A feasibility assessment of low-carbon concrete containing calcined clay and granite powder or dolomitic limestone in Burkina Faso Authors: Philbert Nshimiyimana, Annicette Kouspare Kadere, Iliassou Nouhoun Salou, Kassoum Kanazoe, Yawo Daniel Adufu, Decroly Djoubissié Denouwé, Seick Omar Sore, Henri R. Nemaro, Adamah Messan Published: 2025 ABSTRACT The development of low-carbon concrete (LCC) is crucial for reducing the large carbon footprint of the construction sector, in countries like Burkina Faso where there is a lack of high-purity limestone to produce clinker, resulting in the importation of clinker, which exists in dolomitic limestone form. This study assesses the feasibility of producing such LCC in Burkina Faso by utilizing local resources. It designs a class C25/30 concrete, containing 330 kg of cement per m3, where ordinary Portland cement was partially replaced (up to 50%) with a mix of calcined metakaolin clay (30%) and granite powder or dolomitic limestone (10–20%). Results show that the synergy between metakaolin and granite powder significantly improves performance. The optimal mixture, 60C30M10G, achieved a 28-day compressive strength of 35 MPa, representing a 35% increase over the reference concrete, while also enhancing tensile strength and resistance to acid attack and water absorption. This is due to a densified microstructure resulting from pozzolanic reactions and filler effects. Furthermore, the 50C30M20G design reduced CO₂ emissions by 30% while still meeting the target strength class. In contrast, dolomitic limestone provided a modest improvement. This study confirms the potential for high-performance, low-carbon concrete adapted for the West African context, aligning the circular economy with sustainable infrastructure development.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-32956-1. Introduction Owing to climate emergencies and the need to reduce the carbon footprint of the construction sector, the development of low-carbon concrete represents a major contribution towards decarbonization of the sector. Concrete, the most widely used construction material in the world, accounts for approximately 8% of greenhouse gas emissions1,2, primarily due to the production of clinker for Portland cement. Therefore, the low-carbon concrete is an alternative for mitigating the climate impact caused by cement production. Various studies have led to the development of low-carbon cements with excellent physical, mechanical, and durability properties, which could enable the development of low-carbon cementitious and/or alkaline activated alumino-silicates and geopolymer concretes3–12. However, the applications or feasibility assessments of these innovations to concrete design remain relatively unexplored, particularly in the specific context of sub-Saharan countries. In this context, the application of low-carbon concrete could offer sustainable solutions6,7,13 to meet the infrastructure needs associated with population growth while minimizing environmental impacts and, ultimately, maximizing socio-economic benefits. A study showed that SSRC-L35 concrete containing 35% LC3 (limestone calcined clay cement) and SSRC-L50, containing 50% LC3, exhibited CO2 emission reductions of 32% and 42%, respectively, compared to the NAC-LO reference concrete without LC3. This study showed a 23% to 28% reduction in energy consumption compared with the reference14. In addition, SSRC-L35 and SSRC-L50 concretes achieved acceptable compressive strengths of 42.5 MPa and 38.5 MPa, respectively, compared to 46.4 MPa for the SSC-LO reference concrete14. Another study15 has assessed the decarbonization potential of the construction sector through structural system and concrete mix design choices for mid-rise buildings. The authors have shown that using low-clinker cement in low-binding concrete can reduce CO2 impact by 50% per built-up area. This case study concluded that implementing three strategies: structural design optimization, optimized material selection, and reduction of material carbon content, could reduce the typical carbon footprint of structural elements in a mid-rise building from 232 kg CO2eq/m2 to just 58 kg CO2eq/m2, a four-fold reduction. A study on the fresh properties of quaternary mixture-based pastes and concrete using LC3 explored the incorporation of 10% of different supplementary cementitious materials, including silica fume, fly ash, and sugarcane bagasse ash. Mini-slump and Marsh funnel tests have shown that LC3 pastes require 150% more consumption of ether polycarboxylate superplasticizer than Portland cement paste and exhibit longer setting time2. The above studies highlight the potential to achieve low-carbon concrete with LC3 cement, owing to the synergistic hydraulic, pozzolanic, and filler effects among the clinker, limestone, and calcined clay components. In Burkina Faso and the regional context, similar approaches were used to produce lower-carbon cement containing imported clinker