A deep learning model for controlling the galvanizing process in a production line

In the industry of steelmaking, the process of galvanizing is a treatment which is applied to protect the steel from corrosion. The air knife effect (AKE) occurs when nozzles emit a steam of air on the surfaces of a steel strip to remove excess zinc from it. In our work we formalized the problem to control the AKE and we implemented, with the R&D dept.of MarcegagliaSPA, a DL model able to drive the AKE. We call it controller. It takes as input the tuple : a tuple of the physical conditions of the process line (t,h,s) with the target value of the zinc coating (c); and generates the expected tuple of (pres and dist) to drive the mechanical nozzles towards the (c). According to the requirements we designed the structure of the network. We collected and explored the data set of the historical data of the smart factory. Finally, we designed the loss function as sum of three components: the minimization between the coating addressed by the network and the target value we want to reach; and two weighted minimization components for both pressure and distance. In our solution we construct a second module, named coating net, to predict the coating of zinc resulting from the AKE when the conditions are applied to the prod. line. Its structure is made by a linear and a deep nonlinear “residual” component learned by empirical observations. The predictions made by the coating nets are used as ground truth in the loss function of the controller. By tuning the weights of the different components of the loss function, it is possible to train models with slightly different optimization purposes. In the tests we compared the regularization of different strategies with the standard one in condition of optimal estimation for both; the overall accuracy is ± 3 g/m^2 dal target for all of them. Lastly, we analyze how the controller modeled the current solutions with the new logic: the sub-optimal values of pres and dist can be optimize of 50% and 20%.

Modelling growth dynamics and toxin production in Ostreopsis cf. ovata

The benthic dinoflagellate O. ovata represents a serious threat for human health and for the ecology of its blooming areas: thanks to its toxicity this microalga has been responsible for several cases of human intoxication and mass mortalities of benthic invertebrates. Although the large number of studies on this dinoflagellate, the mechanisms underpinning O. ovata growth and toxin production are still far to be fully understood. In this work we have enriched the dataset on this species by carrying out a new experiment on an Adriatic O. cf. ovata strain. Data from this experiment (named Beta) and from another comparable experiment previously conducted on the same strain (named Alpha), revealed some interesting aspects of this dinoflagellate: it is able to grow also in a condition of strong intracellular nutrient deficiency (C:P molar ratio > 400; C:N > 25), reaching extremely low values of chlorophyll-a to carbon ratio (0.0004). Was also found a significant inverse relationships (r > -0.7) between cellular toxin to carbon and cellular nutrient to carbon ratios of experiment Alpha. In the light of these result, we hypothesized that in O. cf. ovata nutrient-stress conditions (intended as intracellular nutrient deficiency) can cause: i) an increase in toxin production; ii) a strong decrease in chlorophyll-a synthesis; iii) a lowering of metabolism associated with the formation of a sort of resting stage. We then used a modelling approach to test and critically evaluate these hypotheses in a mechanistic way: newly developed formulation describing toxin production and fate, and ad hoc changes in the already existent formulations describing chlorophyll synthesis, rest respiration, and mortality, have been incorporated in a simplified version of the European Regional Seas Ecosystem Model (ERSEM), together with a new ad hoc parameterization. The adapted model was able to accurately reproduce many of the trends observed in the Alpha experiment, allowing us to support our hypotheses. Instead the simulations of the experiment Beta were not fully satisfying in quantitative terms. We explained this gap with the presumed different physiological behaviors between the algae of the two experiments, due to the different pre-experimental periods of acclimation: the model was not able to reproduce acclimation processes in its simulations of the experiment Beta. Thus we attempt to simulate the acclimation of the algae to nutrient-stress conditions by manual intervention on some parameters of nutrient-stress thresholds, but we received conflicting results. Further studies are required to shed light on this interesting aspect. In this work we also improve the range of applicability of a state of the art marine biogeochemical model (ERSEM) by implementing in it an ecological relevant process such as the production of toxic compounds.

Assessing the impact of ammonia content in VFA-rich streams on the PHA-storing MMC acclimatization and the PHA production stage

The use of environmentally friendly products increased the interest in renewable resources as alternatives to petrochemical products. Polyhydroxyalkanoates (PHAs) are examples of such promising products, as they are biodegradable polymers with numerous potential applications. PHA production approach consists of using an open mixed microbial culture (MMC) and inexpensive feedstocks (waste or industry byproducts feedstock). The PHA process generally comprises three stages: (1) acidogenic fermentation (AF) stage (conversion of organic carbon into fermentation products); (2) culture selection stage (enrichment in PHA-storing organisms by applying Feast and Famine regime); and (3) PHA production stage (PHA accumulation up to the culture’s maximum capacity). AF of protein-rich residues results in ammonia-rich fermented streams, which can be presented as a challenge for the PHA production stage. The presence of ammonia during this stage may induce organisms to grow instead of producing PHAs. For this reason, the assessment of the effect of a high content of ammonia on PHA production it is the utmost importance. The main goal of the current project is to select a MMC enriched in PHA-accumulating organisms in conditions of high ammonia content and to evaluate the effects of ammonia presence during PHA accumulation. The culture was selected applying the Feast & Famine strategy, and fed, firstly, using a synthetic mixture of VFAs and later using a fermented stream obtained from the fermentation of protein-rich raw materials. The selected culture could accumulate up to 24% PHA per VSS with the synthetic mixture of VFAs and up to 29% for the real fermented stream. The PHA accumulation resulted in different production in the presence and absence of ammonia. Regarding to the synthetic feed, 59%wt. PHA (VSS basis) in the absence of ammonia, and 55%wt. (VSS basis) in the presence, were obtained. For the real feed, the PHA content was about 40%wt. (VSS basis) in both reactors.

Towards the sustainable production of methyl methacrylate: A comparative analysis of industrial chemical production processes using life cycle assessment methodology

Increasing environmental awareness has been a significant driving force for innovations and process improvements in different sectors and the field of chemistry is not an outlier. Innovating around industrial chemical processes in line with current environmental responsibilities is however no mean feat. One of such hard to overhaul process is the production of methyl methacrylate (MMA) commonly produced via the acetone cyanohydrin (ACH) process developed back in the 1930s. Different alternatives to the ACH process have emerged over the years and the Alpha Lucite process has been particularly promising with a combined plant capacity of 370,000 metric tonnes in Singapore and Saudi Arabia. This study applied Life Cycle Assessment methodology to conduct a comparative analysis between the ACH and Lucite processes with the aim of ascertaining the effect of applying principles of green chemistry as a process improvement tool on overall environmental impacts. A further comparison was made between the Lucite process and a lab-scale process that is further improvement on the former, also based on green chemistry principles. Results showed that the Lucite process has higher impacts on resource scarcity and ecosystem health whereas the ACH process has higher impacts on human health. On the other hand, compared to the Lucite process the lab-scale process has higher impacts in both the ecosystem and human health categories with lower impacts only in the resource scarcity category. It was observed that the benefits of process improvements with green chemistry principles might not be apparent in some categories due to some limitations of the methodology. Process contribution analysis was also performed and it revealed that the contribution of energy is significant, therefore a sensitivity analysis with different energy scenarios was performed. An uncertainty analysis using Monte Carlo analysis was also performed to validate the consistency of the results in each of the comparisons.