Mineralogical, Chemical, and Petrographical Assessment of Fly and Bottom Ashes from Agios Dimitrios Power Plant, N. Greece, for Their Evaluation as Fillers in Concrete Batching

Abstract

Coal combustion residues are often useful components for the cement industry. This study represents a material characterization and screening analysis by focusing on the mineralogical, physicochemical, and petrographic compositions of fly and bottom ash samples from four Greek power plants in order to evaluate their suitability and potential in industrial applications, especially as fillers in cement manufacturing. Proximate analysis revealed LOI values exceeding ASTM C618-22 limits. The sum of SiO₂, CaO, and Al₂O₃ classifies the studied samples as Class C except one. Iron and magnesium oxides are among the major components, while S, Ni, and Sr are also contained in significant amounts. Calcite, quartz, and plagioclases dominate, corresponding to their geochemical profile, while secondary mineral phases (i.e., neo-formed minerals during coal combustion) such as natrolite and gehlenite, were also identified. Relatively high amounts of carbonized organic matter and unburnt organic particles point to the incomplete combustion process, revealing the risk of slagging into the combustion chamber; this is confirmed through the high slagging and fouling indices. The amount of the magnetic fraction is low; magnetic spherules with complex surface structures and a wide range of spherule sizes were observed. While the pozzolanic character of the samples is strong, high values of LOI, S content, and carbonized organic material make them suitable for the cement industry after further treatment only.

1. Introduction

With the rising of the global energy demand, both conventional and renewable energy sources are included in the energy mix, at variable proportions, depending on the region. Even if fossil fuels’ share in the mix is decreasing, they continue playing a dominant role worldwide. The clean energy policy requires optimal management of emissions and wastes. In 2024, coal contributed 34.5% of the total electricity generation [1], while in Greece, coal covered 6% of the country’s electricity supply [2], showing a remarkable decrease compared to 15 years ago (50% in 2012). Fly ash, a very fine, powdery material is the major solid residue of coal combustion in coal-fired power plants; it is produced in significant quantities, causing severe environmental concerns and solid waste management issues [3]. India, China, and the USA produce the largest amounts of fly ash annually, reaching in total almost 300 Mt/yr [4]. Bottom ash, another solid waste residue of the coal combustion, is the coarser fraction that is collected on the bottom of the combustion chambers and typically constitutes 10%–20% of the total ash generated [5].

In addition to fly and bottom ash, boiler slag or fluidized bed combustion ash are also produced during coal combustion. All four coal combustion residues (CCRs) have various physical, chemical, morphological, petrographic, and radioactive properties, which all depend mainly on the geological conditions of coal formation and the combustion process as well [6]. Fly ash particles are dominated by amorphous glassy material of aluminosilicate composition and crystalline phases of silicates, sulfides, clays, and oxides [7,8,9,10,11]. The magnetic fraction in fly ash is also significant, accounting for approximately 2 to 20% of the total and being composed mainly of minerals such as magnetite, maghemite, and hematite [12]. Physicochemical changes to which the feeding material is subjected during combustion, result in a redistribution of the major and trace elements. The occurrence mode of them is based on how they are bound and physically distributed in the coal [13]. Fly ash particles consist mainly of SiO₂, Al₂O₃, MgO, and Fe₂O₃, whereas CaO, Na₂O, P₂O, SO₃, and K₂O are commonly contained; trace elements are usually identified in minor amounts [14], depending on the mineralogical composition of coal; the same pattern is followed also for the bottom ash particles. The difference lies in the fact that the more volatile elements during combustion, such as As, Se, Pb, Hg, Cd, Zn, S, and Sb [15,16,17], tend to accumulate in fly ash; however, less volatile elements, such as Cr, Ni, and V, are commonly present in the bottom ash [18,19]. Most of the time, the triggered component is Ca, as it can influence the behavior of the fly ash during its disposal in the environment, as well as its potential in various applications, reforming its pozzolanic properties [20].

Morphological characteristics (shape, size, porosity) of the fly ash, typically composed of particles with spherical shapes, solid or hollow, and ranging in size between 0.5 and 100 μm [21,22], influence its performance in various applications. However, particles of unburnt coal, chars, and soot are often encountered in the mixture of the fly ash. These are irregular particles with solid or fused structures but may also be spherical (mainly chars). Chars (Ø 1–100 μm) and soot (Ø < 1 μm) are the carbon-rich residues from incomplete fossil fuel combustion [23]. Chars constitute a kind of indicator for the identification of the initial coal [24]. Sooty particles form aggregates and are chemically more stable than chars [25,26].

In more detail, fly ash is composed of three main components: (1) solid organic fraction, (2) solid inorganic fraction (crystalline and amorphous phases), and (3) volatile compounds, which are associated with both organic and inorganic fractions [27]. At the beginning of the temperature rising, the organic part is released into gases. During the burning process, the contained minerals melt at 1100–1500 °C, and molten droplets form. When temperature dropping starts, glassy spherical particles are formed. Heavier inorganic fraction settles as bottom ash, and the finer particles – in form of fly ash – are carried aside by the aid of flue gases and are detained by particulate collectors [7,9].

One quarter of the globally produced fly ash is utilized mainly in cement manufacturing, increasing its durability and workability [28,29]. The diversity characterizing fly ash also makes its characterization difficult in simple terms. Fly ashes with CaO < 18% belong to Class F; they usually derive from bituminous coal and anthracite and possess pozzolanic properties. In contrast, Class C fly ashes are usually produced from lignite and sub-bituminous coal combustion and contain CaO > 18%. In addition to their pozzolanic properties, they also possess a cementitious tendency [7]. Fly ash is also employed in wastewater treatment, solid waste management, the ceramic industry, soil stabilization and rehabilitation, and zeolite synthesis [30]. Its applications include road base construction and use in grout, plaster, wallboard, and other construction materials, enhancing products’ properties [29], as well as a filler in polymers, paints, and composites. Despite its wide range of applications, inadequate management of fly ash (inappropriate storage methods) poses environmental concerns such as groundwater contamination, balance disruption of the soil components and the aquatic ecosystems, increasing gas emissions, and alterations in fauna and flora development [31,32]. The cumulative effect of these perturbations leads to health issues, including respiratory problems (e.g., asthma, bronchitis), cardiovascular diseases, weakened immunity and elevated risk of lung cancer [32].

