Conext XW+ and exploding batteries

aj164 · February 2018

We've had two customers now who have had batteries blow their tops off. Several others have very warm batteries with warped cases (but still intact). In all of these cases, the batteries are at least 6 years old. They are all AGM. In NO case has the XW system halted battery charging due to overtemp. I always use the XW temperature sensor. Sometimes it's taped to the side of a battery, and most of the time I bolt it to a battery terminal.

What happens is one or more of the AGM batteries develops an internal short or some other condition which causes it to accept lots of current. The charger (XW inverter or MPPT charger) happily dumps more current as though it were charging a depleted battery set. The affected batteries heat up, eventually sizzle, and sometimes explode. We get a call. When we show up, the charger is still humming away trying to get those batteries charged up.

1) Does the battery over-temp parameter actually work? Will it force the XW charger and the MPPT chargers, which are all on the Xanbus network, to shut down?

2) The bad battery always seems to be the one that _doesn't_ have a temperature sensor. Should I be using multiple temperature sensors? If so, will the system shut down when one sensor is over-temp? (Does one sensor have priority over another?)

Dave Angelini · February 2018

The temperature sensor is not a an "over temperature" sensor as you wrote. It ramps the voltage up or down very slightly depending on temperature.

There only needs to be one sensor to perform the function that is described in the manual. You might read it.

Blown batteries are from a major problem that you as a installer should know about. Usually the overcurrent protection needs to be correct for the battery. The standard 250A breaker that is in an XW power distribution panel may be too large for your battery. This is something the installer needs to be able to adjust if needed. The set-points and many other installer configured settings need to be correct.

aj164 · February 2018

Thank you for sharing that. Anyone else have some ideas?

This is what I'm referring to (XW 5548/6848 operator manual):

"F44:

Battery Over Temp

F45:

Capacitor Over Temp

Any state.
If inverting, the unit shuts down and waits for the temperature to return to nominal value. If in any of the AC-interactive states (charging, peak load shaving, sell, gen support), the unit goes into AC bypass mode until the temperature returns to the nominal value. If the unit is not in AC bypass, it shuts down until the temperature returns to nominal value. After these faults clear, the unit returns to its previous operating state. "

"F44

Battery Over

Temperature

Automatic
Battery over- temperature shutdown at 60 °C.

Solution:
Clear the fault and attempt restart. Stop charging, check battery voltage and temperature. Check for excessive ambient temperature and adequate ventilation in the battery compartment. Note: Shutdown temperature is above 60 °C. Recovery occurs at 50 °C where the XW will be enabled again."

Regarding the overcurrent protection, batteries can heat up well below their allowable or recommended charge current. I set the maximum charge rate on the inverter-charger to C/10. Even at C/10, if all of that is going into a faulty battery, it gets turned into heat instead of stored energy.

Dave Angelini · February 2018

You said the battery tops were blown off?

Too be of any help you need to state all of the data like the voltage, amps into battery, battery model, set-points and so -on. You can call Schneider if there is warranty. Get a case number.
AGM batteries are very sensitive to set-points being too high. 6 years is where they start having problems.
A combox would log all of this data for you to review later.
Good Luck!

clockmanfran · February 2018

I have used 1300ah 48v AGM batteries for over 9 years now. 5 different manufacturers. The majority were from Mobile phone mast installations UPS back up systems. I got them for scrap at 5 years old and with most got another 4 to 5 years out of them before one cell would internally shorted out. Yes I took them apart to find out why.

Last year I scrapped and got rid of all of them and got SLA at a good price and after 4 years of testing particular SLA's. Hooray!

AGM's, and I mean Glass matt impregnated with acid as the separator between the plates, ....... These are normally sealed with a small expansion blowable chamber above each sell. Acid, its just damp in there, is trapped inside and can not be replenished, ie the battery is really sealed. No actual caps but the expansion chamber has a blow off cap sometimes.

