Abstract

Many attempts have been made to estimate calorific value of bagasse using mathematical equations, which were created based on data from proximate, ultimate, physical and chemical analysis. Questions have been raised on the applicability of these equations in different parts of the globe. This study was initiated to tackle these problems and also check the most suited mathematical models for the Law Heating Value of Cameroonian bagasse. Data and bagasse samples were collected at the Cameroonian sugarcane factory. The effects of cane variety, age of harvesting, source, moisture content, and sucrose on the LHV of Cameroon bagasse have been tested. It was shown that humidity does not change within a variety, but changes from the dry season to the rainy season; the sugar in the rainy season is significantly different from that collected in the dry season. Samples of the same variety have identical LHV. LHV in the dry season is significantly different from LHV in the rainy season. According to the fact that this study was done for cane with different ages of harvesting, the maturity of Cameroonian sugarcane does not affect LHV of bagasse. Tree selected models are much superior tool for the prediction of the LHV for bagasse in Cameroon compared to others. The standard deviation of these validated models is around 200 kJ/kg compared to the experimental. Thus, the models determined in foreign countries, are not necessarily applicable in predicting the LHV of bagasse in other countries with the same accuracy as that in their native country. There was linear relationship between humidity, ash and sugar content in the bagasse. It is possible to build models based on data from physical composition of bagasse using regression analysis.

Keywords

Sugarcane Bagasse, Relative Humidity, Ash, Low Heating Value

Share and Cite:

1. Introduction

To address the energy crisis that began in 1970, special attention has been directed towards alternative fuels to overcome the global energy problem. This shift has been prompted by the rising costs of fuel oil, natural gas, and electricity. Bagasse, which was previously burnt as waste, is now being utilized for combustion in electricity generation, thus becoming a viable boiler fuel. Given the necessity to consider combustion efficiency when using it as boiler fuel, bagasse has emerged as an opportunity to serve as an energy reserve to address deficits in electricity supply ([1] [2] [3]).

Currently, Cameroon owns two (02) sugar factories, which combustion of bagasse is for co-generation steam and electricity just for the sugar process. This situation laid to the accumulation of enough quantity of bagasse around the factories (41.165 tons/year for one factory, Nkoteng), which causes many environmental problems such the accident of bagasse fired ([4] [5]). The electricity deficit in the rural region where the factory is located is less than 5%. Developing the possibility to burn total fuel in the aim to export excess of electricity to the national grid, could make socio-economic benefits and favorable environmental impact for the factory.

In recent years, studies have been carried out over the world to better understand the combustion of this fibrous residue and its calorific impact on boilers efficiency in order to obtain steam. The problem of flame stability in boilers, highlighted by several of these works, is closely linked to the physic, chemical and thermal properties of bagasse ([6]-[23]). These properties have to be studied before the study of combustion process. The previous work in the Cameroon sugar factory has shown that the increasing of water moisture content in bagasse in the raining season causes down time according to the degradation of steam quality and quantity in the rainy season [10]. A fuel having a moisture content of 50% and an ash content of 2% will have an inert to combustible material ratio of 52/48 = 1.08. If the moisture is 50% and the ash content is 5%, the ratio is 55/45 = 1.22. This ratio is important when evaluating acceptable grate heat release rates and the amount of excess air required for complete combustion [24]. Complete combustion which will help to avoid factory down time. Industrial solid fuel boilers were designed for mining coal before being modified and subsequently adapted to other fuels such as biomass. Compared to coal, sugar cane bagasse, like other biomasses, has high volatile matter and moisture content ([25]-[37]). According to the fact that moisture, ash content, and inert ratio could increase during the raining season, it is necessary to make particular research. The increasing of moisture during the raining season which causes down time is due to the fact that a lot of mud usually adheres to the harvested cane and consequently increases the inert ratio in fuel bagasse, which influences the heating value.

