All LCA results are presented using the normalized environmental impact value, therefore, the unit is Average European Citizen Equivalent (AECE).
3.1 Life Cycle Assessment results: Model 1
Figure 3a displays the normalized results of six considered ICs. The results from Model 1 show that by maximizing the use of the recovered energy as high pressure steam, Indaver/Sleco’s WtE plant can avoid an annual environmental impact of approximately 21200, 36800, 6700, 15800, 37000 and 6900 AECE in the ICs ‘climate change’, ‘freshwater and terrestrial acidification’, ‘freshwater eutrophication, ‘photochemical ozone creation’ and ‘respiratory effects, inorganics’, ‘terrestrial eutrophication’, respectively.
The results clearly illustrate the significant variation in normalized impact values between the four scenarios in the IC ‘climate change’. Causative factors of this variation will be discussed in detail in Sect. 3.1.1. For all three ICs – ‘climate change’, ‘terrestrial eutrophication’ and ‘photochemical ozone creation’ – the ‘Steam’ and the ‘Electricity and Steam with CCS’ scenarios result in negative (thus, avoided) or negligible environmental impact, whilst the ‘Electricity and Steam’ and ‘Electricity’ scenarios result in a positive environmental impact value. The remaining ICs – ‘freshwater and terrestrial acidification’, ‘freshwater eutrophication’ and ‘respiratory effects, inorganics’ – all four energy output scenarios resulted in negative environmental impact values.
The results of Fig. 3a illustrate a common ranking of the different scenarios in all IC’s (except for ‘climate change’). From least to most environmental impact this ranking is ‘Steam’, ‘Electricity and Steam with CCS’, ‘Electricity and Steam’, and finally the ‘Electricity’. This means that by coupling a MEA-based CCS process, similar to that modelled in this study, to WtE, only environmental impact in the IC ‘climate change’ will be reduced. In this particular case, the ‘Electricity and Steam with CCS’ scenario further reduces the environmental impact value by ca. 28300 AECE compared to the ‘Steam’ scenario in the IC ‘climate change’. However, for all other IC’s, the ‘Steam’ scenario has the largest avoided impact. These results corroborate that of Lausselet et al. (2017).
There are two explanations for the relatively poor environmental performance of the scenario with CCS compared to the ‘Steam’ scenario. Firstly, the electrical requirements for the capture and conditioning of CO2 were assumed to be provided by the steam turbine on site, therefore diverting steam from being sold to neighboring chemical companies and consequently reducing the avoided volume of conventional production of steam. Secondly, auxiliary inputs and outputs related to the CCS process, such as the production of MEA or the transportation of liquid CO2 to the storage facilities in the North Sea, also contribute to the overall environmental impact. These observations illustrate that reducing the environmental impact associated with global warming and CO2 emissions (e.g., installing a CCS process), may cause larger environmental impacts in other ICs. However, these results do not imply that CCS should not be investigated or considered as a viable addition to WtE but rather indicates that further research is necessary which investigates other state-of-the-art CCS or CCU techniques coupled with WtE, which aligns with the conclusion drawn by Christensen et al. (2021).
To identify sub-processes that contribute to the overall environmental impact value in one IC shown in Fig. 3a, the individualized results in the ICs ‘climate change’, ‘respiratory effects, inorganics’ and ‘photochemical ozone creation’ in Model 1 (shown in Fig. 4a, Fig. 4b, and Fig. 4c respectively) will be discussed in detail in the Sect. 3.1.1, 3.1.2, 3.1.3. These three ICs were selected for further discussion due to the magnitude of their overall result shown in Fig. 3a as well as having a variety of causative pollutants.
3.1.1 Climate change
Figure 4a illustrates the normalized environmental impact values in the IC ‘climate change’ for the six processes that contribute to over 97% of the total environmental impact in the IC ‘climate change’. The ‘Electricity and Steam with CCS’ scenario results in an environmental impact value approximately 26600 AECE and 32600 AECE less than the other three scenarios for the grate and fluidized bed incinerator line processes, respectively. The relatively large environmental impact values, ca. 30900 AECE and 38400 AECE, for the grate and fluidized bed incinerator lines respectively, seen in the other three scenarios, can be mainly attributed to the non-biogenic CO2 that is emitted at the stack. The scenario with CCS captures 90% of the total CO2 before the flue gases are emitted, therefore reducing environmental impact. CO2 is an inherent product during the combustion of MSW in the presence of oxygen, therefore it’s environmental impact can only be minimized by decreasing the fossil waste fraction in the input waste composition, or by capturing the CO2 at the stack. However, the addition of CCS has not reduced the environmental impact to zero for the grate and fluidized bed incinerator line processes, i.e., the impact value is ca. 4300 AECE and 5800 AECE respectively. This remaining environmental impact can be attributed to the 10% of direct CO2 emission that was not captured in the CCS process, as well as the indirect greenhouse gas emissions of auxiliaries required by the CCS process and the incinerator line processes e.g., the greenhouse gas emissions from the production of urea and activated carbon.
