Batterie shown from time to time at 99%

[rath4snps_apc]

This was originally posted on APC forums on 9/26/2016

---

Hi,

I have 2 Smart-UPS SRT 3000 in remote site. both UPS are monitored by my own SNMP based tool, where I create Graphs from. The graphs are based on last hour, last day, last week... and were created with RRDTool.

As you can see, the battery is always between 99% and 100%. UPS is max 3 month old, same for batteries. I can't see any recent power issues from incoming power,

On both pictures the same timeframe is shown. The 99% happen if Incoming power (red line) is below, but also higher than configured 230V output level.

Is there anything wrong with the Batteries? Do I need to be concerned that devices attached to UPS are not 100% covered?

I've quite some more UPS devices (2200, 3000, 6000... series) which I monitor with same commands/tools. and none of them show this behavior.  The 99% is shown between 30 and 60 minutes and can also be seen internally in the logfiles (so no monitoring issue)

Any ideas? Any suggestions?

Thanks for any help in this.

thanks and regards

Stephan

[voidstar_apc]

This was originally posted on APC forums on 9/26/2016

---

I don't think you have to be worried about the battery fluctuating from 99% to 100% state of charge (SoC). The UPS does a weekly self-test to check the battery and if you're really concerned, you could do a runtime calibration and then see if the runtime is too short.

You might consider also plotting battery voltage and load (output or input watts). Also, if the UPS does go to battery, you might not see it on the input voltage plot. Better to look at the event log.

When fully charged, the SoC is based on battery voltage, load and (I think) temperature. Suppose load decreases -- the battery capacity will increase which means it can be charged more which brings state of charge down to 99%.

[rath4snps_apc]

This was originally posted on APC forums on 9/27/2016

---

Hm,

sure, other Batterie Parameters are also monitored. the confusing point:  I can't see such behavior on any other of my UPS devices (all over all ~ 15 pcs.). All show teh charge at 100%, never going to 99%.

It has no relationship to weekly self tests, the flapping between 99 and 100 happens 10-20 times a week in irregular frequency. Sometimes 5 x in 2 days, than nothing in 2 days.

My only idea was so far that the device is hit by some incoming power glitches, which are not recognized by the 5 minutes monitoring frequency. In my eyes, if such a glitch happens, the UPS will immediately switch to battery and stabilize the power, or? Could that be a reason for such behavior?

This is shown in log when battery is on 99%

Date	Time	Vmin	Vmax	Vout	Iout	%Wout	%out	FrqOut	%Cap	Vbat	Tups
09/27/2016	02:54:23	229.72	232.47	229.59	2.78	18.08	21.21	50.00	100.00	109.47	25.00
09/27/2016	02:44:23	230.88	232.83	229.59	2.75	18.03	21.15	50.00	99.00	109.47	25.00
09/27/2016	02:34:23	230.53	232.92	229.67	2.78	18.27	21.23	50.00	99.00	109.47	25.00
09/27/2016	02:24:23	230.69	232.59	229.64	2.78	17.98	21.28	50.00	99.00	109.47	25.00
09/27/2016	02:14:23	230.61	232.55	229.61	2.78	18.08	21.25	50.00	99.00	109.50	25.00
09/27/2016	02:04:23	230.36	232.14	229.64	2.75	18.14	21.22	50.00	99.00	109.47	25.00
09/27/2016	01:54:23	230.34	232.34	229.67	2.78	18.14	21.34	50.00	99.00	109.50	25.00
09/27/2016	01:44:23	230.55	232.33	229.64	2.75	18.00	21.13	50.00	99.00	109.50	25.00
09/27/2016	01:34:23	230.39	232.45	229.64	2.75	17.87	21.23	50.00	100.00	109.50	25.00
09/27/2016	01:24:23	229.86	232.08	229.61	2.75	18.04	21.24	50.00	100.00	109.47	25.00
09/27/2016	01:14:23	229.98	232.23	229.56	2.75	18.02	21.20	50.00	100.00	109.53	25.00

For me all parameter look OK, except of the 99% of Battery. But again the logfile read values in 10 minutes Frequency, and Battery is for at least an hour on 99%?

I need to understand to make sure that in worst case the UPS will cover attached systems.

Thanks for more ideas

regards

Stephan

[voidstar_apc · Accepted solution]

This was originally posted on APC forums on 9/27/2016

---

You'd need to look at the event log if you want to rule out operating on battery. I think it's more likely that the 99% is an artifact of the state of charge calculation. Battery voltage seems fairly stable... 109.47 during the 99% SoC region and 109.50 during the 100% but that pattern might not hold over a larger timespan.