Estimations

Where Net Purpose’s research indicates that a company provides a good or service which delivers measurable impact, the magnitude of that impact will be calculated using lifecycle assessment. These estimations are computed in a transparent way following a standardised framework of datapoints related to each other to drive from either revenue or output to outcomes and impact. For any given estimation, not all roles in the estimation will be populated, for example some estimations can start at Outcome level 1, it is dependent on the theory of change and the availability of data.

Each estimation can have one or more branches. These branches have an impact_theme and represent the multi-dimensional impact a given solution can have. For example, when considering Wind power as a solution, it has an impact on climate due to the avoidance of CO2e. It also, however, impacts Air quality as there is a consumate reduction in particulates such as NOX, SOX and PM2.5. These different branches capture this multi-faceted behaviour.

Each estimation is represented by a set of estimation_lifecycle rows. These row represent a singular node in the lifecycle graph, with the underlying value, source material and derivation of that node being the linked datapoint and datapoint_derivations. The provided lifecycle_index column exists to aid correct sorting of the rows to provide legible estimations.

All views on this page are versioned — filter is_current = TRUE (see Conventions).

`datapoints`

One row per estimated value. When datapoint has required derivation there is a formula (expressed as excel like manipulations of other datapoints) and the populated ast. These are provided for transparency purposes on the underlying calculations.

Column	Type	Description
`datapoint_id`	VARCHAR (UUID)	Primary key
`type`	VARCHAR	Kind of datapoint within the model.
`is_derived`	BOOLEAN	`TRUE` when computed from other datapoints, `FALSE` for model inputs
`question_id`	NUMBER	What the value answers. Joins to `questions`
`canonical_question_id`	VARCHAR	The question’s readable slug, for convenience
`value`	FLOAT	The value in the standard unit
`unit`	VARCHAR	Name of the standard unit
`reported_value`	FLOAT	The originally reported value, where the input came from a disclosure
`reported_unit`	VARCHAR	Unit of the reported value
`source_id`	NUMBER	Source document for reported inputs. Joins to `sources`
`page_num`	NUMBER	Page of the document

`datapoint_derivations`

The formula behind each derived datapoint, for full transparency on how an estimate was computed.

Column	Type	Description
`datapoint_id`	VARCHAR (UUID)	The derived datapoint
`formula`	VARCHAR	Human-readable formula
`ast`	VARIANT	The parsed formula tree (machine-readable)

`estimation_lifecycle`

Ties datapoints together into an estimation chain for one entity, taxonomy activity and year. The taxon_id points at the modelled activity — a technology-level node in the taxonomy, for example a renewable energy generation technology whose outputs and impacts we model.

Column	Type	Description
`entity_id`	NUMBER	Who. Joins to `entities`
`taxon_id`	NUMBER	The modelled activity. Joins to `taxonomy`
`year`	NUMBER	Fiscal year
`lifecycle_role`	VARCHAR	The role this datapoint plays in the chain (`ouput`, `outcome_level_1`, etc.)
`lifecycle_index`	NUMBER	Position of the step within the chain
`impact_theme`	VARCHAR	Impact theme the chain contributes to, where applicable
`datapoint_id`	VARCHAR (UUID)	The datapoint at this step. Joins to `datapoints`

Example: Get all lifecycle estimations for a given entity

SELECT
  el.entity_id,
  el.taxon_id,
  el.year,
  el.impact_theme,
  q.name,
  d.value,
  d.unit
FROM NET_PURPOSE.FOUNDATIONS.ESTIMATION_LIFECYCLE el
JOIN NET_PURPOSE.FOUNDATIONS.DATAPOINTS d
  ON el.datapoint_id = d.datapoint_id
JOIN NET_PURPOSE.REFERENCE.QUESTIONS q
  ON d.question_id = q.question_id
ORDER BY year desc, el.lifecycle_index asc