Moisture Content Effects on Coal–RDF Co-Firing Performance in a Bubbling Fluidized Bed Incinerator

2.1 Materials and sample preparation

Two solid fuels were used in this study: RDF produced from MSW and coal, the latter serving as the primary fuel. The RDF feedstock was obtained from the TPST Kertalangu integrated solid-waste processing facility, which was selected because it is representative of the urban MSW generated in a tourism-driven region, where the waste stream is highly heterogeneous and rich in organic and plastic fractions. At the facility, the combustible fraction of the MSW – mainly paper, wood and plastic residues – was separated from non-combustible materials such as metals, glass and inert debris.

The RDF preparation comprised several mechanical stages, illustrated in Figure 1. The raw MSW (Figure 1(a)) was first manually pre-sorted to remove bulky and clearly non-combustible items. The waste was then size reduced in a mechanical shredder (Figure 1(b)) to obtain smaller, more uniform particles and subsequently screened (Figure 1(c)) to remove oversized and inert fractions and improve fuel homogeneity. The resulting RDF (Figure 1(d)) consisted predominantly of combustible plastic, paper and wood residues.

tu_pian_1.png

The RDF was prepared at two target MC, 25% and 40%, to compare their influence on combustion performance. As shown in Figure 2, sample preparation was conducted using a laboratory oven and a digital balance to determine and adjust the RDF MC gravimetrically. Representative samples were initially weighed using the digital balance and subsequently dried in the oven at 105 ± 5 ℃ until a constant mass was achieved. The MC was calculated using Eq. (1), and the samples were then adjusted to the specified target moisture conditions.

$M C=\frac{m_i-m_d}{m_i} \times 100 \%$               (1)

where, MC is the moisture content of the RDF (%), m_iis the initial (as-received) mass of the sample (g), and m_d is the oven-dried mass (g). The wet mass required to reset each sample to its target MC was then obtained from Eq. (2):

$m_t=\frac{m_d}{1-M C_t}$           (2)

where, m_t is the target mass of RDF at the desired moisture content (g), and MC_t is the target moisture content expressed as a fraction (0.25 or 0.40). Immediately before combustion, the prepared RDF was blended with coal at 10 wt.% RDF and 90 wt.% coal on an as-received (wet) basis, giving a total wet fuel charge of 100 g. Consequently, the two charges did not contain identical dry RDF masses.

Quartz sand was used as the inert bed material, with the same characteristics as in the authors’ previous work on a dual reactor fluidized bed [17]. The sand consisted of 99.4%–99.9% SiO₂ and had a particle density of 2.18 g/cm³, a specific heat of 0.20 cal/g·℃ and a particle size of 0.4–0.5 mm. Its high melting point of about 1800 ℃ makes it well suited to fluidized bed applications. The preparation of the bed material is shown in Figure 3. The proximate composition and calorific value of the prepared fuels are reported in Sections 3.1 and 3.2, respectively.

tu_pian_2.png

Figure 2. Moisture content (MC) determination of the refuse-derived fuel (RDF) samples by the gravimetric oven-drying method, (a) drying of the samples in a laboratory oven and (b) measurement of the sample mass with a digital balance

tu_pian_3.png

Figure 3. Preparation of the quartz sand bed material (adapted from Winaya et al. [17]), (a) raw silica sand, (b) sieving to obtain a uniform particle size, and (c) weighing prior to loading into the reactor

2.2 Proximate and thermogravimetric analysis

Proximate analysis of the RDF samples was performed using a LECO TGA 701 thermogravimetric analyzer, as shown in Figure 4, following ASTM D7582 [18] and the standard procedure applied by the testing laboratory. The nominal MC of 25% and 40% represents the target conditions established during sample preparation, whereas the corresponding measured MC obtained from the analysis was 23.65% and 43.16%, respectively. The programmed analysis comprised three stages. For moisture determination, the furnace temperature was increased from 25 to 107 ℃ at 6 ℃/min under nitrogen and held for 15 min or until constant mass, using the instrument's high-flow setting (10.0 L/min). For volatile-matter determination, the temperature increased from 107 to 950 ℃ at 43 ℃/min under nitrogen and was held for 7 min at the same high-flow setting. For ash determination, the programmed setpoint was changed from 950 to 750 ℃ at 3 ℃/min under oxygen with a stated purity of 98%, using the low-flow setting (3.5 L/min), and the analysis continued until constant mass without a fixed holding period. Moisture, volatile matter, and ash were determined from successive mass changes, whereas fixed carbon was calculated by difference. All proximate-analysis values were reported on an as-received (wet) basis. Each moisture condition was analyzed once (n = 1). Therefore, the reported values represent individual measurements and are presented without standard deviations or inferential statistical analysis.

tu_pian_4.png

2.3 Reactor configuration and experimental procedure

The experiments were conducted in a laboratory-scale BFB incinerator fabricated from stainless steel SS304 and operated through direct co-firing of 10 wt.% RDF and 90 wt.% coal. The cylindrical reactor had a total height of 700 mm and a diameter of 55 mm. Quartz sand containing 99.4%–99.9% SiO₂, with a particle-size range of 0.4–0.5 mm, was used as the inert bed material. The initial static bed height was maintained at 150 mm. Fluidizing air was supplied through an 8.5-mm-diameter bottom inlet pipe at an inlet velocity of approximately 9 m/s. Before fuel feeding, the bed was heated to an operating temperature of approximately 800 ℃. The axial temperature distribution was measured using three thermocouples positioned at T1 in the lower bed, T2 in the splash zone, and T3 in the freeboard. The dimensional configuration and temperature-measurement locations are presented in Figure 5.

tu_pian_5.png

Figure 5. Schematic of the laboratory-scale bubbling fluidized bed (BFB) incinerator

