Development of Low-Carbon Autoclaved Aerated Concrete Using an Alkali-Activated Ground Granulated Blast Furnace Slag and Calcium Carbide Slag

Abstract

The environmental impact of traditional construction materials has led to increasing interest in developing more sustainable alternatives. This study addresses the development of low-carbon autoclaved aerated concrete (AAC) through the complete replacement of ordinary Portland cement (OPC) with ground granulated blast furnace slag (BFS), activated with lime and, in some formulations, supplemented with calcium carbide slag (CCS). Five different AAC mixtures were prepared and evaluated in terms of workability, foaming behavior, compressive strength, phase composition, density, thermal conductivity, and life cycle assessment (LCA). The BFS-based mixtures activated with lime exhibited good workability and foaming stability. After pre-curing, the addition of CCS significantly improved the formation of tobermorite during autoclaving. As a result, the BFS–CCS formulations achieved compressive strengths comparable to the reference OPC-based mix while maintaining low densities (420–441 kg/m³) and thermal conductivities in the range of 0.111–0.119 W/(m·K). These results confirm the technical feasibility of producing structural-grade AAC with a lower environmental footprint.

1. Introduction

Autoclaved aerated concrete (AAC) is a construction material that originated in the early 20th century. It was patented in 1924 by the Swedish architect Johan Axel Eriksson in collaboration with Henrik Kreüger at the Royal Institute of Technology [1,2]. The formulation, often referred to as the “lime mix,” involved aerating a blend of limestone and ground slate. Also known as aerated cellular concrete (ACC) or autoclaved lightweight concrete (ALC), AAC is industrially manufactured and supplied in the form of blocks or precast components for applications such as walls, floors, and roofs. It has achieved widespread adoption across various regions including Europe, South America, the Middle East, and East Asia. Technically, AAC is a lightweight, porous concrete material composed of cement, lime, gypsum (or anhydrite), finely ground silica sand, and aluminum powder. The aluminum acts as the primary foaming agent, generating gas that expands the fresh mix. Subsequent autoclaving under elevated temperature and steam pressure promotes the development of strength, dimensional stability, and durability in the hardened material. Throughout the hydrothermal curing, calcium hydroxide reacts with silica, forming C-S-H phases which later recrystallize into tobermorite as the main binding phase and primary contributor to the mechanical strength of AAC [2,3,4,5].

AAC is increasingly recognized for its sustainability and efficiency in construction owing to its distinctive combination of lightweight characteristics, thermal insulation capacity, and sufficient mechanical performance. The material’s porous nature significantly enhances its thermal behavior, with conductivity values commonly falling below 0.14 W/m·K, which makes it highly suitable for energy-conscious building designs [6,7]. In addition, AAC provides excellent fire resistance, dimensional stability, and acoustic insulation, and it can be easily modified on-site through cutting, drilling, or shaping.

In recent years, growing attention has been given to the use of alternative raw materials in the manufacture of AAC, with the goal of minimizing reliance on non-renewable natural resources. Much of this research has concentrated on replacing silica sand—commonly employed as the main source of reactive silica in AAC formulations—either partially or entirely. Various substitutes, including zeolite, copper tailings, coal bottom ash, fly ash, rice husk ash, and recycled glass, have been examined for their ability to replace silica sand while still supporting the formation of key hydration phases under hydrothermal curing [5,7,8,9,10,11].

Despite this growing body of research on silica sand substitution, relatively few studies have explored the possibility of replacing cement and lime in AAC formulations as a strategy to reduce the overall environmental footprint. These two components are essential for providing calcium sources required for the formation of strength-giving phases such as tobermorite. Some recent investigations have proposed the use of ground granulated blast furnace slag, fly ash, or alkali-activated binders to reduce or eliminate the use of Portland cement [5,11,12,13]. The reduction of lime has also been investigated through the use of calcium carbide slag [14,15]. However, these approaches remain less studied, and further research is needed to fully assess their impact on phase formation, long-term performance, and autoclave compatibility. This article presents a laboratory study on the development of eco-friendly AAC through the use of calcium carbide slag and ground granulated blast furnace slag as substitutes for ordinary Portland cement and lime. The influences on the physico-mechanical and chemical properties of the final products were studied.

