Smart-UPS SUA1000XL + 2 battery packs shows low runtime

[BillP]

This is my first time posting here, but I've been using APC products for over 9 years now and never had any problems until now.
I have a 2003 Smart-UPS SUA1000XL with two (2) XL 24V Battery Packs connected. In early February of this year (2009) I replaced all the batteries in the main unit and both battery packs (a very expensive venture, let me tell you). I replaced the batteries because the "replace battery" light went on (duh!) in late January and both battery packs were very hot to the touch, so I figured something was very wrong (it was difficult to get the bloated batteries out of the housing which should say something). Everything is a cool-to-the touch temperature now.
Anyways, after replacing the batteries, the unit reports 21 minutes of runtime. With 2 battery packs I was expecting it to have several hours. I have a very constant 30% load on it. Recently the power went out (thank you PG&E!) and the unit immediately went from 100% to 0% runtime (full bars to one flashing bar). The software reported the outage and 10 seconds later shut the systems down. This doesn't make sense to me and I'm wondering if the over-heated batteries did some damage to the UPS. If I let it recharge fully, then pull the power, it does the same thing, and takes almost a full day to fully recharge.
The thing it, this unit has always had these two battery packs and has never reported more than 21 minutes of runtime. Back when I bought the unit I set the number of batteries to "2", as per the instructions, and the UPS reports "2" back to me today. What gives? Its like it's not really using the battery packs. From other posts it seems I should attempt a runtime calibration, but that takes time (and movement of systems to an alternate UPS), and something tells me that there's something else wrong. Any thoughts?

[BillP]

Thanks for the followup. Yes, I've adjusted the load and checked what the UPS calculated for runtime and dir a little math myself to verify it looked correct. My current, stable, load is a straight 30% and the runtime shows 396 minutes, which I calculate to be about right. So, thanks again. My issue has been resolved!
Long Live APC!! 🙂

[Erasmus_apc]

Based on the symptoms you describe (calculation of battery runtime incorrect and a suddenly fluctuating capacity), I would certainly try a runtime calibration, and if that does not work, I would be interested in checking the battery constants via hyperterminal that the unit has after the runtime cal (you will need to call into our main tech. support line @ 1-800-800-4272 for this procedure as it needs to be followed to the letter to avoid risk of damage to the UPS).

Additionally, if you were to remove both the battery packs and readjust the number of "packs" in the software back to 0, does this change the runtime? Try doing this, and also running a self test...If the above steps don't fix the issue I'd like to see the main UPS itself calculate runtime correctly, then the packs can be accounted for.

Please make sure you have any smartslot accessories removed when doing all of the above procedures to rule out bad communication due to an accessory.

[BillP]

Thanks for the quick reply.
I've had a quick play with the number of battery packs and the results are this:
Setting the number of batteries to 000 causes all the battery leds to by on and flashing. The reported runtime is 0003.
Setting the number of batteries to 001 causes all the battery leds to be on (no flashing). The reported runtime is 0010.
Setting the number of batteries back to 002 causes all the battery leds to be on (no flashing). The reported runtime is 0021.
So, it seems the setting of the number of battery packs is effective in the calculation of runtime, but it seems the calculation is really, really off.
I'll see about moving everything off of it and performing calibration.
Would you recommend doing this with both battery packs connected or just the unit by itself?
Thanks again for your speedy help with this!

[Erasmus_apc]

You will still need as close to 30% load on the UPS as possible in order for the unit to accurately recalibrate the runtime. That load can be something non-critical, but nonetheless there needs to be load on the unit.

I'd prefer a start with just the UPS by itself, with 0 batts added in the software. Before the runtime cal, once the # packs has been set to 0, see if the UPS will support a load by itself. With 20% load on the UPS (160 watts) you should get about 1:45 minutes or so of runtime. With 50% load on the UPS (400w) you should get about 30 minutes. You don't need to leave the unit on battery with that kind of load for the full time, but it would be nice to see if the UPS by itself with no added packs at least calculates the runtime correctly via the software, and if it is actually capable of holding some load up for at least 5-10 minutes without going into a "low battery" state (constant beeping).

After doing this test, perform the runtime calibration on the unit (still with no batt packs and none added in the software), and make sure you keep 30% load on the UPS while doing so. If this does not resolve the problem you should call into tech support to have them walk you through forcing the battery constants on the unit to their correct values. Once the UPS itself is calculating its own runtime and/or batt capacity correctly we can then quickly add in the battery packs.

Edit: Can you also post some battery voltage information, battery serial #s, as well as data and event logs that have been recorded in the past, and get recorded during your future testing?

Message was edited by: Erasmus

[BillP]

