Each environmental impact category was individually discussed, followed by the top three contributing processes in each ground improvement technique. The sensitivity and uncertainty analyses of selected variables were then introduced. Figures and values are reported per the functional unit defined in this study.
3.1. Global warming potential
The contribution of CO2, N2O, and CH4 emissions in the global warming potential (GWP) of each soil improvement technique is presented in Fig. 2. Carbon dioxide was the highest contributor to GWP in both techniques, accounting for 97 and 70% of the total GWP in the case of using PC and EICP, respectively (Fig. 2a). The CO2 emissions were significantly greater than the carbon equivalent of the N2O and CH4 emissions, although the GWP indices of CH4 and N2O are 28 and 298 times more than CO2, respectively, over a 100-year time horizon. Using PC for soil stabilization resulted in an estimated total GWP of 255 tons of CO2-eq; this was reduced to 247 kg CO2-eq with the EICP application, i.e., a 3% reduction in GWP. This slight reduction can be partially attributed to utilizing carbon dioxide in urea production plants which results in a positive impact on GWP (Armstrong and Styring, 2015).
To study the highest contributors to GWP in both techniques, the top three processes that affected GWP were plotted in Fig. 2b. The majority of PC production emissions was from clinker (86%) with 219 tons of CO2-eq, followed by electricity usage (7%), and hard coal operation and preparation (1.8%). The reason behind the huge GHG emissions of clinker is the massive amount of energy required to heat the mixture to 1450 Co (Zapata and Bosch, 2009). On the other hand, the highest GWP contributors from EICP were ammonia production (during urea production) and onsite emissions, with 31 and 20% of the total GWP, respectively. These results are in agreement with Raymond et al. (2021), in which process emissions were found to be the highest contributor, followed by onsite EICP emissions. In addition, in the urea production process generation of ammonia gas (gasification) consumes 60 to 70% of the total supplied energy (Shi et al. 2020; Zhou et al. 2010). Another key GWP contributor was the non-fat milk powder used as an additive to improve the EICP cementation efficiency, with ~ 15% of the total GWP of the EICP process.
3.2. Acidification potential
Acidification potential (AP) is particularly affected by processes involving SOx, NH3, and NOx emissions (Adghim et al. 2020). As shown in Fig. 3a, such emissions were present with different amounts in both improvement techniques. Although PC production processes were found to largely contribute to AP (Kim et al. 2016), the present results revealed that PC had AP of 517 kg SO2-eq which is 57.7% less than that of EICP. The major contributor is the ammonia emissions from EICP, which equaled 687.2 kg SO2-eq compared to 21.0 kg SO2-eq from PC. On the other hand, EICP produced 50% fewer nitrogen oxides and 40% more sulfur oxides compared to PC. The main contributor to AP in PC was from clinker production, with 60% of total AP from PC. As shown in Fig. 3b, urea production emissions, grass planting at dairy farms, and non-fat milk production had 17.8, 17.7, and 15.5%, respectively, of total AP from EICP. Grass at dairy farms had a high impact on AP due to the usage of fertilizers. In addition to its nitrogen oxide and ammonia emissions to air, grass farming results in direct heavy metal discharges into water ecosystems as stated in Agri-footprint 5.0 (van Paassen et al. 2019).
3.3. Eutrophication potential
Eutrophication potential (EP) is caused by nutrients loadings of mainly nitrogenous and phosphorus compounds, in soil, water, or air that cause rapid algal growth (Kim et al. 2016). As shown in Fig. 4a, the EICP technique had several significant EP-related emissions in water, air, and soil. The total ammonia emissions in water, air, and soil from EICP production and reactions by-product onsite added up to 843 kg-PO4, i.e., 72% of the total EP of EICP. In contrast, PC production and application produced 86% lower EP compared to EICP. This percentage could be decreased by 22.3% by controlling the EICP emissions on-site, as they contribute to around 63.7% of total EICP EP. On-site EICP emissions were the most significant contributor to EP, which is in line with the findings of Raymond et al. (2021). As shown in Fig. 4b, the emissions of PC production were mostly due to spoils from coal mining with 64% of the total EP of PC. Coal spoils are acidic and contain metal contamination that can leach to ecosystems due to their low water holding capacity (Ghosh and Maiti, 2020).
