Live status
Supply position, tanks and today's figures
Measured energy against the billing baseline
Recent trends, other systems
Active alarms
Advisory findings
Alarms of the last 24 hours
Every column in process order with its meaning, unit, block and newest valid value; missing columns are listed as missing.
Newest valid reading of every mapped signal, exactly as collected from the plant (OPC UA subscription or historian export on the collector box). Each value carries its own timestamp - the page never invents liveness the data does not have.
Recommended settings
—
聯合最佳化與逐項設定
Choose the daily production; the parameters that move together in one proposal are listed below against their PLC readings.
每日預計產水量
Model verdict per setpoint
Every setpoint's model verdict: a published value, an unverified value, or why the reading stands - the same snapshot as the PLC view.
Current PLC setpoints
PLC setpoints by system and by kind of setting: the value, the design value, the difference and the selectable range. Readback / calculated rows are not settings.
Two independent advisory sources, each shown as reported; nothing is merged or re-scaled.
Daily basics (measurements and records, served every day)
Membrane health and recommended CIP dates
Details: ERD envelopes, indices, CIP outcomes and records
Every membrane stage on this site, in process order. A stage with no mapped instruments is not shown at all rather than shown empty.
Membrane state and CIP timing (each unit against itself)
Against the manufacturer's own criteria
UF · pre-treatment
SWRO · first pass
BWRO · second pass
Second-pass (brackish) RO monitoring: newest values and 14-day trends for each unit, straight from the mapped instruments. Health indexing for BWRO is on the roadmap - this page shows measurements, and says so.
CIP timing against the makers' triggers
CIP outcomes and membrane performance indicators
Dual gate per train
Cleaning history
Every cleaning on record
The two sections above compare each SWRO train against itself, which is what they are for. This is the plain record — every unit type, one row per cleaning step, as the plant kept it.
Running hours
Maintenance and records
Online instrument calibration
Equipment limits
Per m³ of product since the deployment, against the year before
A withheld advisory is the product working: the reason names exactly what the data could not support.
Physics mode: the membrane transport model is calibrated on each train's own record; every settable value is predicted, contract-gated and priced, and every refusal is named. Read-only: every step is taken by a person.
Where several identical units run in parallel, one of them going wrong can be invisible in its own reading — if the measurement shares a term across units, a fault in one moves all of them together and reads as a shared upstream cause. What shows it is each unit's distance from its siblings, measured against what that unit itself normally does.
The sheet the plant fills in by hand every morning, built from the record instead. Every figure states the basis it was computed on, because a compliance number that cannot say which samples it used is not one anybody can defend.
Chemical dosing monitor: tank levels, dosing-pump status and product-water pH, straight from the mapped instruments. Values carry their own timestamp; a low tank or an out-of-band pH is flagged, nothing is smoothed.
Chemical tank levels
Dosing pumps
Product-water pH
Every number on this dashboard has a stated method behind it. This page is the audit trail for a technical reviewer: what each index is, where its formula comes from, and what the product refuses to compute.
What this deployment can see
Each gap below names the number it degrades.