The Kozani-Ptolemaida-Amyntaio Basin (Northwestern Greece) hosts the largest lignite center in Greece, followed by the Megalopolis Basin (Southwestern Greece; Figure 1). Lignite mining in both regions has faced various environmental issues for decades, mainly regarding the environmental footprint of the fly ash management. During the period 2014–2016, fly ash produced from all power plants in Western Macedonia and Southwestern Peloponnese (Megalopolis) ranged from 6 to 9 Mt/yr, while for bottom ash the corresponding values ranged from 0.6 to 0.4 Mt/yr [33].

Fly ash from Ptolemaida Mining Center is classified as Class C with 15%–35% CaO, in contrast to fly ash from Megalopolis being classified as Class F with CaO < 10% [34,35]. These amounts indicate the pozzolanic properties of the siliceous fly ash from Megalopolis and both the pozzolanic and the hydraulic properties of Ptolemaida fly ash.

Iordanidis et al. [36] studied the past four bottom samples (out of a total of fourteen), focusing on their particle fractions (from <0.18 to >1.25 mm). They conducted TG/DTG, X-Ray Diffraction (XRD), and Energy Dispersive System (EDS) analyses combined with proximate analysis in order to evaluate their potential in concrete manufacturing. The present study constitutes a further step by examining the composition, qualitatively and quantitatively, of fly ash and bottom ash samples from the Agios Dimitrios Power Plant operated by the Greek Public Power Corporation (DEI), by emphasizing the integrated study of the char petrographic and magnetic features, in order to identify their properties and assess their potential for future applications. The focus is on material characterization and screening, based on geochemical, mineralogical, and petrographic parameters for their suitability, especially for cement manufacturing, without providing any performance validation.

2. Origin of Samples

In the Kozani-Ptolemaida-Amyntaio Basin located in Northwestern Greece, the Agios Dimitrios Power Plant is still in operation, whereas due to Greece’s energy transition policy, other plants (Kardia, Ptolemaida, and Meliti) have ceased operating. Today, the Main and South Lignite Fields are supplying the fuel to Agios Dimitrios Power Plant. The ash is transported to its disposal dump in the Akrini area via a conveyor belt system. There is also a modern lignite power plant, Ptolemaida Unit V, which is currently operating on a pilot basis until its conversion to a natural gas unit after a couple of years (Figure 1).

The margins and the basement of the basin consist of crystalline rocks of the Pelagonian and Axios geotectonic Units. From the oldest to the youngest, the sequence includes crystalline basement rocks, granites, Permian-Lower Triassic sedimentary rocks, Triassic-Jurassic carbonates, Upper Cretaceous limestones and flysch, and Neogene sediments, which contain the lignite layers with marls and clays [37]. The basin was formed by Quaternary tectonic movements, which created river systems depositing layers of sand, pebbles, marls, and clays at the lower part of the basin. Recent formations comprise the elluvial mantle and fluvial and lacustrine sediments. The characteristic geomorphological feature of the area is the intense relief with numerous mountains around the basin, with the highest being the Voras Mountain (Kaimaktsalan, +2524 m) at the northeast. The entire basin is over 100 km long and has an average width of 15 km, while the average altitude is 600 m above sea level [38].

Megalopolis Basin is located in the central part of Peloponnese (Southern Greece) and hosts the second largest lignite deposit in Greece. In March 2025, the Megalopolis Power Plant was shut down, and the mines ceased operations and supplying coal. The margins and the pre-Pliocene basement are mainly composed of marine sedimentary rocks such as chert, limestone, dolomite, and flysch deposited from the Jurassic to the Eocene, and belonging to the Tripolis and Pindos geotectonic Units. The Plio-Pleistocene sequence consists of six formations reflecting the climatic alternations. The main lignite-bearing Marathousa Member reflects palaeopeat accumulation under warm climate periods within the Cromerian Stage [39].

3. Samples and Methods

In total, fourteen samples of bottom and fly ash from Agios Dimitrios, Kardia, Meliti (Ptolemaida-Kozani-Amyntaio Basin), and Megalopolis (Megalopolis Basin) Power Plants were studied (

Table 1. The studied bottom (ba) and fly ash (fa) samples (PP: power plant).

Sample Origin	Bottom Ash Samples	Fly Ash Samples
Agios Dimitrios PP	DIMba, DIMba 1, DIMba 2, DIMba 3, DIMba 4, DIMba 5	DIMfa 6, DIMfa 7, DIMfa 8, DIMfa 9, DIMfa 10
Megalopolis PP	MGL
Kardia PP	KAR
Meliti PP	MLT

). Samples were composites of several days’ production and homogenized in order to be more representative for each station. Proximate analysis included air-drying (25 °C), grounding (Ø < 250 μm) and then drying at 105 ± 2 °C for 24 h according to ASTM D3173-03 [40]. The determination of the ash and volatile matter yields was performed according to ASTM D3174-04 and ASTM D3175 standard procedures [41,42], respectively. The ultimate analysis aimed to determine the contents of the elements C, H, N, S, and O based on ASTM D5373-16 [43] and was conducted using a Carlo Erba EAGER 200 Elemental Analyzer (Carlo Erba Reagents, Milan, Italy) at the Center for Instrumental Analysis of the School of Mineral Resources Engineering of the Technical University of Crete, in Chania. Loss of Ignition (LOI) was also determined at 950 °C following standard procedure [44].