The process ‘steam production, as an energy carrier, in the chemical industry’ results in the largest absolute avoided environmental impact: the ‘Steam’, ‘Electricity and Steam with CCS’, ‘Electricity and Steam’ scenario have a value of ca. -80800, -49100, and − 7400 AECE respectively.
The scenario with CCS results in a smaller avoided environmental impact than the ‘Steam’ scenario because a smaller volume of conventional steam production was avoided by the neighboring chemical industry: 6300 TJ in the ‘Steam’ and 3826 TJ for ‘Electricity and Steam with CCS’ scenario (see Fig. 2). However, this difference in avoided emission (ca. 31700 AECE) is more than compensated by the combined reduced impact of both incinerator line processes (ca. 59200 AECE).
In the process ‘market for electricity, medium voltage’, ‘Electricity’, ‘Electricity and Steam’, and ‘Electricity and Steam with CCS’ scenario have a relatively modest environmental impact values of ca. -14500, -13200, and − 2400 AECE respectively. Indeed, nuclear power and renewable sources, which do not directly emit CO2 during electricity production, make up 68.5% of the 2014 Belgian electricity mix (Eurostat, 2021). This can explain the relatively low avoided environmental impact for the IC ‘climate change’. In contrast, the process used for conventional steam production, uses fossil fuels such as natural gas, hard coal and crude oil, which release non-biogenic CO2 into the atmosphere during combustion, resulting in a relatively large avoided environmental impact in the IC ‘climate change’ as described above.
3.1.2 Respiratory effects, inorganics
Figure 4b shows the six processes that contribute to over 94% of the total environmental impact for the IC ‘respiratory effects, inorganics’ with their corresponding pollutants. There are two dominant pollutants: PM2.5 and SO2, the former contributes with a larger magnitude in all processes. The processes ‘reinforcing steel production’ and ‘market for aluminum, primary, ingot’ result in an overall avoided environmental impact value of ca. -5000 AECE and − 4100 AECE respectively in all four energy output scenarios. Combined (9100 AECE), the avoided impact from both substituted metal production processes largely compensates for the combined environmental impact resulting from the operation of both incinerator line processes (ca. 2200 AECE) for all four scenarios.
The ‘Steam’ scenario, in the process for avoided conventional steam production, results in the largest magnitude avoided impact value (ca. -16200 AECE and − 11800 AECE for PM2.5 and SO2 respectively) followed by ‘Electricity and Steam with CCS’ with an environmental impact value of ca. -9800 AECE and − 7100 AECE for PM2.5 and SO2 respectively. Hard coal containing sulphur constitutes approximately 14% of the fuel mix used for the Ecoinvent process ‘steam production, as an energy carrier, in the chemical industry’ used in Model 1. SO2 is released as a pollutant during hard coal combustion and thus appears as a pollutant during conventional steam production in Model 1.
3.1.3 Photochemical ozone creation
Figure 4c illustrates the normalized environmental impact values in the IC ‘photochemical ozone creation’ for the six processes that constitute for over 95% of the total environmental impact in the IC ‘photochemical ozone creation’. The combined environmental impact value for the two incineration line processes is ca. 16000 AECE and is attributed to the pollutant NOx (see Online Resource 2). NOx is formed by either the oxidation of nitrogen compounds contained in the waste (fuel NOx) or the oxidation of N2 in the combustion gas (thermal NOx). Fuel NOx could be reduced by attempting to lower the nitrogen content of the input waste by applying more stringent separating at source regulations or adding waste sorting processes before incineration takes place. The reduction of thermal NOx is more difficult because if air is being used as the source of oxygen for the combustion process, the formation of thermal NOx is unavoidable. For this reason, end-of-pipe technique are installed to reduce the NOx in the flue gas. In the considered Indaver/Sleco WtE plant, selective non catalytic reduction (SNCR) involving injection of urea in the hot flue gases is installed as an end-of-pipe technique to reduce the amount of NOx emitted at the stack. Therefore, a way to further reduce the thermal NOx emitted would be to improve the de-NOx end-of-pipe technique, for example by installing selective catalytic reduction (SCR). However before doing so, it would be important to evaluate whether introducing the SCR technique would indirectly increase the environmental impact of other IC’s due to auxiliary exchanges attributed to the new technique (Van Caneghem et al., 2016).
