The raw materials used in this study are presented in Figure 1 below.

The laterite comes from the eastern region of Burkina Faso, in the city of Fada N’Gourma at the GPS coordinates N12.10020 E0.34710.

The Shea Butter residue (SBr) was collected at the Fédération NUNUNA (FN) production center at the coordinates (11˚6'30.64"N; 2˚5'49.40"O) located in central-western region of Burkina Faso, exactly in the city of Léo. Shea is a tree native to the savannahs of Africa and in 2011, Burkina Faso became the world’s 2^nd producer of shea kernels (338,000 tons) and the world’s 3rd exporter of shea butter [34] .

The hydrated lime (HD) used in the study is an industrial by-product obtained from the production of acetylene by Burkina Industrial Gas (BIG), a company based in Ouagadougou the Kossodo industrial zone at GPS coordinates (N12˚25.935', W001˚29.374') and specialized in the manufacture and distribution of industrial gases.

The laterite soil, the Hydrated Lime (HD) and the Shea butter solid residue (SBSr) were sieved through respectively a 5 mm, 0.8 mm and 0.4 mm sieve ( Figure 1(a) , Figure 1(b) , Figure 1(d) ) to accommodate their granulometry within the spindle recommended by standard CRATerre.

For the formulation and manufacture of CEB, the main raw earth materials must comply with some characteristics, in a particular class, which must be included in the grading range, with normative values enabling the correct property to be assessed.

The laterite soil size distribution was performed using two methods. The coarser fraction (>80 µm) was analyzed by wet sieving and the finer fraction (<80 µm) by sedimentation methods according to standards NF EN 933-1 [35] and NF EN ISO 17892-4 [36] . Because the soil particle size distribution is a fundamental step in evaluating the suitability of the soil for earth construction [4] . The Atterberg’s limits were determined according to standard NF P 94-051 and the methylene blue value was determined according to standard NF P 11-300-GTR [37] .

The laterite soil and the SBr specific weights by air pycnometer test were determined using NF EN 1097-7 standard [38] . The standard Proctor compaction test was carried out on untreated soil, HL stabilized specimens and the specimens with different percentages of SBr for the optimum moisture content (OMC) and Maximum dry density (MDD) under standard NF P 94-093 [39] .

To determine the SBr fat content, we used the Wolff et Castera-Rossignol modified method for liquid sample [40] and the Soxhlet extraction method using hexane as the solvent for solid samples.

To make the CEB, we opted for a formulation in which the proportion of earth matrix added to that of stabilizer gives the final mass of the manufactured sample. Previous studies on HL have used mass contents ranging from 0% - 25% [41] - [43] ; on SBSr contents ranging from 3% - 10% [23] on SBLr 25% - 100% [44] in the earth material stabilization. Based on the above, five (05) formulations with stabilization rates by weight ( Table 1 ) were selected for the present study.

For this purpose, raw materials quantities were evaluated to enable the production of 120 CEBs, including 24 CEBs per formulation, as summarized in Table 2 below.

The materials are first sieved by hand through various sieves. They are then mixed wet and dry and inserted into the mold (29.5 cm × 14 cm × 9.5 cm) for compaction and followed by demolding ( Figure 2 ). Finally, the CEB specimens were hermetically sealed in a plastic bag for 45 days of curing. The curing is one of the parameters that affects the mechanical performance of compressed earth blocks (CEBs) stabilized with calcium carbide residue [45] the oven-dried at 60˚C to a variation of 0.1% in mass for characterization.

The characterizations were carried out on at least three specimens of CEBs for the consideration of average and standard deviation values.

