**3. Test Results and Discussion**

Results of macroscopic analysis are presented in Table 4. The most important observations made in a part of this study was the colour test variation from brownish-grey (5YR4/1) for CFBC fly ash from Siersza to light olive-grey (5Y6) for CFBC fly ash from Turów. It is probably related to the amount of unburned coal in the analyzed specimen and the type of burned material: Siersza (hard coal) and Turów (lignite). The darker colour corresponds to the higher coal content, which is also visible on the basis of LOI results in the analyzed fly ashes.


**Table 4.** Result of macroscopic analysis of the CFBC ashes: S, T and K.

Reaction with HCl, for each of the fly ashes passed with the same intensity of the characteristic light "effervescence". So it should therefore be expected these specimens had similar content of CaOfree. Higher fineness (36.1%) was exhibited by CFBC fly ash (T); it was several age points higher than the other tested fly ashes (K and S).

Observations in the scanning electron microscope showed typical grains corresponding to fly ash from fluidized bed combustion boilers. In Figure 1 a high surface development and irregular shape of the fly ash particles is visible. The largest observed particle size reached 120 μm for T fly ash and 80 μm for K and S fly ash. Also, relics of coal (often like char, which was verified by EDS analysis) are shown, which are disadvantageous from the standpoint of application of these fly ashes in cement (Figure 2). The non-hydroxylated clay minerals and combustion sorbent calcium carbonate have been found.

**Figure 1.** Microstructure of CFBC ashes from different power stations: (**a**) T, lignite burning; (**b**) S, hard coal burning; and (**c**) K, hard coal burning. Magnification 25,000×, scale bar = 1 μm.

**Figure 2.** Microstructure of CFBC fly ash grain with visible unburned coal particle (**a**) and analysis in microarea (**b**); magnification 20,000×, scale bar = 2 μm.

The results of flexural strength and compressive strength of tested pastes are shown in Table 5. The results of flexural strength for all pastes with CFBC were in the range of 5.2–6.4 MPa and differences were statistically negligible. The reference paste revealed a lower value of flexural strength (4.6 MPa). A much higher (about 50%) 28-day compressive strength of pastes made with the 20% replacement of cement by CFBC fly ash from hard coal burning (K and S) was observed compared to the CEM I reference paste. Specimens with CFBC fly ash from lignite burning (T) achieved slightly lower values of compressive strength compared to specimens with fly ash from hard coal burning. The increase in CFBC fly ash content influenced the decrease of compressive strength of pastes. This tendency is much more visible with increase of CFBC addition to 30%. The obtained results are consistent with the literature data [31,32]. Šiler et al. [31] showed that the replacement of cement by 10, 20 and 40 wt % of fluidized bed combustion fly ash influenced on the increase of compressive strength compared to ordinary cement paste in the early age of hydration (up to 28 days). Hanisková et al. [32] analyzed the influence of fly ash from fluidized bed combustion on mechanical properties of pastes. They showed that the highest values for 28-day compressive strength were achieved by pastes with 30–40% replacement of cement by CFBC fly ash, twice as much as the reference paste without fly ash.


**Table 5.** The results of compressive and flexural strength of pastes stored in water in 20 ± 2 ◦C after 28 days of curing (MPa).

The phase composition was evaluated using the thermal analysis method. The loss on ignition and weight loss of analyzed specimens were associated with the relics of clay minerals (temperature up to 350 ◦C) as well as the oxidation of unburned coal residue, and they were used to determine the content of Ca(OH)2 and CaCO3. The total loss on ignition was identified by heating fly ash to a temperature of 1000 ◦C. The results of thermal analysis are presented in Table 6. The obtained measurements show that a total loss on ignition depends not only on the loss of the unburned coal relics, but also the

distribution of carbonates, portlandite and disposal of residual clay mineral water. Therefore, the LOI determined according to PN-EN 450-1 does not clearly reflect a precise content of unburned coal. This phenomena is the most visible for CFBC fly ash from Turów power plant (lignite burning), where the presence of the LOI (at 1000 ◦C) is mainly due to the calcium carbonate content (4.80%), and the relics of unburned coal accounted for only a fraction of a percent in the tested ash (0.8%).


