Biomass higher heating value prediction from ultimate analysis using multiple regression and genetic programming

The heating value is one of the most important properties of biomass materials, as it can be used for design calculations or numerical simulations of the thermal conversion systems of biomass. The heating value can be determined experimentally or numerically. The determination of the heating value experimentally (by using a bomb calorimeter) involves laborious measurements while an ultimate analysis can be performed using automatic equipment. In this study, a multiple linear regression analysis was used to develop an empirical equation for the estimation of the higher heating value of Oil Palm Fronds (OPFs) using an ultimate analysis. An empirical equation based on the main elements carbon, hydrogen and oxygen (HHV = 0.879C+0.321H+0.056O-24.826) was the most accurate model, with a correlation coefficient (R2) of 0.92, Average Absolute Error (AAE) of 1.4% and Average Bias Error (ABE) of 0.16%.

Thirty-nine different species of waste biomass materials that include woody or herbaceous resources as well as nut shells and juice pulps were used to develop empirical equations to predict the calorific value based on the proximate analysis results. Ten different linear/nonlinear equations that contain proximate analysis ingredients including or excluding the moisture content were tested by means of least-squares method to predict the HHV (higher heating value). Prediction performance of each equation was evaluated considering the experimental and the predicted values of HHV and the criteria of MAE (mean absolute error), AAE (average absolute error), and ABE (average bias error). It was concluded that the presence of moisture as a parameter improves the prediction performance of these equations. Also, the samples were classified into two subsets according to their fixed carbon (FC)/ash values and then the correlations were repeated for each subset. Both the full set of samples and the subsets showed a similar trend that the presence of moisture in equations enhances the prediction performance. Also, the FC content may be disregarded from the equation of the calorific value prediction when the FC/ash ratio is lower than a given value.

Interest in biomass has increased due to current environmental issues, and biomass analysis is usually performed using element and proximate analyses to ascertain its fuel characteristics. Mainly, element component prediction models have been developed based on proximate analysis, yet few studies have predicted proximate components based on element analysis. Hence, this study developed a proximate component prediction model following the calorific value calculation. Analysis of Pearson’s correlation coefficient showed that volatile matter (VM) and fixed carbon (FC) were positively correlated with hydrogen and oxygen, and with carbon, respectively. Thus, the model correlation was developed using a combination of the “stepwise” and “enter” methods along with linear or nonlinear regressions. The optimal models were developed for VM and ash content (Ash). The VM optimal model values were: R2 = 0.9402, root-mean-square error (RMSE) = 7.0063, average absolute error (AAE) = 14.8170%, and average bias error (ABE) = −11.7862%. For Ash, the values were: R2 = 0.9249, RMSE = 2.9614, AAE = 168.9028%, and ABE = 167.2849%, and for FC, the values were: R2 = 9505, RMSE = 6.3214, AAE = 18.3199%, and ABE = 15.0094%. This study provides a model to predict the proximate component by element analysis. Contrary to existing method, proximate analysis can be predicted based on elemental analysis, and shows that consume samples can be performed at once.

The biomass higher heating value (HHV) is an important thermal property that determines the amount of recoverable energy from agriculture byproducts. Precise laboratory measurement or accurate prediction of the HHV is essential for designing biomass conversion equipment. The current study combines feature selection scenarios and machine learning tools to establish a general model for estimating biomass HHV. Multiple linear regression and Pearson’s correlation coefficients justified that volatile matter, nitrogen, and oxygen content of biomass samples have a slight effect on the HHV and it is better to ignore them during the HHV modeling. Then, the prediction performance of random forest, multilayer and cascade feedforward neural networks, group method of data handling, and least-squares support vector regressor are compared to determine the intelligent estimator with the highest accuracy toward biomass HHV prediction. The ranking test shows that the multilayer perceptron neural network better predicts the HHV of 532 biomass samples than the other intelligent models. This model presents the outstanding absolute average relative error of 2.75% and 3.12% and regression coefficients of 0.9500 and 0.9418 in the learning and testing stages. The model performance is also superior to a recurrent neural network which was recently developed in the literature using the same databank.

