PM8000 range and FCAS metering

[alannavial]

My customer, owner of service stations in Australia, has some questions below, please help me to answer these:

I am doing some due diligence on the PM8000 range and I see that Schneider has released a manufacturer declaration of conformity for IEC 61557-12:2007.  Is this valid for all PM8000 meters, or just specific versions?

Is there an intent to update this declaration to the latest 2018 version of the standard?  Or is there some aspect of the 2018 version the PM8000 does not meet?

With regard to the declaration, how is this impacted by CT/VT choices?  Specifically around the frequency measurement accuracy.  The MASS calls for less than or equal to 0.01Hz intrinsic uncertainty for local frequency measurements…..if we take the Class 0.2 accuracy for frequency as per the declaration, this is OK, but I expect this changes if you are using VT’s, but is not impacted by the accuracy of CT’s?

Are there specific versions/options of the PM8000 that are required to enable participation in FCAS?

I assume that as the PM8000 is storing the frequency values as 32bit Floats, we can meet the 0.0025Hz resolution requirement for both read and stored frequency measurements.

Lastly, when taking highspeed data recordings, can we:

• Trigger off a number of different frequency thresholds within 50ms?
• Record frequency and power at 50ms for 5 seconds before the event, 10 seconds during the event, and 60 seconds after the event (75s total)?
• How quickly can the PM8000 recover and start recording the next event (flush high speed registers, reset data recorder, etc)?
• Is the above impacted by the method of communications to the meter?
• What happens when a number of events occur in quick succession, say at 2s after event, 15 seconds after, 60 seconds after, etc)?

[Charles_Murison · Accepted solution]

Hello @alannavial ,

I cannot answer all your questions but I can speak to some.

If declaration of conformity was limited to specific versions of PM8000 meters, the limitation would be included with declaration.

Regarding FCAS, have to looked at https://community.se.com/t5/Metering-Power-Quality/PM8000-Frequency-resolution/m-p/406158#M6418
https://community.se.com/t5/Metering-Power-Quality/ION9000-FCAS-window-monitoring/m-p/403065#M6360

Regards,

Charles

[alannavial]

Hi Charles,

Thank you very much for the helpful links.  I think that answered most of my customers questions, except now they want to know about the recovery time after an event, see below:

The last piece of the puzzle is the recovery time after an event and what happens during multiple events.  I’ve had experience with similar spec (non Schneider) meters that take significant time to recover the HS registers to trigger another recording, etc.  It’s pretty important to understand the limitations, so that we can be aware and manage the impacts.

[Charles_Murison · Accepted solution]

Hello @alannavial ,

This really depends on the settings inside the dat recorder. There are multiple settings that do different things. For more information recommend you look at the ION Reference guide.

Depth is the total amount of data the recorder can keep in memory before either FIFO cycled out old data or recorder is full and needs to be reset. This is used the size the memory in non-volatile memory

Buffer depth is the number of records to keep in RAM (volatile) memory, This will be some multiple of 1+ # records before trigger + # records after trigger. All data is written to internal buffer before being transferred to long term storage. The number of record sets determines how many triggers before the buffer is full. The recordings can be triggered back to back, however there is some limitations on how fast the meter can transfer to non-volatile memory.

Say for example to have 2 records before the trigger and 7 records after. The total set would be 1+2+7 records. If the buffer depth was 20, you could have 2 triggers before the buff was full. If the depth was 500, you would have 50 triggers. More triggers, more time to move data from buffer to non-volatile memory.

There are some limits to buffer depth, this buffer is shared with all data and waveform recorders.

Before you ask, no I cannot give you any timings for transferring data as this will greatly depend on what else is happening on the meter.

Regarding triggering during a bust logging in progress I do not know for sure the meter's behavior.

Regards,

Charles