Heating value is also calling calorific value depending of the authors. There are two different heating values: low heating value (LHV) and High heating value (HHV). Hugo [7] empirical analysis shows that, bagasse HHV in the dry basis base is approximately HHV = 19,256 kJ/kg and varies only about two percent (2%) from different countries and region in the world. The LHV is the HHV influencing by the percentage of net hydrogen in bagasse which varies from countries to countries. These percentages of net hydrogen represent also the latent heat of the water formed by the combustion process. LHV gives a more accurate indication of heat practically obtainable. Bagasse as a fuel is usually burn in wet basis, and it’s characterized by fiber, brix, ash and moisture [7]. All these parameters could vary per countries and regions in the word. There are different methods used in the literature to analyzed bagasse: physical, proximate and ultimate analyses. From theses method and parameter obtained, we find a lot of models in the literature used to predict heating value of bagasse in the world ([38] [39] [40] [41] [42]). The heating value of bagasse can also be determined experimentally (direct) by a bomb calorimeter, though this method is reliable, it is not feasible due to the equipment use and consumable cost. Then models are the most method use in the literature and could help to study the effect of extrinsic parameters of the region on these Calorific Values ([43] [44] [45] [46]). As a limitation of these Mathematical models created, performs best in the country/locality in which it is created, while producing over or under prediction when used internationally.

This work describes Cameroonian bagasse characterization, relevant models were selected in the literature used in others countries by authors to determine the most suitable one which could be used to predict the Cameroonian sugar cane bagasse LHV.

2. Material and Methods

2.1. Location

The SOSUCAM sugar complex is located approximately 129 km north of the city of Yaoundé, along national road No.1 connecting Yaoundé to Nanga-Eboko in the Mbandjock-Nkoteng region in the South of Cameroon, between the following geographic coordinates: 11˚51' - 12˚10' East longitude, 4˚20' - 4˚35' latitude North.

Soil coverage of the Mbandjock-Nkoteng region (75% of the surface area), belongs to the domain of ferrallitic soils desaturated soils of the southern Cameroonian plateau alongside soils with little evolved and raw minerals (5%) on residual massifs and soils hydromorphs (20%) which cover the shallows (Figure 1). The climate is subequatorial with four seasons: two dry seasons (from mid-November to mid-March and from July to August) and two rainy seasons (from September to mid-November and from mid-March to July). The temperature is moderate and most often oscillates between 23.5˚C to 26.5˚C and therefore the annual average is 24.7˚C. While the sunshine is 5h20' per day and evapotranspiration of 3.8 mm per day are relatively low yet, the hygrometry remains high (morning value greater than 90%, the rest of the day the humidity does not falling below 70% than during the long dry season). The harvest being manual, it is preceded by spreading fire which eliminates all the leaves on the harvested cane.

Figure 1. Localization of the Cameroon Sugar Company (SOSUCAM); A—Cameroon in Africa; B— Southern Cameroon; C—The SOSUCAM site [47].

2.2. Sampling Design

Sample from Table 1

From the literature, Heating Value of sugar cane bagasse can be predicted by models based on fiber, sugar, ash and moisture content in bagasse. From these parameters, only moisture don’t have heating value (Table 2), but it being absorbs heat in vaporized during the combustion.

Assuming that each of these parameters could be influence by cane varieties, season, age of cane, and types of soil when harvested; data were collected of historical information of the factory (Table 1). These data were collected over two years to make sure that these parameters could be taken during rainy and dry season. Sample from (Table 1) where collected according to Hugo mathematical model of heating value which depend on the proximate parameter (moisture, sugar and inert material). Means, in this table heating value were calculated by the mathematic model given by Equation (4) (Table 3) [7]. The effect of season, variety of can, and soil on humidity, inert ratio and heating value will be study.

Sample from Table 4

Based on the hypothesis that impact of cane variety, ages when harvested, and types of soil on the calorific value of bagasse variation is not significant. The bagasse samples were taken directly from the inlet of the burner taking into account their variations in humidity level with the quality of steam produced. In the plant, when the steam quality is poor, operators often use the mixture of bagasse from storage (dry) (Figure 2(b)) with bagasse coming directly from the mills (Figure 2(a)).

Table 1. Experimental data of bagasse caracteristics acoording to variety and parcel types.

Table 2. Heating value of constituents of bagasse [7].

Table 3. Summary of models used for predicting law heating value of bagasse.

ω: moisture content in bagasse, k: ash content, s: sugar content.

Table 4. Sample collection for analysis.

Figure 2. Bagasse of sugar factory: excess from milling (a) and storage (b).