Many correlations are available in the literature to predict the higher heating value (HHV) of raw biomass using the proximate and ultimate analyses. Studies on biomass torrefaction are growing tremendously, which suggest that the fuel characteristics, such as HHV, proximate analysis and ultimate analysis, have changed significantly after torrefaction. Such changes may cause high estimation errors if the existing HHV correlations were to be used in predicting the HHV of torrefied biomass. No study has been carried out so far to verify this. Therefore, this study seeks answers to the question: “Can the existing correlations be used to determine the HHV of the torrefied biomass”? To answer this, the existing HHV predicting correlations were tested using torrefied biomass data points. Estimation errors were found to be significantly high for the existing HHV correlations, and thus, they are not suitable for predicting the HHV of the torrefied biomass. New correlations were then developed using data points of torrefied biomass. The ranges of reported data for HHV, volatile matter (VM), fixed carbon (FC), ash (ASH), carbon (C), hydrogen (H) and oxygen (O) contents were 14.90 MJ/kg–33.30 MJ/kg, 13.30%–88.57%, 11.25%–82.74%, 0.08%–47.62%, 35.08%–86.28%, 0.53%–7.46% and 4.31%–44.70%, respectively. Correlations with the minimum mean absolute errors and having all components of proximate and ultimate analyses were selected for future use. The selected new correlations have a good accuracy of prediction when they are validated using another set of data (26 samples). Thus, these new and more accurate correlations can be useful in modeling different thermochemical processes, including combustion, pyrolysis and gasification processes of torrefied biomass.

In this work, fuel characteristics such as proximate analysis, ultimate analysis, higher heating value, and ash composition of shells from four oilseed plants (hazelnut, walnut, almond, and sunflower) and olive husk were determined by analytical methods. Fuel properties for the combustion analysis of biomass can be conveniently grouped into physical, chemical, thermal, and mineral properties. The carbon content of selected biomass species varies from about 47.4% to 53.5% due to varying lignin and extractives content. The hydrogen content of the species varies from 5.8% to 6.7%. Oxygen content ranges from 35.5% to 41.6%, S is <0.1% and N ranges from 0.5- 1.6%. Lignin content ranges from 17.0% to 52.3%. The higher heating values (HHVs) of selected samples were correlated with their lignin contents. There was a highly significant linear correlation between the HHVs of the samples and their lignin contents. The HHV (MJ kg -1 ) of the samples as a function of lignin content (L, wt%) was calculated using the following equation: HHV = 0.0889(L) + 16.8218; for which the correlation coefficient (r) was 0.9504. The HHVs calculated with this equation showed a mean difference of 0.056%.

Designing efficient biomass energy systems requires a thorough understanding of the physicochemical, thermodynamic, and physical properties of biomass. One crucial parameter in assessing biomass energy potential is the higher heating value (HHV), which quantifies its energy content. Conventionally, HHV is determined through bomb calorimetry, but this method is limited by factors such as time, accessibility, and cost. To overcome these limitations, researchers have proposed a diverse range of empirical correlations and machine-learning approaches to predict the HHV of biomass based on proximate and ultimate analysis results. The novelty of this research is to explore the universal applicability of the developed empirical correlations for predicting the Higher Heating Value (HHV) of biomass. To identify the best empirical correlations, nearly 400 different biomass feedstocks were comprehensively tested with 45 different empirical correlations developed to use ultimate analysis (21 different empirical correlations), proximate analysis (16 different empirical correlations) and combined ultimate-proximate analysis (8 different empirical correlations) data of these biomass feedstocks. A quantitative and statistical analysis was conducted to assess the performance of these empirical correlations and their applicability to diverse biomass types. The results demonstrated that the empirical correlations utilizing ultimate analysis data provided more accurate predictions of HHV compared to those based on proximate analysis or combined data. Two specific empirical correlations including coefficients for each element (C, H, N) and their interactions (C*H) demonstrate the best HHV prediction with the lowest MAE (~0.49), RMSE (~0.64), and MAPE (~2.70%). Furthermore, some other empirical correlations with carbon content being the major determinant also provide good HHV prediction from a statistical point of view; MAE (~0.5–0.8), RMSE (~0.6–0.9), and MAPE (~2.8–3.8%).

