ANN modelling of the steam reforming of naphthalene based on non-stoichiometric thermodynamic analysis

Steam reforming and carbon deposition evaluation of phenol and naphthalene used as tar model compounds over Ni and Fe olivine-supported catalysts

Optimization of methanol steam reforming over a Au/CuO–CeO2 catalyst by statistically designed experiments

Biomass-derived substrates such as bio-oil and glycerol are gaining wide acceptability as feedstocks to produce hydrogen using a steam reforming process. The wide acceptability can be attributed to a huge amount of glycerol and bio-oil obtained as by-products of biodiesel production and pyrolysis processes. Several parameters have been reported to affect the production of hydrogen by biomass steam reforming. This study investigates the effect of non-linear process parameters on the prediction of hydrogen production by biomass (bio-oil and glycerol) steam reforming using artificial neural network (ANN) modeling technique. Twenty different multilayer ANN model architectures were tested using datasets obtained from the bio-oil and glycerol steam reforming. Two algorithms namely Levenberg-Marquardt and Bayesian regularization were employed for the training of the ANNs. An optimized network configuration consisting of 3 input layer 14 hidden neurons, 1 output layer, and 3 input layer, 5 hidden neurons, and 1 output layer were obtained for the Levenberg-Marquardt and Bayesian regularization trained network, respectively for hydrogen production by bio-oil steam reforming. While an optimized network configuration consisting of 5 input nodes, 9 hidden neurons, 1 output node, and 5 input nodes, 8 hidden neurons, and 1 output node were obtained for Levenberg-Marquardt and Bayesian regularization trained network, respectively for hydrogen production by glycerol steam reforming. Based on the optimized network, the predicted hydrogen production from the bio-oil and glycerol steam agreed with the actual values with the coefficient of determination (R2) > 0.9. A low mean square error of 3.024 × 10−24 and 6.22 × 10−15 for the optimized for Levenberg-Marquardt and Bayesian regularization-trained ANN, respectively. The neural network analyses of the two processes showed that reaction temperature and glycerol-to-water molar ratio were the most relevant factors that influenced the production of hydrogen by bio-oil and glycerol steam reforming, respectively. This study has demonstrated the robustness of the ANN as a technique for investigating the effect of non-linear process parameters on hydrogen production by bio-oil and glycerol steam reforming.

Silica nanowires encapsulated Ru nanoparticles as stable nanocatalysts for selective hydrogenation of CO2 to CO

Hydrogenation of carbon dioxide and carbon monoxide over supported rhodium catalysts under 10 bar pressure

Computational fluid dynamics study of hydrogen generation by low temperature methane reforming in a membrane reactor

The capture and electrochemical conversion of CO2, powered by renewable electricity, is an attractive method of sustainably producing valuable chemicals and fuels (e.g. carbon monoxide (CO)), reducing atmospheric CO2, and storing intermittent renewable energy. Integrated capture and conversion (reactive capture) of CO2 presents a CO2-to-CO electrolysis pathway that eliminates most of the upstream capital and energy costs by releasing CO2 directly inside the electrolyzer using an internal pH-swing. The reactive capture system readily allows for the collection of produced gas products via phase separation, thus minimizing downstream separation costs.Industrial-scale integration of reactive capture systems with upgrading processes require a pure and consistent product stream. Previous studies using bicarbonate electrolytes have demonstrated high selectivity towards CO. However, the limited CO2 capture capacity of bicarbonate electrolytes dilute the cathode product gas stream with excess CO2. This mandates a secondary CO2 capture unit and increases the cost of downstream separation. Other studies using carbonate or carbamate electrolyte as the inlet feed did not simultaneously achieve high CO selectivity and long-term stability.This study sought to improve the Faradaic efficiency (FE) toward CO in our carbonate electrolysis system by engineering a novel membrane electrode assembly structure. We designed a composite CO2 diffusion layer (CDL) between the cathode and the membrane that attains high CO selectivity by simultaneously achieving high alkalinity and sufficient CO2 availability at the cathode. We determined that the thickness, wettability, and permeability of the CDL affected species transport and were important optimization parameters. Applying this strategy, we produced syngas, a mixture of CO and hydrogen (H2), with an industrial H2/CO ratio of 1.16 at 200 mA cm-2. This corresponded to a CO Faradaic efficiency (FE) of 46% and energy intensity of 52 GJ tsyngas-1. The syngas produced in this system was not diluted by CO2 and contained sufficient CO content to meet industrial standards. We further increased the FE towards CO by exploring different capture solutions and designing selective catalysts for energy efficient CO production. System parameters such as temperature and pressure effects were also investigated to improve the CO2 concentration at the cathode. This study illustrated the potential for the industrial implementation of an energy efficient and capital cost effective CO2-to-CO pathway via reactive capture.

