Prediction Of Pollutant Emissions Research Articles

A Deep Learning Method for the Prediction of Pollutant Emissions from Internal Combustion Engines

A Deep Learning Method for the Prediction of Pollutant Emissions from Internal Combustion Engines

Investigation on NO and N2O emissions characteristics in deep peak regulation circulating fluidized bed boilers

Abstract To adapt to the increasing proportion of new energy power generation capacity, coal power must transition from its traditional role as the primary power source to serving as a fundamental backup and system regulation energy source. The circulating fluidized bed technology is known for its wide range of load regulation capabilities; however, emissions of pollutants during load regulation can exhibit significant variability. This study utilized Aspen Plus software to develop a circulating fluidized bed combustion model based on a gas-solid fluid dynamics model, an equivalent coal pyrolysis model, and a multi-phase macroscopic combustion reaction dynamics model of pyrolysis products. This model was used to predict both the temperature distribution within the furnace chamber and the distribution of pollutant concentrations. Predictions of pollutant emissions from 100 % load to 30 % load of the circulating fluidized bed were explored under the original combustion condition and 5 % proportion of recirculated flue gas. Under the primary combustion condition, the emission concentration of NO x showed a decreasing and then increasing trend with decreasing load, while the concentration of nitrous oxide, in contrast to NO x , showed an increasing and then decreasing trend. The effect of recirculated flue gas on pollutant emissions was not significant at reduced loads. This study aims to provide technical support and theoretical guidance for the management of pollutant emissions from the deep peak regulation of actual circulating fluidized beds.

Using a Machine Learning Approach to Evaluate the NOx Emissions in a Spark-Ignition Optical Engine

Currently, machine learning (ML) technologies are widely employed in the automotive field for determining physical quantities thanks to their ability to ensure lower computational costs and faster operations than traditional methods. Within this context, the present work shows the outcomes of forecasting activities on the prediction of pollutant emissions from engines using an artificial neural network technique. Tests on an optical access engine were conducted under lean mixture conditions, which is the direction in which automotive research is developing to meet the ever-stricter regulations on pollutant emissions. A NARX architecture was utilized to estimate the engine’s nitrogen oxide emissions starting from in-cylinder pressure data and images of the flame front evolution recorded by a high-speed camera and elaborated through a Mask R-CNN technique. Based on the obtained results, the methodology’s applicability to real situations, such as metal engines, was assessed using a sensitivity analysis presented in the second part of the work, which helped identify and quantify the most important input parameters for the nitrogen oxide forecast.

Prediction of Pollutant Emissions in Various Cases in Road Transport

The road transport sector is a key source of carbon dioxide and air pollutants. Mathematical modeling is frequently used to assess the sector’s contribution to the total national emissions budget (inventory). The present article focuses on studying the impacts of the fuel used (LPG, CNG, gasoline, diesel, and biofuel), the Euro standard, and the structure of vehicles on CO2, NOX, and PM2.5 emissions. This paper presents the results of mathematical simulations of the influence of the fuel type and Euro standards on pollutant emissions. Two scenarios were considered in terms of the effect on CO2, NOX, and PM2.5 emissions: one focused on changing the current fleet and introducing Euro 6/VI vehicles, and the second scenario focused on cities adding low-pollution zones (only Euro 6 vehicles and PHEVs, HEVs, and BEVs entering the city). The results of the simulations showed that Euro 6/IV vehicles emit significantly less PM2.5 and NOX, and biofuels can substantially contribute to reducing emissions.

Design and qualification of a bench-scale model for municipal waste-to-energy combustion