and locally available natural mineral additions such as calcined clay, granite, and tuff16–18 and agro-industrial by-products such as coal bottom ash, rice husk ash, and glass powder19–21; and even produced high-performance concrete22. However, these local studies could not achieve the common LC3 due to the lack of high-grade limestone; substitution with granite and dolomitic limestone powder offers a locally available, low-cost alternative and supports regional circular economy initiatives. Given the abundant local materials, including kaolinite clay, granite dust, and dolomitic limestone, which can be used to develop high-performance, environmentally friendly concretes, Burkina Faso aims to promote a green, circular economy, focusing on eco-construction. Therefore, this study specifically aims to explore the feasibility of designing low-carbon concrete by substituting cement with calcined metakaolin clay, granite dust, or dolomitic limestone, and to evaluate the influence of these additives on the physico-mechanical properties, durability, and sustainability indicators of concrete. Materials and methods Materials Cement (C) was a CEM I 52.5 N produced by the CIMIVOIRE (Cote d’Ivoire). Its chemical composition is mainly CaO, SiO2, and Al203 (Appendix 1). Calcined clay of metakaolin (M) was produced by calcining kaolinite (61%) clay obtained from a Saaba quarry located approximately 10 km from the city of Ouagadougou. It was crushed and sieved to 80 μm, and the fines were calcined at 700 °C for 3 h. Its chemical composition is mainly SiO2 and Al203 (Appendix 1). CEM was reported to have a Blaine specific surface area of 3299 cm²/g and a specific density of 3.1 g/cm316. Metakaolin (M) was used as the amorphous (70%), highly reactive aluminosilicate pozzolan, with Blaine SSA of 6608 cm²/g and a specific density of 2.63 g/cm316. Granite dust (G) is a residue from a civil engineering company quarry (GLOBEX construction). It is located approximately 20 km away from Ouagadougou. The samples were crushed and sieved to 80 μm. Its chemical composition is rich in SiO2 and Al203 (Appendix 1). The granite dust was used as a non-reactive filler had specific density of 2.31 ± 0.01 g/cm3, compared to 2.65 g/cm3 with a Blaine SSA of 2199 cm²/g from16. Dolomitic limestone (L) was obtained from CIMCARRIERE in Bobo-Dioulasso, about 330 km from Ouagadougou, and was prepared by crushing and sieving to 80 μm. Its chemical composition is rich in SiO2 and CaO (Appendix 1). The dolomitic limestone had a specific density of 2.73 ± 0.07 g/cm3 compared to the reported value of 2.79 g/cm323. The small specific surface area of G, compared to CEM and Mk, could help reduce water demand, thereby improving concrete workability. The superplasticizer was VISCOCRETE, an aqueous solution of modified polycarboxylates24. The local sand with a 0/5 grain size was sieved to 2 mm, and the passers were washed to obtain clean sand with a fineness modulus of 2.16. The crushed granite aggregates, with a Los Angeles coefficient of 15% and a grain size of 5/15 mm, come from the quarry of Groupe Québec Afrique Sarl, located approximately 8 km from the city of Ouagadougou. The grain size curves show that the sand is well spread, with a uniformity coefficient CU of 2.5, whereas the gravel is uniform (CU = 1.7) (Appendix 2). Methods The composition of low-carbon concrete C25/30, dosed with 330 kg/m 3 of cement, 842 kg of sand, and 1161 kg of gravel, is presented in Table 2. It is based on a low-carbon cement design proposed at the LEMHaD laboratory. The reference design is a concrete designed 100 C, containing mainly CEMI cement, sand, gravel, water, and superplasticizer. Partial substitutions of the cement were made by incorporating a mixture of 30% of metakaolin (M) and 10 to 20% granite dust (G) or dolomitic limestone (L), i.e., C60M30G10, … The water-to-binder ratio (W/B) was maintained at 0.48 for all mixes. The superplasticizer was added at 5% of the mass of CEM for the very plastic reference concrete, then increased to 6–7% of binder (CEM + M + G or L) for other mixes to maintain a very plastic consistency class (Table 1). The first step was to mix the aggregates (gravel and sand), cement, and additions (M, G, or L). After thoroughly homogenizing the dry mixture, water was added gradually, and the superplasticizer was added directly into the humidified concrete for efficient interaction, as shown in the preliminary tests. The total mixing time was 15 min to ensure proper homogenization, followed by assessing the consistency. The consistency was assessed by measuring the slump on the Abrams cone in accordance with NF EN 12350-225. The compressive strength was evaluated at 7 and 28 days (Appendix 3), and the splitting resistance was evaluated at 28 days on cylindrical specimens (Ø100 mm × H200 mm), respectively, in accordance with NF EN 12390-326 and NF EN 12390-627. The density, determined by hydrostatic weighing, and the ultrasonic wave propagation velocity, measured