• Korean Chemical Engineering Research
• /
• v.58 no.4
• /
• pp.588-595
• /
• 2020

Prediction and control of nitrogen oxides (NO_x) emission is of great interest in industry due to stricter environmental regulations. Herein, we propose an artificial intelligence (AI)-based framework for prediction of NO_x emission. The framework includes pre-processing of data for training of neural networks and evaluation of the AI-based models. In this work, Long-Short-Term Memory (LSTM), one of the recurrent neural networks, was adopted to reflect the time series characteristics of NO_x emissions. A decision tree was used to determine a time window of LSTM prior to training of the network. The neural network was trained with operational data from a heating furnace. The optimal model was obtained by optimizing hyper-parameters. The LSTM model provided a reliable prediction of NO_x emission for both training and test data, showing an accuracy of 93% or more. The application of the proposed AI-based framework will provide new opportunities for predicting the emission of various air pollutants with time series characteristics.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.22 no.3
• /
• pp.1-11
• /
• 2014

The aim of this study is to investigate the effect of SCR reactor on the exhaust emissions characteristics in order to develop a urea-SCR aftertreatment system for reducing $NO_x$ emissions. The experiments are conducted by using a flue tube LPG steam boiler with the urea-SCR aftertreatment system. The urea-SCR aftertreatment system utilizes the ammonia converted from 17% aqueous urea solution injected in front of SCR catalyst as a reducing agent for reducing $NO_x$ emissions. The equivalence ratio, urea injection amount, ammonia slip and $NO_x$ conversion efficiency relative to boiler load are applied to discuss the experimental results. In this experiment, the average equivalence ratio is calculated by changing only the fuel consumption rate while the intake air amount is constantly fixed at $25,957.11cm^3/sec$. The average equivalence ratios are 1.38, 1.11, 0.81 and 0.57 when boiler loads are 100, 80, 60 and 40%. The $NO_x$ conversion efficiency is raised with increasing urea injection amount, and $NH_3$ slip is also boosted at the same time. Consequently, the $NO_x$ conversion efficiency relative to boiler load should be examined in combination with urea injection amount and $NH_3$ slip. The results are calculated by 89, 85, 77 and 79% for the boiler loads of 100, 80, 60 and 40%. The appropriate amount of urea injection for the respective boiler load can be not discussed by only $NO_x$ emissions, and should be determined by considering the $NO_x$ conversion efficiency, $NH_3$ slip and reactive activation temperature simultaneously. In this study, the urea amounts of 230, 235, 233 and 231 mg/min are injected at the boiler loads of 100, 80, 60 and 40%, and the final $NH_3$ slips are measured by 8.48, 5.58, 11.97 and 11.34 ppm at the same conditions. THC emission is affected by the SCR reactor under other experimental conditions except 100% engine load, and CO emission at only 40% engine load. The rest of exhaust emissions are not affected by the SCR reactor under all experimental conditions.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.14 no.3
• /
• pp.10-21
• /
• 2006

The aim in this study is to develop the combined EGR system with a non-thermal plasma reactor for reducing exhaust emissions and improving fuel economy in turbo intercooler ECU common-rail diesel engines. At the first step, in this paper, the characteristics of performance and $NO_x{\cdot}THC$ emissions under four kinds of engine loads are experimentally investigated by using a four-cycle, four-cylinder, direct injection type, water-cooled turbo intercooler ECU common-rail diesel engine with a combined plasma exhaust gas recirculation(EGR) system operating at three kinds of engine speeds. The EGR system is used to reduce $NO_x$ emissions, and the non-thermal plasma reactor and turbo intercooler system are used to reduce THC emissions. The plasma system is a flat-to-flat type reactor operated by a plasma power supply. The fuel is sprayed by pilot and main injections at the variable injection timing between BTDC $15^{\circ}$ and ATDC $1^{\circ}$ according to experimental conditions. It is found that the specific fuel consumption rate with EGR is increased, but the fuel economy is better than that of mechanical injection type diesel engine as compared with the same output. Results show that $NO_x$ emissions are decreased, but THC emissions are increased, as the EGR rate is elevated. $NO_x$ and THC emissions are also slightly decreased as the applied electrical voltage of the non-thermal plasma reactor is elevated. Thus one can conclude that the influence of EGR in $NO_x$ and THC emissions is larger than that of the non-thermal plasma reactor, but THC emissions are greatly influenced by the non-thermal plasma reactor as the EGR rate is elevated.