ABSTRACT This paper reports the design and qualification of the first purpose-built, bench-scale reactor system to model the municipal waste-to-energy combustion of fluorinated polymers. Using the principle of similarity, the gas-phase combustion zone of a typical municipal waste-to-energy plant has been scaled down to the bench with a focus on chemical similarity. Chemical similarity is achieved in large part through the use of methanol as a surrogate for municipal solid waste (MSW). Review of prior research shows that methanol is one of the major volatile products expected during MSW thermal conversion in the fuel bed of waste-to-energy plants. Like full-scale waste-energy plants, the design of the bench-scale model includes a flame zone and a post-flame zone. Maintaining steady methanol vapor flow premixed with air to the model reactor system ensures stable combustion resulting in bench-scale CO emission levels comparable to those of full-scale waste-to-energy plants. Since investigation of fluorinated polymer combustion includes trace analysis of exhaust gas for perfluorooctanoic acid (PFOA), qualification testing focused on PFOA collection efficiency. High PFOA collection efficiency (>90%) demonstrated the capability of the reactor system in transporting and absorbing PFOA that may be generated during high-temperature combustion testing of fluorinated polymers. Overall, the bench-scale system is qualified for its intended use to investigate potential generation of PFOA from combustion of fluorinated polymers under conditions representative of waste-to-energy combustion. Implications: Decision makers depend on environmental researchers to provide reliable predictions of pollutant emissions from waste combustion of polymers at end of product life. Reliable predictions are especially important with regard to questions about potential PFOA emissions from municipal waste combustion of fluorinated polymers. Results from qualification testing confirm that the novel bench-scale model reactor system is capable of representing gas-phase municipal waste combustion behavior upstream of air pollution control and generating representative exhaust gas samples for off-line trace-level analysis of PFOA.

Low-Order Reactor-Network-Based Prediction of Pollutant Emissions Applied to FLOX® Combustion

Prediction of pollutant emissions is a key aspect of modern combustor design in energy conversion systems. In the presented work, a simple and robust model based on low-order reaction networks is applied to a FLOX® laboratory combustor at atmospheric conditions. The applied approach is computationally cheap and therefore highly suited for design studies. Steady-state CFD RANS simulations are carried out, serving as a basis for the network generation algorithm. CFD results are validated with experimental data for flow field and combustion. Different degrees of fidelity of reactor network models are taken into consideration and findings are opposed to measurements, evaluating the quality of the low-fidelity models. Validation of CO and NOx emission results of reactor network modeling provides accurate qualitative and quantitative reproduction of experimental findings, depending on the degree of heat loss applied on the combustion system. The introduced approach is therefore readily applicable to large-scale, industrial, and gas turbine combustion.

Open burning and open detonation of explosives. Prediction of pollutant emissions

Open burning and open detonation of explosives. Prediction of pollutant emissions

A review on the applications of the chemical reactor network approach on the prediction of pollutant emissions

PurposeThe purpose of this paper is to review the applications of the chemical reactor network (CRN) approach for modeling the combustion in gas turbine combustors and classify the CRN construction methods that have been frequently used by researchers.Design/methodology/approachThis paper initiates with introducing the CRN approach as a practical tool for precisely predicting the species concentrations in the combustion process with lower computational costs. The structure of the CRN and its elements as the ideal reactors are reviewed in recent studies. Flow field modeling has been identified as the most important input for constructing the CRNs; thus, the flow field modeling methods have been extensively reviewed in previous studies. Network approach, component modeling approach and computational fluid dynamics (CFD), as the main flow field modeling methods, are investigated with a focus on the CRN applications. Then, the CRN construction approaches are reviewed and categorized based on extracting the flow field required data. Finally, the most used kinetics and CRN solvers are reviewed and reported in this paper.FindingsIt is concluded that the CRN approach can be a useful tool in the entire process of combustion chamber design. One-dimensional and quasi-dimensional methods of flow field modeling are used in the construction of the simple CRNs without detailed geometry data. This approach requires fewer requirements and is used in the initial combustor designing process. In recent years, using the CFD approach in the construction of CRNs has been increased. The flow field results of the CFD codes processed to create the homogeneous regions based on construction criteria. Over the past years, several practical algorithms have been proposed to automatically extract reactor networks from CFD results. These algorithms have been developed to identify homogeneous regions with a high resolution based on the splitting criteria.Originality/valueThis paper reviews the various flow modeling methods used in the construction of the CRNs, along with an overview of the studies carried out in this field. Also, the usual approaches for creating a CRN and the most significant achievements in this field are addressed in detail.

Estimation and prediction of pollutant emissions from agricultural and construction diesel machinery in the Beijing-Tianjin-Hebei (BTH) region, China☆

