*2.1. The Italian Natural Gas Networks*

Two different kind of networks are operated in Italy: the transportation and the distribution networks. More than 90% of natural gas is imported by foreign countries. The Italian natural gas network is characterized by the presence of seven "Import Points", which are connected to the Italian transportation system for natural gas supply [24]:


Two further connections should be considered, that are between the Italian natural gas transportation system and national natural gas storages, which are located in Campo Collalto (Treviso–Veneto) and Montalfano (Chieti–Abruzzo).

From the operative point of view, the Italian transportation system is operated at a pressure between 24 and 75 bar g, even if submarine pipelines are operated at a pressure up to 115 bar g. Figure 1 shows through different colors and thickness how the transportation system is indeed divided into two networks, the National Transportation (NT) system and the Regional Transportation (RT) system (in blue and light blue, respectively, in Figure 1). Figure 1 includes only the Transportation system managed by SNAM (in Italian "Società Nazionale Metanodotti), that is the most important of the nine Italian Transportation System Operator (TSO) that controls more than the 93.2% of the Italian system [25].

**Figure 1.** The Italian National and Regional Transmission Systems [26].

In accordance to the Decree of the Ministry of Industry and Economic activities 22/12/2000 [27], the NT system consists of networks with a total length of 10,272 km that connects the North with the South of Italy conveying the natural gas from the Import Points to the Interconnection Points with the RT systems and the two storage plants. Thirteen gas compression plants, with a total load of 961 MW el, are installed to compensate the pressure drops along the TN system [26]. Particularly, centrifugal gas compressors are installed. However, due to the high flowrate elaborated, i.e., up to 1,500,000 Sm3/h, a maximum compression ratio up to 1.4–1.5 is available in gas compression plants. Therefore, since gas compression plants have to be able to restore the downstream pressure up to 75 bar g in case of a national peak demand, a minimum upstream pressure of 50 bar g (=75/1.5) is allowed by gas transmission code [28]. In accordance to the Decree of the Economic Ministry 29/9/2005 [29], the RT system, with a total length of 24.700 km and 20 interconnection points with the NT, accounts for the distribution of natural gas though the national territory and, particularly, to power plants and to local distribution networks that are connected through 567 ReMi (Regolazione and Misurazione in Italian) stations at a minimum pressure up to 24 bar g. An updated list of TN and RT networks is available at [30].

The Italian Distribution system is responsible for natural gas supply to final customers. Almost 30 GSm3 of natural gas, equivalent to almost 300 TWh, are annually supplied by more than 200 Distribution System Operators (DSO) to more than 23 million final Italian customers through more than 260.000 km of local networks that are mainly in the Northern of Italy, wherein 70% of the Italian natural gas consumption is concentrated [31]. Respect to the NT and RT systems, gas pressures lower than 5 bar g are operated in the Distribution networks [32]. Due to the lower nominal pressures than TN and RT systems, in addition to steel also polyethylene, iron and copper have been used [33]. Although distribution network is considered as a possible short-term storage for syngas produced in local P2G applications, concerns exist about the implementation of hydrogen blending along the Distribution network. First of all, the presence of multiple hydrogen injection point would be responsible for very different concentrations of the HCNG along the local networks that could impede to DSOs an effective and reliable control of the network operation. Secondly, the high number of DSOs connected to the transportation system could create difficulties in terms of management of the energy fluxes with the transmission system. Particularly, more than 500 "connection points" between distribution and transmission networks are present in Italian gas system [26]. Each "connection point" would become a hydrogen blending point into the transmission system. Therefore, the resulting hydrogen concentration of the transportation system depends on the hydrogen concentrations and on the flowrates entering from each connection point. A very complex coordination between DSOs would be therefore required to not exceed the hydrogen concentration threshold. Therefore, hydrogen blending is assumed only in Italian transmission gas systems while distribution gas networks are not considered as an option in the following sections of the papers for location of P2H plants. Nevertheless, the result of the preliminary assessment in terms of quantification of low percentage hydrogen blending potential is not affected by this choice.