by Pundit, were determined for these specimens before crushing, respectively in accordance with NF EN 12390-728 and NF EN 12504-429. Water absorption by capillary immersion at 15 min, 30 min, 1 h, 2 h, 4 h, 8 h, 12 h, and 24 h, and absorption by total immersion after 24 h were determined to deduce the total water accessible porosity, in reference to22. The resistance to acid attack was tested to evaluate the deterioration of the sample by sulfuric acid (H2SO4) solution with an initial concentration of 3% and weighed at 7, 14, and 28 days to evaluate the mass loss22. A cradle-to-gate life cycle assessment was conducted, referencing the life cycle inventory database developed for Burkina30. The function unit of m3 was used to reflect the common practice in the design of concrete, and following ISO 14,04031and BS EN 15,80432. Table 1Design of C25/30 concrete: quantity (kg) of components per m3 of concrete.Mix designCMGLSandGravelWaterSuperplasticizer (%)W/B(-)100 C330000842116115950.4860C30M10G19899330842116115960.4855C30M15G181.59949.50842116115960.4850C30M20G16599660842116115970.4860C30M10L19899033842116115960.4855C30M15L181.599049.5842116115960.4850C30M20L16599066842116115960.48 Table 1Design of C25/30 concrete: quantity (kg) of components per m3 of concrete.Mix designCMGLSandGravelWaterSuperplasticizer (%)W/B(-)100 C330000842116115950.4860C30M10G19899330842116115960.4855C30M15G181.59949.50842116115960.4850C30M20G16599660842116115970.4860C30M10L19899033842116115960.4855C30M15L181.599049.5842116115960.4850C30M20L16599066842116115960.48 Design of C25/30 concrete: quantity (kg) of components per m3 of concrete. Results and discussion Workability Figure 1 illustrates the evolution of the slump of different flesh concretes containing metakaolin, granite dust (G), and dolomitic limestone (L) in relation to the superplasticizer (SP) dosage. The concretes 60C30M10G, 55C30M15G, and 50C30M20G (Fig. 1a) exhibit slumps of 12.4 cm, 13.5 cm, and 12.5 cm, respectively, despite containing up to 6–7% superplasticizer; compared to the reference 100 C (13.5) containing 5% Sp. The Mk increased water demand and thus Sp dosage at constant W/B, as it increased the normal consistency of binder containing MK, whereas the granite has an opposite effect4. This variation suggests a complex influence of granite dust and metakaolin on the rheology of the concrete. At low dosages, between 10 and 15% G, the low water demand and fineness of the granite dust appear to compensate for the high water demand of Mk, therefore enhancing the lubrication and improving the mobility of the fresh concrete. However, at a 20% G, the increased internal friction and possibly water absorption of the granite dust reduced the slump, necessitating a higher superplasticizer dosage to maintain the very plastic consistency. Fig. 1Evolution of slump of concrete with (a) granite dust, (b) dolomitic limestone at W/B of 0.48 and respective Sp dosage. Evolution of slump of concrete with (a) granite dust, (b) dolomitic limestone at W/B of 0.48 and respective Sp dosage. The concretes containing dolomitic limestone (L) showed greater improvement in workability compared with G. Concrete 60C30M10L, 55C30M15L, and 50C30M20L (Fig. 1b) exhibit slumps of 15.5 cm to 15.2 cm for the SP content of 6%. This positive effect on workability can be attributed to the filler effect of dolomitic limestone, which improves the compactness and dispersion of solid particles in the paste, thereby reducing internal friction and resulting in lower water demand than granite dust, thereby achieving the very plastic consistency at lower Sp dosage. Although the present study did not specifically assess the efficiency of interactions between the Sp and different types of mineral additions, the current results indicate better interaction with dolomitic limestone, as evidenced by its higher slump than with granite at similar W/B and Sp content. This suggests that further assessment should deepen this observation and use various types of Sp. Apparent density Figure 2 shows the evolution of the apparent density of the different low-carbon concretes after 28 days of curing. For the concrete containing granite dust (G), the apparent density evolves in the range of 2604 to 2670 kg/m³ (Fig. 2a), compared with 2495 kg/m³ for the reference concrete 100 C. This increase indicates that incorporating granite dust fills voids in the cement matrix, thereby reducing porosity and improving the compactness of the concrete. This effect is particularly pronounced for the 55C30M15G and 50C30M20G, suggesting optimal granite dust filling. In contrast, for mix designs containing dolomitic limestone (L), the apparent density remained quasi-constant at 2495–2517 kg/m³ (Fig. 2b), which is also lower than with granite. This smaller variation can suggest a less filling effect of dolomitic limestone (L) particles in the cement matrix. Unlike granite dust, dolomitic limestone did not affect the apparent density of concrete. This