Mineralogical analysis was performed by using a Bruker D8 Advance X-ray Diffractometer (Bruker Corporation, Billerica, MA, US), equipped with a LynxEye detector (Bruker AXS GmbH, Karlsruhe, Germany) and EVA software (v12) was used to identify the mineral phases. The scanning area covered the interval 2θ 2–70°, with a scanning angle step of 0.015° and a time step of 0.1 s. Furthermore, the concentrations of the major elements Al, Fe, Ca, Mg, K, S, Si, P, Ti, Na, and Mn, but also the trace elements Sc, V, Co, Cr, Ni, Cu, Zn, Rb, Sr, Y, Zr, Ba, Hf, W, Pb, Ce, La, and Th, were determined using a ZSX PRIMUS II RIGAKU X-Ray Fluorescence spectrometer (WDS-XRF) (Rigaku Corporation, Tokyo, Japan) at the Laboratory of Electron Microscopy and Microanalysis of the School of Natural Sciences, University of Patras. Additionally, polished blocks of the samples were prepared (Ø < 1 mm) according to ISO 7404-2 [45] and examined using a Leica DMRX coal-petrography microscope (Leica Microsystems GmbH, Wetzlar, Germany) under white incident light and blue light excitation and a ×50 oil-immersion objective (total magnification of ×500) at the Laboratory of Geo-Energy Resources Team, Department of Geology, University of Patras. Maceral analysis was conducted based on ISO 7404-3 [46], whereas for the maceral nomenclature, the ICCP System 1994 [47,48,49] was applied, and the quantification and characterization of the fly ash components followed the nomenclature and guidelines established by ICCP [50,51].

Furthermore, the magnetic fraction of the coal combustion residues (fly and bottom ash) seems to affect their properties when used in concrete manufacturing [28]. For the separation of the magnetic fraction of selected samples, 20 g of representative material (Ø < 1 mm) was mixed with 250 mL of deionized water and 10 mL of a dispersing agent [(NaPO₃)₆, 100 g/L], and placed with a magnet in a beaker. A magnetic stirrer, Heidolph MR 3001 (Heidolph Instruments GmbH & Co. KG, Bavaria, Germany), operated at low speed (< 100 rpm) without heating, was used in order to extract the strongly magnetic particles while ensuring proper particle dispersion. The separation process was repeated three times, with intervals of 5, 2, and 1 min, respectively [52]. Finally, for a better result, deionized water was applied with high pressure by using a nozzle to the magnetic fraction extracted from each sample in order to remove any clay material from the surface of the magnetic particles and make them more visible when examined under the SEM. This method was selected as a fast, efficient, and inexpensive procedure for the separation of the magnetic particles. The final magnetic fraction extracted from each sample was dried at 110 °C until reaching a constant weight, and then it was weighed and stored in plastic vials. A scanning electron microscope JEOL JSM6390LV (JEOL Ltd., Tokyo, Japan) of the Electron Microscopy Laboratory of the Aristotle University of Thessaloniki was used to observe the morphology of randomly selected magnetic particles.

4. Results and Discussion

The results of proximate analysis of the studied ash samples are presented in

Table 2. Proximate analysis results (moisture, non-combusted organic matter (NCOM), volatile matter, fixed carbon, and LOI) of the studied ash samples. All values are in wt.%, on a dry basis.

Sample	Moisture	Non-Combusted Organic Matter (NCOM)	Volatile Matter	Fixed Carbon	LOI
DIMba (bottom ash)	5.0	33.0	25.0	28.0	12.5
DIMba 1 (bottom ash)	4.0	31.0	25.0	25.0	31.0
DIMba 2 (bottom ash)	3.0	28.0	22.0	13.0	27.0
DIMba 3 (bottom ash)	3.5	38.0	25.0	13.0	16.0
DIMba 4 (bottom ash)	4.0	31.0	16.0	16.0	22.0
DIMba 5 (bottom ash)	4.0	37.0	22.0	15.0	25.0
DIMfa 6 (fly ash)	4.0	36.5	25.5	11.0	17.0
DIMfa 7 (fly ash)	3.5	26.0	16.0	10.0	26.0
DIMfa 8 (fly ash)	4.0	37.0	21.0	17.0	27.0
DIMfa 9 (fly ash)	3.0	28.0	17.5	11.0	17.0
DIMfa 10 (fly ash)	4.0	34.0	21.0	13.0	25.0
AVERAGE (DIM–DIM 10)	3.8	32.7	21.4	15.6	22.3
MLT (bottom ash)	2.5	20.0	12.0	8.0	21.0
MGL (bottom ash)	4.5	35.0	21.0	14.0	28.0
KAR (bottom ash)	3.5	38.0	22.0	20.0	29.0

. During combustion the moisture of the feeding coal evaporates; hence, low values are expected in the ash (DIM) samples, ranging between 3 and 5 wt.% (on average 3.8 wt.%), while for the MGL, MLT, and KAR samples the values are 4.5, 2.5, and 3.5 wt.%, respectively. The contents of non-combusted organic matter (NCOM) in the DIM samples are generally high and range between 26 and 38 wt.% (dry basis). For MGL, MLT, and KAR samples the values are 35, 20 and 38 wt.%, respectively. The MLT sample presents the lowest value compared to the average values of DIM samples, which have similar NCOM with MGL and KAR values. In addition, the range of volatile matter yield for the DIM samples is 16 to 25.5 wt.% (average of 21.4 wt.%), very similar to those of MGL and KAR with values 21 and 22 wt.%, respectively, but significantly higher than this of MLT (12 wt.%). The MLT sample follows the same pattern with fixed carbon yield, as it shows the lowest value (8 wt.%) compared to KAR and MGL (20 and 14 wt.%) and to the average value of DIM samples, too (15.6 wt.%), which is close enough to MGL value. The LOI for DIM samples, having a range between 12.5 and 31 wt.%, is similar to the MLT value of 21 wt.%, being quite different (i.e., lower) from the KAR and MGL samples (29 and 28.5 wt.%, respectively) and resembling the trend of volatile matter, fixed carbon, and NCOM.

High LOI values indicate high amount of unburnt organic material. Increased LOI might also point to the porous nature of ash, especially of bottom ash particles, which subsequently captivate water. The use of ash with this type of LOI value will lead to a possible inefficiency of the material for the cement industry as durability problems occur and may lead to discoloration, weak air entrainment, segregation, and low compressive strength [53]. Based on ASTM C618-22 [54], LOI values for both Class C and F fly ashes should be <6% in order to prevent the use of fly ashes that will adsorb significant quantities of air entrainer [55]. The LOI values for all the samples (Table 2) are higher than this baseline. Furthermore, high volatile matter results in a poorer performance in cement production, as it affects the admixture compatibility if it is not treated [56]. However, these factors are not the only ones controlling the capacity of ash material for the concrete industry applications.

