A robust framework to predict mercury speciation in combustion flue gases

Temperature Prediction of Seasonal Frozen Subgrades Based on CEEMDAN-LSTM Hybrid Model

Improving the temperature prediction accuracy for subgrades in seasonally frozen regions will greatly help improve the understanding of subgrades' thermal states. Due to the nonlinearity and non-stationarity of the temperature time series of subgrades, it is difficult for a single general neural network to accurately capture these two characteristics. Many hybrid models have been proposed to more accurately forecast the temperature time series. Among these hybrid models, the CEEMDAN-LSTM model is promising, thanks to the advantages of the long short-term memory (LSTM) artificial neural network, which is good at handling complex time series data, and its combination with the broad applicability of the complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) in the field of signal decomposition. In this study, by performing empirical mode decomposition (EMD), ensemble empirical mode decomposition (EEMD), and CEEMDAN on temperature time series, respectively, a hybrid dataset is formed with the corresponding time series of volumetric water content and frost heave, and finally, the CEEMDAN-LSTM model is created for prediction purposes. The results of the performance comparisons between multiple models show that the CEEMDAN-LSTM model has the best prediction performance compared to other decomposed LSTM models because the composition of the hybrid dataset improves predictive ability, and thus, it can better handle the nonlinearity and non-stationarity of the temperature time series data.

A novel support vector machine ensemble model for estimation of free lime content in cement clinkers

Free lime (f-CaO) content is a crucial quality parameter for cement clinkers in rotary cement kiln. Due to lack of hardware sensors, f-CaO content in cement clinker is mostly obtained by offline laboratory measurement, making timely control rather difficult and even impossible. In this work, a soft sensor approach named as support vector machine ensemble (ESVM) model is proposed to estimate f-CaO content. The process data employed to train and test the model were collected from a cement plant in China, covering a time span of about 30 days. The raw data were preprocessed by filters and time-series matching. The processed data were then clustered by fuzzy c-means clustering algorithm to capture process features at different operating conditions. For each individual cluster, a base SVM regressor was trained to estimate f-CaO content. Finally, an ensemble model consisting of four base SVM regressors was established to estimate f-CaO content at multifarious process conditions. The effectiveness of the proposed ESVM model was investigated by comparing it with manual measurements and other models available in literature. The results demonstrate that the proposed ESVM model achieves improvements in model accuracy as well as generalization capability. The proposed ESVM model has a broad application space in cement production process for automatic monitoring of f-CaO content.

Active methods of mercury removal from flue gases

Due to its adverse impact on health, as well as its global distribution, long atmospheric lifetime and propensity for deposition in the aquatic environment and in living tissue, the US Environmental Protection Agency (US EPA) has classified mercury and its compounds as a severe air quality threat. Such widespread presence of mercury in the environment originates from both natural and anthropogenic sources. Global anthropogenic emission of mercury is evaluated at 2000 Mg year^-1. According to the National Centre for Emissions Management (Pol. KOBiZE) report for 2014, Polish annual mercury emissions amount to approximately 10 Mg. Over 90% of mercury emissions in Poland originate from combustion of coal.The purpose of this paper was to understand mercury behaviour during sub-bituminous coal and lignite combustion for flue gas purification in terms of reduction of emissions by active methods. The average mercury content in Polish sub-bituminous coal and lignite was 103.7 and 443.5 μg kg^-1. The concentration of mercury in flue gases emitted into the atmosphere was 5.3 μg m^-3 for sub-bituminous coal and 17.5 μg m^-3 for lignite. The study analysed six low-cost sorbents with the average achieved efficiency of mercury removal from 30.6 to 92.9% for sub-bituminous coal and 22.8 to 80.3% for lignite combustion. Also, the effect of coke dust grain size was examined for mercury sorptive properties. The fine fraction of coke dust (CD) adsorbed within 243-277 μg Hg kg^-1, while the largest fraction at only 95 μg Hg kg^-1. The CD fraction < 0.063 mm removed almost 92% of mercury during coal combustion, so the concentration of mercury in flue gas decreased from 5.3 to 0.4 μg Hg m^-3. The same fraction of CD had removed 93% of mercury from lignite flue gas by reducing the concentration of mercury in the flow from 17.6 to 1.2 μg Hg m^-3. The publication also presents the impact of photochemical oxidation of mercury on the effectiveness of Hg vapour removal during combustion of lignite. After physical oxidation of Hg in the flue gas, its effectiveness has increased twofold.

Method development and validation for total mercury determination in coke oven gas combining a trap sampling method with CVAAS detection

Coke oven gas is one of the by-products of the coal coking process. It is used as a fuel in the coking plant, but also as a raw material in the chemical industry to produce methanol, syngas or environment-friendly, low-CO₂ hydrogen fuel. Due to the reasons mentioned above, the knowledge of coke oven gas pollutants such as mercury is a key issue. To determine the mercury in the cleaned coke oven gas a trap sampling method combined with CVAAS mercury detection was developed, optimized and validated. In order to perform the sampling process the traps filled with activated carbon were used. The determination of mercury in the traps material was performed by means of an MA-2 mercury analyzer. During the optimization of the method one selected the trap material, sample volume and flow rate. The optimal volume of the coke oven gas sample was 3 dm³ and the flow rate was 18 dm³/h (per one trap). The developed method was validated according to the Eurachem recommendation and was applied to determine mercury in the clean coke oven gas. The coke oven gas sampling was performed in a coking plant in Poland. The average concentration of mercury in the clean coke oven gas was 3.2 ± 0.3 μg/m³_N (k = 2) for n = 18.

Study on emission of hazardous trace elements in a 350 MW coal-fired power plant. Part 1. Mercury

Hazardous trace elements (HTEs), especially mercury, emitted from coal-fired power plants had caused widespread concern worldwide. Field test on mercury emissions at three different loads (100%, 85%, 68% output) using different types of coal was conducted in a 350 MW pulverized coal combustion power plant equipped with selective catalytic reduction (SCR), electrostatic precipitator and fabric filter (ESP + FF), and wet flue gas desulfurization (WFGD). The Ontario Hydro Method was used for simultaneous flue gas mercury sampling for mercury at the inlet and outlet of each of the air pollutant control device (APCD). Results showed that mercury mass balance rates of the system or each APCD were in the range of 70%-130%. Mercury was mainly distributed in the flue gas, followed by ESP + FF ash, WFGD wastewater, and slag. Oxidized mercury (Hg²⁺) was the main form of mercury form in the flue gas emitted to the atmosphere, which accounted for 57.64%-61.87% of total mercury. SCR was favorable for elemental mercury (Hg⁰) removal, with oxidation efficiency of 50.13%-67.68%. ESP + FF had high particle-bound mercury (Hg^p) capture efficiency, at 99.95%-99.97%. Overall removal efficiency of mercury by the existing APCDs was 58.78%-73.32%. Addition of halogens or oxidants for Hg⁰ conversion, and inhibitors for Hg⁰ re-emission, plus the installation of a wet electrostatic precipitator (WESP) was a good way to improve the overall removal efficiency of mercury in the power plants. Mercury emission factor determined in this study was from 0.92 to 1.17 g/10¹²J. Mercury concentration in the emitted flue gas was much less than the regulatory limit of 30 μg/m³. Contamination of mercury in desulfurization wastewater should be given enough focus.