AquaTwinOptimizationSyncing
InitializingSimulated plant
—°C
Temperature
—g/L
Salinity
—NTU
Turbidity
—
pH
--:--:--
SIM --:--:--
Optimization

How AquaTwin searches for an operating strategy

Every candidate is predicted with the hybrid twin, screened by AquaGuard at its 90 % interval edge, then ranked. Nothing below is pre-computed.

Candidate strategies

Pareto-optimal Admissible Rejected by AquaGuard Withheld (low confidence) Current
Evaluating candidate strategies…

Strategy

Modeled
Select a candidate in the chart.

Search space and constraints

Decision variables
  • Feed pressure 54–72 bar
  • Feed flow per vessel 7–13 m³/h
  • Recovery (follows from both)
  • Train allocation equal / −3 / −6 bar / offline
  • Production & storage draw (planner)
AquaGuard hard limits
  • Permeate TDS ≤ 400 mg/L
  • Feed pressure ≤ 70 bar
  • Recovery ≤ 50% · flux ≤ 17 LMH
  • Vessel ΔP ≤ 3.5 bar · brine ≥ 3.6 m³/h
  • Storage ≥ 25% · confidence ≥ 50%

Current vs recommended

values from the hybrid model

Objective weights

Constraints are never traded off: weights only rank admissible strategies.

AquaTwin
Digital twin offline