2. Materials and Methods

2.1. Materials

A variety of mixtures were formulated using combinations of sand, lime, synthetic gypsum, OPC, two different kinds of BFS (BFS J and BFS C + I), CCS, green autoclaved aerated concrete cutting waste (GAAC-CW), and aluminum powder. The elemental composition of each material was analyzed using X-ray fluorescence (XRF) with a wavelength-dispersive S8 Tiger Bruker spectrometer (Billerica, MA, USA). The results of chemical composition for each material are shown in

Table 1. XRF chemical composition of OPC, CCS, BFS J, BFS C + I, and GAAC-CW. L.O.I.: loss on ignition.

Oxide (wt. %)	OPC	CCS	BFS J	BFS C + I	GAAC-CW
SiO2	17.36	1.71	30.83	35.40	30.09
Al2O3	4.63	0.86	13.58	9.84	4.03
Fe2O3	4.03	0.10	0.87	1.04	2.25
CaO	60.61	67.67	46.58	40.94	36.2
MgO	1.29	0.14	4.26	5.85	0.649
Na2O	0.32	-	0.17	0.24	0.703
K2O	1.06	-	0.35	0.99	1.39
SO3	3.54	3.97	1.63	1.17	4.59
TiO2	0.31	0.02	0.65	1.63	-
P2O5	0.10	-	0.03	0.02	0.07
MnO2	0.13	-	0.16	1.68	-
ZnO	0.04	-	0.01	-	0.02
Cr2O3	0.02	0.01	0.02	0.02	-
NiO	0.01	-	0.01	-	0.01
CuO	0.02	0.01	0.01	0.01	0.01
SrO	0.05	0.10	0.07	0.09	0.03
ZrO2	0.04	-	0.04	0.05	0.01
BaO	0.03	0.03	0.06	0.12	0.04
Cl-	0.04	0.06	-	-	0.04
H2O	1.22	22.4	-	-	14.09
L.O.I.	5.16	2.87	0.62	0.45	5.51

.

The natural sand exhibited a Blaine fineness of 3931 cm²/g and an average particle diameter (D50) of 52.6 μm. Mineralogically, it consists primarily of quartz (72.7%), followed by albite (16.7%), microcline (7.2%), and other minor crystalline components. The lime used in this study was a CL90-Q (according to EN459-2), supplied by Carmeuse (Louvain-la-Neuve, Belgium), with 94.88% of CaO and a T60 reactivity of 8 min 47 s. Synthetic gypsum, obtained from the Termocentrala 2 coal-fired power plant in Craiova (Romania), served as the calcium sulphate source. OPC type II 52.5R was sourced from the Heidelberg Materials plant (Craiova, Romania) and had a Blaine fineness of 5390 cm²/g, including 10 wt.% limestone filler. The GAAC-CW was incorporated as an internal recycled by-product generated during AAC production at CRH (Craiova, Romania). Aluminum powder pastes SPOR AL 4507 (D50 = 49.5 μm) and AL 4510 (D50 = 29.6 μm) with different particle sizes were purchased from Policolor din 1965.

Two types of BFS were included. BFS J, provided by Cementos de Levante SL (Murcia, Spain), was milled to a Blaine fineness of 3950 cm²/g and exhibited a higher content of CaO and Al₂O₃ compared to BFS C + I. The latter was sourced from Cementos La Cruz SL (Abanilla, Spain) and ground to 3900 cm²/g. The MgO + CaO/SiO₂ + Al₂O₃ ratio was 1.14 for BFS J and 1.03 for BFS C + I, with higher values of this ratio generally correlated with faster setting times in alkali-activated BFS. Lastly, CCS, a by-product of acetylene (C₂H₂) gas production from calcium carbide, was included as an alternative calcium source. This material was supplied by Air Liquide in Zaragoza, Spain.

