An array that produces less than the sun allows does not raise an alarm. The inverter reports MPPT, the meter turns, and the shortfall hides in a monthly total. Array Yield polls each inverter's SunSpec models over Modbus for AC power, lifetime energy, per-input DC power, cabinet temperature, and operating state. It reads hourly irradiance and air temperature for the site from NASA POWER or Open-Meteo, works out what the array should have produced, and keeps the gap per inverter and per input.
When the gap holds past your tolerance for the run of days you set, it drafts a work order in your CMMS: cleaning when the whole array dulls together, inspection when one input trails its siblings, fault when the inverter says so. A maintenance lead approves the draft. Nothing is written to an inverter, ever.
This is a reference listing. It documents what Fibric would read from Array Yield and what it could propose, based on the vendor's published interfaces. Fibric builds it under a managed deployment when you request it; selecting it here installs nothing.
Inputs
AC power, lifetime energy, DC power, and power factor from SunSpec inverter models 101 to 103 and 111 to 113: W, WH, DCW, PF
Operating state St per inverter: OFF, SLEEPING, STARTING, MPPT, THROTTLED, SHUTTING_DOWN, FAULT, or STANDBY, and the Evt1 event flags
Cabinet and heat sink temperature, TmpCab and TmpSnk, from the same model
Per-input DC current, voltage, power, lifetime energy, and state from the Multiple MPPT Inverter Extension Model, model 160, where the inverter offers it
All-sky and clear-sky surface shortwave irradiance in Wh/m² and temperature at 2 metres by hour from NASA POWER, from 2001 to near real time
Global tilted irradiance for the array's tilt and azimuth, and total cloud cover, by hour from Open-Meteo, with up to 16 days of forecast
Open work orders on the array's asset in your CMMS, so a lag already assigned is not proposed a second time
Proposed actions
Target capability: propose a cleaning work order in your CMMS when the array's energy lags expected yield past your tolerance for the days you set
Target capability: propose an inspection work order naming the inverter and MPPT input when one input's DC power trails its siblings under the same sky
Target capability: propose a fault work order when St reads FAULT or an Evt1 flag is set, with the flag named in the draft
Target capability: propose a note on the open work order with the yield before and after, once production returns to expected
Proposed actions are target capabilities. Every action runs propose-first and needs a validated deployment and the appropriate permissions.
What you can build
Send a cleaning crew when the whole array dulls
Every inverter on the roof produces less than NASA POWER's irradiance predicts, and the gap has held for the days you set. The operator drafts one cleaning work order in MaintainX naming the array asset and the gap.
Model 160 shows one MPPT input's DC power trailing the others on the same inverter under the same sky. The operator drafts an inspection work order in Fiix naming the inverter and the input.
Sites without Modbus reach report through SolarEdge or Enphase site energy instead. The operator runs the same comparison per site and drafts work orders in UpKeep, within each vendor's rate limits.
Production fell, but Open-Meteo's cloud cover and the NWS forecast say the sky did too. The operator records the day as weather and proposes nothing, so Limble never fills with work orders for clouds.
Inverters that speak SunSpec Modbus over TCP or RTU, reachable from a gateway, found by the SunS marker at base address 0, 40000, or 50000
A CMMS connector to receive the work order: MaintainX, Fiix, UpKeep, or Limble, with each array registered as an asset
An irradiance source: NASA POWER hourly at the site's coordinates, or Open-Meteo with the array's tilt and azimuth set
A tolerance and a run of days set by you before a lag becomes a proposal; the operator ships no default
Authentication
Reads inverters through the Modbus reach of your gateway or the vendor API credentials you attach, and irradiance from public feeds. It writes only to the CMMS, only after approval.
Limits
Expected yield comes from a gridded irradiance model, not a rooftop pyranometer. A local sensor read over Modbus, where you have one, replaces it
NASA POWER hourly data runs to near real time; the day just ended may not be in it yet. The comparison waits rather than guessing
A string shaded at a fixed hour looks like a dirty one. The operator names the lagging hours; the person who approves makes the diagnosis
Open-Meteo's free tier is non-commercial, under CC BY 4.0, and capped below 10,000 calls a day; a commercial site needs a subscription
Access and pricing
Reference listing. Fibric builds the operator under a managed deployment when you request it. Your quote covers the build, capabilities, usage, and support.
A work order draft: the asset, the inverter and input, measured energy against expected for the days in question, the irradiance and temperature used, the inverter's operating state, and the reason class. The lead approves, edits, or declines. One approved draft becomes one work order.
What record is left after a work order is created?
The work order in your CMMS carries the gap, the hours it lagged, and the readings behind it. The operator keeps the draft, who approved it, when it was created, and the yield before and after, so a cleaning that changed nothing is visible.
Does it ever change the inverter?
No. It reads SunSpec models and never writes a register. Curtailment, limits, and settings stay with your installer and the vendor tools. Its only write is a work order, and only after a person approves it.
Ask about Array Yield
Ask about the capabilities and requirements in this listing.
This operator is developed, published, and supported by Fibric. Third-party names and logos identify the systems an integration connects to; they are the property of their respective owners, who are not affiliated with Fibric and do not sponsor or endorse this listing. Trademark policy