In this study, the steam reforming of naphthalene and pyrene as heavy tar model compounds was investigated experimentally in a horizontal tube reactor and theoretically using the CHEMKIN simulation. The experimental results revealed that the reactivity of naphthalene was higher than that of pyrene in the presence of steam. The carbon content converted to soot is a little more than that converted to light gas during tar steam reforming. The kinetic parameters of the overall reaction were determined, and the pre-exponential factor and activation energy were calculated using the experimental data. The comparison of the numerical simulation with the experimental findings exhibited an excellent agreement for the prediction of the light gas products and soot after eliminating the influence of the water–gas shift. Further, the reaction schemes including the reaction pathway and associated kinetics were determined for the steam reforming of these model compounds. Both naphthalene and pyrene exhibited a similar performance during the reaction in the presence of steam. Benzene and naphthalene, representing the precursors of the light gas product, were confirmed to be the dominant intermediate components of naphthalene and pyrene, respectively. The consecutive reactions of these intermediates subsequently resulted in the generation of the light gaseous products.

ABSTRACTDue to the limitations of existing methods, steam reforming is the most important method of hydrogen production. In this study, we intend to investigate the potential of two Rh/Al2O3 and Cr/Al2O3 catalysts in the conversion of naphthalene in the steam reforming process. For this purpose, the experimental method was first described in a fluidized bed reactor. In the next step, the effect of various parameters such as catalysts, temperature and steam/carbon ratio on naphthalene conversion was investigated. With increase in temperature from 700°C to 850°C, the naphthalene conversion increased from 43.2% to 82.8% for Rh/Al2O3 and from 39.0% to 77.9% for Cr/Al2O3. Hydrogen production increased as the injection of steam into the reactor increased which can be explained based on the principle of Le Chatelier.

The reaction mechanism of carbon monoxide (CO) hydrogenation to methanol has been carried out theoretically in this paper. The atomic configuration was analyzed by an unlimited B3LYP calculation method in density functional theory. The reaction model for CO hydrogenation to methanol was established by using Gaussian 09. The adsorption sites, bond angles, bond lengths, reaction intermediates, transition-state structures, adsorption energy, reaction-energy barriers, and reaction heat of CO hydrogenation to methanol with different amounts of copper-based catalysts were calculated. The calculations provided the elementary reaction of methanol in the synthesis, and the reaction potential-energy diagram for methanol synthesis was plotted. The optimum reaction path for CO hydrogenation to methanol was as follows: CO→HCO*→H2CO*→H3CO*→CH3OH. The rate-limiting step was the hydrogenation of the methoxy (H3CO) species with an activation barrier of 1.28 eV.

The effect of utilising energy derived from fossil sources on the environment has aroused research interest in alternative and sustainable energy sources. Synthesis gas, a mixture of carbon monoxide (CO) and hydrogen can be used as starting materials in hydrogenation reactions to produce chemical intermediates that can be used in various processes. This study investigates the robustness of applying a fine Gaussian support vector machine algorithm for predicting light olefins from catalytic CO hydrogenation using magnesia nanoparticles-based catalysts. The datasets obtained from the CO hydrogenation reaction consist of input parameters such as magnesia nanoparticles contents, reaction temperature, and reactor pressure, and the output parameters which include CO conversions and the selectivity of light olefins (CH4, C2H6, C2H4 C3H8, C4H8, and C3H6). The dataset was trained and employed for the prediction of the light olefins using a support vector machine with an inbuilt Fine Gaussian Kernel function. The performance of the support vector machine was evaluated using the coefficient of determination (R2), root mean squared error (RMSE), mean square error (MSE), and the mean absolute error (MAE). The support vector machine showed significant potential in the prediction of CO conversion, CH4 selectivity, C2H6 selectivity, and C2H4 selectivity as indicated by R2 of 0.770, 0.800, 0.730, and 0.930, respectively. While less predictive performance was obtained for the prediction of C3H8 selectivity, C4H8 selectivity and C3H6 selectivity as indicated by R2 of 0.630, 0.610, and 0.320, respectively.

