R5000 UPSes drop load

[Matt_Hale]

We have inherited a pair of HP R5000 UPSes from another part of our business.  They each came with a network interface modue and a single runtime extension module (aux battery).  We did a full factory reset on them and updated the firmware to the latest available.  When we first put them in service they claimed to be at about 10% load and had a runtime estimate of over 9 hours, but when the power went out they dropped the load and completely shut off after just 5 minutes.

Obviously the batteries were expired, so we diconnected the extension modules and bought a fresh set of HP-branded internal battery packs.  With just those internal batteries, and some additonal load up to about 33% they now calculate ~26 minutes of run time, but again drop the load and shut off entirely after ~7 minutes.  We left them in this state for several weeks but nothing improved.

Frustrated, we bought a third set of internal batteries from a different supplier (exactly the same type as used in the genuine HP battery packs) and have exactly the same result: they calculate 26-27 minutes run time at 33% load but die completely and without warning at roughly 7 minutes.

We could live with 7 minutes uptime if they would calculate runtime correctly and signal the attached servers to shut down gracefully, instead of just suddenly dropping dead.  The battery test function produces either 'abort' or 'OK' as a result but does so instantly it doesn't seem to actually test anything.

How do we calibrate these things so that they have an accurate time estimate?  Our similar APC UPSes have a calibration test that can be triggered manually but I don't see any corresponding function either on the web interface or the front panel.

Thanks.

[AshutoshM]

Suggestions:

1. Ensure that the batteries are charged for at least 48 hours before testing.

 https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c02945229

2. Install and configure the HP Power Protector software to initiate a graceful shutdown on servers when power lost.

https://support.hpe.com/hpsc/swd/public/detail?swItemId=MTX_847b7148e45846848bc3e6c063 

Also review the below advisory and ensure the firmware at the same level or higher as mentioned in the advisory.

https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c03906556

[Matt_Hale]

Thanks for the suggestions.

1.  These batteries have been charged for weeks.

2. I dont think this is an external software issue - they are configured for use with SIM, which receives a notification that they are switching to batteries when the power source is removed, but does not receive anything when they die.  It is not a matter that the UPS is switching off the load segments, the entire UPS including the management card drops dead without warning when it is still estimating about 20 minutes of runtime remaining.

3. The management card frimware is already at 1.10.001 - well beyond what is recommended in that advisory.

Again , comparing to our APC UPSes, they will only conduct a battery life test when they're at 100% charge.  These HP UPSes never reach 100%.  They achieve 95% and the status switches to 'Floating' then they eventually reach 99% and it changes to 'Resting'.  Is it possible they will not recalibrate because, like APC, they want to see 100% charge before the test is run?

[AshutoshM]

Recommend testing with the Power Protector Software to configure graceful shutdown and also will help check the log files.

Alternately raise a support case with HPE to have this looked into in depth.