Changelog
Every change to Mirox, grouped by month. Filter by group or by kind of change.
November 202611
Detected capacity no longer reads above the nameplate
A plant's detected capacity no longer reads above its nameplate when a group of shaded strings at half current makes healthy single strings count as parallel rows. For affected plants the detected capacity moves down toward the nameplate. Plants receive this with their next agent update.
A plant controller can now be added as a data logger
A plant controller can now be added as a data logger, with the on-site agent reading the power, energy, operating state, string positions and fault codes of the plant's string inverters. Plants receive this with their next agent update.
A device's web page over remote access loads when its stored port answers only with an error. A device whose web interface port was stored wrong needs one scan of its web interface first.
- A restarted on-site agent connects to the platform at once instead of waiting up to 3 minutes.
A stopped unit's fault event closes once the unit produces again
A stopped unit's fault event previously stayed open for a day or more after the unit recovered. It now closes once the unit produces again. Plants receive this with their next agent update.
The components page no longer marks a string as not connected
A string that produced no energy on one day but energy within the last 30 days is no longer marked as not connected. The not-connected mark now uses the same 30-day energy window as the "probably not connected" badge on the components page. Plants receive this with their next agent update.
In a plant's event list, the no-communication summary updates again.
Device faults standing for days are raised once, not every night; a silent tracker raises its event after 20 minutes.
- A device fault that stands for days is now raised once, not each night.
- A silent tracker now raises its no-communication event after 20 minutes, not 15.
- A condition the on-site agent retracted no longer announces a resolution.
The month-to-date yield no longer counts the previous day twice
A solar data logger's month-to-date yield previously counted the previous day's production again on the first day of a new month, inflating the lifetime energy sum. The repeated production is no longer published, and figures already issued keep their value. Plants receive this with their next agent update.
False offline and unreachable alerts, false fault events and false data holds no longer fire on loggers, inverters and battery historians. An inverter fault that re-raises after its first raise now opens its fault event again.
- A plant no longer reads offline after a brief network link drop.
- A data logger's month-to-date yield no longer opens a false mismatch on a late step.
- A data logger is no longer alerted unreachable while the plant's network link is pending.
- A data logger outage is now reported once per episode.
- A battery historian's cell voltages are no longer withheld as "not fed".
In a plant's event list, DC measurement-defect, tracker and battery events no longer repeat. In an export, frozen lifetime-energy totals are not shown.
- In a plant's event list, DC measurement-defect descriptions now name the faulty DC reading.
- In a plant's event list, tracker silence and fault-stop events no longer repeat at night.
- In a plant's event list, a battery unit at rest no longer posts a state change every night.
- A data logger whose device list could not be read no longer reports its inverters as missing.
- In an export, a lifetime-energy total frozen at its maximum value is no longer shown.
October 202635
Wind plants now show their production state on the plant monitor
A wind plant's production state on the plant monitor is now computed from the sum of its turbines' power. It no longer reads unknown with a standing warning. Plants receive this with their next agent update.
An open inverter alarm on the plant's Events page keeps its start time when the on-site agent updates. A self-clearing inverter alarm no longer opens a device fault each night.
- On the plant's Events page, an open inverter alarm no longer closes and reopens with a reset start time when the on-site agent updates.
- A self-clearing inverter alarm no longer opens a device fault each night on the plant's Events page.
- On the plant's Events page, inverter alarm start times from a data logger that records local time now show in the plant's timezone.
- A mismatch event between a data logger's counter and its power reading no longer assumes the counter is wrong.
- A solar data logger's live data no longer stops for minutes while the on-site agent reads its history.
- On the plant's Events page, a warning event now names an inverter whose data read failed three times in a row.
An inverter's lifetime energy figure no longer freezes once its counter passes 65,535 kWh. Inverter fault items now show the inverter's own alarm code.
- Inverter fault items now show the inverter's own alarm reason from its fault table.
A wind turbine's values no longer stop after its SCADA restarts. A solar data logger's DC values no longer go missing after its login expires.
- A wind turbine's power, state of charge and frequency no longer stop for 55 minutes of every hour after its SCADA restarts.
- On the production page and in the string view, DC values from a solar data logger no longer go missing after its login expires.
Per-string figures no longer fail on plants whose strings are wired directly to the inverter. The on-site agent scans a network device's web UI once, not every hour.
- A network device's vendor, model and firmware fields no longer take their values from another device.
- The on-site agent scans a network device's web UI once and only again on a plant scan, instead of every hour.
- A web-UI scan of a network device leaves a low-priority event in the plant's event list.
- A device-fault summary for a fault reported at night now holds its notice 30 minutes instead of 5.
A data logger's device list is now read whole
A data logger that announced more devices than it delivered previously left the newest entry out of its list, so that device was never polled. The device list is now read whole, and a short list seen while running falls back to the logger's web interface. Plants receive this with their next agent update.
A curtailed segment no longer marks its inverters as degraded
While a curtailment segment of a plant sat at a zero setpoint, the plant monitor previously marked its inverters as degraded. The inverters no longer open a degraded event. Plants receive this with their next agent update.
Security updates for third-party components are applied.
A no-communication summary groups silent inverters and combiner boxes
In a plant's event list, a no-communication summary now collects silent inverters, combiner boxes, turbines and battery storage into one notification, announced in 20-minute batches. It closes when every component reports again. Plants receive this with their next agent update.