#### *2.2. Premises and Main Hypothesis for Hydrogen Blending Potential Estimation in Italian Natural Gas System*

In general, based on research to date [14], only minor or no issues should arise with limited percentage of hydrogen blends, i.e., less than 5–15% hydrogen by volume. More significant problems would be addressed for higher blends, in the range of 15–50%, such as conversion of household appliances, an increase in compression capacity along distribution mains serving industrial users, and the development of a complex control strategies to monitor hydrogen injection and hydrogen percentage blend into the network. Hydrogen blending above 50% is expected to be possible on through challenging actions across multiple areas, including pipeline materials, safety, and substantial modifications required for end-use appliances or other uses. Nevertheless, up to now the limits for hydrogen blending into the natural gas networks have been usually kept very low, varying between 0.2% up to 6% [34]. Even if Italian regulation allows hydrogen concentration for blending only up to 1.0% [35], as defined for biomethane injection, experimental activities have been already performed in Italy to evaluate the impact of higher concentrations in existing networks: 5% blending has been already tested in a small closed network near the southern city of Salerno [36], while new tests have been planned with the aim of testing 10% hydrogen injection [37].

Moving towards a hydrogen economy will require the design and implementation of a complex and long-term national strategy. While potential targets and techno-economic impact by 2050 of the hydrogen economy in Italy have been already estimated [38], a national strategy is still far from being clearly defined. A fundamental part of the EU hydrogen strategy is the "first step", i.e., the public and private investments to be planned in the next 4 years, targeting 2024. Accordingly, short term actions must be planned to stimulate the growth of the hydrogen market and to start the hydrogen penetration in the Italian energy sector. How to approach the opportunity of hydrogen blending into the natural gas network by 2024 is crucial since Italy has one of the largest natural gas network infrastructures in Europe [39], connected with several foreign and strategic areas like Northern Africa and East Europe. Furthermore, Italy also has a huge potential for renewable power generation via wind and solar: [40] identifies in 18.4 GW the wind potential that can be installed by 2030, which would correspond to an annual electricity production of 40.1 TWh, while [41] estimates in about 127 TWh per year the power production from photovoltaics (PV) integrated in buildings.

From a practical point of view, P2H plants will be needed to blend green hydrogen into the Italian natural gas network. So, the design of a strategy moving towards a growing percentage of green hydrogen injected into the natural gas network requires to plan the design, installation and simultaneous operation of an increasing number of P2H plants year by year. Furthermore, since renewable power is needed to produce green hydrogen, the planning of new P2H plants cannot be realized without taking into consideration the current location of renewable power plants as well as the setting up of new ones, if needed. Table 1 shows the current installed power capacity of PV and wind turbine power plants in Italy by Regions [42].


**Table 1.** Current installed power capacity of photovoltaic (PV) and wind turbine powerplants in Italy by Regions (data updated to June 2020, from [42]).

Therefore, the complexity of the hydrogen economy development will increase with the increasing of green hydrogen percentage injected into the natural gas network due to (i) the impact of hydrogen blending into the existing infrastructures and end-users, and (ii) the interactions between renewable power generation and hydrogen production. However, in a first phase these issues can be minimized if (i) the percentage of green hydrogen is kept relatively low and (ii) the installation of P2H is optimized by taking into account the current locations of both natural gas network and renewable power plants.

The aim of the paper is to identify what is the total amount of green hydrogen that could be produced and injected right now in the Italian natural gas network without compromising its integrity and with no relevant drawbacks for the end-users. The quantification of such a target is fundamental to calculate the P2H installations needed and to evaluate in a first assessment the geographical distribution and the required budget for the realization of these new P2H plants in relation with natural gas network characteristics and current regional distribution of renewable power plants.

#### *2.3. Analytical Description of the Methodological Approach*

The evaluation of the maximum green hydrogen blending capacity to be injected in the Italian natural gas network with no relevant impacts has been done accordingly to the following considerations. The maximum blending threshold (BT), defined as in Equation (1), is the limit to hydrogen blending beyond which many actions are needed to guarantee infrastructure integrity, end-users safety and an effective control of hydrogen percentage flowing in the natural gas network. BT, calculated in (Sm3/h), can be computed if (i) the minimum natural gas (MNG) flowrate in (Sm3/h) measured in the natural gas network is known, and if (ii) the allowed blending percentage (ABP) is fixed. ABP can be defined as the upper limit of hydrogen blending percentage in volume in the natural gas grid under which modifications on the network and its auxiliaries and on the end-users are not required. ABP is defined in (%vol) Natural gas and hydrogen density are respectively defined as ρ*NG* and ρ*H*2, both in (kg/Nm3). A safety factor (SF) in (%) and lower than 1 is also introduced in Equation (1) to take into account of the available data quality.