observation can be justified by higher air content in the concrete containing L than in G at the constant W/B ratio. L-concretes were more fluid than G-concretes, which may have entrained air in the micropore network (Sect.3.3 and 3.6). These results indicate that partial substitution of cement with metakaolin and granite dust has a higher effect on the apparent density of concrete than dolomitic limestone. An optimal dosage of these additions is essential to prevent excessive density reduction, which is partially associated with increased porosity and, consequently, potential reduction in mechanical performance or durability. Fig. 2Evolution of apparent density of concrete with (a) granite dust, (b) dolomitic limestone. Evolution of apparent density of concrete with (a) granite dust, (b) dolomitic limestone. Propagation of ultrasonic pulse velocity Figure 3 shows the evolution of ultrasonic pulse velocity (UPV) of low-carbon concrete containing granite dust or dolomitic limestone at 28 days. For concretes incorporating granite dust (G), UPV significantly increased compared to the control 100 C. The concretes 60C30M10G (5296 m/s), 55C30M15G (5250 m/s), and 50C30M20G (5255 m/s) reached higher velocities than the reference concrete (4929 m/s) (Fig. 3a). This is due to increased compactness and a denser structure resulting from the combined action of metakaolin and the granite filler (Sect. 3.2). This confirms that partial substitution of cement with metakaolin and granite filler promotes a lasting improvement in concrete compactness, attributed to the pozzolanic and filler effects that refine porosity and increase matrix cohesion. Similarly, concretes containing dolomitic limestone dust (L) show the same increasing evolution, but to a limited extent. The UPV of concretes 60C30M10L (5138 m/s), 55C30M15L (5146 m/s), and 50C30M20L (5036 m/s) were higher than that of the control (4929 m/s), indicating a similar improvement in compactness, due to the filling effect and synergy with metakaolin (Fig. 3b). This suggests that dolomitic limestone dust also provides densification of the structure, even though it is limited by a higher content of entrained air (Sect. 3.2). Fig. 3Propagation of ultrasonic pulse velocity of concrete with (a) granite dust, (b) dolomitic limestone. Propagation of ultrasonic pulse velocity of concrete with (a) granite dust, (b) dolomitic limestone. Compressive strength Figure 4 shows the evolution of the compressive strengths of low-carbon concrete containing granite or dolomitic limestone dust at 7 and 28 days. The concretes 60C30M10G, 55C30M15G, and 50C30M20G showed a significant improvement in compressive strength compared with the reference 100 C (Fig. 4a). At 7 days, the strengths were 21 MPa, 21 MPa, and 19 MPa, respectively, compared to 19 MPa for the control 100 C, increasing by 11%, 11%, and 0%. At 28 days, the compressive strengths were 35 MPa (60C30M10G), 29 MPa (55C30M15G), and 26 MPa (50C30M20G); corresponding to the increases of 35%, 12%, and 0%, respectively, compared to the 26 MPa of the control 100 C. This increase is mainly attributed to the pozzolanic reaction of amorphous silica (SiO₂) and alumina (Al₂O₃)-rich in metakaolin and the filler effect of granite dust, which densifies the microstructure by reducing porosity. These results confirm that incorporating granite dust and metakaolin significantly improves the compressive strength5,17,33–35. The concrete containing dolomitic limestone 60C30M10L, 55C30M15L, and 50C30M20L (Fig. 4b) showed similar behavior, but to a lesser extent than granite. At 7 days, the strengths were 23 MPa, 21 MPa, and 20 MPa, representing increases of 21%, 11%, and 5% compared to 19 MPa for the control 100 C. The filler effect of dolomitic limestone could have promoted mechanical development by improving the compactness and intergranular adhesion of the paste. At 28 days, the strengths are respectively 26 MPa, 21 MPa, and 17 MPa, representing 0%, −19% and − 35% compared to 26 MPa for the control 100 C. The early (7 days) improvements in strength with 60C30M10L (21%) and 60C30M10L (12%), which are higher than with granite (11%), can be related to the good filler effect of dolomitic limestone. However, the decreases in strength at 28 days beyond 60C30M10L (0%, −19%, and − 35%), contrary to granite (35%, 12% and 0%), indicate a potential detrimental effect of dolomitic limestone on the reactivity mechanism. This could be attributed to dilution and most likely to the detrimental effect that dolomitic limestone may have on the microstructural development of cured pastes. It is highly possible that dolomitic limestone promotes the formation of non-cementitious and/or degrading products, which destroy the microstructure and degrade macroscopic stability. These results suggest that dolomitic limestone cannot be used in concrete beyond 10% and require a dedicated assessment of the reaction mechanisms and microstructural formation in the cementitious systems