Life Cycle Analysis of a Geothermal Power Plant: comparison of the environmental performance with other renewable energy systems

: A Life Cycle Analysis was performed considering three existing power plants of comparable size operating with different sources of renewable energy: geothermal, solar and wind. Primary data were used for building the life cycle inventories. The geothermal power plant includes emissions treatment for removal of hydrogen sulfide and mercury. The scenario about the substitution of natural emissions from geothermal energy, with specific reference to the greenhouse effect, is also investigated performing a sensitivity analysis. The results are characterized employing a wide portfolio of environmental indicators employing the Recipe 2016 and the ILCD 2011 Midpoint+ methods; normalization and weighting are also applied using the Recipe 2016 method at endpoint level. The results demonstrate a good eco-profile of geothermal power plant with respect to other renewable energy systems and allow for a critical analysis to support potential improvements of the environmental performances.


Introduction
Geothermal energy is an important energy resource, largely contributing to limiting the use of fossil fuels, for both electricity and direct uses (mainly heat for district heating). The world installed electrical capacity is over 12,000 MWe [1][2][3][4], with provision of direct heat of the order of 165,000GWh/yr [5]. The geothermal resource is well distributed around the world [6,7], and several locations are favoured by the presence of hot fluid resources (hydrothermal systems).Recently, the feasibility of Enhanced Geothermal Systems (EGS) has been demonstrated and this technology will allow an even more widespread use of the earth inner heat [8,9]. Experience has demonstrated that geothermal energy can be considered renewable if the resource is correctly managed [10,11], if the sizing of the conversion/utilization plants is compatible with that of the hydrothermal reservoir and if reinjection of the fluids practiced.
Italy has a long tradition of geothermal energy utilization [12], with nearly 1,000 MWe installed in two areas of the Tuscany region (Larderello/Travale and Monte Amiata) operated by Enel GreenPower. Specifically, the plants of the Larderello/Travale region (about 700 MWe) have been in industrial operation for more than 60 years, and this activity has considerably contributed to the local economic growth. An extensive grid exists for the management of fluids, including primary supply to local district heating as well as resource and reinjection fluid distribution. All power plants are equipped with effective emissions treatment equipment, which removes the greatest part of hydrogen sulphyde (H2S) and mercury (Hg) though the application of proprietary technology (AMIS® process [13,14]). The geothermal power plants located in Tuscany have demonstrated a high reliability, with equivalent operation time exceeding 7,500 hrs/yr and with a productivity of more than 6,200GWh/yr [15]. Solar electricity is mainly produced by photovoltaic (PV) power plants. Over the world, the power installed exceeds 500 GWe. Italy represents one of the main players in Europe with more than 20GWe installed and a productivity exceeding 24,000 GWh/yr [15]. Most of the PV plants in Italy are small (<50 kWe), however a significant share of production is done by 6% of the power plants with size > 50 kWe. The productivity data show that the utilization factor of solar PV is much smaller than for geothermal, with an equivalent full-load operability of about 1,200 hrs/yr. This is due to the periodic cycle of solar radiation (daily and seasonal).
Wind energy has had a strong increase with specific reference to Europe (180 GWe installed with a productivity of about 362,000 GWh/yr). In Italy more than10GWeare installed (mainly in the South), with a productivity exceeding 17,000 GWh/yr [15]. The equivalent full-load operability is typically 2,000 hrs/yr, as the wind resource is highly stochastic.
The lower operability identifies solar PV and Wind as Variable Renewable Energy (VREs), raising strong challenges to the grid infrastructure (solar being today more predictable and favoured in this sense). A higher market penetration of renewable energy sources (RES) will entail optimized strategies for production/load matching, and the development of extensive energy storage infrastructures supporting VREs. These latter will entail additional costs and environmental impacts, as well documented by the scientific literature about storage systems. Geothermal energy, which is typically employed as a baseload energy resource, is highly complementary to VREs and can represent a very valuable support, both in countries with limited electric grid infrastructure and in developed countries committed to an ever higher market penetration of electricity with respect to other energy vectors.
This work raises from the consideration that the clean energy does not exist: the only clean energy is the one we do not need to use, namely the saved one with efficiency actions. However, in an environmental sustainable perspective RES are better than fossil ones, but even in the use of RES the only rationale to make a choice should be based on benefit/cost ratio and a rigorous comparison of their environmental advantages and drawbacks. LCA analysis is a optimum tool to make this comparison.
In this study the comparison of the environmental performances of three power plants based respectively on geothermal, solar and wind energy is performed through the life cycle assessment (LCA) methodology grounding on robust and reliable primary data.