Carbon content in high amounts reduces the workability of the concrete mix, as it can absorb water, which results in an unsuitable cementitious mixture to work with. It may also affect the air-entraining agents (AEAs), a critical tool for freeze and anti-cracking resistance [57,58].

ASTM C618-22 [54] sets a quantitative LOI limit of <6% for fly ash material for its use in concrete production. High LOI values often lead to high demand of water and high-range water-reducing admixtures (HRWRAs), so the concrete mix may need specific modifications in order to prevent these requirements [55]. Furthermore, different types of carbon-rich materials (unburnt carbon vs. powder activated carbon additions) present in the concrete mix, reduce the correlation and/or dependence between LOI (representing mostly the organic fraction) and the performance [55]. Experimental data [59] showed that concretes incorporating fly ash with an LOI value of 8%, at replacement levels exceeding 60%, exhibited longer setting times compared to concretes with lower LOI content (5%); in addition, higher superplasticizer addition was needed to achieve similar workability. However, no reduction of the long-term strength was observed when appropriate mix-design modifications were applied. Coppola et al. performed tests with concretes by incorporating fly ashes with LOI content ranging from 4 to 11%; their study showed again increased superplasticizer demand and slightly reduced slump to the mixture with the highest LOI value [60]. No clear negative trend in workability and durability performance was observed up to 11%. These data indicate that LOI and unburnt carbon content alone are not always strict indicators for concrete mix design and should be evaluated in combination with additional parameters.

The results of the ultimate analysis in selected samples are presented in

Table 3. The CHNS results of the selected studied ash samples (all values in %, on a dry basis).

Sample	C	H	N	S
DIMba (bottom ash)	3.80	0.25	0.10	0.54
DIMba 2 (bottom ash)	10.31	0.58	0.22	0.2
DIMfa 10 (fly ash)	19.33	0.39	0.22	0.86
MLT (bottom ash)	1.35	0.29	0.02	1.91
MGL (bottom ash)	2.96	0.12	0.05	0.36
KAR (bottom ash)	10.96	1.02	0.22	0.34

. DIM_fa 10 fly ash sample shows the highest value for C (19.33%), while the lowest corresponds to the MLT bottom ash sample (1.35%); this difference shows a wide value range between the selected samples. Hydrogen values range from 0.12% (MGL) to 1.02% (KAR), nitrogen values range between 0.02% (MLT) and 0.22% (DIM_ba 2, DIM_fa 10, KAR), and sulfur values range between 0.2% (DIM_ba 2) and 1.91% (MLT). According to ASTM C618-22 [54], SO₃ for both Classes C and F should not exceed 5%, whereas there are no restrictions or baselines for nitrogen content. High amounts of sulfur will adversely affect the workability and strength of the cementitious mixture, leading to volumetric expansion (bulging phenomenon) and cracking [53,61]. As it is shown, there is a deviation in all sets of values, which is interpreted to be related to the nature of the samples (fly or bottom ash) but also to the feeding coal properties, as well as to the combustion conditions.

Fly and bottom ashes are heterogeneous mixtures due to the multitude of major and trace elements contained. The heavy metal fraction of fly and bottom ashes is of major concern, given its potential impacts on human health and environmental ecosystems, especially when these ashes are incorporated into various applications, including concrete production. The chemical components of the ashes depend on several factors, including the type of feeding coal, the occurrence of useful or hazardous elements and their association with the organic and inorganic components of the coal, combustion conditions, volatilization-condensation mechanisms, and the particle size of the ash [53,62]. The results of the geochemical analysis of the fourteen studied fly and bottom ash samples are presented in

Table 4. Average data of the chemical analysis for major and trace elements of the fourteen studied samples.

Major Elements (wt%, Dry Basis)	DIMba–DIMba 5 (Average)	DIMfa 6–DIMfa 10 (Average)	KAR Bottom Ash	MLT Bottom Ash	MGL Bottom Ash
SiO2	29.09	27.77	31.95	31.19	35.99
Al2O3	13.21	14.36	12.19	14.26	14.22
Fe2O3	5.52	5.30	5.38	7.22	7.65
CaO	24.49	24.52	16.66	21.03	8.90
MgO	3.07	2.70	2.32	2.72	2.08
Na2O	0.23	0.22	0.36	0.25	0.40
K2O	1.24	1.23	1.30	1.36	1.89
P2O5	0.33	0.93	0.26	0.25	0.22
TiO2	0.56	0.51	0.55	0.67	0.63
MnO	0.02	0.05	0.03	0.03	0.03
Trace Elements (mg/kg, Dry Basis)
Sc	7	1	8	20	22
V	96	104	100	148	165
Cr	123	248	130	134	159
Co	6	7	1	14	bdl
Ni	467	301	356	653	397
Cu	73	31	79	114	108
Zn	68	61	51	61	69
Rb	56	42	58	53	97
Zr	144	118	153	110	172
Ba	198	215	219	90	336
Hf	4	0	5	bdl	5
S	16,505	21,910	6236	17,133	7782
Y	28	45	29	33	51
Sr	313	181	328	199	459
W	5	5	3	6	3
Pb	9	15	13	13	6
La	20	0	35	23	21
Ce	22	41	36	30	70
Th	12	9	12	13	13

bdl: below detection limit.

.

The MGL bottom sample displays the highest SiO₂ value (35.99 wt.%), Fe₂O₃ (7.65 wt.%), and K₂O (1.89 wt.%), while both the DIM_ba and DIM_fa samples present the highest CaO values, 24.49 and 24.52 wt.%, respectively. DIM_ba samples show the highest value for MgO (3.07 wt.%), and DIM_fa the highest for Al₂O₃ (14.36 wt.%). Na₂O, P₂O_5, MnO, and TiO₂ show almost similar values for all the samples ranging from 0.02 wt.% (for MnO) to 0.93 wt.% (for P₂O₅). SiO₂, Al₂O₃, and Fe₂O₃ are the major components, as expected. The range of the CaO is the main differentiated factor for the samples, distinguishing the calciferous properties of the samples, especially of Agios Dimitrios Power Plant (Northern Greece) with the other samples and especially that of Megalopolis Power Plant (Southern Greece). The chemical profile of the studied samples fits the results of studies conducted on ashes from Greece and abroad [10,19,53].

