AI-Driven Control System for Hybrid Plant

Africana Energy Limited is developing a hybrid renewable energy-powered green hydrogen and synthetic fuel production facility in Blantyre, Malawi. We are seeking a capable freelancer or team of developers to implement and customise an open-source AI-based automation, optimisation and control platform using OpenEMS, MLFlow, and related technologies.

This is a mission-critical deployment for Africa’s first integrated solar-wind-hydrogen-Fischer-Tropsch fuel plant, with a focus on cost-effectiveness, modularity, and long-term scalability.

Plant Systems & Components
A. Renewable Energy Generation
• Wind Turbines: Envision Energy 6.25 MW turbines (multiple units)
• Solar PV System: ~40 MWp high-efficiency N-type panels (e.g. DMEGC, LONGi)
• Inverters + MPPTs: High-voltage solar string inverters with monitoring interfaces
• Battery Energy Storage System (BESS): Li-ion or LFP cells with 4-6 hours autonomy

B. Hydrogen Production
• Electrolyser: 15 MW skid-mounted Peric Alkaline Electrolyser
• Water Source: Borehole with on-site purification (RO, UV, filtration)
• Hydrogen Purification + Drying: Included in Peric system
• Oxygen Vent or Capture System

C. CO2 Capture and Scrubbing
• CO2 piped from nearby breweries (~550m distance) and cement plant (~950m distance)
• Scrubber and dryer system for CO2 purification

D. Synthesis and Processing Systems
• Methanol Synthesis Reactor (vendor TBD)
• Fischer-Tropsch Synthesis Reactor (vendor TBD)
• Hydrocracker Unit (vendor TBD)
• Tail Gas Recovery & Flare System
• Product Fractionation System (SAF, diesel, naphtha, wax, LPG)
• LPG Compression and Bottling Skid
• Storage Tanks for each output

E. Utilities & Support Systems
• Wastewater Drainage System
• Instrument Air + Nitrogen Generation Plant
• Site-wide IP-based SCADA system (e.g., ETAP, openHAB, Node-RED dashboards)

Project Objectives
The selected freelancer will:
1. Deploy and configure OpenEMS to manage and optimise:
o Solar PV and Wind Turbine input
o Battery charge/discharge scheduling
o Real-time dispatch to the electrolyser based on power availability
2. Build ML pipelines using MLFlow and TensorFlow/PyTorch to:
o Optimise electrolyser efficiency based on real-time data
o Predict hydrogen production and adapt to power fluctuations
o Optimise synthetic fuel yields once FT and methanol systems are integrated
3. Design IoT integration layer with secure data ingestion and control for all relevant components:
o Sensors: voltage, current, temperature, pressure, flow, humidity, dissolved solids, H2/CO2 purity, O2 content, tank levels, vibration, gas composition
o Actuators: control valves, relays, switchgear, vent stacks, inverter control, scrubber operation, pump throttling, flare ignition
4. Integrate with the site-wide IP-based SCADA system:
o Preferably OPC-UA or MQTT interface
o Node-RED visualisation and rule engine or compatible front-end
5. Establish digital twin capabilities (optional) using OpenModelica or similar to simulate:
o Electrolysis system dynamics
o Thermal profile of FT synthesis
o Load vs. fuel yield optimisation

Deliverables
• Fully configured OpenEMS with local weather API integration
• Integrated MLFlow pipeline for electrolyser and fuel synthesis AI models
• Secure sensor/actuator interface with live data stream and edge buffer
• Integrated control interface to SCADA
• Visual dashboard (Node-RED or similar)
• Documentation and handover of source code + deployment scripts
• Support for testing and commissioning
• The AI platform should facilitate real-time monitoring of the plant's operations.
• The software will need to integrate with various sensor data for the purpose of this real-time monitoring.

Technology Stack
• OpenEMS (Java-based)
• MLFlow, TensorFlow, PyTorch (Python-based)
• MQTT / OPC-UA / Modbus-TCP
• Docker / Kubernetes optional for deployment
• Node-RED or openHAB dashboards

Profile of Ideal Freelancer / Team
• Experience with industrial IoT systems and control
• Past deployment of OpenEMS or SCADA interfaces
• MLFlow or MLOps experience (model training, serving, tracking)
• Understanding of renewable energy, hydrogen electrolysis, or chemical processes
• Ability to work remotely, communicate clearly, and deliver on milestones
• Proven experience in AI and machine learning, particularly in the context of operational management in industrial settings.
• Strong background in developing custom AI platforms.
• Experience in working with real-time monitoring systems and sensor data.
• Knowledge of renewable energy and fuel production processes is a plus.

When applying for this challenging project, please include:
• Your relevant experience (with links to past work if available)
• Technical approach and timeline for development
• Breakdown of deliverables and support period offered
• Budget expectation or hourly rate

We look forward to collaborating with a skilled team to bring this ground-breaking project to life.