An inverter's outage no longer needs daylight after a night fault. Under a zero setpoint, a false outage no longer opens when reporting returns.
- The plant's outage summary holds its notice for 30 minutes when a fault is reported before sunrise.
Battery setpoint no longer reads a flat zero
The battery setpoint figure read a flat zero even while the plant dispatched up to 10 MW. The setpoint figure now matches the plant's dispatch level. Plants receive this with their next agent update.
Battery converter's stopped state and fault alarm now read correctly
A battery converter stopped by a protection previously read as unknown, counted as available, and raised no alarm. Its operating state now reads as stopped, its own fault alarm opens five minutes after the stop, and the plant's summary alarm opens fifteen minutes after the stop. Plants receive this with their next agent update.
A converter lane that reports nothing is withheld with a warning. Collection from it resumes once it reports a change.
Stretches where a feed-in meter sent nothing are now filled from the meter's recordings. They are filled to the minute.
Wind curtailment now uses the same turbine power-curve and wind-speed model as expected production. Lost-energy figures for plants with many strings no longer fail to compute.
A wind plant's expected production is now computed per turbine
A wind plant's expected production is now computed per turbine from the plant's measured weather, each turbine's power curve and hub height, and the air density. Plants receive this with their next agent update.
A data logger that is refused or down no longer interrupts the plant's data collection or its events.
- A data logger that refuses its configured login no longer stops the plant's data collection.
- Per-inverter no-communication events now open while the plant's meters deliver, even when the inverters' data logger is down.
A silent solar inverter no longer joins the plant's outage summary after 12 hours. It joins once the plant's meters show it down.
An inverter can now be added as a data logger
An inverter can now be added as a data logger, reading its live output, energy, status and alarm codes, and backfilling its day curve from the inverter's web interface at 15-minute steps. Plants receive this with their next agent update.
A device's own not-found page no longer reads "Route not ready,". Devices that only answer at their own address now load.
- On embedded dashboards, a battery converter's operating state now reads producing or idle when its controller reports normal operation.
- On plant monitor, a battery unit no longer shows a fault for a status its controller reports at all times.
Settlement months no longer finalize while satellite days are missing
A curtailment settlement month is no longer finalized while satellite data is still missing. It is rebuilt once more if that data arrives up to 14 days after the month's deadline. Plants receive this with their next agent update.
Curtailment figures now include the direct-marketer setpoint
Curtailment figures now include the plant controller's power cap and the direct marketer's active-power setpoint. The curtailment figure no longer reads too low where either ran. Plants receive this with their next agent update.
A faulted inverter's outage now opens at once
A faulted inverter's outage now opens at once, instead of five minutes after the inverter stopped, and closes once the inverter produces again. The plant's outage summary names the inverters still down and counts the ones producing again. Plants receive this with their next agent update.
A battery storage system and a second generation of a solar data logger can now be added. Adding or updating several components of one plant at once no longer fails.
- A second generation of a solar data logger can now be added, reading inverters and environment sensors behind it.
- Adding or updating several components of one plant at once no longer fails for inverters, combiner boxes and strings.
- Combiner boxes and strings now carry their own loss figures beside the inverter loss.
- A feed-in meter's data now comes from the meter configured on the plant.
- A manually added network device now gets one alive check in the network monitor, with an action to run it.
The on-site agent now fills in satellite irradiation days published late. Satellite-derived figures for those days are no longer left empty.
A data logger's polling interval set on the dataflow page now applies to its collection cadence. The interval can be cleared back to the default.
Declared outages now reduce the flat-rate curtailment settlement
The plant's outage register is now reported to the direct marketer, and declared maintenance or administrative outages reduce the flat-rate curtailment settlement. Compensation for a period with a declared outage goes down, and figures already issued keep their value. Plants receive this with their next agent update.
A wind plant's SCADA can now be added as a data logger
A wind plant's SCADA can now be added as a data logger, reading each turbine's production, its stop commands with their cause, the grid operator or the direct marketer, and the grid operator's setpoint. Plants receive this with their next agent update.
Security updates are applied to third-party components.
A solar plant's string data no longer goes stale when its analysis updates run at the same time.
A power quality analyzer can now be added as a data logger.
Curtailment settlement and lost energy no longer read too low
Curtailment settlement and lost energy no longer read too low for a period that started after a quiet stretch. The current and previous month are recalculated on the next run, and older months keep their value. Plants receive this with their next agent update.
Counters at the plant no longer drop to zero or restart after a bad reading. Energy figures keep updating after one unreadable power reading.
- After a logger outage, a battery's lifetime counters no longer drop to zero.
- The expected-production and clear-sky energy counters no longer restart at zero when their starting reading fails.
- Energy figures no longer stop updating after one unreadable power reading.
A power quality analyzer can now be added as a feed-in meter
A power quality analyzer can now be added as a feed-in meter and reads the grid, the PV feeder or the battery feeder. Plants receive this with their next agent update.
A new plant's inverters now take the datasheet they report. The plant's battery status is now confirmed by the first battery reading.
- An inverter that reports its own manufacturer and model now gets the matching datasheet at a new plant.
- A new plant's battery status is now confirmed by the first battery reading instead of starting as none.