Both agricultural and construction machinery are important non-road sources of atmospheric pollution, with total hydrocarbons (THC), nitrogen oxides (NOx) and particulate matter (PM) emissions accounting for more than 60% of the total emissions from all non-road mobile sources in China. However, there exist relatively few efforts to establish the emission inventory for these machineries. This study attempted to estimate and predict air pollutant emissions from agricultural and construction diesel machinery, using the Beijing-Tianjin-Hebei (BTH) region as the case study area. The results show that total emissions of PM10, PM2.5, THC, NOX, CO and SO2 in 2015 were 41.10, 38.80, 86.14, 520.41, 379.01 and 17.32 Kt respectively. The contribution of agricultural machinery was slightly higher than that of construction machinery, accounting for 60–71% of the total. Moreover, emissions of various pollutants (except SO2) from agricultural machinery were mainly distributed in central Hebei (Cangzhou, Shijiazhuang and Baoding), while emissions from construction machinery were mainly distributed in Beijing and Tianjin. The prediction suggest that the total emissions of agricultural and construction diesel machinery in the BTH region would increase by 6% in 2020 and 9% in 2025. Moreover, pollutant emissions from construction machinery would contribute from 29% to 40% in 2015 to 34%–61% in 2025. These results could provide important information for making effective mitigation measures of non-road mobile sources.

Modeling diesel combustion with tabulated kinetics and different flame structure assumptions based on flamelet approach

Computational fluid dynamics analysis represents a useful approach to design and develop new engine concepts and investigate advanced combustion modes. Large chemical mechanisms are required for a correct description of the combustion process, especially for the prediction of pollutant emissions. Tabulated chemistry models allow to reduce significantly the computational cost, maintaining a good accuracy. In the present work, an investigation of tabulated approaches, based on flamelet assumptions, is carried out to simulate turbulent Diesel combustion in the Spray A framework. The Approximated Diffusion Flamelet is tested under different ambient conditions and compared with Flamelet Generated Manifold, and both models are validated with Engine Combustion Network experimental data. Flame structure, combustion process and soot formation were analyzed in this work. Computed results confirm the impact of the turbulent–chemistry interaction on the ignition event. Therefore, a new look-up table concept Five-Dimensional-Flamelet Generated Manifold, that accounts for an additional dimension (strain rate), has been developed and tested, giving promising results.

Corrigendum to “Prediction of Pollutant Emissions from Bluff-Body Stabilised Nonpremixed Flames”

Corrigendum to “Prediction of Pollutant Emissions from Bluff-Body Stabilised Nonpremixed Flames”

Neural network predictions of pollutant emissions from open burning of crop residues: Application to air quality forecasts in southern China

Open burning of crop residues is a strong seasonal source of air pollutants in many parts of China, but the large day-to-day variability of the associated emissions pose a great challenge for air quality forecasts. Here we developed back-propagation neural network (BPNN) ensembles to forecast the daily fire pixel counts in Southern China during the month of January. The BPNN ensembles were trained using daily assimilated surface meteorological data (including air temperature, relative humidity, pressure, and winds) and daily fire pixel observations from the Moderate Resolution Imaging Spectroradiometer (MODIS) for the month of January during the years 2003–2012. We showed that the BPNN ensembles successfully forecasted the day-to-day variability and the interannual variability of fire pixel counts over Southern China in January of the years 2013–2015, with correlation coefficients of 0.6–0.8 against the MODIS observations. We used the forecasted daily fire pixel counts in January 2014 and January 2015 to scale the climatological January biomass burning emissions from the Fire Inventory from NCAR (FINN) and applied the resulting forecasted daily biomass burning emissions to drive the WRF-Chem regional air quality model. The use of BPNN-ensemble-forecasted daily biomass burning pollutant emissions led to significant improvements in the daily forecasts of PM2.5 concentrations in Southern China for January 2014, with the mean bias of the simulated surface PM2.5 concentrations reduced from −9.1% to −1.2%. We repeated the sensitivity simulations for January 2015 and also found a modest improvement when using the forecasted daily biomass burning pollutant emissions (mean bias of the simulated surface PM2.5 concentrations reduced from −5.8% to −2%). Our approach can be applied to other source regions of biomass burning emissions to improve the accuracy of daily air quality forecasts.

A Brief Review of the Combustion Diagnosing Techniques for Coal-Fired Boilers of Power Plants in China

China has a large number of coal-fired power plants which are commonly suffered with many problems like uncertain coal source, frequent variation of coal quality, blended coal combustion and great fluctuation of the load. It is dire need of a system that can conduct combustion diagnosis, which can enables new horizon in optimization and to ensure the safety and efficiency of a unit. The contents of the combustion diagnosis include flame stability evaluation and temperature measurement. The latter provides quantitative information of the furnace, which can be used for unburnt carbon and pollutant emissions prediction. Diagnosing techniques are generally based on flame radiation intensity detect, flame image analysis, acoustic pyrometry and artificial intelligence analysis. The performance and engineering application of these techniques are reviewed and compared in this study. Artificial intelligence analysis based on the flame radiation intensity signal is recommended for flame stability evaluation as it is easily implemented in practical boilers with good accuracy and low cost. The challenge of the method based on flame image is thought to be that the reliability of the hardware system for flame image collection remains to be improved as they are easily affected by the high temperatures and fouling in engineering applications. The particles produced in the furnace is the hinder for application of the acoustic pyrometry in coal-fired boilers.

