# Output format | CUR enrichment reference
Source: https://greenpixie.com/docs/reference/cur-enrichment/output-format

The columns, format and location of your enriched data.

## What comes back [#what-comes-back]

Enrichment returns your billing data with sustainability columns appended. Every original column is preserved exactly as provided, and the row count never changes: enrichment only adds columns.

## Added columns [#added-columns]

| Column                                            | Description                                                      | Unit        |
| ------------------------------------------------- | ---------------------------------------------------------------- | ----------- |
| `estimate_category`                               | Greenpixie's category label for the line item                    | n/a         |
| `usage_electricity_consumption_kwh`               | Electricity consumption, including data centre facility overhead | kWh         |
| `usage_emissions_tonnes_co2e`                     | Usage (operational) emissions                                    | tonnes CO₂e |
| `manufacturing_emissions_tonnes_co2e`             | Embodied (manufacturing) emissions                               | tonnes CO₂e |
| `total_combined_emissions_tonnes_co2e`            | Usage + embodied emissions combined                              | tonnes CO₂e |
| `water_consumption_cooling_litres`                | Water consumed cooling the data centre                           | litres      |
| `water_consumption_electricity_production_litres` | Water consumed generating the electricity                        | litres      |

How each value is calculated is documented in the [Cloud methodology](/methodology/cloud); see [Assessment and Calculation](/methodology/cloud/assessment-and-calculation) for the calculation components behind each column.

## File format [#file-format]

Output files are written in the same format as your input files (`.csv`, `.csv.gz`, `.parquet`, `.snappy.parquet`).

## Where output lands [#where-output-lands]

Enriched files are written to your `output` directory, mirroring the structure of your input data:

```text
// [!code word:input]
// [!code word:output]
<bucket_container_name> / input  / <platform> / <account_id> / <year_month> / billing-part-0000.parquet
<bucket_container_name> / output / <platform> / <account_id> / <year_month> / billing-part-0000.parquet
```

The same mirroring applies to the [partner directory structure](/reference/cur-enrichment/format-and-structure#directory-structure).