The phase composition of the BFS and CCS was analyzed by X-ray diffraction (XRD) using a Bruker D8 Advance high-power diffractometer (Bruker Corporation, Billerica, MA, USA). The XRD patterns of the BFS revealed a predominantly amorphous structure, characterized by a broad diffuse halo centered around 30° 2θ. A distinct crystalline peak corresponding to calcite (CaCO₃) was observed at 29.5° 2θ in the BFS C + I sample, indicating minor crystalline content, while the rest of the slag phases remained largely amorphous (Figure 1). In contrast, the CCS sample exhibited a more crystalline nature. The dominant phase identified was portlandite (Ca(OH)₂), accompanied by smaller quantities of a mixed calcium–magnesium–manganese oxide hydrate, specifically CaO·MgO·MnO₂·(H₂O)·O₄₆. This secondary phase presented a prominent reflection at 9° 2θ (Figure 2), confirming its presence in the crystalline fraction of the material.

2.2. Sample Preparation

Five AAC mixtures were prepared: one conventional AAC mix (Control AAC), two mixes incorporating BFS as a replacement for cement (BFS J, BFS C + I), and two additional mixes using the same slags as cement replacements with the addition of CCS to the formulation (BFS C + I—CCS, BFS J—CCS). All mixtures contained 0.08% aluminum by total weight of the mix, and the water-to-solid ratio in all mixes was 0.86. The composition of all these mixtures is shown in

Table 2. Mixture proportions of the studied AAC samples by weight (wt.%). OPC: ordinary Portland cement; CCS: calcium carbide slag; BFS J: ground granulated blast furnace slag J; BFS C + I: ground granulated blast furnace slag C + I; GAAC-CW: Green autoclaved aerated concrete cutting waste.

Mix	OPC (wt.%)	Lime (wt.%)	CCS (wt.%)	BFS J (wt.%)	BFS C + I (wt.%)	Sand (wt.%)	Gypsum (wt.%)	GAAC-CW (wt.%)
Control AAC	28	9	0	0	0	44	4	15
BFS J	-	9	-	28	-	44	4	15
BFS C + I	-	9	-	-	28	44	4	15
BFS C + I—CCS	0	9	9	0	23	41	4	14
BFS J—CCS	0	9	9	10	0	50	5	17

.

To prepare the mixtures, the procedure began with pre-heating both the raw materials and the mixing equipment (metal bowl and blade) 24 h prior to mixing. Initially, the aluminum powder was manually dispersed in a portion of the mixing water (10% of the total). Subsequently, all dry and wet components—except for the aluminum—were mixed for 60 s. The pre-dispersed aluminum solution was then added, and mixing continued for an additional 30 s. The fresh mixture was poured into standardized stainless steel prismatic molds and subjected to a pre-curing stage in a controlled climate chamber at 56 °C and 85% relative humidity for 40 min. After this initial period, the temperature was raised to 68 °C, and pre-curing continued for another 24 h. Molds had been conditioned beforehand under the same pre-curing conditions. Following this stage, three prismatic specimens (4 × 4 × 16 cm) were extracted from each mold (the excess foamed material that overflowed the mold was removed.). These specimens were then subjected to hydrothermal curing in a vertical laboratory autoclave at 190 °C and 12 bar of saturated steam pressure. The autoclaving process consisted of a 20 min vacuum phase (−1 bar), a heating ramp lasting 40–50 min, a dwell period of 5.5 h at 12 bar, and a 1 h cooling ramp. Finally, the specimens were removed from the autoclave and dried at 70 °C until constant weight.

