Symmerta PX Battery Temperatures and Cooling

[ojfsdi]

I'm examining a PX with 3-4 year old battery packs.  A few packs have reported failed in the last two months, and after removing, allowing time to cool and putting back in a different slot to verify the failure, they did not re-report failure.

Then when another one faulted, I took the time to open it up first, and discovered deformed SLA battery.  We've all seen melty batteries before, but I snapped a thermal pic that is just too cool not to share:

The white area is where it was bubbled out.  Glad I had my PPE on.  NFG at all, I do say.  PROTIP: If you need oven mits to remove a battery cartridge, it's bad.  I wonder if I can get an SE embroidered oven mit from a sales rep next time we order.

What cooling, if any, is the PX 80kva battery cabinet supposed to provide?  I don't feel any moving air, at all, around it so I was wondering if perhaps our fans had failed.

What is the nominal temperature range for battery packs? (Ours are all model SYBTU1-PLP )

I shot some video on a high res FLIR camera if you're interested.  Pardon my voice.  I was at a dentist 30 minutes prior.  The video shows temp on a color scale that updates each frame, but the key/scale is always present in the video,

video 1 (Whole battery cabinet)  video 2 (the culprit pack) 

I'd expect some temperature variation to always be present.  But just how much is normal?   +/- 10%  in celsius/kelvin/faren?  No doubt there's a problem with the pack I pulled.  But the pack on the bottom, only ten or so degrees above the others, is it worth pulling?

[ojfsdi]

What cooling, if any, is the PX 80kva battery cabinet supposed to provide?

[al_buffington · Accepted solution]

Hi,

There is no "cooling" provided by the battery cabinet. There are perforations in the cabinet to allow for proper airflow. Generally speaking, the temperatures reported by the batteries in a PX80 are not much higher than the room ambient. 

[ojfsdi]

When they're sitting unused (tickle charging, I wouldn't expect the batteries to generate much heat.  But during discharge and charging they would, wouldn't they?

How much heating is normal?  I don't think the web interface displays the temp of every battery cartridge.  But I can see some variation on a FLIR camera at times.

The FLIR cam exentuates even the smallest variation, to about a tenth of a degree.  So I want to know what threshold I should be concerned about.  If one back is a degree F above its neighbors, should I replace it?  5 degrees?  10?

In this pic, you can see most of our packs are at 25C. But one is at 27.

the 27 degree one shows up as normal and OK in the management interface.

(sorry it's showing up sideways in this forum.  it's vertical on my pc.

The handles show up as lines in the flir image.  Also, Ignore the two vacant spots in L2 that show up hot.  that is an artifact of the reflective surface inside and FLIR imagery.  It's the same artifact that makes he handles show up btw.

But the elevated temp at position 4-C is not an artifact. 

How much heat, or variation in heat, is normal while not running on battery?

[al_buffington]

Hi,

I've tried to address each of your questions below.

When they're sitting unused (tickle charging, I wouldn't expect the batteries to generate much heat. But during discharge and charging they would, wouldn't they? Yes, this is when the most heat is generated within the cabinet. The battery temp is always monitored though and they can have elevated temperatures while in normal operation. 

How much heating is normal? I don't think the web interface displays the temp of every battery cartridge. But I can see some variation on a FLIR camera at times. I'm not sure that I can answer what is "normal". It greatly depends on the environment where it's installed. The web interface does not display the battery temp. There should be a SNMP OID for the battery temperature, which on this system would show the module with the highest reading. FSRs, using their software tools, can read the individual battery module temp values 

The FLIR cam exentuates even the smallest variation, to about a tenth of a degree. So I want to know what threshold I should be concerned about. If one back is a degree F above its neighbors, should I replace it? 5 degrees? 10? There is no black and white answer to this. The UPS will not issue an alarm for an elevated battery temperature until a module reaches 40C. In addition, the UPS battery charger compensates for temperature. If you start to see a slightly lower float voltage on the batteries perhaps then you can be concerned as the charging to all of the batteries would be impacted. 

How much heat, or variation in heat, is normal while not running on battery? Again this will also depend on the environment. No studies have been done (to my knowledge) to examine temp variations within a system while in normal operation.

[ojfsdi]

What does the charger do to compensate for temperature?

I'd imagine it reduces charging current if a pack is hot, but I don't want to put words in your mouth.  is that correct?

Is there a way to see and log over time the charging current being applied?

I know that I can observe voltage and current in and out of the UPS, so I suppose I could rig something up that would calculate total watts in and total watts out and the difference between that would be basically power to charge the batteries, right? 

Where can I get that software tool that can read temps of all modules?

I'd love to get that on our librenms graphs somehow.  I'd like to know if there is a reliable slow period of temperature growth before a pack goes temp critical and 'fails'

I'd gladly share the data I gather from monitoring the temp of all my cartridges over time.

[al_buffington]

Hi,

I've tried to address each of your questions below....

What does the charger do to compensate for temperature? The battery float voltage is reduced. By how much will depend on how high the temperature is

Is there a way to see and log over time the charging current being applied? Not really. Although the charging current is visible through the software used by Field Service. 

I know that I can observe voltage and current in and out of the UPS, so I suppose I could rig something up that would calculate total watts in and total watts out and the difference between that would be basically power to charge the batteries, right?  I don't know if it would be that exact. I can tell you that each Power Module has a 1kw charger

Where can I get that software tool that can read temps of all modules? This is available only to internal Schneider Electric employees or authorized service providers. 

I'd like to know if there is a reliable slow period of temperature growth before a pack goes temp critical and 'fails' I don't believe such a study has been conducted.