research
Process systems engineering at CEPIMA: modelling and simulation, optimisation, applications and performance assessment.
CEPIMA is a process systems engineering (PSE) group. Our research centres on using and developing methods and tools that support decision making in chemical processes and in the transformation of natural and productive resources, and that make those processes better. We specialise in the modelling, simulation and optimisation of processes (for their design, integration, operation and control, and for the supply chains they belong to) and in systematic, automated procedures for solving the problems that come with them.
Our approach
We develop tools and procedures for design and operation problems that are, in practice, never one-dimensional:
- Multi-objective. Economic, environmental and social impact are weighed together, not one after the other.
- Multi-actor. Most decisions in the process industry involve several stakeholders with different priorities. We analyse how they interact with game theory, and look for solutions that every actor can accept.
- Multi-scale. From the molecular level to the management of whole supply chains, and from the seconds of process control to the years of a design decision.
We then apply them to improving processes, their efficiency and their integration within the paradigm of the circular economy.
Research lines
Our work is organised in four disciplines that feed one another, as the wheel on our home page shows: models are what we optimise, optimisation proposes alternatives, applications put them to the test, and assessment judges them.
Modelling and simulation
Every decision we support starts from a model, built from first principles, from data or from both, and at the scale the decision needs.
- Process simulation. Flowsheets of complete processes, used to compare alternative routes, such as those from CO₂ to methanol, before any of them is built.
- Kinetic modelling. Mechanistic and lumped models of advanced oxidation processes such as photo-Fenton and ozonation, calibrated against our own experiments.
- Surrogates and knowledge models. Machine-learning surrogates that replace expensive simulations, and ontologies that capture knowledge on transformation processes.
Selected work
- Mechanistic modelling of Fe³⁺-EDDS mediated photo-Fenton revisited: lumped radicals and sensitivity analysis. Chemical Engineering Journal, 2023.
- Development of surrogate models for distillation trains. 14th Mediterranean Congress of Chemical Engineering, 2021.
- Synthesis and assessment of waste-to-resource routes for circular economy. Computers & Chemical Engineering, 2021.
Optimisation
With a model in hand, the question becomes which decision is best, and best for whom.
- Mathematical programming. Superstructure and multi-objective formulations that screen design alternatives systematically, including life-cycle optimisation with our PULPO framework.
- Multi-actor optimisation. Frameworks based on game theory that find integration solutions every company involved can accept.
- Uncertainty and control. Machine-learning methods for optimisation under uncertainty and for model predictive control that adapts to the production schedule.
Selected work
- Cooperative multi-actor multi-criteria optimization framework for process integration. Applied Energy, 2025.
- PULPO: a framework for efficient integration of life cycle inventory models into life cycle product optimization. Journal of Industrial Ecology, 2024.
- Development of approximate scheduling-adaptive controllers for multi-products continuous chemical processes using deep learning techniques and model predictive control. Computers & Chemical Engineering, 2026.
Applications
We apply these methods to real processes, combining experiments in our own laboratories with modelling and optimisation. Two lines on the treatment, recovery and valorisation of waste currently stand out, water and gases.
- Water. Advanced oxidation processes (photocatalysis, ozonation, Fenton and photo-Fenton) for recalcitrant contaminants and contaminants of emerging concern, such as pharmaceuticals.
- Gases. Capture, separation and valorisation of CO₂ and fluorinated gases with porous materials and ionic liquids, including methanol production from CO₂ and renewable energy.
- Circular economy and industrial symbiosis. Chemical recycling of plastic waste, shared utility networks between plants, and the recovery of nutrients and metals from polluted streams.
These applications run through projects such as WAPO, P2CE, AIMS and CEPI.
Selected work
- Enhanced degradation of venlafaxine induced by Fe₂O₃/CeO₂ heterostructures through the thermal transformation from type I to direct Z-scheme in heterogeneous photo-Fenton. Journal of Environmental Chemical Engineering, 2024.
- Desarrollo y evaluación de matrices sólidas porosas para la captura, separación y valorización de CO₂. 16th Mediterranean Congress of Chemical Engineering, 2026.
- Application of industrial symbiosis principles to the management of utility networks. Applied Energy, 2022.
Performance assessment
An alternative is only better if it is better on the criteria that matter, measured consistently.
- Economic and environmental indicators. Total cost, net present value, life cycle assessment, CO₂ emissions and toxicity.
- Operational indicators. Controllability and robustness, so that a design that looks best on paper also copes with disturbances and uncertainty.
- Monitoring and supervision. Decision-support systems for plant monitoring, fault diagnosis and the supervision of schedule execution.
Selected work
- Assessing the economic viability of green methanol production: the critical role of CO₂ purity. Systems and Control Transactions, 2025.
- Comparing the climate impact of methanol production in Europe: steam methane reforming vs. plastic waste gasification processes. Resources, Conservation and Recycling, 2024.
- Sulfaquinoxaline oxidation and toxicity reduction by photo-Fenton process. International Journal of Environmental Research and Public Health, 2021.
How the lines connect
The four disciplines are a way of describing the work, not of dividing it: a single question usually runs through all of them. Take methanol from CO₂. The alternative routes are first simulated (modelling); their operating conditions, and where the hydrogen and electricity come from, are optimised (optimisation); the options are tested with real sorbents and real processes (applications); and each is judged on its net present value and its life-cycle climate impact (performance assessment), which sends new questions back to the models. The same loop turns a photo-Fenton experiment into a kinetic model, the model into an optimal dosage policy, and the policy into a treatment that costs less and leaves less behind.
Facilities
CEPIMA’s equipment and facilities are used both in our research projects and to provide services to companies and institutions:
- Pilot plant.
- Analytical laboratory: HPLC, total organic carbon (TOC) and UV–Vis analysis, among others.
- Computing laboratory: dedicated servers for simulation and optimisation.
Tools
Our methodological work is supported by software we develop and maintain in-house. See Tools for the current catalogue, including:
- PULPO: Python framework for life-cycle optimisation.