Symmetra PX: Unstable battery voltage when on battery (reported)

[dakotawallender]

We are seeing decreased runtime compared to what should be available, however during a runtime calibration, i have noticed that the reported battery voltage is unstable. I have verified with a multi-meter that both positive and negative side of the battery bus have stable voltage when measured externally. We have 2 strings of battery in the unit and both strings are new. I will attach a video but the battery voltage can be see swinging from 195.9-197.1 in under 1 minute. the battery voltage never moved from 197.6 on both sides of the bus when measured. this is 15 minutes into the runtime calibration. 

The calibration did not complete successfully and reported canceled via NMC2 after 2hrs 40mins of runtime. I suspect when the battery voltage is reported as falsely low the calibration is ended early. Any insight would be great. 

Thank You.

[Teken]

Based only on the information you have provided the battery calibration failed because you didn’t follow APC guidelines during the procedure.

The image capture shows 17% load vs the minimum of 30% or greater.

As it relates to the battery voltage on the LCD vs what you measured. I’m unclear as to what you’re saying?!?

The displayed voltage is wrong? The DMM is more accurate in terms of measured battery voltage?!? Other??

Lets just ignore everything for now and state some basics as it relates battery voltage and battery calibration for the benefit of others who may want to know.

DMM: It’s assumed the multimeter has been sent out for an annual certification of calibration and meets its specifications. 

If not, you’re simply guessing and using a tool that is injecting variance where it may not exist. All serious installations have and use a reference source for AC, DC, Resistance, Current, Capacitance.

These reference tools allow you to have some level of confidence of in calibration / conformity while in use and before the next certification.

Is the Fluke DMM still in calibration?!? Did you validate the DMM to a reference tool and has that same reference tool been validated and sent out for certification of conformity?!?

As we are talking about tenths of a volt which does matter.

Next, any measurement is impacted by temperature . . . 

In ideal conditions the ambient temperature and environment falls within the range of 20-25’C and compensated for every degree as outlined by the battery makers testing guidelines.

Next, where you take the voltage measurement also determines what value(s) you see and obtain. The measurement you take using the DMM is for all intents is live, immediate, point in time.

If you take the measurements at one end vs the entire string the values will be different. If you were to disconnect every cell and measure them separately you would quickly see they all have different voltage levels especially after 72 hours of being disconnected and measured again.

If there is one bad / weak cell in the string it will take down the entire RBC.

The readings you see on the LCD Menu are measured from only a specific point. Ideally it would measure the entire string vs at a specific point / location.

The value(s) may also be averaged (RMS) and the sampling interval is delayed.

This also doesn’t take into account how the data is refreshed on the screen which is separate and independent of the actual measured voltage reading.

Battery Calibration: During the battery calibration multiple DMM’S would be in place along with a clamp meter. All of the data would be logged and recorded and displayed as a graph.

Why?!?

Because a static voltage measurement can not provide you any context or long term history as it relates to time & performance.

Doing this also provides you independent measurement and data from the main system for validation purposes.

Battery calibration should be done when a new RBC is installed. The overall intent is to teach the system about the new RBC so the displayed operational runtime is as accurate as possible via the LCD, NMC, SNMP, etc.

This same test along with a Lights Out / Grid Down validation test is conducted on an annual basis.

Keeping in mind both procedures are extremely stressful and negatively impact the service life of the RBC! 

The RBC is rated for 260 full depletion cycles before degradation is seen. As such this impacts the overall service life of 2-5 years which is directly affected by usage, temperature, environmental conditions.

Both tests are conducted to determine and validate the operational runtime, system health, and meet specific ISO Standards of compliance.

If your site has dirty power and the system engages (usage) this impacts it’s useful (2-5) year service life.

Regardless of all the above voltage measurements are not used as the only standard of compliance or RBC health!

A conductance and internal resistance test is the industry gold standard. As both tests are more accurate and do not negatively affect the battery cells.

Conducting a load test via a carbon pile for bulk testing or electronic load testing are other methods to determine how individual cells / strings are performing.

This also negatively impacts the health and service life of the RBC / cells.

Lastly, once these validation tests have been conducted the system is tested in a static & dynamic state to determine the true operational performance and health of the system.

You may search this forum and the net to understand what this means, how it’s conducted, and the importance of doing so.

Questions Ask . . . 👍 

[dakotawallender]

Regardless of the multi-meter reading the unstable battery voltage shown on the video (lcd) and the data logs is abnormal. a large battery pack does not swing up and down on voltage regardless of its health with a constant load..

This is a problem on the UPS end, it could not be related to the calibration being cancelled however it is a problem or a warning sign either way.

[Teken]

Provide the following logs for review: event, data, ups  . . . 

What is the standby (floating) voltage when the RBC is 100% SOC?!? What loads are / were connected during the battery calibration?