The minor elements of coals (Si, Al, Mg, Fe, K, Na, Ca, and S) are, in general, associated with the inorganic part and constitute the basic elements of the minerals [63]. In view of the chemical aspect, the presence of SiO₂ and Al₂O₃ in bottom ash contributes to its additional pozzolanic nature, similar to that in fly ash [64]. These compounds react with Ca(OH)₂ during hydration process of cement, to form additional calcium silicates hydrates (C-S-H) and calcium aluminate hydrates (C-A-H) [53]. Based on the ASTM C618-22 [54] standard for a fly ash to be classified as Class C or F, a sum of SiO₂, Al₂O₃, and Fe₂O₃ contents should be >50%; hence, the studied samples meet the criteria for this classification, and according to the Roy ternary diagram (Figure 2), they belong to Class C. MGL sample is spotted out of the Class C field, and this is probably attributed to its low CaO value (15%). Based on ASTM C618-22 [54], this sample is classified as F; however, on the Roy ternary diagram, it is projected neither on the Class F field nor on the Class C. This differentiation is attributed most probably to the low rank of the feeding coal. Both for the sum of SiO₂, Al₂O₃, and Fe₂O₃ and for previous parameters (e.g., LOI), ASTM C618-22 [54] is used as an internationally recognized standard, and its application in this case makes the data easily compared with a large body of research beyond North America. ASTM C618-22 [54] defines classes based on ash origins as Classes F and C, making the classification of the studied ash clearer compared with global datasets. Regarding the EN 450-1 standard, it basically applies to siliceous materials and hence, is not fitting to the studied carbonate-rich materials.

Additionally, during combustion, most of the trace elements of the feeding coal are retained in the fly and bottom ash and are concentrated in the smaller volume of the ashes compared to the original coal. Trace elements are affiliated either with the organic or the inorganic matter of coal, while commonly a combined affiliation is observed [67]. In the organic part, the elements usually form organometallic compounds, while in the inorganic part they may appear in accessory minerals, in ion substitutions within minerals, or adsorbed on the surface of clay minerals [68]. Silicon, Al, K, and Na are the main elements of quartz, clay minerals, and feldspars, while Ca and Mg occur mainly in carbonate minerals. Iron shows a mixed affiliation mode because a part of it is bound to clay minerals (e.g., in the illite structure), while it is a main element of pyrite, hematite, and magnetite, or it is part of the carbonate minerals structure, such as ankerite [69]. Table 4 shows that S, Ni and Sr are the dominant trace elements of the studied samples and range from 6236 (KAR) to 21,910 (DIM_fa average) ppm, 301 (DIM_fa) to 653 (MLT) ppm and 181 (DIM_fa) to 459 (MGL) ppm, respectively. Barium, Cr, Zr, and V are also contained with values > 100 ppm (V average value for DIM_ba is 96 ppm). In this study, statistical analysis with the Pearson simple linear correlation coefficient for DIM samples was used to determine the affiliation of the chemical components and to assess their possible sources. The results are available in the Supplementary Material Table S1.

Silicon, Al, Mg, Na, Ti, and K are associated (weak to strong correlation) with the aluminosilicate minerals such as feldspars and/or clay minerals [63]. Strong correlation of Si and Al oxides also confirms the pozzolanic properties of the material. Phosphorus is mainly identified in minerals such as apatite and monazite [70], as well as in clay minerals [67]. Moderate correlation of P with Na₂O and Y indicates its connection with aluminosilicate phases (plagioclases or glassy material). Titanium shows a strong negative correlation with CaO and S which indicates aluminosilicates, ilmenite, rutile or magnetite as its sources and not the carbonate and sulfur mineral phases of the ashes [71]. Vanadium and Rb are also associated (moderate to strong correlation) with the aluminosilicates and have been found adsorbed on the illite surface [72]. The strong correlation of V with Ti, Cr and Co indicates ilmenite, chromite, titanomagnetite as its secondary source [73]. Rubidium shows a strong correlation with Ti, Sc, Cu, and Ni. Manganese is usually bound in carbonate minerals and to a lesser extent, in clay minerals or even pyrite [67,74]. In the present study, no correlation of carbonates and sulfides is indicated with Mn (r < 0.3), which replaces Fe [75] but is also associated with clay minerals. Barium usually occurs in barite and feldspars [74]; however, in the present work, it does not show a correlation with S (i.e., barite) or P₂O₅, nor with Al₂O₃ and SiO₂. Tungsten (W) shows a strong correlation with Ba. Sulfur shows a strong correlation with Ca probably attributed to the presence of anhydrite or gypsum. Strontium is associated with carbonate minerals in coals [74,76], clay minerals and feldspars [63] or even with phosphate minerals such as apatite [77]. Strontium shows a strong correlation with Hf and La and a moderate one with Cu, while the correlation with SiO₂, Al₂O₃, CaO, and S is very weak. Cobalt is usually associated with clay minerals [67,77] but can also occur in sulfide minerals. Its moderate to strong correlation with Al₂O₃ and SiO₂ and its strong negative correlation with S confirm the aluminosilicate mineral phases as its sources. Cooper and Zn are usually associated with clay minerals [63,71], while other studies report an association with sulfide minerals [78,79]. In the present study, both Cu and Zn do not show a direct correlation with clay or sulfide minerals but present a strong correlation with Sc, Ni and Rb.

Although fly and bottom ashes can be utilized in a range of sectors, including the cement industry, there are still concerns of leachability and ecotoxicity, as many of their constituents limit their use as raw material [80,81]. Aluminum could react with water and release hydrogen (H₂). The rate of the gas produced depends on the smaller particle size, the molarity, and the reaction temperature of ash particles. More commonly, dissolution of soluble salts (CdCl₂, ZnSO₄), metal adsorption on oxides, and redox reactions with the most important of them, Cr⁶⁺ formation (highly mobile), may occur. Washing or chemical treatments may lead to the control of soluble salt, metal, and organic compound formation and enable the separation of smaller particles from the larger. It should be noted that the cement/filler ratio (in this case ashes) plays a significant role on the leaching process of heavy metals and organic compounds.

