Can I skip the equipment step?

Yes, and you can add equipment later. Plant-level analytics work without it, but per-inverter soiling estimation and fault attribution need inverters defined and matched to their SCADA tags.

Can one plant have more than one data source?

Yes. A common setup is a vendor cloud API for live polling plus a CSV upload to backfill history. Each source gets its own credentials, discovery run, and field mapping, and you mark one as primary.

The connection test fails. What should I check?

The wizard lists any required fields you left empty before it lets you test. Beyond that, the most common causes are portal credentials used where API credentials are required, and a wrong regional endpoint. The vendor guides cover the specifics.

Getting started

Onboard your first plant

The connection wizard takes a plant from a name on a form to live analytics in eight steps.

Quick answer

Onboard a plant from the dashboard with the connection wizard: define the plant, describe its equipment, choose a data source, enter credentials, run discovery, match device tags, review field mappings, and confirm. Polling starts on save, first measurements arrive within about 15 minutes, and provisional analytics appear within a day.

Before you start

You need a role that can create plants: ORG_ADMIN or MANAGER. Creating the data connection itself requires ORG_ADMIN.
Have the credentials for your data source at hand, for example a Huawei Northbound account or a SolarEdge API key. The vendor guides in the connecting-data section cover how to get them.
Optional but valuable: plant metadata such as coordinates, module tilt and azimuth, inverter models, and whether the site has a weather station or DustIQ soiling sensor. The more you provide, the better the day-one physics model.

The eight steps

1. Plant: Name, location, coordinates, altitude, timezone, and rated capacity in MW.
2. Equipment: Inverter groups with tilt, azimuth, model, and MPPT layout, plus weather station and soiling sensor details.
3. Data sources: Pick one or more sources that feed this plant, for example a vendor cloud API plus a CSV backfill.
4. Configure: Enter credentials for each source and run the connection test.
5. Discover: NuraVolt probes the source and finds plants, measurement fields, and device tags.
6. Match devices: Match discovered SCADA tags to the inverters you defined in step 2.
7. Map fields: Review the automatic field mappings and correct anything ambiguous.
8. Confirm: Save and activate. Optionally create a starter performance report for the new plant.

What the discovery step does

Discovery reads a sample from your source and works out what is in it. It lists the plants the credentials can see, proposes a mapping from your field names to NuraVolt measurement types, each with a confidence score, and collects device-level tags such as per-inverter power channels. High-confidence mappings are accepted automatically; anything ambiguous is queued for your review in the map fields step.

Matching device tags to the inverters you defined is what unlocks per-inverter analytics. If you skip the equipment step, NuraVolt still works at plant level, but per-inverter soiling and fault attribution need the device match.

What happens after you confirm

Polling starts immediately, on a 15-minute interval by default, so the first measurements land within about 15 minutes.
Provisional analytics appear within a day: a physics digital twin built from your plant metadata plus a climate-zone transfer model for soiling.
Estimates improve automatically as your own operating data accrues. See the cold start guide for what to expect when.

Frequently asked questions

See this on your own plants

NuraVolt turns your SCADA and BMS data into early fault detection, degradation-aware BESS analytics, and audit-ready reporting. A fixed-scope audit shows you what we’d find on your portfolio.

Book an audit Try the live demo