AGM's at 6 years old?. You will need to check each battery in the string for getting warm and then to disconnect a string, let the battery rest, 6 hours is okay. Then check each battery in the string to find the battery that is low. I found with a 12v AGM battery that a cell had gone if the battery showed below 11.5v at rest.

Good battery casing just has the cells dropping out and the casing bulging and getting warm, other manufacturers have the cases splitting if you don't check regularly.

I found that AGM's loved being float charged, but the life was short if they were being constantly heavily used.

Bottom line, the charger will carry on charging, the bad cell will not take a charge, so the other cells get overcharged.

At one time I progressed with a BMS project with the AGM's to constantly watch over them and report issues, but it became a hassle to constantly daily to adjust and swap out at 35kg/80lbs each battery.

Estragon · February 2018

I'm not aware of any inverter/chargers or charge controllers that handle multiple RTS inputs, and the odds of a single sensor catching a shorted cell is 1/n where n is number of cells. Even putting one sensor on each battery wouldn't entirely eliminate risk if more than 1 cell per battery (ie not 2v single cell batteries) as the shorted cell may not trigger the sensor in time. Even a short in a single cell may be too far away from the sensor.

The FET and capacitor temps are internal to the unit. They will only overheat with high ambient temps and/or fan failure etc. - unlikely to catch a battery fault.

IMHO, the best way to catch this is with regular human monitoring, especially with an aging bank. 6 years is getting up there, and failures pretty likely. You could make a custom monitor with a bunch of sensors and an ardiuno or something I guess, but an actual human is more likely to catch something like a funny smell.

Dave Angelini · February 2018

A human is not always possible and that is where the combox comes in and logs data, graphs, and logs faults/warnings in a logical progression for troubleshooting after the event. The device that is acting up will be highlighted in yellow and then when it gets dangerous flashing red lights and buzzers.

A shorted, open, or stuck value on a temp sensor thermistor will only change the output a small amount. And as Clockmanfran stated nicely, the other batteries will just overcharge, gas out, and then the bad things start.

This is such a rare event (never happened to anyone I know) and that is where the human has to step in. Since this is a grid based system there are so many setpoints that have to be exactly right someone/something needs to monitor it all the time.

I wish there were pictures of the install.

BB. · February 2018

Even good AGM batteries seem to have around a 7 year maximum life.

AGMs use a catalyst (like palladium). While the catalyst by deffinition is not used up in reactions (combining Hydrogen and Oxygen back to water in a "low temperature" chemical reaction--basically lower temperature than burning)--In reality, the catalysts do "wear out" (pellet falls apart?) over time. Specifically, there is an "Amp*Hour" life rating (when the battery is fully charge, the float current generate gases based on amps, the gasses recombined at the catalyst).

So, my guess, is that the catalyst has failed. That can be from excessive charging voltage, excessive float charging current over time, or simply age. When the catalyst fails, the recombination rate goes way down. The batteries build up pressure and vent gasses--slowly using up the electrolyte in the cell. And eventually drying out the cell(s). And overheating/failing/possibly causing an explosion or fire.

Another possibility (don't think it is true in your case), excessive charging voltage/gassing current generates lots of H+O gasses, and they have to recombine at the catalyst--Which generates lots of heat (same BTU/Watts as a real fire--just lower temperatures). Causes the catalysts to overheat and fail (fire)?

http://www.power-eng.com/articles/print/volume-109/issue-10/field-notes/catalysts-on-vrla-batteries-save-air-force-millions-of-dollars.html
http://unifiedpowerusa.com/wp-content/uploads/2017/04/vrla_temperature1.pdf

If the batteries are "running" hot (in sun, hot battery box, etc.), for every 10C increase over "room temperature" (18F), battery aging life is cut by 1/2.

In times past, there were companies (still are?) that sold replacement catalyst caps (~$18 a piece?).... Did not find any now with a quick search.

http://hydrocapcorp.com/contact-us/faq/

How long is a HYDROCAPS life? Operating experience indicates 16,000 to more than 30,000 ampere-hours of battery overcharge. In normal application this should be 5 years. Instances of record show HYDROCAPS operating after more than 10 years.
Does catalytic activity deteriorate? Chemical activity after the first 16,000 to 20,000 ampere-hours would normally lose 15% to 20% capacity. To offset a decline and insure proper activity, the reliability factor is 15 to 1. Meaning: The reactive catalyst surface area provided is 15 times the area required for normal activity.