2.3. Characterization

2.3.1. Initial Mixture Product

Prior to casting, the flowability of the mixture was evaluated using a flow test. A cone-shaped mold (7 cm in diameter, 6 cm in height) was completely filled with the mix. Upon vertical removal of the cone, the diameter of the spread on a plate was measured. On the other hand, after mixing with the aluminum and pouring the slurry into the molds, the internal temperature of the material was measured. Foaming behavior was also evaluated using a cylindrical container with an internal diameter of 6.5 cm and a height of 42 cm. A volume of 600 mL of fresh mix was poured into the container, and the rise in height was recorded at one-minute intervals to monitor the expansion over time.

2.3.2. Intermediate Green Product

The green compressive strength of the samples after the pre-curing period was determined in accordance with the UNE EN 772-1 [16] standard. For these tests, specimens were prepared following the same procedure described above but without the addition of aluminum in order to prevent material foaming. In addition, after the pre-curing stage, the samples were analyzed using an X-ray θ-θ Bragg–Brentano diffractometer (PANalytical X’Pert PRO, Malvern, UK, Cu Kα radiation, 45 kV and 40 mA) equipped with a real-time multiple-band detector (RTMS) (X’Celerator by Panalytical, Malvern, UK). Data acquisition was performed through continuous scanning in the range of 3–70°2θ, with virtual scans in steps of 0.02°2θ. The XRF patterns were interpreted using PANalytical’s X’Pert HighScore Plus 3.0 software and version 2.1 of Bruker’s DiffracPlus TOPAS software, allowing for the qualitative reconstruction of the mineral profiles by comparing them with the PDF databases of the International Centre for Diffraction Data (ICDD). The quantification of crystalline and amorphous phases was performed using the Rietveld method, with an internal standard (ZnO 20 wt.%) added to the powdered samples.

2.3.3. Autoclaved Product

The density and compressive strength were determined according to UNE EN 771-4 [17] and UNE EN 772-1 [16] standards, respectively. The thermal conductivity λ was measured using an HFM 446 Lambda Medium de NETZSCH device by means of guarded hot plate and heat flow meter methods according to UNE EN 12664:2002 [18]. All samples were tested in dry conditions after reaching constant mass.

As with the pre-cured samples, the quantification of crystalline and amorphous phases in the autoclaved specimens was carried out using an X-ray θ-θ Bragg–Brentano diffractometer and the Rietveld method, with an internal standard (ZnO 20 wt.%) added to the powdered samples.

Finally, LCA was conducted for the reference mix (Control AAC) and the BFS C + I—CCS and BFS J—CCS mixtures, with the aim of quantifying the associated environmental impacts. The analysis was conducted using Air.e v3.19.0.7 LCA software from Solid Forest S.L. and the ecoinvent v3.11 database according to ISO 14040 [19]. The functional unit was 1 m³ of AAC blocks manufactured. The system boundaries were defined according to a cradle-to-gate approach, which includes the stages from raw material acquisition to the completion of the manufacturing process: (i) extraction and processing of raw materials, including all processes related to extraction, transportation, and any necessary pre-treatment of raw materials used in AAC block production, and (ii) manufacturing, accounting for the resources consumed during the production process itself, such as electricity use. The construction, use, maintenance, demolition, and end-of-life stages were excluded, as the aim is to evaluate the environmental profile associated solely with the material production phase. For waste inputs, differentiated initial impact values were assigned: CCS was assigned zero environmental burden, whereas BFS was allocated 5% of its associated impact. Average transportation distances within Europe between raw materials suppliers and the AAC production site were taken into account.

3. Results and Discussion

3.1. Initial Mixture Product

Five AAC mixtures were prepared: one conventional AAC mix (Control AAC), two mixes incorporating BFS as a replacement for cement (BFS J, BFS C + I), and two additional mixes using the same slags as cement replacements with the addition of CCS to the formulation (BFS C + I—CCS, BFS J—CCS).