The water-accessible porosity can be used to access the CEB’s permeability and mechanical strength. The principle is to determine the dry CEBs mass (Md in kg), the mass of water saturated CEBs after 24 h immersion respectively in water (Msat.wt in kg) and in air (Msat.air in kg), and apparent volume by hydrostatic weighing ( Figure 3 ). The specimen for testing is kept in water at 20˚C to be saturated and then placed in the oven at 60˚C until a constant mass is obtained. Water-accessible porosity (Pa), also used as an indicator of block durability, is obtained by applying the following Equation (1). The bulk density (ρ_b (kg/m³)) of dry CEBs of mass, Md (kg) is determined using Equation (2) after hydraulic weighing [46] . The compressive strength ( Figure 4 ) was tested on the stack of two halves of CEBs, in dry and wet conditions after immersion in water for 2 h, using a hydraulic press (Proeti safr, Madrid, Spain) equipped with a 300 kN capacity load cell at a loading rate of 0.2 mm/s, referring to XP P13-901 standard [47] . The CEB’s compressive strength, Rc (MPa), was calculated using Equation (3), where Fr (kN) is the maximum load at failure and S (cm²) is the area of applied surface:

$P_a\left(\%\right)=\frac{M_{sat.air}-M_d}{M_{sat.air}-M_{sat.wt}}\times 100$(1)

${\rho }_b=\frac{M_d\times {\rho }_{eau}}{M_{sat.air}-M_{sat.wt}}$(2)

$R_c=10\times F_r/S$(3)

Water absorption by capillary action was measured according to standard NF XP 13-901 [47] . The principle is to partialize the blocks to a depth of 5 mm and measure their mass for 24 hours. The CEBs capillary absorption coefficient (C_ac in g/cm²) is expressed by Equation (4) where M_h(g) and M_d(g) are respectively the CEBs humid mass and dry mass and S (m²) the specimen surface:

$C_{ac}=\frac{M_h-M_d}{S}$(4)

Absorption by total immersion was carried out following standard NF EN 14617-1 [48] on normalized dimensions of CEB (29.5 × 14.5 × 9.5 cm³). The absorbed water amount (H_p in %) by the CEBs was determined by Equation (5) below:

$H_p=\frac{M_h-M_d}{M_d}\times 100$(5)

The purpose of the abrasion resistance test is to simulate the CEBs’ behavior concerning the various erosions eventually caused by human activities or wind. This test is carried out using a steel wire brush loaded with 3 kg to simulate these effects. Brushing is applied on the face of the CEB and along its entire length, at the rate of one return stroke per second for one minute (i.e. 60 return strokes) without applying any vertical force to the brush during handling. These effects are simulated and measured and the abrasion coefficient (Cab in cm²/g) is conventionally determined by Equation (6) where M₀(g) and M_f(g) are respectively the initial and final CEBs mass. The erodibility test presented in Figure 5 was carried out following the standard NZS 4297:1998 [49] . Samples are subjected to a constant water pressure of 50 kPa for one hour ( Figure 6 ). The tested CEBs were placed at a distance of 470 mm and then the depth of water penetration is measured.

$C_{ab}=\frac{S}{M_0-M_f}$(6)

The thermal properties, such as conductivity, λ (W/m·K), diffusivity, a (m²/s), specific thermal heat (J/˚C·kg) and specific thermal mass (kJ/kg·K), were measured on dry samples by using the device name “KD2 Pro Thermal Properties Analyzer” ( Figure 7 ) Each test take around 5 minutes and when the progress bar has completely darkened, the results are displayed on device screen.

Figure 8 shows the results of laterite soil (LS) particle size analysis and the spindle recommended by the ARS 680 [50] . Theses analyses show that the used LS contains a particle size fraction of around 30% gravel, 40% sand, 15% silt and 10% clay. We also noted that the LS particle curve is within the range recommended by ARS 680, which defines the appropriate particle size for CEB design and production.

The raw laterite material studied has a liquidity limit of 41.19%, a plasticity limit of 24.2% and a plasticity index of 17%. These properties are close to those of the Morinda soil with a plasticity index of 17.4%, and are already used as a matrix for CEB stabilization with industrial and agricultural waste [16] . Antonelli and Azambuja (2021) mentioned that soils with plasticity indexes between 15% and 30% have a stabilization success rate of 69% [4] . With a plasticity index from 10% - 20% the stabilization with cement is preferable, and according to standard CRATerre the higher the IP, the higher the material strength [51] . The studied raw LS is within the plasticity diagram recommended by standard ARS 680 ( Figure 9 ) and constitutes an appropriate raw material for construction. In addition, the methylene blue value (VBS = 0.68) is range [0.2 - 1.5], so the studied sample is a sandy-loam soil sensitive to water according to the standard GTR 92.