containing calcined clay and dolomitic limestone. Fig. 4Compressive strength of concrete with (a) granite dust, (b) dolomitic limestone. Compressive strength of concrete with (a) granite dust, (b) dolomitic limestone. Splitting tensile strength Figure 5 presents the results of the splitting tensile strength on the concrete specimens. The reference concrete exhibited an average strength of 2.1 MPa, whereas all other concretes reached higher values, indicating the beneficial effect of partial cement substitution. For concrete containing granite dust, the 60C30M10G and 55C30M15G (Fig. 5a) achieve a strength of 2.8 MPa, a 33% increase over the 100 C reference. This highlights that the effect of metakaolin and granite predominates in the mechanical improvement. However, the 50C30M20G had a strength of 2.4 MPa, reflecting the dilution effect of the active binder. An optimal amount of granite dust, which is essentially inert, primarily acts as a physical filler, enhancing density without significantly participating in hydration reactions, while a higher amount has an opposite effect. For concrete containing dolomitic limestone (Fig. 5b), the strength reaches 2.8 MPa with 60C30M10L, a 33% increase over the 100 C reference. This improvement is also attributed to the pozzolanic reactivity of metakaolin, which promotes the formation of secondary CSH products, and to the filler role of dolomitic limestone, which densifies the matrix. The 55C30M15L and 50C30M20L exhibit flexural strengths of 2.5 MPa and 2.7 MPa, respectively, confirming a limited effect of dolomitic limestone at higher substitution. The ratio of compressive to tensile strength ranged from 12 to 11 for concrete containing granite dust and from 12 to 6 for concrete containing dolomitic limestone dust. These results confirm good microstructural cohesion with granite, due to increased densification, reduced capillary pores, and likely tighter paste-aggregate interfacial zones. The comparison of fillers shows that granite dust is optimal at a moderate content of 10–15%, whereas dolomitic limestone content is limited to 10%, owing to the ability to act as a nucleation site for chemical reactivity and to fill the microstructure. Thus, ternary concrete combining metakaolin and granite dust or dolomitic limestone represents a mechanically efficient alternative to Portland cement concrete. Fig. 5Splitting tensile strength of concrete with (a) granite dust, (b) dolomitic limestone. Splitting tensile strength of concrete with (a) granite dust, (b) dolomitic limestone. Capillary absorption and water accessible porosity Figure 6 presents the evolution of the capillary absorption kinetics of concrete. The evolutions of capillary absorption and porosity confirm that the joint incorporation of metakaolin and granite or dolomitic limestone dust improves the transfer properties of the concrete. The reference concrete 100 C reaches the highest absorption kinetics within 24 h, a sorptivity of 0.023 kg/m2.h0.5, and a porosity of 7.13% (Table 2). Conversely, the concretes 60C30M10G, 55C30M15G, and 50C30M20G reach the lowest absorption kinetics (Fig. 6a), less than 0.015 kg/m2.h0.5 and porosity less than 4.44% (Table 2), corresponding to a reduction of 33% compared to the reference. However, concretes containing dolomitic limestone reach moderate improvement of capillary absorption kinetics (Fig. 6b), less than 0.022 kg/m2.h0.5 and porosity less than 5.96% (Table 2), corresponding to a reduction of only 27% compared to the reference, which is less efficient than those containing granite dust. These results confirm that metakaolin combined with granite dust yields a smaller, less connected pore network than with dolomitic, thereby reducing capillary water penetration. The decrease in porosity reduced capillary permeability, resulting in lower water absorption kinetics throughout the monitoring period. The pozzolanic effect of metakaolin contributes to the formation of secondary CSH gels, densifying the matrix, while granite dust acts as a filler, reducing pore connectivity. This was responsible for the densification and strength improvement of concrete (Sect. 3.2–3.4). Similar observations were reported by Malaiškienė, et al.36, who showed that the addition of metakaolin improves the density and reduces the porosity of high-performance concrete. Gao et al.37 also showed that metakaolin reduces sorptivity by refining the capillary structure and increasing the amount of secondary C-S-H. The concrete containing 20% filler exhibited slightly higher porosity and absorptions than those containing 10–15% filler, confirming a dilution effect of the active binder beyond the optimal threshold. Fig. 6Kinetics of capillary absorption of concrete with (a) granite dust, (b) dolomitic limestone. Kinetics of capillary absorption of concrete with (a) granite dust, (b) dolomitic limestone. Table 2Sorptivity from the capillary absorption and water accessible porosity of concrete.Design24-hour capillary