The use of BFS activated with lime as a replacement for OPC can lead to changes in the viscosity of the resulting mixtures due to differences in factors such as the heat generated and setting time. The viscosity and flowability of the mix influence pore size and distribution in the AAC. To evaluate the evolution of mix flowability, the consistency was measured over a period of 10 min. The results are presented in

Table 3. Flow of mixtures.

Mix	Flow, Time Zero (mm)	Flow After Ten Minutes (mm)
Control AAC	341	340
BFS J	330	332
BFS C + I	338	341
BFS C + I—CCS	339	340
BFS J—CCS	340	342

. The target flowability was set to achieve a spread flow of 340 mm ± 10 mm in order to match the flowability of the Control AAC mix. All mixtures complied with this range. Flow was measured again 10 min after the initial test, confirming that no significant changes in flowability occurred. The use of BFS-based mixtures activated with lime from the formulations showed similar flow behavior during the first minutes.

Another parameter that was evaluated was the internal temperature of the mixtures during the first 40 min (Figure 3). The Control AAC mix showed a higher temperature compared to the BFS-based mixtures due to the greater heat of hydration of OPC during the initial 40 min. At 7 min, the Control AAC mix reached a temperature 2 °C higher than the BFS mixtures. In the case of BFS mixtures, all exhibited similar internal temperatures. The use of CCS in the mixtures, being a by-product composed mainly of slaked lime, did not affect the temperatures generated during hydration.

The foaming of the mixtures was also evaluated. The results obtained are presented in Figure 4. All mixtures exhibit very similar foaming behavior, with a rapid initial rise up to 10 min, after which the foaming process begins to slow down. By 35 min, all mixtures have fully expanded. The Control AAC mixture and the two mixtures containing BFS J show similar foaming profiles. Mixtures with BFS C + I exhibit slightly lower expansion compared to the Control AAC reference mixture, although this difference does not result in changes in the final densities. The use of CCS does not lead to an increase in foaming rate, indicating that the available lime in the Control AAC mixtures is already sufficient to achieve similar foaming regardless of the additional hydrated lime added.

Figure 5 presents a visual comparison of the cross sections of the five AAC mixtures evaluated, highlighting the distribution, size, and density of the pores formed after curing. In general terms, differences can be observed between the reference mixture (Control AAC) and those formulated with BFS, both with and without the addition of CCS. The Control AAC sample exhibits a finer and denser porosity, with a homogeneous distribution of smaller pores. In contrast, the mixtures containing BFS display a more open pore structure and a less uniform distribution.

3.2. Intermediate Green Product

The evolution of the green body plays a key role in defining the final properties of AAC. In AAC, tobermorite is formed as a result of the hydrothermal reaction between calcium hydroxide Ca(OH)₂ and amorphous silica or from amorphous CSH [20,21]. During the autoclaving process, Ca(OH)₂ reacts with the dissolved silica under high-pressure steam conditions, leading to the crystallization of tobermorite. This reaction significantly enhances the mechanical strength of AAC by transforming the initially disordered C–S–H gel into a more stable and compact crystalline structure [20].

Table 4. Quantification of portlandite, quartz, and tobermorite phases using XRF and the Rietveld refinement method.

Mix	Pre-Curing			After the Hydrothermal Treatment
	Portlandite (%)	Quartz (%)	Tobermorite (%)	Portlandite (%)	Quartz (%)	Tobermorite (%)
Control AAC	9.0	29.2	0.0	0.2	16.7	40.6
BFS C + I	1.4	29.5	0.0	0.2	21.9	28.5
BFS J	3.6	29.2	0.0	1.3	17.3	27.8
BFS C + I—CCS	8.3	30.6	0.0	0.7	14.5	39.1
BFS J—CCS	7.9	28.5	0.0	0.8	18.2	36.8

