ComputationaL dEsign optimization for unpolluted wAteR

Project Information

CLEAR

Grant agreement ID: 897344

DOI

10.3030/897344

Project closed

EC signature date 16 March 2020

Start date 1 May 2020

End date 30 April 2022

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 196 707,84

EU contribution

€ 196 707,84

Coordinated by

EDEN TECH
France

Periodic Reporting for period 1 - CLEAR (ComputationaL dEsign optimization for unpolluted wAteR)

Reporting period: 2020-05-01 to 2022-04-30

Summary of the context and overall objectives of the project

The increased use of pharmaceutical drugs and agriculture chemicals in the last five decades now results in a global contamination of the water cycle. Antibiotics, endocrine disruptors, pesticides are among the most alarming compounds found in wastewaters and even in treated waters.
Currently, wastewater treatment plants are the most significant barriers to water contamination, but are known to fail in decreasing micropollutants concentrations to reasonable levels. Consequently, there is an urgent need to implement technologies for removal of micropollutants from wastewaters before they reach rivers, groundwater and marine waters.
EDEN offers a compact, easy-to-deploy, cost-effective technology for micropollutant removal, thus giving access to quality water to the EU and beyond. The system uses microfluidics for treatment of high volume of effluent at low pressures, decreasing significantly energy consumption compared to current techniques. But the fluidic design, while efficient, fails to adapt for the wide range of pollutants and their properties. The fluidic architecture should be dependent on pollutant types, sizes, concentration, ... Designing with such a large array of parameters requires expertise in a specific computational method: Computational Design Optimisation.
The objective of the research proposed here is to optimize the design of the devices developed by EDEN by means of computational methods. We will build the necessary physics-based or data-driven models to enable the use of numerical optimization algorithms to determine optimal design solutions for a variety of conditions and requirements. We will employ multidiscisplinary design optimization techniques to account for the interaction of different disciplines entailed in the treatment process, and will also formulate strategies for platform-based design of families of devices to derive different solutions at minimal cost and development lead times.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

The specific objectives for the project were set as follows.
1. Transfer theoretical and practical knowledge of computational design optimization to Eden.
The researcher introduced the principle of computational design optimization to the research methodology and instilled it into the overall development process of Eden. This was accomplished through:
a. The direct application of optimization to the design of the microfluidics-based water cleaning device; the developed optimization models and algorithms can be reused in other applications.
b. Instructing two Eden R&D engineers and four interns on how to incorporate optimization by formulating appropriate numerical problems and selecting fitting algorithms to solve them.
2. Acquire knowledge on the processes and devices developed by Eden as well as the entrepreneurial aspects.
The researcher gained extensive experience on microfluidics devices by collaborating with its R&D engineers on modelling, optimizing a device for removing pollutants from water and from participating in regular meetings of the research, development, and innovation (RDI) team with the chief scientific officer (CSO),and chief executive officer (CEO) related to the entire spectrum of R&D efforts of Eden. The researcher learned how a deep tech start-up company goes about identifying potential markets and what necessary steps must be taken to mature its technology to successfully penetrate these markets.
3. Develop the necessary computational models for conducting multidisciplinary design optimization on the processes and devices developed by Eden.
The researcher collaborated closely with two Eden R&D engineers and one intern, under the supervision of the CSO, in two principal efforts:
a. The first effort concerned the numerical optimization of a family of microfluidics-based devices used to remove different pollutants from water. Different models, both analytical and simulation-based, were developed such that they can be used effectively in a computationally cost-intensive numerical process such as multidisciplinary design optimization (MDO), which requires many executions of these models.
b. A secondary effort considered the development of data-driven surrogate models (also known as metamodels) to be used in lieu of computationally costly physics-based models. The main objective of this effort was to train off-line modular metamodels that are then used on-line in a web-based calculator to yield rough yet useful predictions of microfluid dynamics applications, but was also applied to the MDO of pollutant-removing microfluidics-base devices.
4. Establish commonality strategies to create platforms from which family of devices can be derived to address different needs and situations.
Different architectures and geometries were considered for the pollutant-removing microfluidics-base devices to satisfy a spectrum of performance ranges (in terms of flow rate and contaminants). The developed optimization approach allowed the identification of common elements in these architectures and geometries so that different devices could potentially benefit from sharing some commonality.
5. Validate the developed models and use them as digital twins for continuous improvement.
Several prototypes of the geometrically optimized disks were fabricated by Eden’s manufacturing team. They were then assembled into the pollutant-removing device and tested under different conditions and for different contaminates. The obtained experimental data showed good agreement with the (micro)fluid dynamics predictions of the developed models; data related to the removal of pollutants were less accurate in absolute values but showed qualitative agreement (such as trends and relative values).
6. Disseminate research findings
This objective was partially accomplished as Eden, a deep tech start-up company, needs to protect its intellectual property. Therefore, publications were not produced, and obtained data are not made publicly available. However, the researcher introduced numerical optimization as an integral component of R&D to the company and trained several of its staff in the area, both formally via talks and presentations within Eden and informally through the interactions with Eden’s staff and interns.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Eden is developing numerous innovative and promising solutions with broad applications based on previously non-exploited powerful microfluidics principles. Nevertheless, Eden’s innovative design ideas can be improved and matured through modelling and optimization to capitalize on their full potential. By introducing systematic and rigorous math-based numerical optimization into the R&D process, this project did not only yield significant but short-term results but also influenced the approach and methodology of Eden, which is an ever more important and long-lasting impact.
Moreover, the novel approach of utilizing machine learning approaches for efficient off-line training of surrogate models to be used later in web-based on-line calculations has a broader impact because it can inform non-expert users on the usefulness and power of microfluidics by user-friendly and straightforward experimentation with different architectures for different applications.
Finally, the researcher expanded his knowledge in a field of paramount importance (clean water) and learned valuable lessons in how deep tech innovation is generated and technology is matured in start-up industrial environments.

picture1.png