sorptivity (kg/m2.h0.5)Correlation coefficient R 2Water accessible porosity (%)100 C0.0230.987.13 ± 0.3660C30M10G0.0150.884.05 ± 0.0655C30M15G0.0100.974.25 ± 0.1350C30M20G0.0110.994.44 ± 0.2860C30M10L0.0130.995.22 ± 0.0855C30M15L0.0180.995.23 ± 0.0950C30M20L0.0220.985.96 ± 0.25 Table 2Sorptivity from the capillary absorption and water accessible porosity of concrete.Design24-hour capillary sorptivity (kg/m2.h0.5)Correlation coefficient R 2Water accessible porosity (%)100 C0.0230.987.13 ± 0.3660C30M10G0.0150.884.05 ± 0.0655C30M15G0.0100.974.25 ± 0.1350C30M20G0.0110.994.44 ± 0.2860C30M10L0.0130.995.22 ± 0.0855C30M15L0.0180.995.23 ± 0.0950C30M20L0.0220.985.96 ± 0.25 Sorptivity from the capillary absorption and water accessible porosity of concrete. Resistance to sulfuric acid attack Figure 7 shows the evolution of the mass loss of different low-carbon concrete containing granite dust or dolomitic limestone after immersion in a sulfuric acid solution for 7, 14, and 28 days. The results showed that all the concrete exhibited a progressive increase in mass loss over time, due to the continuous degradation of the cement matrix during the exposure period. For concrete containing metakaolin and granite dust or dolomitic limestone, the mass losses are systematically lower than those of reference concrete 100 C. At 7 days, 60C30M10G, 55C30M15G, and 50C30M20G lost only 0.97%, 0.52%, and 0.76%, respectively (Fig. 7a), representing decreases of 58%, 77%, and 67% from the reference 100 C (2.27%). At 14 days, the mass losses respectively increased to only 2.85%, 2.19%, and 1.16%, reflecting decreases of 42%, 56%, and 76% from the reference 100 C (4.94%). At 28 days, the losses reached 5.09%, 4.94%, and 3.63%, representing a decrease of 49%, 51%, and 64% from the reference 100 C (10.07%). This shows that increasing the granite dust content from 10% to 20% reduces cumulative mass losses by more than 50% due to the filler effect and matrix densification, which impact resistance to acid attack. The concretes containing dolomitic limestone generally reach higher mass losses than those containing granite dust. At 7 days, 60C30M10L, 55C30M15L, and 50C30M20L (Fig. 7b) respectively lost 1.35%, 1.57%, and 0.78%, representing decreases of only 41%, 31%, and 66% compared to the reference 100 C (2.27%). At 14 days, the losses increased to 3.69%, 3.32%, and 1.87%, and at 28 days, they peaked at 7.10%, 6.77%, and 4.98%, reflecting relative decreases of only 29% to 51% compared with the reference 100 C (10.07%). This indicates that concrete containing dolomitic limestone is more sensitive to particle loss under acid attack than that of granite dust, explaining the higher mass losses despite the presence of metakaolin. However, metakaolin contributes to the formation of more stable CASH gels, partially limiting degradation18. Fig. 7Mass loss of concrete, after 28 days in sulfuric acid, with (a) granite dust, (b) dolomitic limestone. Mass loss of concrete, after 28 days in sulfuric acid, with (a) granite dust, (b) dolomitic limestone. Figure 8 shows the residual compressive strength, determined with respect to the remaining cross-sectional areas of each specimen after mass loss, of concrete after 28 days of immersion in a sulfuric acid solution. The reference concrete 100 C (Fig. 8a) exhibits significant strength degradation, with a residual strength of 16 MPa, corresponding to only 61% of the initial strength (26 MPa). This indicates its susceptibility to the acidic solution, which not only degrades the surface particles (mass loss) but also potentially ingresses through the water-accessible porosity, reducing the cohesion of its structure. The concrete 60C30M10G had a residual strength of 31 MPa, representing up to 87% of the reference (35 MPa). For the 55C30M15G (24 MPa) and 50C30M20G (26 MPa), the residual strengths were 83% and 100% of the initial values, respectively. These results confirm that a higher granite content results in lower porosity and less ingress of acidic solution, in which the matrix compactness and the pozzolanic effect of metakaolin combine to limit acid degradation17. Fig. 8Resistance to sulfuric acid attack, after 28 days of immersion of concrete with (a) granite dust, (b) dolomitic limestone. Resistance to sulfuric acid attack, after 28 days of immersion of concrete with (a) granite dust, (b) dolomitic limestone. For concrete containing dolomitic limestone (L), the residual strengths were generally lower (Fig. 8b) than those of the concrete with granite dust, although they remained better than the reference. The concrete 60C30M10L, 55C30M15L, and 50C30M20L reached 27 MPa, 24 MPa, and 20 MPa, respectively, representing 0%, + 14%, and 0% evolution from the initial values. These results indicate that although the incorporation of dolomitic limestone impacts more on the particle loss in acidic environment than granitite dust, it offers better residual strength compared to granite dust. This contradicting behavior can be