and Figure 6 show the main phases identified after pre-curing, as determined by XRF. The portlandite present in the mixtures may originate from the GAAC-CW, from CCS, or from the hydration of calcium oxide (CaO), as well as from the BFS and OPC. The Control AAC sample contains 9% portlandite, a value that serves as a reference target for the other BFS-based formulations. In the BFS formulations where no additional lime source was used (BFS J and BFS C + I), the portlandite content after the pre-curing period was low: 1.4% for BFS C + I and 3.6% for BFS J. In the lime-activated BFS, the major hydration products are calcium silicate hydrate, monosulfate (kuzelite), calcium hydroxide, ettringite, and hydrotalcite [22], with the final portlandite content governed by the raw materials and the reaction mechanisms involved [23]. To increase the free portlandite content prior to the hydrothermal process, an alternative source of calcium hydroxide was employed. The incorporation of CCS into both BFS-based mixtures (BFS C + I—CCS and BFS J—CCS) resulted in a portlandite content comparable to that of the reference mix (Control AAC): 8.3% for BFS C + I—CCS and 7.9% for BFS J—CCS.

Another important variable is the green compressive strength of the material prior to hydrothermal treatment, as it is critical for ensuring the structural integrity of the material during autoclaving. Figure 7 shows the green compressive strength of the samples after the pre-curing period. All the BFS-based mixtures activated with lime exhibited higher compressive strength than the reference mix made with OPC (Control AAC). Notably, the mixtures incorporating BFS J showed outstanding performance, with compressive strength values approximately twice as high as the reference mix in the case of the BFS J formulation. The difference in performance between the two slags is attributed to the higher MgO + CaO/SiO₂ + Al₂O₃ ratio of BFS J (1.14 compared to 1.03 for BFS C + I), indicating greater initial reactivity and faster setting in alkali-activated BFS, which leads to mixtures with rapid early hardening. In alkali-activated BFS, setting is largely governed by calcium availability: rapid Ca²⁺ release and build-up promote the early precipitation of Ca-rich aluminosilicate gels, thereby accelerating the initial setting [24,25]. A higher CaO content in the BFS increases reactivity, whereas greater SiO₂ and Al₂O₃ tends to suppress early reactions and delay setting [24,26].

3.3. Autoclaved Product

3.3.1. Compressive Strength, X-Ray Diffraction, and Density

In general, the compressive strength development is inversely proportional to the porosity and density of AAC. However, the shape, size, and distribution of pores also significantly affect the mechanical properties of samples [12]. Beyond the physical structure, the formation of specific hydration products during hydrothermal curing plays a critical role. In particular, tobermorite is recognized as the key binding phase responsible for strength development in AAC, providing structural cohesion within the porous framework [2,3,4,5].

The compressive strength and density results are presented in

Table 5. Compressive strength, apparent density, and thermal conductivity of mixtures.

Mix	Compressive Strength (MPa)	Apparent Density [kg/m3]	Thermal Conductivity [W/(m K)]
Control AAC	2.51	425	0.111
BFS C + I	1.52	428	0.119
BFS J	1.33	433	0.112
BFS C + I—CCS	2.27	420	0.113
BFS J—CCS	2.49	441	0.119

. All measured densities are very similar, ranging from 425 to 441 kg/m³. This indicates that despite differences in porosity, where the control mixture exhibits finer and more uniformly distributed pores, while the BFS-containing mixtures display a coarser and more heterogeneous pore structure (Figure 5), the final densities remain comparable. In contrast, the compressive strength values show a clear distinction between the reference AAC mix (Control AAC) and those incorporating BFS and CCS, as opposed to the BFS-only mixtures. The Control AAC, BFS C + I–CCS, and BFS J–CCS mixtures reached similar compressive strength values of 2.51 MPa, 2.27 MPa, and 2.49 MPa, respectively. Meanwhile, the BFS C + I and BFS J mixtures without CCS achieved lower compressive strengths of 1.52 MPa and 1.33 MPa, respectively. These differences are primarily attributed to the varying tobermorite contents formed, consistent with previous studies [3,4,5].