Large eddy simulation/dynamic thickened flame modeling of a high Karlovitz number turbulent premixed jet flame

Due to the complex multiscale interaction between intense turbulence and relatively weak flames, turbulent premixed flames in the thin and broken reaction zones regimes exhibit strong finite-rate chemistry and strain effects and are hence challenging to model. In this work, a laboratory premixed jet flame in the broken reaction zone, which has recently been studied using direct numerical simulation (DNS), is modeled using a large eddy simulation (LES)/dynamic thickened flame (DTF) approach with detailed chemistry. The presence of substantial flame thickening due to strong turbulence-chemistry interactions, which can be characterized by a high Karlovitz number (Ka), requires the DTF model to thicken the flame in an adaptive way based on the local resolution of flame scales. Here, an appropriate flame sensor and strain-sensitive flame thickness are used to automatically determine the thickening location and thickening factor, respectively. To account for finite-rate chemistry and strain effects, the chemistry is described in two different ways: (1) detailed chemistry denoted as full transport and chemistry (FTC), and (2) tabulated chemistry based on a strained premixed flamelet (SPF) model. The performance of the augmented LES/DTF approach for modeling the high Ka premixed flame is assessed through detailed a posteriori comparisons with DNS of the same flame. It is found that the LES/DTF/FTC model is capable of reproducing most features of the high Ka turbulent premixed flame including accurate CO and NO prediction. The LES/DTF/SPF model has the potential to capture the impact of strong turbulence on the flame structure and provides reasonable prediction of pollutant emissions at a reasonable computational cost. In order to identify the impact of aerodynamic strain, the turbulent flame structure is analyzed and compared with unstrained and strained premixed flamelet solutions. The results indicate that detailed strain effects should be considered when using tabulated methods to model high Ka premixed flames.

Numerical Study of Natural Gas/Diesel Reactivity Controlled Compression Ignition Combustion with Large Eddy Simulation and Reynolds-Averaged Navier–Stokes Model

In the current study, a comparative study is performed using Large Eddy Simulation (LES) and Reynolds-averaged Navier–Stokes (RANS) turbulence models on a natural gas/diesel Reactivity Controlled Compression Ignition (RCCI) engine. The numerical results are validated against the available research work in the literature. The RNG (Re-Normalization Group) k − ε and dynamic structure models are employed to model turbulent flow for RANS and LES simulations, respectively. Parameters like the premixed natural gas mass fraction, the second start of injection timing (SOI2) of diesel and the engine speed are studied to compare performance of RANS and LES models on combustion and pollutant emissions prediction. The results obtained showed that the LES and RANS model give almost similar predictions of cylinder pressure and heat release rate at lower natural gas mass fractions and late SOI2 timings. However, the LES showed improved capability to predict the natural gas auto-ignition and pollutant emissions prediction compared to RANS model especially at higher natural gas mass fractions.

Turbulent flow-field effects in a hybrid CFD-CRN model for the prediction of NOx and CO emissions in aero-engine combustors

The paper presents a numerical study where a hybrid CFD-Chemical Reactor Network (CRN) approach is used to predict pollutant emissions in a tubular combustor for aero-engine applications. A fully-automated clustering of the simulated flow field with the generation of a reactor network representative of the main flow features is exploited. Similar cells are detected and grouped using a two step approach, the first one based only on aerodynamic criteria for turbulent flows followed by a chemical refinement based on mixture fraction. A formulation for turbulent diffusion fluxes is introduced in the reactor code to model species and energy exchanges between reactors. Three different operating conditions are studied for which measured NOx and CO are available. Results highlight the importance of including turbulent diffusion in the network solution. The accurate prediction of pollutant emissions at different load points confirms that CFD-CRN is a valid and flexible approach for preliminary assessment of aero-engine combustor emissions in the design phase.