Life Cycle Assessment
LCA is a method to evaluate the environmental load associated with a product, process, or activity. LCA allows to quantify the used amount of energy and materials and of emissions and waste released in the environment, allowing for the evaluation of the associated potential impacts. The assessment is performed over the entire life cycle of the product, processor activity covering extraction and processing of raw materials; manufacturing, transportation and distribution; use, reuse, maintenance; recycling, and final disposal. The results of LCA can be expressed via a large number of environmental indicators and, generally, several impact categories are used to circumspectly detect the full range of ecological burdens associated with the investigated process or activity over the three environmental compartments (atmosphere, soil and water), thus aiming at avoiding burden shifting. The LCA methodology is regulated according to the general guidelines described in the International Standard series 14040 [16,17] and consists of four phases: 1. Goal and scope definition: in this phase the goal of the study, the system boundaries, the quality requisites of the data sources are described, and the functional unit of the analysis is specified.
2. Life Cycle Inventory Analysis (LCI): the purpose of this phase is to collect the input/ output data with regard to the system studied; generally robust and reliable LCI are built on primary data, that's to say specific data that highly characterize the system under study.
3. Life Cycle Impact Assessment (LCIA): this phase evaluates the significant potential environmental impacts using the LCI results; the process involves associating inventory data with specific environmental impact categories and the calculation of indicators values using accepted characterization factors. 4. Life Cycle Interpretation: it is the final phase of an LCA study in which the results of the LCI and LCIA steps are presented and discussed; interpretation includes conclusions and recommendations in accordance with the goal and scope of the study.
LCA was born as a detailed and quantitative approach for the evaluation of environmental sustainability [18]. The regulatory approach described in the ISO standards and in the more completely elaborated ILCD Handbook Guidelines [19] claims for the development of an LCA study till the characterization of the environmental impacts at a midpoint level. With this approach the LCIA method looks at the impact earlier along the cause-effect chain of the environmental mechanism and can refer to a relevant number of impact categories characterized by a low uncertainty but, on the other hand, difficult to interpret. In principle, it represents a good approach for the characterization of the eco-profile of the product or activity under study to use several wide-scope LCIA methods and check if findings are consistent in all of them. If so, it is possible to claim that findings appear robust.
But when this is not the case, the LCA practitioner might have to delve into the particularities of the LCIA methods and find out why the results are dissimilar, which can be a good learning experience about the characteristics of the applied LCIA methods.
The environmental evaluation at the endpoint level is a non-mandatory part of LCA, which includes normalization and weighting steps that allow to express the results referring to a limited number of damage categories, typically resources availability, human health and ecosystem quality.
Endpoint results provide insight on the environmental impact at the end of this cause-effect chain of the environmental mechanism, thus with larger uncertainty. If interpretation at this level provides less details, it is recognized that it is more suitable for the presentation of results to non-technical audiences. The various LCIA methods apply different impact category grouping, normalization and weighting factors thus it is necessarily recommended to refer to the same methodology when comparing different technologies dealing with the same product or process.
Energy conversion and utilization is one of the most famous and important fields of application of LCA calculations. LCA indeed offers a powerful approach to analyse systems overarching the complete life cycle of a system (from cradle to grave) which is necessary when considering the substitution of fossil fuels with renewables. When applying LCA to energy conversion systems, for fossil fuel-based technologies it is common to find high impacts connected to the use of fossil resources in the operational phase [20][21][22]; on the other hand, RES, which minimizes the use of consumables such as fossil fuel, entail a consistent use of materials because of the diffuse nature of renewable energy, some of which are rare, or whose extraction and/or production entails direct or indirect negative effects on the environment. In general, RES scores better environmental performance than fossil fuel systems in most impact categories with respect to the use of fossil Several LCA studies are available on solar PV energy conversion systems [23][24][25][26][27][28][29][30][31][32]; in general, the results indicate that a significant impact is coming from the manufacture of the PV modules, with the current silicon technology performing definitely better than CdTe, notwithstanding substantial advantages for thin-film manufacturing [33]. A significant fraction contribution to the overall ecoprofile (20-30%) comes from the structural materials and glazing. The environmental footprint is lower than the best fossil fuel-based technologies in most categories, with a weighted score typically 4-8 times smaller. The relatively standard production process has led to the development of accepted guidelines [34], which have determined an improved homogeneity in the results and better comparability of the studies. Wind energy has also attracted several LCA studies [35][36][37] however, they mostly rely on previously published LCA studies and, in most cases, on the use of literature data. There is a substantial lack of primary data (produced by the plant owner or operator), which are definitely more reliable as the source of the information can be completely tracked. Utilities such as Enel Green Power have a good opportunity to access these primary data (often gathered with the purpose of economic analyses, or of commitment of construction work or trusting of maintenance services) , and to use them to document the environmental quality of their product (electricity). This represents a key passage in the environmental evaluation, both in terms of company, services and products (possibly leading to an ECO-Label) and is also a primary motivation behind the present study.
The case studies described in the following were analysed using the OpenLCA 1. All data presented for the LCI inventory are primary data resulting from checked information about materials employed for construction. Secondary data were used for common materials (e.g. steel, concrete, copper, plastics, etc.) and for upstream processes (e.g. transport). The LCI reports also data for operation and maintenance, including replacement of equipment, consumables etc.