The raw LS specific weight results gave for raw LS 28.345 ± 0.36 kN/m³, so around double that of Shea butter solid residue (SBSr) which is 14.8 ± 0.57 kN/m³ and for Hydrated lime (HL) 27.15 ± 0.64 kN/m³. This means that LS has a higher solid grain density than SBSr. Table 3 shows the Optimum Moisture Contents (OMC) and the Maximum Dry Density (MDD) of the different composites formulated. The OMC ranges from 19.5% to 26.3% and the MDD from 1.79 g/cm³ to 1.31 g/cm³ depending on the type and the rate of stabilization. We noted that the addition of SBr, especially the SBSr results in lower CEBs densities and higher water requirements for the different formulations. This could be explained by the SBr low density compared to raw LS.

The fat content of the Shea Butter residues (SBr) sampled was 0.461 g/L for Shea butter solid residues (SBSr) and 0.586 g/L for Shea butter liquid residue (SBLr).

The physical properties as well as mechanical strength were affected by the stabilization with by-product binders such as HL and SBr. However, the stabilized CEBs with 25 wt% SBSr were friable and completely degraded in contact with water; so, it was impossible to quantify their water-accessible porosity, due to their high sensitivity to water.

The addition of 5 - 25 wt% SBr decreases the bulk density of CEB in the range of 1580 - 1370 kg/m³ following the increase of the porosity accessible in the range of 31.58% - 41.47% ( Figure 10 ). The increased accessible porosity with the addition of SRSr in the range of 8.04% - 31.31% can be related to the OMC for the production of stabilized CEB and the organic nature of the stabilizer. The partial substitution of 25% LS by the SBLr slightly decreased the bulk density from 1580 to 1540 kg/m³ and the accessible porosity from 31.58% to 32.53% (i.e. 3.01%). This indicates that the SBSr affect strongly the stabilized CEB densities and the accessible porosity than the SBLr. Moreover, the stabilized CEB with HL is less porous (31.58%) and denser (1580 Kg/m³) than the HL and SBr combined stabilized CEB. The results of stabilized CEBs with HL could be mainly due to the good cohesion and the different physical and chemical reactions between the LS and the HL stabilizer [52] . These results corroborate those of Zoma et al. (2020) [12] . The low density and high-water-accessible porosity observed with the addition of SBr could be explained by the poor cohesion between LS and SBr and also the formation of lumps during the mixing with the water added.

Figure 11 details the evolution of compressive strength in dry and wet conditions of CEBs stabilized with HL and SBr. The average dry compressive strength significantly decreased by the addition of SRSr from 3.26 MPa (Stabilized CEBs with HL) to 0.3 MPa (25 wt% SBSr) and 2.3 MPa (25 wt% SBLr). Additionally, the wet compressive strength of stabilized CEBs also decreased with the addition of SBr from 0.75 MPa (Stabilized CEBs with HL), to 0.10 MPa (10 wt% SBSr) and 0.60 MPa (25 wt% SBLr). The wet compressive strength of stabilized CEBs with 25 wt% SBRs could not be determined as they immediately degraded in water.

For the stabilized CEBs with 5 wt% HL, the optimum values obtained may due to the creation of bonds between the earth particles and the ions in solution during the hydration of the hydrated lime. These results are in agreement with 1.33 MPa [53] and 1.48 MPa [54] .

The sharp decrease in the strength of CEBs with 5 - 25 wt% SBSr and the slight decrease in strength of stabilized CEBs with 25 wt% BLr could be due to their high-water accessible porosity and low density.

However, these results present similar dry compressive strength with previous studies using 3 - 10 wt% SBSr [23] . According to previous studies, most standards recommended the minimum dry compressive strength of 2 MPa for CEBs-based construction. According to African Standards ARS 674 [55] , the stabilized CEBs with 5 wt% HL and stabilized CEBs with 25 wt% SBLr with dry compressive strength in the range of 2 - 4 MPa, could be used for non-load bearing walls. However, the other stabilized CEBs with dry compressive strength of less than 2 MPa did not meet this standard for each construction and should be excluded.

