Data-Driven Dynamic Modelling of WTPs for short-term operations and long-term infrastructure planning

Data-Driven Dynamic Modelling of WTPs for short-term operations and long-term infrastructure planning

Overview

CWT was engaged by Seqwater to support the development of a data-driven dynamic modelling framework for water treatment plant optioneering, planning, process and capacity assessments, and plant optimisation. The project combines a comprehensive process unit library of first-principles engineering equations with long-term historical plant data to support decision-making across both short-term operations and long-term infrastructure planning. Full adoption of the DDM is pending completion of Seqwater’s assurance review, which is scheduled for completion calendar year 2026.

What the DDM is and does:

The Data-driven Dynamic Model (DDM) is a data-driven process model of Seqwater’s water treatment plants, brought together in a single application. The DDM framework is intended to provide the capability required to streamline process and capacity assessments across Process Engineering teams. Users assemble a plant from calibrated unit-process building blocks: coagulation, clarification, filtration, disinfection, etc, and run it against measured or forecast raw water quality to predict treated water quality, plant capacity, chemical demand and operating cost.

Rather than relying on fixed assumptions, every result traces back to a documented engineering equation (e.g. mass balance, reaction kinetics, hydraulics). The performance of each process unit (i.e. the relevant engineering equations) is calibrated against each plant’s own operational records (plant data logs, laboratory data), so plant capability, water quality outcomes and costs can be assessed on demand for any raw water quality envelope, flow and operating mode. This enables replacement of periodic, consultant-led studies with a data-driven tool that produces auditable calculations and reports, and that can be extended with new processes, cost data and water quality limits without code.

Services Offered

Process assessment methodology review and improvement:

  • Review and gap analysis of existing Process Assessment Framework
  • Capturing and linking the project’s user requirements via a ‘Voice of the Customer’ approach
  • Modelling approach development

Process unit library development:

  • Treatment performance functions for 45 processes, from core processes such as clarification, media filtration, and UV disinfection through to more novel processes such as ceramic membranes and magnetic ion exchange
  • Process datasheets and treatment effectiveness functions spanning 34 water quality constituents
  • Wastewater and residuals (sludge) handling process modelling
  • Water quality envelope definitions, disinfection log-reduction (C.t) credit logic, and assessment against water quality and regulatory limits (e.g. ADWG, HACCP)

Cost modelling:

  • Development of CAPEX and OPEX costing models to support planning and optimisation
  • Project- and site-specific cost adjustment factors (greenfield/brownfield context, infrastructure complexity, integration difficulty, site access, etc.), plus OPEX factors across chemical, power, sludge, maintenance and labour
  • P50 cost curves with P10/P90 bands, supporting Class 5 estimates and whole-of-life costs
  • Delivery of whole-of-life costing methodology, costing report and Process Cost Register

Verification and validation:

  • Independent (RPEQ) peer review of the DDM methodology and models, including stress-testing, feature testing and review against the requirements register (in progress)
  • Pre-delivery review of key deliverables and calculation development reports, with identification of targeted corrections

Project delivery and stakeholder support:

  • Scope development and task prioritisation for staged, multi-task delivery
  • Stakeholder engagement via weekly and monthly meetings, presentations and workshops

Need

Seqwater’s existing steady-state process assessment approach had limited ability to anticipate rare or unprecedented water quality events, conditions expected to become more frequent under climate change. A dynamic modelling framework was needed to predict treatment plant performance across a broad range of scenarios, including extreme events that are poorly represented in historical data.

Solution

CWT’s engagement covered two workstreams running in parallel.

The first focused on reviewing Seqwater’s Process Assessment Framework, assessing the needs of future users, running a gap analysis, and generating synthetic datasets for training and validation. This was followed by development of the modelling methodology, research and development of engineering calculations and models, and work on the model with the Seqwater team.

The second built a structured library covering 45 process units including conventional, advanced, and novel treatment technologies across 34 water quality constituents. For each process unit, CWT developed datasheets, treatment effectiveness functions, and CAPEX/OPEX cost curves, with project-specific adjustment factors for greenfield or brownfield context, infrastructure complexity, integration difficulty, and access constraints.

Benefit

The integrated framework gives Seqwater a flexible, predictive platform for assessing treatment resilience and vulnerability under a wide range of raw water conditions, including future climate scenarios. The process unit library and costing module provides a standardised, evidence-based resource for scenario analysis, treatment optimisation, and asset investment planning across Seqwater’s network.

More broadly, the DDM converts periodic, consultant-led, spreadsheet-based assessments into a data-driven tool that is faster, cheaper, more transparent and more trusted, and that serves operations, planning and water security teams from a single source of truth.