Skip to content

Shapes & profiles

Time-series behavior is driven by shapes: named multiplier curves stored with the circuit and managed in the Shapes tab of the bottom panel. They compile to OpenDSS LoadShape objects and are saved inside your .oneline.json project (and exported inline in .dss files).

The library has two kinds, in two sub-tabs:

  • Load shapes — demand multipliers. Assignable to loads (scales the rated kW each step) and to storage dispatch in follow mode (positive = discharge, negative = charge).
  • Irradiance shapes — solar multipliers. Assignable to PV systems, where they scale the base irradiance parameter each step.

Element dropdowns are filtered accordingly — loads see load shapes, PV systems see irradiance shapes, storage may follow either. Assigning the wrong kind (possible via imports) raises a validation warning, and referencing a shape that doesn't exist is an error.

A shape's points can be at 60- or 15-minute intervals. The same shape drives both daily and yearly runs: a 24-point daily shape repeats across a year, and a full-year shape is read from its start in a daily run. OpenDSS interpolates between points when the run's step size is finer than the shape's interval.

Creating and editing shapes

New creates a flat 24-point shape in the current tab; Duplicate, Delete, and inline rename work as expected — deleting or renaming a shape updates or clears every element that references it.

  • CSV paste — paste values into the text area and load them: one value per line, or comma/semicolon/whitespace-separated, or two-column time,value rows (a monotonically increasing first column is treated as a time axis and dropped). A single header row and # comment lines are tolerated.
  • Curve editor — shapes up to 96 points (a 15-minute day) get draggable point handles on the preview chart; each drag is one undo step. Larger shapes show a read-only preview and are edited via CSV.
  • Normalize — scale the shape so its peak is 1.0, or its average is 1.0.
  • Interval — 60 or 15 minutes per point.

NREL/NLR building load profiles

Import from NREL… (under Load shapes) pulls real building-stock demand curves from the End-Use Load Profiles for the U.S. Building Stock dataset — simulated against actual 2018 weather:

  • Sector & building type — 5 residential (ResStock) and 14 commercial (ComStock) types, from single-family detached to hospitals and warehouses.
  • Climate zone — ASHRAE/IECC zones (1A–7B residential, through 8 commercial); pick the one covering your service territory.
  • Resolution — hourly (8,760 points, the default) or native 15-minute (35,040 points).
  • Normalize — peak = 1.0 means the load's kW rating is its peak demand; average = 1.0 means the kW rating is its average demand (peaks will exceed the rating). Peak-normalized is the conventional choice.

The first fetch of a given zone/type downloads a 10–30 MB CSV from the public dataset (no account needed); it's cached on disk after that, so subsequent imports are instant.

NSRDB irradiance

Fetch NSRDB irradiance… (under Irradiance) pulls hourly global horizontal irradiance for any location in the Americas from the NLR National Solar Radiation Database — deliberately for weather year 2018, the same year the building load profiles were simulated against, so PV output stays physically correlated with demand in yearly runs (the same cold snap, the same cloudy week).

  • Location — type a place name and search, or enter latitude/longitude directly.
  • Scaling — kW/m² (÷1000) keeps the shape in physical units: leave the PV system's irradiance parameter at 1.0 and the shape peaks around 0.7–1.0 depending on climate (recommended). peak = 1.0 normalizes the shape and the import message tells you what to set the irradiance parameter to instead.

NLR API key

The NSRDB requires a free API key from the NLR Developer Network (takes a minute). The app prompts for the key and a contact email on the first fetch and remembers them in your browser only; they are sent to the NLR API per request and never stored server-side. Fetched data is cached on disk per location.

Under the hood

Shapes are emitted before any element that references them, as new loadshape.<name> … commands. Small shapes are inline; shapes over 288 points are passed to the engine as CSV side files (very long inline commands destabilize the DSS parser), while .dss exports always inline everything so an exported file stays a single portable script.