Selective oxidation of CO in excess hydrogen over [formula omitted] catalysts

Introduction During the last two decades, the potential impact of indoor air quality on human health has stimulated an interest in hazardous compounds survey such as carbon monoxide (CO) [1]. The French Institute for Health Surveillance (InVS) reports that accidental domestic poisoning by the CO affects about 1000 households in France each year [2], and is responsible for about 100 deaths. The detection of this compounds has consequently become a need. To address it, we here report results on the capability of functionalized Surface Acoustic Wave (SAW) devices for the selective detection of CO. Here we insist on the necessity to detect CO in presence of interferent such as O2 that is obviously present in the air and CO2 present in significant quantity in urban area. Material SAW delay lines based on Love waves, shown in figure 1, are used to probe mass of sensitive materials deposited as a thin layer on its surface. These devices consist in two-port delay lines built on quartz. The Love wave is generated and detected using interdigited transducers (IDTs) and the frequency operation is in the vicinity of 125 MHz. A 1.5 µm thick silica guiding layer is deposited onto the IDTs providing a propagation path which permit the guidance of the acoustic wave. For the functionalization of the sensor, we take advantage of the great capabilities of cobalt corroles to trap CO molecules at the sensor’s surface with selectivity [3]. Because of the structure of the corroles, small molecules such as N2, O2 and CO2 can be trapped within by mean of weak interactions. In the particular case of CO, stronger chemical interactions are involved reaching high selectivity for this gas. In addition to this intrinsic selectivity, a second corrole has been selected to be used as a reference. This corrole exhibits the same structure as the cobalt corrole, so that it interacts with interferents in a similar way, but has no affinity to the target gaz. Method Two SAWs delay lines (A and B) shown in figure 1 are respectively coated with cobalt- and copper-corroles by mean of a spray coating method. The quantity of corroles deposited on the device is monitored during the process to avoid the deposition of an excessive amount of corroles that would damage the sensor signal compromising its proper functioning. They are then heated at 90°C for an hour under vacuum to remove the ammonia ligands present on the cobalt to prevent the corrole’s degradation in solid state. A gas-mixing bench composed of mass-flow controllers allow for generating various gas mixtures. The so prepared mixtures are sent to a dedicated test chamber that may feature up to six delay lines. Once the sensors are in the chamber, a primary vacuum and a heating is applied to remove any contaminant. A carrier gas composed of different gas of interest (N2, O2, CO2) then flows at 500 sccm through the chamber and CO is then injected at different concentrations. The shift of the synchronous frequency of the delay lines consecutive to gas adsorption is then monitored by mean of a dedicated electronics that delivers information similar to those from a classical network analyzer. Results and Conclusions As we can see in figure 2 the differential acquisition using a copper corrole allows stabilizing the sensor’s signal. Indeed, the drift of the signal on the delay line functionalized with cobalt corrole can be compensated by subtracting the one from the delay line functionalized with copper corrole. This differential acquisition results in a stable basic level of the phase signal and also an improved repeatability of the measurements (figure 3).As expected from previous work [4], we noticed a linear correlation between CO concentration, in the 100 to 7000 ppm range, and the phase shift velocity undergone by the sensors regardless of the composition of the carrier gas (figure 4). From there, we determined the sensitivity of the sensors in presence of interfering gas. It appeared that the presence of 20 % of oxygen, known as the main interferent in this measurement, in the carrier gas tends to lower by 5.5 % the sensitivity in comparison with a pure nitrogen carrier gas (figure 4). The presence of CO2 at 20 000 ppm, which represents 50 times the mean concentration in Paris, lowers the sensitivity by 19.3 % (figure 4). From these results, the selective interaction between the cobalt corrole and CO has been verified and the possibility to use it as part of a selective CO sensor has been evidenced.

Hydrogen is a promising clean energy strategy that can be obtained from the steam reforming of hydrogen-rich materials such as dimethyl ether (DME). In this study, the results of thermodynamic analysis of DME steam reforming were utilised as a framework for the development of correlations for predicting percentage molar yield of each species given a known set of inputs. Response surface methodology—historical data design was used for the work. The effect of temperature, pressure and steam to carbon ratio was shown to be significant in the process. Combinatorial effects of factors and the yield of various chemical species in the product stream were elucidated. The key contribution of this paper is the development of model predictors of all chemical species in the steam reforming of DME. All models developed were shown to be significant. These can afford for quick predictions given a known set of process inputs and can play an important role in the design of DME reforming systems.

The quantity of ash yield and carbon monoxide (CO) emitted during co-combustion of empty fruit bunch (EFB), palm kernel shells (PKS) and kaolin in a grate furnace depend on the fuels mixing ratio, the combustion temperature and duration. These factors can be tuned to minimize ash deposition and CO emission which is partly responsible for the greenhouse effect. In this study, seventy-three (73) data points were obtained from combustion of EFB, PKS and kaolin mixtures based on D-optimal design. Artificial neural network (ANN) model, optimized with Taguchi technique, was developed to predict ash yield (AY) and CO emission from the combustion of the fuel mixture. The data were divided into training, validation and testing in a 2:1:1 relative proportion. The optimized ANN architecture for AY and CO emission were 5-11-3-1 and 5-6-3-1, respectively, with scale conjugate gradient training algorithm and a learning rate of 0.1. Results of the ANN model agreed significantly with the experimental results with coefficients of determination (R2) of 0.96 and 0.93 for ash yield and CO emission, respectively. The mathematical models for the ash and CO emission using the D-optimal design indicate a good fit with R2 of 0.916 and 0.906, respectively. Parametric studies based on the two models showed that ash yield and CO emission reduced with increased combustion temperature and increased fraction of PKS within the temperature range of 800-1000 °C. These results indicated that both ANN and D-optimal can be deployed to select mixture with minimal ash yield and CO emission.