Plant-wide monitoring and control of data-intensive processes with AI functions

Erik Dahlquist at Mälardalen University provides further insight about the FUDIPO project concerning plant-wide monitoring and control of data-intensive processes including AI functions

FUDIPO is a project funded by the European Commission under the H2020 programme, SPIRE-02-2016: “Plant-wide monitoring and control of data-intensive processes”, which started on October 1st, 2016 and ends on 31st January 2021. The project is coordinated by Mälardalen University, and the consortium consists of energy experts, applied mathematicians, and software engineering experts to face the SPIRE topic.

The process industry needs solutions to reduce operating costs, environmental performance footprint and improve the quality of the products. Thus, FUDIPO is developing and testing (in five case studies) advanced dynamic physical (complemented with soft sensors) and statistical models, like Bayesian networks and machine learning models, to form advanced diagnostic, decision support, optimisation and model predictive control. The system developed will optimise all levels in a factory, integrating the different control levels from the separate production units to mill level by building blocks. Thus, the project aims to provide energy and resource savings, as well as better environmental performance in EU industries. The developed system is implemented in full-scale, and validated in five case studies as described below:

Oil refinery (Tüpras)

Problematic: Diesel is produced in a unit where the focus is increased production within European standards for distillation point, Sulphur content, flash point, etc., whose variation is unmeasured.

FUDIPO: The project brings better process control, reducing give-away product below or above European limits.

Large heat and power plant (Mälarenergi)

Problematic: The heterogeneity of the waste used for cogeneration plant causes operational problems and challenges in emissions control.

FUDIPO will improve the control, decreasing downtime, fluctuations, corrosion, fouling and agglomeration.

Waste-water treatment plant (ABB)

Problematic: The aeration demand constitutes 50% of the electric energy demand.

FUDIPO: The development of control algorithms for better performance, measuring the quality of incoming waste.

Pulp and paper industry (Billerud-korsnäs)

Problematic: The plant has three fibre lines with different pulp qualities. The most important parameter is the Kappa number, which measures how much lignin is left in the pulp after the digester, and whose control is difficult.

FUDIPO: More stable process and fault diagnostics due to better control of Kappa number.

Micro heat and power turbine (MTT)

Problematic: The EnergyTwin system has high maintenance costs and needs to improve its diagnosis efficiency and decision support to installers and operators.

FUDIPO: Increasing efficiency in supporting clients with scheduled and predictive maintenance support and planning.

FUDIPO: Results to date

Oil refinery – oil properties are predicted from NIR and used to optimise hydrocracker, desulphurisation and catalyst function, as well as distillation and separation using on-line advanced control.

Wastewater treatment plant – process- and sensor diagnostics are used together with model predictive control to optimise energy use and treatment results.

Large heat and power plant – a “digital twin” is used for process- and sensor diagnostics and advanced model predictive control to avoid agglomeration in the fluidized bed, as well as avoiding the formation of dioxins. Also, inventory management is made in a production planning tool.

Pulp and paper factory – a digital twin and NIR spectra predict lignin content and reactivity of incoming wood. This is used to minimise kappa number variation (=color) of fibres after the digestion, but also reducing the use of energy and chemicals and increasing the production rate. This means a higher value of fibres, increased income, but also reduced cost per tonne.

Micro heat and power turbine – fleet management of micro gas turbine CHP plants is implemented. This means there is a possibility to control hundreds or thousands of small plants in a large geographic area remotely to optimise performance and reduce downtime.

FUDIPO: Potential earnings

The estimations of how much energy could be saved if the proposed and similar techniques and are implemented everywhere in the EU show the strong impact that FUDIPO would have. Total potential savings by industries are: Pulp and paper (some 86-95 TWh/y), Waste combustion (some 130-190 TWh/y), Oil refineries (some 120 -200 TWh/y), Wastewater treatment (some 17-41 TWh/y). Joined, the total savings are 353-526 TWh/y, which is in the range of 3% of all energy used within the EU for all purposes.

FUDIPO: Open-source toolbox

One platform with primarily commercial software and another with only an open-source software have been developed within the FUDIPO project. The open-source software has the advantage that the user does not need to pay any license fees, but at the same time, the user doesn’t get support if adjusted codes or improved functions are needed, which the commercial suppliers can provide but to a specified cost. If own functions and its integration with a platform are desired, it may be easier in the open-source platform, but if just the functionality and support for this are desired, it is better to use the commercial software.

In both cases, new artificial intelligence (AI) functions are continually being added. This includes e.g. ANN (artificial neural nets), BN (Bayesian nets), PLS models and other regression models, multivariate analysis, ML (machine learning) etc. It also includes simulation tools like OpenModelica for building physical models for any application.

 

Contact: erik.dahlquist@mdh.se

Stay informed of our activities and results!

This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme Under Grant Agreement No. 723523.

 

Please note: This is a commercial profile