Open to industry roles & PhD collaboration — 2025

Thang
Nguyen

Process Engineer · Digital Twin Developer · Applied Researcher

16+ years designing and commissioning wastewater treatment systems from 500 to 19,200 CMD across Germany, Singapore, Turkey, Vietnam, and Japan. Currently building real-time controllers and digital twins deployed in production — reducing energy consumption by 25–35% for Jacobs, Veolia, and ThamesWater.

Download CV See Data Work ↓ Research Interests ↓
16+
Years in Water Engineering
19,200
CMD — Largest plant (Turkey)
35%
Max energy reduction — DT
4
Countries with live RTCs

Who I Am

Three disciplines. One workflow.

Most engineers understand the plant or the data. I design both — and I know how to turn operational reality into a model that actually works.

Field-Proven Engineer

EPC project management from Pre-FEED to handover. P&ID development in COMOS, FAT/SAT supervision, biological process design across industrial and municipal scales. I have stood on the plant floor when things go wrong.

EPCCOMOSP&IDASM ProcessesCommissioning
Digital Twin Developer

Built and deployed real-time controllers (RTC) integrating ASM mechanistic models, machine-learning soft sensors, and Kalman-filter state estimators. Connected to live PLCs via OPC UA and Modbus in production across four countries.

PythonOPC UAInfluxDBSUMOASM1/2d/3
Applied Researcher

MSc Environmental Engineering, TU Munich. Published in peer-reviewed journals. Certified in Aerobic Granular Sludge Technology (TU Delft). My field experience generates real-world datasets and open questions that advance the science.

TUM MScTU DelftPublished 2025ORCID

Data Processing · Process Modelling · Analytics

What the data actually looks like

These charts are drawn from real operational data patterns. They show how digital twin control, model calibration, and energy optimisation look in practice — not in textbooks.

R² 0.94
ASM2d model fit (calibration)
10 min
RTC update cycle (live plants)
4
Live WWTP deployments
−31%
Average energy reduction
Digital Twin Real-Time Control — NH₄-N Effluent Over 24 Hours
Measured sensor data vs. ASM2d model prediction vs. uncontrolled baseline · 10-min intervals · simulated from operational patterns
Without RTC — peak exceedances at morning & evening load
Measured NH₄-N with RTC active
ASM2d model prediction (digital twin state)
Effluent quality permit limit 2.0 mg/L
ASM2d Model Calibration
Predicted vs. measured NH₄-N effluent · 90-day sampling campaign (n=30)
Calibration data points
1:1 reference line · R² = 0.944
Energy Impact of Digital Twin
Aeration energy (kWh/m³) before vs. after RTC deployment — per project
Before digital twin
After digital twin

System Architecture

Digital Twin Architecture — Wastewater Treatment

The five-layer architecture I design and implement. Data flows from the physical plant upward into the digital twin core, and optimised control signals feed back down to the plant in real time.

LAYER 1 — PHYSICAL PLANT Bioreactors MBR / CAS / SBR Clarifiers Primary / Secondary Aeration System Blowers / Diffusers Pump Stations RAS / WAS / Feed Sludge Handling Thickener / Digester Laboratory BOD / COD / N / P 4–20 mA · Modbus RTU · HART · Profibus LAYER 2 — SENSING & FIELD INSTRUMENTATION DO / pH / ORP Online sensors Flow / Level Electromagnetic / US Turbidity / MLSS Online sensors Online Analyzers NH₄ · NO₃ · PO₄ Temp / Conductivity Process quality Soft Sensors (ML) BOD / VSS estimators Modbus TCP/IP · OPC UA · MQTT LAYER 3 — EDGE COMPUTING & COMMUNICATION Siemens S7-1200/1500 PLC / Controller OPC UA Server Kepware / S7.Net MQTT Broker Mosquitto / EMQX Edge Gateway Pre-processing VPN / Firewall Network security OPC UA · REST API · MQTT · Python client LAYER 4 — DIGITAL TWIN CORE ★ (Core Innovation) Process Models ASM1 / ASM2d / ASM3 ADM1 · SUMO · BioWin GPS-X · custom Python State Estimator Kalman Filter / UKF Parameter estimation Real-time calibration ML / Soft Sensors Gradient boost / ANN Hybrid mechanistic-ML scikit-learn · TensorFlow Data Historian InfluxDB (time-series) PostgreSQL (relational) Grafana dashboards Optimisation Engine MPC / rule-based RTC Energy optimisation Setpoint generation Optimised setpoints · Control commands LAYER 5 — CONTROL & OPERATOR INTERFACE RTC Algorithms SND · SRT · P · N/DN SCADA / HMI WinCC · Ignition Operator Dashboard Grafana · Power BI Alarm Management Permit compliance Reporting / Energy KPIs Continuous improvement Control feedback Physical / DT layers Sensing Edge / IT Control / HMI Feedback control loop ★ = Core innovation zone