2.3. Analytical Methods

Effect of humidity, season, sugar Sample from Table 2

The relatively unknown effects of many different parameters on the LHV of Cameroonian bagasse will be tested in this study by experimentation under controlled conditions. The parameters on which the LHV depends and the LHV itself were divided into various groups which would isolate one or two controlled variables and a factorial analysis of variance (F ratio test) was applied to the samples within the groups. This is the case of the test of the effect of cane variety, season and the ages when harvested on the sugar content, humidity and LHV.

Bagasse characterization: Sample from Table 4

Bomb calorimeter (1520 kPa of oxygen pressure, 6 V circuit voltage, ca. 0.7 g of bagasse sample)

The experimental calculation of the Calorific value (LHV and HHV) was made according to Standard NF MO3-005/EN 14918/ISO 1928. So, the sample was first ground using the cutter mill. Then sieved through a sieve whose mesh opening is between 0.2 mm and 1mm. Before starting the analysis, the sample was mixed thoroughly for approximately one minute using a spatula. The calorimeter was conditioned by obtain a temperature difference between the thermometer of the adiabatic chamber and the calorimetric vessel thermometer less than or equal to 1˚C.

Effect of moisture, ash and sugar on the LHV of bagasse

Each bagasse corresponding sample from bomb calorimeter was analyze to determine moisture, ash and sugar content. These values were used to calculate the LHV of bagasse with the selected literature models. Results from calculations using formulas (Table 3) and from bomb calorimeter were statistically analyzed with multiple variable analysis method. To study the impact of inert ratio which characterized the region of culture (sludge and sand) [37], effective moisture of bagasse was evaluated and his impact on the calorific value study. in fact, if a lot of mud sticks to the bagasse during harvesting, it will not be crushed well and will arrive at the burner with a lot of mud and humidity, thus increasing the rate of inert matter. This parameter depends on the type of soil and the harvesting technique (describe in Section 2.1).

3. Results and Discussion

3.1. Effect of Variety, Season, Moisture Age and Region of Cane Harvested on the LHV

Table 5 shows the summary statistical test of the effect of season and variety change on the humidity content, sugar content and LHV. From these first three analyses, it is emerging that the constituent elements (humidity and sugar content) of the model used to calculate the LHV depend on the season and the variety, the LHV depend also on the season and the variety. The results obtained also showed that humidity does not change within a variety. It would therefore be interesting or even essential for an application data of bagasse to indicate the harvest season and the variety used.

Table 6 presents the Multiple Range Testing for Moisture Content by Variety. Five Homogeneous varieties are identified using columns of X. In each column, the levels containing X form a group of means within which there are no statistically significant differences. The results obtained showed that humidity does not change within a variety. The samples of the varieties (Co997, B46364, FR81258, Cr87339 and B82333) have identical humidity. This can also mean that the types of soil do not really have an influence because these samples of the same variety come from different types of soil.

Table 7 is the result of the multiple comparisons procedure to determine the means which are significantly different from each other, are there varieties which have similar humidity. The upper part of this table displays the estimated differences between pairs of means of two varieties. A star (*) was placed next to 12 pairs, indicating that these pairs have statistically significant differences at the 95.0% confidence level. We therefore see that the B46364 variety has the same moisture content as the FR81258 variety. Similar conclusion for B43664 and R585, B43664 and R570, Co997 and R570, Co997 and R585, Cr87339 and R570, FR81258 and R570, FR81258 and R585, R570 and R585.

In Table 8, two homogeneous groups are identified using columns of X. In each column, the levels containing X form a group of means within which there are no statistically significant differences. The conclusion from these last two tables is that the varieties in the dry season have the same humidity. In the rainy season, we also have the same observation. Then, the humidity of different bagasse in the dry season is different from the humidity in the rainy season.

Table 9, Table 10 and Table 11 gives the averages of the sugar content for each level of the factors (season and variety). The question asked here is whether the sugar content of samples collected in the dry season is different from that obtained in the rainy season.

Table 5. Summary of statistical tests of the effect of season and variety change on the sugar and humidity content in bagasse, and the LHV.

Table 6. Multiple range testing for humidity content by variety.

Table 7. Multiple comparison between cane varieties and humidity level.

*Denote a statistically significant difference.

Table 8. Multiple range testing for humidity content by variety.

Table 9. Multiple range testing for sugar by cane variety.

Table 10. Multiple comparison between sugar content of the cane varieties.