Every material that originates from an industrial process presents a unique morphology, the characterization of which provides insights into the origin of the material and the forming conditions. The petrographical analysis of selected studied samples indicates that carbonized organic particles surpass the rest of the fly ash components, ranging from 18 (MLT) to 26 vol.% (DIM), as reasonably expected. The contents of huminite and inertinite macerals reflecting unburnt (raw) lignite particles range from 3 to 4 to 7 vol.%, whereas the liptinite maceral group follows with values between 1 and 3 vol.% (

Table 5. Results of the maceral group analysis for the studied samples in vol.%, on whole-sample basis (DIM: Agios Dimitrios, MLT: Meliti, KAR: Kardia, MGL: Megalopoli, COM: carbonized organic material).

Samples	Huminite	Inertinite	Liptinite	COM	Inorganic Fraction
DIMba (<1.25) 1	3	7	1	26	63
DIMba (>1.25)	5	5		19	71
DIMfa 10 (<1.25)	4	4	2	24	66
DIMfa 10 (>1.25)	3	5	1	21	70
KAR (<1.25)	6	5	2	24	63
KAR (>1.25)	5	5	3	21	66
MLT (<1.25)	7	5	1	18	69
MLT (>1.25)	5	4	2	22	67
MGL (<1.25)	5	4		25	66
MGL (>1.25)	3	5		23	69

¹ 1.25 mm refers to the grain size of the studied sample.

). However, inorganic particles significantly outnumber all other particle groups, ranging between 63 and 71 vol.%.

The presence of the NCOM in the fly and bottom ash samples (Figure 3a–d) [82] is a proof of the incomplete combustion. Temperature ranges, the residence time of particles in the combustion zone, or deficient oxygen supply may lead to the incomplete oxidation of the huminite and rarely liptinite, as well as the thermal subsistence of the inertinite group macerals [83].

However, carbonized organic material (Figure 4a–f) dominates over the unburnt coal particles mainly composed of chars and secondary soot aggregates. The minerals commonly contained in the fly ash mixture and being identifiable under the organic-petrography microscope are quartz, as well as carbonate, sulfate, and clay minerals.

The composition and the morphological features of the fly and bottom ash particles depend on various factors, including the type of the feeding material, the process conditions, and the scale of the combustion unit/boiler type [82]. In addition, the presence of soot aggregates indicates a lower combustion efficiency since they need higher temperatures to completely burn [84]. In

Table 6. Fly ash component characterization of selected samples.

	DIMba (>1.25 1)	DIMba (<1.25)	DIMfa 10 (>1.25)	DIMfa 10 (<1.25)	KAR (>1.25)	KAR (<1.25)	MGL (>1.25)	MGL (<1.25)	MLT (>1.25)	MLT (<1.25)
Crassisphere
Tenuishpere
Crassinetwork	10	14	13	8	7.5	10	9	12	8	6
Tenuinetwork	10	9	12	12	19	8	7	11	19	7
Skeletal
Mesosphere	2	4	5	3.5	7.5	3	3	4	7	7
Mixed dense	4	5		6			5.5	6	2
Mixed porous	19	26	27	18.5	16	33	11.5	22	28	22
Inertoid	10	9	5	14	14	6	17	6	7	11
Solid
Fusinoid	6		6	3		6		7
Soot	39	33	32	35	36	34	47	32	29	47
Total (chars and soot)	100	100	100	100	100	100	100	100	100	100

¹ 1.25 mm refers to the grain size of the studied sample.

, a characterization for these fly ash components is presented based on the organic-petrography examination; the terminology followed the Atlas of Fly Ash Occurrences by Suarez-Ruiz et al. [50].

Char is a product of organic matter pyrolysis that can be produced directly from laboratory reactors or is included in the fly ash and bottom ash obtained from full-scale reactors. The char particles that did not burn inside the combustion chamber, may fall to the bottom of the furnace, get mixed with the mineral matter of the bottom ash or the fly ash particles and carried away by the flue gas stream [85]. Tenuinetwork chars dominate over the types of chars, followed by mixed porous particles. Crassisphere and crassinetwork particles also have a strong contribution to the fly ash mixture.

Tenuinetworks and crassinetworks show an internal structure, but in contrast to crassinetworks, tenuinetworks, display a porosity > 70% (due to the degassing process) and more than 50% of the wall area is <3 μm. In Polish char samples produced by fluidized-bed gasification of lignite from the Szszercow deposit, crassinetworks and inertoid morphotypes dominate in the residue mixture, while tenuinetworks and fusinoids are included in significant amounts [86].

Textinite (huminite group) is a porous maceral, which contains various amounts of residual cellulose and lignin. While it is heavily altered, its initial cellular structure is frequently preserved, mostly forming tenuous works [86]. Attrinite (huminite group) represents fragments (with a diameter of 10 μm or less) of usually soft plant tissues [82] and is highly reactive because of its low aromaticity. Morga and Bielowicz [86] suggested that in the lignite residues from fluidized-bed gasification, the morphotype of crassinetwork originates from detrohuminite. Misz-Kennan [87] claims that liptinite from feeding coal may generate detrital particles during the combustion process and its association with inertinite may form networks char morphotypes (i.e., crassinetworks, tenuinetworks).

Gehlenite, an epigenetic mineral typically identified in fly ash mixtures, was found in significant amounts (6 wt.%); it was formed by calcite, quartz, and feldspars at temperatures up to 700–800 °C [88]. Its formation depends on the furnace conditions and the crystallization during the cooling process [89]. At temperatures up to 400–500 °C [89], calcite probably formed anhydrite and lime, which are traced with values up to 3 wt.%. This is supported by the strong correlation Ca shows with S. Illite, hematite, and natrolite were also identified in <3 wt.% and justify the Fe₂O₃, NaO, and K₂O from the geochemical results. However, it should be noted that the amorphous phase, a very common component of fly ash mixture, was found even at higher amounts than calcite.

During coal combustion, the mineral matter undergoes chain reactions leading to physical and chemical changes. The results from the mineralogical study are presented in

Table 7. Mineralogical composition (in wt.%) of the studied samples.