**Table 6.** Selected properties of CFBC fly ash separated for the fractions below and above 0.045 mm.

\* Loss on ignition at 1000 ◦C.

The X-ray analysis performed on whole fraction of fly ash revealed the presence of crystalline phases such as anhydrite II (CaSO4), portlandite (Ca(OH)2), quartz (SiO2) and small quantities of clay minerals and calcium (CaO) and magnesium (MgO) oxide. A clear difference between the CFBC fly ash from the hard coal (K and S) and lignite (T) combustion is visible. There is much more portlandite (Ca(OH)2) and calcite (CaCO3) in the fly ash from the combustion of lignite (fly ash T) compared to hard coal combustion. Diffraction patterns of the analyzed specimens are presented in Figure 3.

For the further study two CFBC fly ashes were selected: K and T as representatives of the waste materials from the fluidized bed combustion of hard coal and lignite, respectively, with a low content of SO3. The analysis of the phase composition of cement pastes with 20% and 30% replacement of the cement mass by CFBC fly ash after 28, 200 and 400 days of maturation in water was conducted.

Based on the results of thermal analysis of cement paste, the chemical-bound water content in hydration products (HI) was determined.

The volume of HI is considered as the loss of water at temperatures up to 400 ◦C, which is associated with hydrated calcium silicates (C–S–H) and calcium aluminosulfate. Additionally, the presence of calcium carbonate (CC) was detected in the hardened cement paste, and it was considered as the mass loss at 800–1000 ◦C. XRD analysis of the above hydration products revealed that small amounts of ettringite and calcium sulfates mainly in the form of gypsum were present. Diffraction patterns of the cement paste specimens are shown in Figure 4. After the qualitative analysis in order to present the quantitative changes occurring during the hydration process, the change index (WZ) was introduced, which is used as standard procedure [33]. The above method was described in [12].

**Figure 3.** X-ray (XRD) patterns of CFBC fly ash from (**a**) S (hard coal); (**b**) K (hard coal); and (**c**) T (lignite).

**Figure 4.** XRD patterns of cement pastes matured in water for 400 days. (**a**) Without any addition; (**b**) with 20% CFBC fly ash T; (**c**) with 30% CFBC fly ash T; (**d**) with 20% CFBC fly ash K; (**e**) with 30% CFBC fly ash K.

The XRD pattern revealed portlandite as a major phase in the reference cement paste specimen analyzed after 400 days curing in water. The products of the cement hydration in the forms of portlandite, gypsum and ettringite, and products of the carbonation of the hydrated cement in the form of calcite were found in the specimens made with addition of the CFBC fly ash. The content of calcite was increased with increasing fly ash content.

Value of change index WZ was defined as the ratio of selected characteristic of the tested cement paste with CFBC fly ash to the same characteristic of the reference cement paste, expressed as a percentage. As a reference, a cement paste without addition of CFBC fly ash matured in the same period of time as specimens with addition of fly ash was chosen. The results of the comparison of phase changes in the tested cement pastes are presented in Table 7. Due to the increased sulfate and calcium ion content in the analyzed fly ashes, the most relevant in the examination of cement paste are ettringite content and amount of calcium carbonate introduced. It was revealed that the increase of ettringite and calcium carbonate content in the hardened cement paste was related to the increase in CFBC fly ash content.

**Table 7.** XRD estimated composition of hardened cement paste after 28, 200 and 400 days of curing (CFBC T—paste with fly ash from lignite burning. CFBC K—paste with fly ash from hard coal burning. 20/30–percent content of fly ash addition).


\* The ordinary cement paste without fly ash addition after 28, 200 and 400 days is equal to 100%; \*\* Absolute intensity ettringite planes of symmetry (d = 9.73 i d = 5.61) in conventional units.