Case Studies
The case studies examined represent three power plants of similar nominal capacity (about 20 MWe): the geothermal power plant Chiusdino 1, the solar photovoltaic power plant SerrePersano (SP) and the wind farm in Pietragalla (P).    1 Distances are calculated from the two platforms (Montieri and Travale). 2 Only 53,5% of the flow rate from Montieri is used by the Chiusdino power plant.

Chiusdino Geothermal Power Plant
The Chiusdino 1 power plant is equipped with an AMIS ® emissions treatment system, which removes H2S and Hg with measured efficiencies of respectively 99,8% and 82,2%. A soda solution is currently used for acid gas treatment, while Hg is captured by a solid adsorption reactor. Details on the pollutant streams emitted, according to measured values certified by the regional authority (ARPAT), are provided in Table 2.
For Chiusdino 1, two scenarios are documented in order to consider the effects of emissions treatment: the real scenario GEO featuring the AMIS ® process and the hypothetical scenario GEO_NA representing the power plant as if no AMIS ® process were operating. Moreover, a third case was considered (GEO_AS), considering emissions treatment plus the partial substitution of natural emissions (several scenarios were investigated: the GEO_AS assumes that 40% of the power plant emissions would anyway reach the atmosphere as natural emissions).
The whole liquid condensate of Chiusdino 1 is re-injected using a complex network of pipelines connecting to the Larderello reinjection sites, with an overall estimated length of about 20 km.

Pietragalla Wind Farm
The   Table 3 resumes the features of the two PV fields.      The results of the comparison are synthesized graphically (impact category indicators) in the spidernet diagrams (Figure 9 for ILCD Midpoint 2011+ and in Figure 10 for Recipe 2016). Finally, a contribution analysis of the midpoint impact categories for the Recipe 2016 method is presented in Figure 11. The contribution analysis shows that the relative impacts for each of the energy technologies considered take place in different categories: for example, for geothermal the dominant categories are terrestrial acidification and fine particulate matter formation (for the scenario of power plants not equipped with emissions treatment, GEO NA), water consumption, marine and freshwater ecotoxicity. Wind and solar PV score high, in relative terms, for marine and terrestrial eco-toxicity. The NEM scenario produces large impacts for water consumption, for the marine and freshwater environments, for land use and fossil resource scarcity.

