Field Parameters

Hi powerbiexpert22​,

A semantic model full refresh should not recreate or remove an existing field parameter. The field parameter is part of the semantic model definition, while the refresh only updates the data.

So if the field parameter already exists, a normal/full data refresh should leave it as it is.

The main thing to check would be whether the fields or measures used by the parameter have been renamed, removed, or changed. Also, if the semantic model itself was replaced or redeployed, that's a different situation.

You can test it with a full refresh and check whether the field parameter still behaves as expected afterward.

If it works in your case, please mark the answer as the solution so others who have the same question can benefit from it.

Hi powerbiexpert22​ 

Semantic model refresh wont change existing field parameter. It is calculated table with NAMEOF DAX so refresh recalculate and return same rows with no change of slicers and visuals. It can break if referenced field is renamed from Tabular Editor or deleted. You can test after model change

Hello powerbiexpert22​,

No. A semantic model full refresh should not recreate an existing field parameter.

Refreshing just pulls in new data, it won't delete your parameter. I will only dig into it if something looks off afterward, like a field getting renamed or removed at the source. That's usually what breaks it, either the parameter itself or the visuals built on it.

So if you are only doing a full data refresh, the existing field parameter and its slicer selections should remain. If the model was replaced or republished with a different definition, that's a separate case.

Microsoft docs:
Use field parameters to let report readers change visuals

Data refresh in Power BI

Hi powerbiexpert22​ ,

adding to what ShahRukhSameer said: a full data refresh (via Refresh Now, scheduled refresh, or Refresh API) only reprocesses the data tables, field parameters are metadata/calculated table objects defined in the model schema, so they persist through normal refreshes.

Where field parameters do get affected or lost:

• Republishing the .pbix from Desktop overwrites the entire model definition, including field parameters, if the new version doesn't include them.
• Deployment pipeline promotions can overwrite field parameters if the target stage's model differs from source.
• XMLA/Tabular Editor scripted deployments will drop field parameters if they're not included in the deployed schema.
• Renaming or removing the underlying fields/measures referenced by the parameter breaks it, even though the parameter object itself still exists.

So the safe rule: scheduled/manual data refresh = safe, any full model redeployment = check first.