3.4. Marine aquatic ecotoxicity potential
Marine aquatic ecotoxicity potential (MAETP) is defined as the impact on organisms in seawater due to toxic substances emitted to ecosystems. As shown in Fig. 5a, the EICP had nearly twice the impact on MAETP compared to PC. Beryllium discharged in water was found to be the highest contributor to MAETP in the EICP and PC techniques with 39.8 and 46.1%, followed by hydrogen fluoride emissions to air (22.7 and 20.2%) of the total, respectively. The main contributing process for such high emissions in EICP was the sulfidic tailings (~ 30.4% of EICP MAETP) resulting from mining sulfidic minerals. The sulfidic tailings are one of the worst environmental impacts to the mining industry and have been considered as the largest environmental liability of the mining industry (Nehdi and Tariq, 2007). On the other hand, similar to EP, the largest contributor to MAETP in PC production was the spoils from coal mining with a total contribution of 26.5%, as shown in Fig. 5b.
3.5. Abiotic depletion potential
Abiotic depletion potential (ADP) is the depletion of resources from the non-organic, non-living materials, e.g., air, land, freshwater (Rasul and Arutla, 2020). As shown in Fig. 6a, EICP outperformed the PC, with nearly 90% less ADP. The most contributing process was the co-production of lime in zinc mine operation, which accounts for 98% of total ADP for PC production as shown in Fig. 6b. Previous studies have discussed the significant amount of lime co-produced with zinc concentrate in mining and beneficiation processes (Genderen et al. 2016). On the other hand, the zinc concentrate production processes from mining operation resulted in the highest EICP contribution to ADP, where 90% of this ADP was from the production of urea.
3.6. Impacts of external processes
The external processes refer to onsite operations and transportation of final products from local suppliers to site location. Overall, all materials were available locally near the study area. Using the PC in the soil improvement involved more transported weights, compared to EICP which was assumed to have longer travel distances. In both cases, the ozone layer depletion potential (ODP) was the most affected environmental impact category, with a 16% higher impact from PC compared to EICP despite shorter distances assumed in the PC case. The next two categories that were severely impacted by external processes were the AP and photochemical ozone creation potential (POCP). PC transportation had a higher impact on AP and POCP by 21.7 and 28.2%, respectively, compared to EICP. The highest difference between PC and EICP was found in GWP, where the external processes of PC produced 71.1% higher GWP compared to those of EICP. The main reason for the lower impact of EICP in overall external processes is the lighter weights of EICP constituents. The total raw materials weight transported in the case of the EICP ground improvement technique was nearly 1/3 compared to those in the PC case.
3.7. Sensitivity analysis
The different sensitivity analysis scenarios were proposed to evaluate the relative weight of each of the hotspots in the environmental impact of the EICP process. As shown in Fig. 7, the most affected environmental impact categories by those changes were the GWP and EP, respectively. The GWP has decreased by 13.4% when the lowest emissions were assumed, whereas applying the highest emissions increased the overall GWP by 38.5% compared to baseline, as shown in Fig. 7a. The high GWP in the case of the highest EICP emissions would be 34.2% higher than that of the PC technique. In the no emissions and waste non-fat milk scenario, the GWP decreased by 38.5% compared to the PC scenario. Furthermore, EP has significantly increased in the highest emission scenario (Fig. 7b); it nearly doubled the EICP baseline scenario, which is 7.2-fold greater than the case of PC treated soils. In contrast, in the no emissions scenario adopting waste non-fat milk, the EICP impacts on EP would decrease to be similar to that of PC. Overall, the sensitivity analysis corroborates the high effect of EICP onsite emissions and non-fat milk on the GWP and EP; reducing those emissions would favor EICP over PC. In terms of AP, using waste non-fat milk would reduce the AP of EICP by 38.1%, which will make the EICP 49% higher than PC in AP compared to 140% higher in the case of using fresh non-fat milk. The individual contribution of non-fat milk was identified to be 39 tons CO2-eq in GWP, 249 kg PO4-eq in EP and 465 kg SO2-eq in AP.