$$\text{BT} = \text{SF} \times \frac{\text{ABP} \times \rho\_{\text{H2}}}{(1 - \text{ABP}) \times \rho\_{\text{NG}}} \times \text{MNG} \tag{1}$$

An energy density correction factor (EDF) is also needed to take into account the reduction of the Lower heating value (LHV) of the HCNG volumetric flowrate (QHCNG) respect to the pure natural gas case. This reduction depends on the energy density of natural gas and hydrogen, in accordance to the respective higher heating Values (HHVNG = 9.70–12.58 kWh/Sm<sup>3</sup> [43]) and HHVH2 = 3.36 kWh/Sm3). In fact, since the total energy demand by the end-users (EDemand) does not change, an increase of HCNG flowrate is required proportionally to the reduction of the energy content of the gas mixture resulting from the hydrogen blending. The HCNG volumetric flowrate is the sum of the natural gas (QNG) and hydrogen (QH2) volumetric flowrates (Sm3/h) as reported in Equation (2):

$$\mathbf{Q\_{HCNG}} = \mathbf{Q\_{NC}} + \mathbf{Q\_{H2}} \tag{2}$$

where, considering wNG and wH2 as the volumetric concentrations of natural gas and hydrogen in the HCNG, Equations (3)–(5) apply:

$$\mathbf{Q\_{H2}} = \mathbf{Q\_{HCNG}} \times \mathbf{w\_{H2}} \tag{3}$$

$$\mathbf{Q\_{NC}} = \mathbf{Q\_{HCNG}} \times \mathbf{w\_{NG}} \tag{4}$$

$$\mathbf{w}\_{\rm NG} + \mathbf{w}\_{\rm H2} = 1 \tag{5}$$

Since end-users' energy demand does not depend on the composition of the gas supplied, the same amount of energy in case of pure natural gas flowrate has to be delivered through HCNG. Particularly, if QNG' is the natural gas flowrate when no hydrogen is blended in (Sm3/h), the existing energy demand EDemand (kWh) of the end-users can be calculated as in Equation (6):

$$\text{E}\_{\text{Demand}} = \text{Q}\_{\text{NG}'} \times \text{LHV}\_{\text{NG}} \tag{6}$$

where LHVNG is the lower heating value of the natural gas in (kWh/Sm3). The same amount of energy has to be transported by HCNG mixture. Defining the energy delivered by the HCNG flowrate as EHCNG (kWh), Equation (7) has to be considered:

$$E\_{\rm HCNG} = E\_{\rm Demand} \tag{7}$$

The energy transported by the HCNG flowrate can be calculated as in Equation (8):

$$\rm E\_{HCl\%} = Q\_{HCl\%} \times \rm LHV\_{HCl\%} \tag{8}$$

Where LHVHCNG (kWh/Sm3) is the lower heating value of the HCNG flowrate and it is calculated as in Equation (9):

$$\text{LHV}\_{\text{HCNG}} = \text{LHV}\_{\text{NG}} \text{w}\_{\text{NG}} + \text{LHV}\_{\text{H2}} \text{w}\_{\text{H2}} \tag{9}$$

From Equation (7) and by the use of Equations (6), (8) and (9), the HCNG flowrate required to supply the same amount of energy that end-users require is calculated as in Equation (10):

$$\mathbf{Q\_{hCNG}} = \mathbf{Q\_{NG'}} \frac{\mathbf{LHV\_{NG}}}{\mathbf{LHV\_{NG'}w\_{NG} + \mathbf{LHV\_{H2}w\_{H2}}}} \tag{10}$$

In accordance to Equation (10), the HCNG flowrate increases as the hydrogen concentration in the HCNG mixture rises due to the lower volumetric energy density of hydrogen respect to natural gas. Therefore, EDF, which is greater than 1 and defined as in Equation (11), is introduced in Equation (2) to calculate an energy corrected blending threshold (BTcorr) in accordance to Equation (12):

$$\text{LEDF} = \frac{\text{LHV}\_{\text{NG}}}{\text{LHV}\_{\text{NG}} \text{w}\_{\text{NG}} + \text{LHV}\_{\text{H2}} \text{w}\_{\text{H2}}} \tag{11}$$

$$\text{BT}\_{\text{corr}} = \text{SF} \times \text{EDF} \times \frac{\text{ABP} \times \rho\_{\text{H2}}}{(1 - \text{ABP}) \times \rho\_{\text{NG}}} \times \text{MNG} \tag{12}$$