As shown in Table 4 and Figure 6, the main phases formed after hydrothermal treatment reveal that the Control AAC, BFS C + I–CCS, and BFS J–CCS samples contained similar tobermorite contents of 40.6%, 39.1%, and 36.8%, respectively. By comparison, the BFS C + I and BFS J samples exhibited lower tobermorite contents of 28.5% and 27.8%. These variations explain the compressive strength trends. Additionally, portlandite content after hydrothermal treatment was below 1.3% in all cases, indicating that most of the portlandite reacted with silica sand to form tobermorite. Therefore, the initial lack of portlandite in the BFS C + I and BFS J mixtures appears to be a key limiting factor in tobermorite formation and compressive strength development. The use of CCS as an additional source of portlandite emerges as a viable strategy to achieve AAC products with both adequate mechanical strength and appropriate density, particularly in formulations based on lime-activated blast furnace slags.

According to the UNE EN 771-4 standard [17], the compressive strength of AAC blocks for load-bearing masonry must exceed 1.5 MPa. In order to calculate this strength, the values obtained in this study must be multiplied by 0.8 since the tests were carried out on samples dried to constant mass [27]. Both the Control AAC sample and the mixtures containing BFS and CCS meet this minimum requirement.

3.3.2. Thermal Conductivity

Thermal conductivity is presented in Table 5. In this study, the thermal conductivity values obtained follow the same trend as density, consistent with observations reported in previous works [4,9], with no significant differences observed among the different mixtures tested. The use of lime-activated BFS as a complete replacement for cement did not result in any noticeable changes in thermal conductivity or density values. In contrast, other studies in which BFS was used to partially replace aggregate reported no clear correlation between density and thermal conductivity. This behavior has been attributed to the differences in raw material densities and the thickness of the pore walls, which affect thermal conductivity [11]. A similar observation was reported by other researchers when sand was replaced with coal bottom ash, where thermal conductivity decreased as the proportion of coal bottom ash increased [9].

3.3.3. Life Cycle Assessment

In this study, the LCA was conducted following the Environmental Footprint (EF) 3.1 methodology, encompassing a broad set of environmental impact categories. This analysis followed a “cradle-to-gate approach”, including the stages from raw material acquisition to the completion of the manufacturing process.

The LCA of AAC blocks was conducted only on the mixtures that meet the minimum strength requirements established by the UNE EN 771-4 standard [17].

Table 6. LCA results for AAC Control, AAC BFS J–CCS, and AAC BFS C + I–CCS mixtures.

Impact Category [EF 3.1]	Unit	AAC Control	AAC BFS J—CCS	AAC BFS C + I—CCS
Acidification	mol H+ eq	0.43	0.22	0.22
Water resource depletion	m3 world eq. deprived	48.33	39.99	38.38
Ozone depletion	kg CFC-11 eq	2.67 × 10−6	2.28 × 10−6	2.27 × 10−6
Fossil resource depletion	MJ	1628.24	1208.32	1198.79
Mineral and metal resource depletion	kg Sb eq	2.00 × 10−4	1.00 × 10−4	1.00 × 10−4
Climate change (total)	kg CO2 eq	206.13	107.62	107.39
Freshwater ecotoxicity (total)	CTUe	163.59	104.28	107.95
Eutrophication, freshwater	kg P eq	2.83 × 10−2	1.82 × 10−2	1.80 × 10−2
Eutrophication, marine	kg N eq	1.14 × 10−1	5.09 × 10−2	4.89 × 10−2
Eutrophication, terrestrial	mol N eq	1.19	0.51	0.49
Photochemical ozone formation, human health	kg NMVOC eq	0.41	0.22	0.21
Particulate matter	Disease incidence	3.42 × 10−6	2.28 × 10−6	2.10 × 10−6
Ionizing radiation, human health	kBq U235 eq	10.97	7.92	7.92
Human toxicity, cancer effects (total)	CTUh	1.22 × 10−7	5.42 × 10−8	5.43 × 10−8
Human toxicity, non-cancer effects (total)	CTUh	9.88 × 10−7	3.67 × 10−7	3.60 × 10−7
Land use	pt (points)	377.65	303.01	267.49

summarizes the life cycle assessment results across all impact categories and material formulations evaluated.

