ICALEPCS2023 - List of Authors (Schofield, B.)

Paper	Title	Page
MO3AO06	Energy Consumption Optimisation by Using Advanced Control Algorithms	145
	F. Ghawash, E. Blanco Viñuela, B. Schofield CERN, Meyrin, Switzerland
	Large industries operate energy-intensive equipment and energy efficiency is an important objective when trying to optimize the final energy consumption. CERN utilizes a large amount of electrical energy to run its accelerators, detectors and test facilities, with a total yearly consumption of 1.3 TWh and peaks of about 200 MW. Final energy consumption reduction can be achieved by dedicated technical solutions and advanced automation technologies, especially those based on optimization algorithms, have revealed a crucial role not only in keeping the processes within required safety and operational conditions but also in incorporating financial factors. MBPC (Model-Based Predictive Control) is a feedback control algorithm which can naturally integrate the capability of achieving reduced energy consumption when including economic factors in the optimization formulation. This paper reports on the experience gathered when applying non-linear MBPC to some of the contributors to the electricity bill at CERN: the cooling and ventilation plants (i.e. cooling towers, chillers, and air handling units). Simulation results with cooling towers showed significant performance improvements and energy savings close to 20% over conventional heuristic solutions. The control problem formulation, the control strategy validation using a digital twin and the initial results in a real industrial plant are reported together with the experience gained implementing the algorithm in industrial controllers.
	Slides MO3AO06 [3.101 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO3AO06
About •	Received ※ 04 October 2023 — Revised ※ 09 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 29 November 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP102	Leveraging Local Intelligence to Industrial Control Systems through Edge Technologies	793
	A. Patil, F. Ghawash, B. Schofield, F. Varela CERN, Meyrin, Switzerland D. Daniel, K. Kaufmann, A.S. Sündermann SAGÖ, Vienna, Austria C. Kern Siemens AG, Corporate Technology, München, Germany
	Industrial processes often use advanced control algorithms such as Model Predictive Control (MPC) and Machine Learning (ML) to improve performance and efficiency. However, deploying these algorithms can be challenging, particularly when they require significant computational resources and involve complex communication protocols between different control system components. To address these challenges, we showcase an approach leveraging industrial edge technologies to deploy such algorithms. An edge device is a compact and powerful computing device placed at the network’s edge, close to the process control. It executes the algorithms without extensive communication with other control system components, thus reducing latency and load on the central control system. We also employ an analytics function platform to manage the life cycle of the algorithms, including modifications and replacements, without disrupting the industrial process. Furthermore, we demonstrate a use case where an MPC algorithm is run on an edge device to control a Heating, Ventilation, and Air Conditioning (HVAC) system. An edge device running the algorithm can analyze data from temperature sensors, perform complex calculations, and adjust the operation of the HVAC system accordingly. In summary, our approach of utilizing edge technologies enables us to overcome the limitations of traditional approaches to deploying advanced control algorithms in industrial settings, providing more intelligent and efficient control of industrial processes.
	Poster TUPDP102 [3.321 MB]
DOI •	reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP102
About •	Received ※ 06 October 2023 — Revised ※ 21 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 12 December 2023
Cite •	reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

Paper

Title

Page

Energy Consumption Optimisation by Using Advanced Control Algorithms

145

F. Ghawash, E. Blanco Viñuela, B. Schofield
CERN, Meyrin, Switzerland

Large industries operate energy-intensive equipment and energy efficiency is an important objective when trying to optimize the final energy consumption. CERN utilizes a large amount of electrical energy to run its accelerators, detectors and test facilities, with a total yearly consumption of 1.3 TWh and peaks of about 200 MW. Final energy consumption reduction can be achieved by dedicated technical solutions and advanced automation technologies, especially those based on optimization algorithms, have revealed a crucial role not only in keeping the processes within required safety and operational conditions but also in incorporating financial factors. MBPC (Model-Based Predictive Control) is a feedback control algorithm which can naturally integrate the capability of achieving reduced energy consumption when including economic factors in the optimization formulation. This paper reports on the experience gathered when applying non-linear MBPC to some of the contributors to the electricity bill at CERN: the cooling and ventilation plants (i.e. cooling towers, chillers, and air handling units). Simulation results with cooling towers showed significant performance improvements and energy savings close to 20% over conventional heuristic solutions. The control problem formulation, the control strategy validation using a digital twin and the initial results in a real industrial plant are reported together with the experience gained implementing the algorithm in industrial controllers.

Slides MO3AO06 [3.101 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-MO3AO06

About •

Received ※ 04 October 2023 — Revised ※ 09 October 2023 — Accepted ※ 14 November 2023 — Issued ※ 29 November 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)

TUPDP102

Leveraging Local Intelligence to Industrial Control Systems through Edge Technologies

793

A. Patil, F. Ghawash, B. Schofield, F. Varela
CERN, Meyrin, Switzerland
D. Daniel, K. Kaufmann, A.S. Sündermann
SAGÖ, Vienna, Austria
C. Kern
Siemens AG, Corporate Technology, München, Germany

Industrial processes often use advanced control algorithms such as Model Predictive Control (MPC) and Machine Learning (ML) to improve performance and efficiency. However, deploying these algorithms can be challenging, particularly when they require significant computational resources and involve complex communication protocols between different control system components. To address these challenges, we showcase an approach leveraging industrial edge technologies to deploy such algorithms. An edge device is a compact and powerful computing device placed at the network’s edge, close to the process control. It executes the algorithms without extensive communication with other control system components, thus reducing latency and load on the central control system. We also employ an analytics function platform to manage the life cycle of the algorithms, including modifications and replacements, without disrupting the industrial process. Furthermore, we demonstrate a use case where an MPC algorithm is run on an edge device to control a Heating, Ventilation, and Air Conditioning (HVAC) system. An edge device running the algorithm can analyze data from temperature sensors, perform complex calculations, and adjust the operation of the HVAC system accordingly. In summary, our approach of utilizing edge technologies enables us to overcome the limitations of traditional approaches to deploying advanced control algorithms in industrial settings, providing more intelligent and efficient control of industrial processes.

Poster TUPDP102 [3.321 MB]

DOI •

reference for this paper ※ doi:10.18429/JACoW-ICALEPCS2023-TUPDP102

About •

Received ※ 06 October 2023 — Revised ※ 21 October 2023 — Accepted ※ 05 December 2023 — Issued ※ 12 December 2023

Cite •

reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)