I'll post what I can.
Battery Voltage: the logs seem to indicate it oscillates between 27.40 and 27.54 and that's after the batteries were all replaced (on 2/8/09).
I don't have any data earlier than 2/26, sorry, but I do have an event log going back to 8/16/08. Here's a sample of the data that I do have (this is from PowerChutePlus):
02/26/09,06:20:06,115.9,118.8,117.3,27.54,60.00,029.9,029.2, ,
02/26/09,06:30:06,116.6,118.8,116.6,27.54,60.00,029.9,029.7, ,
02/26/09,06:40:06,113.7,118.8,117.3,27.54,60.00,029.9,029.7, ,
02/26/09,06:50:07,115.2,118.0,115.9,27.54,60.00,029.9,029.7, ,
02/26/09,07:00:06,112.3,118.0,115.2,27.54,60.00,029.9,029.7, ,
02/26/09,07:10:06,114.4,118.0,117.3,27.54,60.00,029.9,029.7, ,
02/26/09,07:20:06,113.0,118.0,116.6,27.54,60.00,029.9,029.7, ,
02/26/09,07:30:07,113.0,118.0,116.6,27.54,60.00,029.9,029.7, ,
02/26/09,07:40:06,115.2,119.5,118.8,27.54,60.00,029.9,029.7, ,
02/26/09,07:50:07,115.2,119.5,118.8,27.54,60.00,029.9,029.7, ,
02/26/09,08:00:06,115.9,119.5,118.0,27.54,60.00,029.9,029.7, ,
02/26/09,08:10:06,115.2,118.0,116.6,27.54,60.00,029.9,029.7, ,
02/26/09,08:20:06,113.0,118.8,117.3,27.54,60.00,029.9,029.7, ,
02/26/09,08:30:07,112.3,118.0,115.2,27.54,60.00,029.9,029.7, ,
02/26/09,08:40:06,110.8,116.6,114.4,27.40,60.00,029.9,030.1, ,
02/26/09,08:50:06,110.8,115.9,114.4,27.40,60.00,029.9,030.1, ,

It seems we have poor quality line voltage in our area, 'cause I see lots of the following in the log:
200200 08/16/08 11:20:07 UPS enabling SmartBoost
100300 08/16/08 11:22:03 Normal power restored: UPS on line

About a week after replacing the batteries, I ran a self-test:
300301 02/17/09 11:36:24 UPS battery is discharged: 001.0%
100403 02/17/09 11:36:34 Self-test at UPS passed
100701 02/17/09 12:05:33 UPS returned from low battery condition: 022.0%
300301 02/17/09 12:07:14 UPS battery is discharged: 023.0%
100701 02/17/09 12:07:21 UPS returned from low battery condition: 023.0%
300301 02/17/09 12:07:22 UPS battery is discharged: 023.0%
100701 02/17/09 12:07:30 UPS returned from low battery condition: 023.0%
300301 02/17/09 12:07:34 UPS battery is discharged: 023.0%
100701 02/17/09 12:07:34 UPS returned from low battery condition: 023.0%
300301 02/17/09 12:09:15 UPS battery is discharged: 024.0%
100701 02/17/09 12:09:16 UPS returned from low battery condition: 024.0%
300301 02/17/09 12:11:14 UPS battery is discharged: 026.0%
100701 02/17/09 12:11:20 UPS returned from low battery condition: 026.0%
300301 02/17/09 12:13:15 UPS battery is discharged: 027.0%
100701 02/17/09 12:13:19 UPS returned from low battery condition: 027.0%

This is pretty typical of what I saw over the last few years (that I recall) and what I see today, running a self-test.
I can't get you the serial numbers of each battery just yet (that is if I can get them out without destroying them...they were a bit difficult to put in), but if I have to open it up for some reason, I'll be sure to record those. However, I do have the serial numbers of the main unit and the two battery packs themselves:
SUA1000LX: AS0315330062
SUA24XLBP: WS0318130541
SUA24XLBP: WS0318130543

I'm moving my loads off this unit but it's taking some time to do. Hopefully I'll be able to give it a calibration test tomorrow.

[BillP]

Ok, I've run the calibration test and the results are in:
I've put a 35% load on the unit, disconnected the extra battery packs and set the pack count to 000.
Just sitting there the unit reported 4 minutes runtime. I pulled the plug and monitored it.
Pulling the plug the runtime dropped to 0 immediately and started constantly beeping.
The battery voltage started at 24.34v. It steadily declined to 22.68v and powered off after 57mintes.
I turned off the load and plugged it back in to main power.
Surprisingly, while it turned off power to the load, the UPS was still active and able to take commands.
It reported 4% of battery reserve at this point.
Its recharging now. Once it's fully recharged, I'll reconnect the load and see what it thinks it's runtime is then.
Once it's recharged, and if it reports a decent runtime, I reconnect the battery packs.
Do I need to perform another calibration?

[BillP]

As a followup, the UPS has reached full battery charge. Applying the same 35% now shows a runtime of 56 minutes.
So, it looks like the calibration test worked.
Now, my next question is, do I need to perform the calibration test again with the two battery packs attached or can I just attach them and get back to business?
Using the Extended Runtime calculator here:
http://www.apc.com/resource/include/ups_runtime_chart.cfm?base_sku=SUA1000XL&family_id=164
Shows that with two packs attach I should expect about 6 hours 15 minutes (approx).
So, if I setup everything as before and it doesn't report about this amount of time does that mean I have to perform the full calibration test again?
6 hours is a long time!
Again, thanks for your help on this!

[BillP]

As a follow up, I've reattached the 2 battery packs and set the number of packs to 002.
Runtime reported is 390 minutes (or 6 hours 30 minutes).
The question is: can I trust it?
Would you recommend doing another full calibration test or do you think it's Ok now?

[Erasmus_apc · Accepted solution]

Based on all of your above posts I would say that you do not need to perform another runtime calibration. Once the UPS itself is properly calibrated it will know how to calculate runtime for the battery packs as they are added into the UPS EEPROM (memory). As long as the 390 minutes it has given is the UPS estimate with your load attached (and not some "empty" reading), I would say that this issue has been resolved.

Edit: As an FYI APC typically recommends against constant runtime cals. since these put stress on the battery and over time can weaken them as if they had actually been used. Only do another runtime calibration if this same issue comes up on this UPS or (another one) in the future.

Message was edited by: Erasmus