The construction materials are used for structural or envelope components in one or multiple-story buildings, taking into account their structural efficiency. The coefficient of structural efficiency (CSE) is an important physical-mechanical parameter to assess the contribution of the strength and density of CEBs toward the load-bearing capacity in building construction [18] . The CSE of the stabilized CEBs is obtained by the ratio between the dry compressive strength and the bulk density. The stabilization with SBr did not  improve the CSE which decreased by 34.84%, i.e. from 2063.29 Pa·m³/kg (J/kg) for CEBs stabilized with 5% HL to 718.95 Pa·m³/kg (J/kg) for CEBs stabilized with 5 - 25 wt% SBSr (i.e. 52.5%) and 1493.506 Pa·m³/kg (J/kg) for CEBs stabilized with 25 wt% SBLr (i.e. 27.6%). This indicates also that the shea butter liquid residue had more effect than shea butter solid residue on CEBs structural efficiency by around 1.9 times. Finally, this suggests that the decrease in bulk density has negative effect on the structural efficiency of stabilized CEBs.

Figure 12 shows the evolution of abrasion coefficients ( Figure 12(a) ) and the mass loss ( Figure 12(b) ) of stabilized CEBs with HL and SBr. The abrasion coefficient (Ca) ranges from 14.49 cm²/g - 3.9 cm²/g depending on the addition rate of SBr and is 13.1 cm²/g for stabilized CEBs with 25 wt% SBLr versus a value of 13.03 cm²/g for stabilized CEBs with 5 wt% HL. The highest Ca (14.49 cm²/g) is obtained on the stabilized CEBs with 5 wt% HM and the lowest Ca (3.9 cm²/g) for stabilized CEBs with 25 wt% SBSr. We noted that the addition of SBr improved the Abrasion resistance of the stabilized CEBs with 10 - 25 wt% SRr. According to standard XP P13-901 concerning the normative value of Ca, the stabilized CEBs with (5 wt% HL, 5 wt% - 10 wt% SBSr and 5 wt% HL - 25 wt% SBLr with Ca > 7 cm²/g, are classed as CEB 60 used for external wall [47] . However, the stabilized CEBs with 25 wt% SBLr which Ca > 2 cm²/g are classed CEB 20 used for internal walls.

Furthermore, the mass loss of studied stabilized CEBs ranges from 0.45% to 0.08% depending on the type and the rate of stabilizer. The stabilized CEBs with 5 wt% HL - 25 wt% SBSr have the highest mass loss while the stabilized CEBs with 5 wt% HL - 5 wt% SBSr have the lowest, so the mass loss is inversely proportional to SBr rate in the mix. From these results, the mass loss is less than 10%, and then all the stabilized CEBs reach the recommendation limits of the standard ARS 675 [56] .

Figure 13 shows the evolution of erodibility of stabilized CEBs, in particular depth ( Figure 13(a) ) and eroded surface ( Figure 13(b) ). We noted that the stabilized CEBs with 5 wt% HL-10 wt% SBSr, 5 wt% HL, and 5 wt% HL - 25 wt% SBLr have the greatest depths respectively 4.37 mm/h; 3.25 mm/h and 3.12 mm/h versus the lowest value of 2.88 mm/h for stabilized CEBs with 5 wt% - 5wt% SBSr. The stabilized CEBs 5 wt% - 25 wt% SBSr could not be tested due to their sensitivity to water. The literature, previous studies of stabilized CEB with 7 wt% cement, 5/7 wt% cement/lime and 1 wt% fibers found respectively eroded depth of 1 mm/h, 20 mm/h and 55 mm/h [57] . According to standard NZS 4298 [58] , the limit value of erosion depths must be less than 120 mm/h [42] . Consequently, the studied stabilized CEBs can be classified as non-erodible blocks except the stabilized CEBs with 5 wt% HL - 25 wt% SBSr.