The results show that AAC formulations incorporating BFS and CCS significantly reduce environmental impacts across all categories assessed under the EF 3.1 methodology when compared to the OPC-based AAC control mix. The most notable reduction is observed in the climate change category, where emissions decrease by nearly 48%. Similar trends are evident in other impact categories; for example, acidification potential is nearly halved, and fossil resource depletion drops by over 26%. Reductions are also observed in toxicity-related categories: freshwater ecotoxicity decreases by up to 36%, while human toxicity (non-cancer effects) is reduced by more than 60%. Additionally, land use impact decreases significantly, from 377.65 to 267.49 pt, a reduction of nearly 30%.

OPC-based AAC blocks exhibited the highest greenhouse gas emissions, reaching 206,2 kg CO_2eq, due to the significant contribution of OPC. In contrast, BFS-based mixtures achieved substantial emission reductions, with values of 107.5 kg CO_2eq and 107.3 kg CO_2eq for the AAC BFS J—CCS and AAC BFS C + I—CCS mixtures, respectively, highlighting the effectiveness of material substitution strategies.

The amount of BFS used does not lead to significant differences in the resulting emissions. Although the AAC BFS J—CCS mixture contains 57% less BFS compared to the AAC BFS C + I—CCS mixture, both exhibit similar final emission values. This outcome is attributed to the low environmental impact associated with both the transportation and production of the BFS material. Although it does not result in further emission reductions, using mixtures with lower BFS content does lead to cost savings, which encourages the adoption of such BFS-based formulations.

The emission calculations were carried out considering two stages: extraction and processing of raw materials and manufacturing (Figure 8). Focusing specifically on CO₂ emissions as this is one of the most widely used and policy-relevant environmental indicators, the reference mixture shows that the production stage accounts for 22% of total emissions, while the remaining 78% originates from the raw materials used, among which OPC and lime are the main contributors. Therefore, the choice of materials has the greatest impact on the final emission values. Notably, the elimination of OPC alone resulted in a 48% reduction in total emissions. Similar findings have been reported by other researchers evaluating the construction process, where OPC was identified as the main source of emissions [28]. For the BFS-based mixtures, the same production process was followed as in the conventional AAC formulation, leading to no significant differences in emissions associated with the manufacturing stage. These results emphasize the potential of locally sourced waste materials to promote more sustainable and cost-efficient construction practices in line with circular economy principles.

4. Conclusions

This study explored the development of low-carbon AAC by replacing OPC with BFS, activated with lime and complemented in some cases with CCS as an additional source of calcium. The following conclusions can be drawn:

• The substitution of OPC with BFS, when activated with lime, yields AAC with comparable workability and foaming, as evidenced by consistent flow values, foaming expansion, and internal temperature profiles across all formulations.
• BFS-based mixtures incorporating BFS demonstrated enhanced green compressive strength, reaching levels similar to or higher than the reference AAC mix.
• After hydrothermal curing, AAC samples made with BFS and CCS achieved compressive strength values above the minimum threshold for load-bearing masonry units (≥1.5 MPa), confirming the feasibility of producing structural-grade AAC without OPC. Notably, these mixtures also developed tobermorite contents comparable to that of the reference mix, highlighting the effectiveness of lime activation and CCS addition in promoting the formation of strength-giving phases.
• The thermal conductivity of the AAC samples ranged between 0.111 and 0.119 W/(m·K), with no significant differences observed among the different formulations. These values follow the same trend as density, which remained between 420 and 441 kg/m³.
• The use of BFS and CCS did not lead to any detrimental effects in terms of pore structure or final mechanical and thermal performance. On the contrary, it presents a viable pathway for reducing the carbon footprint associated with AAC production, achieving reductions of up to 48%.