Physicochemical and radiological characterization of flue gas desulfuration waste samples from Brazilian coal-fired power plants

.


INTRODUCTION
Coal-fired power plants are a major source of emissions of air pollutants including SO2, NOx, particulate matter, among others. The sulfur dioxide (SO2) combines with water and air forming sulfuric acid, which is the main component of acid rain.
Flue gas desulfurization (FGD) is the most used technology for reducing sulfur oxide emissions and remove up to 99% of the sulfur oxides in flue gas for a typical conventional coal-fired power plant [1].
FGD systems are generally classified as wet, semi-dry or dry processes. The wet FGD process is based on an acid-base reaction that takes place under oxidation conditions. Sulfur oxides in flue gas react with CaCO3 and MgCO3 to produce 90% gypsum (CaSO4.2H2O) and hannebachite (CaSO3·½H2O) in forced oxidation. FGD by-product of wet process is also also known as FGD gypsum [1][2][3].
In dry and semi-dry FGD processes, the sulfur oxide in the flue gas reacts with Ca(OH)2 (or CaO) and an excess of water to produce a slurry. This slurry is nebulized in a spray dry absorber, where SO2 is removed from the flue gas. The resulting by-product of this process is a dry mixture of hannebachite (CaSO3·½H2O) with minor amounts of gypsum [1].
Large volumes of FGD wastes from wet, dry or semi-dry FGD process are expected to be continuously produced with the increasing activity of coal-fired power plants.
FGD waste is discarded in landfills on the surrounding of the coal-fired thermal plant. For FGD gypsum, exists a market demand as a substitute of natural gypsum. However, dry or semi-dry FGD by-products are currently not widely used due to its high content of calcium sulfite. There have been some reports on the use as cement additives and in the manufacture of sulfoaluminate cement [4].
Presently, continuous research is ongoing on the use of FGD wastes for more advanced applications, such as zeolite synthesis [5,6]. FGD waste belong to complex composite materials with variable physical properties, particle morphology and chemical composition. FGD waste contain also trace quantities of naturally occurring radionuclides from the uranium and thorium series, as well as other naturally occurring radionuclides such as 40 K derived from the original coal matrix that tend to become enriched in their by-products. Thus, it is important to analyze their properties in order to address the potential reuse of FDG wastes [7][8][9]. The aim of this investigation is to compare the physicochemical characteristics of FGDs byproducts produced using wet and semi-dry process from three Brazilian thermal power plant. In addition, radiation characterization was performed to assess the possible radiological risks to human health due to the use of such materials.

Material
All chemicals that are used in this study were of analytical grade. Three types of samples of different origin were studied here. The waste produced from the wet flue gas desulfurization process was collected at President Medici coal-fired power plant, located in Candiota, Rio Grande do Sul, Brazil (FGD-C). The semi-dry samples were collected at Pecem coal-fired power plant (FGD-P) located in São Gonçalo do Amarante, Ceará, Brazil and Itaqui coal-fired power plant (FGD-I) located in São Luís, Maranhão, Brazil.

Characterization of materials
Samples were characterized in terms of semi-quantitative chemical composition by X-ray fluorescence (Rigaku -RIX 3000). A scanning electron microscope was used to verify the morphology of the samples (Philips XL 30). The mineralogical composition was determined by X-ray diffraction (Rigaku -Multiflex) using Cu Kα radiation at 40 kV and 20 mA. The particle size distribution for the samples was determined using a laser diffraction particle size analyzer (Malvern -Mastersizer 2000). Leaching and solublization experiments were performed following the Brazilian Association of Technical Standards Norm [10,11]. The concentration of elements was determined in the resulting leachate and solubilized extracts by inductively coupled plasma optical emission spectrometry using a Spectro ARCOS ICP OES (Kleve, NRW, Germany).
Neutron activation analysis was used to determine the concentrations of the elements U and Th.  the irradiation, two sets of measurement were done. The first, after a one week period of cooling, to determine U concentrations and, the second, after two weeks, to determine Th concentration. Samples were counted for a period of 5000 s, in the first and second measurement, by using an EG&G Ortec Ge high pure Gamma Spectrometer detector (AMETEK Inc., USA) and associated electronics, with a resolution of 0.88 and 1.90 keV for 57Co (122 keV) and 60Co (1332 keV), respectively [12].
Activity concentrations of U and Th were obtained by their specific activity using the conversion factor of 24.5 and 4.05, respectively.
Instrumental neutron activation analysis was also applied to determine trace elements. Ge Highpure Gamma Spectrometer detector (AMETEK Inc., USA) and associated electronics, with a resolution of 0.88 and 1.90 keV for 57 Co and 60 Co, respectively. The analysis of the data was done by using an in-house gamma ray software, VISPECT program, to identify the gamma each energy transition. Methodology validation was performed by analyzing the same reference materials cited above. The results presented are mean values (mg kg -1 ) for the duplicate and standard deviation for concentrations of elements. The relative standard deviations and relative errors were lower than 10% for most of the elements determined.
The concentration of the natural radionuclides 226 Ra, 228 Ra, 210 Pb and 40 K were carried out by non-destructive γ-ray spectrometry. Samples were packed in 50 cm 3 polypropylene cylindrical containers and they were kept sealed for at least 30 days in order to reach radioactive equilibrium between 226 Ra and 222 Rn progenies. A HPGe EG&G Ortec detector with 40% of relative efficiency and 2.09 keV resolution at 1.33 MeV and associated electronic devices were used, with live counting time of 80,000 s. The spectra were acquired by multichannel analyzer and, for the analysis, Genie for self-absorption according to the method described in [13]. The detector efficiency was calibrated by using the reference material IAEA-RGU-1, IAEA-RGTh-1 and IAEA-RGK-1.