Table 11. Multiple range testing for sugar content by season.

To compare the average sugar content of different varieties, the multiple comparison procedure was used. Table 9 shows the multiple comparison of sugar content of different varieties of sugarcane. The upper part of this table displays the estimated differences between pairs of means. A star was placed next to 2 pairs (B46364 - B82333 and B46364 - Cr87339), indicating that these pairs have statistically significant differences at the 95.0% confidence level. With the exception of these, the sugar contents are significantly equivalent.

Table 11 presents the multiple range tests of sugar cane content by season of harvest. The upper part of this table displays the estimated differences between pairs of means. A star has been placed next to one pair, indicating that this pair has a statistically significant difference at the 95.0% confidence level. At the top of the table, 2 homogeneous groups are identified using columns of X. In each column, the levels containing X form a group of means within which there are no statistically significant differences. Through this, we see that the sugar content of the samples of sugar cane varieties collected in the rainy season is not significantly different. Similar conclusion in the dry season. On the other hand, the sugar content of samples collected in the rainy season is significantly different from that of samples collected in the dry season.

Table 12, Table 13 and Table 14 give the averages of the LHV for each level of factors (season and variety). At the top of this table, 4 homogeneous groups are identified using columns of X. In each column, the levels containing X form a group of means within which there are no statistically significant differences, means that samples of the same variety have identical LHV.

From Table 12 and Table 13, part of this table displays the estimated differences between pairs of means. A star was placed next to 11 pairs, indicating that these pairs have statistically significant differences at the 95.0% confidence level. In other words, 11 pairs of varieties (B46364 - B82333, B46364 - Cr87339, B82333 - Co997, B82333 - Cr87339, B82333 - FR81258, B82333 - R570, Co997 - Cr87339, Co997 - FR81258, Cr87339 - FR81258, Cr87339 - R570 and FR81258 - R570) have significantly different LHV.

Table 14 presents the multiple range test for the comparison of LHV collected in the rainy season and the dry season. A star has been placed next to a pair, indicating that this pair has a statistically significant difference at the 95.0% confidence level. The LHV of samples collected in the dry season is significantly different from that obtained in the rainy season.

Table 12. Multiple range testing LHV content by variety.

Table 13. Multiple comparisons between LHV of the cane varieties.

Table 14. Multiple range testing for LHV by season.

The results obtained showed that humidity does not change within a variety, the humidity of different bagasse in the dry season is different from the humidity in the rainy season, the sugar content of samples collected in the rainy season is significantly different from that of samples collected in the dry season. Samples of the same variety have identical LHV. The LHV of samples collected in the dry season is significantly different from that obtained in the rainy season. According to the fact that this study was done for cane with different age of harvesting, the maturity of Cameroonian sugarcane does not affect LHV of bagasse.

3.2. Validity of Selected Formulas Based on Sample Analysis

Table 15 shows the minimum, maximum, average, and standard deviation (SD) of the ash contents, sugar and humidity contents, and the heating values of the normal milling operation bagasse. The ash contents of the bagasse samples covered a wide range, and the humidity contents varied in a moderate range.

A P-value statistical test was done to check the relationship between ash contents and heating value, sugar contents and heating value, and humidity contents and heating value.

Table 16 shows a P-value which tests the statistical significance of the estimated correlations. P-values below 0.05 indicate statistically significant non-zero correlations at the 95.0% confidence level. The following pairs of variables have P-values below 0.05: Ash (%) and sugar (%), Ash (%) and humidity (%), Ash (%) and LHV (kJ/kg), Ash (%) and HHV (kJ/kg); sugar (%) and humidity (%), sugar (%) and LHV (kJ/kg), sugar (%) and HHV (kJ/kg), humidity (%) and. LHV (kJ/kg), humidity (%) and HHV (kJ/kg), LHV (kJ/kg) and HHV (kJ/kg). It is means that the variability observed in HV was explained by moisture, ash and sugar.

Table 17 presents the summary statistics of LHV obtained experimentally compared to the values obtained by calculation using different models presented from the literature (Table 3). The statistical evaluations are shown namely the average, standard deviation and the range between the maximum and minimum error that occurred, which will indicate the performance of the model.