	Sample	DIMba	DIMba 1	DIMba 2	DIMba 3	DIMba 4	DIMba 5	DIMfa 6	DIMfa 7	DIMfa 8	DIMfa 9	DIMfa 10	Avg	Std	Min	Max
Mineral
Amorphous phase		34	34	35	36	33	31	29	31	31	34	32	33	2	29	36
Quartz		15	17	19	18	14	12	16	14	15	13.5	14	15	2	12	19
Calcite		32.5	31	32	24	30	33	32	30	29	27	30	30	3	24	33
Plagioclase		10	11.5	11	11	11	11	9	11	11	12	12	11	1	9	12
Hematite		1	1	<1	<1	<1	<1	1	1	1	1	1	1	0	1	1
Gehlenite		6	5.5	<1	5	6	8	5	6	5	6	6	6	1	5	8
Anhydrite		<1	<1	<1	3	3	2	3	4	3	2.5	3	3	1	2.5	4
Lime		<1	<1	3	<1	3	3	2	3	2	2	2	3	0	2	3
Natrolite		1.5	<1	<1	<1	<1	<1	0	<1	<1	<1	<1	2	0	1.5	1.5
Illite		<1	<1	<1	<1	<1	<1	3	<1	3	2	<1	3	0	2	3

Avg: average value, Std: standard deviation, Min: minimum value, Max: maximum value.

. In both fly and bottom ashes of the DIM samples, calcite was identified as the primary mineralogical phase (30 wt.%), followed by quartz (15 wt.%) and plagioclases (11 wt.%) in major amounts, presenting a typical mineralogical profile of coal combustion residues [9,10,11,63,71] and corresponding to the geochemical profile of the samples with SiO₂, CaO, and Al₂O₃ as major components and the presence of K₂O and NaO.

Calcite was identified as the major mineral phase for KAR, MGL, and MLT samples too, in lesser amounts (

Table 8. Mineralogical composition (in wt.%) of the studied samples KAR, MLT, and MGL.

Mineral	KAR	MLT	MGL
Amorphous phase	30	34	38
Quartz	17	15	16
Calcite	29	25	23
Plagioclase	15	19	12
Hematite	1	1	<1
Gehlenite	8	7	7
Anhydrite	<1	<1	<1
Lime	<1	<1	<1
Natrolite	<1	<1	2
Illite	<1	<1	2

), and corresponds to the lower CaO percentage in this sample set than in DIM samples. Quartz and plagioclases were identified in significant amounts, while gehlenite is present at slightly higher amounts than those of DIM samples (7–8 wt.%). Hematite was again identified, and natrolite and illite were only in MLT samples, while anhydrite and lime are totally lacking from their mineralogical composition. Again, amorphous phases are identified in similar amounts as in samples from Agios Dimitrios.

Iron is contained in most of the coals in carbonates, sulfides, and sulfates. During the combustion process, iron is transferred to multiphase systems, including magnetite, maghemite, hematite, etc., which eventually constitute the magnetic fraction of the fly ash [90]. Hycnar et al. [91] indicated that in silicate-rich fly ash samples from Polish power plants, the magnetic fraction was 10 wt.%, whereas coal combustion may form magnetite in the total fly ash mixture, up to 16 wt.% [12,91]. Magnetic properties of the coal combustion residues affect their use in the industry and especially in concrete production. The magnetic fraction may reduce the long-term strength performance of the concrete and increase its thermal conductivity [92]. Additionally, it affects the final shade of the product, perhaps making it difficult to use for decoration and cladding purposes. Based on the results of the separation of the magnetic fraction from selected samples, DIM_ba-1 (bottom ash) sample’s magnetic fraction is 1.19 wt.%, DIM_fa 9 (fly ash) 0.80 wt.% and this of MGL (bottom ash) accounts for 2.2 wt.%. These values seem to fluctuate at low levels, regarding the possible use of the ashes at the concrete manufacturing. Furthermore, a higher amount of the magnetic fraction seems to be gathered at the bottom ash samples, connecting it with coarser grain sizes.

Scanning electron microscopy (SEM) combined with petrographic techniques is one of the most useful and widely used proxies for the characterization of the physical features of coal combustion residues, especially for fly ash. Coal source, combustion temperature, and cooling rate constitute the main formation pillars of these types of particles, controlling the size, the texture, and the presence or absence of pores [93,94,95]. The same function of variables applies for magnetic particles, the morphological characteristics of which were observed under the SEM. Multiple types of magnetic spheres were identified, presenting various textures and sizes (Figure 5a–g).

These spherules are mentioned in the literature under different terms such as ferrospheres and ferrispheres (depending on magnetite or hematite and limonite composition, respectively), magnetite globules or magnetic microsphreres [85]. Further classifications are based on the iron content and the spherules’ morphology [95]. Thread-type (a), foamy-type (b), orange peel-type (c), and smooth (d) surface structures of the magnetic particles in bottom ash samples (DIM_ba 1 and DIM_ba 3) were observed (Figure 5). The spherule’s size ranges from 40 to 570 μm with an average size of 197 μm, indicating an apparent heterogeneity among individual particles and aggregates. The same size heterogeneity is observed on the magnetic fly ash particles too (DIM_fa 10), with their size ranging from 15 to 180 μm (average 94 μm), slightly smaller than the bottom ash particles. Spongy-type (e), orange peel-type (f), and dendritic (g) surface structures were identified, as well as cenospheres with microspheres as infilling (h). These skeletal crystalline patterns of the studied samples are commonly observed in the coal combustion residues, indicating the high temperature, pressure, and cooling effect during the combustion process [95].

Fly ash is a great substitute for Portland cement, ranging from 15 to 35% of the total product [96]. Using Class C fly ash for Portland cement provides a variety of advantages, such as increasing early and late compressive strength, resistance to alkali silica reaction > 15%, less heat generation during hydration, etc. [96]. However, these advantages do not mean that there are no limitations and specificities that should be taken into account [97]. Also, the fine aggregate fraction of the concrete should be modified because fly ash has a lower specific gravity than Portland cement and therefore, occupies more volume for the same mass. If fly ash has a high calcium content, it should not be used in applications exposed to sulfates. Trial mixes are essential to ensure the correct ratio for achieving the desired properties.