Radiological Properties
The activity concentrations of naturally occurring radionuclides in different FGD samples is shown in Table 1. The data give an idea of the variability found in FGD samples coming from different sources. FDGD-I and FDG-P were generated in plants that use coal mainly from Colombia, while FGD-C is generated in a plant that uses coal from southern Brazil. In general, the activity concentrations of 40 K is higher in FGD samples of semi-dry process (FGD-P. FGD-I) than in FGD sample of wet process gypsum (FGD-C). Also, 210 Pb, 238 U and 232 Th showed the highest activity in FGD-P. Bq kg -1 for 40 K for fly ash [14].
There are few studies in the literature on the radionuclide activity concentrations in FGD samples.

Concentration of trace elements in FGD samples
During the coal combustion process, trace elements (e.g., As, Se and Hg) are decomposed by heat and travel with the flue gas. These trace elements may be captured during the desulfurization process and become part of FGD by-product. In addition, trace elements are captured by fly ash.
The elements of greater concern are As, Cd, Hg, Pb, Se and Zn, due to their potentially detrimental impact on human health and the environment.  In general, trace element concentrations were higher in the fly ash than in the FGD gypsum (FGD-C) collected from the same power plant [21].

Chemical composition
The chemical composition of FGD wastes and loss on ignition (LOI) obtained by X-ray fluorescence is given in Table 3. As can be seen from Table 3, the major constituents were Ca, Si, S, Al and Fe. The high magnesium content in FGD-C sample is justified because in the wet process Mg 2+ such as slaked lime (Ca(OH)2) or magnesium-containing slaked lime (Ca(OH)2 and Mg(OH)2) is used to improve the oxidation efficiency of coal combustion products, thus increasing the amount of gypsum produced. The percentage of elements is within the range usually encountered in flue gas desulfurization (FGD) by-product [20]. An important parameter relating to the unburned carbon content is LOI (loss of ignition). As can be seen from Table 3, the highest value of LOI has FGD-C. The large percentage of loss of ignition may be an advantage over some applications that require a high coal content. Figure 1 shows the differential and cumulative particle size distributions for FGD samples. The particle size distributions of material are given in Table 4. The differential size distribution of the material is relatively uniform with the expected normal bell-shaped distribution curve for FGD-C and FGD-I samples. The FDG-P sample showed a bimodal particle distribution curve.

Particle size distribution
These distributions specify that the majority of particles (90%) lie below 90 µm for semi-dry process (FGD-P and FGD-I) and below 194 µm for wet process (FGD-C). The particle size distribution for FGD-C samples is broader than for FGD-P and FGD-I.
Particle size is affected by the FGD process, and the ash size from the semi-dry process is lower than that from the wet process. In wet process, forced oxidation, which is a separate step after the    Figure 2-4 shows X-ray diffraction patterns (XRD) of the FGD samples, and the respective chemical compositions of crystalline phases are identified in Table 5-7.

Mineralogical composition
The FGD-C sample (Figure 2) presented an ettringite phase, which contains calcium and sulfate in its structure, and brucite phase, which are commonly found in residues from the process of wet desulfurization. Ettringite is formed due to hydration reactions that start between 10 and 50 days after

Morphological characterization
The scanning electron micrographs of the FGD wastes are shown in Figure 5. The leaching procedure is similar to the toxicity characteristic leaching procedure (TCLP) described in EPA method 1311. Solubilization test is applied to non-hazardous waste to establish it as inert or noninert, i.e., whether or not the parameters, after solubilization, are below maximum limits of potable water. According to these tests, the waste material can be classified as hazardous (Class I) and as non-hazardous (Class II), being that Class II is divided into Class II A-not inert and Class II B-inert. Table 8 and 9 shows the results of leaching and solubilization tests, respectively, and the maximum limit values determined by the Brazilian Standard Norm [25].
All element concentrations in the extract leachate of FGD samples (Table 8) were found to be below the maximum limit allowed (Annex F of the Norm ABNT NBR 10004), so the waste was considered Class II, non-hazardous waste.  Table 9 showed that concentrations of some elements of wet FGD and semi-dry FGD samples were above the maximum limit allowed (Annex G of the Norm ABNT NBR 10004). Therefore, according to solubilization results, the FGD samples can be classified as Class II A (non-inert) materials. The three combustion residues are non-hazardous solid waste that can be safely put into landfills.

CONCLUSIONS
The natural radioactivity and trace elements concentrations in wet FGD and semi -dry FGD by-products arising from three different coal-fired power plants were evaluated. In general, the activity concentrations of 40 K is higher in FGD samples of semi-dry process. These results suggested that wet FGD can be used in production of eco-friendly cementitious matrices. Furthermore, due to its chemical composition, the three FGD samples can be used as raw material for the synthesis of value-added sorbents like zeolite and calcium silicate hydrate compound.