Skip to main content

Integrated Process Control based on Distributed In-Situ Sensors into Raw Material and Energy Feedstock

Periodic Reporting for period 1 - DISIRE (Integrated Process Control based on Distributed In-Situ Sensors into Raw Material and Energy Feedstock)

Reporting period: 2015-01-01 to 2016-06-30

DISIRE has been inspired by the real existing needs of multiple industrial sectors, including the world leading partners in the non-ferrous, ferrous, chemical and steel industries that are highly connected and affiliated with the SPIRE PPP and its objectives. The primary clear and measurable objective of DISIRE is to evolve the existing industrial processes by advancing the European Sustainable Process Industry through an overall resource and energy efficiency paradigm based on the technological breakthroughs and concepts of the DISIRE technological platform in the field of Industrial Process Control.

DISIRE is focusing in an overall improvement of the product quality, a reduction in the consumed energy and a corresponding reduction on the environmental impact. This problem is addressed by the application of the DISIRE that has the vision to insert in the production processes novel concepts on modeling, control and big data processing.

Overall the aim of DISIRE is to create novel sensors that will be able to follow the stages of the process or the products through the supply chain or the processing stages and measure specific characteristics of the products or the processes. In a secondary stage these sensory readings will be integrated with the current and vast amount of previously logged data from all the processes in order to be uploaded in a cloud infrastructure. The cloud will make use of the bid data analytics concept in order to process in real time this vast amount of information and to create conclusions towards the process model improvements and the overall control scheme. This process will be analyzed in the sequel by advanced cloud based control schemes that will result in a final reconfiguration of the local controllers’ tuning variables. The described DISIRE enabled processes will have the ability to be continuously updated with the most realistic status of the process, while will enable a better utilization of the energy consumption and the knowledge of the processes internal dynamics and thus produce products of better quality and more greener.

The industrial sectors that DISIRE is focusing are all characterized as heavy energy consumption processes and thus the successful execution of DISIRE and the application of the findings in the existing industrial processes will have a direct impact on the consumed energy, the quality of the products and the impact on the environment. As an example, for only the DCI North facilities, the consumption of the fuel gas is reaching up to 250K Ton/year and thus it is evident that small improvements will result into a very big impact in the field, thus 1% improvement of the DCI’s cracking furnaces, based on the DISIRE technology, will result in 2K Ton/year savings.

For achieving these visions, DISIRE has the following technological objectives:

O1. Develop miniaturized PAT technologies capable of being inserted into flows of raw materials and thus enabling the concept of “Intelligent Raw materials” and delivering a PAT based Swarm Sensing and Data Analysis

O2. Introduce Multi-objective In line Sensor Technology for Real Time Sensing & Networking in Industrial Environments

O5. The focus of DISIRE in the non-ferrous processes includes cross-sectorial processes in Mines, Processing Plants, Mining, Machine Producers, etc. A key objective for the DISIRE project is to improve efficiency and competitiveness related to the copper production. A way to progress towards this objective is to decompose the whole production process into many sub-processes and their identification and better understanding.

O6. The focus of DISIRE in the Ferrous Processes lies on demonstrations and experiments in real industrial environments in order to establish a rugged sensor platform capable of withstanding the harsh industrial conditions of ferrous mineral processing, constituting of abrasive wear and high temperatures. Ferrous mineral processing presents several challenges that have to
The up to now work performed in DISIRE is presented in the sequel with respect to the corresponding WPs.

WP1 - Application Scenarios, Impact Goals and Benchmarking

WP1 has been the base for all the developments in DISIRE since it contained all the specifications and the plans that all the technological WP2-4 and the demonstration WPs WP5-8 should follow. In this WP the initial and end-user specifications for the sensing, controlling and data analyzing techniques have been specified, while the targeted demonstration processes in WP5-8 have been also defined. Specific focus was maid in highlighting the current status of the industrial processes and the way that DISIRE will evolve and impact. Overall, the DISIRE demonstration actions will be categorized in open loop processes (WP5 and WP6) and closed loop processes (WP7 and WP8). Finally in this WP, specific and realistic end-user driven measurable key performance indices for the DISIRE developments have been produced in order to maximize the impact of the project.

WP2 - Process Modeling and Control

In this WP there have been produced novel methodologies for the analysis, modelling and control of industrial systems, with a special focus on the industrial processes of DISIRE. A fast online modelling algorithm (operating on a stream of data coming from the process) has been introduced, which turned out to be one order of magnitude faster than other existing approaches. Furthermore, a Newtonian algorithm has been introduced for the fast solution of the identification problems for piecewise affine dynamics where we simultaneously identify the polyhedral partition and the individual system dynamics. This algorithm can also operate on a stream of data. Additionally, WP2 presented an algorithm for the selection of a (sparse) controller configuration based on the system's relative gain array and a stochastic model predictive control scheme using scenario trees.