Table 17 confirms that Equation (1), Equation (2) and Equation (3) are better than Equation (4) due to a smaller standard deviation. Thus Equation (1), Equation (2) and Equation (3) are much superior tool for the prediction of the LHV values for bagasse in Cameroon. The standard deviation is around 200 kJ/kg comparing experimental values to those gives by the models. Thus, the models determined in foreign countries, such as that by the Conventional Equations above, are not necessarily applicable in predicting the LHV value of Bagasse in other countries with the same accuracy as that in their native country.

Figure 3 shows the average with the maximum and minimum LHV value, which ably shows the divergence of the averages, maximum, and minimum values from the experimental data. It’s strengthening the arguments that Equation (1), Equation (2) and Equation (3) gave good prediction as compared to Equation (4).

Table 15. Characteristics of bagasse from analysis.

Table 16. The statistical tests significance of the estimated correlations.

Table 17. Summary Statistics of the LHV values of final mill bagasse from experience compare to the calculation value.

Figure 3. Average, maximum and minimum values of LHV for models and experiment.

Lastly, this finding also proves that the best suited models depend not only on the humidity and sugar content but also on the ash content. Figure 4 shows the LHV predicted by Equation (1), Equation (2) and Equation (3) decreases with the increase of ash content in sugarcane bagasse.

The previous paragraph shows that the Calorific Value of sugar cane bagasse depends on the harvest season, and the moisture content of the bagasse obtained after crushing the cane. The joint influence of the plot and the harvest season would be linked to the sand content (not studied here) which could limit the action of the mill on the cane and therefore would favor a very sandy bagasse and with a more ash after combustion. To choose the study models, it becomes important to evaluate the influence of the rate of ash on the LHV.

Table 18 shows the ash content, effective humidity content and LHV of sugar cane bagasse analyzed. From this table inert rations increases with moisture content in our factory locality. Figure 5 shows the influence of the effective humidity (inert ratio) on the LHV. When the inert ration (effective humidity) increase, LHV decrease.

Figure 4. Comparison between LHV values estimated for literatures formulas.

Figure 5. Profile of LHV according to the effective humidity, compared to the profiles of [40] for the rate of ash at 0% and 10%.

Table 18. Ash content, moisture and effective humidity of sugar cane bagasse.

4. Conclusions

The objective of this work was to select the most efficient model to predict LHV of Cameroonian bagasse. We have recorded data related to the characteristics of the bagasse such as humidity rate, sugar content, different varieties of harvest plots, and a sample according to whether or not severe combustion conditions were sampled. Statements linked to characteristics of the fuel were analyzed to study the influence of the various parameters on the LHV.

It emerged from this analysis that: humidity does not change within a variety, the humidity of different bagasse in the dry season is different from the humidity in the rainy season, the sugar content of samples collected in the rainy season is significantly different from that collected in the dry season. Samples of the same variety have identical LHV. The LHV of samples collected in the dry season is significantly different from that obtained in the rainy season.

The link observed between the harvest season and the humidity rate is explained by the existence of sand on the cane during the harvest, which should reduce the work of the mills and promote the production of very humid bagasse. Consequently, the mixture of sands in the bagasse of sugarcane increased the ash content.

Equation (1), Equation (2) and Equation (3) are better than Equation (4) when using to predict the Cameroonian LHV due to a smaller standard deviation. Thus Equation (1), Equation (2) and Equation (3) are much superior tool for the prediction of the LHV values for bagasse in Cameroon. The standard deviation is around 200 kJ/kg comparing experimental values to those given by the models. Thus, the models determined in foreign countries, such as that by the Conventional Equations above, are not necessarily applicable in predicting the LHV value of bagasse in other countries with the same accuracy as that in their native country.

When the inert ration (effective humidity) and ash increase, LHV decrease thus the effect of ash in the LHV is not negligible as by those models in literature. Then, there is linear relationship between humidity, ash and sugar content in the bagasse. It is possible to build models based on data from physical composition of bagasse using regression analysis. It is therefore necessary to ameliorate the clean method of the sugarcane before processing it, or use mechanical harvesters instead of manual harvesting.

Data Availability

The article contains all the relevant data. These data, used to support the findings of this study were obtained from the Cameroonian Sugar Society, Nkoteng sugar Factory.

Acknowledgments

Authors acknowledge to Cameroon Sugar Society through Nkoteng Sugar factory for providing facilities.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

References