Transformer Loading
Estimating distribution-transformer utilization from downstream meter data
Transformer loading in Texture estimates how much of a distribution transformer's rated capacity is being used, by aggregating the telemetry of the meters downstream of it. Distribution transformers rarely report their own telemetry, so Texture derives their load from eligible downstream meter data. Texture selects one reading per consumer element, aggregates those readings through grid topology, and compares the result against the transformer's nameplate rating. Selecting a single reading per consumer element ensures that duplicate series or meter changeouts don't multiply the load.
Why transformer loading?#
Most utilities don't meter every distribution transformer directly. Transformer loading gives visibility into utilization anyway:
- See utilization without new hardware — load is derived from meters you already collect, not from added telemetry on the transformer itself.
- Catch over-utilization early — aggregate load is compared continuously against rated capacity, so trends above the threshold are visible before they become a field issue.
- Surface bad nameplate data — because utilization is only as good as the rating it's measured against, results that aren't physically plausible point back to a rating that needs correction.
How it works#
The system operates through three steps:
- Meter-to-transformer association — grid topology data connects each meter to the transformer it's downstream of.
- Aggregation of downstream load — one selected reading per eligible consumer element is aggregated for the transformer; duplicate series and meter changeouts do not multiply load.
- Comparison against rated capacity — the aggregate is measured against the transformer's nameplate kVA rating to produce a utilization value.
Load is calculated as kVA = √(kW² + kVAR²), using reactive power where it's available, or a configurable assumed power factor (typically around 0.95) when it isn't.
The rated capacity is the gating input
Accurate nameplate ratings from the GIS model are essential — without a reliable rating, Texture can still display load trends, but it can't tell you whether the transformer is over- or under-utilized.
Exploring transformer loading at scale#
Beyond the load curve for an individual transformer, Texture provides a list view for reviewing utilization across an entire transformer fleet — built to stay responsive across tens of thousands of transformers, with server-side sorting and cursor-based pagination rather than loading and sorting everything in the browser.
The list adds columns purpose-built for utilization review:
| Column | What it shows |
|---|---|
| Peak kVA | The highest aggregated kVA magnitude observed in the reporting window, whether import or export. |
| Peak export kVA | The highest aggregated back-feed (export) observed in the same window. |
| Utilization % | Peak kVA expressed as a percentage of the transformer's rated capacity. |
| % time above rating | The share of the reporting window during which aggregated load exceeded the rated capacity. |
The reporting window (for example, a rolling multi-day period) is shown at the top of the view, and any column can be sorted — so, for instance, sorting by utilization surfaces the most heavily loaded transformers across the whole fleet.
Back-feed and data-quality signals#
Back-feed detection. When the largest interval in the window is export rather than load, the view annotates it as such — surfacing back-feed through the transformer, which is typically a sign of downstream generation (such as distributed solar) rather than a metering error.
Implausible-value flags. Because utilization is derived from a nameplate rating that isn't always accurate, the view also flags values that aren't physically credible for a healthy asset — for example, a utilization figure that's many multiples of 100% of the transformer's rating. These flags aren't reports of real overload; they're a prompt to check the rating in the GIS model. Texture applies a hardcoded threshold of 120% utilization — based on industry standard — as the point beyond which a value is treated as implausible and flagged for rating review.
Filtering by substation#
The list view supports filtering down to the transformers belonging to one or more specific substations, using a search-and-multi-select control. This narrows the fleet-wide list to a working set — for example, to find the most heavily loaded transformers within a single substation rather than across the whole territory — without losing any of the sorting, columns, or annotations described above.
Relationship to other concepts#
- Meters — supply the interval energy that is summed.
- Grid Topology & GIS — supplies the meter-to-transformer matching and the nameplate ratings.
- SCADA — provides the substation-level control-system view; transformer loading complements it by giving distribution-transformer visibility where SCADA does not.