On the other hand, the eroded surface areas are 7.42; 10.72 and 8.58 cm² respectively for the various stabilized CEBs with 5 wt% HL - 10 wt% SRSr, and 5 wt% HL - 25 wt% SBLr, versus stabilized CEBs with 5 wt% HL (9.12 cm²). These results testify the addition of SBr increases the stabilized CEBs erodibility. Despite this finding, the stabilized CEBs with 5 wt% HL - 10 wt% SRSr have higher eroded depth and surfaces (i.e. 17.5% more than stabilized CEBs with 5 wt% HL). So, the rate of stabilizer SBr influences the stabilized CEBs erodibility resistance. This could be explained by the poor cohesion between LS-HL-SBr, and also the rougher surfaces of these stabilized CEBs and the areas of brittleness displayed during their manufacture.

Figure 14 shows the evolution of stabilized CEBs capillary absorption as a function of the square of time. We noted that the absorption varies according to the nature and the rate of the stabilizer. The slope, line gave us the sorptivity, which represents the rate of water absorption by capillary. Sorptivities range from 0.046 - 0.097 g/cm²·min^1/2 for stabilized CEB with 5% HL - 5 - 25 wt% SBr, versus 0.077 g/cm²·min^1/2 stabilized CEB with 5% HL. The low sorptivity is attributed to stabilized CEBs with 5% HL - 5 wt% SBr (0.046 g/cm²·min^1/2) and the greatest for stabilized CEB with 5% HL - 10 wt% SBr (0.097 g/cm²·min^1/2. This CEB behavior is probably linked to the water-repellent (moisture-preserving) properties of Shea butter residue. Thus, it is the presence of flat in shea butter that makes stabilized CEB more or less impermeable [41] . The results of durability studies on geomaterials with shea decoction have shown similar trends [44] .

Figure 15 shows the stabilized CEB water absorption by total immersion after 2 hours and 24 hours. It can be seen that the stabilized CEB with 5 wt% - 5 wt% SBSr have lower water absorption of 4.06% and 11.94 % versus 6.26 et 24.52% for stabilized CEB with 5 wt% - 10 wt% SBSr. However, the ratio Ab_2H/Ab_24H of stabilized CEB varies from 0.26 - 0.46. Consequently, more than 26% - 46% of the porous microstructure is highly water accessible over a short period. Izemmouren et al. (2019) carried out studies on stabilized CEB with 6 - 10 wt% lime and presented a water absorption by total immersion from 10.1% - 13.5% [59] . Moreover, similar values (13.6% - 16.5%) were observed for stabilized CEB with 8 wt% cement and 4/4 wt% cement/lime by Bogas et al. (2019) [60] .

The thermal properties of CEBs were also improved by the stabilization with by-product Shea butter residues. Figure 16 shows the stabilized CEB thermal parameters results.

The average value decreased in the range of 0.492 to 0.233 W/m·K, for thermal conductivity (i.e. 52.64%), 0.217 to 0.131 mm²/s, for thermal diffusivity (i.e. 39.63%), 2.567 to 1.774 MJ/m³·K, for specific heat capacity (i.e. 30.89%), 1.537 to 1.185 kJ/kg·K for specific thermal mass (i.e. 22.91%), measured on CEB stabilized with 0 - 25 wt% SBr. These results indicated that the CEB stabilized with HL/SBr has excellent thermal parameters than CEB stabilized only with HL, so the Shea butter residue improve the CEB thermal properties.

The thermal properties change with the rate of HL in the mix and their values decrease inversely proportionally to the HL rate. This could be due to these stabilized CEB low density and the porosity. The effect on the agricultural by-products gave similar conclusion on CEB stabilized with (1 wt% fiber + 1 wt% lime) and (1 wt% fiber + 3 wt% lime) with the respective value of 0.80% ± 0.30% W/m·K and 0.69% ± 1.73% W/m·K for thermal conductivity and 708.97% ± 0.24% and 876.29% ± 0.60% J/˚C·kg for specific heat capacity [12] . It’s the case of similar values of thermal conductivity range from (0.667 to 0.798 W/m·K), thermal diffusivity range from (2.24 × 10⁻⁷ m²/s to 3.055 × 10⁻⁷ m²/s) and the specific thermal mass (1.508 to 1.584 kJ/kg·K) obtained on CEB stabilized with 3 - 10 wt% SBSr [23] .