Impact assessment at endpoint level: Recipe 2016 normalized and weighted results
In this section results calculated at endpoint level are presented. For the reasons outlined in Section 3, supported by the better suitability of the method in the field of energy conversion (with specific reference to transition from fossil to renewable resources), Recipe 2016 is applied in the following for normalization and weighting calculations. These operations allow to group the impact assessment in three areas of protection: Ecosystem quality, Human Health, and Resources. Normalization for Recipe is done referring to the European population and leads to the calculation of results in function of: (i) DALY unit (disability adjusted life years), for human health, representing the years that are lost or that a person is disabled due to a disease or accident; (ii) species per year unit, for ecosystem quality, representing the local species loss integrated over time; (iii) dollar unit (USD 2013), for resources scarcity, representing the extra costs involved for future mineral and fossil resource extraction. The overall results of the environmental impact evaluation at endpoint level are summarized in Table 4; results are also illustrated graphically in Figure 12.   weighted by a factor of 300. This perspective is a common assumption in the field of energy conversion system. The Recipe 2016 weighted results are shown in Figure 14 in terms of Ecopoints referred to the functional unit (1 kWh of electricity).

Conclusions
Geothermal energy conversion was benchmarked by LCA methodology in comparison with other RES and with the Italian national energy mix. Calculations were performed based on specific power plant info built on primary data, taking advantage of life cycle inventories available through the plant operator (Enel Green Power). Three options were considered for Geothermal energy: the current power plant with AMIS® emissions treatment (GEO); the same system without emissions treatment (GEO_NA); and a hypothetical case where 40% of the emissions could be considered substitute of the natural emissions (GEO_AS). The GEO cases were compared with a wind energy farm (W) and with a large solar PV power plant (PV) having similar capacity.
Midpoint calculations were performed comparing the ILCD 2011 Midpoint+ and Recipe 2016 methods. The results were similar in terms of identifying the most impacting categories: terrestrial acidification, human toxicity, marine and freshwater eco-toxicity for the geothermal power plant (with a notable improvement in the case of emissions treatment); marine and freshwater eco-toxicity for wind and solar PV. The national energy mix impacts mainly in water consumption and fossil fuel depletion. In absolute term, wind energy emerged as the least impacting technology in most categories. It was evident already for the impact evaluation at midpoint level that the Recipe 2016 method can provide a higher degree of detail, accounting for relevant issues when comparing RES and fossil fuels (for example, mineral and fossil resources depletion).
The impact evaluation at the endpoint level was performed with the Recipe 2016 method allowing to cluster the results in three significant damage categories (ecosystem quality, human health and resources). All RES technologies scored definitely better than the national energy mix, as this last includes the use of considerable amount of fossil fuel resources (mostly gas and coal). Geothermal scenarios even with emissions treatment (GEO and GEO_AS) resulted to have a lower performance compared with wind and solar PV for ecosystem quality and human health damage categories; however, it represents a definite step forward with respect to the NEM and it compares well with respect to other RES for the resources damage category. This result is a direct consequence of the high productivity of geothermal power plants (over 7500 hrs/yr operation at nominal load, compared to 1300 for solar PV and 2300 for the wind farm). These results were confirmed by the contribution analysis performed on the Recipe 2016 normalized results at the endpoint level.
Finally, weighting allowed to calculate a final synthetic indicator that can be used to compare the environmental performances of the different electricity generation systems. To this end, a Hyerarchist cultural perspective was applied to the normalized Recipe 2016 results. Wind resulted to be the best technology with a value of 0,0012 Ecopoints/kWh, a result in line with previous documented LCA studies; however, the geothermal power plants achieved values of about 0,0177 Ecopoints/kWh which were close to solar PV (0,0087 Ecopoints/kWh) and much lower than those of the national energy mix (0,1240 Ecopoints/kWh).
These results, which should also be interpreted in terms of real availability of the RES and of economic profitability, demonstrate that geothermal energy conversion is a good option for sustainable development. Author Contributions: D.Frosali was responsible for LCIbuilding from primary data. The LCI data was collected using primary data by F. Sansone and D. Frosali. R.Bonciani provided expertise on power plant and emissions treatment processes. G.Manfrida, R. Basosi and M.L. Parisi organized the validation and presentation of the results. All authors contributed to writing and interpreting the results.

AP
Funding: This research received no direct funding through research projects, in which the authors are however involved.

Acknowledgments:
The Authors applied for LCI data gathering prospects developed within the GEOENVI H2020 project (Grant agreement no. 818242). R. Basosi and M.L. Parisi acknowledge MIUR Grant -Department of Excellence 2018-2022.

Conflicts of Interest:
The authors declare no conflict of interest.

Resource characteristics
Global annual radiation on the normal surface 2131 kWh/m2