WP2 then focused on the three case studies of DISIRE, namely, the walking beam furnace of MEFOS, the ethylene/LPG cracking furnace of Dow Chemical and the network of conveyor belts at the mines of KGHM. In particular, for the walking beam furnace we provide a complete analysis starting from dynamical modeling of the involved uncertainties using scenario trees, modelling of the combustion quality using machine learning techniques and the formulation of a control problem with relevant closed-loop simulations under uncertainty. For the cracking furnace of Dow Chemical we identified dynamical models using process data. For the network of conveyor belts we provided a description of the system characteristics and specifications and a problem statement, we used models based on cellular automata to describe the flow of mass inside the ore bunkers and we present a probabilistic analysis of the ore mass moving on the conveyor belts and the pertinent waiting times (in terms of their probability density functions). The ore flow with continuous charging from the retention bunkers was also modelled using time series models.

WP3 – Sensor and Electronics

WP3 created the inline sensing technologies of DISIRE. More specifically, positioning RFID tags have been developed for tracing the iron ore, with corresponding RFID readers, where the position was measured both using 13.56 MHz RFID as well as using 125 kHz beacon technology, with the aim to continue the work with other sensor technologies towards oxygen sensing. Furthermore, in WP2 studies were performed towards the applications of the positing measurement in transportation chains in the underground transport system in KGHM. For the experimental blast furnace and the walking beam furnace specific sensor inline modules have been developed that can be inserted in the hot processes, transmit data and retain their operation up to 900 Degs. For the batch sensor development, WP3 has investigated possible solutions for the batch processes, including spatial temperature distribution and s
The DISIRE results up to the M18 month of reporting have lead to the following progress beyond the sate of the art and with the corresponding expected potential impact.

Mineral Processing

In the hot side, the up to now activities of DISIRE have increased the knowledge about the heating products in the grate zones and the logistics to the customer. Also the ability of inline sensing has created a real potential for reducing the environmental impact, while increasing the energy efficiency and the throughput, with a corresponding reduction of the waste products. The developed sensing technologies regarding the blast furnace although still cannot survive in extremely high temperatures, are in position to provide important data sets for the first time ever in the corresponding industry and to further improve the knowledge of the process. This information will be combined in the current stages of DISIRE in order to investigate potential IPC solutions for investigating the further optimization of the process. The hot stage of the mineral processing at LKAB is the largest consumer of energy. With the embedded sensors measuring temperature inside of the material, rather than measuring the surface temperature, it will be possible to optimize the combustion process.

The up to M18 envisioned impact in the hot side of DISIRE is including the generation of a better knowledge of the heating process inside the LKAB’s ovens that will produce further possibilities to reduce the energy consumption. As an example it should be stated that a 2% decrease in energy consumption or reduction of oil of 0.1l/ton produced pellet could save up to 50K to 100K Euros per pelletizing plant. Furthermore, more precise temperature control will improve the product quality and will reduce the slag residues. By knowing each sensor’s position, the throughput of the different experimental batches will be increased.

Similarly, in the cold side, before any modifications of the hot process are possible the transportation conditions of the final product should be mapped since a large amount of the material is damaged and the strength of the pellets are directly correlated to the amount of heat added in the grate and kiln. During transportation with rail from the mine site in Kiruna to the harbour in Narvik a considerable amount (in order of several per cent) of the product is damaged and has to be sieved from the premium product that are shipped to the customers. Considering that the annual production is about 25 million tones this corresponds to a significant amount and improved tracing and mapping of the transport conditions using embedded sensors that could considerably improve the competitiveness of the business. Every reduction in the percentage not sieved off before shipping reduces energy consumption by the same amount, and more if the losses due to transportation is considered. DISIRE has performed extended real life trials in this direction and the consortium is working in demonstrating the capabilities of this technology in the remaining time of the project.

The up to M18 identified impact of DISIRE in the cold side considers the creation of a better understanding of how a large number of pellets is spread out in the transport chain and how this is affecting the creation of virtual product batches within the continuous flow of the iron ore pellets. With a simulation model, LKAB is able to handle production batches that are produced outside the specifications in a more efficient approach, within the existing transportation systems, otherwise these batches will need to be handled in a separate approach, a fact that will significantly increase the production cost. Thus, it is expected that if LKAB meets the potential of a reduction of separate handling by 50% this would directly mean that around 250000-350000 euros per pelletizing plant in saving per year, while the establishment of the virtual batches will have a direct effect on improving the handling