explained by the microstructural evolution which needs to be assessed to fully apprehend the performances of these low carbon concrete. Although the dolomitic concretes resist acid better than the reference, they show higher particle loss and lower durability compared to granite concretes. This suggests that limestone’s filler effect is less effective in maintaining microstructural integrity under aggressive environments. Similar observations were reported on concrete after exposure to high temperature (650 °C) using granite and limestone as aggregate38. The synergy between the pozzolanic reactivity of metakaolin and the filler effect of granite dust creates a dense microstructure, reducing capillary porosity and acid ingress. Dolomitic limestone improves workability but introduces chemical vulnerability under acidic conditions, likely due to the formation of soluble products during attack. This highlights that durability performance (Fig. 8) strongly correlates with initial microstructural densification and strength (Figs. 2, 3 and 4), validating the importance of mix design optimization for aggressive environments. However, properly dedicated microstructural analyses (XRD, SEM) still need to be performed on cementitious pastes to confirm these trends and especially qualify and quantify the initial microstructural formation and residual after exposure to different environments. Environmental performances The analysis of CO₂ emissions reveals its progressive reduction with the increasing substitution of cement by mineral additions M and G (Fig. 9a) or M and L (Fig. 9b). The reference concrete 100 C emits up to 354 kg CO₂ eq./m³, whereas the alternative concrete, containing granite dust, 60C30M10G (282 kg CO₂ eq./m³), 55C30M15G (264 kg CO₂ eq./m³), and 50C30M20G 248 kg CO₂ eq./m³) reduced emissions by 20%, 25%, and 30%, respectively. Similarly, the incorporation of dolomitic limestone reduces emission, where the 60C30M10L (324 kg CO₂ eq./m³), 55C30M15L (304 kg CO₂ eq./m³), and 50C30M20L (291 kg CO₂ eq./m³)respectively reduced emissions by 16%, 20%, and 24%. Dolomitic limestone offers less reduction than granite dust because it was extracted from the quarry located a long distance (approximately 350 km) from the capital Ouagadougou, compared to granite, which is not only collected from around Ouagadougou (20 km) but also as a by-product from the production of crushed granite aggregates. This decrease is primarily attributed to the reduction in clinker content in the CEMI, which accounts for 90% (319.67/354.03)30 of the emissions in reference concrete, as indicated in the supplementary data (Appendix 4). The incorporation of CEMI by 30% metakaolin generates only 28.2 kg CO₂eq., whereas the same quantity of Portland cement would have generated 106.2 kg CO₂eq., resulting in a reduction of 78 kg CO₂eq. The incorporation of granite dust (10%, 15%, 20%) generates negligible emissions of CO2 (0.12, 0.18, 0.24 kg CO₂ eq.), while replacing equivalent proportions of cement that would have emitted approximately 35.4 kg, 53.1 kg, and 70.8 kg CO₂eq. This explains the gradual decrease in environmental impact up to 50C30M20G. Therefore, the incorporation of dolomitic limestone (10%, 15%, 20%) also follows a similar trend, but with low reductions. Fig. 9CO₂ emissions and environmental performance indicator of concrete with (a) granite dust, (b) dolomitic limestone. CO₂ emissions and environmental performance indicator of concrete with (a) granite dust, (b) dolomitic limestone. The environmental performance indicator was determined by the ratio of CO2 emissions of each concrete design to its compressive strength at 28 days, reflecting the carbon efficiency of the concrete and quantifying CO₂ emissions per unit of mechanical performance. It revealed that the 60C30M10G was the most efficient, emitting only 8 kg CO₂ eq./m³ per MPa of compressive strength. The indicators for 55C30M15G and 50C30M20G also showed comparable performance (≤ 10 kg CO₂ eq./m³/MPa). This indicator also allows to confirm that 50C30M10L is the optimum with limestone with 12 kg CO₂ eq./m³/MPa, beyond which it increases slightly. For practical applications, 60C30M10G, which can be extended up to 50C30M20G, represents the best compromise between reducing emissions and maintaining the mechanical performance of the C25/30. The current results reveal strong synergy between metakaolin (M) and granite powder (G), resulting in a denser and more durable composite than the metakaolin-dolomitic limestone (L) system. The improved performance in mixes containing G is not due to a single factor but results from a series of interconnected structural enhancements. An increase in ultrasonic pulse velocity (UPV) and apparent density in mixes like 55C30M15G directly corresponds with reduced capillary porosity and water absorption. This improved pore structure, resulting from the combined pozzolanic reaction of M and the physical filler effect of G, is the main driver of higher