I am not sure I see the advantages of a VRLA (valve regulated lead acid battery) that does not have a catalyst "cap". The recombination of oxygen appears to occur in the negative plate, causing problems there--And shortening battery life. You can get Catalyst Caps for VLRA batteries that are supposed to extend VRLA battery life from ~5-7 years out to 20 year design life:

http://www.sbsbattery.com/products-services/by-product/battery-maintenance-safety-accessories/vent-caps-etc/bi-directional-recombination-vent-caps.html

-Bill

BB. · February 2018

I should add that AGM batteries have very low float charging current (probably in the 0.001 to 0.0001 of the 20 hour battery capacity (0.1% to 0.01%). Flooded cell lead acid batteries are 10x that amount.

Note that for any lead acid battery that floats at 2% or more of 20 hour current (100 AH with 2 amp of constant float current) is probably on the way to "exploding" and/or catching fire. Even a "good battery" that is being equalized at 2% rate of charge can eventually overheat (if fed 10+ hours of 2%+ charging current when the battery is at 100% state of charge).

-Bill "in my humble opinion" B.

Dave Angelini · February 2018

The standard settings for AGM's in a battery inverter often float the battery below normal manufactures recommendation when selling to avoid what Bill is describing.

The other workaround to avoid this kind of Grid thing is to use "charge block" settings, or timers to keep the battery from ever just being floated "forever". This is why this rarely happens to an offgrid solar system. They typically just lose capacity and power outages result.

There are many other modes of battery failure. The OP here is talking about at least 4 cases of battery damage and they could all be from commissioning errors. (wrong set-points) They could all be bad quality batteries or incompetence. An AGM in a grid battery system with casual users is very scary thing to me. Flooded Batteries are much safer as long as the water level is maintained.

Bill's system "Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset" is what should be used unless the user knows exactly what risks they are taking.

BB. · February 2018

Note that battery caps with catalysts should get a little warm when charging near full state of charge (when they are gassing a bit). The cap (or top of the battery) will get warm while the battery is gassing and recombining. If you have some cells getting warm and others not--Then that may be a battery problem (failing catalyst, lower state of charge cell, shorted cell, not charging current through that cell/battery, etc.).

-Bill

jonr · February 2018

> Bottom line, the charger will carry on charging...

Things like only one battery temperature sensor and a single voltage reading for the entire string are pretty primitive.

Possibly impending AGM failure could be detected with a hydrogen gas alarm.

Dave Angelini · February 2018

You can have as many sensors as devices and there would not be any difference. The gas alarm would only help one type of battery failure. They are primitive but any system that does not have a very smart device or human walking by and taking a look is going to have risk.

If the system was set-up right the charger would not have kept charging. A charge block timer, 2 stage/no float should have been used and lower than recommended float settings would have helped. The OP never added what those were. He insisted the sensor was there to prevent over temperature and that is just not so.

It would be great if it was and Outback, Schneider, Magnum and all the others could eliminate another weak link in the process and allow people to make mistakes. It clearly states in many places that you have to know what you are doing though.

Here is a description of a temp sensor.

The BTS mounts on a terminal post of a battery in a bank or on the side of a battery and measures its temperature. It sends precise information to the charger or charge controller, which automatically adjusts voltage to help ensure full battery charge depending on the ambient temperature of your battery installation.

aj164 · February 2018

I'll double-check the settings on one such failed system in the next week or so. The charger is always 2-stage no float, but the solar MPPT chargers need to be 3-stage, I believe, for Grid Sell to work properly. Charge Block is a good idea to let everything rest at night (grid-connected systems).