It was noted that the specimens without the addition of CFBC fly ash (i.e., reference specimens) had less than 15% and 13% ettringite content after 400 days of maturing than specimens with 30% fly ash, T30 and K30. respectively. It was surprising that the largest difference in ettringite content between reference specimens and those with addition of CFBC fly ash was visible for the first 28 days of the maturation period. Ettringite decline in subsequent periods is probably associated with the progress of hydration and transformation of ettringite in monosulfate or connected the SO4 2– ions to the rising C–S–H phase with a low C/S ratio. Hydrated calcium silicates formed by the pozzolanic reaction decrease the content of the portlandite in the cement paste. This is consistent with the results regarding siliceous fly ash. It is known that the addition of the siliceous fly ash to Portland cement generally reduces the amount of portlandite and this is often accompanied with an increase in the amount of C–S–H with reduced Ca/Si ratio and AFm phases due to a higher content of Al2O3 in fly ash. The AFm phase of Portland cement refers to a family of hydrated calcium aluminates based on the hydrocalumite-like structure of 4CaO·Al2O3·13–19H2O, [34]. Also, the content of ettringite varies depending on the reactivity of the siliceous fly ash used. Studies to characterize the microstructure of concrete modified with addition of calcareous fly ash were performed by Glinicki et al. [35]. They showed that the addition of calcareous fly ash reduced the content of portlandite in the matrix by 45–74%. Results of tests conducted by Tishmack et al. [36] showed that the products of cement hydration incorporating calcareous fly ash included lower amounts of ettringite and higher content of AFm phases, including mainly monosulfates.

Microphotographs of the reference specimens and cement paste with addition of CFBC fly ash after 28, 200 and 400 days of maturation are presented in Figures 5 and 6. It is visible as a fine-grained and fine-porous microstructure in micro-areas occupied by C–S–H, verified by EDS analysis. The C–S–H phase created a spongy mass of the conformation of small grains, generally forming single fibrils with lengths less than 0.1 μm. Ettringite needles and micro-tubes with lengths of up to 2 μm occurred in air-voids.

The addition of the CFBC fly ash, both from hard coal and lignite burning, caused an increase in the content of ettringite, assessed on the basis of the X-ray diffraction analysis. After 28 days of curing the ettringite content in the specimens with 30% T fly ash was 31% higher and with 30% K fly ash, 24% higher than in reference specimens. The content of portlandite decreased with increasing CFBC fly ash content. In specimens with 30% ash from hard coal K burning, an increase of the content of calcium

carbonate over time is clearly visible. Similar observations regarding the microstructure of paste with CFBC fly ash from hard coal burning were made by Lee and Kim [37], who investigated the hydration reactivity of the CFBC fly ash. They concluded that the microstructure of pastes with CFBC fly ash (the mixing ratio of CFBC fly ash to water was set at 1.0) after 1 day of hydration consisted mainly of fibrous ettringite and various sizes of hexagonal-plate portlandite. After 91 days, the CFBC fly ash was hydrated to a considerable degree, as the reaction ratio of the anhydrous gypsum was more than 80%. The microstructure of CFBC fly ash pastes contained portlandite, ettringite, gypsum and C–S–H [37].

The phase composition of the hardened cement paste was not affected by prolonged exposure in water and temperatures of 20 ± 2 ◦C. Irrespective of the content of CFBC fly ash in concrete, the microstructure of the presented cement pastes is similar to the reference specimen. It can be assumed that the efficiency of mineral additives, like fluidized bed combustion fly ash in cement paste, can be similar to other non-standard fly ash (e.g., calcareous fly ash). It was found in [38] that the addition of calcareous fly ash resulted in an improvement of concrete durability. A beneficial reduction of chloride migration coefficient was observed, while the effect on the water and air permeability was similar to its effect on the compressive strength of concrete.

**Figure 5.** Microphotographs of hardened cement paste microstructure without CFBC fly ash addition, matured in water for 400 days. (**a**) Empty air void without crystalline hydration products; (**b**) C–S–H, ettringite, relics of clinker; (**c**) C–S–H and cluster of elongated ettringite needles; (**d**) C–S–H, ettringite, relics of clinker.

**Figure 6.** Microphotographs of hardened cement paste with CFBC fly ash (K30), matured in water for 400 days. (**a**) Cement paste; (**b**) C–S–H and relics of clinker, microcracks; (**c**) carbonated hydration products; (**d**) C–S–H and relics of clinker.
