Life cycle inventory

LCI data for the main processes at the foreground level was collected from own measurements and combined with technical process parameters for the biogas and methanol plants. The energy efficiency was calculated at 35.58% based on the individual energy efficiency rates shown in Table 4.17. The process of methanol production is the most intensive in the use of electricity, followed by the electricity

demand of the BGP and the CO2 separation by membrane. Yet, only 25% of the electricity production of the CHP and 35% of the produced heat is used within the system, given that H2 production is wind-based. The methanol synthesis data was obtained from a process design application for small plants in Aspen Plus as described in chapter 4.2.1. Primary data was combined with secondary data for the energy demand of the biogas upgrading, as well as certain emission factors of the BGP. It must be noted that the production of capital goods for the main processes considered in the foreground system are not included within the system boundaries due to data limitations at this stage. Manure application and transport are also not included since it is assumed that the integrated system is located on-farm. The main assumptions made and data sources employed through the LCI, are described as follows, while the LCI is shown in Table 4.18:

Table 4.17: Energy efficiency of the main sub-processes included in the system boundaries to produce 1 kg of methanol by means of the Power-to-Fuel system proposed.

Biogas production plant

(incl. CHP)

Polymer electrolyte membrane electrolysis

CO2

recovery plant

Methanol synthesis

plant

η𝑠𝑢𝑏−𝑠𝑦𝑠𝑡𝑒𝑚 0.65^a 0.70^b 0.92^c 0.85^d

a RAU (2019), ^b SCHIEBAHN et al. (2015), ^c Sun et al. (2015), ^d Schemme et al. (2020).

a) Biogas production: The LCI data was taken from a BGP located in Eastern Germany, via personal communication with the Technical University Bergakademie Freiberg (RAU, 2019). The plant’s characteristics and assumptions have already been described in section 4.3.1 and also in more detail in chapter 4.1.

Its feedstock is considered waste from livestock farming and entails zero emissions according to the RED II. As the raw manure is only shortly stored in a covered tank below a concrete floor, the emissions of pre-storage are thus expected to be inexistent. Therefore, CH4 and ammonia (NH3) emissions occur only during the stages of AD and digestate storage. N2O emissions from AD are neglected according to IPCC (2006b) as data is scarce, while the process releases negligible quantities of H2, H2S, H2O and other trace gases. We further considered CH4 losses in the form of both rest gas potential arising from the open storage of digestate and leakages as well as overpressure security valves from the fermenter. These are estimated at 1 kg/MWh equivalent to 1.4%, mainly coming from the digestate storage facility (FNR, 2016, RAU, 2019). The leakage of NH3 from the fermenter is lower than 0.05% of the N content in the digestate and hence excluded (EMEP/EEA, 2016, WULF et al., 2019). NH3 emissions of 2.66% of the N in the

digestate, occurring during storage, were included in the inventory based on a Tier 2 approach from EMEP/EEA (2016). The values for the N content as well as other components in digestate arise from on-site measurements from the digestate storage tank of the larger plant, which is fed with the same manure, though slightly different additives, and have the same retention time of 150 days.

b) CO2 recovery: The energy demand for the membrane technology of the standard case capacity is provided for a capacity of 40 m³/h raw gas by the Apex AG (OESTER, 2019). It is stated in kWh/Nm³ raw gas. Since the exact amount of raw gas per hour is known within the system, this can be converted to MJ/h. The electricity demand is in fact higher than that used in EGGEMANN et al. (2020) because that paper used data for more industrial, large-scale upgrading processes.

Here, a closed upgrading system and thus no emissions are assumed. As biogas upgrading includes upstream desulphurisation and dehydration, it should guarantee that the biogas no longer contains sulphur when entering this stage (ADLER et al., 2014a). The membrane technology by OESTER et al. (2018) also includes fine desulphurisation via activated carbon. As desulphurisation is part of the upgrading technology described by VIEBAHN et al. (2018), we assumed that the LCI includes the demand of electricity for desulphurisation as part of the CO2 separation process.

