In this paper, multi-layer feed forward neural networks are used to predict the lower heating value of gas (LHV), lower heating value of gasification products including tars and entrained char (LHVp) and syngas yield during gasification of municipal solid waste (MSW) during gasification in a fluidized bed reactor. These artificial neural networks (ANNs) with different architectures are trained using the Levenberg-Marquardt (LM) back-propagation algorithm and a cross validation is also performed to ensure that the results generalise to other unseen datasets. A rigorous study is carried out on optimally choosing the number of hidden layers, number of neurons in the hidden layer and activation function in a network using multiple Monte Carlo runs. Nine input and three output parameters are used to train and test various neural network architectures in both multiple output and single output prediction paradigms using the available experimental datasets. The model selection procedure is carried out to ascertain the best network architecture in terms of predictive accuracy. The simulation results show that the ANN based methodology is a viable alternative which can be used to predict the performance of a fluidized bed gasifier.
According to World Bank data, about 4 billion tonnes of waste is generated per year, out of which cities' alone contribute 1.3 billion tonnes of solid waste. This volume is forecast to increase to 2.2 billion tonnes by 2025. Three-fourths of this waste is disposed of in landfills, with only one fourth being recycled. It is expected that in lower income countries waste generation will double in the next 25 years (Hoornweg and Bhada-Tata, 2012). With rapid industrial growth and growing world population, most developing countries are facing acute disposal problem for municipal solid waste (MSW). MSW refers to the discarded materials from household wastes such as kitchen garbage, paper, wood, food waste, cotton as well as materials derived from fossil fuels such as plastic, rubber etc. (Cheng and Hu, 2010). In urban areas significant environmental problems are arising from the disposal of MSW which have led to major concerns regarding human health and environment. These issues are common to both developed as well as developing countries (Pires et al., 2011). Furthermore, these issues are stimulating the need for further development of treatment technologies to meet these global challenges. The new European sustainable development strategy (EU, 2009) promotes thermal treatment processes to recover energy from MSW while tackling the issues related to climate change.
There are several processes that could treat MSW including thermal, biochemical and mechanical processes. Incineration technology is widely used to process MSW, but the control of NOx, SOx, nano-particle, dioxins and furans emissions are challenging (Cheng and Hu, 2010). In a quest for a sustainable waste treatment technology, waste to energy (WtE) technology has been reviewed by (Brunner and Rechberger, 2015). The study concluded that due to the advancement in combustion and air pollution control technologies WtE plants are useful for energy and material recovery from waste without having adverse effects on environment. The impact on the environment of thermal treatment of waste with energy recovery was evaluated by Pavlas et al. (2010) who concluded that thermal treatment of MSW with energy recovery was undoubtedly one of the best techniques. WtE not only offers an alternative to treat the waste but also produces clean energy which can offset primary energy consumption in conventional heat and power units. In general, WtE plants are considered as carbon neutral but they are not. The total carbon content present in the MSW is bound with various materials present in the waste. It was found that more than half of the carbon present is biogenic in nature but the remaining part originates from fossil fuels which cannot be considered as biogenic carbon (Gohlke, 2009). As per the EU’s new directive, each WtE plant has to report how much electricity was produced from the renewable sources present in the waste feed. The measured biogenic CO2 fraction in the flue gas from an incinerator plant in Netherlands was between 48-50% (Palstra and Meijer, 2010) whereas, in Austria the ratio of biogenic to anthropogenic energy content in MSW was reported in the range 36-53% (Fellner et al., 2007).
Thermal treatment technologies for MSW have been extensively reviewed by Arena (2012); Leckner (2015); Lombardi et al. (2015); Malkow (2004) and it was proposed that an alternative to combustion is to gasify the MSW for energy recovery. To date, gasification processes have been investigated by several contemporary researchers and extensively reviewed by Gómez-Barea and Leckner (2010). Thermal gasification provides flexibility for the production of heat and power based on clean biomass derived syngas (Basu, 2010). In addition, thermochemical conversion technologies can reduce the original volume of wastes disposed by 80-95% along with energy recovery (Rand et al., 1999). Lately, gasification of solid wastes which originates from the household or industrial sectors have received increasing attention by researchers. The syngas from MSW can be used for heating and production of electricity to offset the use of fossil fuels. However, gasification of MSW is not widespread. The major barrier that has prevented the widespread uptake of advanced gasification technologies for treating MSW has been the higher ash content in the feed making the gasification operation difficult. In addition, high amounts of tar and char contaminants in the produced gas make it unsuitable for power production using energy efficient gas engines or turbines.
A comprehensive review of fluidized bed biomass gasification model was presented by Gómez-Barea and Leckner (2010). In the past, different modelling approaches starting from black box modelling to thermodynamic equilibrium, kinetic rate, fluid-dynamics, neural network and genetic programming models (Pandey et al., 2015;Puig-Arnavat et al., 2010) and Gaussian process based Bayesian inference (Pan and Pandey, 2016) were applied for modelling gasificatio
This content is AI-processed based on open access ArXiv data.