In the process ‘steam production, as energy carrier, in chemical industry’, the normalized environmental impact value for the ‘Steam’ scenario is ca. -23700 AECE, NOx being the most contributing pollutant for this process. A fuel present in the average fuel mix used for this process is hard coal, which can account for the emission of SO2 as it is the product of the oxidation of sulphur in the coal during the combustion process.
Due to the relatively large environmental impact of the grate and fluidized bed incinerator line processes (ca. 8900 AECE and 7100 AECE respectively), only the scenarios which avoid a large volume of conventional steam production result in overall negative impact values for this IC. For example, the combined avoided impact value for all pollutants in the avoided conventional steam production process is ca. -23700 AECE from the ‘Steam’ scenario. It thus compensates for the combined environmental impact caused by both incinerator line processes. Whereas the largest avoided environmental impact value caused by the ‘market for electricity, medium voltage’ process is ca. -4400 AECE and thus cannot compensate for the environmental impact caused by both incinerator line processes.
3.2 Life Cycle Assessment results: Model 2
Figure 3b displays the normalized LCA results for the same six ICs used for the evaluation of Model 1. The scale on the y-axis in Fig. 3a and Fig. 3b is equivalent illustrating the clear difference between the results of Model 2 compared to Model 1. That is, 19 out of 24 LCA results from Model 1 varied by more than 50% when compared to Model 2. The results of Model 1 show that the ‘Steam’ scenario has an environmental advantage over the ‘Electricity’ scenario for all six ICs measured, however contrarily, in Model 2, the ‘Steam’ scenario results in less avoided environmental impact compared to the ‘Electricity’ scenario in the ICs ‘freshwater and terrestrial acidification’, ‘photochemical ozone creation’ and ‘terrestrial eutrophication’. There are two causes of this disparity: firstly, the environmental impact of producing 1MJ of heat with natural gas is lower in all six ICs compared to the production of 1MJ of steam using the European average steam production mix (see Online Resource 3). For example, the environmental impact of producing 1MJ of heat in a natural gas industrial furnace (as in Model 2) is 17%, 38% and 34%, relative to European steam production mix (as in Model 1) for the ICs ‘freshwater and terrestrial acidification’, ‘photochemical ozone creation’ and ‘terrestrial eutrophication’, respectively (see Online Resource 3). Secondly, the three mentioned ICs result in an larger avoided environmental impact for the production of 1kWh using a natural gas fueled power plant compared to the Belgian national electricity mix. For example, in the IC ‘freshwater and terrestrial acidification’, the production of 1kWh in a natural gas power plant (as in Model 2) has a relative impact that is 1.28 larger than the Belgian electricity mix (as in Model 1), which, indeed, results in a reduced avoided impact in Model 2 compared to Model 1 (see Online Resource 3). Therefore, in Model 2, scenarios that produce steam rather than electricity will result in smaller avoided environmental impacts. Consequently, both the ‘Electricity’ and ‘Electricity and Steam’ scenarios show improved overall environmental impact values for four of the six ICs in Model 2 compared to Model 1 (as displayed in Fig. 3). For instance, the normalized impact value, in the IC ‘climate change’, was ca. 45000 AECE for the ‘Electricity’ scenario in Model 1 and was reduced to ca. 14200 AECE in Model 2 and in the IC ‘photochemical ozone creation’ the normalized impact value was ca. 3200 AECE in Model 1 and was reduced to ca. -8300 AECE.
These results illustrate how the LCA practitioner’s selection of predetermined substitution processes, available on the Ecoinvent or other databases, has a significant effect on the overall results of the LCA. They show the importance of using processes that best model reality, thus reducing as much uncertainty as possible. Furthermore, the results and discussion above make it clear, in a quantitative way, that there is no general rule for the determination of the environmental performance of WtE plants since it depends largely on the (assumed) avoided energy production processes. This corroborates with conclusions drawn by Boesch et al. (2014).