Selected Deployments

Live systems. Real clients. Measured results.

Every project ran in production with permit-compliance obligations — not pilots, not simulations.

Jacobs Engineering, USA2022–2025
Wellington WWTP — SND, Swingzone, IRC, RAS & SRT Controllers

Full RTC suite for biological nitrogen removal. ASM-based soft sensors integrated with OPC UA-connected Siemens PLCs. Remote delivery from Germany to USA including commissioning support.

5
RTC modules live
Remote
DE→USA delivery
PythonOPC UAInfluxDBASM models
Veolia, France2023–2024
Sausheim WWTP — MOV, N/DN, SRT & P Real-Time Controllers

Multi-parameter RTC for nitrogen and phosphorus control with dynamic setpoint adjustment based on influent load prediction. Reduced chemical dosing through model-predictive logic.

4
RTC modules
N + P
Dual nutrient control
Modbus TCPPythonPostgreSQLSoft sensors
MERI Environmental / SÖKE Turkey2020–2022
SÖKE Industrial WWTP — 19,200 CMD Full EPC Design

Complete P&ID design and commissioning for Southeast Europe's largest industrial wastewater facility. Advanced biological treatment, Siemens PCS7 DCS integration.

19,200
CMD capacity
EPC
Full lifecycle
COMOSSiemens PCS7P&IDASM Bio
Kobelco / PUB Singapore2013–2014
Jurong Water Reclamation Plant — 4,500 CMD UASB & Biogas

Led UASB anaerobic treatment and biogas recovery for Singapore's national water agency. Delivered on time and under budget, exceeding performance targets by 12%. Zero safety incidents.

+12%
Above target
0
Safety incidents
UASBBiogasEPCSingapore PUB

For Academic Collaborators

Research interests & open questions

Commercial wastewater digital twins are being built — but rarely scientifically validated. The gap between ASM model calibration in simulation and real-time parameter estimation under variable influent loads remains largely unsolved in practice.

My specific interest is in hybrid modelling that combines first-principles ASM kinetics with data-driven state estimation — not as a curiosity but as deployable real-time tools for plant operators. I have worked with sparse, noisy, real-plant data for three years and know exactly where the models break down.

RQ 01
How can Kalman filter or observer-based state estimation be made robust for ASM models under the sensor sparsity typical of real municipal WWTPs?
RQ 02
Under what conditions do hybrid ML–mechanistic models outperform pure mechanistic models for real-time soft sensing of BOD, TSS, and nitrification rate?
RQ 03
What minimum sensor infrastructure is required to support a credible real-time digital twin for a mid-size WWTP? How should sensor placement be optimised?
RQ 04
How can energy optimisation across aeration, recirculation, and sludge handling be coordinated in real time without violating effluent quality constraints?

Publications

Peer-reviewed work

2025
Energy efficiency evaluation of a centralised wastewater treatment plant in an industrial zone of Binh Duong province, Vietnam
Vietnamese Journal of Science, Technology and Engineering · DOI: 10.31276/vjste.2024.0117
View publication ↗
Digital twin methodology for real-time process control in municipal WWTP — in preparation
Target: Water Research / Journal of Water Process Engineering
ORCID: 0009-0008-6318-742X ↗

Get in Touch
Open to the right industry role or research collaboration.

If you work in wastewater digital twins, real-time process control, or water infrastructure — let's talk. I respond within 48 hours.

For headhunters: German and Vietnamese nationality. Based in Braunschweig, Germany. Open to remote and hybrid EU roles. CV ready to download above.

For professors: Interested in PhD or research collaboration on applied digital twin development for water/wastewater. Happy to share anonymised operational datasets or discuss project ideas.