Emissions are deleted when refreshing separate data imports in Microsoft Sustainability Manager

Eiber Alejandro Torres Galvan 0 Reputation points
2026-08-14T14:27:39.42+00:00

Hello,

I have two separate activity data imports in Microsoft Sustainability Manager:

  • Stationary combustion
  • Purchased electricity

The activity data is loaded correctly for both imports and remains available in the corresponding activity tables.

However, when I refresh Stationary combustion, its emissions are generated, but previously generated Purchased electricity emissions disappear from msdyn_emission. When I refresh Purchased electricity, the opposite occurs: its emissions are generated, but the Stationary combustion emissions disappear.

Only the generated emissions are affected. The underlying activity records remain loaded correctly.

In Synapse Link, the missing emission rows do not appear with IsDelete = 1; they are completely absent from the replicated msdyn_emission table.

Both calculation profiles were initially configured with Automatically run this calculation when data is refreshed.

Could the automatic calculation be replacing emissions for the same organizational unit, reporting period, or calculation boundary? Should automatic calculation be disabled and both calculation profiles executed sequentially after all imports have completed?

Also, is there a supported method or audit table that can identify which calculation job removed or replaced the emissions?

Thank you.

Microsoft 365 and Office
Microsoft 365 and Office

A comprehensive suite of productivity tools and cloud services that enhance collaboration, communication, and efficiency. Combining classic Office apps with advanced Microsoft 365 features, it supports both personal and business needs

1 answer

Sort by: Most helpful
Most helpful Newest Oldest

  1. AI answer

    2026-08-14T14:27:57.87+00:00

    Yes. The documented behavior matches what is happening.

    For each calculation profile, when a new calculation is run, the previous calculation errors and the emissions related to the activities are deleted. That means running one profile can remove previously generated emissions before recalculating its own output.

    Also, if calculations are configured to run automatically after data ingestion, each connection refresh can trigger a calculation job as soon as that import finishes. The documented guidance is to separate imports and control the order when calculations run automatically, because calculations can start before all required data is in place.

    Based on that, the supported approach is:

    1. Review each calculation profile in Calculation profiles.
    2. Edit the profile and check whether Automatically run this calculation when data is refreshed is enabled.
    3. If the goal is to keep emissions from multiple activity imports available together, disable automatic calculation on the separate imports.
    4. Complete all required data imports first.
    5. Then run the calculation profiles manually in sequence from Calculation profiles by selecting the profile and choosing Run calculation.
    6. After each run, use History on the profile to review the associated calculation jobs, last run time, success rate, and any errors.

    For tracing what removed emissions, the supported visibility in the documentation is the calculation profile job history:

    1. Go to Calculation profiles.
    2. Select the profile.
    3. Open History in the right pane.
    4. Select the most recent calculation job to view error codes and line-item details when available.

    The provided documentation does not describe a separate supported audit table that identifies which calculation job deleted or replaced rows in msdyn_emission.

    One more relevant point: after existing data is deleted and new data is imported, calculations must be run again so emissions are recalculated for the new activity data.

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.