Based on the analysis of the studied samples, deposition indices were calculated, indices that are used in combination with pilot power plant tests to predict the behavior of ash and its tendency to form ash deposits in combustion systems. These indices are as follows:

B/A (base/acid) = (Fe₂O₃ + CaO + MgO + Na₂O + K₂O)/(SiO₂ + Al₂O₃ + TiO₂),

(1)

Rs (Slagging index) = B/A × S,

(2)

Fu (Fouling index) = B/A × (K₂O + Na₂O),

(3)

Sr (Slag viscosity index) = SiO₂ × 100/(SiO₂ + Fe₂O₃ + CaO + MgO),

(4)

where B/A is proportional to the deposition rate. The Rs index refers to the deposition of fly ash on the surfaces exposed to heat radiation, and the Fu index to the depositions gathered in the heat recovery surfaces [98]. At

Table 9. Limit values and slagging and fouling tendencies for Rs, Fu, and Sr indices [88].

	Risk	Low	Medium	High	Extremely High
Indices
B/A		<0.5	0.5–1.0	1.0–1.75	>1.75
Rs (Slagging)		0.6	0.6–2.0	2.0–2.6	>2.6
Fu (Fouling)		0.6	0.6–40	>40	-
Sr (Slag viscosity)		72	65–72	<65	-

, limit values and slagging/fouling indices are presented according to data from Garcia-Maraver et al. [98].

Generally, the B/A ratio shows the slagging tendency of the ash, while the Fu index shows the fouling tendency. Low values of the Sr index indicate low slag viscosity and hence, high slagging risk. Thus, when Sr values are < 65, the tendency to form impurities is very high. Furthermore, the slagging index Rs indicates the slagging propensity [98].

From the values reported in

Table 10. Slagging and fouling values for the studied samples DIM (average), KAR, MLT, and MGL.

Indices	DIM (Average)	KAR	MLT	MGL
B/A	0.82	0.58	0.71	0.41
Rs (Slagging)	1.77	0.36	1.21	0.32
Fu (Fouling)	1.20	0.97	1.14	0.94
Sr (Slag viscosity)	46.8	56.8	50.2	65.9

it is indicated that KAR and MGL samples display the lowest slagging propensity, whereas DIM and MLT samples have a medium slagging tendency. The Fu index for all the samples shows medium values, while slag velocity values are high for all the samples except for MGL samples, in which they are medium to high. These values indicate the increased risk for reducing the heat transfer coefficient resulting in increased fuel consumption. Apart from the chamber combustion, cement manufacturing is affected by high values of slagging and fouling indices, resulting in sintering/clogging in the kiln and the formation of build-ups and rings. While rings grow thicker, they form a dam in the freeboard of the kiln, impeding the flow of the material and flue gases through the kiln. In addition, the heat transfer is disrupted, and the efficiency in precalcinators or preheaters is reduced [99]. High slagging values may produce more Si-Al gels (amorphous/glassy phases) and can enhance the formation of C-H-S affecting the hydration of the cement, while high fouling values may harm the durability of the concrete mix.

According to the slagging and fouling indices, but also to the results from the chemical analyses, magnetic fraction separation, and the proximate analysis, several factors, such as high LOI values or the raised amount of COM, should be treated in order to make the studied samples suitable for use in concrete manufacturing. Washing, electrochemical, and hydrothermal treatments are commonly proposed. Chemical activation could be carried out either through alkali or sulfate activation. Alkali activation may transform the powdery fly ash to a material with even better cementitious properties, resulting in higher mechanical strength. Even if magnetic fraction is not high in the studied samples, magnetic separation could remove the iron-rich particles and the heavy metal loads of them, leading to an increased chemical stability and reduction of corrosion and possible leaching. Froth floatation could also be a useful treatment method, with which the unburnt particles could be separated from the mixture based on their hydrophobicity [100].

The recycling economy is a valuable key for waste management. In order to reduce the impact of the wastes on the environment as much as possible, their utilization is increasing in various sectors of industry. From an industrial perspective, the findings highlight the significance of the petrography as a useful tool for characterization of the initial material (in our case, fly and bottom ash samples) and set the path for further experimental studies. It enables the identification of carbonized organic matter and the inorganic phases. When integrated into routine quality control, petrography can support mix design adjustments and suitable application domains (structural vs. non-structural concrete). Its approach enables the use of fly ash in concrete production.

5. Conclusions

Fourteen bottom and fly ash samples from four Greek power plants (Agios Dimitrios, Kardia, Megalopolis, and Meliti) were studied in order to assess their potential use in the concrete industry. Based on the obtained results, the samples could be suitable for the cement industry under restrictions and prerequisites in order to avoid harmful slag depositions in the combustion chambers.

LOI values among the studied samples exceed the limits of ASTM C618-22 regulations affecting the kiln operation and concrete quality, indicating the need for beneficiation treatments as well as optimization of the initial furnace conditions. High sulfur content is another hindrance, as it will lead to expansion, durability problems, and kiln issues.

A higher amount of CaO in DIM samples than in the MGL, KAR, and MLT samples is a main differentiation parameter. The SiO₂-Al₂O₃-Fe₂O₃ sum combined with the Roy’s diagram classifies DIM, MLT and KAR samples as Class C; whereas based on the CaO content MGL is classified as Class F. The strong correlation of SiO₂ with Al₂O₃ indicates the pozzolanic identity of the studied samples, implying increased, long-term durability, sustainability benefits and improved strength. The notable identified amount of amorphous phase possesses some possible beneficial characteristics in concrete manufacturing. Calcite, quartz, and plagioclases dominate in the mineralogical composition of the studied samples, clearly corresponding to their geochemical components, while the presence of anhydrite verifies the strong correlation of S with CaO. Moreover, the relatively low amount of the magnetic fraction for selected samples enhances the possibility of their use in concrete batching.

Based on the petrographic analysis, the presence of carbonized organic matter is quite high with chars and sooty aggregates. Unburnt and carbonized organic particles were mostly identified in DIM samples, considerably limiting the applications of these samples. This suggests an inefficient combustion process that compromises overall energy efficiency and affects the concrete quality and manufacturing process. Additionally, high slagging and fouling indices show a tendency for samples DIM for impurity formations.

Overall, the presented characteristics and screening results indicate that the samples could be suitable for the cement industry under restrictions and prerequisites in order to avoid malfunctions in kiln operation and combustion chambers. Washing, floating, and chemical and thermal activation could be possible beneficiation treatments. Follow-up studies involving performance validation shall be conducted for definitive conclusions.