Even if different operative conditions in terms of operative mixture pressure and temperature could verify during the years, it should be noted that the density ratio (ρH2/ρCH4) can calculated as follow. In fact, in accordance to the real gas law, Equations (13) and (14) apply:

$$\frac{\mathbf{P\_{NG}}}{\rho\_{\rm NG}} = \mathbf{Z\_{NG}} \frac{\mathbf{R\_0}}{\mathbf{M\_{NG}}} \mathbf{T\_{NG}} \tag{13}$$

$$\frac{\mathbf{P}\_{\rm H2}}{\mathbf{P}\_{\rm H2}} = \mathbf{Z}\_{\rm H2} \frac{\mathbf{R}\_0}{\mathbf{M}\_{\rm H2}} \mathbf{T}\_{\rm H2} \tag{14}$$

where pNG and pH2 are natural gas and hydrogen pressures [Pa], ZNG and ZH2 are natural gas and hydrogen compressibility factors in [#], R0 is the universal gas constant in [kJ/kmol K], MNG and MH2 are natural gas and hydrogen molecular weights (kg/kmol) and TNG and TH2 are the natural gas and hydrogen operative temperatures [K]. Even if operative annual temperature of natural gas conveyed in buried pipelines changes during the year [44], the variation can be considered negligible for the purpose of the following evaluations. However, the same consideration is not valid for pressure that depends on the specific point of the network. However, in the reported pressure range, i.e., [25 bar g, 75 bar g], the ratio between the hydrogen and methane compressibility factor can be considered almost constant. In fact, assuming an operative temperature of 285.15 K, the reduced temperature of hydrogen is equal to 6.8, resulting in a compressibility factor ZH2 equal almost to 1, independently from the reduced pressure. Concerning natural gas, assuming the same properties of methane, a reduced temperature of 1.49 and a reduced pressure between [0.04, 0.13] is calculated. A compressibility factor ZNG between 1 and 0.96 is obtained from available diagrams [45]. Therefore, compressibility factors are neglected in following evaluations. Equations (13) and (14) can be elaborated as reported in Equations (15) and (16):

$$\mathbf{p\_{H2}} = \mathbf{p\_{H2}} \times \left(\frac{\mathbf{R\_0}}{\mathbf{M\_{H2}}} \mathbf{T\_{H2}}\right)^{-1} \tag{15}$$

$$\rho\_{\rm NG} = \mathbf{p}\_{\rm NG} \times \left(\frac{\mathbf{R}\_0}{\mathbf{M}\_{\rm NG}} \mathbf{T}\_{\rm NG}\right)^{-1} \tag{16}$$

Therefore, the density ratio is calculated as in Equation (17) based on Equations (15) and (16):

$$\frac{\rho\_{\rm H12}}{\rho\_{\rm NG}} = \left(\frac{\rm P\_{\rm NG}}{\rm P\_{\rm H2}} \times \frac{\rm M\_{\rm CH4}}{\rm M\_{\rm H2}}\right)^{-1} \tag{17}$$

where also TNG is assumed equal to TH2 since natural gas and hydrogen are in the same mixture.

But why the authors define the MNG as natural gas flowrate reference for hydrogen blending? The hypothesis is that if the P2H blending capacity is calculated starting from the lowest capacity of the current natural gas flowrate, i.e., when the natural gas flowrate delivered by the national transportation system is at the minimum level, some important benefits occur:


After the BT has been identified, it is part of the strategy to define how much fast the threshold should be reached, i.e., how many MW of P2H plants are planned to be realized every year up to 2024 as schematically shown in Figure 3. The cumulative green hydrogen blending capacity is influenced by policy makers and energy planners' decisions, since the slope of the cumulative curve may allow to reach the threshold before (α1 in Figure 3) or close to the deadline (α2 in Figure 3). The second step of the strategy will start once the BT has been overcome, and will require relevant actions, as synthetized in Figure 3, as well as the practical implementation of actions over the time (curve slope β1 or β2 in Figure 3). Therefore, it is crucial to properly set the first phase timing to not reach too early the blending threshold, thus avoiding the risk of dead time waiting for the revamping/adaptation needed to increase the hydrogen blending percentage.

15/16 August 2020 16/17 January 2020 23/24 March 2019 9/10 September 2019

**Figure 2.** Natural gas flowrate variation respect to the daily average flowrate. Original elaboration based on data from [46].

**Figure 3.** Design of the first phase of implementation of the hydrogen blending strategy.
