Snapshots failing with Duplicate DBT_SCD_IDs

The problem I’m having

My snapshots are failing due to duplicate rows already existing in the snapshot table.
However, when I query the table, I see duplicate dbt_SCD_IDs.
My source data does not have duplicates.
I’m using a check_col strategy, and the columns are not changing.

What I’ve already tried

I’ve ensured I have no duplicates in the source.
I’ve ensured the the columns are called out that I want to check.
I’ve previously tried the timestamp method and it also failed.

Same here, using BigQuery. It may be worth noting that columns not within the check_cols condition ARE changing, but we don’t want changes in those columns to constitute a new row version.