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.4
• /
• pp.387-392
• /
• 2015

A method for measuring and analyzing the $NO_x$ in ships is described in $NO_x$ Technical Code 2008. The analysis device, as required by the Code, has been to use a Chemi-luminescence detection method or Heated Chemi-luminescence detection. on the other hand, selective catalytic reduction using $NH_3$ as a reducing agent has an interference effect on the analyzer, and causes measurement error. In this study, the Chemi-luminescence detection method was examined according to how it affects the concentration of $O_2$, CO, $SO_2$, $NH_3$. Fourier transform infrared spectrometry analysis equipment and measurement methods were compared. In order to confirm the effect of the physical interference of the measuring device, it was confirmed by decomposing a measuring device. Consequently, white precipitate and moisture were generated inside the chemiluminescence detection system and I found that affecting interference. The influence of interference highlights the need to consider the minimized $NO_x$ measurement method.

• Applied Chemistry for Engineering
• /
• v.33 no.6
• /
• pp.600-605
• /
• 2022

In this study, nitrogen oxide (NO_x) removal experiments were performed using a graphene based ceramic filter coated with a V₂O₅-WO₃-TiO₂ catalyst. Graphene oxide (GO) was prepared by Hummer's method using graphite, and the reduced graphene oxide was produced by reducing with hydrazine (N₂H₄). Vanadium (V), Tungsten (W), and Titanium (Ti) were coated by the sol-gel method, and then a metal oxide-supported filter was prepared through a calcination process at 350 ℃. A NO_x removal efficiency test was performed for the catalytic ceramic filters with UV light in a humid condition. When graphene oxide (GO) and reduced graphene oxide (rGO) were present on the filter, the NOx removal efficiency was superior to that of the conventional ceramic filter. Most likely, this is due to an improvement in the adsorption properties of NOx molecules on graphene coated surfaces. As the concentration of graphene increased, higher NO_x removal efficiency was confirmed.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.14 no.10
• /
• pp.861-867
• /
• 2001

In this paper, in order to investigate the catalytic effect of the sludge exhausted from waterworks on NO$_{x}$ removal, we measure NO removal characteristics with and without sludge pellets in BaTiO$_3$-sludge packed-bed reactor of plate-plate geometry. NO initial concentration is 50 ppm balanced with air and a gas flow rate is 5ι/min. Gas temperature is changed from 25 to 10$0^{\circ}C$ to investigate the role of sludge pellet on removing active oxygen species and NO$_2$. BaTiO$_3$pellets is filled for coronal discharge at upstream of reactor and sludge pellets is filled for catalytic effect at downstream of reactor. The volume percent of sludge pellets to BaTiO$_3$pellets is changed from 0% to 100% and AC voltage is supplied to the reactor for discharging simulated gases. In the results, when sludge pellets is put at the downstream of plasma reactor, NO removal rate is slightly increased. However, NO$_2$and $O_3$ as by-products during NO removal is significantly decreased from 51ppm without sludge pellets to 5 ppm with sludge pellets and from 50 ppm without sludge pellets to 0.004ppm with sludge pellets, respectively. Therefore, NO$_{x}$(NO+NO$_2$) removal rate is increased up to 93%. It is thought that sludge pellet maybe react with active oxygen species and NO$_2$ generated by corona discharge in surface of BaTiO$_3$pellets, the then NO$_2$O$_3$as by-products are considerably decreased. When we increase gas temperature from room temperature to 10$0^{\circ}C$, NO removal rate is decreased, while NO$_2$ concentration is independent on gas temperature. These result suggest that the removal mechanism of active oxygen species and NO$_2$in sludge pellet is not absorption, but chemical reaction. Therefore we expect that sludge pellets exhausted for waterworks could be used as catalyst for NO$_{x}$ removal with high removal rate and low by-product.oduct.

