incremental refresh and data partitioning in pbi service

[ThiagoShiverzz]

Good morning everyone!

I have a dataset configured to store 1 year of archive data and import only the current day (increment).
This dataset connects to an external database where the data source "only has the current date." My idea is to create a history of this table's data in PBI, but strangely, PBI is deleting previous days, even though the incremental load is configured to only fetch the current day's data.
Why does this happen? Even though the increment is for the current day, does it recheck to see if the data from previous days still exists in the source?
I've done several tests, but it keeps deleting the imported data from previous days, as if it were a full load.
Technically, why does this happen?
At what point does PBI append the data from the increment partition to the archive partition?

[v-prasare]

We would like to confirm if you've successfully resolved this issue or if you need further help. If you still have any questions or need more support, please feel free to let us know. We are more than happy to continue to help you.

Thank you for your patience and look forward to hearing from you.
Best Regards,

Prashanth Are

[v-prasare]

Hi @ThiagoShiverzz,

We would like to follow up to see if your query got resolved? Please let us know if you need any further assistance.

Thanks,

Prashanth Are

MS Fabric community support

[v-prasare]

Hi @ThiagoShiverzz ,

We would like to follow up to see if the solution provided by the super user resolved your issue. Please let us know if you need any further assistance.

@powerbidev123 , thanks for your prompt response.

Thanks,

Prashanth Are

MS Fabric community support

If our super user response resolved your issue, please mark it as "Accept as solution" and click "Yes" if you found it helpful.

[powerbidev123]

Hi @ThiagoShiverzz 

What you're observing is a common misunderstanding with how incremental refresh works in Power BI, especially when configured on import mode with archive + increment logic.

Yes — Power BI does revalidate and potentially delete historical partitions if it cannot find the corresponding data in the source anymore, even if they are supposed to be archive partitions.

What’s Actually Happening Technically

When you configure Incremental Refresh in Power BI (Import Mode), here’s the typical flow:

 During a scheduled refresh:

For "Increment" Range (e.g., today):

Power BI queries the source for only today’s data.

That result is appended to the existing table, replacing the current day's partition.

For "Archive" Range (e.g., previous 364 days):

Power BI may requery archive partitions depending on the config or certain triggers (like schema change, or "detect data changes" logic).

If the data source returns no data for those date ranges, Power BI assumes those partitions are now empty and deletes them.

Conditions That Trigger Historical Deletion:

Data source no longer contains old data (e.g., only has current day's data).

Detect data changes is enabled and Power BI cannot validate partition data.

Data source query fails or returns empty for archive partition — even if the partition exists in Power BI.

So, if your source only has today’s data, then Power BI cannot validate partitions for the previous days and assumes they are gone → it deletes them.

 At What Point is Data Appended?

During refresh, each partition (based on date range) is queried and rebuilt individually.

If partition range = today → only today is queried.

If other partitions are re-evaluated and return no data, they are considered obsolete and removed.

Appending does not mean “just stick today’s data onto the end of existing file.” It means:

Each partition (e.g., one per day) is refreshed individually.

Only if Power BI successfully queries and returns data for that partition does it keep it.