mechanical strength and better resistance to sulfuric acid attack. The acidic resistance simply reveals this durability, as mixes with the lowest porosity (and capillary absorption) had the least mass loss and retained the highest residual strength. Conversely, L-based concrete, while improving workability, exhibited only modest densification (less UPV and density increase) and was more susceptible to acid dissolution and particle loss, but sustained better residual strength. This demonstrates that the chemical stability of the filler and its particle packing efficiency are just as important as pozzolan reactivity for long-term performance. These results strongly support and expand upon principles from previous studies on low-carbon cementitious systems. The filler effect and particle packing, central to the performance of granite powder, are fundamental to the performance of LC³ (Limestone Calcined Clay Cement) technology39,40. The significant improvement in strength and durability from metakaolin aligns with studies that emphasize its role in enhancing microstructure and pore networks35. The non-linear response to substitution levels, with an optimal point beyond which a “dilution effect” occurs (60C30M10G and 60C30M10L), is well-known in ternary and quaternary mixes41,42. Furthermore, environmental performance results indicate that cement replacement is the primary method for reducing CO₂ emissions, aligning with life-cycle assessment approaches43,44 and decarbonization strategies for concrete45. The high performance of the optimum mixes in reducing carbon footprint (kg CO₂/MPa) supports the focus on functional units that connect environmental impact with mechanical strength46,47. This study affirms the global LC3 approach while highlighting a specific, practical application using Burkinabe granite powder and dolomitic limestone as effective, local supplementary cementitious materials. Therefore, this study contributes to the regional decarbonization of the construction sector by valorizing locally available mineral resources and industrial by-products, aligning with circular economy principles and SDG 9 (Industry, Innovation and Infrastructure) and SDG 13 (Climate Action). Conclusions and perspectives This study aimed to assess the feasibility of a low-carbon concrete in the context of Burkina Faso specifically and west African region as whole by evaluating the physical, mechanical, and durability performance of a concrete containing down to 50% of Portland cement and a mixture of 30% metakaolin (M) and up to 20% granite (G) or dolomitic limestone (L) powder. At 28 days, the 60C30M10G reached 35 MPa, compared to 26 MPa for reference 100 C, representing a 35% increase. Meanwhile, 60C30M10L only reached 26 MPa (+ 0%), indicating a better reactivity-filler synergy with G than with L. In splitting tension, 60C30M10G and 50C30M10L both reached 2.8 MPa, compared to 2.1 MPa for 100 C (+ 33%). In terms of durability indicators, water absorption was reduced by 33% with 60C30M10G and by 27% with 60C30M10L, reflecting a more compact microstructure, than the reference 100 C. When exposed to H₂SO₄ solution, mass losses were reduced by 64% with G and 51% with L, illustrating better resistance to acid attack. This confirmed the beneficial synergy between the pozzolanic reactivity of metakaolin and the filler effects of granite and limestone dust. It also shows the possibility of designing 50C30M20G concrete that maintains similar mechanical performance (26 MPa) and better durability indicators than reference concrete. The designed concrete offers notable mechanical performance and durability while contributing to a 30% reduction in CO₂ emissions. It also emits ≤ 10 kg CO₂ eq./m³/MPa compared to 14 kg CO₂ eq./m³/MPa for the reference. These conclusions demonstrate the potential to produce low-carbon concrete, aligning with circular-economy principles and supporting climate-resilient construction in Burkina Faso and across West Africa. The present study, therefore, supports the achievement of environmental and socio-economic goals. This exploratory study assessed several substitution contents, maintaining a constant water/binder ratio and limiting optimization, which will be investigated in the future. Durability, evaluated over short-term durations (28 days), should also be tested in the long term and across multiple environments (different acidic environments, high temperature,…). Microstructural analyses should also be performed to confirm reaction mechanisms, especially on cement pastes containing dolomitic limestone to assess the reaction mechanisms and formed products as it seems to induce a detrimental effect. Addressing these constraints will allow a better understanding of the performance, mechanisms, and sustainability potential of locally produced low-carbon LC3-like cement and concrete, and especially adding value to granite by-products for the circular-economy. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1