Prediction of NOx Emissions and Fuel Consumption of a City Bus under Real Operating Conditions by Means of Biharmonic Maps

The prediction of pollutant emissions and fuel consumption under real operating conditions of any motor vehicle requires the use of complex mathematical models and experimental tools. In the present research, biharmonic maps (BM) were used to predict NOx (nitrogen oxides) emissions and relative fuel–air ratio (Fr) of a passenger city bus. For the collection of experimental data, an instrumented city bus was tested during real passenger transportation. The data were classified into four dynamic sequences: acceleration, idling, deceleration with fuel consumption, and deceleration without fuel consumption. Among them, the acceleration sequence was selected due to high NOx emissions and high fuel consumption. Experimental results were in good agreement with BM predictions. Significant parameters for predicting NOx concentration were vehicle velocity and relative fuel–air ratio (Fr). While for predicting Fr, significant parameters were the exhaust gas flow (EGF) rate, vehicle velocity, and NOx concentration, BM is a good technique for predicting NOx emissions and Fr of a city bus under transient operation.

Prediction of Pollutant Emissions of Biomass Flames Through Digital Imaging, Contourlet Transform, and Support Vector Regression Modeling

This paper presents a method for the prediction of NO x emissions in a biomass combustion process through the combination of flame radical imaging, contourlet transform and Zernike moment (CTZM), and least squares support vector regression (LS-SVR) modeling. A novel feature extraction technique based on the CTZM algorithm is developed. The contourlet transform provides the multiscale decomposition for flame radical images and the selected operator based on Zernike moments is designed to provide the well-defined structure for the images. The resulted image features are a variable structure, which is originated from the CTZM. Finally, the variable features of the images of four flame radicals (OH*, CN*, CH*, and $\text{C}^{\ast }_{2}$ ) are defined. The relationship between the variable features of radical images and NO x emissions is established through radial basis function network modeling, SVR modeling, and the LS-SVR modeling. A comparison between the three modeling approaches shows that the LS-SVR model outperforms the other two methods in terms of root-mean-square error and mean relative error criteria. In addition, the structure of the image features has a significant impact on the performance of the prediction models. The test results obtained on a biomass-gas fired test rig show the effectiveness of the proposed technical approach for the prediction of NO x emissions.

A fully coupled, parallel approach for the post-processing of CFD data through reactor network analysis

In this paper we describe and apply a parallel code, named KPPSMOKE, for the prediction of pollutant emissions from combustion devices operating in turbulent conditions. The approach is based on the kinetic post-processing of CFD simulations, which are transformed into equivalent networks of perfectly stirred reactors and solved using a detailed kinetic mechanism (hundreds of species). The numerical algorithm is based on a fully-coupled technique, in which the highly non-linear mass balance equations are solved together, by alternating different resolution methods in order to ensure high accuracy and fast convergence. As a result of KPPSMOKE parallel structure, large reactor networks characterizing industrial devices (105–106 reactors) can be solved in reasonable times (∼hours).The accuracy and the reliability of the algorithm was demonstrated on a lab-scale burner and on a full-scale industrial device, i.e. a combustor for aircrafts. The numerical performance was also assessed in terms of parallel efficiency and speedup.

A simplified approach for predicting NO formation in MILD combustion of CH 4–H 2 mixtures

The direct implementation into CFD codes of large kinetic mechanisms for the prediction of pollutant emissions is still unfeasible, due to computer time limitations which become particularly relevant when considering the typical scale of the industrial applications. Therefore, simplified modeling approaches are generally adopted, as they allow reducing the computational effort associated with the numerical simulations. With regard to NO formation, simple one-step mechanisms are used to describe each of the relevant routes contributing to the overall generation of NO, i.e., thermal, prompt. The main drawback associated to a simplified NO formation approach lies, however, in the extreme sensitivity of the lumped rates on the thermo chemical state which define the combustion system of interest. Then, a proper description of turbulence/chemistry interactions must be employed in the CFD model, to provide a realistic background for the estimation of NO emissions. This becomes particularly important in MILD combustion regime, which generally requires an accurate description of the gas-phase oxidation, due to the kinetic control on the overall combustion process. The present paper discusses key aspects and requisite for predicting NO formation in MILD combustion regime. The approach is based on the direct coupling of simplified NO mechanisms to the CFD calculation and is applied to different MILD conditions. Simulations are carried out for a set of experimental runs performed on a self-recuperative MILD burner, varying the hydrogen content in the fuel stream from 0% up to 50% by wt.