3.8. Limitations and recommendations
Similar to all LCA studies, the findings of this assessment have to be carefully interpreted taking into consideration the various assumptions and project-specific conditions. To enable proper cross-comparison of the study outcomes, the following limitations and considerations must be accounted for:
This study was based on a specific EICP mix; addition, removal, or altering the ratios of constituents may change relevant outcomes. Besides, the FU defined in this study stated a specific performance, i.e., average UCS of treated soil up to 1.5 MPa in a 2-week period over one cycle. A different FU would significantly change the proportions of constituents required for both techniques.
The lifecycle inventory analysis was computed in this study considering the sub-tropical arid climate in Dubai. Changing the project geographic location would vary certain inputs, e.g., emission factors, possibly leading to different outcomes.
The scarcity of data in the literature, particularly those related to the field onsite emissions of EICP, has led to adopting the IPCC recommendations. The uncertainty analysis of EICP emissions showed a substantial impact on GWP and EP. To avoid this critical assumption, laboratory and field measurements for those emissions and their leaching rates are highly recommended for future studies.
Although the amounts of EICP constituents were significantly less than PC, the costs associated with the implementation of EICP would be substantially higher than those of the conventional PC technique. This unmatched benefit of PC is due to the numerous technological enhancements and cost optimization achieved in the cement industry for decades. As the EICP technique becomes gradually commercialized, its economic performance would eventually improve compared to PC.
Based on the environmental hotspots indicated in this study, it is clear that the overall performance of the EICP soil improvement technique can be significantly improved by applying sustainability and mitigation measures to reduce the emissions of selected sub-processes, as follows:
Urea production was one of the key environmental hotspots in the EICP technique. Several techniques could decrease the energy consumed by the gasification process in the urea production, e.g., heat recovery of primary reformer in the natural gas reforming process (Wang et al. 2006) which employs highly efficient catalyst to reduce steam use in gasification.
The contribution of onsite emissions was quite high in the overall EICP performance. This is because of high levels of nitrogen dioxide and ammonia gas produced during the hydrolysis process of urea. Studies in the field of reducing nitrous oxide and ammonia emission from the EICP technique are scarce. Cheng et al. (2019) investigated the atmospheric ammonia produced from “low-pH treated” microbial induced calcium precipitation (MICP), which is a hydrolysis process similar to EICP. A 90% reduction of ammonia was achieved by reducing the initial pH of the solution which holds the ammonia ions in liquid form.
The onsite emissions were found to be the second contributor to the GWP in the case of EICP treated soils. Several studies had focused on the removal of ammonium from soils. For example, Wang et al. (2019) have shown that using electro-kinetics was effective in electricizing ammonia in soils which significantly reduces the EP. However, the electricizing technique still needs additional studies on large-scale field studies to improve its applicability.
The non-fat milk powder contributed to ~ 15% of the total GWP from the EICP processes. Reduction of GWP in dairy farms could be achieved by adopting sustainable manure management techniques, such as anaerobic digestion that could reduce the GWP of dairy farms up to 25% compared to conventional techniques (Adghim et al. 2020). The use of expired milk as a raw material in the EICP cementing may reduce the cost and improve the environmental sustainability of the EICP technique. From another perspective, using non-fat milk in EICP can be fully eliminated. Cui et al. (2020) achieved ~ 1.5 MPa compressive strength of EICP treated specimens (similar to the present FU) without using milk, however, in order to achieve such strength, the researchers applied three cycles of the EICP solution.
In the EICP technique, four components are manufactured in different factories and agro-industrial processes, which is not the case with centralized cement production. Having multiple entities producing various components to be combined into one product would reduce the overall environmental efficiency of the system (Gatimbu et al. 2020); this is on top of the additional transportation-related environmental burdens.
The potentially improved performance of EICP treated soils has not been considered in this study. EICP treated soils were proven to perform well under certain conditions such as sulfate contamination (Arab et al. 2021). Comparative laboratory experiments on EICP and PC-treated soils under harsh environmental conditions, e.g., heavy metal leaching, freeze and thaw cycles, wetting and drying cycles, and sulfate contamination, are required to assess the effect of durability on the LCA analysis.