Higher heating value (HHV) is an important property defining the energy content and thereby efficiency of fuels. In this paper, the correlation between the proximate analysis/ultimate analysis of some rice husk and HHV was discussed. It was found that the correlation between HHV and proximate, ultimate analysis was nonlinear. Therefore, two models were developed with artificial neural networks to predict HHV of rice husk by its proximate and ultimate analysis. A total of 25 samples of rice husk selected from the literature and experiments were used as the training data to build and train the two nets. Then, several samples selected randomly were used as predicting samples to check the accuracy of the two nets, respectively. A higher precision of 1.8 % relative error in the prediction results was obtained through this method, while the relative error of linear empirical equations given in the literature was more than 12.7 %. By this method, HHV can be estimated directly from the proximate analysis and ultimate analysis of rice husk when the HHV measurement equipment was not available.

Energy contents analysis of rice-husk derived from indigenous rice paddy species as sustainable solid biofuel

The clinical and hemodynamic factors that influence the concentrations of biomarkers of myocyte injury measured by high sensitive assay PATHFAST

In recent decades, the generation of Municipal Solid Waste (MSW) is steadily increasing due to urbanization and technological advancement. The collection and disposal of municipal solid waste cause considerable environmental degradation, making MSW management a global priority. Waste-to-energy (WTE) using thermochemical process has been identified as the key solution in this area. After evaluating many automated Higher Heating Value (HHV) prediction approaches, an Optimal Deep Learning-based HHV Prediction (ODL-HHVP) model for MSW management has been developed. The objective of the ODL-HHVP model is to forecast the HHV of municipal solid waste, based on its oxygen, water, hydrogen, carbon, nitrogen, sulphur and ash constituents. In addition, the ODL-HHVP model contains a Deep Support Vector Machine (DSVM) regression component that can accurately predict the HHV. In addition, the Beetle Swarm Optimization (BSO) method is utilised as a hyperparameter optimizer in conjunction with the DSVM model, resulting in the highest HHV prediction accuracy. A comprehensive simulation study is conducted to validate the performance of the ODL-HHVP method. The Multiple Linear Regression (MLR), Genetic Programming (GP), Resilient backpropagation (RP), Levenberg Marquardt (LM) and DSVM approaches have attained an ineffective result with RMSEs of 4.360, 2.870, 3.590, 3.100 and 3.050, respectively. The experimental findings demonstrate that the ODL-HHVP technique outperforms existing state-of-art technologies in a variety of respects.

Prediction model of higher heating value of torrefied biomass based on the kinetics of biomass decomposition

Laboratory Assessment to Correlate Strength Parameter from Physical Properties of Subgrade

Diarrhea is a frequent and potentially severe complication in solid organ transplant (SOT) recipients. One of the most common infectious etiologies of diarrhea in these patients is Clostridium difficile. Our objective was to investigate the association of C. difficile infection (CDI) with the outcomes of hospitalized SOT patients. We extracted all adult cases with discharge diagnoses of SOT or CDI from the United States Nationwide Inpatient Sample, Healthcare Cost and Utilization Project, Agency for Healthcare Research and Quality 2009 database. We collected outcome variables (mortality, length of hospital stay [LOS], hospitalization charges, complications of the transplanted organ, and colectomy), demographic information, and comorbidity data for each of the cases. The data were evaluated using univariate and multiple variable regression analyses. We identified 49,198 cases with SOT of which 2.7% had CDI. Univariate comparisons of cases with SOT+CDI to those with SOT-only revealed significant differences in the evaluated outcomes including in-hospital mortality (7.4% vs. 2.4%, P<0.001), LOS (median 9days vs. 4days, P<0.001), charges (median $53,808 vs. $31,488, P<0.001), organ complications (38.1% vs. 33.9%, P<0.001), and colectomy (1.1% vs. 0.3%, P<0.001). Using multiple variable regression analyses, in the SOT cohort (SOT-only and SOT+CDI), CDI was independently associated with greater mortality (adjusted odds ratio [aOR] 2.48, 95% confidence interval [CI]=2.22, 2.76, P<0.001), longer LOS (difference 9.6days, 95% CI=9.3, 9.9, P<0.001), higher charges (difference $69,647, 95% CI=$66,190, $73,104, P<0.001), more complications of the transplanted organ (aOR 1.36, 95% CI=1.28, 1.44, P<0.001), and increased need for colectomy (aOR 3.10, 95% CI=2.35, 4.08, P<0.001). Our results demonstrate that CDI is associated with overall significantly worse outcomes in hospitalized patients with SOT.

Review of higher heating value of municipal solid waste based on analysis and smart modelling