I know the primary purpose of the battery temperature sensor, but there's no other way the XW System would "know" about battery over-temp except from this same temperature sensor. It's a secondary use of the sensor, and clearly it isn't fail-safe due to the high probability that it isn't mounted onto the battery with the shorted cell.

If you have multiple charging devices (e.g. multiple inverters, PV chargers) on the network, it looks like you _can_ have multiple temperature sensors. The wording is a bit odd, but have a look at this paragraph from the XW+ manual:

"Batt Temp | Battery temperature as read by the BTS, connected to the Conext XW+. If the BTS is
connected to a different device on the Xanbus network, please see the Meters screen for
that device for temperature data. All Xanbus devices on the Xanbus network will share
temperature data for temperature compensated charging. However, the temperature is
only displayed on the meters screen for the device with the BTS installed. The BTS which
is monitoring the device displaying the highest temperature will take priority. If the BTS is
not installed, it shows NotAvailable."

Also I found this online:

"XW6048 F44 / System with multiple battery temp sensors (BTS)
XW system with multiple battery temp sensors (BTS)

Even though the reported temp will show up in the METERS menu of the device it is connected to, a single BTS will compensate charge voltage for all devices in the Xanbus network. It is possible that a bad BTS will report a high battery temp and cause F44 in the XW inverter. When this happens, it is also possible the erroneous value will be reported in the METERS menu of the device that has a good BTS. For this reason, it is necessary to remove all but one BTS at a time to see what temp is being reported and isolate the bad BTS. "

Lastly, the ComBox didn't exist when we installed these systems (that are starting to have failed batteries). These days, we install a ComBox on every system we build.

clockmanfran · February 2018

Reading this topic thread. I think it reaffirms my insistence of a BMS, (Battery Monitor system) with AGM batteries.

Does anyone do a independent off the shelf commercial BMS for AGM.s and even FLA. ?

As I said earlier, in my opinion and my empirical evidenced, AGM's are not a long term solution for Off/On Grid applications where the AGM's take a heavy loading.

In 2010 we started a BMS for my AGM'S in an effort to save money by using used/second hand AGM's , and this worked well for a few years until I got fed up of spending precious time looking at the SD card and swapping the batteries in and out.

The BMS runs with an Arduino and we used every pin input /output we could and got up to monitoring 32off 12v 110amp AGM's.

'Eric W' helped out with the coding and design.

"So ... our new circuit monitors every battery with a small box of tricks, and also contains a minicomputer that will examine every battery’s voltage, every minute in relation to the overall system voltage and power in/out. This info is stored on a data chip and analysed on the PC to see which offending batteries need special care and attention"

Heres a few old pics, pre phone camera days.

Circuit layout.

The completed BMS circuit ready to be installed into its enclosure box with all 32 individual battery small monitor cables.

I trust this helps?

Dave Angelini · February 2018

You really will never know what caused this problem and from what my buds that install AGM's tell me they are not doing well at this time.
The Concords have been pretty bad the last few years and the only ones they are using are the chinese Full River brand.

The few grid-tie XW's I take care of are set for 2 stage no float, and lower battery V, and the Charge block is set so the grid never can overcharge.
I have them do a sanity check once a month and shut off the grid and make sure the system can run the way it is suppose to.
If they are not doing that why in the heck would you have a battery system. I mean what good is it for them?
I bill them for my making sure that they never have a problem like this or I make sure they know what to look for. I would never sleep well if an AGM was not being looked at and replaced every 5 years or so. Offgrid is so much easier and far safer but you know that!
All battery makes have their dangers and risks! My LG Resu10 was on a handtruck the first year because of it's unknowns!

What is the temperature compensation range for my MPPT charge controller?

Product Line:
Conext MPPT60-150 or MPPT80-600

Environment:
MPPT installation with possible extreme temperatures

Cause:
Issue not addressed in product documentation

Resolution:
The temperature range where compensation is applied is between 0°C and 50°C. Outside of this temperature range, the compensation value is clamped at the corresponding value for either 0°C or 50°C.

Good Luck AJ

Conext XW+ and exploding batteries

Comments

Categories