The main components of the experimental system are shown in Figure 5. The initial bed height was set at 150 mm in the lower section of the cylindrical reactor. Three axial side ports were used for the T1, T2, and T3 temperature measurements. The thermocouples (1) were connected to a control panel (2), which was used to monitor the combustion chamber temperature.  Fluidizing air (3) entered through a line located beneath the reactor to drive the fluidization process. The flue gas left the reactor through a flue-gas nozzle (4), passed through a filter (5), and was then cooled by a cooler (6) before entering the gas analyzer (7), where it was analyzed. Fuel was introduced into the reactor for combustion through a fuel feeder (8). Before fuel feeding, the reactor was preheated by a heater (9) until the operating temperature of 800 ℃ was reached. A distributor plate (10) supported the silica sand bed and ensured uniform air distribution and fluidization. The fluidizing air was supplied by an air-supply blower (11) through the bottom inlet, allowing the bed to reach the bubbling fluidization regime during combustion.

One combustion run was conducted for each target moisture condition (n = 1). The combustion gases were analyzed with a Gasboard-3100P gas analyzer, which recorded the lower heating value (LHV) together with the CO, CO₂, and CH₄ concentrations. At the selected reference temperature of 0 ℃ and with the result expressed in MJ/m³, the LHV was calculated using the factory coefficients given in the instrument manual [19]: LHVgas = 12.620 yCO + 35.818 yCH₄ + 91.180 yC_nH_m + 10.777 yH₂, where each y is the measured gas volume fraction expressed as a decimal. The principal operating conditions are summarized in Table 1.

Table 1. Operating conditions of the bubbling fluidized bed (BFB) incinerator

2.4 Net plant heat rate calculation

The NPHR was calculated from the ratio of the fuel-energy input rate to the useful thermal power released during combustion. Following the definition widely adopted for coal-fired thermal systems [20, 21], it is expressed as:

$N P H R=\frac{m_{\dot{fuel}} \cdot L H V_{\text {in}}}{P_{\text {th}}}$           (3)

where, $\dot{m}_{\text {fuel}}$ is the fuel consumption rate $\left(\mathrm{kg} \cdot \mathrm{h}^{-1}\right), L H V_{\text {in}}$ is the lower heating value of the incoming fuel $\left(\mathrm{kcal} \cdot \mathrm{kg}^{-1}\right)$, and $P_{\mathrm{th}}$ is the useful thermal power output (kW_th) derived from the net thermal energy released during combustion, defined as the difference between the chemical energy supplied by the fuel and the residual chemical energy carried by the outlet gas.

$P_{\text {th}}=\frac{Q_{\text {net}}}{t_{\text {comb}}}$           (4)

$Q_{\text {net}}=\dot{m}_{\text {fuel}} \cdot L H V_{\text {fuel}}-V_{\text {gas}} \cdot L H V_{\text {gas}}$      (5)

where, $V_{\text {gas}}$ is the total gas volume discharged during combustion $\left(\mathrm{m}^3\right)$, $L H V_{\text {gas}}$ is the lower heating value of the outlet gas $\left(\mathrm{kcal} . \mathrm{m}^{-3}\right)$, and $t_{\text {comb}}$ is the combustion duration. Appropriate unit conversion was applied to express $P_{\text {th}}$ in $\mathrm{kW}_{\text {th}}$. Accordingly, the NPHR reported in this study represents a thermal-equivalent heat rate based on combustion and outlet-gas parameters rather than a conventional plant-level NPHR based on measured net electrical power.

2.5 Combustion efficiency

Combustion efficiency is defined, following the energy-balance approach used for fluidized bed combustors [20], as the ratio of the heat energy effectively utilized during combustion to the total heat energy available in the fuel supplied to the combustion chamber:

$\eta=\frac{Q_{\text {in}}-Q_{\text {out}}}{Q_{\text {in}}} \times 100 \%$         (6)

where, Q_in is the total heat energy supplied (MJ), and Q_out is the heat energy lost (MJ). The result, expressed as a percentage, reflects the proportion of fuel energy converted into useful energy.

This study quantified the influence of RDF MC on the combustion performance of coal–RDF co-firing in a laboratory-scale BFB incinerator. Increasing the RDF target MC from 25% to 40% lowered the peak bed temperature and reduced the fuel GCV by 27.6%, increased the NPHR by 27.5% (from 926.1 to 1180.6 kcal/kWh_th) and decreased the combustion efficiency from 92.92% to 72.89%. The higher moisture level also promoted incomplete combustion, as evidenced by the elevated CO and CH₄ emissions and the 52.5% longer burnout time.

Taken together, these results identify MC as a critical parameter governing the thermal performance, reaction stability and energy efficiency of fluidized bed co-firing systems. The best performance was obtained at target moisture 25%, consistent with the upper limit recommended by SNI 8966:2021; maintaining the RDF moisture at or below this value is therefore essential for efficient operation. Future work should extend the investigation to a wider range of moisture levels and co-firing ratios and incorporate detailed emission and ash-behavior analyses to support the scale-up and design of full-scale co-firing systems.

I Made Agus Putrawan: Conceptualization, methodology, investigation, data curation, formal analysis, visualization, and writing-original draft. Samuel Alfred Manumpil: Investigation, data curation and validation. I Nyoman Suprapta Winaya: Conceptualization, methodology, supervision, project administration, funding acquisition, and Writing-review and editing. Made Suarda: Validation, supervision, and writing-review and editing. I Ketut Gede Wirawan: Methodology, validation, and resources. I Putu Angga Yuda Pratama: Investigation, data curation, and visualization. I Gusti Ngurah Putu Tenaya: Formal analysis, validation, and writing-review and editing.