charging which I think is plain wrong. Then again, what do I know?

IF you're willing to accept its few shortcomings, the odds favour this CC surpassing your expectations. I particularly love the way it varies the charging current with the SOC and at times occssionally shuts off charging (even in the brightest of days) when the batteries are full and no load is connected, periodically reentering the float phase as needed. Most times though, it applies an average of 2A - 4A consistently at float voltage (26.4V for me). Unless in bulk mode (when the batteries are pretty much flat), it NEVER delivers more than the recommended 13% charging current, regardless of what the PV panels are generating.

Saipro

I have slightly modified my system by replacing the iTracer with a Midnite Classic 150. Nothing much has changed apart from the readings. Greater accuracy thus I'm "seeing" slightly less power from the panels (due to voltage offset adjustments) but it's what it is. No complaint.

The problem: in the new string of four Trojan T105-RE 225 Ah batteries (I added them to an older string of four), two batteries maintain the same voltage profile always while another two keep growing in voltage berth. For example, a few minutes ago, they measured 7.27V, 7.42V, 7.27V and 6.26V. I'm thinking of pairing them up as 7.27+7.42 and 7.27+6.26 and linking them in parallel overnight. Tomorrow, I equalize at 15V till I'm satisfied. Is this the right way to go about it?

vtmaps

Saipro wrote: »

I'm thinking of pairing them up as 7.27+7.42 and 7.27+6.26 and linking them in parallel overnight. Tomorrow, I equalize at 15V till I'm satisfied. Is this the right way to go about it?

Not quite.... best to equalize them one 12 volt pair at a time.

I haven't seen a word in this thread about using an hydrometer. An infrared temperature probe can point to a hot cell. A DC clamp ammeter might show you high current in the string with the low voltage battery. You may have a shorted cell in the low voltage battery.

I usually try to make a diagnosis before I begin treatment. Equalizing a shorted cell is not recommended.

--vtMaps

Saipro

I see. Point taken. I do have a standard hydrometer. Perhaps I would start by following the established flow of diagnosis. I guess I must have become irrational somewhere along the line. Thanks.

EDIT: Err, ..... how do you recommend I proceed?

vtmaps

Saipro wrote: »

I see. Point taken. I do have a standard hydrometer. Perhaps I would start by following the established flow of diagnosis. I guess I must have become irrational somewhere along the line. Thanks.

EDIT: Err, ..... how do you recommend I proceed?

Use the tools you have... use your hydrometer. Do you have a DC Clamp ammeter? That could be useful.

One of the problems with parallel batteries is that if you have a shorted cell, when you try to figure out what the resting voltage is, you can't do it. Even if you turn off all loads and charging sources. The reason is that one battery may be a load on the parallel batteries. One side of parallel circuit can be discharging into the other side.

First thing you must do is separate your battery banks... that is connect only one string at a time to your system.

When you hook up the battery with the low voltage, be careful charging it... watch it for excessive temperature or boiling from one cell. Do you have a digital volt meter? Watch its voltage under load. After a load, watch the voltage to see how far it rebounds. After a couple of hours without a load or charge, what is its resting voltage?

--vtMaps

BB.

A couple of nice (reasonably priced) diagnostic tools:

http://www.solar-electric.com/midnite-solar-battery-hydrometer.html (nice "different" hydrometer--Be sure to rinse with distilled water before putting away)
http://www.sears.com/craftsman-digital-clamp-on-ammeter/p-03482369000P ("good enough" for our needs AC/DC Current Clamp DMM from Sears--currently $60)

As vtMaps says--paralleling 12 volt batteries is a pain... You cannot use a volt meter to find "problem" cells/batteries. And if AGM/Sealed batteries, you cannot even use a hydrometer.

If you use 6 volt (or 4 or 2 volt) batteries on a 12 volt bank, you can, at least, do a quick check of voltages across each battery.

And if you use a DC current clamp meter, you can measure the discharge/charging current for each battery string (suggest heavy loads or high charging current when measuring).

With lots of cells/batteries--You are looking for differences. A "high or low reading" is not necessarily good or bad--It is just an indicator that you need to investigate and find the "root cause" of the variations.

-Bill

Saipro

While wainting for reply posts, I tried a little looking. Seems the positive terminal cable on the lower voltaged battery wasn't properly crimped ergo some heating as well. I have a hydrometer only and everything seems fine. Connecting the highest and lowest in parallel for 3 hours (risky but worked) seemed to get them closer to each other. Late hour charging with sunlight seemed to regularise things a bit. The margin is now narrower.

I think the battery is simply at the receiving end of a bad connection and I didn't diagnose it on time. Let me take two days off to parallel-dump/equalise the string in different 12V battery cluster combinations then do a final equalization run on the entire string. I hope that's somewhat a logical solution. The annoying thing is the second string (the older one) has all batteries in voltage sync, down to the second decimal place while the new string has voltages running amock within it.

inetdog

When one battery in a string is significantly hotter than the others, there is a good chance that it will require more total amp hours to make up for each amp hour of discharge.
That can lead to a charging imbalance in the string.

Saipro

It's actually colder than the others and is hardly bubbling while the high voltage battery is significantly warmer and vigorously bubbling, literally boiling over.

That on one side, I working on a way of raising its status to par with the others. I've attained some degree of success but it's hard. Feedback on the way.

BB.

"Cool string" could have open cell/bad or corroded wiring. Could also be just lower state of charge.

"Hot string" could have shorted cell/better wiring (shorter/heavier wiring).

Check voltage drop on wring/each connection with a volt meter set to 2 volt or 200 mVolt full scale (under heavy load/charging current). Look for high voltage drops in "cool string".

Check cell (or battery) voltages. Check specific gravity of each cell.

-Bill

Saipro

Saipro wrote: »

I have slightly modified my system by replacing the iTracer with a Midnite Classic 150. Nothing much has changed apart from the readings. Greater accuracy thus I'm "seeing" slightly less power from the panels (due to voltage offset adjustments) but it's what it is. No complaint.

The problem: in the new string of four Trojan T105-RE 225 Ah batteries (I added them to an older string of four), two batteries maintain the same voltage profile always while another two keep growing in voltage berth. For example, a few minutes ago, they measured 7.27V, 7.42V, 7.27V and 6.26V. I'm thinking of pairing them up as 7.27+7.42 and 7.27+6.26 and linking them in parallel overnight. Tomorrow, I equalize at 15V till I'm satisfied. Is this the right way to go about it?

Did everything as instructed. Now the difference during the absorb phase has narrowed down to a max of 0.20V between the lowest and higest batteries in the string. During float, it's often less than than 0.1V (typically 08.V or less) and at rest, they are of the same voltage of 6.33V.