• New & Renewable Energy
• /
• v.16 no.3
• /
• pp.13-26
• /
• 2020

This study analyzed whether the diffusion of new and renewable energy contributed to mitigating emissions of various air pollutants, including particulate matter, using panel econometric models. The theoretical foundation of such econometric models is based on the Environmental Kuznets Curve (EKC) hypothesis, which assumes an inverted U-shaped relation between national income and environmental pollution, as originally proposed by Grossman and Krueger. We examined whether there are inverted U-, U-shaped, or N-shaped relations between national income and air pollution. We demonstrate that increases in new and renewable energy significantly mitigated emissions of CO, NO_X, and PM2.5. Additionally, we included NO_X, SO_X, PM10, and VOCs as secondary emission sources of PM2.5 and found that emission of PM10 resulted in the highest PM2.5 emissions, followed by NO_X and SO_X emissions. The impact of new and renewable energy on air pollution varied across regions. Increase of new and renewable energy in the Honam region significantly mitigated CO, NO_X, and TSP emissions, while that in the Youngnam and metropolitan areas did not significantly mitigate air pollution overall. There was a U-shaped relationship between air pollution and national income for CO, NO_X, PM2.5, and SO_X, while an inverted N-shape was observed for PM10.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.3452-3457
• /
• 2007

Experiments were carried out in an atmopheric pressure, lab-scale gas turbine combustor to see the effect of partial premixing on unstable flame structure and $NO_X$ emission characteristics. The swirl angle is 45 deg., fuel-air mixing degrees were varied 0, 50, and 100% respectively at equivalence ration ranging from 0.53 to 0.79. The evaluation of phased-locked OH chemiluminescence images were acquired with an ICCD. $NO_X$ emission characteristics were also investigated at each experimental condition. The effect of the fuel-air mixing degree on the flame structure was obtained from phase-locked $OH^*$ images. And it was obtained from local heat release characteristics that the information about the region which the combustion instability was amplified or damped. It also could be confirmed that ${\sigma}$ has greatly influence on $NO_X $emission characteristics at lean regimes. It would be expected that it could provide invaluable data for understanding the mechanism of combustion instability

• Journal of ILASS-Korea
• /
• v.21 no.4
• /
• pp.200-206
• /
• 2016

Recently, the certification procedure for exhaust emission regulation of LDV has tested with the NEDC mode in the laboratory. But the on-road exhaust emissions exceed the standard emission limits. Therefore, it is important to analyze the real-driving emissions (RDE) with a portable emissions measurement system (PEMS). In present study, the on-road emissions were measured with a PEMS and evaluated by moving averaging window (MAW) method. Also, it was compared with the $CO_2$ and $NO_x$ emissions for real-driving and test modes from euro-6 light-duty vehicles equipped with SCR and LNT systems. In results, on-road $NO_x$ emission has been 2.3-10.0 times higher than the standard $NO_x$ emission limit on NEDC mode. The reason was that the test modes did not reflect traffic and various real-driving patterns sufficiently.

• Proceedings of the KSME Conference
• /
• 2001.06d
• /
• pp.37-42
• /
• 2001

This paper computes the bluff-body stabilized jet and flame. This study numerically investigates the nonpremixed $C_{2}H_{4}-air$ jet for the nonreacting case and the nonpremixed $CH_{3}OH-air$ turbulent flames for the reacting case using the laminar flamelet model on modified KIVA2 code. And this study predicts $NO_{x}$ formation characteristics using Eulerian Particle Flamelet Model. In the present study, the turbulent combustion model is applied to analyze both nonreacting and reacting case. And both standard $k-{\varepsilon}$ model and modified $k-{\varepsilon}$ model are used in nonreacting case. Calculations are compared with experimental data in terms of velocity, mixture fraction, mixture fraction Root Mean Square and Temperature. The present model correctly predicts the essential features of flame structures and $NO_{x}$ formation characteristics in the bluff-body stabilized flames.