Power Volumes¶

15-minute granularity volume and price data for EWE's short-term power desk and controlling functions. Two layers are available:

EWE wrapper views — fct_ewe_volume_pwr_* with EWE-specific business-unit and portfolio filters baked in.
Generic numera-core views — fct_volume_pwr_15m[_deal] and fct_volume_pwr_floating_15m[_deal] with no embedded EWE filters; you apply them in your query.

The generic views are the strategic direction. EWE wrappers remain available during the transition.

Migrating from the EWE views?

Start with How to migrate from EWE views for the view-by-view mapping and side-by-side SQL.

Documentation by purpose (Diátaxis)¶


Learning	Doing	Looking up
Getting Started — run your first 15m query	How-to guides — task-oriented recipes	Reference — view-by-view column specs
Migrate Your First Query — hands-on EWE→numera-core walkthrough
		Understanding
		Explanation — concepts, gap-fill, architecture

How-to guides¶

Migrate from EWE views — the canonical migration playbook
Join dimensional views for filtered queries — portfolio, location, BU, counterparty
Compute P&L and net position — VWAP-based analysis recipes
Handle time zones and DST — UTC vs CET, spring/fall transitions
Filter and aggregate efficiently — partition pruning, date-bound patterns
Query floating / index-priced volumes — PWR Index instruments

Reference¶

EWE views — 4 short-term desk + 3 controlling unfiltered + 2 quarterly variants
Numera-core views — 4 generic 15m customer views
Semantic views — sv_power_volumes, sv_power_volumes_floating
Feature flag & deployment — time_series.15m, build gating

Explanation¶

Concepts — 96 periods/day, business case, grid, VWAP vs index price
Gap-filling — why every (combo, day) has 96 rows
Architecture & lineage — from raw arrays to unrolled views
EWE vs numera-core layers — why two layers exist, which to use

"I just want to…" quick lookup¶

Goal	Page
See what columns a view exposes	EWE views or Numera-core views
Reproduce an EWE view's output using the generic view	Migrate from EWE views
Filter by portfolio	Join dimensional views
Filter by counterparty	Join dimensional views
Compute P&L	Compute P&L and net position
Understand why I see zero-volume rows	Gap-filling
Understand why my query is slow	Filter and aggregate efficiently
Handle the spring-forward / fall-back hour	Time zones and DST
Use natural-language queries (Cortex Analyst)	Semantic views

Feature flag¶

All 15-minute models are gated by time_series.15m. If queries fail with "relation does not exist", the flag is off on your stack. See Feature flag & deployment.