The required activated carbon of 0.085 kg/h is also considered, which was not considered in EGGEMANN et al. (2020). In addition, the process of obtaining pure CO2 would require some sort of post-combustion. If the combustion process is run with pure O2, the formation of harmful NOx emissions can be avoided. The O2 from the H2 production is sufficient, as the process only requires 2.4% of it. The amount is subtracted from the O2 occurring in the modelled process, assuming it has already been transmitted to the post-combustion process.

c) CHP: Emissions were measured for the 75 kW CHP. When calculating the CH4

losses due to biogas production and upgrading, we assumed that an additional 1.5%

of biogas is required to guarantee a full-load drive of the CHP. This translates into additional feedstock requirements in input that can be easily operated with the existing fermenter, as presented in chapter 4.1.4. Emissions data for the CHP gas engine without a catalyst was measured at the plant in Eastern Germany in November of 2018. As the CO2 in the flue gas is biogenic, it is excluded from the inventory. If it was included it would have an impact of 0.26 kg per kg methanol.

The NOx emissions by the CHP were adjusted as compared to those used in EGGEMANN et al. (2020), as there was a mistake and they were actually lower.

They were corrected by one decimal.

Table 4.18: Life cycle inventory of all inputs and outputs associated with the production of 1 kg of methanol by means of the Power-to-Fuel system proposed.

INPUTS OUTPUTS

Methanol synthesis Methanol synthesis

CO2 (kg) 1.37 Methanol (kg) 1.00

H2 (kg) 0.19 Water (kg) 0.56

Electricity (MJ) 1.79 Process heat (MJ) 2.99

CO2 recovery CO2 recovery

Biogas (m³) 1.58 CO2 in flue gas (kg) 1.37

Electricity (MJ) 0.15 H2O (m³) 1.96E-05

Activated carbon (kg) 0.0033 Biomethane 95vol.% (kg) 0.66

Biogas production Biogas production

Electricity (MJ) 0.85 Biogas (m³) 1.56

Heat (MJ) 4.86 Urea as N (kg) 0.69

Potassium chloride as K2O (kg) 1.09 Single superphosphate as P2O5 (kg) 0.49 NH3 emissions from digestate storage (kg) 7.87E-04 CH4 losses from AD and digestate storage

(kg) 8.07E-03

CHP CHP

CH4 (kg) 0.66 Electricity (MJ) 10.63

Heat (MJ) 13.89

Emissions

SO2 (kg) 2.20E-03

NOx (kg) 1.31E-02

CO (kg) 6.97E-03

NMVOC (kg) 1.50E-04

CH4 (kg) 2.10E-02

H2 production

(PEM) H2 Production (PEM)

Electricity (MJ) 32.27 Oxygen (kg) 1.33

Water (kg)

1.69E-03 H2 (kg) 0.19

CHP = Combined heat and power plant.

d) H2 production: Production data for the wind turbine and H2 electrolysis as well as the methanol production were obtained from own simulations performed by the Institute of Electrochemical Process Engineering within the Institute of Energy and Climate Research at the Forschungszentrum Jülich. Assuming 2000 FLH of the wind turbine, a factor of 4.25 was considered for 8500 FLH for methanol synthesis and the AD process.

e) Methanol production: The methanol production process was carried out under 250°C at 80 bar inside an isothermal reactor that uses 1.37 kg of CO2 per kg of methanol, as also described by BILLIG et al. (2019). The 138.38 kWth,LHV methanol synthesis plant uses 34.29 kg of CO2 per hour, and 4.7 kg of H2. The treatment of the discharged water from the process is also considered in a waste water treatment process, an additional process that is introduced to the analysis in EGGEMANN et al. (2020). Another difference compared to that study is that the excess heat from the methanol process is not accounted for. The process simulation showed that it is only about 10 kW and temperatures lie far below low pressure steam level. It would also not be profitable from an economic point of view, considering the additional infrastructure that would have to be built. Furthermore, the electricity demand was adjusted compared to the LCA in EGGEMANN et al. (2020) by a factor of 3.2 higher than the one used before.

For the reference process, we consider that conventional methanol production in Germany is carried out via the steam reforming of NG, importing NG from Russia as one of the main import partners. According to KEHLER et al. (2016), the CH4

emissions for transport in and from Russia vary between 0.32-0.97% of the gas produced. Hence, losses of 0.65% are assumed. This fits with the optimist values stated by SIMLA et al. (2019, p. 54). Associated LCI data for the local NG extraction is taken from the process in Ecoinvent 3.5 (Wernet et al., 2016) by considering energy consumption only and neglecting capital goods, same as in the proposed system. The LCI data related to NG and energy demand for the methanol synthesis is also taken from Ecoinvent and presented in Table 4.19.

Table 4.19: Utilities required for conventional methanol production according to WERNET et al. (2016).

Utility Value Unit

Electricity 0.266 MJ/kg MeOH

Heat 6.930 MJ/kg MeOH

Natural gas 0.652 m³/kg MeOH